Communication with Signals Produced by Abdominal Vibration
Transcription
Communication with Signals Produced by Abdominal Vibration
BEHAVIOR Communication with Signals Produced by Abdominal Vibration, Tremulation, and Percussion in Podisus maculiventris (Heteroptera: Pentatomidae) ALENKA ŽUNIČ,1,2 ANDREJ COKL,1 META VIRANT DOBERLET,1 AND JOCELYN G. MILLAR3 Ann. Entomol. Soc. Am. 101(6): 1169Ð1178 (2008) KEY WORDS Podisus maculiventris, substrate-borne communication, song repertoire, mating behavior Communication with substrate-borne signals is widespread in insects (Čokl and Virant-Doberlet 2003, Virant-Doberlet and Čokl 2004, Cocroft and Rodriguez 2005), and over the past few years, our understanding of the basic phenomena underlying substrate-borne communication by insects on plant substrates has increased substantially. The consequences of using different types of plant substrates for vibrational communication have been studied previously (Cocroft et al. 2006), and it has been demonstrated that nondispersive waves are used by insects communicating with higher frequency signals through larger stems (Casas et al. 2007). Furthermore, the spectral characteristics of vibrational signals produced by vibration of the abdomen by one group of insects, the stink bugs, have been conÞrmed to be tuned to the resonant properties of eight species of herbaceous plant substrates on which they are typically found (Čokl et al. 2005). Herbaceous plants act as low pass Þlters (Michelsen et 1 Department of Entomology, National Institute of Biology, Večna pot 111, SI-1000 Ljubljana, Slovenia. 2 Corresponding author, e-mail: [email protected]. 3 Department of Entomology, University of California, Riverside, CA 92521. al. 1982, Barth 2002, Cocroft et al. 2006, McNett and Cocroft 2008), and during transmission, the amplitude and spectral characteristics of higher frequency stridulatory signals are changed signiÞcantly during passage through the plant substrate (Čokl et al. 2006). Thus, frequency characteristics of the signal play only a minor role in determining species speciÞcity. However, temporal characteristics have been shown to be more species speciÞc (Hrabar et al. 2004, Miklas et al. 2001). Our increasing knowledge of insect-plant interactions in the context of substrate-borne communication follows investigations of the signals and their production mechanisms in different insect species. In true bugs in the infraorder Pentatomorpha, substrateborne signals are produced by stridulation and/or vibration of the abdomen (Gogala 2006). For these bugs, the interaction between the insect and the plant substrate during communication has been studied only for signals of the fundamental frequency, which is typically in the range of 90 Ð150 Hz. Insect-produced signals of frequencies below 50 Hz generally have been ignored, and plant vibrations in this frequency range typically have been described only in the context of 0013-8746/08/1169Ð1178$04.00/0 䉷 2008 Entomological Society of America Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 ABSTRACT Mating behaviors of the predaceous spined soldier bug, Podisus maculiventris (Say) (Pentatomidae: Asopinae), were observed, and vibrational signals used in intraspeciÞc communication were recorded and analyzed. Only males produced vibrational signals, using three different vibrational modes: vibration of the abdomen, percussion with the front legs, or tremulation of the body. When recorded on a nonresonant substrate (a loudspeaker membrane), the mean dominant frequency of signals produced by abdominal vibration varied between 90 and 140 Hz, and of tremulatory signals between 8 and 22 Hz. Percussion signals were broad-band, with the dominant frequency ⬇97 Hz and lower amplitude spectral peaks between 1,500 and 3,000 Hz. In the courtship phase of mating behavior, males emitted pulse trains composed of an abdominal vibration after a tremulatory pulse. Females stimulated by signals of abdominal vibration and composite signals expressed searching behavior and orientation toward the source of vibration, whereas no speciÞc reaction to tremulatory signals was observed. Observations of mating behavior and playback experiments showed that the abdominal vibration signals had both long range calling and short range recognition functions. Tremulatory signals were emitted at close range and were essential for copulation to occur. The sequences of fast and repeated percussion signals were emitted both as individual signals during calling and courting or between vibratory and tremulatory pulses. However, unlike the other vibrational signals, percussion did not seem to be emitted in a particular behavioral context, and its possible function in eliciting responses from females is unclear. ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Male * Female 20 - 25 cm Laser vibrometer * Playback device Materials and Methods Insects and Plants. A laboratory colony of spined soldier bugs was started from eggs purchased from Rincon Vitova Insectaries, Ventura, CA. The bugs, separated by sex and kept individually in plastic containers (15 cm in diameter and 6 cm in height) were reared on green beans, Phaseolus vulgaris L.; raw sunßower, Helianthus anuus L., seeds; and larvae of either almond moth, Caudra cautella (Walker); beet armyworm, Spodoptera exigua (Hübner); or navel orangeworm, Amyelois transitella (Walker). The containers were held in an environmental chamber at 26 ⫾ 1⬚C, 55% RH, and photophase of 16:8 (L:D) h. Experiments were conducted with adult males and females at least 4 d after their Þnal molt. Each experimental animal was used only once per day. Behavioral and playback experiments were conducted on cuttings of plumbago (leadwort) Plumbago auriculata Lam. (Plumbaginaceae) on which P. maculiventris may be found in the Þeld (J. S. McElfresh, personal communication). Cuttings of Y-shaped branching were cut fresh daily and supported in a 13-cm-high glass cylinder (7 cm in diameter) Þlled with vermiculite, with 20 Ð25 cm of the stem (2Ð 4 mm in diameter) protruding above the vermiculite surface (Fig. 1). Recording of Vibrational Signals and Behavior. All experiments were made in an enclosed room between 0900 and 1600 hours (3Ð10 h after the start of photophase) under daylight conditions, and relatively constant room temperature (23 ⫾ 1⬚C) and humidity. We observed no correlation between frequency of recorded vibrational signals and temperature. Three types of experiments were conducted: 1) recording of P. maculiventris vibrational signals on the loudspeaker membrane; 2) behavioral experiments on the plumbago cuttings; and 3) playback experiments on the plumbago cuttings. Fig. 1. Schematic drawing of experimental setup. When observing mating behavior male and female were placed on opposite branches of plumbago cutting; in the playback experiments (*) a female was placed on one branch and stimulus was applied to the stem of the plant. A laser vibrometer was used for recording the emitted signals in both types of experiment. Loudspeaker Recordings. To obtain recordings of emitted vibrations that were minimally inßuenced by substrate properties, a male and a female bug were placed on the membrane of a 10-cm-diameter lowmidrange loudspeaker (40 Ð 6,000-Hz frequency response, 8-⍀ impedance; RadioShack, Taipei, Taiwan), and the resulting signals were ampliÞed by a microphone ampliÞer (Sonifex Redbox ampliÞer, type RBMA, Sonifex Ltd., Irthlingborough, Northamptonshire, United Kingdom), digitized, and stored via a sound card (24-bit, 96-kHz, 100-dB signal-to-noise ratio; Sound Blaster Extigy, Creative Laboratories Inc., Milpitas, CA) on a laptop computer using Cool Edit Pro version 2.0 software (Adobe Systems Inc., San Jose, CA). Behavioral Experiments. Behavior was observed with 49 pairs of bugs, each placed on a freshly cut plumbago shoot (Fig. 1). A female and a male were placed on the leaves of opposite branches at a distance of 8 Ð12 cm apart, and behavior was observed for 20 min, or until copulation occurred. While observing males and females on the plant, vibrational signals were recorded from the plant with a portable digital laser vibrometer (PDV-100, Polytec GmbH, Waldbronn, Germany), directly digitized, and stored on the laptop computer. The laser beam was directed perpendicular to the stem or leaf long axis, parallel to the direction of propagation of the vibrational signals through the plant. To obtain better reßection of the Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 environmental noise caused, for example, by rain or wind (Casas et al. 1998). As a member of the predatory bug subfamily Asopinae, Podisus maculiventris (Say) has received extensive attention as a potential biological control agent and numerous studies have examined predatorÐprey interactions (De Clercq 2000). However, little is known about use of vibrational signals in this group, apart from one study in which P. maculiventris was shown to use substrate-borne vibrations produced by chewing caterpillars (a common prey species) as a cue for prey location (Pfannenstiel et al. 1995). The occurrence of intraspeciÞc vibrational signaling in Asopinae has to date received virtually no attention in the literature (Gogala 2006). Here, we describe the results of a thorough investigation of the vibrational song repertoire of P. maculiventris, focusing particularly on low-frequency substrate vibrations produced by mechanisms other than stridulation and abdomen vibration. To our knowledge, this is the Þrst detailed study of substrate-borne vibratory signaling for any predacious bug species. Vol. 101, no. 6 3 cm 1170 November 2008 ŽUNIČ ET AL.: SUBSTRATE-BORNE COMMUNICATION IN P. maculiventris 1171 Table 1. Frequency and temporal characteristics of songs of male P. maculiventris signals produced by abdominal vibrations tremulatory signals and composite and male N. viridula calling song and courtship song used in the playback experiments Song type Na Dominant frequency (Hz) Tremualtory signal Vibratory signal Composite signal N. viridula male calling song N. viridula male courtship song 10 25 11 30 12 9⫾2 126 ⫾ 12 142 ⫾ 3 109 ⫾ 4 109 ⫾ 4 Pulse train duration (ms) Pulse train repetition time (ms) 254 ⫾ 20 449 ⫾ 15 3901 ⫾ 474 15654 ⫾ 3663 Pulse durationb (ms) Pulse repetition time (ms) 181 ⫾ 49 224 ⫾ 18 77 ⫾ 10 (T) 177 ⫾ 19 (V) 88 ⫾ 8 299 ⫾ 37 407 ⫾ 37 546 ⫾ 223 Means ⫾ SD are shown. Number of pulses or pulse trains within a sequence repeated within 15 min. Durations of the composite pulse trains are shown separately for the tremulatory (T) and vibratory (V) pulses. a b walking had amplitude peaks at 51, 117, and 179 Hz, with the amplitudes being 19, 28, and 43 dB, respectively, below that of the dominant peak, which was at 13 Hz. Each bug was tested only once per day with only one type of stimulation program, and a freshly cut plumbago shoot was used for each bug tested. Each bug was tested with all types of stimulation programs, and only once with each type; the order in which the bugs were tested and the order of the stimulation programs used in a particular experiment were randomized. We recorded vibrational responses, and analyzed movements of stimulated bugs by measuring the number of individuals reaching the source of vibration (playback device), the number of individuals staying at the source for ⬎2 min, and the time bugs needed to reach the source of vibration when stimulated with signals from abdominal vibration and composite signals. In control experiments, bugs were tested under the same conditions and experimental setup but without any stimuli transduced to the plumbago stem (Fig. 1). Signal Analysis and Statistics. Pulses were deÞned as unitary homogenous parcels of vibrations of Þnite duration (Broughton 1963). Pulses arranged into repeatable and temporally distinct groups were termed a pulse train. Loudspeaker recordings of vibrational signals were used for analyzing the frequency, velocity, and temporal characteristics (duration; repetition time deÞned as time interval between the start of two sequential signals) (n ⫽ number of analyzed signals; N ⫽ number of bugs from which the signals were obtained). Analyses were done using Sound Forge, version 6.0 (Sonic Foundry Inc., Madison, WI) or Raven, version 1.2 (Charif et al. 2004) (Fig. 2). Velocity is a vector quantity that speciÞes time rate of change of displacement. The velocity of vibrations was measured with the laser vibrometer (at 5 mm/s/V sensitivity) and determined at the maximum amplitude of the pulse (millimeters per second). Data are reported as minimal and maximal individual means. In analyses of playback experiments, we compared the behavioral responses of controls to the behaviors recorded under each of the different stimulation conditions with two-tailed FischerÕs exact tests. The difference between the time needed to reach the source of vibration with vibratory and composite signal stimulation was compared with the nonparametric MannÐ Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 laser beam, a small piece of reßective tape (⬃1.43 mg, ⬃1 mm2) was attached to the plant surface, or ⬃1 mm2 area was painted with white typewriter correction ßuid. To investigate possible maleÐmale interactions, experiments were carried out as described above, except that two males were used instead of a male and female (20 pairs tested). Recordings from plants were made to determine whether speciÞc signal was associated with a speciÞc sequence of mating behavior. Moreover, with recordings on plants, the possible effects of the plant substrate on the spectral properties of emitted signals could be observed. Playback Experiments. In playback experiments, bugs were stimulated by signals of males prerecorded from the loudspeaker membrane, or by vibrations produced by a female walking on a plant and recorded with the laser vibrometer (Fig. 1). Vibrational stimuli were transduced to the plumbago stem (3 cm above the vermiculite surface) via a playback device that consisted of nonresonant loudspeaker membrane (10cm-diameter low-midrange, 40 Ð 6,000-Hz frequency response, 8-⍀ impedance; RadioShack), with a 3.5-cm plastic rod (8 mm in diameter) connecting the membrane and the plant (Fig. 1). The velocity of stimulation was adjusted to the level of the speciÞc signal prerecorded from a test plant with a singing male (0.5Ð1.6 mm/s, at 3-cm distance from the vibrating male). Noise produced by a female walking on the plant also was reproduced at the recorded velocity level (0.1Ð 0.01 mm/s). Six stimulation programs were synthesized. Each program included only one type of vibrational signal: 1) P. maculiventris male signal produced by abdominal vibration (vibratory signal) (30 females and 10 males tested); 2) tremulatory signals of males (30 females and 10 males tested); 3) composite pulse trains (composed of a tremualtory pulse followed by a pulse train of abdominal vibrations) (30 females tested); 4) the calling song (24 females tested) and 5) courtship song (20 females tested) of heterospeciÞc male Nezara viridula L. (pentatomid and potential prey species; De Clercq et al. 2002); and 6) vibrations produced by a conspeciÞc female walking on a plant (30 females tested). Each stimulation consisted of a short sequence of one speciÞc natural signal replayed in a loop for 15 min. Frequency and temporal characteristics of stimulatory signals are shown in Table 1. The spectrum of vibrations produced during 1172 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA MALE FEMALE FEMALE or Walk Search Song (vibratory,percussion) Approach Approach Song (tremulatory) Song (composite) Mount Copulatory attempt Walk Stationary Walk Approach Song (tremulatory,antenn.) Song (composite) Stationary Mount Copulatory attempt Copulate Copulate Whitney U test. A P value ⬍0.05 was considered signiÞcant. Results Mating Behaviors. Mating behaviors were observed with pairs of males and females placed on opposite branches of a plumbago shoot (Fig. 1). Of 98 tested bugs (49 pairs), 94 individuals walked around on the plant, with only two females and two males remaining motionless. The main steps of successful courtship behavior are described in Fig. 2. In 57% of the pairs tested (28/49), the female approached the male. Antennation was observed in 16 pairs. In nine cases the male antennated the female, and the antennated female walked away in three cases, whereas in six cases the male antennated the stationary female and then mounted and copulated with her. In seven cases, the female antennated the male. Five of seven antennated males moved away and no copulation was observed, whereas in one case the female again approached the male and they copulated. In subsequent experiments with pairs of males, a similar reaction (antennated male walking away) was observed after male-male antennation (see below). In several cases, the approaching female extended and touched the stationary male with her proboscis. In no case did we record any vibrational signals from females, whereas 80% of tested males (39/49) did emitted vibrational signals, produced by abdominal vibrations, body tremulation, or percussion with legs. Of 39 singing males, 29 (74%) successfully copulated, whereas copulation did not occur in the absence of vibrational signals. Before copulation, males always emitted tremulatory or combined tremulatory and vibratory signals; therefore, we assumed that signaling is essential for copulation to occur (Fig. 2). The behaviors of 20 pairs of males were observed under the same experimental conditions. In 11 cases, males remained motionless on the same branch or walked but did not meet, whereas in nine cases they encountered each other on the plant. When standing next to each other, Þve males emitted tremulatory signals, and two males emitted abdominal vibrations and/or percussion signals. Signals were always produced only by one male, and we recorded no regular alternation of signaling, as is typical for the rivalry songs produced by other pentatomid bugs and by Auchenoryncha (Kon et al. 1988, Claridge and Morgan 1993, Čokl and Virant 2003, Čokl et al. 2004). Vibrational Signals Produced during Mating Behaviors. P. maculiventris males produced vibrational signals by dorso-ventral vibration of the abdomen, by striking one or more legs on the substrate (percussion signal), or by tremulation, characterized by right and left movements of the body with the legs kept Þrmly on the substrate. These different signals were emitted in different behavioral contexts. Abdominal vibration signals (Fig. 3) were recorded as the Þrst song emitted by males in 12 cases (of 39 singing males). In nine cases, it was produced spontaneously when a male was standing alone on the plant, and in three cases it was produced after a pair had met, but the female had moved away from the male. Thus, abdominal vibration seems to have a calling function, triggering female movement toward the singing male, and triggering searching behavior which includes stopping at stem junctions, waiting for the next male signal, and testing possible paths with legs and antennae. Analogous searching behavior has been described for males of phytophagous stink bug N. viridula (Ota and Čokl 1991). Pulses of abdominal vibrations recorded from the loudspeaker had a mean duration of 190 ⫾ 30 ms (n ⫽ 1890, N ⫽ 13) and a repetition time of 449 ⫾ 78 ms (n ⫽ 60, N ⫽ 6), with the dominant frequency varying among individuals from 90 ⫾ 13 (n ⫽ 10) and 140 ⫾ 2 (n ⫽ 10) Hz. Compared with the spectra of abdominal vibration recorded from the loudspeaker membrane, signals recorded from plants did not exhibit a second higher frequency component peak that was seen on the loudspeaker recordings (Fig. 3). When a female approached a male emitting abdominal vibrations, he started to tremulate (Fig. 4). In 57% (22/39) of cases, the male produced tremulatory signals without prior emission of abdominal vibration signals, and in 82% (18/22) of these, this reaction was triggered by the close vicinity (1Ð2 cm) of a female. The tremulatory song (recorded on the loudspeaker) was composed of sequences of single pulses of several seconds duration (Fig. 4), with the mean duration varying among individuals from 96 ⫾ 15 (n ⫽ 10) to 138 ⫾ 28 (n ⫽ 10) ms. The repetition time varied between 286 ⫾ 46 (n ⫽ 10) and 347 ⫾ 77 (n ⫽ 10) ms, and the dominant frequency varied between 8 ⫾ 2 (n ⫽ 10) and 22 ⫾ 12 (n ⫽ 10) Hz. Before copulation, in the majority of cases (22/39), males produced regularly repeated pulse trains (Fig. 5) composed of a tremulatory pulse followed by a pulse of abdominal vibration. Composite pulse trains recorded on the loudspeaker were 260 ⫾ 32 ms in duration (n ⫽ 30, N ⫽ 3), with 473 ⫾ 57 repetition time (n ⫽ 30, N ⫽ 3), and a dominant frequency (deter- Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 Fig. 2. Main steps of successful courtship behavior observed in male and female P. maculiventris on a plumbago cutting. Vol. 101, no. 6 November 2008 ŽUNIČ ET AL.: SUBSTRATE-BORNE COMMUNICATION IN P. maculiventris 1173 mined for the vibratory unit) varying between 107 ⫾ 1 (n ⫽ 10) and 152 ⫾ 9 (n ⫽ 10) Hz. In eight cases, males emitted percussion signals (Fig. 6) in connection with abdominal vibrations, either in the pause between vibratory pulses or as a longer sequence after the abdominal vibrations terminated. Two males also emitted percussion signals between tremulatory signals, and in Þve cases per- cussion was recorded as the Þrst emitted signal. The percussion signal was produced by the insect tapping on the loudspeaker membrane with one or both of the Þrst legs, producing pulses the duration of which in three males ranged from 100 ⫾ 23 (n ⫽ 30) and 134 ⫾ 41 ms (n ⫽ 30), with repetition time varying from 464.3 to 510.4 ms. The spectra of 10s-long sequences of percussion pulses recorded on Fig. 4. Male tremulatory signal recorded on the loudspeaker (left) and on the plumbago plant (right). (A) Oscillogram of several signals. (B) Frequency spectrum of several pulses. Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 Fig. 3. Male abdominal vibration signal recorded on the loudspeaker (left) and on a plumbago plant (right). (A) Oscillogram of six pulses. (B) Frequency spectrum of one pulse. 1174 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 101, no. 6 the loudspeaker showed broad-band characteristics, with a narrow dominant frequency peak around 97 Hz, and a broader peak of lower amplitude at frequencies between 1,500 and 3,000 Hz (Fig. 6). In contrast, percussion signals recorded from plant substrate showed only the peak ⬇100 Hz. These results are consistent with observations of insectproduced vibrational signals on other herbaceous plants (Čokl et al. 2005), and they indicate that spectral properties of signals are inßuenced by the substrate, with plants acting as low-pass Þlters. The mean velocity of signals recorded from the stem 3 cm from where the male was tapping on the leaf was 0.19 ⫾ 0.10 mm/s (n ⫽ 20). When recorded (from the plumbago plant) simultaneously at the same distance from the singing male with vibratory and tremulatory signals, the velocity of percussion was 9.4 and 18.7 dB lower compared with vibratory (0.56 ⫾ 0.11 mm/s; n ⫽ 20) and tremulatory (1.63 ⫾ 0.46 mm/s; n ⫽ 20) signals respectively. Fig. 6. Male percussion signal recorded on the loudspeaker (left) and on the plumbago plant (right). (A) Oscillogram of several pulses. (B) Frequency spectrum of several pulses. Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 Fig. 5. Male-produced composite signal (composed of tremulatory and vibratory signal) recorded on the loudspeaker (left) and on the plumbago plant (right). (A) Oscillogram of nine pulses. (B) Frequency spectrum of one pulse. November 2008 ŽUNIČ ET AL.: SUBSTRATE-BORNE COMMUNICATION IN P. maculiventris Discussion The predaceous spined soldier bug, P. maculiventris, is common in North America, ranging from Mexico, the Bahamas, and parts of the West Indies in the south, and into Canada in the north (De Clercq 2000). It is found in a variety of natural habitats, including woodlands and shrubs, orchards, and Þeld crops, where it feeds on ⬎90 insect species from eight orders (McPherson 1980, De Clercq 2000). This species uses visual, chemical, tactile, and vibrational cues to locate prey (Coppel and Jones 1962, Mukerji and LeRoux 1965, Tostowaryk 1971, Pfannenstiel et al. 1995), and these sensory modalities also can be involved in intraspeciÞc communication. The secretions of the dorsal abdominal glands of males function as long-range pheromones that attract both adults and immatures of both sexes (for review, see SantÕAna et al. 1999). The smaller dorsal abdominal glands of females produce secretions that have been suggested to act as closerange aggregation pheromones (Aldrich et al. 1984). P. maculiventris reacts to moving prey at distances up to 10 cm (Heimpel and Hough-Goldstein 1994), indicating that visual signals are probably important at close range. Vibrational signals had been previously noted in two species of the pentatomid subfamily Asopinae, Picromerus bides (L.) and P. maculiventris (Gogala 2006). The results of the current study show that the predaceous spined soldier bug emits vibrational communication signals by three different mechanisms: vibration of the abdomen, tremulation of the whole body, and percussion with the front legs. To our knowledge, this is the Þrst report of vibrational signals being produced by three independent modes for the whole suborder Heteroptera. It had been shown previously that many heteropteran species communicate with signals produced by two mechanisms, stridulation and abdominal vibration (Gogala 1984, Gogala 2006), with the latter signals being tuned to the resonant properties of herbaceous plants and adapted for long range calling (Čokl et al. 2005). Long-range communication through monocotyledonous plants has been described in spiders, exchanging vibrational signals at several meter distance (Barth 2002). However, experiments with the burrower bug species Scaptocoris castanea (Perty) and Scaptocoris carvalhoi (Becker) (Čokl et al. 2006) demonstrated that stridulatory signals are signiÞcantly modiÞed and attenuated during transmission through herbaceous plants, and as such, are effective only for close range interactions such as courtship or rivalry. Of the three vibrational signals produced by different mechanisms by P. maculiventris, behavioral observations and playback experiments support the calling function of the signal produced by abdominal vibration. Considering its behavioral context, the song is analogous to the female-produced calling songs of N. viridula (Čokl et al. 2000), Acrosternum hilare (Say) (Čokl et al. 2001), Thyanta pallidovirens (Stål) and Thyanta custator accerra (McAtee) (McBrien et al. 2002), and many other pentatomid species (Čokl et al. 1978, Kon et al. 1988, Moraes et al. 2005). However, unlike the phytophagous pentatomids, to date, P. maculiventris is the only pentatomid species in which males rather than females call from one place for an extended period with a steady signal repetition rate, and silent females walk toward them with clearly expressed searching behavior at junction points indicating that they are tracking the vibrational signal. Surprisingly, females exhibited searching responses to both conspeciÞc and heterospeciÞc calling signals, but in the latter case more than half of them did not remain on the vibration source. Repetition time constituting Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 Playback Experiments. Females responded differently to stimulation with different vibrational signals. Eighty percent (24/30) and 69% (19/30) of females stimulated with abdominal vibration signals or composite signals, respectively, expressed typical searching behavior at junction points, reached the source of stimulation (playback device), and remained there for ⬎2 min. In control experiments, signiÞcantly fewer females (10%, 3/30; P ⬍ 0.001, two-tailed Fisher exact test) walked to the silent source by chance, and none of them remained on the source for ⬎2 min. Females stimulated with abdominal vibration signals reached the source of vibration signiÞcantly faster (median ⫽ 257 s, min ⫽ 56 s, max ⫽ 508 s) than those stimulated by composite signals (median ⫽ 392 s, min ⫽ 55 s, max ⫽ 730 s) (P ⬍ 0.05, two-tailed MannÐWhitney, z ⫽ ⫺2.58). Interestingly, responses also were elicited from females stimulated with N. viridula male calling song (of characteristic similar to that of conspeciÞc vibratory signal), with 56% (14/25) of the females reaching the source, versus 10% (3/30) reaching the source in the controls (P ⬍ 0.01, two-tailed Fisher exact test). There is no signiÞcant difference between the percentage of females that reach the source of vibration when presented with the two conspeciÞc versus heterospeciÞc calling song (vibratory versus hetrospeciÞc calling song P ⫽ 0.07; composite signal versus heterospeciÞc calling song; P ⫽ 0.59, two-tailed Fisher exact test). Upon reaching the source of vibration all females, when presented with conspeciÞc signals, remained on the source for ⬎2 min. However, only six females that were stimulated with the heterospeciÞc calling song remained at the source for ⬎2 min, whereas the remaining eight resumed walking within ⬍2 min of reaching the source (P ⬍ 0.001, two-tailed Fisher exact test). Only 20% (6/30) of females reached the source when stimulated with either tremulatory or walking noise, and only one and two females, respectively, remained on the source for ⬎2 min (P ⬍ 0.01, two-tailed Fisher exact test). Females did not respond to heterospeciÞc (N. viridula) male courtship song (of temporal characteristics very different from those of the conspeciÞc signal) and remain motionless during stimulation with this heterospeciÞc signal. No vibrational responses were produced by males stimulated with tremulatory and vibratory signals. Males remained motionless during the whole stimulation period or ßew off the plant. 1175 1176 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Percussion resulting from contact with the substrate has been described in many arthropod groups (Ewing 1989). For example, stoneßies tap against the substrate with the abdomen (Stewart 1997), jumping spiders produce percussive signals with the forelegs and abdomen (Elias et al. 2003), crabs and Orthoptera tap with their appendages (Salmon and Horch 1972, Sismondo 1980), and termites and beetles tap with their heads (Howse 1964, Connétable 1999, Birch and Keenlyside 1991). The percussion described here for P. maculiventris does not seem to be connected with a speciÞc signal or behavioral context, because percussion pulses were produced as the Þrst emitted signal before, after, or between emission of vibratory or tremulatory signals. When males emitted percussion signals no speciÞc responses by females were observed: nevertheless the percussion signals may carry information in the pulse repetition rate, because the spectral and amplitude pattern characteristics of broad-band signals change signiÞcantly with increasing distance along a plant substrate (Čokl et al. 2006, 2007). Thus, for example, regularly repeated percussion signals with amplitude different from vibrational signals might aid in choosing the correct direction to the source at plant crossings. Antennation was observed in femaleÐmale and maleÐmale interactions. Antennation of a male by either a female or another male triggered a speciÞc male reaction, expressed as retreat. Cuticular extracts of P. maculiventris males and females did not show any obvious sex speciÞc differences (J.G.M., unpublished data), indicating that antennation is probably connected with touch as a speciÞc mechanical stimulation rather than the detection of sex-speciÞc contact cues. The results of the current study highlight several differences between the roles of vibrational signaling in P. maculiventris and our previous knowledge of vibrational communication within the Pentatomidae. SpeciÞcally, male P. maculiventris produce vibratory signals by three different modes: abdominal vibration, body tremulation, and percussion with the legs. The signals are emitted only by males, and females search for the calling male. These major differences between the behaviors of the predatory P. maculiventris and the phytophagous Pentatomidae indicate the need for further comparative investigations of vibrational communication within the Asopinae, and other true bug subfamilies. Acknowledgments We are grateful to Mariana Krugner and J. Steven McElfresh (Department of Entomology University of California, Riverside) for rearing P. maculiventris. We thank anonymous reviewers for helpful suggestions to improve the original manuscript. This work was Þnancially supported by the Slovenian Research Agency (Programme P1-0255-0105 and Project L1-71299-0105), and by Hatch Project CA-R*ENT5181H. Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 the most reliable parameter of plant-transmitted signals is similar to N. viridula calling song (Čokl et al. 2000) and P. maculiventris vibratory and composite signals. The signiÞcant distortion of both the spectral properties and the duration of vibrational signal during transmission through the plant (Michelsen et al. 1982, Miklas et al. 2001) may explain the lower species speciÞcity of female P. maculiventris responding with searching behavior when stimulated by vibrational signals from heterospeciÞcs on plant substrates. In contrast, females were seen to discriminate between conspeciÞc and heterospeciÞc calling signals when they reached the source of vibration (playback device), where the signal integrity was preserved in duration, amplitude pattern, and frequency. These results are analogous to those from the studies of N. viridula vibrational signaling (Miklas et al. 2001). However, none of the tested P. maculiventris females responded to heterospeciÞc male courtship songs with temporal characteristics very different from the songs of their own species, and very few females responded and remained at the source of vibrations when stimulated by walking vibrations. These results indicate that P. maculiventris distinguish between vibrational signals of different origins, and respond selectively to particular substrate-borne vibration. Details of signal distortion and transmission through substrate and species recognition system will be discussed in a subsequent article (in preparation). Tremulation has been described in many insect groups but not within Heteroptera. For example, tetigoniids vertically oscillate the body without touching the ground (Busnel et al. 1955, Morris 1971, Gwynne 1977), and in katydids, tremulatory signals produced by males encode information on body size (De Luca and Morris 1998). Tremulation in P. maculiventris was different: when triggered by the presence of females, males oscillated the body parallel to the substrate. Tremulatory signals of P. maculiventris emitted alone or in combination with vibratory pulses within composed pulse trains were associated with the later stages of pair formation, as part of short range courtship. Tremulation also was characteristic of maleÐmale interactions within rivalry. Playback experiments excluded any calling function of tremulatory signals, but they were essential for copulation to occur. Thus, we hypothesize that tremulatory signals play an important part in close range species and sex recognition, a multimodal process that probably also includes antennation, visual contact, and chemical signaling. In fact, vigorous movement of the body during tremulation may induce suprathreshold air particle movement that could be detected by sensilla sensitive to air movement. Furthermore, tremulatory signals may provide information about body size, as shown for Conocephalus nigropleurum (Bruner) (De Luca and Morris 1998). Comparative investigations within the Asopinae and further behavioral experiments are needed to elucidate the level of species speciÞcity and the full biological role of tremulation in the mating behaviors of P. maculiventris and other asopine species. Vol. 101, no. 6 November 2008 ŽUNIČ ET AL.: SUBSTRATE-BORNE COMMUNICATION IN P. maculiventris References Cited their transmission through plants. Physiol. Entomol. 29: 372Ð380. Čokl, A., M. Zorovič, A. Žunič, and M. Virant-Doberlet. 2005. Tuning of host plants with vibratory signals of Nezara viridula L. (Heteroptera: Pentatomidae). J. Exp. Biol. 207: 1481Ð1488. Čokl, A., C. Nardi, J.M.S. Bento, E. Hirose, and A. R. Panizzi. 2006. Transmission of stridulatory signals of the burrower bugs, Scaptocoris castanea and Scaptocoris carvalhoi (Heteroptera: Cydnidae) through the soil and soybean. Physiol. Entomol. 31: 371Ð381. Čokl, A., M. Zorović, and J. G. Millar. 2007. Vibrational communication along plants by the stink bugs Nezara viridula and Murgantia histrionica. Behav. Processes 75: 40 Ð54. De Clercq, P. 2000. Predaceous stinkbugs (Pentatomidae: Asopinae), pp. 738 Ð751. In C. W. Schaefer and A. R. Panizzi [eds.], Heteroptera of economic importance. CRC, Boca Raton, FL. De Clercq, P., K. Wyckhuys, H. N. De Olivera, and J. Klapwijk. 2002. Predation by Podisus maculiventris on different life stages of Nezara viridula. Fla. Entomol. 85: 197Ð202. De Luca, P., and G. K. Morris. 1998. Courtship communication in meadow katydids: female preference for large male vibrations. Behaviour 135: 777Ð793. Elias, D. O., A. C. Mason, P. W. Maddison, and R. R. Hoy. 2003. Seismic signals in a courting male jumping spider (Araneae: Salticidae). J. Exp. Biol. 206: 4029 Ð 4039. Ewing, A. W. 1989. Mechanisms of sound production, pp. 16 Ð22. In ed. A. W. Ewing [ed.], Arthropod bioacoustics, neurobiology, and behaviour, Edinburgh University Press, Edinburgh, United Kingdom. Gogala, M. 1984. Vibration producing structures and songs of terrestrial Heteroptera as systematic characters. Biol. Vestn. 32: 19 Ð36. Gogala, M. 2006. Vibratory signals produced by HeteropteraÐPentatomorpha and Cimicomorpha, pp. 275Ð296. In S. Drosopoulos and M. F. Claridge [eds.], Insect sounds and communication: physiology, behaviour, ecology, and evolution. CRC, Taylor & Francis Group, Boca Raton, FL. Gwynne, D. T. 1977. Mating behaviour of Neoconocephalus ensiger (Orthoptera: Tettigoniidae) with notes on the calling song. Can. Entomol. 109: 237Ð242. Heimpel, G. E., and J. A. Hough-Goldstein. 1994. Search tactics to cues by predatory stink bugs. Entomol. Exp. Appl. 73: 193Ð197. Howse, P. E. 1964. The signiÞcance of the sound produced by the termite Zootermopsis angusticollis (Hagen). Anim. Behav. 12: 284 Ð300. Hrabar, N., A. Čokl, and M. Virant-Doberlet. 2004. Species speciÞcity of males southern green stink bug reactions to the female calling song. Acta Zool. Sin. 50: 566 Ð575. Kon, M., A. Oe, and H. Numata. 1988. Comparison of the mating behavior between two sympatric species Nezara antennata and N. viridula (Heteroptera: Pentatomidae) with special reference to sound emission. J. Ethol. 6: 91Ð98. McBrien, H. L., A. Čokl, and J. G. Millar. 2002. Comparison of substrate-borne vibrational signals of two congeneric stink bug species, Thyanta pallidovirens and T. custator accerra (Heteroptera: Pentatomidae). J. Insect Behav. 15: 715Ð737. McNett, G. D., and R. Cocroft. 2008. Host shifts favor vibrational signal divergence in Enchenopa binotata treehoppers. Behav. Ecol. 19: 650 Ð 656. Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 Aldrich, J. R., W. R. Lusby, J. P. Kochansky, and C. B. Abrams. 1984. Volatile compounds from the predatory insect Podisus maculiventris (Hemiptera: Pentatomidae): male and female metathoracic scent gland and female dorsal abdominal gland secretions. J. Chem. Ecol. 10: 561Ð568. Barth, F. G. 2002. A spiderÕs world: senses and behavior. Springer, Berlin, Germany. Birch, M. C., and J. J. Keenlyside. 1991. Tapping behavior is a rhythmic communication in the death-watch beetle, Xestobium rufovillosum (Coleoptera, Anobiidae). J. Insect Behav. 4: 257Ð263. Broughton, W. B. 1963. Methods in bioacoustic terminology, pp. 3Ð24. In R. G. Busnel [ed.], Acoustic behaviour of animals. Elsevier, Amsterdam, The Netherlands. Busnel, R. G., F. Pasquinelly, and B. Dunortier. 1955. La trémulation du corps et la transmission aux supports des vibrations en résultant comme moyen dÕinformation a courte portee des Éphippigerrès mâles et females. Bull. Soc. Fr. 80: 18 Ð23. Casas, J., S. Bacher, J. Tautz, R. Meyhőfer, and D. Pierre. 1998. Leaf vibrations and air movement in a leafminerparasitoid system. Biol. Control 11: 147Ð153. Casas, J., C. Magal, and J. Sueur. 2007. Dispersive and non-dispersive waves through plants: implications for arthropod vibratory communication. Proc. R. Soc. B 274: 1087Ð1092. Charif, R. A., C. W. Clark, and K. M. Fristrup. 2004. Raven 1.2 userÕs manual. Cornell University Laboratory of Ornithology, Ithaca, NY. Claridge, M. F., and J. C. Morgan. 1993. Geographical variation in acoustic signals of the planthopper Nilparvata bekeri (Muir) in Asia: species recognition and sexual selection. Biol. J. Linn. Soc. 48: 267Ð281. Cocroft, R. B., and R. L. Rodriguez. 2005. The behavioral ecology of insect vibrational communication. Bioscience 55: 323Ð334. Cocroft, R. B., H. J. Shugart, K. T. Konrad, and K. Tibbs. 2006. Variation of plant substrates and its consequences for insect vibrational communication. Ethology 112: 779 Ð 789. Connétable, S., A. Robert, F. Bouffault, and C. Bordereau. 1999. Vibratory alarm signals in two sympatric higher termite species: Pseudacanthotermes spiniger and P. militaris (Termitidae, Macrotermitinae). J. Insect Behav. 12: 329 Ð342. Coppel, H. C., and P. A. Jones. 1962. Bionomics of Podisus spp. associated with the introduced pine sawßy. Diprion similis (Htg.) in Wisconsin. Wisc. Acad. Sci. Arts Lett. 51: 31Ð56. Čokl, A., M. Gogala, and A. Blaževič. 1978. Principles of sound recognition in three pentatomidae bug species (Heteroptera). Biol. Vestn. 26: 81Ð94. Čokl, A., M. Virant-Doberlet, and N. Stritih. 2000. The structure and function of songs emitted by southern green stink bugs from Brazil, Florida, Italy and Slovenia. Physiol. Entomol. 25: 196 Ð205. Čokl, A., H. L. McBrien, and J. G. Millar. 2001. Comparison of substrate-borne vibrational signals of two stink bug species, Acrosternum hilare and Nezara viridula (Heteroptera: Pentatomidae). Ann. Entomol. Soc. Am. 94: 471Ð 479. Čokl, A., and M. Virant-Doberlet. 2003. Communication with substrate-borne signals in small plant-dwelling insects. Annu. Rev. Entomol. 48: 29 Ð50. Čokl, A., J. Prešern, M. Virant-Doberlet, G. M. Bagwell, and J. G. Millar. 2004. Vibratory signals of harlequin bug and 1177 1178 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Pfannenstiel, R. S., R. E. Hunt, and K. V. Yergan. 1995. Orientation of a hemipteran predator to vibrations produced by feeding caterpillars. J. Insect Behav. 8: 1Ð9. Salmon, M., and K. W. Horch. 1972. Acoustic signalling and detection by semiterrestrial crabs of the family Ocypodidae, pp. 61Ð96. In H. E. Winn and B. L. Olla [eds.], Behavior of marine animals. Plenum, New York. Sant’Ana, J., R.F.P. da Silva, and J. C. Dickens. 1999. Olfactory reception of conspeciÞc aggregation pheromone and plant odors by nymphs of the predator, Podisus maculiventris. J. Chem. Ecol. 25: 1813Ð1826. Sismondo, E. 1980. Physical characteristics of drumming of Meconema thalassimum. J. Insect Physiol. 26: 209 Ð212. Stewart, K. W. 1997. Vibrational communication in insects. Epitome in language of stoneßies? Am. Entomol 43: 81Ð91. Tostowaryk, W. 1971. Life history and behaviour of Podisus modestus (Hemiptera: Pentatomidae) in boreal forests in Quebec. Can. Entomol. 103: 662Ð 674. Virant-Doberlet, M., and A. Čokl. 2004. Vibrational communication in insects. Neotrop. Entomol. 33: 121Ð134. Received 5 February 2008; accepted 24 August 2008. Downloaded from http://aesa.oxfordjournals.org/ by guest on October 14, 2016 McPherson, J. E. 1980. A list of the prey species of Podisus maculiventris (Hemiptera: Pentatomidae). Great Lakes Entomol. 13: 17Ð24. Michelsen, A., F. Fink, M. Gogala, and D. Traue. 1982. Plants as transmission channels for insect vibrational song. Behav. Ecol. Sociobiol. 11: 269 Ð281. Miklas, N., N. Stritih, A. Čokl, M. Virant-Doberlet, and M. Renou. 2001. The inßuence of substrate on male responsiveness to the female calling song in Nezara viridula. J. Insect. Behav. 14: 313Ð332. Moraes, M.C.B., R. A. Laumann, A. Čokl, and M. Borges. 2005. Vibratory signals of four Neotropical stink bug species. Physiol. Entomol. 30: 175Ð188. Morris, G. K. 1971. Aggression in male conocephaline grasshoppers (Tettigoniidae). Anim. Behav. 19: 132Ð137. Mukerji, M. K., and E. I. LeRoux. 1965. Laboratory rearing of a Quebec strain of the pentatomid predator, Podisus maculiventris (Say) (Hemiptera: Pentatomidae). Phytoprotection 46: 40 Ð 60. Ota, D., and A. Čokl. 1991. Mate location in the southern green stink bug, Nezara viridula (Heteroptera, Pentatomidae), mediated through substrate borne signals on ivy. J. Insect Behav. 4: 441Ð 447. Vol. 101, no. 6