* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Species-Recognition in the Field Cricket
Document related concepts
AMER. ZOOL., 22:597-607 (1982) Species-Recognition in the Field Cricket, Teleogryllus oceanicus: Behavioral and Neural Mechanisms1 RONALD R. H O Y , GERALD S. POLLACK, 2 AND ANDREW MOISEFF 3 Section of Neurobiology and Behavior, Langmuir Laboratory, Cornell University, Ithaca, New York 14850 SYNOPSIS. Field crickets depend on acoustic organs to detect the presence of potential predators as well as conspecific crickets. Predators are recognized largely on the basis of spectral frequencies that are contained in their acoustic signals. Puffs of air and very low frequencies activate a cricket's cereal receptors and ultrasonic frequencies activate their tympanal organs. Both of these acoustic stimuli release "escape behavior," in the form of evasive movements. An identified neuron sensitive to ultrasound is described. Crickets recognize singing conspecifics by both frequency and temporal properties of cricket songs; however species recognition requires specific temporal information in calling songs. While previous studies have emphasized the role of songs on female behavior, males also recognize conspecific songs; sexual differences in recognition behavior occur. ' From the Symposium on From Individual to Species Recognition: Theories and Mechanisms presented at the the necessity of recognizing categories of signallers—being able to tell predators from crickets, for example. In the latter case we refer to discrimination within a category— homospecific from heterospecific crickets, for example. In principle, crickets could also perform individual recognition, but thus far such has not been demonstrated. The final question, "Where is it?" is not directly related to the matter of recognition and will not be dealt with here. The cricket's conspicuous acoustic signals are vitally important, for they mediate social behavior—the calling song of the adult male is stereotyped and species-specific and acts as a beacon to attract distant conspecific females to his territory for mating. It is by discriminating between the calling songs of different species that females find reproductive partners; this discrimination behavior serves as our operational definition of species "recognition." In this brief review we do not have space to describe the extensive and excellent work on the many aspects of acoustic behavior in the singing orthoptera. We must restrict ourselves to acoustic behavior in field crickets, and furthermore, we will focus primarily on our own work on the Austra- Annual Meeting of the American Society of Zoologists. 27-30 December 1980, at Seattle, Washington. 2 Current address of G. S. Pollack is: Department of Biology, McGill University. Montreal, Canada. 3 Current address of A. Moiseff is: Division of Biology. California Institute of Technology, Pasadena, California. interested reader should consult the excellent reviews of acoustic behavior in other crickets and insects that can be found in Michelsen and Nocke (1974), Eisner and Popov (1978), and Huber (1978). INTRODUCTION Field crickets are primarily active nocturnally and depend heavily on acoustic signals by which they monitor their environment. Like other acoustically dependent animals they must solve three problems: 1) What is producing the acoustic stimulus? Among the relevant sources of sound in the cricket's world are other crickets, other kinds of singing insects, and acoustic predators of various species. 2) Who is producing the sound? Crickets must be able to discriminate among different species of sympatrically singing crickets in order to differentiate their own species signals from those of others. 3) Where is the signaller located in the environment? Obviously, crickets must be able to localize the sources of sounds, whether they emanate from conspecific crickets or from potential predators. The questions "What is it?" and "Who is it?" are related to the issue of signal recognition. In the former case we refer to lian field cricket, Teleogryllus oceanicus. The 597 598 HOY ETAL. T H E PRODUCTION AND RECEPTION OF ACOUSTIC SIGNALS Among crickets only males sing; they do so by rubbing their forewings together, a behavior called stridulation. Each closing stroke of the wings produces a relatively pure tone (field crickets songs are pitched in a range between 3 and 8 kHz, depending on species) due to the frictional mechanics and resonant properties of the wings themselves (Michelsen and Nocke, 1974). The opening stroke is silent. Thus, the calling song is composed of a train of tones in a rhythm determined by the timing of the contraction cycles of opener and closer muscles; this rhythm is neurogenic, stereotyped, and species-specific (Huber, 1962, 1964, 1978; Bentley and Kutsch, 1966). Differences among calling songs of different species focussed attention on the calls as species-isolating mechanisms (Alexander, 1962). As an example, the call rhythm of Teleogryllus oceanicus is shown in Figure 1. Both males and females hear by means of specialized tympanal organs that are located on the forelegs. This "ear" lies internal to a distinct tympanal membrane upon which sound waves act. The membrane has a peak resonant vibration that corresponds to the fundamental frequency of the male's call (Johnstone etal., 1970; Dragsten etal., 1974; Larsen and Michelsen, 1978). These vibrations activate a group of about 70 auditory receptor neurons whose axons project to the prothoracic ganglion (Young and Ball, 1974; Eibl, 1978; Esch et al., 1980). There they activate an unknown number of interneurons that carry auditory information to the brain and to lower motor centers that control locomotion; several auditory interneurons have been identified (Stout and Huber, 1972; Rheinlaender et al., 1976; Casaday and Hoy, 1977; Popov etal., 1978; Wohlersand Huber, 1978; Elepfandt and Popov, 1979). RECOGNIZING PREDATORS BY EAR: T H E PITCH IS THE SWITCH In its world a cricket hears not only other crickets, it hears potential predators. Crickets are sensitive to a surprisingly wide range of frequencies; in fact, the frequen- - Phrase- Chirp -Trill tnn B mi*——ii' Intra Htrillii ii M I WHIHIH -Intrachirp -Inlertrill FIG. 1. Temporal structure of the species-specific calling song of T. oceanicus shown as an oscillograph. Calling song is composed of a series of repeating stereotyped phrases. Each phrase is composed of an introductory series of pulses (chirp), followed by a series of doublet pulses (trills). The three types of interpulse intervals are defined in the oscillograph. The time bar equals one second. cy band devoted to social communication is only a narrow one considered in light of the insect's auditory capabilities. Figure 2 shows a frequency range from zero Hz to 100,000 Hz—from infrasound to ultrasound. Crickets are sensitive over a good part of this range; compare this with the auditory sensitivity of humans, which spans a range of 50 Hz to 15,000 Hz. Crickets achieve this sensitivity to airborne vibrations through two kinds of auditory organs. They detect very low frequencies with a pair of appendages called cerci that project from the tip of the abdomen. The sensitivity of the cerci extends from between nearly zero Hz in the infrasound to almost 1,000 Hz. Depending on species, the rest of the frequency range of sensitivity begins at 1-2 kHz and extends to at least 100 kHz. This wide band of higher frequencies is detected by the tympanal organ, or "ear," in the tibia of each foreleg. Toads are insectivorous predators that capture prey by flicking out their tongues at their victims. These prey-capture movements set up a pressure wave (low frequency vibrations) that activates particular sensory hairs on a cricket's (or cockroach's) cerci. Precisely which hairs become stimulated depends on the direction of the wind or vibrational source. These receptors in turn activate giant interneurons in the CNS that mediate directional escape responses in crickets and cockroaches (Roeder, 1967; Matsumoto and Murphey, 1977; Camhi, 1980). In addition to cerci, other low frequency receptor organs exist but their behavioral role is uncertain (Dambach, 1972). Other predators produce ultrasound. SPECIES-RECOGNITION IN CRICKETS S&9 FIG. 2. The range of hearing in field crickets. The frequency line is drawn logarithmically, from zero Hertz (Hz), through the infrasound and terminating in the ultrasound at 100,000 Hz (100 kHz). Low frequency stimuli are detected by the cricket's cerci, and higher frequencies are detected by its tympanal organ or "ear." Below the frequency line are drawn typical sources of sound that fall within the cricket's range of audition: terrestrial predators such as frogs produce low frequency vibrations, crickets produce middle frequency vibrations, and flying bats produce ultrasound. Many moth species have ears, and are sensitive to the ultrasonic echolocation sounds of insectivorous bats; the acoustically mediated evasive behavior of moths to bats is well-known through the classic work of Kenneth Roeder (1967). Some species of crickets also fly at night (Ulagaraj and Walker, 1973; Popov etal, 1975) and would be vulnerable to bat predation. While critical observations of bats actually hunting down crickets on the wing are lacking, studies from our laboratory indicate that if crickets were pursued by echolocating bats, they would be able to make evasive maneuvers to escape them (see also Rheinlaender et ai, 1976). Crickets can be induced to fly in the lab by placing them on the end of a tether and directing a stream of wind past them; while these "flying" crickets flap their wings and respond to directional sources of sound by making directed movements of their abdomens and hind legs that can be interpreted as steering movements (Moiseff et al., 1978) that are analogous to comparable steering behavior in other tethered, flying insects (Dugard, 1967). Figure 3A-C shows diagrams taken from photographs of flying crickets responding to acoustic stimuli. In the absence of sound a cricket flies with a symmetrical flight posture, with its longitudinal body axis perfectly straight. When a series of sound pulses consisting of pure 5 kHz tones is played from a speaker on the cricket's left, the insect bends its abdomen and legs to the left, a rudder-like action that would propel the insect toward the speaker, were the insect not tethered. However, when the sound pulses are composed of 40 kHz tones, the cricket's abdomen and legs abruptly veer to the right; in free-flight this would propel the cricket away from the sound source. Thus, 5 kHz tones elicit positive phonotactic movements and 40 kHz cause negative phonotactic movements. The sign of the movements makes behavioral sense: 5 kHz is the carrier of fre- 40 kHz i ill i iii II I I 3 4 56 8 0 20 Frequency (kHz) j_ 1 i i i i r 40 60 80 100 FIG. 3. Crickets shown during an instant of tethered flight. Drawings are made from photographs taken with a strobe-flash. In A, the cricket is shown flying in the absence of acoustic stimulation. In B, the cricket is played a 5 kHz calling song from a speaker on its left; note that the abdomen, legs, and one antenna are oriented toward the sound source. In C, the acoustic stimulus was played at 40 kHz from the same speaker as in B. Note that the abdomen and legs are now oriented away from the sound source; such movements of the appendage would steer a free-flying cricket away from the sound. In D, the frequency sensitivity of steering behavior is demonstrated; the data are drawn from experiments on 12 animals. The filled circles represent the median threshold intensities and the bars indicate the ranges. The animals are most sensitive to narrow frequency band between 4 and 6 kHz; a secondary area of sensitivity ranges broadly in the ultrasound between 20 and 100 kHz. From Moiseff et ah, 1978. SPECIES-RECOGNITION IN CRICKETS quency of the calling song of Teleogryllus oceanicus; female crickets are attracted to this frequency. Forty kHz is in the ultrasonic range and occurs in the vocalizations produced by insectivorous bats (Griffin, 1974); crickets attempting to escape from echolocating bats would be expected to react to 40 kHz aversively. The frequency sensitivity of steering behavior can be ascertained by examining a behavioral tuning curve (Fig. 3D) made by measuring the threshold sound intensities required to elicit a phonotactic response as a function of the tone frequency. T. oceanicus is most responsive to tones in the range 4-6 kHz, with peak sensitivity at 5 kHz; a second area of sensitivity occurs in the ultrasound, from 25-100 kHz. While these data reflect the spectral sensitivity of phonotactic behavior, it is from the direction, or sign, of phonotaxis that its behavioral importance can be inferred. In Figure 6 we see that crickets steer toward middle frequency tones (3-10 kHz) but steer away from ultrasonic tones (above 15 kHz). We believe that these data define the information content contained in the frequency spectrum of a sound: middle frequencies are interpreted by crickets as other crickets, and are therefore approached; ultrasonic frequencies are interpreted as something to avoid (predacious bats?). The data in Figure 3, provide insight into the way that simple acoustic information (in this case, frequency) is translated into adaptive behavior. In addition, flight phonotaxis is a powerful behavioral assay for the investigation of the mechanisms that underlie species recognition (Pollack and Hoy, 1979, 1981). To summarize, the spectral frequency in a sound pulse carries important information having to do with predation: 1) Very low frequency sounds activate the cereal system and mediate escape behavior in walking crickets. 2) Ultrasonic frequencies activate the tympanal system and also mediate escape movements in flying crickets. Thus, particular frequency bands are "labelled" and stimuli falling within them elicit escape responses. The task of "recognizing" predators is relatively easy and 601 is done by frequency analysis of a sound. In a narrow frequency band ("middle" frequencies), approximately 2-8 kHz, are found most cricket calls. The role of frequency in species recognition will be dealt with later. ULTRASOUND SENSITIVITY: A NEURAL ANALYSIS Escape behaviors in invertebrate animals have been fertile ground for a neuroethological analysis, for example: crayfish swimming (Wiersma, 1947; Wine and Krasne, 1972); grasshopper jumping (Pearson, 1980); moth flight (Roeder, 1967); and cockroach running (Roeder, 1967; Camhi, 1980). We have made an attempt to analyze the ultrasound mediated "escape" behavior of Teleogryllus oceanicus (Moiseff and Hoy, 1979; Moiseff, 1980). An identified auditory interneuron in the CNS of the cricket Teleogryllus oceanicus (in- terneuron-1—Casaday and Hoy, 1977) has been shown to be highly sensitive to ultrasonic stimulation, including artificial pulse trains designed to resemble bat echolocation signals. Figure 4 shows a diagram of the neuron in the prothoracic (auditory) ganglion, and its response to a train of ultrasonic stimuli delivered from an electronic sound generator arranged to simulate a bat echolocation signal; this identified interneuron is sensitive to ultrasonic stimuli from 20-100 kHz (Moiseff and Hoy, 1979; Moiseff, 1980). Its response resembles the sensitivities of neurons in the auditory systems of many moths known to be preyed upon by bats (Roeder, 1967). Perhaps the most interesting feature of this neuron is its dichotomous response characteristic: while it is excited by ultrasound, its activity appears to be suppressed by lower frequencies—4 to 10 kHz—precisely those frequencies used by crickets for social communication. In fact, peak inhibition occurred at 5 kHz, the carrier frequency of the oceanicus call. This was shown by performing a two-tone inhibition experiment (Fig. 5) which was suggested by our colleague, R. Capranica. First, the neuron was excited by a pulse of ultrasound to establish a standard excitatory response (Fig. 5A). Then, a 5 kHz tone was pre- 602 HOY ETAL. E only E+145 B 60 80 LULL WLL W1L JULL lllllll FIG. 4. lnterneuron-1 and its response to ultrasonic pulses. At the top is shown a camera-lucida drawing of interneuron-1, made from a whole mount of the prothoracic ganglion (see Casaday and Hoy, 1977) stained with cobalt. The neuron has a major dendritic arborization in the auditory neuropile (AN) and it sends its axon (A) to the brain; its cell body (S) is seen in the contralateral hemiganglion. In the lower panels are shown the responses of int-1 to simulated bat echolocation pulses. The first panel shows int-1 responding to 0.5 msec, 30 kHz-tone pulses delivered above threshold (here, 80 dB SPL); the repetition rate is 10 pulses per second. In the second panel, the pulse repetition rate was increased to 300 pps, a rate found in the terminal phase of echolocation. The recordings were made from int-1 intracellularly, with a dye-filled micropipette. sented simultaneously with the 40 kHz excitatory tone (Fig. 5B-D). When played loudly enough, the 5 kHz tone added to the 40 kHz tone reduced the cell's excitatory discharge; the combination tone was inhibitory compared to the excitatory tone alone. Thus, int-l's activity is inhibited by frequencies that occur in the social signals of its own species. The hypothesis that it functions to detect the ultrasonic calls of bats is attractive and we present Figure 6 to support it; the lower graph shows the phonotactic performance of tethered flying crickets as a function of stimulus tone frequency; the upper graph shows the effect of tone frequency on the interneuron's state of activity. At those frequencies that "attract" the cricket (promote positive phonotaxis) the interneuron is inhibited from firing, whereas at those frequencies that stim. - ^ ^ — • FIG. 5. Two-tone inhibition in interneuron-1. In A, the neuron was presented with an excitatory tone alone (30 kHz at an intensity 10 dB above threshold). In B, C, and D, the excitatory tone (same parameters as in A) was paired with a 5 kHz tone at varying levels of intensity, 45, 60, and 80 dB SPL respectively. The effect of the combination tone was a diminished response from the neuron—the 5 kHz tone is inhibitory. The stimulus bar denotes the duration of the stimulation period—220 msec, for both excitatory and inhibitory tones. "repel" the cricket (promote negative phonotaxis) the interneuron is excited. These observations are consistent with the hypothesis that the interneuron plays a role in the cricket's negative phonotactic behavior to ultrasound. SPECIES RECOGNITION IN CRICKETS A long history of phonotaxis experiments on crickets has shown that the temporal pattern of the calling song carries critical information upon which speciesspecific discrimination is based (Walker, 1957; Zaretsky, 1972; Popov and Shuvalov, 1977; Pollack and Hoy, 1979, 1981). While it is clear that carrier frequency is of importance (Hill, 1974; Moiseff et ai, 1978), temporal pattern is the primary cue. We will review the phonotactic behavior of just one species, Teleogryllus oceanicus, the subject of our own studies (Moiseff et al., 1978; Pollack and Hoy, 1979, 1981). In their natural habitat in Australia, 7Yleugryllus oceanicus females are able to discriminate between the calls of oreanims males and those of a sympatrk speiies, 7". SPECIES-RECOGNITION IN CRICKETS 100 Inh 0 Exc 100 100 Pos 0 Neg 100 1 1 8 O 20 30 40 Frequency (KHz) FIG. 6. Possible role of interneuron-1 in phonotactic behavior. The neural response (Top) of int-1 as a function of frequency is compared with the phonotactic behavior (Bottom) of a tethered, flying cricket responding to acoustic stimuli at various frequencies. In the range 4—6 kHz, the response of int-1 is inhibited, and the flying cricket approaches these frequencies. From about ID kHz to 40 kHz, the interneuron is excited, and at these frequencies flying crickets steer away. The bars in the int-1 data indicate the percentage (of 8 experimental preparations) of neurons that showed excitation or inhibition (measured by the two-tone inhibition criterion). The bars in the behavioral data indicate the percentage (of 12 animals tested) of animals that steered toward the sound source (positive) at a given frequency, or away (negative). commodus (Hill et al., 1972). We have established laboratory colonies of both species and investigated their phonotactic behavior, especially during tethered flight. We have shown that T. oceanicus females prefer the calls of oceanicus to those of commodus in both walking and flight phonotaxis (Hoy and Paul, 1973; Pollack and Hoy, 1979). Hereafter, we shall discuss only data from flight phonotaxis experiments. The role of temporal pattern in speciesspecific song recognition was studied by making electronic song models of normal and altered oceanicus calling song (Pollack and Hoy, 1979, 1981). In Figure 7 the results of several phonotaxis experiments are summarized. Figure 7B-D shows electronic song models of normal oceanicus call (B), "altered" oceanicus call (C), and the call of the sympatric species, T. commodus (D). The orderly procession of chirp intervals and trill intervals are species-specific characteristics and are distinctly different in ocean- 603 icus and commodus (Leroy, 1963; Bentley and Hoy, 1972; Hill etal, 1972). The electronic song models were constructed from the data of Bentley and Hoy (1972), and the females responded as well to synthetic as to real songs. The temporal stereotypy of the normal species calls is manifested in two ways. First, the pulse-to-pulse orderliness of pulse intervals is apparent from the sequential interval plots (E, F). Second, the relative proportions of each interval type can be shown in a conventional histogram plot (H-J). Clearly, oceanicus and commodus differ from each other in both of these ways. The altered oceanicus song (C, F, I) differs from normal oceanicus song in that the sequential orderliness of pulse intervals has been "randomized," by repeatedly shuffling the ordinal positions of the song intervals throughout an experiment. However, it is identical to normal oceanicus song in that the "dosage" of each kind of interval is unchanged; the same interval histogram is generated for oceanicus and shuffled oceanicus songs. The relative attractiveness of these songs for oceanicus females during tethered flight is shown in Figure 7, K-M, which summarizes the results from 36 crickets (12 for each experiment). The ordinate shows the proportion of trials in which one song was chosen over another, although both songs were presented to the cricket simultaneously. Crickets in tethered flight show their song preference by flexing their abdomens and hindlegs in the direction of the speaker playing the "favored" song; speakers are set at 90° to the body axis. When oceanicus females were presented with the choice of oceanicus and commodus songs they overwhelmingly chose oceanicus (K); in a few cases they did not make a choice (no postural change), and these are labelled "Nd" on the abscissa. This result was expected, for it corroborates the result for walking phonotaxis (Hill et al, 1972; Hoy and Paul, 1973; Pollack and Hoy, 1979) in Teleogryllus. More surprising was the inability of oceanicus females to make a phonotactic choice between oceanicus song and the shuffled version of oceanicus song (Fig. 7L); the proportion of females that chose one or the other was the same, and the major- 604 HOY ETAL. HIM iniii ii ii IMI mi m i n i nit LL Intrachirp B ^J— Intratnl C Intertnl D MWHHHNHWHII HI HUN I I I I I 1 Ml HI INI 6 I" 200 E 150 10 20 30 Interval number ra 100 so V u a> a. 10 40 0 20 30 40 100 ISO 200 H 75 50 • 25 . 1 . 0 SO 100 . 1 . ISO 200 0 50 100 ISO 200 0 Interval duration (msec) M 1.0 1.0 1.0 0.5 0.5 0.5 0 L Oc Nd Co 0 L Oc Nd Sh Sh Nd Co FIG. 7. In A, oscillograph of T. oceanicus calling song showing the three types of interpulse intervals. In B—D, oscillographs of electronically synthesized songs showing oceanicus, shuffled song, and T. commodus song, respectively; time bars represent 1 sec. In E-G, sequential interval plots for the three songs shown directly above; the abscissa represents the ordinal positions of intervals in each song and the ordinate shows the duration of each interval; the orderliness of interval sequence seen in the T. oceanicus calling song is absent from the shuffled song. In H—J, histograms showing the proportions of the three interval types in the three songs; note that the oceanicus song and the shuffled song generate the same interval histogram, which is clearly different from that for commodus. In K—M, phonotactic preferences of T. oceanicus females during two-choice discrimination tests. Tests were made on tethered, flying animals; two different calling songs were presented simultaneously, one from the animal's right, the other from its left. The results are given as the proportion of trials in which the songs were selected (choices shown on abscissa). Abbreviations: Oc = oceanicus song preferred. Co = mmmndm song preferred, and Nd = no discrimination, which meant that the cricket did not make a steering movement toward either speaker during the experiment. Data drawn from 36 animals. (From Pollack and Hov. 1979.) SPECIES-RECOGNITION IN CRICKETS 605 ity of females failed to choose at all. The of such behavior differs depending on sex inability to choose between normal and (Dawkins and Krebs, 1978). Whatever the shuffled oceanicus songs could be inter- status of current theories, it seems likely preted as their being equivalent. To test that further investigation of sex-linked difthis, females were presented with a choice ferences in acoustic communication in between shuffled oceanicus and normal crickets will provide insight for behavioral commodus songs. Again, females over- theorist and neuroethologist alike; e.g., whelmingly chose oceanicus (even when Cade, 1979. shuffled) over the heterospecific commodus SUMMARY (Fig. 7M). The inability of oceanicus females to dis- It is clear that acoustic behavior in the tinguish between normal and shuffled cricket requires that its auditory system be oceanicus songs, but to clearly prefer shuf- able to analyze spectral frequency as well fled oceanicus to commodus songs, can be in-as temporal pattern in a signal. Spectral terpreted as meaning that oceanicus fe- frequency is "labelled" in the sense that an males recognize the temporal structure by extremely broad range of frequencies is dithe dosages of the three interval types con- vided into several categories of sender-antained in the song. Thus, the strict sequen- imals (Fig. 2). Very low frequencies actitial ordering of each interval within a song vate receptors on the cerci, prominent phrase is not necessary for recognition; it abdominal appendages, and signal the is sufficient to provide intervals in species- presence of possible land predators. At specific proportions. high frequencies (ultrasound), we have A recent and important turn in this story shown that flying crickets make vigorous warrants mention. All of the phonotaxis efforts to fly away from the sources of ulexperiments reported above were per- trasound. This behavior is mediated by formed on female oceanicus. However, male tympanal organs located on the forelegs. oceanicus also fly and perform in-flight Presumably, such a behavior evolved from phonotaxis. Like females, they are attract- predation by echolocating bats which hunt ed to the song of homospecific oceanicus flying insects at night; we make this specover the heterospecific commodus. How- ulation by analogy with the clearly demever, unlike females, males prefer normal onstrated acoustically-mediated escape beoceanicus over shuffled oceanicus songs and havior of moths when chased by bats they prefer shuffled oceanicus to commodus (Roeder, 1967). Using ultrasound we have songs; in experiments thus far, 11 male been unable to provoke escape responses, oceanicus have preferred normal oceanicus such as hopping or running away, from song to none for shuffled song, and 27 crickets that are walking or standing still. failed to disciminate; however, when tested Presumably, on land, predators give themagainst commodus song, 11 males preferred selves away by issuing forth infrasound vithe shuffled song to none for commodus; brations or puffs of air, but not by ultraagain, 27 males failed to make discrimi- sonic signals. However, to a cricket on the nations in the tests (Pollack, unpublished wing, it is ultrasound by which a predator results). Thus, sex differences occur in the announces its presence. acoustic communication in crickets, as have Cricket sounds occur in a narrow band been reported in the singing behavior of of "mid-range" frequencies (3-8 kHz). songbirds (Nottebohm and Arnold, 1976; While different species of crickets may inNottebohm, 1980). These and other sex- deed call with different carrier frequendifferences in song attractiveness (Pollack, cies, the divergence between two sympatric 1980) may reveal important differences in species may be as little as 500-1,000 Hz. the information content or meaning of the There is no doubt that carrier frequency species-specific calling song that depend on in the sound pulses of the call carries imthe sex of the recipient of the signal. Re- portant information. However, differcent speculations on the evolution of com- ences in temporal pattern are critical for munication behavior stress that the pay-off species-specific recognition. Thus, in social 606 HOY ETAL. communication, crickets utilize both cues, frequency and temporal pattern of acoustic stimuli. It has been easier to make a neural analysis of behaviors due to low frequency auditory stimulation (Matsumoto and Murphey, 1977; Edwards and Palka, 1974) and ultrasonic stimulation (Moiseff and Hoy, 1979) than of social communication. Undoubtedly this is due in some measure to the ease with which escape responses can be reproduced in the laboratory and the participation of "giant" neurons in the neural circuits that underlie escape in the cricket as well as other animals. However, studies on crickets are going on internationally to understand precisely what it is that crickets are saying to each other and what is going on in their nervous systems that explains that elusive behavioral act called "recognition." An important contribution of modern ethology has been to demonstrate the utility of the comparative approach in the study of animal behavior. It is to be expected that comparative studies will prove equally enlightening in neuroethology, and that understanding mechanisms of recognition in the auditory behavior of crickets will be informative for similar analyses in higher animals. ACKNOWLEDGMENTS We wish to acknowledge the contributions of our neuro-behavioral colleagues at Cornell, particularly in Dr. Bob Capranica's and Dr. Jeff Camhi's lab groups. Drs. Bruce Land and Melvin Kreithen, and Margaret Nelson provided facilities and advice at critical times. Drs. Margaret Nelson and Jean Fraser commented on the manuscript. We acknowledge, gratefully, the support of the NIH (NINCDS) for a postdoctoral award to G.S.P., and research and RCDA grants to R.R.H. REFERENCES Alexander, R. D. 1962. The role of behavioral study in cricket classification. System. Zool. 11:53—72. Bentley, D. R. and W. Kutsch. 1966. The neuromuscular mechanism of stridulation in crickets (Orthoptera: Gryllidae). J. Exp. Biol. 45:151-164. Cade, W. 1979. The evolution of alternative male strategies in field crickets. In M. S. Blum and X. Camhi, J. M. 1980. The escape system of the cockroach. Sci. Amer. 243:158-172. Casaday, G. B. and R. R. Hoy. 1977. Auditory interneurons in the cricket Teleogryllus oceanicus: Physiological and anatomical properties. J. Comp. Physiol. 121:1-13. Dambach, M. 1972. Der vibrationssinn der Grillen. J. Comp. Physiol. 79:281-304. Dawkins, R. and J. R. Krebs. 1978. Animal signals: Information or manipulation? In J. R. Krebs and N. B. Davies (eds.), Behavioral ecology, Blackwell Sci. Publ., Oxford. DeVoogd, T. J. and F. Nottebohm. 1980. Golgi architectonics of a region involved in song production in the canary. Soc. Neurosci. Abstr. 5:625. Dragsten, P. R., W. W. Webb, J. A. Paton, and R. R. Capranica. 1974. Auditory membrane vibrations: Measurements at sub-angstrom levels by optical heterodyne spectroscopy. Science 185:55-57. Dugard, J. J. 1967. Directional changes in flying locusts. J. Insect Physiol. 13:1055-1063. Edwards, J. S. and J. Palka. 1974. The cerci and abdominal giant fibres of the house cricket, Acheta domesticus. I. Anatomy and physiology of normal adults. Proc. R. Soc. London B 185:83-103. Elephandt, A. and A. V. Popov. 1979. Auditory interneurons in the mesothoracic ganglion of crickets. J. Insect Physiol. 25:429-441. Eisner, N. and A. V. Popov. 1978. Neuroethology of acoustic communication. Adv. Insect Physiol. 13:229-355. Esch, H., F. Huber, and D. Wohlers. 1980. Physiological and anatomical properties of sensory auditory fibers in crickets, Gryllus bimaculatus DeGeer, Gryllus campestris L. J. Comp. Physiol. 137:27-38. Griffin, D. R. 1974. Listening in the dark. Dover Publ., New York, New York. Hill, K. G. 1974. Carrier frequency as a factor in phonotactic behavior of female crickets (Teleogryllus commodus). J. Comp. Physiol. 93:7-18. Hill, K. G., J. J. Loftus-Hills.and D. F. Gartside. 1972. Pre-mating isolation between the Australian field crickets Teleogryllus commodus and T. oceanicus ( O r t h o p t e r a : Gryllidae). Aust. J. Zool. 20:153-163. Hoy, R. R. and R. C. Paul. 1973. Genetic control of song specificity in crickets. Science 180:82-83. Huber, F. 1962. Central nervous control of sound production in crickets and some speculations on its evolution. Evolution 16:429—442. Huber, F. 1964. The role of the central nervous system in Orthoptera during the coordination and control of stridulation. In R. G. Busnel (ed.), Acoustic behavior of animals. Elsevier, Amsterdam. Huber, F. 1978. The insect nervous system and insect behavior. Anim. Behav. 26:969-981. Johnstone, B. M., J. C. Saunders, and J. R. Johnstone. 1970. Tympanic membrane response in the cricket. Nature 227:625-626. Larsen, O. N. and A. Michelsen. 1978. Biophysics of the ensiferan ear III. The cricket ear as a fourinput svstem. ]. Comp. Phvsiol. 123:217-227. A. Blum (eds.). Sexual selection and reproductive competition in insects, Academic Press, New York. Leroy, Y. 1963. Le chant de cour chez les grillons SPECIES-RECOGNITION IN CRICKETS 607 des genres Gryllus L. et Teleogryllus Chop. (Or- Popov, A. V. and V. F. Shuvalov. 1977. Phonotactic behavior of crickets. J. Comp. Physiol. 119: thopteres, Ensiferes). C. R. Acad. Sci. Paris 111-126. 257:3223-3226. Matsumoto, S. G. and R. K. Murphey. 1977. The Popov, A. V., V. F. Shuvalov, and A. M. Markovich. cercus-to-giant interneuron system of crickets. IV. 1975. Spectrum of the calling songs phonotaxis Patterns of connectivity between receptors and and the auditory system in the cricket Gryllus bithe medial giant interneurons. J. Comp. Physiol. maculatus. J. Evol. Biochem. Physiol. 2:453-460. 119:319-330. Rheinlaender, J., K. Kalmring, A. V. Popov, and H. Michelsen, A. and H. Nocke. 1974. Biophysical asRehbein. 1976. Brain projections and inforpects of sound communication in insects. Adv. mation processing of biologically significant Insect Physiol. 10:247-296. sounds by two large ventral-cord neurons of Gryllus bimaculatus DeGeer. J. Comp. Physiol. Moiseff, A. 1980. Auditory interneurons and phonotactic behavior in the Australian field cricket, 110:251-269. Teleogryllus oceanicus. Ph.D. Diss., Cornell Uni- Roeder, K. D. 1967. Nerve cells and insect behavior. versity, Ithaca, New York. Harvard Univ. Press, Cambridge. Moiseff, A. and R. R. Hoy. 1979. Frequency de- Stout, J. F. and F. Huber. 1972. Responses of central auditory neurons of female crickets (Gryllus pendent response properties of an identified interneuron in the cricket Telegryllus oceanicus. Soc. campestris L.) to the calling song of the male. Z. Neurosci. Abstr. 4:255. Vergl. Physiologie 76:302-313. Moiseff, A., G. S. Pollack, and R. R. Hoy. 1978. Ulagaraj, S. M. and T. J. Walker. 1973. Phonotaxis of crickets in flight: Attraction of male and feSteering responses of flying crickets to sound and male crickets to male calling songs. Science ultrasound: Mate attraction and predator avoid182:1278-1279. ance. Proc. Natl. Acad. Sci. U.S.A. 75:4052-4056. Nottebohm, F. 1980. Age and hormones influence Walker, T. J. 1957. Specificity in the response of female tree crickets (Orthoptera, Gryllidae, growth of vocal control stations of the canary Oceanthinae) to calling songs of the males. Ann. brain. Soc. Neurosci. Abstr. 5:633. Ent. Soc. Amer. 50:626-636. Nottebohm, F. and A. P. Arnold. 1976. Sexual dimorphism in vocal control areas of the songbird Wiersma, C. A. G. 1947. Giant nerve fiber system of the crayfish. J. Neurophysiol. 10:23-38. brain. Science 194:211-213. Pearson, K. G., W. J. Heitler, and J. D. Steeves. 1980. Wine, J. J. and F. B. Krasne. 1972. The organization of escape behavior in the crayfish. J. Exp. Triggering of locust jump by multimodal inhibBiol. 56:1-18. itory interneurons. J. Neurophysiol. 43:257-278. Pollack, G. S. 1980. Sexual differences in song rec- Wohlers, D. W. and F. Huber. 1978. Intracellular recording and staining of cricket auditory interognition in crickets. Amer. Zool. 20:853. neurons. J. Comp. Physiol. 127:1 1-28. Pollack, G. S. and R. R. Hoy. 1979. Temporal patterns as a cue for species-specific calling song rec- Young, D. and E. Ball. 1974. Structure and develognition in crickets. Science 204:429—432. opment of the auditory system in the prothoracic leg of the cricket Teleogryllus commodus (Walker). Pollack, G. S. and R. R. Hoy. 1981. Phonotaxis to individual rhythmic components of a complex I. Adult structure. Z. Zellforsch. 147:193-312. cricket calling song. J. Comp. Physiol. 144:367- Zaretsky, M. D. 1972. Specificity of the calling song 373. and short term changes in the phonotactic rePopov, A. V., A. M. Markov, and A. S. Andjan. 1978. sponse by female crickets (Scapsipedus marginatns) Auditory interneurons in the prolhoracic gan(Gryllidae). J. Comp. Physiol. 79:153-172. glion of the cricket, Gryllus bimaculatus DeGeer. J. Comp. Physiol. 126:183-192.