J. exp. Biol. 130, 425-432 (1987) 425 Printed in Great Britain The Company of Biologists Limited 1987 SHORT COMMUNICATION A NEW MICROCOMPUTER-BASED METHOD FOR MEASURING WALKING PHONOTAXIS IN FIELD CRICKETS (GRYLLIDAE) BY JOHN A. DOHERTY* AND ANTHONY PIRES Section of Neurobiology and Behavior, Seeley G. Mudd Hall, Cornell University, Ithaca, NY 14853, USA Accepted 13 March 1987 Calling songs of male crickets attract sexually mature, conspecific females for mating (for recent reviews see Eisner & Popov, 1978; Huber & Thorson, 1985; Doherty & Hoy, 1985). This communication system has been the subject of many behavioural studies on the relevant properties of the male's calling song for recognition by females (e.g. Walker, 1957; Popov & Shuvalov, 1977; Pollack & Hoy, 1979, 1981; Thorson, Weber & Huber, 1982; Stout, DeHaan & McGhee, 1983; Doherty, 19856; Doolan & Pollack, 1985). The behavioural investigations have required some means of measuring the female's phonotaxis (her locomotion or turning towards a sound source). These include 'closed-loop' methods in which the animal moves freely in acoustic space and 'open-loop' methods in which the animal is tethered and is not allowed to experience changes in sound intensities as it runs in the direction of the sound source. Closed-loop methods include free walking in arenas (e.g. Murphey & Zaretsky, 1972; Paul, 1976; Hoy, Hahn & Paul, 1977; Pollack & Hoy, 1979; Stout et al. 1983); open-loop methods include tethered flight (Moiseff, Pollack & Hoy, 1978; Pollack & Hoy, 1979, 1981; Pollack, Huber & Weber, 1984), tethered walking on Y-maze globes (Hoy & Paul, 1973) and free walking on a spherical locomotion compensator (Wendler, Dambach, Schmitz & Scharstein, 1980; Thorson & Huber, 1981; Thorson et al. 1982; Pollack et al. 1984; Doherty, I985a,b,c; Schmitz, 1985). Here we describe a new open-loop method for quantifying cricket phonotaxis. This method uses inexpensive microcomputer technology, is completely automated and therefore rapid and objective, and could be adapted for studying locomotory movements of other animals. The data obtained are in such a form that they are easily compared to data generated by the spherical locomotion compensator (Kramer treadmill, see Kramer, 1976; Wendler^ al. 1980; Weber et al. 1981), which has been used for quantifying phonotaxis in the field cricket, Gryllus bimaculatus (Doherty, 1985a, b,c). Present address: Department of Biology, Villanova University, Villanova, PA 19085, USA. Key words: crickets, phonotaxis, communication, microcomputer.
426 J. A. DOHERTY AND A. PlRES The spherical locomotion compensator or Kramer treadmill was first developed by scientists in West Germany (Kramer, 1976; Wendlere* al. 1980; Weber et al. 1981; Thorson et al. 1982). It uses an infrared detection system to monitor movements of an untethered cricket on top of a sphere. This positional information feeds back to servomotors that move the sphere in the opposite direction. These compensatory movements 'fix' the cricket at the top of the sphere as it performs walking phonotaxis. The colloquial name for this locomotion compensator is 'Kugel' (German translation of the word 'sphere'). Because our device for measuring walking phonotaxis utilizes the Apple Macintosh computer and is similar in appearance to the German Kugel, we call our new device the 'MacKugeP. The main mechanical difference between these two devices is that in the MacKugel system the cricket is tethered and its own walking movements provide the power to rotate the sphere. The MacKugel is a complete stimulus control and data acquisition system for online studies of cricket phonotaxis. Data on cricket movements during phonotaxis experiments are passed directly to an Apple Macintosh microcomputer, which can also synthesize acoustic stimuli or send control signals to an external acoustic synthesizer. In designing this system, we took advantage of ROM routines in the Macintosh for quantifying movements of the Apple 'mouse'. The mouse is a mechanical, optoelectronic device that converts rotational movements of a rubber ball to changes in the coordinates of a cursor displayed on the computer's CRT monitor. The ball bears against the operator's desk surface, and against three rollers inside the ball's housing. One roller is for mechanical support only. The other two, representing orthogonal x- and y-axes, are connected to rotating vanes which interrupt light beams between LED phototransistor pairs. The phototransistors produce electronic signals which are sent to the computer. We adapted the mouse to study cricket phonotaxis by expanding the distance between the rollers and replacing the small rubber ball (2-5 cm diameter) with a larger and much lighter sponge-rubber ball (10 cm diameter, 8-7g). This larger ball (a hollowed-out, toy 'Nerf ball' from Parker Bros, Beverly, MA, USA) was mounted in a frame along with the roller/phototransistor modules from the mouse (Fig. 1). In this way the rollers could be actuated by a tethered cricket walking on top of the ball. As the cricket ran in one direction (i.e. towards an attractive acoustic stimulus), the ball was moved in the opposite direction and the movements were transduced into changing pixel coordinates on the Macintosh screen. The minimum force required to move the Nerf ball ranged from 1-117 to 2-793xlO~ 2 N. A plastic sphere (10 cm diameter, 12-8 g) is currently under development and this improved version only requires forces ranging from 0-69 to 1-05x10" N. High-level computer languages (Macintosh Pascal ) were used for data acquisition and analysis (program available from author). By sampling pixel coordinates once every second, we were able to calculate instantaneous velocity and direction profiles for the cricket's movements. These profiles, along with their corresponding 'vector plots' (see below), were displayed simultaneously on the Macintosh screen, and the data were saved on micro-floppy disks for subsequent data analysis.
Measurement of phonotaxis in crickets RT U 427 **- F ^ t KF rv X 1 TTTjll S t @ Fig. 1. Side (A) and top (B) view of MacKugel device. TM, tether mount; F, frame; RT, rotating tether; PR, passive roller; T,, x-axis tachometer (composed of phototransistor module, extension axle and roller); T y, y-axis tachometer; FB, frictionless bearing (composed of a 1 cm Teflon ball riding on a film of air); AS, air supply (air routed through a threaded aluminium cylinder used to adjust the height of the Nerf ball); S, sphere. To establish the usefulness and validity of our method, we repeated some locomotion compensator experiments run previously on the field cricket, G. bimaculatus (Doherty, 1985a,b). In both methods, walking phonotaxis in crickets can be quantified by measuring time profiles of the cricket's walking velocity and direction. Fig. 2 shows these time profiles for one MacKugel experiment in which a synthetic calling song was presented to a female G. bimaculatus from one of two loudspeakers separated by 180. This calling song was played from the left speaker for 2min and then from the right speaker for another 2min. A 30-s silent period preceded and followed this playback. The female tracked the stimulus, as shown by her narrow meandering about the direction of the active speaker. Her walking velocity also increased when the stimulus was presented, compared to the preceding silent period. The time profiles shown in Fig. 2 were generated by sampling and storing the position of the animal once each second during an experiment. This method of data collection easily lent itself to another way of representing phonotactic movements visually. For each 1-s interval, a vector was calculated that showed the animal's walking direction relative to the direction of the active speaker (vector angle) and its velocity (vector length). These vectors were cumulatively plotted as shown in Fig. 2. These 'vector plots' show that when the stimulus was presented from either the left or right speaker, the vectors were clustered about the direction of the active speaker.
428 J. A. DOHERTY AND A. PlRES The direction component of cricket walking phonotaxis is a sensitive indicator of the attractiveness of an acoustic stimulus. To investigate this sensitivity further, we ran phonotaxis experiments in which acoustic stimuli with different pulse periods were presented in a sequential, to-and-fro paradigm. As in earlier experiments of cricket phonotaxis on a locomotion compensator (Doherty, 1985c), pulse periods ranging from 30 to 50 ms yielded the best tracking on the MacKugel (i.e. best orientation to active speaker direction, see Fig. 3). CP PD "a. 1313 76%.^ T 4 Angular d1ion 2-0 0' RS LS 360' 1 min r Fig. 2. Phonotaxis of a female Gryllus bimaculatus on the MacKugel device. A synthetic calling song composed of 4-pulse chirps repeated at a constant rate (40 ms pulse period, PP; 4 pulses per chirp, PN; 22 ms pulse duration, PD; 400 ms chirp period, CP; shown in box) was presented from one of two speakers for 4 min. The cricket's movements were translated into cursor movements on the CRT of a microcomputer. The cursor's position (in x y pixel coordinates) was sampled once every second and vector plots and time profiles of the cricket's velocity and angular direction were calculated and plotted during the experiment. The two vector plots show the accumulation of vectors during stimulus playbacks from the two loudspeakers. Each vector has an angle and a length. Vector angle is the direction the animal was moving during the sample interval, relative to the position of the active speaker. Vector length is directly proportional to the animal's velocity. Every vector was plotted with its origin at the centre of the vector plot circle. The radius of this circle corresponds to a velocity of 3-4cms~'. The solid dot beside each vector plot indicates the position of the active speaker (left or right speaker; LS, RS), which corresponds to the active speaker position (serrated horizontal line at either 90 or 270 ) in the angular direction profile below. The numbers above each vector plot are the vector score (left-hand number, defined in text) and the percentage of the stimulus presentation time that the animal was moving (right-hand number). The vector score (2426) and the percentage of time spent moving (97 %) were both higher when the stimulus was played from the right speaker.
Measurement of phonotaxis in crickets 429 In these same phonotaxis experiments on the MacKugel we found slight effects of pulse period on the walking velocity profiles. Females walked for a greater percentage of the stimulus presentation time when pulse periods were optimal (i.e. 30-50ms) 0-6 -i 39 40 0-4- 0-2- c O 00-0-2 10 20 30 40 50 60 Pulse period (ms) 70 80 Fig. 3. Effects of synthetic calling songs with different pulse periods on the accuracy of orientation to the speaker. Orientation was quantified as follows: for each 1-s interval of a trial in which the animal was moving, the cosine of the vector angle was calculated. A mean cos (angle) was then obtained for all such intervals within a trial. An average cos (angle) was computed for all trials of each pulse period and plotted as orientation to speaker. If the cricket ran directly towards the loudspeaker (defined as 0 ), the cos (angle) was 1; this value was 1 when it ran directly away from the speaker and 0 when it ran perpendicular to the sound source. The points are means for each pulse period and the vertical lines are standard errors. Data were based on to-and-fro scans of pulse period for 21 female Gryllus bimaculatus. Sample sizes are shown. 70-i 600-1 B 20 30 40 50 60 i -100 70 80 10 20 Pulse period (ms) 30 40 50 60 70 80 Fig. 4. Comparison of two different methods of quantifying stimulus efficacy in sequential, to-and-fro scans of pulse period (see text and Doherty, 1985a,c). The two methods were (A) manual measurements of the percentage of stimulus presentation time that females tracked the stimulus and (B) microcomputer measurements of vector score. The criteria for manual scoring of stimulus tracking are found in Doherty (1985a,c). The points are mean values and the vertical lines are standard errors. Nine Gryllus bimaculatus females were tested. Sample sizes are the same in A and B.
430 J. A. DOHERTY AND A. PlRES and walked less in response to stimuli with pulse periods outside this range. Furthermore, mean walking velocity was highest in response to pulse periods of 35 and 40 ms, whether or not pauses in walking were included in the calculation. Because both walking velocity and direction were affected by acoustic stimuli, we devised a numerical 'vector score' which incorporates both of these components and serves as a quick measure of the phonotactic efficacy of different sound stimuli. This score is defined as: Vector score = 2[cos(vector angle) X vector length]. In calculating the vector score, the angular direction of the active speaker is always defined as 0 and the summation is over all the sample intervals within a trial. When the animal runs towards the loudspeaker the vector score increases, when it runs away from the loudspeaker the score decreases. By definition, movements perpendicular to the speaker axis (cos 90 or cos 270 ) and no movements at all (vector length = 0) result in vector scores approaching zero. For example, in Fig. 2, the vector score in response to calling song from the left speaker was lower than that for the right speaker. This difference was due to less accurate tracking, wider meanderings about the speaker direction and more pauses in walking. The vector score is comparable to other measures of stimulus efficacy in cricket phonotaxis. In earlier studies of G. bimaculatus phonotaxis on a locomotion compensator, the attractiveness or efficacy of an auditory stimulus was quantified by measuring the percentage of the stimulus presentation time that the female clearly 'tracked' the stimulus. Tracking has been defined as meandering within a certain angle 'window' (±60 ) about the angular direction of the active speaker (see Thorson et al. 1982; Doherty, 1985a,b,c). These same criteria were used in our study for measuring the percentage of time spent tracking in a group of nine G. bimaculatus females. In to-and-fro sequential experiments, tracking scores and vector scores in response to calling songs with different pulse periods were comparable (Fig. 4). Both scores showed that songs with pulse periods between 35 and 50 ms were most effective in eliciting positive phonotaxis. The MacKugel system is also useful for quantifying cricket phonotaxis in twostimulus (choice) playback experiments. The total vector score was a sensitive, objective measure of the relative attractiveness of alternative acoustic stimuli. When given a choice between alternating chirps with different pulse periods (PP), female G. bimaculatus clearly preferred the standard chirp stimulus (40 ms PP) to the alternative chirp stimuli with pulse periods of 20, 30, 50 and 60 ms. The strength of the female's preference for the standard over the alternative pulse period was also reflected in the mean vector score. These results were consistent with those of previous choice experiments on a locomotion compensator (Doherty, 1985c). The MacKugel system is an efficient, rapid and objective method for measuring the efficacy of different acoustic stimuli in eliciting walking phonotaxis of crickets. Using this system the data can be displayed to give a qualitative, visual representation of an individual cricket's movements. Complicated time profiles of movement velocity and direction can be reduced objectively to a numerical score that can be
Measurement of phonotaxis in crickets 431 used as a rough basis of comparison between individuals and between experimental treatments. Data analysis can be done 'on-line' as the animal is behaving. Other strengths of the MacKugel are its low price, simplicity of design, and efficient implementation of existing microcomputer technology. This research was supported by grants from NSF and NIH to JAD, AP and Ronald R. Hoy. We thank Alan Cohen and Rick Nicoletti for technical assistance and Wendy Sussdorf for preparing Fig. 1. Ronald Hoy, Peter Brodfuehrer and Jud Crawford read earlier drafts of this manuscript. REFERENCES DoHERTY, J. A. (1985a). Temperature coupling and 'trade-off phenomena in the acoustic communication system of the cricket, Gryllus bimaculatus De Geer (Gryllidae). J. exp. Biol. 114, 17-35. DOHERTY, J. A. (19856). Trade-off phenomena in calling song recognition and phonotaxis in the cricket, Gryllus bimaculatus (Orthoptera, Gryllidae). J. comp. Physiol. 156, 787-801. DOHERTY, J. A. (1985c). Phonotaxis in the cricket, Gryllus bimaculatus De Geer: Comparisons of choice and no-choice paradigms, jf. comp. Physiol. 157, 279-289. DOHERTY, J. A. & HOY, R. R. (1985). Communication in insects. III. The auditory behavior of crickets: Some views of genetic coupling, song recognition, and predator detection. Q. Rev. Biol. 60, 457-472. DOOLAN, J. M. & POLLACK, G. S. (1985). Phonotactic specificity of the cricket Teleogryllus oceanicus: intensity-dependent selectivity for temporal parameters of the stimulus. J. comp. Physiol. 157, 223-234. ELSNER, N. & POPOV, A. V. (1978). Neuroethology of acoustic communication. Adv. Insect Physiol. 13,229-335. HOY, R. R., HAHN, J. & PAUL, R. C. (1977). Hybrid cricket auditory behavior: Evidence for genetic coupling in animal communication. Science 195, 82-84. HOY, R. R. & PAUL, R. C. (1973). Genetic control of song specificity in crickets. Science 180, 82-83. HUBER, F. & THORSON, J. (1985). Cricket auditor)' communication. Scient. Am. 253, 60-68. KRAMER, E. (1976). The orientation of walking honeybees in odour fields with small concentration gradients. Physiol. Entomol. 1, 27-37. MOISEFF, A., POLLACK, G. S. & HOY, R. R. (1978). Steering responses of flying crickets to sound and ultrasound: Mate attraction and predator avoidance. Proc. natn. Acad. Set. (J.SA. 75, 4052-4056. MURPHEY, R. K. & ZARETSKY, M. D. (1972). Orientation to calling song by female crickets, Scapsipedus marginatus (Gryllidae). J. exp. Biol. 56, 335-352, PAUL, R. C. (1976). Species specificity in the phonotaxis of female ground crickets (Orthoptera: Gryllidae: Nemobiinae). Ann. ent. Soc.Am. 69, 1007-1010. POLLACK, G. S. & HOY, R. R. (1979). Temporal pattern as a cue for species-specific calling song recognition in crickets. Science 204, 429-432. POLLACK, G. S. & HOY, R. R. (1981). Phonotaxis to individual rhythmic components of a complex cricket calling song. J. comp. Physiol. 144, 367 373. POLLACK, G. S., HUBER, F. & WEBER, T. (1984). Frequency and temporal pattern-dependent phonotaxis of crickets {Teleogryllus oceanicus) during tethered flight and compensated walking. J. comp. Physiol. 154, 13-26. POPOV, A. V. & SHUVALOV, V. F. (1977). Phonotactic behavior of crickets. J. comp. Phvsiol. 119, 111-126. SCHMITZ, B. (1985). Phonotaxis in Gryllus campestris L. (Orthoptera, Gryllidae). III. Intensity dependence of the behavioural performance and relative importance of tympana and spiracles in directional hearing. J. comp. Physiol. 156, 165-180.
432 J. A. DOHERTY AND A. PlRES STOUT, J. F., DEHAAN, C. H. & MCGHEE, R. W. (1983). Attractiveness of the male Acheta domesticus calling song to females. I. Dependence on each of the calling song features..7. comp. Physiol. 153, 509-521. THORSON, J., WEBER, T. & HUBER, F. (1982). Auditory behaviour of the cricket. II. Simplicity of calling-song recognition in Gryllus, and anomalous phonotaxis at abnormal carrier frequencies. J. comp. Physiol. 146, 361-378. WALKER, T. J. (1957). Specificity in the response of female tree crickets (Orthoptera, Gryllidae, Oecanthinae) to calling songs of the males. Ann. ent. Soc. Am. 50, 626 636. WEBER, T., THORSON, J. & HUBER, F. (1981). Auditory behavior of the cricket. I. Dynamics of compensated walking and discrimination paradigms on the Kramer treadmill..7. comp. Physiol. 141, 215-232. WENDLER, G., DAMBACH, M., SCHMITZ, B. & SCHARSTEIN, H. (1980). Analysis of the acoustic orientation behavior in crickets (Gryllus campestris L.) Naturwissenschaften 67, 99 100.