J. exp. Biol. (1982), 99. 469-472 469 JVith 2 figures printed in Great Britain SHORT COMMUNICATIONS SOUND PRODUCTION BY ABDOMINAL STRIDULATION IN THE AUSTRALIAN MURRAY RIVER CRAYFISH, EUASTACUS ARMATUS BY DAVID C. SANDEMAN* AND LON A. WILKENSf Department of Neurobiology, Research School of Biological Sciences, The Australian National University, Canberra, A.C.T. 2601, Australia Received 27 November 1981 - Accepted 12 March 1982) The Murray crayfish, Euastacus armatus, is a large, spiny, freshwater crayfish (Fig. 1 A), from south eastern Australia. Like the northern hemisphere astacid crayfish, Euastacus and other parastacids assume a defensive posture when they are provoked: the chelae are raised, the legs are spread out, and the abdomen is extended. In Euastacus, the abdominal extension is accompanied by a clearly audible hiss. This sound is produced by abdominal stridulatory organs in which bristles projecting from the posterior edges of the abdominal tergites rub over ridges on the surfaces of the tergites of the next posterior segment (Fig. 1B, C). Extension of the abdomen causes the 6 mm wide field of tergal ridges to rotate forwards and under each row of bristles. Sound production either in air or in Water can be silenced by coating the hairs with vaseline or by shaving them off with a sharp scalpel. If the crayfish are held lightly by the cephalothorax, on the ground, or suspended in water, they will often attempt to escape by flipping their tails. Grasped strongly, so that the cephalothorax is compressed, the animals will give a defensive response. Although compression of the cephalothorax is known to inhibit escape responses in Procatnbarus, mediated either by giant or non-giant neuronal pathways (Wine & Krasne, 1975), this stimulus had not been previously reported to elicit the defensive reflex in crayfish. A rapid extension of the abdomen to this stimulus may be unique to the Murray crayfish where it has the adaptive value of reinforcing an auditory alarm signal with the display of a heavily armed tail. We have analysed the stridulatory response of animals in both air and water environments, and under a variety of conditions. Recordings in air were made with a microphone (Bruel and Kjaer, Type 4135) positioned about 5 cm from the animal. The signals were amplified, filtered (with bandpass limits of 100 Hz and 30 khz) and fed directly to a Nicolet MED-80 computer for spectral analysis. Using a 50 khz maximum sampling rate the stridulation sound was analysed over the range from 0-25 khz, and found to contain a broad spectrum of frequencies (4-18 khz) but with Present address: School of Zoology, University of New South Wales, Kensington. New South Wales 2033, Australia. t Present address: Department of Biology, University of Missouri, St. Louis, Missouri 63121 U.S.A.
470 DAVID C. SANDEMAN AND L. A. WILKENS major peaks in the range of 5-12 khz. The absence of major components of the signai above 12 khz suggests that errors introduced by inadequate filtering (i.e. betwees 25 and 30 khz) have not been introduced in this case. The sound pressure level of the stridulation in air was measured with a sound level meter (Bruel & Kjaer 2209, 2 Hz-70 khz). Sound pressures 5 cm from the animal in the 5 to 12 khz range were between 62 and 63 db re 20 /ipa. Stridulation under water was recorded from hand-held animals, where the auditory signal could be evoked at will by a sudden squeeze to the cephalothorax. Animals were suspended in a polyethylene tank (40 by 40 cm and 30 cm deep) 15 cm from a hydrophone (Bruel and Kjaer, Type 8103) so that their legs were prevented from touching the sides or the floor of the tank. The auditory signals were amplified and filtered as above, and recorded with a Nagra IV-SJ tape recorder. These signals, replayed at one-tenth the recorded speed, were photographed with a continuous recording camera and frequency analysed with the computer. The stridulatory signal under water varied from 150-350 ms in duration and produced a sound pressure of 30-45 Pa at a distance of 15 cm. By comparison, the sound pressure produced by various species of West Indian snapping shrimps is in excess of 60 Pa, at a similar distance (Fish, 1964). An abdominal extension/stridulation response was sometimes followed by repeated tail flips as the animal tried to escape by swimming. The abdominal extension phase during swimming bouts produced much smaller stridulatory signals. No signal was produced when the abdomen was flexed (Fig. 2A). Sequentially blocking the stridulatory organs on each segment with vaseline revealed that the anterior abdominal segments produced most of the signal (Fig. 2 B-E). Spectral analysis of stridulation under water also revealed a broad range of frequencies, with peaks ranging from 4-12 khz when all segments were active. When the number of contributing segments was reduced, by greasing them, the frequency range was narrowed. For a single segmental organ, the frequencies lay between 5-9 khz (Fig. 2F, G) and the signal had a more sinusoidal wave form than that recorded when all five segments operated together (Fig. 2F, G). The fundamental frequency of the underwater stridulatory signal was 6-5 khz, i.e. four to ten times the measured rate at which the ridges were crossed by the bristles during abdominal extension. Trimming the bristles to about half their length removed the higher frequency components of the signal and shifted the fundamental frequency from 6-5 khz to about 4-5 khz. This is surprising as one would expect the shortened bristles to have a higher resonant frequency. The bristles are not simple structures however but have fine side branches, (Fig. 1B), which may produce the higher frequency component of the signal in the intact animal. In the absence of other obvious structures, we conclude that part, if not all, of the stridulatory signal comes from the tergal bristles which vibrate as they move over the ridges. A word of caution must be added here because underwater acoustic signals in small tanks can be significantly distorted (Silver & Halls, 1980). Thus the pressure levels, and even frequencies of the stridulatory signal we report here must remain tentative until they have been confirmed by measurements done in a free field. Stridulation in the Murray crayfish, coupled as it is with the extension of the armed
Journal of Experimental Biology, Vol. 99 Fig. i Fig. i. (A) Euastacus armatus, the Australian Murray River Crayfish. B-D. Scanning electron micrographs showing (B) bristles projecting from the caudal edge of the first abdominal tergite and (C and D) the ridges of the second abdominal segment at two magnifications. The bristles are viewed from beneath. Scale bars: (A) 25 mm, (B) 250 /tm, (C) 250 fim, (D) 25 ion. D. C. SANDEMAN AND L. A. WILKENS {Facing p. 47 )
Sound production by abdominal stridulation 471 0-1 s 1 ms 0 t 12 24 khz Fig. 2. The stridulatory signal in water, recorded with a hydrophone and photographed from an oscilloscope. The top trace (A) shows a signal evoked by suddenly squeezing the cephalothorax of a hand-held animal. Stridulation was followed by a brief 'swim' in which the abdomen was flipped beneath the body and re-extended. Traces, B, C, D and E show a decrease in amplitude of the signal as the stridulatory organs of the abdomen were progressively silenced with vaseline, beginning with the rostral segments. In za, the stridulatory organs on all segments were active. In 2 B, C, D and E, four, three two and one organ, respectively, were active. When all segments were greased no signal could be recorded during abdominal extension. The traces in F and G show parts of the expanded stridulatory signal (left-hand side) and associated power spectra (right-hand side) from an animal in which all the organs were active (F 1-3) and when only the organ of segments 1 and 2 was active (G 1-3). The three traces in each series represent samples, respectively, from the beginning, middle and end of the stridulatory burst. The power spectra, which give the normalized relative power of the frequencies contained in each sample, show that the beginning and end of each stridulatory burst contain fewer sound frequencies than the middle of the burst (F). Also, eliminating all but one of the organs results in the loss in intensity of most of the frequencies above 9 khz (G). Amplifier sensitivity and gain settings are constant throughout the series of traces in A-E and for all records in F, G. abdomen during agonistic behaviour, appears to be a good example of acoustic aposematism (a warning noise). This phenomenon is well known from studies on insects (Masters, 1979, 1980). Known underwater predators of the crayfish are the large Murray cod, and stridulation may be directed at these animals. However, crayfish also stridulate during natural underwater encounters with each other. In one instance, two animals were lured at night to a bait anchored near a hydrophone. The first animal to discover the bait was feeding on it when the second individual approached. The first one turned toward the newcomer, lunged and stridulated with the effect that the approaching animal was driven off. Crayfish have also been observed to stridulate while driving off others in a large tank in the laboratory, but other crayfish in the tank not directly involved in the encounters gave no visible response to the stridulation. We have, therefore, no evidence to suggest that conspecifics are able to detect stridulation, for the reaction of one animal to the attack by another could have been released by either visual or tactile cues. Stridulation in air could be useful to Euastacus, for these animals do leave the water temporarily, in search of food, although a puzzle remains with regard to the evolution of stridulation in air since none of the extant indigenous Australian mammals
47 2 DAVID C. SANDEMAN AND L. A. WILKENS are likely predators. The Tasmanian tiger {Thylacinus cynocephalus) and the Tasmanian devil (Sarcophilus harrisii), however, both now confined to Tasmania, were onca widespread over the Australian continent and may be candidates. The airborne stridulatory signal could be a vestige of a previous age, perhaps again important as a deterrent against more recently introduced predators, such as water rats, foxes and dingoes. We thank Ray Stone, for taking the scanning electron micrographs, and Mick Shea and Don Macgillvray for field observations on the biology of Euastacus. Supported in part by a grant from the Whitehall Foundation. REFERENCES FISH, M. P. (1964). Biological sources of sustained ambient sea noise. In Marine Bio-acoustics (ed. W. N. Tavolga), pp. 175-194. New York: Pergamon Press. MASTERS, W. M. (1979). Insect disturbance stridulation: its defensive role. Behav. Ecol. Sociobiol. 5, 187-200. MASTERS, W. M. (1980). Insect disturbance stridulation: characterization of airborne and vibrational components of the sound. J. comp. Physiol. A. 135, 259-268. SILVER, S. C. & HALLS, J. A. T. (1980). Recording sounds of hydropsychid larvae - a cautionary tale. J. comp. Physiol. 140, 159-161. WINE, J. J. & KRASNE, F. B. (1975). Extrinsic modulation of crayfish escape behaviour. J. exp. Biol. 63, 433-450.