Squeaking with a sliding joint: mechanics and motor control of sound production in palinurid lobsters
Duke University, Biology Department, Durham, NC 27708, USA
Address for correspondence: University of California, Department of Integrative Biology, 3060 VLSB 3140, Berkeley, CA 94720-3140, USA (e-mail: patek{at}socrates.berkeley.edu )
Accepted 15 May 2002
![]() |
Summary |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: stridulation, Palinuridae, spiny lobster, Panulirus argus, Palinurus elephas, joint, biomechanics, sound production, bioacoustics, stringed instrument
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Spiny lobsters (Palinuridae) also produce sound
(Fig. 1). At first glance, the
sound-producing mechanism found in some palinurids appears similar to typical
arthropod washboard mechanisms, such as those used by crickets, in which a
hard pick (the plectrum) rubs over hard macroscopic ridges (the file).
However, a closer look at the spiny lobster's morphology reveals that the
plectrum is made of soft tissue and the file lacks macroscopic ridges
(Phillips et al., 1980). The
pulsed sound, called the `rasp', is made by rubbing two macroscopically smooth
surfaces together, which produces sound through frictional interactions
between the surfaces (Patek,
2001b
).
|
These frictional interactions are analogous to the frictional
stick-and-slip mechanism in stringed instruments
(Patek, 2001b), whereby a bow
rubs over a string and elicits vibrations in the string
(Benade, 1990
). In a model of
stick-and-slip sound production in bowed stringed instruments (e.g. violins),
a bow sticks and slips over the surface of the string as a result of static
and sliding friction between the surfaces
(Benade, 1990
)
(Fig. 2A). With slight
modification, this violin model can be applied to the spiny lobster's
apparatus in which the soft tissue plectrum is a mobile mass/spring system
that moves over the stationary surface of the file
(Fig. 2B)
(Patek, 2001b
). In this case,
the plectrum consists of a mass connected between two springs, and the
mass/spring system is pulled over the file. As it is pulled, static friction
between the plectrum and file surfaces causes the two surfaces to stick
relative to each other. The two plectrum springs eventually extend/compress to
a point at which the static friction between the plectrum and file surfaces is
exceeded and the two surfaces slide relative to one another. The soft, elastic
tissue of the plectrum resists compression and tension and probably stores
energy during the stick phase and releases energy during the slip phase. This
type of mechanism has not been reported in any other biological system and may
be unique to the sound-producing palinurids
(Patek, 2001b
).
|
The stick-and-slip mechanism is found at an intriguing location compared
with most arthropod acoustic mechanisms that use adjacent rubbing surfaces,
such as those mentioned above. The sound-producing apparatus is located at the
proximal antennal joints which, in other taxa, are limited to one degree of
freedom and do not permit translational motion between two adjacent surfaces.
Arthropod joints are usually hinges with two articulations, one degree of
freedom and a simple flexorextensor muscle arrangement
(Alexander, 1983). However, the
sound-producing palinurids generate sound by rubbing the plectrums over the
files through a translational movement of the proximal antennal joints
(Fig. 1).
In this study, I address three questions: (i) what structural modifications permit the translational movement of the proximal antennal joint; (ii) how do muscles control sound production given the unique architecture of the antennal joint; and (iii) how do the acoustic surfaces and antennal muscle anatomy vary across palinurid lobsters? To address these questions, I examined the antennal joint mechanics and associated muscle anatomy in both non-sound-producing and sound-producing lobsters. I measured antennal muscle activity and kinematics in the Caribbean spiny lobster Panulirus argus during rasp sound production and other antennal movements. The muscle activity patterns were used to assess the functional implications of evolutionary variation in palinurid antennal muscles. The acoustic surfaces of sound-producing lobsters were compared using scanning electron microscopy.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Muscle and joint anatomy
Muscles attaching to the base of the antennae control movement of the
proximal antennal joint. The muscles attaching to the first segment of the
antenna had not previously been described in a sound-producing lobster,
although they had been described in a non-sound-producing lobster (Jasus
lalandii) in the palinurid family
(Paterson, 1968). I dissected
15 Panulirus argus and examined joint morphology and the anatomy of
muscles attaching to the posterior edges of the first segment of the antennae.
The muscles were identified using the nomenclature from Paterson
(1968
). The artist Natalia
Rybczynski observed three of these dissections and illustrated the anatomy, as
seen in Fig. 4. Palinurus
elephas and Jasus verreauxi specimens were also dissected. Joint
morphology was inspected in forty-one preserved palinurid species.
|
Acoustics and kinematics
Rasps were recorded to test for correlations between sound, movement and
muscle activity. A hydrophone (HTI-94-SSQ, High Tech Inc.) was placed 0.3 m
deep and at least 0.15 m from the lobster. Output from the hydrophone was
recorded digitally using an analog/digital (A/D) acquisition board
(AT-MIO-16E-10, National Instruments).
To measure the timing of plectrum movements over the file, I attached a
Hall-effect sensor (0.06 g, A3515LUA, Allegro MicroSystems) just posterior to
the file and a samarium/cobalt magnet (0.03 g) to the plectrum. The sensor
tracked changes in magnetic flux density as the plectrum moved over the file
(Deich et al., 1985). The
magnet was glued to the plectrum using a combination of dental repair resin
(Hygenic Corp.) and cyanoacrylate glue. I wired the sensor and coated it with
aquarium silicone sealant to exclude water. The sensor was attached with a
sticky rosin/wax mixture, and the wires were led from the cephalothorax of the
lobster to the A/D board. The voltage output was measured using a customized
LabVIEW virtual instrument (National Instruments) sampling at 3750 or 5000
samples s-1. Several trials were run at 10 000-25 000 samples
s-1 to verify that lower sampling rates collected accurate temporal
data. Acoustic and kinematic data were recorded simultaneously to correlate
rasp pulses and plectrum movements. I conducted kinematic experiments with 97
rasps produced by six Panulirus argus. The onset and offset times of
the acoustic pulses and the plectrum movement were measured.
Electromyography
Electromyography was used to measure muscle activity during sound
production and other antennal movements in Panulirus argus
(techniques described in Basmajian and
Luca, 1985; Loeb and Gans,
1986
). Recordings were made from the medial (Pm) and lateral (Pl)
lobes of the promotor, the lateral levator (LL) and the depressors (Da, Dc,
Dd) (see Results and Fig. 4 for
a description of muscle anatomy). Small holes were drilled through the
exoskeleton using a Xacto hand drill, and five or six electrodes were inserted
into the muscles attaching to the first segment of an antenna. Bipolar hook
electrodes with 1 mm bared ends were inserted into each muscle using a 23
gauge needle and 74 µm diameter Teflon-coated silver wire (California Fine
Wire). The wires were glued to the exoskeleton with a hygenic
resin/cyanoacrylate mixture and coated with a sticky violin rosin/beeswax
mixture. The wires were gathered as a bundle on the dorsal surface of the
cephalothorax and run from the sea water to an amplifier (A-M Systems
differential AC amplifier) and then to an A/D board (National Instruments
AT-MIO-16E-10). The electromyographic (EMG) signals were notch- and band-pass
filtered (10 Hz and 1000 Hz) at the preamplifier and were sampled at 3750 Hz
by the A/D board. The EMG signals were recorded simultaneously with acoustic
and kinematic data in a LabVIEW virtual instrument. Experiments were conducted
in six lobsters with a range of 30-50 rasps per lobster. Periodic difficulties
with electrode recordings resulted in recordings not being made from some
muscles in all lobsters. Following each experiment, electrode placement was
verified through dissection.
A digital high-pass Blackman -60 dB filter (100 Hz) removed movement artifacts from the EMG signals. Since EMG signals typically range from 20 to 2000 Hz, removing lower-frequency movement artifacts does not appreciably distort the EMG signal. I processed the signals using a wide range of high-pass filter frequencies and filter types and verified that the choice of filters did not affect the timing of EMG traces at a scale relevant to the sound production events.
The timing, duration and intensity of muscle activity were measured. For
measurements of muscle onset relative to rasp onset and percentage duration of
muscle activity within the rasp, the EMG traces were rectified, and an
activity threshold was then applied in which a muscle was considered `on' when
the trace was at least two times greater than baseline measures (the baseline
was calculated during times of no movement). A muscle was considered `off' if
the muscle showed no activity for a minimum of 5 ms. However, when measuring
single units of motor activity during the rasp, the onset/offset of the muscle
was considered regardless of the time between bursts. Muscle activity was
compared in non-sound-producing movements and sound-producing movements to
assess the mechanism by which the sound is turned `on' and `off'. Muscle
intensity (µV) was calculated by rectifying the high-pass-filtered signal,
measuring the area (µV s) under the trace during the event and dividing it
by the duration (s) of the event. The non-parametric KruskalWallis test
(Sokal and Rohlf, 1981) was
used to determine whether the mean intensity of muscle activity differed more
than expected by chance across no movement (F), non-sound-producing and
posteriorly directed movements (P) and sound-producing movements (R) within
each lobster. Significance threshold was set at P<0.05.
Scanning electron microscopy
The files and plectrums of Panulirus argus and Palinurus
elephas were viewed using scanning electron microscopy. Using freshly
dead lobsters, I removed the file and plectrum by cutting the carapace with a
dental drill. The tissue was placed immediately in 2% glutaraldehyde in
phosphate-buffered saline and fixed for 1.5 h (see
Dykstra, 1992). The tissue was
rinsed in distilled water and dehydrated in an ethanol series. The specimens
were stored in 100% ethanol until critical-point drying and were then
sputter-coated (60:40 gold:paladium mixture, Anatech Hummer V) and observed at
up to 6000x magnification with a scanning electron microscope (Philips
501 SEM).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The proximal antennal muscles varied in position and size between sound
producers and non-sound producers and varied in size within sound producers. I
illustrated and compared the muscles of the sound-producing Panulirus
argus with those described previously for the non-sound-producing
palinurid lobster Jasus lalandii
(Paterson, 1968). I also
contrasted the muscles between two sound producers, Panulirus argus
and Palinurus elephas.
In Panulirus argus, the promotor muscle attached at the medial
edge of the antenna, just lateral to the plectrum, and extended to the dorsal
surface of the carapace (Fig.
4A). In sagittal view (Fig.
4B), the promotor was pinnate, with two lobes, one attaching to
the lateral portion of the carapace and a second attaching medially below the
rostral horns, just lateral to the midline of the carapace. The promotor
muscle of Palinurus elephas also attached at the medial edge of the
antenna, lateral to the plectrum; however, it had only one pinnate lobe,
attaching on the lateral region of the dorsal carapace. In Jasus
lalandii and Jasus verreauxi, the promotor muscle attached at
the midpoint between the two joint articulations and extended dorsally to the
carapace (Paterson, 1968).
In all four species, the lateral levator attached to the dorso-lateral edge
of the first antennal segment and extended ventrally to attach to the dorsal
and posterior surfaces of the infolding of the epistome (the ventral surface
of the antennular plate, extending between the antennae)
(Fig. 4A,B). The remotor muscle
described by Paterson (1968)
in Jasus lalandii was not apparent in Panulirus argus and,
in any case, was unlikely to be involved in antennal movement since it
attaches to two immovable structures [Paterson
(1968
) suggested that it might
function in bladder control or structure].
The depressor muscles were more difficult to discern as suggested by both
Paterson's (1968) and
Berkeley's (1928
) descriptions
in crustaceans. In P. argus, three of the four depressor muscles
described by Paterson in Jasus lalandii were readily identified
(Fig. 4B,C). Depressor A
attached to the medio-ventral edge of first antennal segment and extended to
several areas of epistomal infolding. It was pinnate and stretched both
posteriorly and ventrally to the epistome and to the lateral facet of a thin,
midline extension of the epistome. Depressor B, described by Paterson
(1968
) as extending from the
`postero-medial corner of the first segment' to the `dorso-medial joint of the
first segment', was not apparent in P. argus. The absence of
depressor B may be related to the fact that P. argus lacks the
dorso-medial joint articulations of the antennae that are present in J.
lalandii. Depressor C attached to the middle of the postero-ventral edge
of each antenna and spread to the floor of the epistome. Depressor D, oriented
slightly differently from the description by Paterson
(1968
), attached the
medio-ventral edge of the antenna to the medial floor of the epistome. The
depressor muscles of Palinurus elephas were generally similar to
those of Panulirus argus. However, the epistome's median process was
larger in P. elephas and reached almost completely across the
epistome, forming a clear separation between the left and right sections of
the epistome.
Kinematics and acoustics
Sound pulses and plectrum movements occurred simultaneously in all
experiments (N=97 rasps, six Panulirus argus). Sound was
produced only during posteriorly directed movements when the plectrum rubbed
against the anteriorly projecting shingles on the file. The plectrum movement
consisted of a series of steps in which there was a still period and a sliding
period; sound was emitted only during the sliding movement
(Fig. 5). Average signal
features are listed in Table
1.
|
|
Electromyography
Table 1 summarizes results
for the activity of different antennal muscles during sound production in
Panulirus argus. Only the promotor muscle (Pl and Pm,
Fig. 4) was active during 100%
of rasps; other muscles were active only during a proportion of the rasps
measured (LL, Da, Dc and Dd; Fig.
4). The depressor muscles (Da, Dc and Dd;
Fig. 4) were active most often
during anterior movements of the plectrum over the file. The lateral levator
(LL, Fig. 4) was active during
many different movements and often during sound production (98.4% of
rasps).
In a typical sound-producing event (Fig. 6), the depressor muscles were activated prior to sound production and pulled the plectrum to an anterior starting position on the file. The promotor muscle lobes were activated just prior to the onset of sound production. The lateral levator (LL, Fig. 4) often was active during sound production, but with substantially higher variability than the promotor muscle as represented by the standard errors of the mean (Table 1).
|
The promotor invariably generated a single, tonic contraction during the
rasp with no evidence of pulsed activity correlating with the ratcheted
movements of the plectrum. Single electrical units of activity occurred in
some recordings of the depressor muscles
(Fig. 7), with maximum
occurrence in depressor C muscle with 25% of sound pulses occurring
simultaneously with electrical units of activity (786 pulses, five lobsters).
These are probably products of stretch receptor activation when promotor and
levator muscles pull upon the opposing depressor muscles during sound
production (Ache and Macmillan,
1980). The occurrence of these electrical units of activity
primarily during movements, rather during than the stationary times between
pulses, is consistent with this interpretation. If these units of activity
functioned to stop plectrum movement during the `stick' part of the rasp, the
electrical activity should have been visible during the stationary periods
between pulses of sound. However, the activity occurred during sliding, which
suggests an electrical response due to pulling on the stretch receptors of
muscles opposing movement. While it is possible that these spikes are due to
movement artifacts, their clear depolarization and hyperpolarization suggest a
biological signal, not a mechanical depolarization due to electrode
movement.
|
Electromyographic results suggested that the promotor muscle activates the
sound-producing mechanism; however, the promotor was also active during
posterior movements of the plectrum that do not produce sound. To determine
whether the promotor muscle activated differently during sound-producing
movements from during non-sound-producing movements, I simultaneously recorded
activity in both the medial and lateral lobes of the muscle (Pm and Pl;
Fig. 4). Within each
individual, the mean intensities of the medial and lateral lobes were
significantly different across no movement (F), non-sound-producing
posteriorly directed movements (P) and sound-producing movements (R) of the
plectrum, with higher intensities during sound-producing movements (lateral
promotor lobe, 228 events, five lobsters; 2 for each
individual=24.52, 37.94, 29.65, 31.81, 37.00; d.f.=2; all
2
values indicated a significance of P<<0.001; medial promotor lobe,
191 events; four lobsters;
2=38.31, 29.93, 36.78, 31.71;
d.f.=2; P<<0.001) (Fig.
8A,B). The mean intensity differences between the two lobes
(medial lobe intensity minus lateral lobe intensity) were also significantly
different across the three categories of plectrum movement, with highest
values during sound-producing movements (191 events; four lobsters;
2=36.49, 30.06, 36.59, 7.15; d.f.=2; P<0.05)
(Fig. 8C). However, no clear
threshold of activity was apparent in lateral, medial or lateral minus medial
promotor intensities to distinguish between sound-producing and
non-sound-producing movements (Fig.
8).
|
Scanning electron microscopy
In Panulirus argus, file shingles were similar to those observed
in previous studies using air-dried preparations, having a mean width of 10
µm (Fig. 9C). However, the
plectrum ridges were distinctly different from previously published air-dried
preparations and showed smooth surfaces at magnifications up to 6000x
with no evidence of the accordion-like wrinkles as seen by Meyer-Rochow and
Penrose (1976) and Smale
(1974
). The plectrum ridges
showed no microscopic structures at the scale of the shingles; they have a
smooth uniform surface (Fig.
9D). The shingles of Palinurus elephas lacked the
transverse ridge found on each shingle in Panulirus argus, and the
plectrums were similar in most respects
(Fig. 9A,B).
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key structural changes are associated with the origin of the stick-and-slip sound-producing mechanism in palinurid lobsters. The loss of the medial joint articulation at the base of each antenna is a major mechanical transition to sound production (Fig. 3). In non-sound producers, the proximal antennal joint is limited to dorsal and ventral movements. In contrast, sound producers can move the proximal antennal joint antero-posteriorly as well as dorso-ventrally, allowing the plectrum, a medial process extending from the base of each antenna, to produce sound with a translational motion over the file.
In sound producers, the translational plectrum movements are often silent and are used in positioning the antenna. Preliminary data show that the plectrum can be planted at a particular location on the file, and the antenna then rotates as if there were two joint articulations with one degree of freedom, although with less range of motion. This suggests a sliding external joint that can be reoriented as necessary and, through repositioning, retains the stability of a typical arthropod hinge joint with two articulations. Such an external, repositionable articulation in a sliding joint that is also used in sound production has not, to my knowledge, been described previously in biological systems.
The plectrum flap, a fleshy extension from the posterior edge of the
plectrum, is found in a subset of sound-producing taxa with especially
elongate files (Panulirus, Justitia and Palinurus)
(Patek, 2001a) (Figs
1,
9). When this process was
removed in specimens of Panulirus argus, the plectrum could still be
manually manipulated over the file to produce sound
(Patek, 2001a
). The structure
may form a pivot point for the sliding joint articulation when the plectrum is
planted in place on the file and the antenna moves dorso-ventrally. The
elongate file of some taxa may necessitate more stability when the plectrum
can move a long distance along the file and thus permits more play in the
system.
Associated with these changes in joint mechanics are dramatic changes in
the position and size of the promotor muscle. The functional implications of
these differences can be inferred by the motor control of sound-producing and
non-sound-producing movements in Panulirus argus. The
electromyographic and kinematic results from Panulirus argus show
that the promotor functions to pull the plectrum over the file with and
without producing sound (Fig.
8). Rasps are produced when the plectrum rubs over the file in a
ratcheted movement (Fig. 5).
During each sliding period within that ratcheted movement, a pulse of sound is
produced. A higher intensity of promotor muscle activity generates
sound-producing movements compared with non-sound-producing posterior
movements and no movement of the plectrum, but there appears to be no specific
threshold of muscle intensity at which sound production occurs
(Fig. 8). There are several
possible explanations for this observation. First, there might be localized
neural control over parts of the promotor muscle or different muscle fiber
types responding to stimulation, such that one muscle region or fiber type is
activated only during sound production
(Ache and Macmillan, 1980). The
recordings of the medial and lateral promotor lobes do not show localized
areas of activation, but the type of electrode design used here may not have
been localized or sensitive enough to distinguish between activity of these
two lobes. More localized recordings of muscle activity or recordings from
motor nerves may be necessary to discern any actual threshold or localization
of control.
Another explanation is that a second muscle, in addition to the promotor muscle, is activated during sound production and not during simple posterior movements. The lateral levator muscle seems a likely candidate since it was active during 98 % of rasps. However, the levator attaches between the lateral facet of each antenna and an interior process of the epistome providing a line of action that should rotate the antenna ventro-laterally and counter to the posterior movement of the plectrum. Also, the high variability in the onset/offset times of the levator muscle relative to the rasp suggests irregular patterning of muscle activity, unlike the regularity observed in the promotor muscle (Table 1). The lateral levator is probably active during many kinds of antennal movement and, while it is often active during sound production, it does not serve to turn the production of sound on or off during posterior plectrum movements.
The promotor muscle in non-sound producers attaches to the midline of the
posterior edge of each antenna and pulls the antenna dorsally
(Paterson, 1968). In contrast,
the promotor muscle is located more medially in sound producers and pulls the
antenna both dorsally and posteriorly. Panulirus argus, with a long
file, has a large promotor muscle with two lobes that spread over most of the
available surface of the dorsal carapace
(Fig. 4). Palinurus
elephas, with a shorter file, has a relatively small promotor muscle with
only one lobe. The shift of the promotor attachment towards the plectrum in
sound producers is probably necessary to control the antero-posterior
translation of the plectrum over the file. This is in contrast to the simple
dorso-ventral movements of non-sound producers with two joint articulations,
which can be controlled by simple extensorflexor muscle attachments. In
this latter case, the promotor attachment in the center of the antennal base
makes sense: a central attachment point pulls equally on both sides of the
antenna.
The increase from one promotor lobe in Palinurus elephas to two lobes in Panulirus argus probably reflects the need for more motor control of the antennal base over the longer file in Panulirus argus. The longer file gives more translational area for the plectrum and, to manipulate the sliding joint at multiple plectrum positions, the promotor probably needs to generate both posteriorly and laterally directed lines of contractions. This issue could be resolved by measuring localized patterns of promotor activity across multiple palinurid species, particularly when the sliding joint produces non-sound-producing motions.
The hemisphere of soft-tissue plectrum ridges is similar across sound
producers, but the file's surface varies considerably. The plectrum appears to
be derived from arthrodial membrane, and the shingles are modifications of the
exoskeletal surface features of the antennular plate. Differences in shingle
structure suggest modification in frictional properties: in Panulirus
argus, each shingle has a prominent ridge, whereas the surfaces of
Palinurus elephas shingles are smooth
(Fig. 9A,C). Meyer-Rochow and
Penrose (1976) suggest that,
in Panulirus longipes, shingles may fuse during development to form
the additional ridge on each shingle. Shingle surface features could affect
the amount of sliding and static friction and may have important implications
for the `stick-and-slip' properties and motor control of the system. Greater
sliding or static friction may require more muscle force to drag the plectrum
over the file. More friction between the plectrum and file could result in
higher-amplitude rasps by generating larger displacements of the vibrating
surfaces.
The mechanical and muscular changes that occurred with the reconstruction
of the proximal antennal joint are significant in explaining both the
evolutionary origin of sound production and the subsequent variation of the
stick-and-slip mechanism. They also provide an example of how arthropods can
modify limiting joint architecture to allow a wider range of movements. These
observations raise the question of whether the loss of the joint articulation
occurred during the origin of the sound-producing apparatus or whether it
preceded sound production for the purposes of increasing the range of antennal
motion. Palinurid lobsters use their antennae for mechanical defense, and the
subset with a sound-producing apparatus generates rasps when interacting with
predators (Lindberg, 1955;
Meyer-Rochow and Penrose,
1976
; Mulligan and Fischer,
1977
; Smale,
1974
). The single origin of stick-and-slip sound production and
the function of antennae for both mechanical and acoustic defense leaves open
the issue of whether the initial mechanical changes in this system were for
antennal maneuverability or solely for sound production.
![]() |
Acknowledgments |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ache, B. W. and Macmillan, D. L. (1980). Neurobiology. In The Biology and Management of Lobsters: Physiology and behavior, vol. 1 (ed. J. S. Cobb and B. F. Phillips), pp. 165-213. New York: Academic Press.
Alexander, R. McN. (1983). Animal Mechanics. Boston, MA: Blackwell Scientific Publications.
Baisre, J. A. (1994). Phyllosoma larvae and the phylogeny of the Palinuroidea (Crustacea: Decapoda): a review. Aust. J. Mar. Freshwater Res. 45,925 -944.
Basmajian, J. V. and Luca, C. J. D. (1985). Muscles Alive, their Functions Revealed by Electromyography. Baltimore: Williams & Wilkins.
Benade, A. H. (1990). Fundamentals of Musical Acoustics. New York: Dover Publications.
Berkeley, A. A. (1928). The musculature of Pandalus danae Stimpson. Trans. R. Can. Inst. 16,181 -231.
Davie, P. J. F. (1990). A new genus and species of marine crayfish, Palibythus magnificus and new records of Palinurellus (Decapoda: Palinuridae) from the Pacific Ocean. Intervert. Taxon. 4,685 -695.
Deich, J. D., Houben, D., Allan, R. W. and Zeigler, H. P. (1985). `Online' monitoring of jaw movements in the pigeon. Physiol. Behav. 35,307 -311.[Medline]
Dumortier, B. (1963). Morphology of sound emission apparatus in Arthropoda. In Acoustic Behaviour of Animals (ed. R. G. Busnel), pp. 277-345. New York: Elsevier Publishing Company.
Dykstra, M. J. (1992). Biological Electron Microscopy: Theory, Techniques and Troubleshooting. New York: Plenum Press.
Ewing, A. W. (1989). Arthropod Bioacoustics: Neurobiology and Behaviour. Ithaca: Cornell University Press.
George, R. W. and Main, A. R. (1967). The evolution of spiny lobsters (Palinuridae): a study of evolution in the marine environment. Evolution 21,803 -820.
Lindberg, R. G. (1955). Growth, population dynamics and field behavior in the spiny lobster, Panulirus interruptus (Randall). Univ. Calif. Publ. Zool. 59,157 -248.
Loeb, G. E. and Gans, C. (1986). Electromyography For Experimentalists. Chicago: The University of Chicago Press.
Meyer-Rochow, V. B. and Penrose, J. D. (1976). Sound production by the western rock lobster Panulirus longipes (Milne Edwards). J. Exp. Mar. Biol. Ecol. 23,191 -209.
Mulligan, B. E. and Fischer, R. B. (1977). Sounds and behavior of the spiny lobster Panulirus argus.Crustaceana 32,185 -199.
Parker, T. J. (1883). On the structure of the head in Palinurus with special reference to the classification of the genus. Nature 29,189 -190.
Patek, S. (2001a). Signal producing morphology and the evolution of palinurid lobster communication. PhD dissertation in Biology, Duke University, Durham, North Carolina, USA.
Patek, S. N. (2001b). Spiny lobsters stick and slip to make sound. Nature 411,153 -154.[Medline]
Paterson, N. F. (1968). The anatomy of the cape rock lobster, Jasus lalandii (H. Milne Edwards). Ann. S. Afr. Mus. 51,1 -232.
Phillips, B. F., Cobb, J. S. and George, R. W. (1980). General biology. In The Biology and Management of Lobsters: Physiology and Behavior, vol.1 (ed. J. S. Cobb and B. F. Phillips), pp.1 -82. New York: Academic Press.
Smale, M. (1974). The warning squeak of the Natal rock lobster. S. Afr. Ass. Mar. Biol. Res. Bull. 11, 17-19.
Sokal, R. R. and Rohlf, F. J. (1981). Biometry. New York: Freeman.
Wainwright, S. A., Biggs, W. D., Currey, J. D. and Gosline, J. M. (1976). Mechanical Design in Organisms. Princeton, NJ: Princeton University Press.