An Electrophysiological Investigation of Power-Amplification in the Ballistic Mantis Shrimp Punch

Abstract

Mantis shrimp are aggressive, burrowing crustaceans that hunt using one the fastest movements in the natural world. These stomatopods can crack the calcified shells of prey or spear down unsuspecting fish with lighting speed. Their strike makes use of power-amplification mechanisms to move their limbs much faster than is possible by muscles alone. Other arthropods such as crickets and grasshoppers also use power-amplified kicks that allow these animals to rapidly jump away from predator threats. Here we present a template laboratory exercise for studying the electrophysiology of power-amplified limb movement in arthropods, with a specific focus on mantis shrimp strikes. The exercise is designed in such a way that it can be applied to other species that perform power-amplified limb movements (e.g., house crickets, Acheta domesticus) and species that do not (e.g., cockroaches, Blaberus discoidalis). Students learn to handle the animals, make and implant electromyogram (EMG) probes, and finally perform experiments. This integrative approach introduces the concept of power-amplified neuromuscular control; allows students to develop scientific methods, and conveys high-level insights into behavior, and convergent evolution, the process by which different species evolve similar traits. Our power-amplification laboratory exercise involves a non-terminal preparation which allows electrophysiological recordings across multiple days from arthropods using a low-cost EMG amplifier. Students learn the principles of electrophysiology by fabricating their own electrode system and performing implant surgeries. Students then present behaviorally-relevant stimuli that generate attack strikes in the animals during the electrophysiology experiments to get insight into the underlying mechanisms of power amplification. Analyses of the EMG data (spike train burst duration, firing rate, and spike amplitude) allow students to compare mantis shrimp with other power-amplifying species, as well as a non-power-amplifying one. The major learning goal of this exercise is to empower students by providing an experience to develop their own setup to examine a complex biological principle. By contrasting power-amplifiers with non-power-amplifiers, these analyses highlight the peculiarity of power amplification at multiple levels of analysis, from behavior to physiology. Our comparative design requires students to consider the behavioral function of the movement in different species alongside the neuromuscular underpinnings of each movement. This laboratory exercise allows students to develop methodology, problem-solving and inquisitive skills crucial for pursuing science.

Full text

The Journal of Undergraduate Neuroscience Education (JUNE), Spring 2019, 17(1):T11-T18
TECHNICAL PAPER
An Electrophysiological Investigation of Power-Amplification in the Ballistic
Mantis Shrimp Punch
Daniel J. Pollak1,2, Kathryn D. Feller3,4, Étienne Serbe1,5, Stanislav Mircic1, and Gregory J. Gage1
1Backyard Brains, Ann Arbor, MI 48104; 2University of Massachusetts Amherst, Amherst, MA 01003; 3Department of
Physiology, Development and Neuroscience; University of Cambridge, Cambridge, CB2 1TN, United Kingdom; 4Grass
Foundation Grass Laboratory, Marine Biological Laboratories, Woods Hole, MA, 02543; 5Max Planck Institute of
Neurobiology, Martinsried, D-82152, Germany.
Mantis shrimp are aggressive, burrowing crustaceans that
hunt using one the fastest movements in the natural world.
These stomatopods can crack the calcified shells of prey or
spear down unsuspecting fish with lighting speed. Their
strike makes use of power-amplification mechanisms to
move their limbs much faster than is possible by muscles
alone. Other arthropods such as crickets and grasshoppers
also use power-amplified kicks that allow these animals to
rapidly jump away from predator threats. Here we present
a template laboratory exercise for studying the
electrophysiology of power-amplified limb movement in
arthropods, with a specific focus on mantis shrimp strikes.
The exercise is designed in such a way that it can be applied
to other species that perform power-amplified limb
movements (e.g., house crickets, Acheta domesticus) and
species that do not (e.g., cockroaches, Blaberus
discoidalis). Students learn to handle the animals, make
and implant electromyogram (EMG) probes, and finally
perform experiments. This integrative approach introduces
the concept of power-amplified neuromuscular control;
allows students to develop scientific methods, and conveys
high-level insights into behavior, and convergent evolution,
the process by which different species evolve similar traits.
Our power-amplification laboratory exercise involves a
non-terminal preparation which allows electrophysiological
recordings across multiple days from arthropods using a
low-cost EMG amplifier. Students learn the principles of
electrophysiology by fabricating their own electrode system
and performing implant surgeries. Students then present
behaviorally-relevant stimuli that generate attack strikes in
the animals during the electrophysiology experiments to get
insight into the underlying mechanisms of power
amplification. Analyses of the EMG data (spike train burst
duration, firing rate, and spike amplitude) allow students to
compare mantis shrimp with other power-amplifying
species, as well as a non-power-amplifying one. The major
learning goal of this exercise is to empower students by
providing an experience to develop their own setup to
examine a complex biological principle. By contrasting
power-amplifiers with non-power-amplifiers, these analyses
highlight the peculiarity of power amplification at multiple
levels of analysis, from behavior to physiology. Our
comparative design requires students to consider the
behavioral function of the movement in different species
alongside the neuromuscular underpinnings of each
movement. This laboratory exercise allows students to
develop methodology, problem-solving and inquisitive skills
crucial for pursuing science.
Key words: electrophysiology, behavior, convergent
evolution, power amplification, motor control, muscle activity
The opportunity to discover insights into the natural world
through behavior can inspire undergraduates to select a
neuroscience laboratory course or choose neuroscience as
a major. Laboratory exercises that investigate the
extraordinary behaviors of animals can capture the
imagination of students and promote a deeper engagement
with the field (Pokala and Glater, 2018). For decades, the
amazing hunting abilities of mantis shrimp (stomatopods)
have attracted the attention of the public (e.g., British
Broadcasting Corporation Television, 1985). While
measuring only a few inches long, mantis shrimp can throw
the fastest punch of any animal with a peak impact force
equivalent to a tiger’s bite ( Patek et al., 2004; Kagaya and
Patek, 2016; Crane et al., 2018). Smaller mantis shrimp are
popular with aquarium hobbyists and can be easily obtained
for relatively low cost (~$30-50). Teachers can harness the
novelty of these animals to motivate students to learn
concepts in behavioral neuroscience and electrophysiology.
Undergraduate and high school labs must often make
use of available organisms and low-cost resources to bring
ideas from neurobiology into the classroom (Gage, 2019).
Invertebrates are an asset to the biology/neuroscience
teaching arsenal, as they are widely available, affordable,
and offer insights into how neural activity encodes
information that would otherwise be too invasive in human
or other vertebrate systems (Linder and Deyrup-Olsen,
1991). Cockroaches and crickets have an established
history as tractable and exemplary systems of invertebrate
nervous systems for classroom laboratory experiments
(Ramos et al., 2007; Dagda et al., 2013; Nguyen et al.,
2017). Unlike labs done with humans, these systems are
not limited to non-invasive investigations. One such lab
uses grasshoppers to demonstrate the role of the
descending contralateral movement detector circuit in
eliciting muscular responses to looming visual cues (Nguyen
et al., 2017). As more sophisticated techniques in teaching
laboratories gain momentum, it has become increasingly
important to make neurobiology as inexpensive and
accessible as possible to minimize educational inequities.
An excellent overview of cost-saving suggestions is

Pollak et al. Mantis Shrimp punch investigation T12
provided in Wyttenbach et al. (2018).
A range of arthropods, including crickets, mantis shrimp,
trap-jaw ants, and other jumping insects use a latch and
spring mechanical process to create power-amplification, a
phenomenon that allows some arthropods to perform tasks
with strength disproportionate to their size. These feats
include the trap jaw ant’s mandible strike, the locust’s jump,
and the mantis shrimp’s appendage strike (Patek et al.,
2004; Sutton and Burrows, 2011; Kayaga and Patek, 2016;
Burrows, 2016; Ilton et al., 2018). Though the underlying
mechanisms of power amplification across all species are
multifarious and not fully characterized, they all generally
use the same resilient properties of chitin (exoskeleton) as
a spring to load potential energy via muscle contraction.
This potential energy is translated into kinetic energy and
fast appendage motion with the removal of a latch that holds
the spring in place. Though the muscular and
electrophysiological control of power-amplified movements
are well characterized within certain species, such as
locusts (Burrows, 2016; Bayley et al., 2012), much remains
to be understood about the electrophysiology of these
movements across different taxa. Of particular interest is
the question of whether the convergent evolution of
muscular power-amplification yields similar or different
mechanisms of energy buildup and release. A carefully
designed laboratory exercise for bringing high school or
undergraduate students in contact with these questions
could place them on the scientific frontier alongside
professional scientists, sparking enthusiasm and curiosity
for the scientific process.
Here we present a method for recording EMGs in
arthropods with and without power-amplification. While the
focus of the surgical protocol described is specific to mantis
shrimp, our design provides a low-cost way to teach muscle
electrophysiology in other common organisms such as
cockroaches and crickets. Using common materials, we
made a chronic, modular implant, or “backpack”, capable of
recording EMG activity reflecting the buildup of energy
preceding cricket leg movement or the famed mantis shrimp
strike. It is simple enough to be fabricated by students
during the laboratory exercise and readily modified for new
uses. We provide students an opportunity to quantify
muscle bursts in power-amplified behaviors that can be
used to compare muscle activity across different types of
limb locomotion. We provide a sample analysis
investigating the muscle-driven movements of the
cockroach, and power-amplified kicks and strikes.
Hands-on fabrication of lab equipment has been shown
to improve students’ engagement with the laboratory
exercise (Crisp et al., 2016). Making the equipment allows
for an intuitive understanding of how the electronics and
physiology interact, and builds a foundation of skills for
investigating other questions about physiology and behavior
in the future. This laboratory exercise emphasizes student
participation, in particular animal handling, making and
implanting EMG probes, and hands-on behavioral
experimentation.
Arthropod electrophysiological investigations are usually
terminal (Kayaga and Patek, 2016), which can serve as an
obstacle to bring this technique into the classroom. This
device is chronically implanted, which means the recording
subject can return to its tank at the end of the exercise and
be recorded again after days or weeks.
Analyses used in this paper are available as Python
scripts online, and described in the Mantis Shrimp Punch
Instructor’s Guide, available in the supplemental materials.
Results we report here are intended to serve as a scaffold
for student- and instructor-driven exploration of EMG and
audio data.
MATERIALS AND METHODS
The following methods are reported as an easy protocol to
guide students when performing this lab. The entire lab
protocol has three distinct sections: fabrication, surgery, and
experimentation. These sections can be completed in a
single 3-4 hour session; however, if time is limited, each
section can be separated across consecutive labs. Due to
the cost and time devoted to managing each station, we
recommend creating groups of 2-4 students each to reduce
these factors. Each section of the materials and methods
can lead the instructor through the laboratory exercise, from
fabrication of each part, to surgery, to experimentation.
Instructors should have all groups complete each part of the
fabrication before moving on so students who finish first can
help groups that may be struggling; facilitating cooperativity
between groups simulates collaboration between labs for
troubleshooting methods. The Mantis Shrimp Punch
Instructor’s Guide explains the procedure more narratively
and includes a table with all the materials necessary for the
exercise. We have also provided a video of the
experimentation process is available at
https://youtu.be/NKdECK_3sJQ.
While we used a Muscle SpikerBox Pro (Backyard
Brains, Ann Arbor, MI) for these experiments, we note that
any EMG amplifier will be able to transduce mantis shrimp
muscle activity. Also, we use silver wire throughout this
procedure, but platinum, iridium, or steel wire would be
acceptable substitutions.
Fabrication
Previous labs have used hands-on fabrication steps as a
pedagogical device for facilitating a deeper understanding of
electrophysiology rigs for EMG (Crisp et al., 2016). In this
lab, students develop a plug-and-play recording system with
electrodes that stay positioned in muscles throughout the
probe’s lifetime. The fabrication process takes 30-60
minutes and allows the students to experience the
development of EMG methods by creating the probe from
scratch (including the backpack itself and the backpack
plug, which connects the backpack to the SpikerBox).
These electrodes could be made in advance and re-used in
later courses. The electrode system consists of an electrode
backpack which will remain on the animal, and a connection
adapter that will connect the recording system to the
backpack before each experiment.
Electrode Backpack
To make the chronic implant (the backpack), cut three ~1.5”
lengths of insulated silver wire. Deinsulate a few millimeters
of both ends of the silver wire by holding the end above a

The Journal of Undergraduate Neuroscience Education (JUNE), Spring 2019, 17(1):T11-T18 T13
flame, causing the insulation to retract from the end. Cut
three pins from a line of dip sockets and solder each female
lead to one end of each length of stripped silver wire. Using
liquid electrical tape, completely insulate the soldered leads,
and allow 10 minutes to dry. Make sure the entire insulated
length is covered. If the metal inside of the dip socket pins
are exposed on the side when cut from the line, they must
be covered in liquid electrical tape as well. Finally, the flame
deinsulation process can leave little balls, or prills, of silver
at the tips of the wires, which come together as a result of
the liquified metal’s surface tension. Cut off the prills,
leaving deinsulated silver wire.
Backpack Plug
The backpack plug and hydrophone will be connected to the
recording system through a Muscle SpikerBox electrode
cable. This electrode cable is a 1m cord that has a 3.5mm
stereo audio connector on one end which terminates to
three cable wires with alligator clips: two red (Left and Right
audio) and one black (ground). The backpack plug is made
of two parts: a three-pin row of dip sockets, and a muscle
electric cable composed of three separate cable wires.
Each pin and cable wire is connected by a deinsulated silver
wire. For a visual, see the Mantis Shrimp Punch Instructor’s
Guide.
Cut three pins from a line of dip sockets and sand the
female side with coarse sandpaper until metal is exposed on
each socket. Then, cut three more ~8” lengths of insulated
silver wires, deinsulating both ends of each silver wire with
flame as described above. Solder the metal part of each dip
socket to a deinsulated silver wire end using flux paste.
On the electrode cable, cut off the alligator clips and strip
off a few millimeters of insulation on each cable wire. Solder
the remaining ends of the deinsulated silver to each cable
wire, and ensure the black wire is not soldered to the middle
socket. Insulate all exposed and uninsulated metal with
liquid electrical tape.
Hydrophone
To capture precise timestamps of the strike, EMG events are
cross-indexed with corresponding audio ‘pop’ waveforms,
recorded with a custom hydrophone (Figure 1). Another
Muscle SpikerBox electrode cable is used to make the
hydrophone, alligator clips removed and wires stripped. The
microphone is a passive ~4W speaker as might be found in
a toy speaker.
To prepare the speaker, solder one lead of the 4W
microphone to one of the cable’s red wires. Then, solder the
other speaker lead to both the black and the remaining red
cable wire. Insulate all metal and soldered parts with liquid
electrical tape. Finally, place the microphone in a plastic
glove.
Stereotax
For the surgery, the animal is restrained on a stereotax
(Figure 2a) with at least one degree of freedom, allowing the
slab to be tipped downward into a shallow pool of water.
Make a slab measuring approximately 0.1” x 1.5” x 7”
out of a strong material like acrylic. Affix a 7” x 1.25” x 0.75”
piece of Styrofoam to the bottom of the slab with superglue
Figure 1. Experimental setup. a) demonstration of the proper
grasping technique to minimize injury from animal strike. The
backpack is placed on the carapace. The EMG wires are implanted
in the merus, targeting the extensor muscle. The antennal scale is
the feather-like appendage extending away from the carapace. b)
hydrophone and amplifier recording locations. c) example traces
of audio (gray) and EMG (Black). Large punch artifact in EMG
(large spike end of trace) was clipped for visibility. Scale bar =
250ms.
or marine epoxy. The Styrofoam acts as a pliant “corkboard”
material into which pipe cleaner or jumper wire restraints can
be securely plugged.
The body of the stereotax must support the weight of the
slab and be able to rotate. 3D printed parts for the stereotax
as well as CAD and 3D print files for printing
(https://backyardbrains.com/products/micromanipulator)
and instructions for building your own
(https://backyardbrains.com/products/files/SearcherBuildin
gInstructions.pdf). Alternatively, a magnetic gimballed stand
such as the Nootle Magnetic foot mini ball head camera
stand sold by Grifiti (http://www.grifiti.com/nootle/grifiti-
nootle-magnetic -mini-ball-head-camera-stand.html) can be
securely mounted to a slab. Using a magnetic stereotaxic
mount also requires a thin slab of iron or steel to function as
a wide and heavy base.
Surgery
Mantis shrimp are available for purchase from most
aquarium supply stores and should be maintained in a
cycled saltwater tank with aeration, filtration, and feeding
every other day with frozen and/or live bait (e.g., snails,
small crabs). We used Odontodactylus scyllarus,
Gonodactylus smithii, and Squilla empusa ranging in size
from 3” to 8” long. A 6” length of 2” diameter PVC piping
should be included in the tank for the mantis shrimp to reside
in.
Anesthesia and Restraint
The size of the mantis shrimp dictates the method of
restraint. Large animals are defined as being greater than
5” long, whereas small animals are defined as less than 5”
in length. To anesthetize a small animal, place it in a cup
containing ice; for large animals, use a bucket containing a

Pollak et al. Mantis Shrimp punch investigation T14
layer of ice at the bottom. Allow the animal to cool for
approximately two minutes, or until the animal stops moving.
Ice should be kept nearby throughout the surgery to chill the
animal if it becomes too active.
For large mantis shrimp, at least three points along the
body need to be restrained: below the carapace, in the
middle of the abdominal region, and at the bottom of the
abdominal segment, above the uropods. These points of
restraint can be held with either pipe cleaners or jumper
wires, inserted in the Styrofoam glued to the bottom of the
slab (Figure 2a).
For small mantis shrimp, assemble two pieces of silly
putty on both sides of the slab’s bottom. Place the shrimp
at approximately 45° to the slab on its rostrocaudal axis.
Gently pinch the silly putty against its sides. The silly putty
will hold the animal in this angled position. Again, at least
three points along the body are restrained with pipe
cleaners/jumper wires inserted in Styrofoam (Figure 2a).
Once restrained, dip the slab into a shallow pool of water so
that the mantis shrimp’s pleopods are mostly immersed in
aerated water.
Adhering the Backpack
Lightly score the carapace on the side of the raptorial
appendage to be implanted with coarse sandpaper. Apply
superglue on top of the scored carapace using a needle.
Place the backpack on the superglue, the female face of the
dip socket toward the anterior. Use either dental cement or
marine epoxy to further adhere the backpack to the
carapace. The sides, top, and back of the backpack should
be covered by adhesive, but the holes in front must not be
obstructed.
Dental cement is mixed in a glass or silicone container,
cleaned after each mixing with ethanol. Dental cement fully
hardens after five to ten minutes. If dental cement is
unavailable, marine epoxy (Loctite; Düsseldorf, Germany)
can be substituted. Marine epoxy begins to set after ten
minutes and dries over the course of a few hours, reaching
maximum hardness after 24 hours.
Implanting Wires
To implant the ground, place the higher gauge (smaller tip)
needle against the carapace, caudal to the backpack. To
open the cuticle, roll the needle between thumb and
forefinger, gently keeping it against that same spot on the
cuticle. Fingers will periodically travel down the syringe due
to the pressure exerted. To fix this, hold the syringe in place
with the non-dominant hand (i.e., touching the top) while the
rolling hand is brought back toward the top, and again
grasped between thumb and forefinger. If the needle cannot
pierce the carapace, the lower gauge (larger tip) needle may
work better. The higher gauge needle might poke through
into the flesh, in which case it may be gently removed. The
lower gauge needle will make a hole from which the animal
will bleed slightly.
Using two pairs of forceps in each hand, make a
millimeter-long bend “anchor” in the silver wire ground,
which should be the most medial lead (Figure 2b). This
anchor ensures the ground will stay in the tissue. Angle the
tip of the bent anchor into the hole in the cuticle, making sure
the entire length of bare silver is below the surface. Apply
superglue at the hole, followed by a thin layer of dental
cement or marine epoxy. Adhesive must not be allowed to
harden and should be quickly wiped away from the animal’s
joints.
The same technique for implanting the ground is applied
to implanting the two wires of the EMG electrode leads.
Score the merus with sandpaper, and open a single hole in
the cuticle above the extensor muscle using the technique
described above. To locate the extensor and carapace, see
the placement of the wires in Figure s 1a, 2b, and 2c. For
the remaining silver wires, make an anchor and angle the
wire into the hole, making sure to insert them in different
directions to avoid shorting. Apply adhesive to the hole
using the technique described above. The wires must not
be covered with dental cement or marine epoxy except at
the base, as they may become brittle. The antennal scale
(Figure 1) is wont to fold back against the merus. Thus, it
may be held out of the way with a needle skewering into the
silly putty or with a pipe cleaner/ jumper wire.
At the end of the surgery, return the animal to its tank.
Data Acquisition
Connect the backpack plug to channel 1 on the Muscle
SpikerBox Pro and the hydrophone to channel 2. EMG and
audio data are acquired with SpikeRecorder, an open-
source desktop app for Windows, MacOS, and Linux
(Backyard Brains, Ann Arbor, MI). To make the mantis
shrimp’s restraint, cut a strip of fabric, 2” long, 0.5” wide, with
a notch in the middle, and wrap it around the mantis shrimp
(Figure 2c,d). For a large animal, the restraint should be
bigger: 3” long, 1” wide. Position the hydrophone half-in,
half-out of the water.
The animal’s tail should have enough space to maneuver
and stretch. The anesthetized animal’s anterior abdomen
should be wrapped such that the legs are pointing away from
the merus and clamped using a helping hands tool. Lower
the restrained animal into an aerated salt water bath with its
pleopods in the water and the carapace out of the water
(Figure 2c,d).
Stimulus Delivery
The stimulus can be a pen, pencil, cotton swab, frozen or
live bait, or a rolled-up piece of paper towel. The stimulus
can be presented in a variety of ways, the simplest of which
is to place it in front of the mantis shrimp, close enough for
it to strike. This is particularly suitable for soft stimuli such
as frozen bait or a rolled up paper towel. Rigid stimuli such
as a pencil can be placed in front of the animal, or run rapidly
back and forth against the pleopods three or four times to
elicit a strike. This second method will induce stress, so
employ it as a last resort. During stimulus presentation, the
experimenter’s hand should not touch the salt water.
Additionally, the experimenter should use electrically inert
objects as stimuli. We found a three-inch long length of
rolled up paper towel to be the best stimulus for eliciting a
strike for Gonodactylus smithii. Space stimulus
presentations out by 5-10m, and make a single recording for
each stimulus presentation.
The trace of a backpack with a short will have a large

The Journal of Undergraduate Neuroscience Education (JUNE), Spring 2019, 17(1):T11-T18 T15
Figure 2: Mantis shrimp surgical and recording restraints. a) a large mantis shrimp (Squilla empusa) in restraint during surgery. The
stereotaxic apparatus tilts the animal into a pool of water, allowing it to breath, while minimizing movement. b) Close-up of placement of
the backpack with dental cement and placement of wires in merus. c) Small mantis shrimp (Gonodactylus smithii) in restraint during
experimentation with callouts highlighting anatomical features: i Telson. ii Pleopods. iii Merus. The extensor muscle is shown in below
the cuticle in green and the flexor muscle is shown in red. iv Antennal scale. v Carapace with backpack and ground. d) Photo of
experimental setup with callouts highlighting individual parts: i Cloth restraint. ii Dental cement encasing the backpack with wires going
below the water into the merus. iii The plug connected to the backpack with wires. iv Marked with black and white arrowheads, these
callouts demonstrate that the water covers most of the dorsal surface of the animal, leaving only the backpack dry. Note that only the
pleopods need to be submerged.
noise band, obscuring the true EMG signal. When the
backpack recording is shorted by water, dry it with paper
towels.
Data Analysis
Analysis was done in Python, using some of the analyses
from Kayaga and Patek (2016) for mantis shrimp data. For
all organisms, we examined spike shapes in addition to
spike times, as well as the number of spikes in the first and
second halves of each burst.
RESULTS
We present analyses of EMG data from one individual of
each species. These analyses should serve as a template
for more rigorous analyses, as opposed to true novel
findings. Code for running these analyses is available at
https://github.com/backyardbrains/MantisShrimpEMG. We
hope that students and instructors alike will adopt and
expand on the ideas and techniques shown here.
EMGs were acquired from two species with power-
amplified movements: the mantis shrimp (Gonodactylus
smithii, N=1, n=21 and the cricket Acheta domesticus, N=1,
n=10) and in one non-power-amplifying organism (the
cockroach Blaberus discoidalis, N=1, n=16). Figure 3
shows example EMG traces from each species. Implants
were successful for other mantis shrimp species (S. empusa

Pollak et al. Mantis Shrimp punch investigation T16
Figure 3. Examples of EMG bursts across three species. Burst
events in mantis shrimp were distinguished by the presence of a
sufficient audio signal at the end of the burst epoch (Figure 1c), in
addition to unpublished video evidence and notetaking. Bursts in
crickets and cockroaches did not correspond to behavioral criteria,
instead they were selected based on the presence of bursting. On
the left, example EMG trace recordings with spike times are
indicated by colored stars above each trace. The color changes
from blue to teal linearly across time, and this chronological color
gradient is reflected in the peak-aligned spike waveforms on the
right, with earlier spikes dark blue and later spikes teal.
and O. scyllarus); however, we generated insufficient data
for statistical analyses. Therefore, behavior and associated
muscle twitches were characterized in the G. smithii mantis
shrimp during the ballistic strike; however, behavior was not
quantified from the cricket and cockroach. Instead, we
included EMG activity from the hindleg extensor muscles
from crickets and cockroaches that took the form of short
bursts (<3.5s) for our analysis (Figure 3), as bursting is
indicative of power-amplified movement in crickets (Burrows
and Morris, 2003). Power-amplified EMG bursts from the
extensor muscle from the second raptorial appendage of G.
smithii were included based on their resemblance to
“canonical” power-amplified bursts in mantis shrimp
(Kayaga and Patek, 2016).
We found anecdotal distinctions between power-
amplifying and non-power-amplifying animals. First, EMG
spike shape changed within bursts (Figure 3). Early spikes
tended to have larger waveforms which diminished in size
over the course of the burst in power-amplifying subjects
(Figure 3a,b), but tended to stay the same size in non-
power-amplifying animals (Figure 3c).
The firing rate of spikes within the EMG traces changed
over the course of each EMG burst. Figure 4 compares
burst lengths (time between first spike and last spike) and
spike counts within the first and last half of the burst across
3 different animals (Figure 4a). Power amplifying animals
such as mantis shrimp and crickets had shorter average
burst durations than cockroaches. Tukey’s post-hoc test
showed that though there was a significant difference
between mantis shrimp and cockroaches, there was no
significant difference between cockroaches and crickets,
and there was a significant difference between mantis
shrimp and crickets (Kruskal-Wallis test: χ2(2)=16.85,
p=0.0002, (Figure 4b). The variance in data between mantis
shrimp and cockroaches differed across organisms as well
(Levene’s test for equal variance: F(2,43)=9.72, p=0.00033).
One salient feature of power-amplifying EMG bursts in
our data is the tendency to increase in firing rate from the
beginning to the end of the recording (Figure 4c). This
increasing trend held across 87% of all recordings of power-
amplifying animals, and exhibits a stark difference from
cockroaches, in which only 25% of the recordings have a net
increase in spikes between the beginning and end of the
burst. Individual changes in spiking rate between the first
and second half of a burst have not been examined in the
literature.
DISCUSSION
Here we have outlined a multi-phase teaching laboratory
that introduces several concepts in behavioral
neuroscience. The mantis shrimp is only one of many
power-amplification examples found in nature. Students
should be encouraged to consider similar mechanisms of
energy build up and release with exoskeletons (trap jaw
ants, grasshoppers, froghoppers, crickets) or without
(Achilles tendons, frog legs, salamander tongues; see:
Roberts and Azizi, 2010), and how this capability relates to
behavior.
This lab requires students to first make and then use the
equipment necessary for discovering these principles.
Fabricating probes enables the class to delve into the
electrical properties of muscle movement by implementing
the devices that measure them. Students learn practical
skills for animal surgery, from anesthetizing and restraining
the animals to actual operations. In the final phase of the
lab, students integrate insights from the preceding steps to
acquire recordings from muscles that drive an innate
behavior.
Since the laboratory exercise is broken into three parts,
there are several natural break points for instructors. These
could be used as stopping points for shorter labs (<1hr), to
ensure that longer lab sessions are on schedule, or to
deepen engagement with the material through discussion.
The first break point occurs after probe fabrication. After
developing a functional EMG electrode, there could be a
discussion on how basic electronics concepts (e.g.,
operational amplifiers) apply to biological processes.
Animal surgery represents the second breakpoint. If an
animal breaks its restraints, it can be easily re-anesthetized
and placed back on the stereotaxic apparatus, and the
surgery can continue as normal. The surgery phase affords
students the opportunity to make fixable mistakes, a rare

The Journal of Undergraduate Neuroscience Education (JUNE), Spring 2019, 17(1):T11-T18 T17
accommodation for surgeries. It also incorporates tactile
learning of mantis shrimp anatomy (i.e., gills, eyes,
antennae, reproductive organs). The experimentation and
analysis phase is the third break point. This phase explicitly
leaves room for students to troubleshoot and innovate, two
crucial skills in STEM. Because students design their own
stimuli for eliciting a strike, they can reason in real time about
why certain features of a stimulus are more salient to the
animal than others. Given the sophisticated visual system
of some mantis shrimp, behavioral contexts for striking, and
individual differences, students are faced with a highly
tractable puzzle in behavioral neuroscience. Therefore,
groups will have to collaborate, sharing methods and ideas
to successfully evoke striking behavior and collect data.
We present this EMG setup as a low-cost alternative to
expensive electrophysiology suites; however, a significant
limitation to our design is that it is not waterproof, which we
believe makes it an ideal candidate for innovation. We found
it more than feasible to collect data while monitoring for
shorts, but a waterproof design would allow for experiments
that take place underwater, perhaps even in burrows. Gruhn
and Rathmayer (2002) implemented a chronic and modular
crustacean electrophysiology system for crayfish and used
a waterproof socket. We hope researchers and students will
build upon our and Gruhn and Rathmayer’s designs to
develop a DIY technique that is waterproof and inexpensive.
In addition to a teaching laboratory, we present a novel
paradigm for doing electrophysiology in mantis shrimp and
other arthropods. Non-terminal chronic implants allow for
greater return on investment per individual and for more
robust statistical analysis of data, namely within-subject
experimental designs.
Finally, we present preliminary analyses of inter-species
comparisons of power-amplification. Low subject counts
render our findings anecdotal, but we found evidence of
commonalities in EMG traces between mantis shrimp and
crickets. In particular, the variances of the mantis shrimp
and cricket burst durations were closer to each other than to
that of the cockroach. The mantis shrimp and cricket tended
to increase firing rates from the beginning to the end of the
burst or co-contraction, while the cockroach showed no
pattern of firing rate modulation across the recording. The
difference between the mantis shrimp and the cricket burst
durations were, however, statistically significant, while the
burst durations of the cockroach were not significantly
different from those of either power-amplifying species.
Power amplification is usually investigated in terms of the
interplay between a set of extensor and flexor muscles in a
particular organism. Mantis shrimp have two extensors and
two flexors; we only recorded from the largest, the lateral
extensor. Our design could be easily expanded to record
from both the lateral extensor and the lateral flexor,
providing a near complete view of the power-amplification
system , though analysis of extensor co-contractions
between power-amplifying and non-power-amplifying
species suggests that examination of extensor spikes may
be sufficient to differentiate between these categories.
These similarities between the mantis shrimp and the cricket
suggest a possible commonality in muscular mechanisms
for power-amplification.
Figure 4. Burst analysis of recording sessions across three
species. a) Burst of extensor spikes in the Mantis Shrimp. Burst
duration is time difference between first and last spikes. To show
temporal change in spiking, we calculate the number of spikes in
the first vs second half of each burst, in this example 9 and 17
respectively. b) Characteristic EMG bursts from each organism.
Individual spike times are highlighted by stars. c) Waveform
summary of spikes for each organism.
The mantis shrimp species used in this study,
Gonodactylus smithii, is not necessarily representative of all
mantis shrimp species. To test for variation among
smashing type mantis shrimp species, we performed a
meta-analysis of EMG data reported from a second species,
N. bredini (Kayaga and Patek, 2016) against our data from
G. smithii. We found differences in muscular activity during
the strikes of these two species. Though the number of
extensor spikes in G. smithii’s burst (mean= 23.9 spikes)
falls within the range of means of its counterpart (mean
range: 13.4 – 68.8 spikes), its average extensor co-
contraction duration (mean=722 ms) was far above the
mean range (mean: 243-383 ms) of values in the literature.
The EMG spike analysis performed on N. bredini measured
the number of spikes in the first and last 100 ms of the strike
(Kagaya and Patek, 2016). For our study, we used an
alternative measure that counted the number of spikes in the
first and second halves of the strike. This method was
developed because we found that it better predicted
differences between power-amplifying and non-power-
amplifying species. This distinction between power-
amplifiers and non-power-amplifiers lies at the center of this
laboratory exercise and the data analysis approach was
tailored to highlight this distinction. Burst spike count and
burst duration provide a sense of the power these species
can muster to store energy for the spike. Assuming

Pollak et al. Mantis Shrimp punch investigation T18
approximately equal energy, a longer duration and an equal
number of spikes would indicate a slower rate of energy
storage (lower power) in the meral saddle. However, to
measure power, strike force must be measured as well
(Patek et al., 2004).
Variations in ecological and local environmental factors
often give rise to morphological and physiological
phenotypes among closely related species. The
stomatopod strike system and its variability among species
specialized for feeding on different categories of prey
provides an excellent system in which to study this biological
principle (DeVries et al 2016). Given the simplified and
inexpensive methods laid out here, we provide a framework
in which the muscular physiology of stomatopods with
different feeding morphologies and ecological requirements
can be evaluated. Our study is in no way an exhaustive
comparison between the strikes of different mantis shrimp
species, though we did observe evidence for a significant
difference in strike parameters between two mantis shrimp
species. Different strike parameters among species will
likely reflect specializations particular to each species’
ecology, representing an unexplored scientific milieu in
comparative ecology that is now highly accessible in the
classroom. These analyses are easy to perform at the
undergraduate level, requiring a low level of programming
skills. Additionally, code for performing these analyses is
available online.
Again, the analyses presented here are meant to be
examples for the kinds of insights students can make from
this preparation. They may also produce hypotheses
students can confirm or reject for themselves. Groups in the
classroom can pool their data and yield a high N. Indeed,
data from multiple years of teaching different class cohorts
can be saved and further compiled to produce a robust and
publishable data set.
REFERENCES
Bayley TG, Sutton GP, Burrows M (2012) A buckling region in
locust hindlegs contains resilin and absorbs energy when
jumping or kicking goes wrong. J Exp Biol 215:1151–1161
British Broadcasting Corporation Television (1985) The Fastest
Claw in the West. In: Wildlife on One. Bristol, United Kingdom:
BBC Television. Available at
http://bufvc.ac.uk/dvdfind/index.php/title/7509.
Burrows M (2016) Development and deposition of resilin in energy
stores for locust jumping. J Exp Biol 219:2449–2457.
Burrows M, Morris O (2003) Jumping and kicking in bush crickets.
J Exp Biol 206:1035–1049.
Crane RL, Cox SM, Kisare SA, Patek SN (2018) Smashing mantis
shrimp strategically impact shells. J Exp Biol 221:1–11.
Crisp KM, Lin H, Prosper I (2016) Breadboard amplifier: Building
and using simple electrophysiology equipment. J Undergrad
Neurosci Educ 14(2):A124–A131.
Dagda RK, Thalhauser RM, Dagda R, Marzullo TC, Gage GJ
(2013) Using crickets to introduce neurophysiology to early
undergraduate students. J Undergrad Neurosci Educ 12(1):A66-
A74.
DeVries, MS, Stock BC, Christy JH, Goldsmith GR, Dawson TE.
(2016) Specialized morphology corresponds to a generalist diet:
linking form and function in mantis shrimp crustaceans
Oecologia 182:429 – 442.
Feller KD, Cronin TW (2016) Spectral absorption of visual
pigments in stomatopod larval photoreceptors. J Comp Physiol
A 202:215–223.
Gage GJ (2019) The Case for Neuroscience Research in the
Classroom. Neuron 102(5):914-917.
Gruhn M, Rathmayer W (2002) An implantable electrode design for
both chronic in vivo nerve recording and axon stimulation in
freely behaving crayfish. J Neurosci Meth 118:33–40.
Ilton M, Saad Bhamla M, Ma X, Cox SM, Fitchett LL, Kim Y, Patek
SN (2018) The principles of cascading power limits in small, fast
biological and engineered systems. Science 360:1–11.
Kagaya K, Patek SN (2016) Feed-forward motor control of
ultrafast, ballistic movements. J Exp Biol 219:319–333.
Linder TM, Deyrup-Olsen I (1991) Use of invertebrate animals to
teach physiological principles. Am J Physiol 260:S22–S24.
Nguyen DMT, Roper M, Mircic S, Olberg RM, Gage GJ (2017)
Grasshopper DCMD: An Undergraduate Electrophysiology Lab
for Investigating Single-Unit Responses to Behaviorally-
Relevant Stimuli. J Undergrad Neurosci Educ 15(2):A162–A173.
Patek SN, Korff WL, Caldwell RL (2004) Biomechanics: Deadly
strike mechanism of a mantis shrimp. Nature 428:819–820.
Pokala N, Glater EE (2018) Using Optogenetics to Understand
Neuronal Mechanisms Underlying Behavior in C. elegans. J
Undergrad Neurosci Educ 16(2):152–158.
Ramos RL, Moiseff A, Brumberg J (2007) Utility and Versatility of
Extracellular Recordings from the Cockroach for
Neurophysiological Instruction and Demonstration. J Undergrad
Neurosci Educ 5(2):A28-A34.
Roberts TJ, Azizi E (2010) The series-elastic shock absorber:
tendons attenuate muscle power during eccentric actions. J Appl
Physiol 109:396–404.
Sutton GP, Burrows M (2011) Biomechanics of jumping in the flea.
J Exp Biol 214:836–847.
Wyttenbach RA, Johnson BR, Hoy RR (2018) Reducing the Cost
of Electrophysiology in the Teaching Laboratory. J Undergrad
Neurosci Educ 16:A277–A281.
Received November 16, 2018; revised April 3, 2019; accepted March 13,
2019.
Author GJG declares a commercial interest in the SpikerBox used here as
a co-owner in Backyard Brains. Authors ES and SM are employed by
Backyard Brains. DJP and GJG were supported by a National Institute of
Mental Health (NIMH) Small Business Innovative Research (SBIR) award
#R44MH093334. Author KDF is funded by European Commission Marie
Sklodowska-Curie Independent Postdoctoral Research Fellowship and the
Grass Foundation.
The authors are grateful to All Hands Active Makerspace in Ann Arbor, MI
for providing a venue to perform this research. We wish to thank Heather
Burke for the husbandry of mantis shrimp and Paloma Gonzalez-Bellido for
thoughtful discussions and for suggesting the collaboration between
Backyard Brains and KDF. All code and data for this manuscript can be
found online at https://github.com/BackyardBrains/MantisShrimpEMG.
Waveform analysis was completed in Python using Numpy, Matplotlib,
Scipy, Pandas, and Seaborn. Statistical analysis was completed in
MATLAB.
Address correspondence to: Dr. Gregory J Gage, Backyard Brains 308 ½
S. State Street, Ann Arbor, MI 48104. Email:
gagegreg@backyardbrains.com
Copyright © 2019 Faculty for Undergraduate Neuroscience
www.funjournal.org