See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/260446606

Portable conduction velocity experiments using earthworms for the college and
high school neuroscience teaching laboratory

Article in AJP Advances in Physiology Education · March 2014

DOI: 10.1152/advan.00088.2013 · Source: PubMed

CITATIONS READS
41 8,404

5 authors, including:

Kyle Michael Shannon Gregory J Gage

University of California, San Diego Backyard Brains
5 PUBLICATIONS 42 CITATIONS 49 PUBLICATIONS 2,110 CITATIONS

SEE PROFILE SEE PROFILE

W Jeffrey Wilson
Albion College
25 PUBLICATIONS 309 CITATIONS

SEE PROFILE

All content following this page was uploaded by Gregory J Gage on 21 January 2015.

The user has requested enhancement of the downloaded file.

 Adv Physiol Educ 38: 62–70, 2014;
Sourcebook Of Laboratory Activities In Physiology doi:10.1152/advan.00088.2013.

Portable conduction velocity experiments using earthworms for the college

and high school neuroscience teaching laboratory
Kyle M. Shannon,1 Gregory J. Gage,1 Aleksandra Jankovic,1 W. Jeffrey Wilson,2
and Timothy C. Marzullo1
1
Research and Development, Backyard Brains Incorporated, Ann Arbor, Michigan; and 2Psychological Science, Albion
College, Albion, Michigan
Submitted 6 August 2013; accepted in final form 4 December 2013

Shannon KM, Gage GJ, Jankovic A, Wilson WJ, Marzullo TC. Teaching inquiry-based neuroscience can be difficult for
Portable conduction velocity experiments using earthworms for the the biology/physiology teacher, as the combination of live
college and high school neuroscience teaching laboratory. Adv Physiol animals paired with concepts of electrical signaling and
Educ 38: 62–70, 2014; doi:10.1152/advan.00088.2013.—The earth-
amplification can be complicated to implement under the
worm is ideal for studying action potential conduction velocity in a
classroom setting, as its simple linear anatomy allows easy axon time and budgetary constraints of the high school/small
length measurements and the worm’s sparse coding allows single college classroom. Some universities have had notable suc-
action potentials to be easily identified. The earthworm has two giant cess in building neuroscience training laboratories, such as
fiber systems (lateral and medial) with different conduction velocities Cornell University with its CRAWDAD program (19, 25)
that can be easily measured by manipulating electrode placement and and the University of Minnesota BrainU high school teacher
the tactile stimulus. Here, we present a portable and robust experi- training program (9).
mental setup that allows students to perform conduction velocity

Downloaded from on January 21, 2015

Neurophysiology instructional laboratories are usually
measurements within a 30-min to 1-h laboratory session. Our im- expensive to maintain, requiring preparation by knowledge-
provement over this well-known preparation is the combination of
behaviorally relevant tactile stimuli (avoiding electrical stimulation) able faculty members with extensive experience in bioin-
with the invention of minimal, low-cost, and portable equipment. We strumentation and/or electrical engineering. We (6, 20) have
tested these experiments during workshops in both a high school and previously developed experiments that introduce students to
college classroom environment and found positive learning outcomes the concept of action potentials and rate coding using
when we compared pre- and posttests taken by the students. cockroaches and crickets; here, we developed experiments
conduction velocity; giant fiber; cable theory; electrophysiology; using earthworms on the more challenging concept of con-
earthworm; SpikerBox; LS1.D-information processing duction velocity.
We build on the work of other groups that have used the
earthworm as a teaching platform (7, 17); our focus was on
STUDENTS DOING THESE EXPERIMENTS will learn how to measure the portability and minimization of equipment. Typically,
the conduction velocity of neural action potentials using an action potentials are evoked in anesthetized worms with an
earthworm preparation and will get an introduction to a basic electrical stimulator and tabletop amplifiers; such “rigs” can
electrophysiology laboratory setup using simple amplifiers and often be intimidating for first-time users and may turn
laptop computers to amplify and record neural data. Students students away from electrophysiology. To combat this, we
will also learn how the earthworm’s lateral and medial giant designed a simple, handheld two-channel amplifier (the
nerve fibers transmit different sensory information from dif- two-channel SpikerBox) that, coupled with tactile stimula-
ferent parts of the worm (allowing the escape withdrawal reflex tion in lieu of electrical stimulation, makes for a compelling
in awake, behaving worms). Advanced students can also learn conduction velocity experiment in the earthworm Lumbricus
1) axonal cable theory and 2) statistical hypothesis testing. terrestris (colloquially called the “nightcrawler”).
Background The earthworm possesses one median giant fiber (MGF)
and two lateral giant fibers (LGFs), both of which run the
Learning neurophysiology, in particular the electrical prop- length of the worm and are located in the worm’s ventral
agation of action potentials, can be challenging for students. nerve cord (Fig. 1). Through careful experiments with nerve
Having a hands-on electrophysiology component to comple- cuts, it has been shown that the LGF transmits sensory
ment neuroscience lectures and laboratories makes the lessons information from the posterior end and the MGF transmits
both more compelling and increases learning (20). Moreover, sensory information from the anterior end of the worm (31),
evidence suggests that an active, inquiry-based learning peda- ultimately resulting in the “escape withdrawal reflex” of
gogy for science learning at the high school and university succinct muscular contractions (8, 26, 28 –30). The diameter
level (12, 21, 24) improves student comprehension and reten- of the giant fibers is ⬃0.05 mm for the LGF and 0.07 mm for
tion of scientific concepts, causing a sustained level of interest the MGF. The two LGFs are fused together via frequent
in pursuing science-, technology-, engineering-, and medicine- interconnects and thus are considered one functional giant
based careers. axon (5, 13).
In this report, we describe our experiments designed with the
Address for reprint requests and other correspondence: T. C. Marzullo,
intent to be easily followed by a student or teacher. We also
Research and Development, Backyard Brains Inc., 308 ½ S. State St., Suite 35, discuss the potential troubleshooting and pitfalls that may
Ann Arbor, MI 48104 (e-mail: [email protected]). occur while doing the experiments.
62 1043-4046/14 Copyright © 2014 The American Physiological Society
 Sourcebook Of Laboratory Activities In Physiology
CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS 63

Prerequisite Student Knowledge or Skills

Before doing this activity, students should have a basic
understanding of neuron and earthworm anatomy. They should
also be introduced to the concepts of action potentials and that
they progress down an axon with a conduction velocity that can
be measured. These can be discussed with the class during a
25- to 50-min lecture before the experiment session.
Students should know how to use an audio recording pro-
gram (such as Audacity, as described below) to record earth-
worm action potentials (“spikes”) and to subsequently measure
the time difference between spikes.
Time Required
An instructor experienced with this preparation can perform
it in one worm in ⬃10 –20 min during a lecture demonstration.
For a student laboratory, a student can perform the experiment
with supervision in ⬃45 min. The creation of a data set of five
to seven worms to allow for statistical hypothesis testing of the
different speeds will take a student ⬃3– 4 h.
METHODS

Downloaded from on January 21, 2015

Equipment and Supplies
Fig. 1. General earthworm anatomy (top) with a cross-section (bottom) zoom-
ing in on the ventral nerve cord with a view of the lateral giant fibers (LGFs) Anesthetics. All experiments in this report were performed on
and medial giant fiber (MGF). worms under a 10% by volume ethanol anesthetic solution. The 10%
ethanol solution was prepared by mixing 30 ml of tap water with 10
ml of 80 proof (40% ethanol) vodka. We placed the earthworms in the
Learning Objectives alcohol anesthetic for ⬃5 min, briefly rinsed them off in tap water,
and then began the experiments.
Using the National Research Council’s Next Generation Carbonated water can also be used as an anesthetic if ethanol is not
Science Standards (22) as a guideline, these lessons fall under available. Carbonated water (60%) can be prepared by mixing 30 ml
Life Sciences 1–From Molecules to Organisms: Structures and of sugar-free seltzer water (also called “club soda” or “sparkling
Processes, subgroup LS1.D–Information Processing. Our ex- water” at grocery stores) with 20 ml of tap water.
periments teach how an organism (the earthworm) detects and Anesthetic effectiveness can be determined by observing a lack of
processes information about a tactile stimulus (a head or tail worm movement and a cessation of the escape withdrawal reflex. The
tap) by converting the tactile stimulus into an electrical signal escape withdrawal reflex can be observed by tapping the tail and head
(action potentials of neurons) and then propagating these elec- with a plastic probe. An alert worm will exhibit a shortening muscle
trical information signals throughout the animal’s body, all of contraction in response to this stimulus, but an anesthetized worm will
which can be observed via the measurement of neural conduc- not have this reflex. The typical time in the alcohol or carbonated
water solution for sufficient anesthesia is ⬃5 min.
tion velocity. Students can use this knowledge to observe how After the anesthetic had taken effect, the worm was removed and
an unanesthetized worm responds to a light touch on its head placed in a container of tap water for several seconds to wash off the
or tail with the escape withdrawal reflex: a touch (the signal anesthetic. The effects of the anesthetic typically last 5–10 min. The
detection) causes neuronal spiking (information processing), worm was periodically kept moist during experiments with a wet
which sends a signal to the muscles to contract (the behavior). cotton swab brushed along the worm. It is important to not leave the
Students will also gain general process knowledge, such as worms in the anesthetic solution excessively, as the worms will not
how to observe, collect, and analyze physiology data. Enthu- produce action potentials and can also perish.
siastic students can also learn 1) statistical hypothesis testing Equipment and software. Figure 2 shows our experimental setup.
between two populations of data and 2) novel experiment Earthworms were placed on a balsa wood or styrofoam board with
design (see the DISCUSSION for some suggestions). measurement marks drawn on the board. This board was then
placed in a small Faraday cage with open ports to access the head
Activity Level and tail of the earthworm. The recording electrodes [electrode 1,
electrode 2, and reference (sometimes also called “ground”)]
This activity is suitable for high school students and above inserted into the worm connected to our two-channel SpikerBox,
who have a basic conceptual understanding of what an axon which is a custom 880⫻ gain two-stage amplifier using the AD623
and an action potential are. Since electrophysiology is novel instrumentation amplifier chip (Analog Devices, Norwood, MA)
for many students and the methods are sensitive to electromag- on the first stage and the TLC2272 operational amplifier chip
(Texas Instruments, Dallas, TX) on the second stage (see Ref. 3 for
netic noise that can easily confuse and frustrate students, close the schematics). The first stage has a 2-G⍀ input impedance and a
observation and help by the teacher during experiments (espe- gain of 4⫻, and the second stage has 220⫻ gain and a band-pass
cially during data collection for the first worm) is essential (see filter from 300 to 1,300 Hz. The audio component of the amplifier
Troubleshooting below). A teacher may find doing a quick uses a standard configuration of the LM386 audio chip (Texas
demonstration of one conduction velocity reading at the be- Instruments). The output of the two-channel SpikerBox is a stan-
ginning of class to be helpful. dard 3.5-mm (1/8 in) stereo audio jack that can then go into the line

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

Sourcebook Of Laboratory Activities In Physiology
64 CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS

Human or Animal Subjects

Human participants. We held workshops in both a neuroscience
class at a selective liberal arts college and a biology class at a
suburban high school in Michigan. Our 2-h workshops were a mix of
lectures and demonstrations that covered neural communication, con-
duction velocity, and cable theory. Students observed and assisted in
live demos of the experiments listed in this report and then actively
engaged in discussion about the theory and experiments.
Before and immediately after the workshops, on the same day,
students were given multiple-choice tests to examine their knowledge
of conduction velocity concepts. The pre- and posttests were exactly
the same. The college survey had 13 questions that covered conduc-
tion velocity physiology and cable theory along with some mathemat-
ical questions dealing with time constant and length constant equa-
tions. The high school survey had eight questions and did not have the
mathematical questions dealing with time constant and length con-
stant equations. Of the 15 respondents at the college, 57% were
women and 43% were men; 85% identified themselves as Caucasian,
7.5% as Asian, and 7.5% unreported; and the average age was 21.3 yr
Fig. 2. Experimental setup and equipment. We had a laptop with a two-
channel USB sound card (iMic, Griffin Technology) and used Audacity, an (SD: 0.63, range: 20 –22 yr). Of 25 respondents at the high school,
open-source recording program. The electrodes were map pins soldered to 52% were women and 48% were men; all but 1 respondent identified
speaker wire. We used a Faraday cage to reduce noise and had an earthworm themselves as Caucasian (1 respondent was of mixed race); and the
on a styrofoam recording platform marked with 2.5-cm hashes. We grounded average age was 15.5 yr (SD: 0.5, range: 15–16 yr).
the SpikerBox to the Faraday cage using an alligator clip. The plastic stick or All survey data collection was performed with informed consent

Downloaded from on January 21, 2015

glass rod was our means of providing tactile stimuli to the earthworm. Some from teachers and students. Our testing policies were approved by the
laptops have stereo input and do not need a two-channel USB converter. Institutional Review Board of Albion College.
Animal models. The earthworm L. terrestris, commonly known as
the Canadian nightcrawler, was our model of choice given its large
in of a laptop computer (preferably running on battery power to size and ease of collection from local pet/sporting goods stores. Boxes
reduce any potential power line electrical noise). If the computer of 12–18 worms typically cost US$3– 4, and these worms can even be
does not have a stereo line-in jack, a USB sound card can be used, harvested in the field after a heavy rainfall.
such as the iMic (US$40, Griffin Technologies). USB sound cards Worms were stored in a refrigerator in a styrofoam container filled
can sometimes generate a 1-kHz ringing noise artifact, but this can with soil, and we poured a small amount of water in the box every
often be grounded out by carefully positioning the earthworm in couple of weeks to keep the soil moist. The worms are very easy to
the Faraday cage. care for; we have observed that they survive for ⬃1–3 mo in the
To record the signals and measure the time difference between the refrigerator.
two electrodes, we used the open-source audio processing program The worm’s clitellum can be used as an anatomic landmark since
Audacity, which has versions available for all platforms (Mac, Win- it is always closer to the anterior end (head) than the posterior end
dows, Linux, etc). In the preferences of Audacity, the input signal was (anus; Fig. 1). Because the clitellum becomes more pronounced as the
set to “line in,” the hardware playthrough checkbox was enabled (so earthworm reaches sexual maturity, the experiments are best done on
as to simultaneously hear the recordings as the signals are coming in), large, sexually mature adults. If the worm has not yet reached sexual
the number of channels was set to two (“stereo”), and the “record” red maturity and the clitellum is hard to identify, close examination of the
button was clicked to begin a recording session. posterior end (anus) will show a flattening, whereas the anterior end
(head) will be cylindrical and cone shaped.
The earthworm preparation can be sensitive to noise, and a Faraday
Earthworms, as invertebrates, do not require institutional oversight
cage is essential to reduce spurious electromagnetic interference. A
or approval for experimentation in a classroom or research laboratory.
Faraday cage (such as the one shown in Fig. 1) can be easily However, we opened the experiments in the classrooms by stating the
assembled at low cost (less than US$10) using components from a earthworms can recover from the experiment, discussing the use of
local hardware store. Students can cut out an 8 ⫻ 16-in. rectangle of anesthesia for humane treatment, mentioning the broader role of
screen door metal mesh and five 8-in. lengths of wood strips (for animals in physiology research, and pointing out an alternative non-
example, 3/8-in. tall ⫻ 3/16-in. wide basswood). The wood strips can invasive experiment (see Experiment 2 below).
then be stapled to the mesh (with the first strip at one end of the of
mesh, the second strip 5.5 in. away from the first strip, the third strip Instructions
2.5 in. away from the second strip, the fourth strip 5.5 in. from the
Experiment 1: comparing the speed of two fibers. After the worm
third strip, and the fifth strip at the other end of the mesh). The mesh
was anesthetized, we placed the worm dorsal side up [the dorsal side
can then be folded into a box, and an alligator clip can be connected of worm is darker (brown) than the ventral side (gray)] on the wood/
from any point on the metal mesh to the ground metal pad on the styrofoam platform and inserted the three electrodes into the worm
SpikerBox circuit board (or ground on any other system). Further slightly off its centerline so as to try to avoid piercing the intestine or
details can found online (4). Enterprising students can also find that ventral nerve cord. The metal pin electrodes traversed through the
metal containers, such as antique lunchboxes, work as well. worm and into the wood/styrofoam base. Figure 3 shows the electrode
Although we developed our own equipment and used an open- positions. We placed the worm into the Faraday cage, and we
source software package for the experiments described here, equip- grounded the Faraday cage to the SpikerBox using an alligator clip
ment from other suppliers, such as from iWorx (two NA-100 ampli- cable. Using a ruler or the hash marks drawn on the styrofoam board,
fiers with the ETH-402 interface), AD Instruments (PowerLab 15T), we recorded the distance between electrode 1 and electrode 2 (where
Biopac (MP36R System), and Grass (two LP511 amplifiers), will all dLGF is the LGF electrode distance measurement and dMGF is the
work as well. MGF electrode distance measurement).

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

 Sourcebook Of Laboratory Activities In Physiology
CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS 65

Downloaded from on January 21, 2015

Fig. 3. LGF and MGF action potentials and the corresponding electrode placement. Channel 1 and channel 2 are shown at a 10-ms timescale (left) and a
zoomed-in 4-ms timescale (right). The LGF response was recorded by applying a mechanical stimulus to the posterior end, whereas the MGF response was
recorded by applying a mechanical stimulus to the anterior end (see “tap here” arrow). Note the different electrode positions in both cases. To observe a time
lag difference between channels 1 and 2, the waveforms must be zoomed in. Electrode 1 corresponds to channel 1, and electrode 2 corresponds to channel 2.
dLGF, LGF electrode distance measurement; dMGF, MGF electrode distance measurement; Ref, reference electrode.

To first record from the LGF, we turned on the SpikerBox and For a good data set, evoked multiple spikes from the MGF and LGF
started our Audacity recording. We then lightly tapped the very can be collected and measured both in a single worm and across
posterior end of the earthworm with a glass or plastic probe. In the multiple worms (Fig. 4). This experiment takes ⬃30 –120 min to set
seven worms, this mechanical stimulus typically evoked one to three up and complete depending on the number of worms used. Worms can
spikes per tap (Fig. 3). We then turned off the SpikerBox but left recover from the experiment and be used again at a later date.
Audacity running to create an easily recognizable flat line area that Experiment 2: invasive versus noninvasive recording. Although the
can be used as a separation marker between our LGF and MGF spikes, use of electrodes inserted in the earthworm keeps the earthworm
making data analysis easier. While the SpikerBox was off, we re- firmly in place, some students may experience emotional discomfort
moved the worm from the Faraday cage and repositioned the elec- inserting needles into the creature. Fortunately, it is possible to record
trodes in the anterior end of the worm to record from the MGF (Fig. spikes ex corporeal using a similar setup as that in experiment 1.
3). Again, we measured the distance between electrode 1 and elec- Instead of placing the electrodes into the worm, the electrodes were
trode 2 using a ruler (dMGF). We placed the worm back into the taped down to the wood or styrofoam recording platform. The metric
Faraday cage and turned the SpikerBox back on. We then caused the placement and distance between the electrodes was maintained as
spiking response in the MGF by tapping the anterior end of the worm described above. The worm was placed on the recording platform in
(head) with the plastic or glass probe. After several spikes were the same way as in experiment 1, dorsal side up, but resting on top of
recorded, we removed the electrodes from the worm and placed the the electrodes. Care must be taken to ensure that the worm’s skin is
worm back in its moist, soil-laden case in the refrigerator. Cleaning resting well on the electrodes.
consisted of dumping out the anesthetic solution and wiping the To test whether the amplitude of action potentials recorded inside
electrodes and styrofoam with a wet cloth. versus outside the worm’s body was different, we built a custom
The Audacity file can then be saved in Audacity’s format or four-channel SpikerBox. Having four channels required the use of a
exported as a .wav file to be analyzed in a program of choice (such as US$199 four-channel USB Audio Mixer (Maya44, Leonberg, Ger-
Matlab). To analyze the data after a recording in Audacity, we many). Audacity was again used to record the spikes. The earthworm
zoomed in on the evoked spikes until we were able to see time was placed on two electrodes, and the two remaining electrodes were
differences between the two channels (Fig. 3). Using the time markers inserted into the worm as close as possible to the surface electrodes
on the Audacity window, we measured the time difference between without physically touching and shorting (Fig. 5). The ground elec-
the first large negative deflections of the spikes (where tLGF is the LGF trode was inserted a distance away and can be either inserted into the
time difference and tMGF is the MGF time difference). We then worm or rest below the worm.
calculated the conduction velocity by dividing the ruler measurement
by the time measurement, as follows: Troubleshooting
dLGF dMGF Sometimes electrode placement between the LGF and MGF can
CVLGF ⫽ CVMGF ⫽ confuse students. The easiest way to resolve this is to take out the
tLGF tMGF
electrodes, turn the worm 180° around, and then simply place the
where CVLGF and CVMGF are the conduction velocity measurements electrodes back in the worm. The MGF was always faster than
for the LGF and MGF, respectively. the LGF in our experiments, so students can in fact use their conduc-

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

Sourcebook Of Laboratory Activities In Physiology
66 CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS

Finally, the biggest issue that students may face is electromagnetic

noise interfering with the recordings and drowning out the earthworm
spikes. The use of a Faraday cage is critical in reducing such noise,
but occasionally noise will still dominate, especially on the upper
floors of tall buildings near radio or cellular transmission towers. In
these situations, laptops should run on battery power alone and the
SpikerBox and earthworm should both be placed in the Faraday cage.
All wires should be folded together as much as possible to avoid
creating antenna loops. Such intense noise is rare and is typically only
observed near transmission equipment; this can be solved by changing
the room where the experiment is being held.
Moreover, if the electrodes are not firmly in contact with the worm,
the recordings may reveal popping and chirping when the worm is
tapped due to the disruption of the electrode-worm interface. It is
critical to ensure that 1) the electrodes are in contact with the worm
(with the worm either laying on top of the electrodes or the electrodes
inserted into the worm), 2) the audio cable from the two-channel
SpikerBox is fully plugged into the laptop, and 3) plastic or glass
probes instead of metal objects are used to tap the worm. The use of
a lightly anesthetized worm that is robustly spiking to head or tail taps
will help debug noise issues. Spurious noise can also occur if a part of
the worm is in contact with the metal mesh of the Faraday cage. No
part of the worm should touch the metal mesh. Also, if excessive
water is on the platform (styrofoam or wood) that the worm is resting

Downloaded from on January 21, 2015

Fig. 4. Intra- and interworm measurements of LGF and MGF conduction
velocities. The two right box plots show the variance of the data for all seven
worms. A paired t-test with an ␣-level of 0.05 gave P ⬍ 0.0001 between the
LGF and MFG conduction velocities.

tion velocity measurements to confirm the anatomy of the worm (for

example, if the clitellum is hard to identify in a young worm).
One limitation of our experiments is when a nonspiking worm
preparation occurs. If the worm is in an electrically noisy environment
or is overanesthetized, spikes often cannot be either evoked or
discriminated. As the amplitude of the spikes from the earthworm can
be small due to their myelination (14), they can be easily buried in
electrical noise. In addition, the very lack of many spikes, which
makes the experiment compelling for conduction velocity measure-
ments, also simultaneously makes the experiment hard to debug (is it
my equipment, or is it the worm, that is not working?). We have found
that starting the experiment using only lightly anesthetized worms can
help students get used to the preparation and what the spikes “look
and sound like” before experimenting on more deeply anesthetized
preparations, where spikes may sometimes not occur. A lightly
anesthetized worm will be moving slightly, and muscle electrical
activity will be present in the recordings, occasionally masking the
spikes, but a student can much more easily hear and identify the spikes
before moving on to a more anesthetized worm to achieve a more
accurate measurement of conduction velocity.
Another limitation of our experiments is that the two-channel
measurement requires the use of a computer/laptop with stereo input
instead of more portable mobile devices, such as tablets or smart-
phones. The use of the computer [given our previous work using only Fig. 5. Invasive versus noninvasive electrode placement. Recording action
mobile devices (20)] makes the conduction velocity experiments potentials both invasively or noninvasively in the same worm showed little
slightly more cumbersome in cases where tablespace may be limited. difference in amplitude and wave shape. The reference electrode (black) was
We are researching wireless solutions for our amplifiers so that the inserted into the animal, but it can also be placed under the animal and
two channels can be recorded and displayed on a mobile device. functions as a reference similarly.

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

 Sourcebook Of Laboratory Activities In Physiology
CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS 67

on, and if such water leaks down the side of the platform and makes MGF and LGF recordings that students and instructors can use
contact with the metal mesh of the Faraday cage, spurious noise will as a reference when attempting to replicate these experiments.1
also result, which can be difficult to identify. We sometimes place the
wooden or styrofoam platform on top of a piece of acrylic plastic Experiment 2: Invasive Versus Noninvasive Recording
inserted into the Faraday cage to avoid this.
When tapping the worm, care should be taken to not be too Although the earthworm can often recover from the insertion
vigorous with the tapping, such that the head or tail “flops up and of the electrodes, recording ex corporeal can ease any students’
down” with each tap. Such movement of the worm can cause spurious potential emotional discomfort. Incidentally, the earthworm
noise transients that look like action potentials on the recording but was actually the first preparation where ex corporeal recording
are not (spurious transients will happen at the exact same time on both of action potentials was demonstrated (29). Figure 5 shows
channels and are thus not biological). recordings and electrode placement from simultaneous inva-
sive and noninvasive recordings. Notably, the amplitude and
Use of Other Worms waveform of the invasive versus noninvasive neural recordings
California red worms (Eisenia fetida) are commonly used in was similar. Recordings in which the earthworm was lying on
composting and are easier to raise in self-sustaining colonies than L. the ground electrode or the ground electrode was inserted into
terrestris; we have had self-sustaining colonies of such red worms for the earthworm were not different (data not shown). The only
3 yr. We attempted to replicate the conduction velocity experiments in disadvantage of this ex corporeal recording is that the worm
red worms, but results were mixed. If the worms are anesthetized such has to be under deeper anesthesia (placed in the anesthetic
that they are no longer moving, we only occasionally could elicit solution 1–3 min longer than our recommended 4 min), as any
spikes in the MGN and never in the LGN. To record spikes success- slight movement of the worm over the electrodes causes the
fully from the LGN, we had to anesthetize the worm only very lightly,
recordings to be unstable. With deeper anesthesia, the risk is
such that it was still moving substantially, making measurement
difficult. We recommend using the smaller E. fetida worms only when higher that the worm will not generate spikes when mechani-
L. terrestris is not available, as getting stable recordings is difficult cally stimulated. The instructor should do a quick demonstra-

Downloaded from on January 21, 2015

and can really only be done by a determined student or experienced tion to all the students before the individual laboratory ses-
instructor. sions, so that the students can learn how to mechanically
Our experiments in keeping L. terrestris in large plastic bins to stimulate the worm and see what the spikes look like on the
make our own self-sustaining colonies were usually not successful, computer screen.
with the worms dying within 1–2 mo. However, one of us (W. J.
Wilson) has two colonies that have survived well for 6 mo in Misconceptions
temperature-controlled environments; time will tell whether these
worms reproduce and the colonies become self-sustaining. For all The action potentials recorded in this study are extracellular,
experiments in this work, we simply bought L. terrestris from bait meaning that the waveforms will not look like the action
supply stores. potentials depicted in textbooks. The shape of an extracellular
action potential depends on the conductivity and electric fields
Safety Considerations surrounding the nerve and are on the order of ⬃1 mV, much
less than the 100-mV range of an intracellularly recorded
The needles we insert into the worm are commonly available map
action potential. Intracellular recordings are technically more
pins; students should take normal precautions when handling the pins
(always picking up by plastic ball end and not the sharp end). The difficult; the first successful intracellular recordings were ac-
anesthetic we prepare is 10% ethanol, which can be prepared by the tually done by inserting a glass microelectrode into the 1-mm-
teacher before the experiment to avoid any connotations with alcohol wide giant axons of a squid (16).
consumption. In schools that do not allow experiments that use It should be clearly stated to the students that they are not
ethanol as a reagent, carbonated water can be used as an alternative inserting electrodes in nerves and that they are viewing an
anesthetic. action potential recorded outside the nerve. For a comparison
of the difference between intracellular versus extracellular
RESULTS recording traces, we recommend Fig. 1 in Ref. 15 as well Ref.
2 for a primer on voltage measurements in neural tissue.
Experiment 1: Comparing the Speed of Two Fibers Due to the earthworm spikes traveling past the recording
We recorded the conduction velocity from seven worms for electrodes and then passing the ground electrode, the spikes
the data in this report but have since replicated this result many recorded from both channels have different initiation times but
times during classroom lectures. Figure 3 shows sample re- identical termination times [see the electrophysiological traces
cordings and traces from an individual earthworm; Fig. 4 shown in Fig. 3, where there is a time delay in spike initiation
shows the compiled data from all seven worms. In each worm, (first negative deflection) but not in spike termination (final
we took five measurements from different spikes in both the positive deflection)]. This should be pointed out to the students
LGF and MGF. Within each worm, the difference between the that they can only use the time delay in spike initiation to
MGF and LGF was immediately apparent and statistically accurately measure conduction velocity.
significant (P ⬍ 0.05 by t-test). Across all worms, the average Even though the axons in the earthworm are relatively large
speed of the LGF was 7.6 ⫾ 1.2 m/s (mean ⫾ SD) and the (0.05 mm for the LGF and 0.07 mm for the MGF) and students
MGF was 22.8 ⫾ 4.5 m/s (mean ⫾ SD). The differences are are commonly taught that invertebrates have large diameter
large enough that even with a low number of samples this axons to increase their conduction velocity during escape
exercise can also serve as an introduction to basic t-tests and
statistical hypothesis testing for undergraduate students (1). In 1
Supplemental Material for this article is available at the Advances in
the Supplemental Material, we have included a .wav file of the Physiology Education website.

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

Sourcebook Of Laboratory Activities In Physiology
68 CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS

behavior, students may be surprised to learn that the conduc- for them to study will probably increase their understanding of
tion velocities are actually quite slow even in the large axons. these concepts.
We measured the conduction velocity at 22.8 m/s for the
0.07-mm MGN, or equivalent to ⬃50 miles/h. Even the large Inquiry Applications
1-mm-diameter unmyelinated giant axon of the squid only has Comparison with other published data. Our results are close
a similar speed of ⬃20 m/s (27). To give students some to agreement with Kladt et al. (17), whose group measured
perspective, the myelinated ␣-motor neurons of mammals can conduction velocities on the order of 16.9 m/s for the MGF and
reach 80 m/s, or 180 miles/h, and are only ⬃20 ␮m in diameter 6.9 m/s for the LGF using electrical stimulation in chlorobu-
(10, 11). tanol-anesthetized worms. Our experiments are not in agree-
However, it should be pointed out that earthworm nerves are ment with Drewes et al. (8), who measured MGF speed at 32.2
also myelinated (14); teachers can explain this is why the m/s and LGF speed at 12.6 m/s using tactile stimulation.
conduction velocity of the MGN in the worm is roughly the Drewes et al. importantly, recorded conduction velocity in
same speed as the unmyelinated squid axon, although the earthworm unanesthetized worms moving around a circular track covered
nerves are ⬃10 times smaller (see the APPENDIX for a mathe- with electrodes. We are beginning to prototype equipment,
matical description of cable theory). This raises the interesting similar to Drewes et al.’s experimental apparatus, that will
question as to why earthworm conduction velocity is not as fast allow us to compare conduction velocities in awake behaving
(or faster) than much smaller (20 ␮m) mammalian myelinated worms versus anesthetized worms to determine if the anesthe-
␣-motor neurons, which can conduct at 80 m/s. We do not have sia is slowing conduction velocities.
a satisfactory answer for this, but it can lead to an interesting Drewes et al. noted in their awake behaving recordings that
discussion with students. when only one spike was elicited by tactile stimulation, a
Teachers can also point out that although it is a general rule muscle contractile response was never observed. When two
of thumb that invertebrates have unmyelinated axons and spikes were elicited, a contraction sometimes occurred, and

Downloaded from on January 21, 2015

vertebrates have myelinated axons, vertebrates have a mix of when three or more spikes were elicited, muscle contractions
both unmyelinated and myelinated axons in their bodies and were “consistently observed.” This would be an interesting
some invertebrates, such as annelid worms and certain species result for students to try to replicate if they would like to build
of shrimp, have myelinated axons as well (14). their own circular track. Moreover, Drewes et al. found 20%
increases in conduction velocities (“facilitation”) of the second
Evaluation of Student Work spike when multiple spikes were elicited by tactile stimulation
(8). MGF second spikes were an average of 39 m/s, and LGF
College. Our survey consisted of 13 multiple-choice ques- second spikes were an average of 14 m/s.
tions, and the knowledge score was computed as the sum of Although there does not appear to be a difference between
correct answers. There was a significant difference in pretest electrical stimulation and tactile stimulation conduction veloc-
(mean: 5.6, SD: 0.34) and posttest (mean: 8.8, SD: 0.53) ities (our data compared with Kladt et al.’s data), there is a key
knowledge scores [t(9) ⫽ ⫺4.4956, P ⫽ 0.0015], suggesting difference between the two methods. During electrical excita-
that students improved their knowledge of core concepts of tion, spikes from both the MGF and LGF spikes are elicited
conduction velocity with a 25% average increase in test scores. regardless of electrode position, but, in our experiments with
Students notably increased their knowledge on questions on tactile stimulation, only MGF spikes are elicited when the head
earthworm anatomy and general conduction velocity theory but is tapped and LGF spikes when the tail is tapped. This is due
did not noticeably increase their correct responses to the more to the sensory receptors in the head and tail only connecting to
difficult questions of what changes in capacitance and resis- the MGF and LGF, respectively. A student could combine
tance across the neuron membrane will do to time constants electrical and tactile stimulation to compare these two types of
and length constants. This is somewhat expected, as the math stimuli.
behind conduction velocity can be difficult to grasp over a Experiments with electrical stimulation in our setup were not
2-h-long workshop for novices being exposed to it for the first successful; the stimulus artifact would cause our SpikerBox
time. We are now working on handouts to give to the students circuit to become unstable (swamped) with a recovery time of
before or after the lectures so they can study cable theory in 10 ms, masking the elicited spikes. We are planning to develop
more detail. amplifiers with blanking to allow electrical stimulation exper-
High school. The knowledge survey consisted of eight iments.
multiple-choice questions, and the knowledge score was com-
puted as a sum of correct answers. There was a significant Wider Educational Applications
difference in pretest (mean: 2.95, SD: 0.21) from posttest The experiments presented here can be used for a wide range
(mean: 3.81, SD: 0.23) knowledge scores [t(21) ⫽ ⫺3.3563, of high school and undergraduate physiology classes. The main
P ⫽ 0.0030], suggesting that students increased their knowl- experiment, in which students examine the different conduc-
edge of conduction velocity concepts with an 11% average tion velocities of two different nerve systems in the earthworm,
increase in test scores. Again, when we examined correct can be used as a basic teaching tool for action potential
versus incorrect responses, students increased their knowledge propagation and cable theory. The instructor can discuss how
on questions relating to earthworm anatomy and general con- axonal diameter and myelination have specific electrical effects
duction velocity theory but did not noticeably increase their on how the spike travels down the axon (the MGF has a larger
correct responses to questions relating to the nodes of Ranvier diameter than the LGF and thus has a faster conduction
or sparse coding. Handouts given to students after the lecture velocity). With the addition of mathematical and electrical

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

 Sourcebook Of Laboratory Activities In Physiology
CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS 69

principles, this experiment can easily lead to modeling work on If the neuron has a time constant of 1 ms, that means if a current
cable theory in interested students. The APPENDIX is a brief change is applied across a neuron, after 1 ms, 63% of the new voltage
primer on cable theory and its relationship to LGF and MGF is reached, after 2 ms, 86% of the new voltage is reached, and after 3
conduction velocity. Students can also begin exploring histol- ms, 95% of the new voltage is reached. The smaller rm and cm
become, the lower the time constant is and the less amount of time is
ogy techniques to get physical measurements of the diameters needed to change an axon’s voltage.
of the fibers in each worm. We have also recently found that An “ideal neuron” would have an infinitely high ␭ value and an
placing a segment of the worm over a block of ice in between infinitely low ␶ value. Thus, any voltage change anywhere in the
the two recording electrodes will lower the conduction veloc- neuron would instantly affect the voltage everywhere else in the
ities by ⬃50% and that the conduction velocities return to neuron.
normal when the worm is back at room temperature. A student Relation to conduction velocity. MYELIN. While commonly
could carefully plot the relationship between temperature and taught as a purely vertebrate invention, myelin-like coverings are used
conduction velocity, similar to Kladt et al. (17). in some invertebrate animals, such as annelids and certain species of
Experiments students can try that we have not tested are as prawn shrimp (see Ref. 18; for an extensive review, see see Ref. 14).
follows: 1) Ruston and Barlow (29) noted that motor responses Covering the neurons with myelin makes the inside and outside of the
in awake worms to tactile stimulation were much more robust neural membrane farther apart from each other, reducing cm, but this
covering also substantially increases rm. The result of this simultane-
in a dark environment rather than a light environment and 2)
ous reduction in cm and increase in rm is hypothesized to cause no net
Bullock (5) noted that that conduction velocity decreased as the change in ␶, although direct experimental evidence in the literature is
animal was harmlessly stretched longitudinally; this is suppos- lacking.
edly due to the diameter of the nerve cord changing (getting However, since myelin does increase rm, it has a dramatic effect on
thinner due to the stretching). ␭. The result is such that the relationship between myelin thickness
and conduction velocity is linear. Doubling the myelin thickness
APPENDIX doubles the conduction velocity, tripling the myelin thickness triples
the conduction velocity, and so on.

Downloaded from on January 21, 2015

A Primer on Cable Theory and Earthworm Conduction Increasing ␭ reduces the number of times that the action potential
Velocity Measurements needs to be regenerated by voltage-gated ion channels as it travels
down the axon; the ion channels each take ⬃1 ms to open in response
Time and length constants. Cable theory, along with the to voltage changes, so the less often ion channels have to open as an
Hodgkin-Huxley model of ion channel kinetics, is one of the most action potential propagates down an axon, the faster the conduction
celebrated biophysical modeling accomplishments in neuroscience. velocity will be.
Cable theory was originally developed in the 1800s, when engineers LARGER-DIAMETER AXONS. The other way that axons can increase
were trying to understand signal transmission across long-distance their conduction velocity is by increasing the diameter of the axon and
telegraph lines. While cable theory is occasionally covered in under- thus increasing ␭ as well. rm and ri are dependent on the constants Rm
graduate neuroscience lectures, it can seem fairly abstract. These and Ri, which are based on the composition of the neural membrane
earthworm experiments can help give context to cable theory. The and axoplasm. rm depends on the circumference of the axon, whereas
most important values relating myelin, axon diameter, and conduction ri depends on the cross-sectional area of the axon, as follows:

冑 冑
velocity are the length constant (␭) and time constant (␶).
rm Rm Ri
␭ can be determined as follows: ␭⫽ ⫽ ⁄

冑
ri 2␲ radius ␲ radius2
rm
␭⫽ By removal of the constants, the equation simplifies such that ␭ is
ri proportional to the radius of the axon, as follows:
where rm is the resistance across the axonal membrane and ri is the ␭ ⬇ 兹radius
internal axon resistance. rm is a measure of how “electrically leaky”
the axonal membrane is. The higher rm is and the lower ri is, the This square relationship of axon radius to ␭ means that the axon
higher ␭ will be. ␭ (sometimes called the “space constant”) is a would have to increase its radius four times to have an increase in ␭
measure of how far a voltage change at one point in an axon travels of two times (and, thus, a corresponding 2-fold increase in conduction
down the axon before it decays. The voltage decays according to the velocity).
following relationship: Thus, myelin scales much faster than increasing axon diameter due
⫺x to the myelin thickness’ linear relationship to conduction velocity as
e ␭
opposed to the axon diameter’s square root relationship to conduction
velocity.
where x is the distance away from the voltage change. If ␭ is 1 mm,
then that means that at 1 mm away from where the initial voltage Predictions in the earthworm. The diameter of earthworm
change occurred, 37% of the voltage magnitude remains. At 2 mm giant fibers has been measured at ⬃0.05 mm for the LGF and 0.07
away from the cell body in an axon, 14% of the magnitude remains; mm for the MGF (5, 13). As these fibers are myelinated, we would
at 3 mm away, 5% remains. expect the MGF to be 1.4 times faster than the LGF (0.07/0.05 ⫽ 1.4).
␶ can be determined as follows: As we measured the LGF speed to be 7.6 m/s, we would predict the
MGF to then be 10.6 m/s. However, we experimentally measured the
␶ ⫽ r mc m MGF to be 22.8 m/s (3 times larger velocity vs. 1.4 times larger
expected difference) and currently cannot explain this higher than
where cm is the capacitance across the neuron membrane.
predicted difference. We are beginning experiments to histologically
␶ is a similar exponential function but applies to time (t) as follows:
reexamine the size of the two fibers.
⫺t
1⫺e ␶ ACKNOWLEDGMENTS
If current flows across a neuron due to an ion channel opening, it takes The authors thank Cristina Mezuk for help with literature searches and Gina
time for the neuron to fully “charge” and reach a new stable voltage. M. Bello for expert proofreading and editing assistance.

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

Sourcebook Of Laboratory Activities In Physiology
70 CONDUCTION VELOCITY EXPERIMENTS USING EARTHWORMS

GRANTS 11. Gasser HS, Graham HT. Potentials produced in the spinal cord by
stimulation of dorsal roots. Am J Physiol 103: 303–320, 1933.
Support for this work was provided by National Institutes of Mental Health
12. Geier R, Blumenfeld PC, Marx RW, Krajcik JS, Fishman B, Soloway
Small Business Innovation Research Grants 1-R43-MH-093334-01 and 2-R44-
E, Clay-Chambers J. Standardized test outcomes for students engaged in
MH-093334-03: “Backyard Brains: Bringing Neurophysiology Into Secondary
inquiry-based science curricula in the context of urban reform. J Res Sci
Schools.” Our education research testing on high school and college students
was reviewed and approved through the Institutional Review Board of Albion Teach 45: 922–939, 2008.
College. 13. Günther J. Overlapping sensory fields of the giant fiber systems in the
earthworm. Naturwissenschaften 60: 521–522, 1973.
DISCLOSURES 14. Hartline DK, Colman DR. Rapid conduction and the evolution of giant
axons and myelinated fibers. Curr Biol 17: R29 –R35, 2007.
K. M. Shannon, G. J. Gage, A. Jankovic, and T. C. Marzullo are employees 15. Henze DA, Borhegyi Z, Csicsvari J, Mamiya A, Harris KD, Buzsáki
of Backyard Brains. W. J. Wilson has occasionally received equipment for G. Intracellular features predicted by extracellular recordings in the
testing/review from Backyard Brains. hippocampus in vivo. J Neurophysiol 84: 390 – 400, 2000.
16. Hodgkin AL, Huxley AF. Action potentials recorded from inside a nerve
AUTHOR CONTRIBUTIONS fiber. Nature 144: 710 –711, 1939.
Author contributions: K.M.S., G.J.G., A.J., and T.C.M. conception and 17. Kladt N, Hanslik U, Heinzel H. Teaching basic neurophysiology using
design of research; K.M.S., G.J.G., W.J.W., and T.C.M. performed experi- intact earthworm. J Undergrad Neurosci Educ 9: A20 –A35, 2010.
ments; K.M.S., A.J., and T.C.M. analyzed data; K.M.S., G.J.G., W.J.W., and 18. Kusano K. Electrical activity and structural correlates of giant nerve fibers
T.C.M. interpreted results of experiments; K.M.S. and T.C.M. prepared fig- in kuruma shrimp (Panaeus japonicus). J Cell Physiol 68: 361–384, 1966.
ures; K.M.S. and T.C.M. drafted manuscript; K.M.S., G.J.G., A.J., W.J.W., 19. Land BR, Wyttenbach RA, Johnson BR. Tools for physiology labs: an
and T.C.M. edited and revised manuscript; G.J.G. and T.C.M. approved final inexpensive high-performance amplifier and electrode for extracellular
version of manuscript. recording. J Neurosci Methods 106: 47–55, 2001.
20. Marzullo TC, Gage GJ. The SpikerBox: a low cost, open-source bioam-
REFERENCES plifier for increasing public participation in neuroscience inquiry. PLos
One 7: e30837, 2012.
1. Backyard Brains. Experiment: Debunking the P-Value With Statistics
21. Minner DD, Levy AJ, Century J. Inquiry-based science instruction–
(online). https://www.backyardbrains.com/experiments/p-value [26 June
what is it and does it matter? Results from a research synthesis years 1984

Downloaded from on January 21, 2015

2013].
to 2002. J Res Sci Teach 47: 474 – 496, 2010.
2. Backyard Brains. Experiment: Referencing Your Spikes (online). https://
22. National Research Council. A Framework for K–12 Science Education.
backyardbrains.com/experiments/referencing [13 September 2013].
3. BackyardBrains.2-ChannelSpikerBoxBundle(online).https://backyardbrains. Washington, DC: National Academies, 2012.
com/products/twochannelspikerboxbundle [30 October 2013]. 23. Nicol JC. The giant axons of annelids. Q Rev Biol 23: 291–323, 1948.
4. Backyard Brains. Experiment: the Faraday Cage (online). https://backyardbrains. 24. Prince MJ, Felder RM. Inductive teaching and learning methods: defi-
com/experiments/faraday [22 October 2013]. nitions, comparisons, and research bases. J Eng Educ 95: 123–138, 2006.
5. Bullock TH. Functional organization of the giant fiber system of Lumbr- 25. Pulver SR, Hornstein NJ, Land BL, Johnson BR. Optogenetics in the
icus. J Neurophysiol 8: 55–71, 1945. teaching laboratory: using channelrhodopsin-2 to study the neural basis of
6. Dagda RK, Thalhauser RM, Dagda R, Marzullo TC, Gage GJ. Using behavior and synaptic physiology in Drosophila. Adv Physiol Educ 35:
crickets to introduce neurophysiology to early undergraduate students. J 82–91, 2011.
Undergrad Neurosci Educ 12: A66 –A74, 2013. 26. Roberts MB. The giant fibre reflex of the earthworm, Lumbricus terrestris
7. Drewes CD. Non-invasive recording of giant nerve fiber action potentials L. I. The rapid response. J Exp Biol 39: 219 –227, 1962.
from freely moving oligochates. In: Tested Studies for Laboratory Teach- 27. Rosenthal JJ, Bezanilla F. Seasonal variation in conduction velocity of
ing, edited by Karcher SJ. Chapel Hill, NC: Proceedings of the 20th action potentials in squid giant axons. Biol Bull 199: 135–143, 2000.
Workshop/Conference of the Association for Biology Laboratory Educa- 28. Ruston WA. Reflex conduction in the giant fibers of the earthworm. Proc
tion, 1999, vol. 20, chapt. 2, p. 45A66 – 62. R Soc 133: 109 –120, 1956.
8. Drewes CD, Landa KB, McFall JL. Giant nerve fibre activity in intact, 29. Ruston WA, Barlow HB. Single-fibre response from an intact animal.
freely moving earthworms. J Exp Biol 72: 217–227, 1978. Nature 152: 597–598, 1943.
9. Dubinsky J. Neuroscience education for prekindergarden-12 teachers. J 30. Stough HB. Giant nerve fibers of the earthworm. J Comp Neurol 40:
Neurosci 30: 8057– 8060, 2010. 409 – 463, 1926.
10. Gasser HS. The classification of nerve fibers. Ohio J Sci 41: 145–159, 31. Stough HB. Polarization of the giant nerve fibers of the earthworm. J
1941. Comp Neurol 50: 217–229, 1930.

Advances in Physiology Education • doi:10.1152/advan.00088.2013 • http://advan.physiology.org

View publication stats