Introduction
Training
dolphins - it’s the career
of many people’s dreams! Almost everyone has
either read about dolphins or seen them on television.
A few people may have been fortunate enough to observe
them in their wild habitat. Many of you probably
have seen trained dolphins in person at an aquarium
or marine
mammal facility. The intelligence and
creativity of dolphins make working with them both
challenging and a lot of fun. As you watch trained
dolphins gracefully perform various behaviors on
cue, the trainer’s job may look easy. It seems
as though the trainer “directs” and the
dolphin “performs”. What you don’t
see is all the time, planning, and expertise that
go into this unique kind of dolphin-human
relationship.
As you read this text, remember
that the same basic principles that apply to shaping
dolphins’ behavior
also apply to other animals, including people. Try using
some of these techniques with your family or friends.
You may be surprised at how well they work!
Why We Train Dolphins at Dolphin Research Center
The DRC dolphins demonstrate
their individual talents and abilities for our visitors
daily in training sessions that take place during public
sessions. While educational and fun for our guests,
this is not the main reason that we train. The dolphins’ health
and happiness is our main priority. Training the dolphins
helps us to accomplish this in several ways:
Physical Stimulation:
Dolphins are very strong animals and capable of performing
amazing physical feats. Many of the behaviors that you
see the DRC dolphins do are natural behaviors for dolphins
in the wild. However, wild dolphins spend a lot of time
and energy swimming and trying to catch their food. At
times, they must swim fast to avoid predators like sharks.
When their stomachs are full and they are safe, they
then have time to frolic and play.
Although the dolphins at DRC occasionally chase and,
sometimes, eat the fish in their lagoons, they do not
have to work hard at catching food because they realize
that it comes to them from the trainers. They also do
not have to worry about safety; the perimeter fences
that separate their lagoons from the open water are designed
to keep sharks out. Since the DRC dolphins do not have
to use their energy for everyday survival, it is important
that they get the proper physical exercise. An important
purpose of training is to keep them active and healthy.
Mental Stimulation:
Dolphins and humans both exhibit high levels of cognitive
functioning. Just imagine if you never had the opportunity
to stimulate your brain by reading a book, playing a
game, solving a puzzle, or learning a new skill; you
probably would become very bored. Almost everyone loves
an occasional challenge, and dolphins are no different.
They constantly amaze us with their eagerness and ability
to learn new things and to solve problems. The mental
stimulation that makes life interesting and fun is also
very good for the dolphins. The stimulation of continually
learning new things develops a healthy ability to adapt
to changes in the environment. Adaptability is important
for all animals, including humans!
Medical/Husbandry Applications:
Without regular medical checkups, the only way we could
assess the dolphins’ physical well being would
be by how they act and how much they eat. This is not
reliable, since it is instinctive for most animals to
hide any illness or injury; many predators take advantage
of the sick or weak. Routine checkups can tell us how
healthy the dolphins are and allow us to take preventative
steps to ensure that they stay healthy.
We have learned at DRC that
the dolphins will accommodate many medical procedures
voluntarily once the trainers teach them these procedures.
Blood, urine, fecal, stomach and blowhole samples are
collected from the dolphins by DRC staff and analyzed
in the lab. In order to get samples, the dolphins are
trained to do certain things, such as holding the tail
very still in a trainer’s
lap while a blood sample is taken from the large vessels
found in the tail flukes. Just as with people, medical
samples can tell us a lot about what is going on inside
the body.
As is the case with the frozen food we eat, analysis
of the fish we feed the dolphins tells us that some nutrients
are depleted in the freezing process. To ensure that
the dolphins are getting all the proper nutrients, we
supplement their diet with vitamins. For some dolphins,
we have to stuff the vitamins into their fish, but, through
training, many of our dolphins have learned to swallow
their pills directly!
Education and Research
Applications: We are only just beginning to
understand the needs and capabilities of the bottlenose
dolphin, let alone all the other species of dolphins
and whales with which we share this planet. At DRC,
we believe it is important to learn all we can from
and about the dolphins in our care, and teach what
we know, so that people will have a better understanding
and appreciation of these fascinating animals.
While we can learn some things from just observing
dolphins, training
dolphins allows us to learn much more about
their physical and mental functioning. In order to give
the dolphins a way to “answer questions,” a
system of communication must be established.
For example, you had to learn to read in order to understand
this text. In order to communicate with dolphins, we
train them to understand and interpret various cues.
Training cues can be many things, such as objects, symbols,
hand signals, or sounds.
Principles of Training
There is a lot to learn before a prospective trainer
is ready to interact with dolphins in a training capacity.
Apprentice trainers at Dolphin Research Center learn
their job by working through a series of progressive
stages.
Rapport:
The cornerstone of all successful training is a strong
dolphin-trainer relationship. All DRC trainers begin
by simply making friends with the individual members
of our dolphin family. An effective trainer will work
to develop a bond of trust and will learn what kinds
of things each dolphin enjoys. Rapport is built through
feedings, play sessions, and dock time. In the early
stages of the relationship, trainers must be very observant
of the dolphin’s behavior. Trainers
should strive to be approachable, gentle, creative, and
FUN! The dolphin is also discovering what a particular
trainer enjoys. If a trainer has a loud, cheerleader-type
personality, the dolphin may respond by becoming loud
and animated, as well!
Operant Conditioning: Our
method of training dolphins is based on operant conditioning.
In the training situation, the dolphin is an “operator,” an
active participant who exhibits a behavior to which the
trainer then responds. The consequences which follow
a behavior directly influence the frequency with which
the behavior will be repeated. Training is very much
a two-way communication process.
The trainer’s response to the dolphin will either
increase or decrease the likelihood that a given behavior
will reoccur. If a behavior is followed by something
that the dolphin likes, the probability that the behavior
will occur again is increased. If the trainer’s
response is not agreeable, the dolphin eventually will
tend to avoid that particular behavior.
If your mother wants to “train” you to
take out the trash, she may use a variety of methods.
She may remove privileges or get upset if you don’t
take out the trash, or reward you with praise about how
nice it is when you do take out the trash. Either punishment
or praise may influence how often you take out the trash.
However, using a method that incorporates positive reinforcement
is most likely to encourage you to take out the trash
on a regular basis.
Positive reinforcement has proven to be the most effective
means of influencing behavior. If, after you take out
the trash, your mother rewards you with a warm smile,
a piece of cake, or the car for the day (depending on
your preferences), you will be more likely to take out
the trash again without being asked. Your behavior has
been modified through operant conditioning.
Note: You
will notice that we refer to “behaviors” when
discussing what we teach the dolphins. When we train
dolphins at DRC, we do not teach them to
do “tricks”. Tricks imply illusions, things
that are not real, but intended to puzzle or amuse. Dolphins
are incredible
athletes and intelligent mammals; their accomplishments
are real. We teach them to do specific behaviors on cue,
which we request by giving verbal or hand signals.
Basic Training Tools & Techniques:
Reinforcement: There
are two types of reinforcement: positive and negative.
Both increase the probability that a behavior will reoccur
because both provide a positive consequence for the behavior,
although in different ways.
Positive Reinforcement: provides
a positive stimulus, something the dolphins enjoy and
desire, as a consequence of the behavior. Positive reinforcement
may be PRIMARY (something inherently positive, requiring
no learning experiences) or SECONDARY (something that
acquires its positive effects through learning experiences).
For dolphins and humans alike, food is a primary reinforcement.
Therefore, the primary reinforcement for dolphins is
usually, but not always, fish.
For many people, a paycheck becomes a powerful secondary
reinforcement, because money is associated with so many
things we may desire (food, shelter, clothing, power).
After learning to read and discovering the pleasure it
can provide, a book may become a secondary reinforcement.
A dolphin may learn to enjoy playing with a certain toy,
and so the toy becomes a secondary reinforcement. At
DRC, the sound of the crowd cheering for the dolphins
has become a unique secondary reinforcement for many
of them.
Some people find affection and hugs rewarding; others
might find quiet and solitude rewarding. Everyone is
different. Knowing the personality of each dolphin is
important because individual preferences vary. Praise
and applause seem to be highly rewarding for many dolphins.
Some like to play with toys, others prefer back rubs
or having their trainer go for a swim with them, while
others may enjoy doing their favorite behavior.
Primary and secondary reinforcements are randomly intermixed
throughout the session, keeping it unpredictable and
fun for both trainer and dolphin.
An intermittent or partial schedule
of reinforcement, with variable or random
amounts of reinforcement given in a random way, produces
the strongest results. One fish after every behavior
is boring and predictable, so instead, trainers vary
frequency, size and type of reinforcer. Sometimes trainers
surprise the dolphins with “Jackpots”—handfuls
of fish. Your attention level stays higher if you never
know when you might hit the jackpot.
Negative Reinforcement increases
the future probability of a behavior because a negative
stimulus is removed as a consequence of the behavior.
A dolphin can achieve a reinforcing consequence by
performing some action, which causes an undesirable
stimulus to be removed. The situation may be as simple
as the dolphin making a move to avoid an object they
don’t like.
For example, a trainer might be attempting to teach
a dolphin to go through a gate into an adjoining pool.
He might wish to positively reinforce any movement toward
the pool; but if the dolphin never moves in that direction,
there will be nothing to positively reinforce.
However, if the dolphin tends to avoid things, the
trainer might try using an unusual object, such as a
plastic chair, to create a situation in which the dolphin
can receive negative reinforcement. When the trainer
splashes the chair into the water, the dolphin may retreat
into the next pool due to the novelty of this object.
The behavior is achieved because the dolphin wants to
avoid the undesirable stimulus: contact with the unusual
object. The dolphin receives a desirable, reinforcing
consequence (a pool with no chair in it) for the behavior.
Positive reinforcement (such as fish or praise) for the
desired movement can then also be offered, further reinforcing
the behavior.
Nonreinforcement
Even with all of the correct
tools for training, there is no guarantee that all
of the dolphin’s behavior
will be positive, or something the trainer wants to reinforce.
In dolphin training, you look for the behavior or steps
in a behavior that you can positively reinforce, and
ignore behavior that is undesirable or “not on
the right track”. If the dolphin is offering behavior
that is undesirable for some reason, not allowing any
stimulus, positive or negative, to follow the behavior
tends to extinguish that behavior.
Nonreinforcement decreases
the likelihood that a behavior will be repeated because
the behavior is not followed by a positive consequence.
Punishment is
nonreinforcing a behavior by presenting a negative
stimulus as a consequence of the behavior. We choose
not to utilize this method in working with the dolphins.
As dolphin/human relationships at DRC are based on
trust and the enjoyment of each other’s
company, we do not believe punishment would be an appropriate
way of influencing dolphin behavior and, therefore, has
no place in our work with dolphins.
Time Out: Sometimes,
to decrease an
undesirable behavior, trainers utilize a procedure called
a time out. A time out is nonreinforcing a behavior by
removing any positive stimuli from the trainer. The time
out serves to decrease the future probability of the
behavior it follows. One situation in which a time out
may be used is when a dolphin is exhibiting inappropriate
behavior such as chasing another dolphin away from the
dock during a session. The trainer will calmly turn their
back to avoid even the reinforcement of eye contact,
or will leave the dock altogether. The dolphins will
not be as likely to chase one another at the dock if
they lose all attention and other reinforcement from
the trainer as a result.
Another situation in which
a time out might be implemented could involve a dolphin
that refuses to do a behavior it apparently knows.
Rather than try to re-train the behavior (which would
be an insult to the dolphin’s
intelligence), the trainer may simply decide to take
a time out. In this case the time out is communicating, “We
both know you can do this. Let’s take some time
to think about it and we’ll try again later.”
Time outs vary in length. When used appropriately,
time outs can be a very effective training technique,
because we have seen the dolphins try to elicit our attention.
They may vocalize or even throw seaweed at us, to get
us to come back. It may be hard not to respond, but if
we did, the dolphins would be shaping our behavior and
it would defeat the purpose!
It is interesting to note
that dolphins also use the time out technique with
us. For instance, if during dock time (a non-training,
free time visit to a dolphin’s
dock) a person exhibits behavior the dolphin doesn’t
like (such as reaching out to him repeatedly when the
dolphin is not in the mood to be touched), the dolphin
will respond by removing itself from the vicinity of
the dock and ignoring the visitor.
Training Tools
Bridge:
In order for operant conditioning to be successful,
you must somehow inform the dolphin that they are doing
the appropriate thing, while they are doing it. If
a dolphin is jumping twenty feet away from the trainer,
it would be almost impossible to offer a fish or some
other reinforcement at the exact moment the dolphin
does the behavior correctly. Instead, in early training,
the sound of a whistle is associated with rewards of
fish and praise. The dolphins quickly learn that when
they hear the whistle, they have done the behavior
correctly and they can return to the dock for their
reward. Therefore, the sound of the whistle becomes
a conditioned reinforcer that serves to BRIDGE the
gap in time between the instant the dolphin does the
correct behavior and when the dolphin receives its
reward. That is why we call the sound of the whistle
a “bridge”.
Learning to bridge correctly
with a whistle may be one of the biggest challenges
for a new trainer. The whistle must be blown at the
precise instant the behavior is correct in order for
it to be effective in letting the dolphin know, “YES! That’s great! That’s
exactly what I’m looking for!”
Recall: Another
important acoustic tool that the trainer uses is the
recall. The recall can come in many forms. It can be
a tap on a bucket or cooler, an underwater sound, or
even a splash on water by your hand. There are times
during the training session when the trainers need
to use the recall sound. The dolphins may be away from
the dock for a variety of reasons—the
trainer may have misdirected them, they may be sidetracked,
or they may be performing a behavior that is incorrect
or below bridging criteria. The trainer can utilize the
recall to ask the dolphin to return to the dock for redirection.
One of the strongest underwater
sounds that we train is a pinger recall. The pinger is
a special frequency sound that travels long distances
underwater. The dolphins are taught that if they hear
this sound, they are to respond no matter what they were
doing. Because of our location in the Gulf of Mexico,
we want to be able to call the dolphins back to us if
they should ever find themselves outside of our lagoons
and in open water. This could happen in the event of
a hurricane, for example. If our perimeter fences were
damaged and the dolphins swam out and got “lost”,
the pinger would be their guide to lead them to safety
and home.
Stationing:
Before giving a hand or verbal signal, the trainer
first must get the dolphin’s
attention. At DRC, the dolphins learn to STATION when
the trainer positions an index finger up in front of
them. The trainer expects the dolphin to look up at them,
head vertical, eyes above the water. This tells the trainer
they have the dolphin’s complete attention.
Target:
A target is used to indicate location or position to
a dolphin. The dolphin is taught to touch a target
such as a trainer’s palm or a
ball on the end of a pole (known as a target pole). Moving
the target and bridging when the dolphin touches it pinpoints
a critical location or point of the behavior. The target
pole can be used as an extension of the trainer’s
arm, helping to shape behaviors farther away from the
dock.
Symbol:
Although not a standard training tool, symbols are
frequently used at DRC. Dolphins have a personal symbol,
a shape we put in the water to let the dolphin know
where they are going to be. The symbol can be any shape
or material and is used to denote where (i.e. at which
dock or pool location) each dolphin’s
session will begin.
Methods Of Training
There are several general methods used to teach behaviors
to a dolphin. In order to get to the end product, the
trainer usually must find a way to prompt the dolphin
into doing the behavior, or a step toward the behavior.
Some methods work best for certain behaviors and some
methods work best with particular dolphins. Deciding
which training method to use is at the discretion of
the trainers and depends on what they feel will work
in a given situation.
Direct Manipulation,
by physically maneuvering a dolphin through the desired
motion, then reinforcing the behavior, is one method
of training. For example, if the trainer wants two
dolphins to touch their rostrums in a “kiss,” he/she
may gently push the dolphins’ rostrums together
until they touch, then bridge at the touch. Direct
manipulation is not used often with dolphins because
of the difficulties presented by their size and in-water
habitat.
Imitation or Modeling: A
trainer may model the behavior for the dolphins and
reinforce imitation of that behavior. The imitation
method might be used to train a dolphin to mimic sounds
or behaviors such as bobbing up and down. Once the
concept of imitating on cue is learned by the dolphins,
you can then ask the dolphins to imitate other dolphins,
sea lions, or people—at
DRC, sometimes members of our public tours.
Targeting uses
a tool such as the trainer’s hand or the target pole to shape a behavior.
Once a dolphin has been taught to touch targets, the
trainer can use the hand or pole to maneuver the dolphin
into doing various movements that can then be reinforced.
For example, to teach a dolphin to nod his head “yes,” the
trainer asks the dolphin to target on their hand, and
then moves the hand in an upward and downward motion
while the dolphin’s rostrum is targeted on it.
By reinforcing the movement and pairing with a signal,
the dolphin will eventually move his head up and down
on signal without the hand target present.
Successive Approximations shape
a behavior by encouraging and reinforcing (often using
targeting and bridging) small steps that lead in the
general direction of the final behavior. The dolphin
is rewarded for behavior that gradually looks more and
more like the desired end result, while behavior that
deviates from the desired end-point is ignored. The trainer
may use successive approximations to refine behavior
already known (such as encouraging higher dives) or to
begin new behavior training (refer to the “Sample
Behavior Chain” which follows for an example).
Opportunistic Training/Capturing:
Some of the first behaviors dolphins learn to exhibit
on signal are spontaneous behaviors that do not involve
any initial prompting by the trainer. The trainer sees
the dolphin exhibit a desirable behavior that occurs
spontaneously, and immediately bridges and reinforces
it, showing the new signal they wish to pair with that
behavior. Trainers must be ready at all times because
the dolphins often exhibit such behaviors between session
times.
As you will read in “Training A New Behavior” (below),
generally one trainer will work on a new behavior with
a dolphin, but in the case of opportunistic training,
everyone helps out. It is important to try to catch the
behavior as many times as possible, and the dolphin may
offer the behavior at any time, in and out of sessions.
If the behavior is reinforced and the new signal shown
consistently, soon the dolphin will associate the new
signal that has been given with the behavior it has just
exhibited, and you now have a new behavior added to that
animal’s repertoire.
Opportunistic training is often used to get vocals
on signal, since it is impossible to physically manipulate
a dolphin to produce a certain sound. For example, all
the trainers began to show one of our dolphins the same
hand signal every time she offered her signature whistle,
and we now can ask her to give this unique whistle on
cue. Since dolphins are extremely innovative, the opportunistic
method of achieving new behaviors is very popular with
them (and fun for us).
Training Session Techniques
Before beginning a session on the dock, a trainer should
be aware of the following techniques that will help ensure
a successful experience for both trainer and dolphin.
The trainer must go down onto the dock both physically
and mentally prepared for the session. This
means having all the essential tools, plus any additional
props that may be needed. They must leave any excess
mental baggage (worries, stress) behind and have their
attention focused on the dolphin. The trainer should
have good, erect posture, and a positive “we
are going to have a great time!” attitude, which
set the stage for a positive learning experience.
Trainers working simultaneously with different dolphins
in the same lagoon begin and end each
session at the same time. Most dolphins are curious and
social creatures. If one dolphin is getting a trainer’s
undivided attention, other dolphins that are not occupied
may vie for it as well, which can cause the training
session to end in chaos. Therefore, each dolphin or team
of dolphins in a pool works with their trainer simultaneously
to create harmony during sessions.
Good trainers avoidroutine,
and never start or end sessions in the same way. Trainers
try to be “consistently inconsistent,” which
challenges both the dolphin and the trainer.
If a session is not going well, the trainer can always
leave and come back later. It is better for a trainer
to leave and collect her thoughts than to continue a
session feeling frustrated. Good trainers try to end
the session on a positive, high note.
A new step in training successfully mastered, or a known
behavior performed beautifully are both good ways to
end a session. The goal is for both the trainer and the
dolphin to feel great about the session, and want more
interaction. This way the dolphin will eagerly anticipate
the next training experience, and so will the trainer.
Training A New Behavior
Initially, an apprentice trainer practices the techniques
he is learning by maintaining behaviors that the dolphins
already know. After the apprentice has mastered the basics
and understands the concepts we have discussed thus far,
he is ready to train a new behavior.
The first step in any new training is to write a behavior
chain. A behavior chain is a specific plan
of action that details each training step leading to
the desired final behavior. Writing a behavior chain
requires the trainer to think through the entire process.
Planning is important because once a trainer is on
the dock training the dolphin, they will not want to
take precious time and risk losing the dolphin’s
attention by stopping to think out the next step in
the progression. Sometimes a dolphin learns quickly,
so a good trainer will be ready to move on. If, on
the other hand, the dolphin is confused, the trainer
must be ready to offer an understandable alternative.
By creating a behavior chain ahead of time, the trainer
is prepared and confident when teaching the dolphin.
Generally, a trainer works
one-on-one with a dolphin until a new behavior is completed.
In initial training, progress may be in small increments,
and it is difficult to communicate the nuances of the
dolphin’s actions
effectively between two or more trainers. Having different
people training the same behavior to the same dolphin
could be confusing for the dolphin. There are some exceptions
to this rule, as in the case of opportunistic training
when all the trainers work to associate a spontaneous
behavior with the chosen signal.
In new training, trainers
tend to rely mostly on primary reinforcement, but many
dolphins get more excited over the trainer’s secondary reinforcement. Secondary
reinforcement includes praise, applause, and often the
trainer’s own excitement over the dolphin’s
new accomplishments.
How long does it take to teach a new behavior? It depends
on the dolphin, the complexity of the behavior, the trainer,
and how often sessions are available to work on the behavior.
It is usually better to teach many short lessons than
a few long ones. Some dolphins take to learning certain
kinds of behaviors more quickly than others; they all
have their preferences.
When learning a new behavior
that they particularly like, the dolphins often will
spontaneously deliver, or “ad-lib,” the
behavior during other sessions, or even between sessions.
If you would like to write your own behavior
chain, follow these steps :
1. Decide on a behavior you would like to train. Remember
to be realistic, especially if this is your first training
attempt. Many creative behaviors can be trained easily,
making the early training experience positive for both
trainer and dolphin.
2. Devise a signal for the new behavior that does not
resemble any signals the dolphin already knows. Some
trainers find this to be one of the more difficult steps!
3. Visualize what the final product will look like.
It is important that you know exactly what to look for
so there will be no hesitation in your bridging. Timing
is everything!
4. Break down the behavior to determine what steps
will be necessary to lead to the final result. List the
steps numerically, from the first, most basic building
block to the finished behavior.
5. Decide what method you
will use to teach each step. Keep in mind what behaviors
the dolphin already knows that may help you train the
new behavior. For example, when DRC’s Delphi learned to paint, the trainer
first taught him to hold a special brush in his mouth.
The next step was to have Delphi move his head while
holding the brush. Delphi already knew the “yes” and “no” behavior,
so the trainer simply gave the signal for the known behavior
while Delphi was holding the brush. These shortcuts can
help speed training time, and give the dolphins confidence
because they are secure with known behaviors.
6. Write down alternative methods in case your first
method is unsuccessful. A back-up plan is always necessary
because you cannot be sure how the dolphin will react.
What seemed perfectly easy to understand written out
at your desk may result in the dolphin giving you a blank
stare during a session. Be prepared to offer another
alternative route to the same end goal.
Sample Behavior Chain
DOLPHIN: Pax BEHAVIOR: Flip
TRAINER: Lindsay START DATE: 4/6/00
SUPERVISOR: Linda DATE COMPLETED: 1/12/01
Signal: With right
hand, touch tip of index finger to tip of thumb (forming
a circle), other three fingers extended straight up.
Should look like the “Okay” sign.
Steps for Behavior
1. Station and show new signal.
2. Tap long target pole on
the water’s surface
to the left of the dock. When Pax touches it, bridge.
3. Repeat Steps 1 and 2, except this time, as soon
as Pax touches the target, pull the target along the
surface to the front of the dock. When Pax follows, push
the target under the water and make a circular motion.
As Pax tucks his head to follow the target, bridge.
4. Repeat Steps 1-3. Encourage Pax to follow the target
and actually do an underwater somersault with it. Bridge
when he tucks and continues to follow the target around
under water.
5. Repeat Steps 1-4. Once Pax is going all the way
around in a circular motion, begin raising the target
slightly out of the water. Pax should raise his head
up to the target, then follow the target as it is pulled
forward and down fast to do the underwater somersault
motion. Bridge for his head movement forward and down.
6. Repeat Steps 1-5. Since
every step begins with Pax starting out to the left
some distance, he should get the idea that he will
need a run to gain momentum for this behavior. Start
phasing out the tap to the left to see if Pax will
do the swim to the left without the target. (This is
the beginning of teaching a “run,” a
pass to build up speed for an aerial behavior.)
7. Repeat Steps 1-6. Start gradually raising the target
higher and higher. Bridge when Pax comes up to the target,
does his chin tuck, and starts to fall over (i.e. at
the peak height of his tuck.)
8. Begin bridging for a good tuck position. This will
result in a fast, good revolution flip.
9. Once Pax is doing the somersault out of the water,
while following the target, the next step is to remove
the target. Signal for the flip, and as Pax approaches
quickly, move the target out of the way before he can
reach it. Bridge if he tucks on his own. At this point,
reduce other criteria (tight tuck) until Pax is initiating
the tuck on his own without the target cue.
10. As the target pole is being phased out, start slowly
expecting better form (tight tuck, good revolution).
Bridge for form first, and once that is mastered, shape
height by bridging for the highest flips.
11. Final step is to shape the placement of the flip
in the lagoon. To encourage Pax to do the flip away from
the front of the dock, give the signal from the boardwalk
or beach area. Bridge flips done in center of pool. If
this is not successful, try training placement of the
behavior using ice cubes as markers. Train Pax to go
where the ice cubes hit the water, and to do the flip
in that location. Write another behavior chain for that
plan if it becomes necessary.
Note: You can see from this behavior chain
that Pax is an experienced dolphin. He has already learned
a strong target-follow behavior, allowing the trainer
to move right into the shaping of the flip.
Also note the two alternatives for placement of the
flip at the end of the behavior chain. This is far along
in the training of the behavior, so the trainer can plan
details of the alternatives as the training progresses.
Incorporating New Training Into A Session
New training can be incorporated
into any training session, at any time in the session.
The trainer may want to begin with the new training,
catching the dolphin’s
initial eagerness for the session. Other times he may
decide to work on known behaviors first, working on the
new behavior later in the session. This decision is totally
up to the trainer. Many good trainers assess the dolphin’s
attention and energy levels at the start of the session
and base their decision on that. Remember, variety makes
a session interesting.
To begin training, the trainer
will station the dolphin and show the new signal. The
dolphin generally will give the trainer a blank stare,
or the dolphin’s eyes
may widen with curiosity. If the dolphin offers a behavior
after the signal, then the new signal may be too close
to one he already knows.
From this point on the trainer will implement the behavior
chain she has written. One of the hardest things to remember
is to quit while you are ahead. The trainer must remember
there are many sessions available to complete the behavior,
and not to rush it. The rule of ending each session on
a high note is especially important with new training.
Advanced Training
As you become more comfortable working with the dolphins,
new training techniques become easier. You can begin
to challenge yourself and the dolphins with more complex
sessions.
Alternate Bridging
At times the whistle we use
as the conditioned reinforcement may not be appropriate
(for instance, in situations where the dolphin’s
ear is very close to the trainer), so we must find
another method to bridge the gap in time between the
peak of the behavior and its reinforcement. At DRC
we teach a dolphin to recognize a double tap of our
hand on his back or side as a substitute-conditioned
reinforcement.
Initially, the double tap is given with the whistle
bridge immediately following it. Once the dolphin responds
to the double tap, we start eliminating the whistle.
Group Behaviors
Group behaviors are fun and exciting to watch. The
training of them can be just as exciting. There are two
methods we use at DRC when asking a group of dolphins
to perform a behavior together for the first time.
a) Start with each dolphin having his or her own trainer
at the same dock. The trainers then give the dolphins
the same signal simultaneously.
b) One trainer gives the signal with all the dolphins
at the same dock.
Even though each dolphin knows the behavior, there
are other elements involved when asking him or her to
execute the behavior in a group with other dolphins.
The first goal is for all the dolphins to perform the
correct behavior, whether they are coordinated with each
other or not. We might then start adding other criteria.
For example, if we want to be able to ask a group of
dolphins to execute beautifully synchronized high dives
out of the water, at first we would bridge when they
all did a dive on signal, no matter where they were in
the pool. We might then look for them to be closer together,
going in the same direction, getting the form better,
and finally we may look for height. Each criterion is
shaped individually before moving on to the next criterion.
When introducing something new to an established behavior,
you must initially relax your previous standards for
bridging that behavior.
Medical Behaviors
Medical behaviors may be the most important part of
training at DRC. Medical behaviors usually involve asking
the dolphins to lie still on the surface of the water.
Medical behaviors therefore may take more time and patience
to train.
Trust is the key to training
medical behaviors. Desensitizing a dolphin to the prick
of a needle, a tube sliding down the throat, or holding
still for an external exam can be very challenging.
We never trick the dolphins. We always let them see
what we are going to do—such
as showing the needle to get a blood sample. As the dolphins
learn what we are trying to do, they relax and cooperate
with the procedure. As with all behaviors, it is of the
utmost importance to understand the preferences of the
individual dolphin, and incorporate that understanding
into how you proceed.
Initially during the training of medical behaviors,
we simulate how the final behavior will be accomplished.
For instance, to take a blood sample, we generally need
two people aside from the trainer to help in the process,
a cooler to store the blood, and a box containing the
medical equipment. We include the introduction of these
items in our behavior chain so that the dolphins get
used to having them around.
Research Training
Research studies usually involve some of the most challenging
training for both dolphin and trainer. It is critically
important to design the study so you know that the question
you are trying to ask is the one the dolphin is trying
to answer. A great deal of time and thought goes into
how we will get the question across.
As in any research, in a
study with dolphins you must control as many variables
as possible to achieve reliable results. Dolphins,
for instance, may look for “social
cuing”; that is, they will watch their trainers
very closely and figure out the answer from even the
subtlest body language. For example, if you have two
paddles in the water, and the paddle on the right indicates
the correct answer, the trainer’s eyes may unconsciously
move toward that paddle. The dolphin is watching the
trainer’s eyes so closely he sees that cue, and
presses the paddle on the right. In this case, to avoid
social cuing the trainer would need to be behind a screen
or wearing goggles so the dolphin couldn’t see
her eyes.
Another situation in which research results may be
invalidated is when the subject, in this case, the dolphin,
develops a bias. One way that bias can creep into a research
study is by varying the reward for correct responses.
Inadvertently giving a larger reward for moving in a
certain direction may encourage the dolphin to develop
a preference for moving in that direction, resulting
in inaccurate data. A consistent reward (such as two
fish for every correct response) prevents the dolphin
from developing a bias towards any particular response.
The challenges in research training are many, but the
rewards are great. Very gradually, we are learning more
and more about the capabilities and needs of our dolphin
friends. The dolphins seem to have a great deal of fun
figuring out the puzzles we present to them.
REFERENCES
Blackman, Derek. 1974. Operant Conditioning: An
Experimental Analysis of Behavior.
Methuen & Co: London,
England.
Freedman, Russell and James E.
Morriss. 1969. How
Animals Learn. Holiday House: New York, NY.
Pryor, Karen. 1975. Lads
Before the Wind: Adventures in Porpoise Training..
Harper and Row,
New York, NY.
Pryor, Karen. 1985. Don’t
Shoot the Dog.
Bantam.
Pryor, Karen. 1995. On Behavior: Essays and Research.
Sunshine Books: North Bend, WA.
Ramirez, K. 1999. Animal Training: Successful Animal
Management Through Positive
Reinforcement . Shedd Aquarium: Chicago, IL.
Reynolds, G. S. 1975. A Primer
of Operant Conditioning.
Scott, Foresman, and Co: Palo Alto, CA.
Sutherland, N. S. and N. J. Mackintosh. 1971. Mechanisms
of Animal Discrimination Learning.
Academic Press: New York, NY.
 Copyright © 2007 Dolphin Research Center
Information Can Be Used with Proper Citing
Dolphin Research Center, 58901 Overseas Highway, Grassy Key, FL 33050-6019
305-289-1121 www.dolphins.org
|
