The Development of an Operant Conditioning Training Program for New World Primates at the Bronx Zoo

JOURNAL OF APPLIED ANIMAL WELFARE SCIENCE, 6(3), 247–261
Copyright © 2003, Lawrence Erlbaum Associates, Inc.
The Development of an Operant
Conditioning Training Program for New
World Primates at the Bronx Zoo
Gina Savastano, Amy Hanson, and Colleen McCann
Mammal Department
Bronx Zoo/Wildlife Conservation Society
Bronx, New York
This article describes the development of an operant conditioning training program
for 17 species of New World primates at the Bronx Zoo. To apply less invasive tech-
niques to husbandry protocols, the study introduced behaviors—hand feeding, sy-
ringe feeding, targeting, scale and crate training, and transponder reading—for for-
mal training to 86 callitrichids and small-bodied cebids housed in 26 social groups.
Individual responses to training varied greatly, but general patterns were noted among
species. With the exception of lion tamarins, tamarins responded more rapidly than
marmosets, Bolivian gray titi monkeys, and pale-headed saki monkeys in approach-
ing trainers and learning behaviors. Marmosets, in comparison to most tamarins, had
longer attention spans. This meant that fewer, lengthier sessions were productive
whereas shorter, more frequent sessions were most successful for tamarins. Among
the cebids, pale-headed saki monkeys needed relatively few sessions to perform basic
and advanced behaviors whereas Bolivian gray titi monkeys were less responsive and
progressed at a deliberate pace. Marked changes in the animals’ behavior during daily
husbandry procedures, their voluntary participation in training activities, and the dis-
appearance of aggressive threats toward care staff indicated that training reduced
stress and improved the welfare of the animals. During daily training displays, zoo
visitors experienced interactive animals while learning the importance of low-stress
animal husbandry.
Despite their abundance in captive collections, to date relatively few operant condi-
tioning training programs involving callitrichid primates (marmosets, tamarins, and
Goeldi’s monkeys) have been developed. This may be due partly to their flighty na-
ture and small physical stature (Epple, 1975; McKinley, Buchanan-Smith, Basset,
Requests for reprints should be sent to Gina Savastano, Mammal Department, Bronx Zoo, Wildlife
Conservation Society, 2300 Southern Boulevard, Bronx, New York 10460. E-mail: gsavastano@
wcs.org

---

248
SAVASTANO, HANSON, MCCANN
& Morris, 2003/this issue; Mittermeier, Rylands, Coimbra-Filho, & Fonesca, 1988;
Rylands, 1993). Traditional methods of transporting or monitoring health status in
the callitrichids often required physical restraint, with concomitant stress, resulting
in animals becoming fearful or aggressive toward their caretakers (Brownie &
McCann, 2003; Farmerie, Neffer, & Vacco, 1999). In studies investigating indica-
tors of stress in marmosets, a significant increase in locomotor (Smith,
McGreer-Whitworth, & French, 1998) and scent-marking behaviors (Barros,
Mello, Huston, & Tomaz, 2001) was found when animals were presented with a neg-
ative stimulus. Additionally, in a comparative study of trained and untrained mar-
mosets, Bassett, Buchanan-Smith, McKinley, and Smith (2003/this issue) demon-
strated that stress imposed by invasive husbandry procedures was mitigated by
exposure to operant conditioning training, illustrating the positive benefits of apply-
ing operant conditioning training to captive husbandry techniques.
The ability to detect signs of illness, weight loss, and pregnancy, as well as
monitoring injuries, medicating and transporting individuals is essential for the ap-
propriate care and management of captive collections. The ability to conduct nec-
essary husbandry procedures in a low stress manner while building a positive
rapport with each individual should be a primary goal for all captive primate care-
takers (Colahan & Breder, 2003/this issue; Laule & Desmond, 1995; Reichard,
Shellabarger, & Laule, 1992).
The Bronx Zoo’s (BZ) New World primate collection includes 86 callitrichids
and small-bodied cebids, totaling 17 species, housed in three separate facilities.
The primary objective for developing a formal operant conditioning program for
the BZ’s New World primates was to decrease the level of stress involved in typi-
cal husbandry routines, and consequently, improve the welfare of the animals in
our collection. In addition to advancing basic husbandry protocols, positive rein-
forcement training has the added benefit of providing a stimulating, enriching, and
trusting environment for the animals (Laule, 1992). This in turn enhances the zoo
visitor experience by exhibiting animals who are engaged in their environment,
spend less time retreating to nestboxes and other hidden spaces, and can be viewed
actively participating in training sessions (Laule & Desmond, 1998).
In this article, we describe the development of a formal training program for a
large and diverse New World primate collection involving various care staff, mak-
ing note of important elements that formed the foundation of the program and the
results of the first year of the program.
METHODS
Study Animals and Housing Conditions
Eighty-six individual animals of 17 species of New World primates participated
in the first year of the training program (see Table 1). Animals are housed in so-

---

TABLE 1
Participating Study Animals Within the Training Program
Group Composition
No. of
No. of
Group
Group
Speciesa
Common Name
Males
Females
Classification
1
Callithrix jacchus
Common marmoset
1
4
Family
Pithecia pithecia
Pale-headed saki monkey
1
1
Breeding
2
Callithrix kuhlii
Wied’s tufted-eared
0
2
Single-sex female
marmoset
Callicebus
Bolivian gray titi monkey
2
3
Family
donacophilus
3
Callithrix kuhlii
Wied’s tufted-eared
0
2
Single-sex female
marmoset
4
Callithrix argentata
Silvery marmoset
2
1
Family
5
Callithrix argentata
Silvery marmoset
1
2
Family
6
Callithrix argentata
Silvery marmoset
3
3
Family
7
Callithrix pygmaea
Pygmy marmoset
2
0
Single-sex male
8
Callithrix geoffroyi
Geoffroy’s tufted-eared
0
2
Single-sex male
marmoset
9
Callithrix geoffroyi
Geoffroy’s tufted-eared
1
3
Family
marmoset
10
Saguinus mystax
Mustached tamarin
1
2
Single-sex female
11
Saguinus oedipus
Cotton-top tamarin
1
4
Family
12
Saguinus bicolor
Pied tamarin
1
1
Breeding
13
Saguinus geoffroyi
Geoffroy’s tamarin
3
3
Family
14
Saguinus geoffroyi
Geoffroy’s tamarin
2
1
Family
15
Saguinus geoffroyi
Geoffroy’s tamarin
1
1
Breeding
16
Saguinus midas
Golden-handed tamarin
0
3
Single-sex female
17
Saguinus midas
Golden-handed tamarin
1
1
Breeding
18
Saguinus midas
Golden-handed tamarin
1
1
Breeding
19
Saguinus imperator
Emperor tamarin
2
0
Single-sex male
Callicebus
Bolivian gray titi monkey
2
1
Family
donacophilus
20
Leontopithecus
Black lion tamarin
1
1
Breeding
chrysopygus
21
Leontopithecus
Golden lion tamarin
2
3
Family
rosalia
22
Leontopithecus
Golden lion tamarin
2
0
Single-sex male
rosalia
23
Leontopithecus
Golden-headed lion
2
0
Single-sex male
chrysomelas
tamarin
24
Callimico goeldii
Goeldi’s monkey
1
1
Breeding
25
Pithecia pithecia
Pale-headed saki monkey
1
1
Breeding
26
Pithecia pithecia
Pale-headed saki monkey
1
1
Breeding
Note.
Two species listed under the same group number indicate a mixed-species group.
aTaxonomy follows Groves (1993).
249

---

250
SAVASTANO, HANSON, MCCANN
cial groups, which we classify as either a breeding pair (one male and one fe-
male), family group (breeding pair with one or more offspring), or single-sex
group. In some cases, the primates are housed in mixed-species groups.
Animals in the training program are housed in indoor glass-fronted naturalistic
exhibits. These exhibits are viewable to the public from 1000h to 1600h daily. En-
closures vary in size but are approximately 2 m wide × 1.5 m deep × 3 m high. Exhibit
furnishings include natural branches, natural and artificial vines, a nestbox, plastic
plants, and a pine-bark mulch substrate over a concrete floor. Exhibit floors, glass,
and plants are spot-cleaned daily; the mulch substrate is removed and the enclosures
disinfected weekly. Most animals have access to an off-exhibit enclosure overnight,
measuring approximately 1.5 m wide × 1 m deep × 3 m high. These enclosures are
furnished with natural branches and enrichment items (puzzle feeders, foraging
boxes, and gum-arabic feeders) and are cleaned daily. The animals are fed twice
daily, in the morning between 0830h and 1000h and in the afternoon between 1400h
and 1600h. Training sessions are conducted in the animals’ exhibit spaces during
public viewing hours. Frequently, there are visitors observing the sessions.
Materials
Equipment utilized in the training program is listed in Table 2. The training
crates have mesh sides and two plexiglass guillotine doors: one on one end and
one side of the crate (see Figure 1). When in the crate, the animals receive their
food rewards through the mesh sides of the crate. Some crates are configured
with clips so that two crates can be attached along side each other with the side
doors lined up to each other. This set-up works well for larger groups as animals
that come into the crate can be locked into one side, leaving the other side open
for additional animals to enter.
Because of the callitrichids’ small physical size, food rewards and amounts
used in the training program were determined by consultation with the zoo’s nutri-
tionist. The most commonly used rewards include small pieces of banana or grape,
apple sauce, gum arabic, crickets, waxworms, and mealworms with amounts
equaling 10% of the caloric value of the total diet. Food given during training ses-
sions is removed from the animal’s daily diet to avoid over feeding and skewing
the recommended diet.
Behaviors
The behaviors that the animals are trained to do, and their defining criteria, are
listed in Table 3. Basic behaviors identified to be most important for animal
management include hand feeding, syringe feeding, targeting, scale and crate

---

TABLE 2
Equipment Used in Training Program
Item
Use
Clicker (Click and Treat™)
Bridge
Wooden dowel—1 cm diameter × 20 cm long
Target
Quick-draw Training Pouch™
To hold food rewards
1 cc plastic syringes
To dispense food rewards
Beaded pony-tail holders of different colors
Stations
Small plastic battery-operated scale, platform
To obtain weights on callitrichids
size 14cm × 14cm (Ohaus model LS200™)
Large metal battery-operated scale, platform
To obtain weights on callitrichids, titi and saki
size 30cm × 40cm (Weigh-tronix model
monkeys
QC3265™)
Wooden crates with two plexiglass doors and
For crate training callitrichids
mesh sides (30cm × 30cm × 40cm)
Wooden crates with two plexiglass doors and
For crate training titi and saki monkeys
mesh sides (40cm × 40cm × 50cm)
Aluminum platforms 40cm × 50cm, 1m high
For setting the large scale and training crates on
Plexiglass platform 14cm × 19cm, 1m high
For setting the small plastic scale on
Transponder reader (Avid Power Tracker IV™)
To detect identifying transponder microchips in
animals
FIGURE 1
Crate training with a family group of cotton-top tamarins (Saguinus oedipus).
Note the two crate set-up that is advantageous for training with larger groups (see text). (Photo
credit: Julie Larsen/Wildlife Conservation Society)
251

---

TABLE 3
Behaviors Trained in the Training Program
Behaviors
Verbal Cue
Visual Cue
Criteria
Basic behaviors
Hand feed
—
Food in hand
Animal takes the treat either in their hand
or mouth directly from the trainer’s hand
Syringe feed
—
Presence of syringe
Animal takes liquid from a syringe
Target
Target
Point to target or extend
Animal touches nose to the tip of the target
target to animal
and holds until released by bridge
Station
Station
Point to ponytail holder
Animal sits within one body length of their
specific colored ponytail holder
Scale
Scale
Point to scale
Animal sits on the scale and stays until
released by bridge
Crate
Box
Point to crate
Animal enters crate and waits while door is
closed
Transponder
—
Transponder reader
Animal stands on all fours while
read
transponder wand is passed along their
back and shoulder blades
Advanced behaviors
Up
Up
Index finger pointed up
Animal stands up on legs and holds until
released by bridge
Palpate
Belly
Index finger pointed up
Animal stands up on legs and holds while
trainer manipulates their hand along the
animal’s abdomen
Back
Back
Keeper holds own hand
Animal sits while trainer runs his/her hand
above animal’s back
down the length of the animal’s back
Tail
Tail
Keeper holds own hand
Animal sits while trainer runs his/her hand
above animal’s tail
down the length of the animal’s tail
Hand
Hand
Keeper holds own index
Animal places the appropriate hand on
finger sideways in
trainer’s finger (appropriate hand is
front of animal
determined by which side of the body
the trainer’s finger is on)
Stethoscope
—
Presence of stethoscope
Animal sits while stethoscope is placed on
their chest, abdomen, and back
Otoscope
Ear
Presence of otoscope
Animal sits while an otoscope is placed in
their ear
Ultrasound
Up/belly
Presence of ultrasound
Animal holds in an Up position on a t bar
equipment
while their abdomen is prepped with gel
and an ultrasound wand placed and
moved around on their abdomen
Note.
An em dash (—) = no verbal cue used.
252

---

PRIMATE TRAINING AT THE BRONX ZOO
253
training, and transponder reading. After a group has learned to perform all of the
basic behaviors, advanced behaviors that include tactile manipulations are intro-
duced (see Table 3). Behaviors are trained through positive reinforcement; the
animals receive rewards for performing desired behaviors, whereas undesired
behaviors are ignored. Standard operant conditioning techniques using clickers
as bridges and successive approximations are used (Laule, Bloomsmith, &
Schapiro, 2003/this issue; Pryor, 1999).
Program Organization
Six keepers form the core group of trainers in the program. Each trainer is
scheduled to work a minimum of 3 days each week with the New World primate
collection. Trainers are allotted two 30-min training sessions per day. Within
each session, approximately 5 min is spent on preparation, 10 min on training, 5
min on equipment removal and clean-up, and 10 min on record keeping.
Each trainer serves as the primary trainer for two or three groups of monkeys.
Primary trainers are responsible for introducing new behaviors to the animals (see
Figure 2). Once a behavior is consistently performed by an animal according to es-
tablished criteria, other keepers act as secondary trainers. The secondary trainers
assist in maintaining established behaviors and are available to work with the ani-
mals in the primary trainer’s absence. Each team member involved serves as a pri-
mary trainer on some groups as well as a secondary trainer on others. Daily records
FIGURE 2
A Geoffroy’s tamarin (Saguinus geoffroyi) takes a food treat from a syringe while
being palpated. (Photo credit: Julie Larsen/Wildlife Conservation Society)

---

254
SAVASTANO, HANSON, MCCANN
are kept and bi-monthly meetings are held to track training progress and to facili-
tate communication among the trainers and animal department managers.
RESULTS
The training program is ongoing and continues to develop as both the staff and
animals advance their skills. Here we present the results from the first year of
the program.
Logistical Challenges
Initiating and maintaining a training program for New World primates at the BZ
posed several challenges due to the size of the primate collection, the unavoidable
rotation of keeper staff throughout the various animal facilities, and the movement
of animals between facilities for exhibit and husbandry purposes. Maintaining train-
ing consistency between keepers and animals required significant attention. To ad-
dress this, uniform training criteria were created, written records of training sets
were kept, and weekly meetings were held for communication.
Animal Challenges
In addition to logistical challenges, animal challenges were, and always will be,
encountered. Table 4 lists some of the animal challenges that we encountered
and the techniques used to overcome them.
The trainers observed that the animals’responses to training varied greatly
among individuals, groups, and species (see Table 5). The results show a wide
range in the number of sessions conducted prior to a behavior being successfully
accomplished by most of the animals in the groups as well as differences among
species in which behaviors were learned successfully. Hand feeding from a keeper
took from 1 to 150 training sessions, syringe feeding from 1 to 10 sessions, target-
ing from 1 to 8 sessions, entering a crate from 1 to 20 sessions, going onto a scale
from 1 to 75 sessions, going to a color-coded station from 1 to 40 sessions, and al-
lowing their implanted microchip identification transponders to be read from 1 to 3
sessions. Ten of the groups responded particularly well to training and were taught
advanced behaviors (see Table 6).
DISCUSSION
Although training session length varies between species, the data in Tables 5
and 6 are useful in providing an indication of species differences in the time in-

---

PRIMATE TRAINING AT THE BRONX ZOO
255
TABLE 4
Animal Challenges Within the Training Program
Challenge
Solution
Food rewards: avoiding
Consult with the staff nutritionist to establish approved reward
obesity
items and quantities
Identifying animals and their
Maintain records describing physical characteristics of
motivations for training
individuals and their reward preferences
Training large groups
Establish control by training individuals to station
Timid individuals
Use a single trainer to develop a trust bond, introduce bolder, or
previously trained animals to the group
Overeager, dominant, or
Ask overeager animals to station at a distance so that the trainer
aggressive individuals
can focus on others; also, offer a time-consuming reward such
as nuts with shells, super mealworms, or whole grapes that
keep the overeager animal occupied
vestment required for training of basic and advanced behaviors. Although some
animals participated in the training program immediately and performed all of
the basic behaviors within five training sessions (e.g., Group 6, silvery marmo-
sets), others required months even to accept hand feeding (e.g., Group 21,
golden lion tamarins).
Tamarins
In general, lion tamarins were relatively slow to become comfortable with train-
ing. Building a rapport with each individual animal was a lengthy process, and
relatively long training sessions were needed. On the contrary, tamarins
(Saguinus spp.) responded more quickly than all marmosets and cebids with re-
gard to approaching trainers as well as learning behaviors. Individuals would be-
come engaged immediately as the trainer entered the enclosure and set up
materials, volunteering to begin the training session. However, Saguinus spp.
lost interest in sessions more rapidly than did marmosets or cebids. Shorter,
more frequent sessions throughout the day proved most productive for Saguinus
spp. Pied tamarins stopped responding to the trainer after just a few minutes into
the sessions. However, if the trainer left the enclosure and then re-entered a
short time (even < 1 min) later, they generally regained interest. Thus, a sched-
ule that provided up to 10 short training sets a day was the most productive for
this species.
Marmosets
In general, marmosets took longer than tamarins or cebids to begin interacting
with the trainer. However, once a bond was formed and a behavior established,

---

TABLE 5
Time Scale for Training Basic Behaviors
No. of Training Sessions for Animals to Perform Behavior
Hand
Syringe
Transponder
Group
Speciesa
Feed
Feed
Target
Station
Scale
Crate
Read
1
Callithrix jacchus
60
2
1
40
1
1
1
Pithecia pithecia
1
b
c
c
2
10
0
2
Callithrix kuhlii
20
5
4
c
3
7
2
Callicebus donacophilus
30
c
c
c
2d
10
c
3
Callithrix kuhlii
10
5
5
8
3
5
1
4
Callithrix argentata
1
1
2
b
1
3
c
5
Callithrix argentata
2
1
6
c
1d
2
3
6
Callithrix argentata
1
2
2
5
2
4
1
7
Callithrix pygmaea
90
c
c
c
2d
4
c
8
Callithrix geoffroyi
1
3
1
c
1
6
1
9
Callithrix geoffroyi
2
2
b
b
2
2
c
10
Saguinus mystax
5
2
1
c
1
5
c
11
Saguinus oedipus
5
3
c
c
10
10
c
12
Saguinus bicolor
1
1
8
c
1
1
1
13
Saguinus geoffroyi
4
1
1
c
1
1
1
14
Saguinus geoffroyi
1
10
1
c
7
8
c
15
Saguinus geoffroyi
20
b
b
c
9
c
c
16
Saguinus midas
2
1
c
c
4
4
c
17
Saguinus midas
1
1
1
2
1
1
1
18
Saguinus midas
1
2
1
c
1
1
1
19
Saguinus imperator
1
1
2
2
3
5
1
Callicebus donacophilus
20
c
c
c
3d
b
c
20
Leontopithecus chrysopygus
120
1
1
15
2
20
1
21
Leontopithecus rosalia
150
3
b
b
25
10
c
22
Leontopithecus rosalia
60
b
b
c
3d
b
c
23
Leontopithecus chrysomelas
15
c
c
c
3d
c
c
24
Callimico goeldii
50
b
b
b
75d
c
c
25
Pithecia pithecia
1
b
4
c
2
8
1
26
Pithecia pithecia
1
c
2
15
1
4
c
Note.
This table indicates the number of training sessions it took for most of the animals in the groups to
perform the behaviors.
aSee Table 1 for the common names. bBehavior is being trained. cTraining for this behavior has not been started.
dBehavior is done while trainer is outside enclosure.
256

---