4.4. Enforced Restraint

Restraint during clinical procedures and sample collection is a distressing experience not only for human primates (Figure 54a; Selekman and Snyder, 1996; Tomlinson, 2004; Folkes, 2005; Melhuish and Payne, 2006; Bland et al., 2007; Brenner, 2007) but also for nonhuman primates, who unlike humans are usually restrained with force without their consent (Figure 54b).

Published information provides scientific evidence that traditional, involuntary restraint techniques of research non-human primates are intrinsically a source of distress resulting from fear (Reinhardt et al., 1995, p 221). Research data collected from a distressed monkey are "distressed" and hence of little scientific value (Reinhardt, 1998, p 18). There is no scientific evidence that the animals adequately habituate to involuntary restraint (Reinhardt et al., 1995, p 221). Physical restraint procedures should be used on awake animals only after alternative procedures have been considered and found to be inadequate. If a restraint will be utilized the animal should be trained or conditioned to the restraining device, using positive reinforcement, prior to the beginning of the experiment (Prentice et al., 1986).


Figure 54a,b Restraint during unpleasant procedures can be a distressing experience for human primates (a) and for nonhuman primates alike (b).

4.4.1. Signs of Distress

Handling practices of primates traditionally bear two serious stressors for the individual subject:

  1. Being forcefully caught and removed from the home cage triggers behavioral distress responses and significant endocrinological and cardiovascular stress reactions (Mitchell and Gomber, 1976; Phoenix and Chambers, 1984; Herndon et al., 1984; Line et al., 1987; Reinhardt et al. 1990b; Line et al., 1991; Crockett et al., 1995; Jorgensen et al., 1998; Gerber et al., 2002; Davenport et al., 2007).
  2. Being forcefully restrained results in behavioral and emotional distress responses and significant hematological, endocrinological and cardiovascular stress reactions (Ives and Dack, 1956; Ackerley and Stones, 1969; Manning et al., 1969; Berendt and Williams, 1971; Quadri et al., 1978; Goosen et al., 1984; Golub and Anderson, 1986; Wheeler et al. 1990; Line et al., 1991; Brockway et al., 1993; Schnell and Wood, 1993; Fowler, 1995; Klein and Murray, 1995; Reinhardt and Reinhardt, 2001).

There are numerous scientific articles mentioning that caged primates can be trained or were trained to cooperate during common procedures in order to reduce or eliminate data-biasing stress reactions (Michael et al., 1974; Elvidge et al., 1976; Byrd, 1977; Rosenblum and Coulston, 1981; Herndon et al., 1984; Wall et al., 1985; Whitney and Wickings, 1987; Jaeckel, 1988; Suleman et al., 1988; Hein et al., 1989; Scallet et al., 1989; Chambers et al., 1992; Reichard and Shellaberger, 1992; Eaton et al., 1994; Hernándes-López et al., 1998; Hrapkiewicz et al., 1998; Nelms et al., 2001; Bentson et al., 2003; Grant and Doudet, 2003; Iliff et al., 2004; Koban et al., 2005). There are only a few reports describing and evaluating the techniques used to achieve the goal of such training.

4.4.2. Refinement Training to Cooperate during Injection and Venipuncture

Levison et al. (1964) developed a technique by which a large, aggressive male baboon was trained to offer his arm to receive an injection, rather than being forcefully chair-restrained during this routine procedure. The front wall of the baboon's cage contained a 9 cm-diameter porthole.

The training procedure was begun by holding a slice of fruit in front of the hole and giving it to the male when he extended his arm through the opening. Then, the fruit was given only when the arm was fully extended, and later, held quietly for a number of seconds.

On the trials that followed, the baboon was required to maintain this behavior while the experimenter touched his arm in a progression of closer approximation to drug injection. The baboon was given fruit after each satisfactory extension. Reinforcement was withheld if the wrong arm was extended, or if the arm was bent or withdrawn in any degree in response to tactile stimulation. The trainer:

  1. touched, and later held the baboon's wrist with his left hand
  2. touched the biceps with his right hand, and then with the syringe, while holding the animal's wrist firmly;
  3. made injection contacts in which the syringe and needle were placed against the arm and finally inserted into the muscle.

Only two training sessions were required before the needle could be held against the animal's biceps. Emotional displays and withdrawal of the arm occurred more frequently after the first penetration of the needle; however, the behavior was brought well under control when a special procedure for inserting the needle was begun.

The experimenter would press down progressively harder on the biceps muscle with the side of the needle, then slowly slide the point forward into the muscle while maintaining the lateral pressure. The point of the needle was not in contact with the skin until the forward move to insert it was made. After insertion, the needle was held in the muscle for successively longer periods; then, an actual injection was performed.


Figure 55 This, originally aggressive male baboon has been successfully trained to voluntarily present his arm for test drug injection in his home cage. Note that the male is not forced with a squeeze-back to tolerate the procedure. Photo by P.K. Levison; retouched by Annie Reinhardt; reproduced with permission from Levison PK, Fester CB, Nieman WH and Findley JD 1964 A method for training unrestrained primates to receive drug injection. Journal of the Experimental Analysis of Behavior 7: 253-254; Copyright 1964 by the Society for the Experimental Analysis of Behavior, Inc.

Satisfactory injection was reliably obtained after approximately three weeks of one hour-training sessions on alternate days. The baboon continued to cooperate when both active and control compounds were injected by two different researchers (Figure 55).

Priest (1990, 1991a) provides a detailed description and video-document about how he trained an adult single-caged diabetic drill (Mandrillus sp.) to cooperate during insulin injection and blood collection in the subject's home cage:

Because of Loon's medical condition, our first training priority was to condition him to accept his insulin injections voluntarily. This was begun in July, 1989, at the Zoo's veterinary hospital by hospital technicians. Necessary daily injections were being administered using a squeeze cage. By simply pairing a food reward with his daily injection, we began to establish the medical procedure as a positive event. In the early stages of conditioning, it was necessary to continue to use the squeeze cage to immobilize him. However, Loon quickly learned to recognize the injection as a precursor to food. By pairing his afternoon meal with an injection, while at the same time fading the use of the squeeze cage, the need for immobilization quickly became unnecessary. Within a few days, Loon learned to offer his back for the injection in anticipation of the reward. In addition to the food reward, Loon was being positively reinforced by the physical freedom made possible by his compliance.

Our [next] priority was to train him to allow venipuncture for blood sampling. Loon was trained to reach into a stainless steel tube, cut to the exact length of his arm, and to grasp a steel rod positioned crosswise at the end of the tube. As long as the drill was grasping the rod, he could not easily grab the trainer. Within three days of his exposure to a formal program of operant conditioning, Loon was grasping the rod and holding this position until a bridging stimulus (a clicker) was sounded, signaling termination of the behavior and presentation of a food reward.

Through an ellipse cut in the tube, I began to desensitize the drill to touch on his shaved forearm while he grasped the rod at the end of the tube. I began by reinforcing his allowing me to groom his arm and, on a separate command, his back. In addition to the social rewards baboons attach to grooming, Loon was also being rewarded with food items.

As training progressed I would occasionally drag different items over the bare skin of his forearm. This procedure desensitized him to a variety of stimuli, and simultaneously provided an occasion to reward him for grasping the rod.

During the first several weeks of training, Loon was very aggressive. He would snatch the food reward and, if I were not quick enough in removing my hand, take a swipe at me. On several occasions he succeeded in tearing the surgical glove off my hand. To reduce his aggression, we rewarded him with additional treats when he took the reward gently.
About six weeks into his training, Loon's medical condition required a blood sample. He was given the command to place his arm in the tube and grasp the rod. Within moments, a veterinarian had withdrawn the blood sample. Loon continued to wait patiently for the bridging stimulus to terminate rod-holding. The blood withdrawal had apparently been of no concern to him as he focused on holding the rod.

As a result of the need to test Loon's blood frequently, the veins in both of Loon's forearms have become heavily scarred. Loon has tolerated up to six failed attempts to draw blood from these battered vessels, without ever once pulling his arm away from the tube and rod. We responded to this new problem by training Loon to offer the vessels on the ventral side of both of his legs for venipuncture. Now venipuncture sites are rotated to help reduce damage to any single vessel site.

In nearly one year of training, Loon has never failed to voluntarily accept his insulin injection or to allow the veterinarians access to blood vessels in exchange for a good back scratch and a food reward (Priest, 1991b).

Figure 56 The "protected" blood collection training technique, originally developed by Priest (1990), is here successfully applied with a chimpanzee who is rewarded with fruit juice for cooperating during blood collection. Photo by Margaret Whittaker.

Laule and Whittaker (2001), Schapiro (2005) and Pranger et al. (2006) applied the same venipuncture training technique successfully with adult chimpanzees (Figure 56) living in pairs and small group, and adult individually housed rhesus macaques of unspecified gender. It took an average of:

McGinnis and Kraemer (1979) and Laule et al. (1996) used a less protective positive reinforcement training technique to obtain cooperation of adolescent female chimpanzees. While McGinnis and Kraemer (1979) document their success with a photo (Figure 57), Laule et (1996) describe their training technique:

Allie was nursery-raised and, hence, extremely tractable prior to the onset of the formal training, which initially implied that she had to sit upright and allow her arm to be manipulated and held by the trainer.
Next, she was desensitized to having her arm touched by, first, the trainer's finger, then a cotton swab, and then a syringe without a needle, with a blunt needle, and finally with a sharp needle. Throughout the process Allie was rewarded for being calm and for tolerating each stimulus for increasingly longer periods of time.

The first attempt to actually draw blood occurred during the 18th training session, with a total of 275 minutes of training time invested prior to that. The attempt was successful; Allie showed no visible signs of stress or discomfort, sat quietly, watched the entire procedure, and eagerly accepted rewards. During subsequent blood draws, she has never refused or disrupted the procedure.

Figure 57 Adolescent chimpanzee Joe is rewarded with apple juice for his cooperation during blood collection. Photo by P.R. McGinnis; reproduced with permission from Comfortable Quarters for Laboratory Animals (Seventh Edition), 20-27, Animal Welfare Institute, Washington, DC, 1979.

Reinhardt and Cowley (1992) worked with six adult female stump-tailed macaques who were pair-housed for more than one year in double-cages, each provided with a privacy panel, two perches, and one restraint mechanism. The animals were accustomed to being restrained with the squeeze-back for husbandry-related procedures. The door of the restraint compartment was equipped with a sliding transparent Plexiglas panel. Its opening allowed an animal to comfortably extend a leg out, yet was small enough to prevent the animal from protruding the head out of the cage (Figure 58a-d). The panel was also used as a safeguard for the person performing the venipuncture. An animal could be trapped by pulling the squeeze-back past the passage hole of the privacy panel (Figure 46a). The companion had free access to the rear portion of the squeeze-back, allowing visual contact.

The animals were used to having blood collected in a restraint apparatus away from their home cages. They were familiar with the authors who trained four and two of them, respectively. The training protocol comprised the following steps:

  1. The subject is enticed with favored food to enter the restraint compartment of the double-cage.
  2. By pulling the rods of the squeeze-back, the subject is restricted to the front quarter of the restraint compartment. This restricts her freedom of movement but still allows her to turn round and climb up the mesh walls of the cage. The animal is gently scratched through the mesh and food-rewarded.
  3. The spatially restricted subject is enticed with food to face the left or right side of the cage. Her back is gently scratched through the opening of the Plexiglas panel. This again is followed by a food reward.
  4. The subject's back and thighs are scratched. One of her legs is gently lifted and firmly pulled toward the opening of the panel. A food reward follows.
  5. The subject's leg is pulled through the opening of the panel and a blood sample taken by means of saphenous venipuncture. The procedure is again concluded with a food reward.
  6. Once the subject passively tolerates the above procedure with no signs of resistance, she is restrained in one third of the compartment rather than one quarter, thus allowing free movement. Venipuncture is carried out and the animal rewarded. This exercise is repeated on different occasions until the animal spontaneously cooperates (Figure 58a,b).
  7. Restrained in one third of the cage, the subject actively cooperates, i.e., voluntarily presents a leg behind, or through, the opening of the Plexiglas panel and accepts venipuncture (Figure c); this is followed by a food reward and praise (Figure 58d).



Figure 58a-d Adult stump-tailed macaque Zora has been trained to voluntarily present a leg for blood collection in her familiar home cage (a-c). She is praised with "Good Girl!" and rewarded with raisins, her favored treats (d).

All six stump-tailed macaques were successfully trained within a two-week period to actively cooperate during blood collection in their home cages. Nine to 23 training sessions per monkey were necessary to achieve this. Sessions were scheduled according to a subject's progress, although individuals were trained on no more than three occasions per day. The monkeys were trained with firm gentleness, but no sessions were terminated before the goal of the training step was achieved. The duration of individual sessions was therefore not constant but varied between 49 and 351 seconds. Animals who resisted the conditioning process (e.g., were unwilling to turn to one side, climbed up the cage wall to avoid having the leg grasped, struggled while having the leg pulled out of the cage) were never punished but treated with special patience.

Table 1. Cortisol response to voluntary blood collection and time investment to train pair-housed adult female stump-tailed macaques to cooperate during blood collection in their home cages.

On average, 16 minutes of training time was invested until the monkeys passively tolerated in-homecage venipuncture (steps 1-5) an additional 18 minutes were then required to ensure active cooperation during the procedure (steps 6 and 7). Total average training time was thus 34 minutes, ranging from 15 to 45 minutes (Table 1). It is sometimes argued that the training of nonhuman primates to cooperate during procedures has the disadvantage of requiring considerable time to be executed successfully (Klein and Murray, 1995; Hrapkiewicz et al., 1998). The investment of less than one hour per animal suggests that this does not hold true in all cases and, therefore, should not discourage qualified animal care personnel to train primates in their charge.

Once trained, the six stump-tailed macaques no longer displayed behavioral signs of distress during blood collection: They did not resist and struggle and they did not try to scratch or bite the handler in self-defense. In order to evaluate possible physiological stress reactions, serum cortisol concentrations were measured. For this purpose, two 0.5 ml blood samples were collected from each animal one week after the last training session, at 13:00 and at 13:15. The subjects were undisturbed by human activity for 90 minutes prior to the first venipuncture at 13:00. The first sample was used to assess basal cortisol concentrations and the second to assess the magnitude of cortisol response 15 minutes after venipuncture.

Basal serum cortisol concentrations from the samples taken at 13:00 were not significantly different from those taken 15 minutes later, indicating that the animals experienced no stress while cooperating during blood collection (Table 1).

Reinhardt (1991d) and Reinhardt (2003b) applied this training technique with ten pair-housed adult male and 12 pair-housed adult female rhesus macaques. These animals were also used to being restrained with squeeze-backs for husbandry-related procedures and having blood samples taken under enforced manual or mechanical immobilization in a designated treatment area. On average:

Figure 59a-d Rhesus macaque Max voluntarily present a leg for blood collection while his companion Ray attentively watches (a,b). Cooperation is always reinforced with "Good Boy!" and a food reward (c). Ray is also rewarded because he has not disrupted the handling procedure that took place with Max (d).

Average total training time for the females was the same as for the males, i.e., 39 minutes. This time investment does not seem unreasonably high when considering the long-term benefits of working with cooperative animals rather than against resisting animals.

There were no behavioral indications that the trained animals experienced apprehension or fear during the blood collection procedure; all males cooperated not only with the trainer, but also with the attending care personnel, as well as with experienced personnel from other facilities.

Empirical experience has shown that animals who have been successfully trained to cooperate during venipuncture require hardly any extra formal training to obtain their cooperation also during intra muscular injection (Figure 60a-c).



Figure 60a-c Rhesus macaque Rocky presents a leg and holds still during blood collection (a). She and also her cage mate Tora, who did not make any fuss during the procedure, are rewarded with grapes (b). Rocky required little extra formal training to cooperate also during injection (c).

Reinhardt (1992b) applied this training technique to six pair-housed juvenile (13 to 18 months old) female rhesus macaques. The training was successful in only one pair; the two juveniles required 46 and 47 minutes of training distributed over 38 and 37 sessions until they extended their legs through the cage opening for venipuncture. The training of the other four animals was discontinued after more than 40 sessions when it became clear that they were unduly distressed by being at such close quarters with a human "predator."

Stringfield and McNary (1998) successfully trained a high-strung, suspicious, cautious red-tailed moustached guenon (Cercopithecus cephus cephus) to accept daily insulin injection. David lived with two other monkeys. He was moved to a large squeeze-back cage to undergo training during two daily sessions. A clicker and a colored target were used, with food rewards being given for proper behavioral responses.

Within two months David was expert at stationing and putting his arm through the bars to touch the target. He never became comfortable having his arm held or manipulated, and would retreat when his arm was handled. However, when he would approach in a less formal manner, it became apparent that he liked to present by lying down with his back facing the trainer. He would then allow his back and other parts of the body to be scratched. Training was adjusted accordingly and rapidly progressed within another two months from scratching his back, to pinching his skin, to poking with a needle, to injecting a small volume of saline, and finally to injecting insulin.

Bayrakci (2003) developed a technique to achieve active cooperation during injection from three individually housed, adult male lion-tailed macaques (Macaca silenus).

  1. The first step in the training process was to help the monkeys recognize the clicker as an indicator of a correct response and an upcoming food reward. This was accomplished by calling the subject "come here!" and then click while saying "good," and finally offering a food reward. While the animal was sitting attentively in front of the trainer, the trainer continued to click and food-reward. It took only a few sessions for the macaques to expect a reward after hearing the click, and a few more sessions to adjust to this relationship with the trainer.
  2. Before starting injection training, a 5 cm-diameter hole was cut in the mesh wall 30 cm above the floor. Training sessions were conducted in front of this opening, so the macaques were comfortable sitting in front of it. The subject is shown a treat and rewarded for extending his arm outside the cage. The treat is then given through the mesh, not in front of the hole. The trainer quickly learned that if the macaque was rewarded through the hole, the arm extensions are too brief. When the subject reaches for the treat with one arm at a good distance away from the hole, it is easy to gradually increase the duration of the other arm's extension through the hole to allow enough time for an injection.
    The macaques were willing to extend their arms through the hole on command "Touch!" right from the beginning, so training sessions focused on increasing the length of arm outside the caging and the duration of that extension.
  3. Once arm extension was established, the trainer added a bamboo stick poised above and to one side of the hole, and began to gently press on the arm when fully extended. The macaques rapidly got used to the stick and the trainer began to press harder. The stick was then replaced with an empty syringe without a needle, then with the plastic needle tip, then with a long blunted needle. The clicker and "good" followed by a food reward was used to reinforce full arm extension beyond the moment when the syringe was pulled away.
  4. The trainer requested the arm extension behavior performed several times before injecting with a sharp needle. In the beginning, the animals reacted to this with a surprised squeak, but usually remained seated and were willing to continue extending their arms.
    For the first male, 50 training sessions distributed over 15 weeks passed before he cooperated during injection. This was a time investment of approximately five hours of actual training. The training progressed more rapidly with two other subjects. One of them reached the goal after 90 minutes, the other after four hours of training distributed over 18 sessions and 43 sessions, respectively.
    The three trained males did not show signs of fear or resistance during the injection procedure, and they all cooperated not only with the trainer but also with other personnel. Training to Cooperate during Sample Collection from Vascular Access Ports

Friscino et al. (2003) surgically instrumented three female and nine male rhesus macaques with biliary and venous catheters that could be accessed in a pouch located on the back of the subjects' jackets. The animals were then trained - using an unspecified positive reinforcement protocol - in their home cages to present the pouch and to remain stationary while the catheters were accessed. Three to four training sessions spread over a two-week period were required to achieve cooperation. The successful training precluded the need to subject the animals to enforced manual restraint or chair-restraint during sample collection. Training to Cooperate during Saliva Collection

Tiefenbacher et al. (2003) presented nine individually housed adult male squirrel monkeys in their home cages with a thin 10 cm long PVC pipe to which a braided cotton dental rope was attached on one end, and a plastic-coated cable - for retrieving the device - on the other end. The dental rope was flavored by soaking it in a solution of one part Kool-Aid®, one part sugar, and three parts water; it was then baked to dryness.

Seven of the nine monkeys readily acquired the task of chewing on the cotton rope for at least 30 seconds after which the device was retrieved and the subjects were rewarded with a food treat. The saliva obtained in this manner was sufficient to permit cortisol analysis by RIA (radio-immuno assay). Two monkeys required the addition of peanut butter and/or jelly to the dental rope to elicit sufficient chewing; only one monkey refused to cooperate in this saliva collection technique. Repeated saliva samples could be obtained reliably from the other eight animals.

This technique may also lend itself to the non-invasive assessment of other hormones and compounds in saliva. It was originally developed for rhesus macaque infants (Boyce et al., 1995) and adapted to adult rhesus macaques by Lutz et al. (2000) who found that 21 of 23 subjects cooperated, but only 16 (76 percent) produced saliva samples that were sufficiently large (0.4 ml) to allow cortisol analysis.

Cross et al. (2004) found in four adult male and five adult female marmosets (Callithrix sp.) that adequate saliva samples for RIA assessment of cortisol can be obtained reliably, without any extra training, by presenting the animals nine times for a cumulative total of approximately five minutes a cotton-wool bud coated with a thin layer of fresh banana. The animals spontaneously lick and chew on the bud. Many substances such as fruit-drink crystals, gum arabicum, honey, sugar water and crushed mealworm were tried as an alternative to banana to tempt the marmosets to lick and chew the cotton-wool buts, but banana was found to be the only substance that reliably encouraged chewing. Training to Cooperate during Semen Collection

Brown (1998) and Brown and Loskutoff (1998) document and describe how they trained three adult male gorillas (Gorilla gorilla gorilla), living together as a bachelor group, to cooperate during semen collection rather than subjecting the animals to electro-ejaculation under general anesthesia:

The gorillas were not forced into the training area nor did they have to cooperate with the trainer. The training area was an off-display holding cage with a 7.5 x 15-cm opening covered by a solid plate steel sliding door at ground level. The training was based on shaping behavioral responses with positive reinforcement using verbal and food rewards.

  1. The first behavior introduced was "Station." The trainer said "Station," the gorilla approached and took a treat with his lips. As training progressed, rewards were withheld until the gorilla approached, sat down directly in front of and facing the trainer, and accepted a treat in response to the "Station" prompt. All the remaining behaviors were taught with the gorilla in the "Station" position.
  2. The verbal prompt "Target" was used to associate an object with a desired response. The prompt was given while touching a ping pong paddle to the gorilla's fingers opportunistically, when the gorilla placed his hands on the wire mesh. Soon, the gorilla touched the paddle as a response to the "Target" prompt.
  3. The verbal prompt "Hold" was added to the "Target" behavior. When the gorilla touched the paddle through the mesh, the trainer said "Hold," while lifting the paddle off the mesh and moving it out of sight. The "Hold" behavior was shaped so the gorilla remained in the "Target" position until the trainer gave the bridge "Okay," while administering a reward.
  4. The cue "Knee" was shaped with the gorilla in the "Hold" position. The trainer passed a 70 cm long, 2.5 cm-diameter PVC pipe through the mesh and touched the knee when the verbal prompt was given. As training progressed, the gorilla moved the requested body part to the pipe. Eventually, he responded by moving the knee to the finger tips of the trainer's hand. This completed the shaping of the "Knee."
  5. Before semen collection was attempted each gorilla performed reliably the following "set up" procedure: "Station," "Target," "Hold," "Knee," and "Hold."
  6. Originally, an artificial vagina, constructed using a PVC pipe, was placed on the gorilla's penis. After numerous attempts, however, it was not tolerated by the gorillas nor did it stimulate ejaculation. It was decided that the trainer needed to reach through the small door with the left hand and stimulate the genital area directly. If the animal broke the "Hold" position, the trainer immediately withdrew, closed the door, and repeated the "set up" procedure. Eventually, with continuous administration of treats and repetition of the "Hold" prompt, the gorilla allowed penile massage periods long enough to result in ejaculation. As soon as the ejaculate was collected, the trainer's hand was withdrawn, the sliding door was shut, and verbal praise and treats were given to the gorilla.
  7. One of the goals of the training program was to provide a positive experience for the gorillas. Nevertheless, when they were unruly or uncooperative, two types of discipline were used. The first, the most common, was verbal. Verbal discipline included stating the gorilla's name and saying "No" in a low, strong voice. Never was the verbal reprimand shouted. The other type of discipline used was "time out," given when the verbal reprimand failed twice. In these instances, training ceased, and all personnel exited and remained silent and out of visual contact with the animal for 1-3 minutes. Upon return, training resumed normally. The gorilla usually cooperated with the trainer after having a "time out," but if he did not, the trainer continued to give prompts until the gorilla performed a requested behavior. The gorilla was immediately rewarded, and the training session ended on a positive note.

Training sessions were 10 to 20 minutes long on three days per week. The first semen samples of the three gorillas were obtained five, 12 and 14 months after starting the initial training. Training to Cooperate during Blood Pressure Measurement

Mitchell et al. (1980) trained three single-caged adult male and one adult female baboons to voluntarily submit to self-initiated blood pressure measurement in their home cages. Traditionally, blood pressure measurement involves considerable distress for the animals who first have to undergo surgery for arterial catheterization and are then chair-restrained against their will during data collection.

A cage-mounted oscillometric instrument with adjustable cuff assembly and banana-flavored pellet rewards was used for the training. Subjects were rewarded for extending their forearms into the cuff and depressing a lever to initiate measurement and maintain arm position throughout the blood pressure measurement sequence. Releasing the lever or withdrawing the arm too early caused immediate venting of cuff pressure and withholding of the reward.

Initially, the animals' tasks was simplified by mounting the lever directly against the front panel of the cage. This caused the lever to protrude slightly into the cage, where the subjects could reach it with minimal arm extension. In addition, the lever depression time required to earn one pellet was set at about 0.1 second and was then gradually increased to about 35 seconds. Only then was the cuff assembly installed and were the subjects rewarded for fully extending their forearms in the cuff and depressing the lever during a normal blood pressure determination.

All four baboons were trained successfully to cooperate during this procedure. The number of training sessions, which averaged 60 minutes each, ranged from 35 to 51 with a mean of 43.

Figure 61a,b Adult baboon Jim cooperates during manual auscultatory blood pressure measurement; note that there is no squeeze-back forcing the animal to sit still at the front of the cage! Jim voluntarily extends his arm and holds the post at the end of the shelf; the cuff is placed on the arm, the stethoscope on the brachial artery (a), and the cuff is inflated (b). Photo by Jaylan S. Turkkan.

Turkkan et al. (1989) and Turkkan (1990) trained ten adult male baboons to cooperate during blood pressure measurements in their home cages. Training occurred before the daily pellet ration was distributed to ensure that food rewards during training were salient reinforcements. The following training protocol was applied:

After an arm shelf with a 12-cm post at one end is attached to the subject's cage, the animal is rewarded with a food pellet for the following actions in progression:

  1. Extending an arm onto the shelf.
  2. Extending the left arm as far as the post.
  3. Touching the post.
  4. Grasping the post with the left hand.
  5. Holding the post for increasing durations.
  6. Allowing the arm that is holding the post to be touched.
  7. Allowing the arm to be stroked with the blood pressure cuff.
  8. Allowing the cuff to be placed briefly around the arm. At this stage, the cuff is opened and closed repeatedly so that the animal will habituate to the sound of the Velcro fastening and unfastening. At each step, food rewards are given freely.
  9. With the cuff in place, allowing the stethoscope to touch the extended arm (Figure 60a).
  10. With the cuff, stethoscope, and aneroid manometer in place, the trainer slowly inflates the cuff while delivering frequent food pellets (Figure 60b). It is important to keep the training session short so that aversion to the cuff inflation does not have time to develop. Most animals begin to pull on the blood pressure apparatus at this stage, and the trainer needs a quick hand to rescue all the paraphernalia before the animal can pull them into his cage. Also at this stage, training is facilitated by switching from food pellets to fresh fruit chunks, or apple sauce dispensed to a food nozzle by means of an infusion pump. The apple sauce has the added advantage of providing immediate termination of a continuous stream of reinforcement when inappropriate behavior such as arm withdrawal occurs.
  11. The rate and degree of cuff inflation is progressively increased over successive sessions, with termination of apple sauce reinforcement for arm withdrawal, which occurred frequently at this stage.
  12. An episode of uncooperativeness must never be allowed to end a training session, because then the animal quickly learns to avoid discomfort by acting aggressively. When aggressive acts such as scratching the trainer occur, the reinforcement is withheld, and the training resumed after a few minutes.
  13. Once the baboon accepts full cuff inflation, the cuff is deflated slowly. The animal is rewarded for sitting through a period of non-reward while the trainer attends to blood pressure measurement.
  14. After completion of the final measurement the baboon is rewarded with fresh fruit.

The duration of training until the first systolic and diastolic blood pressure measurements were obtained was an average 12 weeks (range 2 to 36 weeks). Systolic and diastolic blood pressure measurements of a trained baboon required approximately five minutes, which included set-up of the shelf and food reward delivery system. Training to Cooperate during Oral Drug Administration

Oral drugs are traditionally delivered via gavage, which is one of the most distressing procedures to which nonhuman primates are subjected.

Turkkan et al. (1989) habituated 11 adult baboons of unspecified gender to voluntarily drink a bitter-tasting solution of quinine which could mask the taste of various test drugs. Initially, the subjects were offered 100 ml of an orange-flavored juice that they all drank avidly. Over daily sessions, increasing amounts of quinine sulfate were added to this orange drink until a concentration of 0.325 mg/ml was reached. It was then possible to add test drugs and to obtain complete dose-effect curves with a number of benzodiazepines, barbiturates and other sedative/anxiolytic drugs. Unfortunately, the authors do not indicate how much time was invested to successfully habituate the baboons to drink the quinine solution.

Baumans et al. (2007) quote a report on vervet monkeys (Cercopithecus aethiops) who voluntarily swallow drugs when these are mixed with the animals' regular diet, consisting of pre-cooked maize fortified with vitamins and minerals. The dry ingredients are blended with water and form a stiff putty-like paste, which is an ideal vehicle for mixing in test substances. If the flavor needs to be masked, there are a variety of possibilities, such as honey and syrup, depending on what the protocol permits:

We usually administer the compound in about a third of the morning feed. The bulk of the food is offered after this portion has been consumed. Some substances we even mix into the entire bulk of the morning feed. Keeping the compound too long in cheek pouches or spitting it out has never been a problem. We have used this simple oral administration technique for pharmacokinetic studies very successfully. Over a time period of 20 years, we have not had to deal with any substance that we could not feed to the vervets, including bitter herbal mixtures in fairly high concentrations. Training to Cooperate during Topical Drug Application

Reinhardt and Cowley (1990) trained adult stump-tailed macaques to actively cooperate during drug application on their foreheads in the home cages. The animals were used to being removed from their cages and subjected to enforced mechanical restraint during this procedure in a treatment area.

There were 17 males and three females living in ten compatible pairs in double-cages equipped with sturdy, replaceable plastic plates that fitted into the cage door openings. Each plate had a face-shaped hole fitting the head of an animal and two smaller circular holes fitting the forearms. The arrangement of the holes was such that an animal could reach out for raisins and eat them while presenting his or her forehead (Figure 62). For the treatment the pairs were temporarily separated by means of a cage divider so that one partner could be treated without the other interfering. The animals required one to 14 training sessions, each lasting one to five minutes, to present their foreheads and allow topical drug application while retrieving raisins from the handler's hand.

Figure 62 Stump-tailed macaque Stan cooperates during topical drug application in his home cage. Pole-and-Collar-and-Chair Training

It has been repeatedly stated that monkeys can be trained to voluntarily cooperate in their home cages to have a pole or leash attached to a collar and allow themselves to be subsequently guided to and securely placed in restraint chairs (Barrow et al., 1966; Nahon, 1968; Anderson and Houghton, 1983; Schmidt et al., 1989b; McCully and Godwin, 1992; Klein and Murray, 1995; Marks et al., 2000; Sauceda and Schmidt, 2000; Scott et al., 2002; Down et al., 2005). This claim is supported with data in only one case.

Skoumbourdis EK (2008) has trained adult and juvenile rhesus macaques and adult long-tailed macaques to cooperate during the capture with the pole and the transfer to and placement in the restraint chair:

All the monkeys I have pole/collar/chair trained have gone through an initial phase of resistance both when the pole was being attached to the collar, and when they were first put into the chair, but for the most part they finally did settle down and cooperate. All it takes is patience and gentle determination on the part of the trainer.

I always collar my animals at least a week or two before the first training session so they get used to wearing the collar. If they're not comfortable with the collar, it really sets the training back because they will spend most of their time pulling at the collar and scratching at their neck.

To start the training, I first make sure that the trainee is comfortable enough with me that he/she is willing to take treats from my hand. I subsequently include the pole, offering treats with one hand, while holding the pole close to the cage in the other. The animals readily get used to this little ceremony and soon seem to ignore the pole, but focuse more on the treats.

The poles come with that handy little clip that opens and closes for collar attachment. The clip is a great place to hook treats, which the monkey has to retrieve directly from the "dreaded pole." I like to stuff a marshmallow tightly into the clip. This makes it a little harder for the animal to get the treat, and extends the time the animal is in contact with the pole. Once the monkey retrieves treats consistently, without signs of apprehension or fear, I start moving the un-baited pole very carefully in the cage, and finally, also touch the animal with it. In subsequent sessions, I gently tap the collar with the pole. When the training session is over, I hang the pole outside on the front of the cage so that the animal gets more and more acquainted with it. Needless to say that I always distribute extra rewards - jackpot if it's deserved! - before I leave the room.

I have trained animals living in both, cages equipped with squeeze-backs and those without. If the animal's cage has a squeeze-back, I use it only with the tougher customers. But, generally, I try to avoid using it so that the trainee is always in control of the situation. I believe this greatly helps the animals to stay relaxed, continue to trust me, and learn quickly what is expected from them in each training session. I also consistently reward cooperation with a treat and verbal praise. If the animal doesn't cooperate, patience from my part replaces the reward. I feel that this strategy helps to create a tension-free ambience for the monkey and for the trainer.

The first few times the pole is actually attached to the collar can be quite dramatic. The trainees usually "freak out" the moment they realize what is happening to them. However, there is no reason for panic. I simply leave the pole attached, maintain a firm grip, and talk reassuringly to the animal who will gradually calm down, stop squirming, and remain quiet long enough so that I carefully unhook and remove the pole. This interaction is always followed by a generous treat reward which, in my experience, is never refused.

During the next sessions, I get the trainee to sit still with the pole attached to the collar for progressively extended periods of time, until he/she "forgets" about the pole and takes treats from me. I repeat this step several times. Some animals adjust better to this situation than others, but they all end up remaining reasonably still with the pole attached to the collar.

Coaxing the poled monkey to come out of the cage is always a big challenge. After all, the familiar home cage is a relatively safe haven for these animals. With patience, and many reassuring words, the trainee does finally stop resisting, follows the pull of the pole, and comes out of the cage. Should the animal begin to thrash about once outside of the cage, I take the pole and carefully, but firmly, push the animal's head to the floor. To be clear, I do not throw him/her down but rather use the pole to turn the collar up towards the animal's head and then apply some forward and downward pressure in a determined manner. The monkey is now fixed and can get his/her bearings while remaining safe from causing himself or herself any serious harm. I have noticed over and over again that you can help the animal to calm down when you speak to him/her reassuringly with a gentle whisper-like voice. When the animal has settled down, I carefully start to walk him or her again; I will drop a few treats on the floor for the animal to pick up as he/she moves along the floor. After a few sessions, most trainees will feel confident enough to walk, rather than struggle, on the pole. If a monkey continues to resist after two or three sessions, I'll call in reinforcements. Most collars have two sides where a pole can be attached. By adding a second pole, directed by a second person, the animal is easier to guide in a forward motion.

I've found that it takes about one week of training until a monkey will cooperate and walk on the pole in a reasonably calm manner and pick up treats from the floor as a reward for good behavior. My goal is to get the trainees to walk, because after they come out of their cages they have a lot of pent-up energy that they like to release; especially the younger animals. I treat this solely as a reward for good behavior. If the poled animals walk calmly, I let them do so for a few minutes, but if they start playing "super man," I pull them straight back into their cages. If you don't have enough space, or the racks are enticingly close for climbing and rattling, or if you are a little new at this and do not have a second person around who can help you control the monkey if need arises, the pole walking isn't a good idea.

Now, onto the chair:

  1. Push the chair up against a wall, with the entrance facing out, and put all the brakes on. This keeps the chair stable and makes it impossible for the monkey to walk straight through - a situation that isn't any fun when you're on the other end of the pole!
  2. Allow the monkey to explore the chair, touch it, climb it, walk around it, and perhaps retrieve a treat or two that you have placed somewhere on the chair.
  3. After a day or so, coax the monkey into the sitting position in the chair. Do this by gently lifting the animal's neck into position and get the collar into place. If another person, who is also on very good terms with the trainee, can help you, the situation becomes less of a challenge, especially when you are dealing with a strong and extremely stubborn monkey. Once you have your monkey in place, let him/her adjust for a few minutes. Don't forget the treats! Some animals will be initially restless and try to push your hand away; but with gentle patience they will settle down and finally accept the food reward.
  4. Gradually extend the time the trainee remains in the chair over the next few days. Always be sure to remain close by to serve as a comforting social support. Should the animal show any signs of discomfort, try giving him/her further treat rewards. If he or she continues to be restless, abort the training session; you do not want the animal to relate the chair with discomfort and/or distress.

I have found that each "big step" involves an initial struggle, but I have also found that with consistency and patience, the animals learn quite quickly what I expect them to do. I have had several animals who were fully trained and just came up to the front of the cage without being squeezed. They actually presented their necks so that their collar loop was exposed for me to attach the hook of the pole. All of these monkeys struggled a great deal when I first started working with them. It is amazing how these animals gradually relax into the training sessions and finally start working with you, rather than against you.

Trust in the trainer is the ultimate key for success. Nonhuman primates are intelligent; when they are free of apprehension or fear, they quickly figure out that it is much easier and even rewarding for them to cooperate with you rather than resist. A successfully trained monkey will have developed so much trust in you that he/she will never fight against you when you pole and chair him/her.

When I train animals, I work with them once or twice daily, five days a week -with additional weekend sessions if needed - until the goal of the training has been achieved. I have found that if I don't work with them on a consistent schedule, they tend to get "rusty" rather quickly. The faster you can get them over the initial struggling, the easier the whole training sequence. If you try to pole a monkey who vigorously resists on a Monday, and decide to wait and try again on Friday, chances are that the struggle will be the same, if not worse. However, if you are persistent and repeat the training step over and over again every day, you will definitely notice progress by the end of the week. I imagine that without consistency and patience, the training would be a rather frustrating experience, both for the trainer and for the trainee.

To pole-collar-chair train a monkey can be a very rewarding process that is not necessarily time-consuming. I have successfully trained 19 animals:
two adult female rhesus,
four adult male rhesus,
five juvenile male rhesus,
four adult female cynomolgus, and
four adult male cynomolgus.

My quickest subject took just five days of training to reliably cooperate (I should mention that he was two years old and an angel!), while other animals have taken me well over a month to get going - especially older rhesus who can be very stubborn and hard to food-motivate. Also, I have had some animals who were just never meant to be put in a chair. This is a reality that both you and the investigators must acknowledge. You cannot force a monkey to cooperate and be relaxed in the chair. It's impossible. Sure, you can try, but you're not going to win. Training to Cooperate for Weighing

McKinley et al. (2003) trained six heterosexual pairs of marmosets to cooperate for weighing in the animals' home cages rather than being caught by gloved hands and transferred to a small cage to be weighed:

  1. The target (a plastic spoon) was held at the front of the cage with the food reward (marshmallow, cornflakes or chopped dates) held behind it. Males were offered a black target placed on the left-hand side and females a white target placed on the right. A reward was given when the correct target was touched. Incorrect responses were ignored.
  2. The target was presented without the reward held behind it. The animals were rewarded when they touched the target.
  3. The time the target had to be held before the reward was given was gradually increased.
  4. Scales for weighing were placed in the cage and the target held in front of them. The marmoset was rewarded for climbing onto the scales and holding the target while her or his body weight was recorded (Figure 63).

Figure 63 This target-trained female marmoset sits on scales in the familiar home cage while her body weight is recorded; her partner waits until the target is presented to him. Photo by Jean McKinley; reproduced with permission from Journal of Applied Animal Welfare Science 6(3), 209-220, 2003.

The cumulative time per animal to achieve the goal of the training ranged from 20 to 120 minutes with a mean of 64 minutes. The time investment for successful training did not differ between females and males. Training to Cooperate for Capture

Traditionally, mechanical force (movable squeeze-back), threats (display of net) and vocal intimidation are used to overcome the reluctance of primates to leave their familiar home cages while these are sanitized or for routine procedures such as weighing. It has been reported that monkeys can be trained to voluntarily exit into transfer boxes (Clarke et al., 1988; Heath, 1989; Sainsbury et al., 1990; Reinhardt, 1992c; Erkert, 1999; White et al., 2000; Coke et al., 2007); detailed training protocols have yet to be published.

Figure 64a,b Paired rhesus macaques entering a transfer box - one at a time - on vocal commands (Reinhardt, 1992c). Note that the animals are not forced with a squeeze-back or a stick to leave their cage. 


Table of Contents