Exposure Therapy Trials

Summary: This describes a pilot study of the effects of high intensity exposure therapy (flooding) on the taming of two anxious donkeys. We used latency to hand-feed as a measure of tameness before and after petting each donkey in a chute for five minutes. The results are limited by a small sample size but suggest that this protocol may be fruitful and does not foster suppression.

Background: There are four goals for taming an animal: 1) to calm the animals; 2) to reduce their flight distance; 3) to make them realize they need not fear humans; and 4) to let them become familiar with their trainer (Conroy & Barney, 1986). Each of these goals reflects a lower state of anxiety about humans.

Some individual equines are extremely anxious, interfering with their ability to respond to the opportunities for reinforcement. Such inability to respond is called suppression. A debilitating fearful pattern of behavior may represent either a learned (but now habitual) avoidance response or a innate predisposition to extreme anxiety (Singewald & Holmes, 2019). Avoidance behaviors are quickly learned and easily reinforced. Predisposing factors could include such things as diet, stress, genetics, brain damage, or malfunctions of the neural network. These two sources of anxiety can be combined, posing an even greater challenge to reducing fear.

Traditional methods of taming donkeys commonly involve touching the animal until it loses its fear of the trainer. This method is considered "flooding" in the academic animal training community and is generally considered to be problematic and therefore to be avoided. Animal trainers tend to rely on systematic desensitization, counter-control, or gradual habituation instead of flooding.

Studies of fear disorders in humans can provide animal trainers a continually developing paradigm to adapt to animal training. Intolerance of uncertainty is a concept from human studies that may also apply to animals. An individual with intolerance of uncertainty is less able to discriminate threat from safety cues and tends to interpret any stimulus pessimistically as a threat (Milne et.al, 2019). This matches our experience of trying to tame animals with debilitating fear levels. For them, every stimulus they contact is a threat. The human studies show that variable exposure therapy, which builds new associations between stimuli and a lack of danger, is the best treatment for intolerance of uncertainty (Knowles & Olatunji, 2019; Craskey et. al, 2019). At it's most intense application, exposure therapy is flooding.

We had previously used flooding successfully to tame burros. We wanted to document the process and demonstrate it's effectiveness. This study was a pilot project to evaluate latency to feed as a measure of fearfulness and to look for initial indications that flooding alone would impact their tameness or welfare.

Methods: The burros in this study had been captured by the Bureau of Land Management (BLM) from Havasu, Arizona in 2014. They were provided to us from a BLM holding facility in Florence, Arizona through the Trainer Incentive Program of the Mustang Heritage Foundation on 5/2/2016. Sangria (SANG) was estimated by the BLM to be 18 years old, while Chardonay (CHAR) was estimated to be 12. During their stay at the BLM holding facility, they were branded and vaccinated. Once at our facility, trainers had unsuccessfully attempted to get them to hand feed at a rate of ten bites per minute.

For this experiment, they were individually isolated in a familiar pen adjacent to a 6 ft. high iron panel chute with bars about 12 inches apart. Each day from 5/12 to 5/14 (3 days) the donkeys were subjected to the flooding intervention and the pre- and post- tests, but no other training was done with them.

For the pre-test and post-tests, two trainers took turns walking into the pen, standing at the middle of the west wall, and offering food by extending a hand toward the donkey. The food used was Equine Senior from Purina. The offer of food was timed, and if after 1 minute the donkey had not taken it, the hand was withdrawn and the trainer left the pen. Ten offers were made in each pre-test and post-test.

After the pre-test, the donkey was confined in the chute by following the donkey until it walked into the gateway, then using the gate to close it in. The subsequent intervention protocol was slightly modified after the first donkey. A dish of chopped alfalfa was placed on the ground at the end of the chute and the first donkey was petted through the bars by trainers on each side. When the donkey stepped forward and started eating the alfalfa, the gate was opened and the donkey was released. This first donkey (SANG) started eating after 1:36 minutes of petting and was released. The protocol was then revised to have 5 minutes of petting before the chopped alfalfa was placed in the chute, then more petting until the donkey started to feed, at which time it was released.

To determine whether the intervention (or other factor) caused a change in the latency to take food, the median of each set of 10 trials was used in a Sign Test, the non-parametric test of differences between medians of related samples.

Analysis: A decrease in latency was taken to indicate less anxiety in the animal. The reduction of anxiety may be a result of the intervention, the animal becoming accustomed to the procedure, or approaching the trainer being reinforced in a controlled environment.

Discussion: Null results in this pilot test do not discourage us from this approach. If the animals were more anxious after the intervention, that anxiety did not persist past five trials. For one of the donkeys, the intervention was always followed by consistently low latency scores. This argues against the idea that flooding causes more suppression in the short-term. The animals were quicker to feed on subsequent days, suggesting that flooding did not create suppression associated with long-term trauma either.

This study show that this approach may have value for diminishing habitual fear and does not compromise the welfare of fearful mature donkeys.

References

Conroy, D., & Barney, D. (1986). The oxen handbook. Butler Pub & Tools.

Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: An inhibitory learning approach. Behaviour research and therapy, 58, 10-23.

Knowles, K. A., & Olatunji, B. O. (2019). Enhancing Inhibitory Learning: The Utility of Variability in Exposure. Cognitive and behavioral practice, 26(1), 186–200.

Milne, S., Lomax, C., & Freeston, M. H. (2019). A review of the relationship between intolerance of uncertainty and threat appraisal in anxiety. the Cognitive Behaviour Therapist, 12.

Singewald, N., & Holmes, A. (2019). Rodent models of impaired fear extinction. Psychopharmacology, 236(1), 21-32.

Acknowledgements: Thank you to Dr. Melissa Shyan-Norwalt for assisting in this project and Dr. Eduardo J. Fernadez for reviewing and improving this report.