Fagen et al - Elephant Training Study

Psychology being Investigated

• Operant Conditioning: Used positive reinforcement to encourage specific trunk movements in elephants.

  • Positive Reinforcement: Rewarding correct behavior to increase its frequency.

  • Negative Reinforcement: Increasing the frequency of behaviour by removing something negative.

  • Positive Punishment: Adding a negative thing to decrease frequency of a behaviour.

  • Negative Punishment: Removing something good in order to decrease the frequency of a behaviour.

• Reinforcers in the Study:

  • Primary Reinforcer: Chopped bananas, which fulfill the elephants’ hunger (direct need).

  • Secondary Reinforcer: A whistle blast paired with bananas. The whistle itself becomes a reinforcer through classical conditioning (association with food).

Background of the Study

• Researcher: Dr. Ariel Fagan, veterinary behaviourist at the American College of Veterinary Behaviour.

• Context: Captive elephants need to be tested for tuberculosis (TB).

• Goal: Reduce stress by training elephants to voluntarily participate in trunk wash procedures, instead of forcing them, for TB testing.

Aim of the Study

• To investigate if secondary positive reinforcement can train elephants to voluntarily complete a trunk wash for TB testing.

Sample

• 5 Female Elephants:

  • 4 juveniles (ages 5-7) and 1 adult (50+ years old).

  • All elephants were born in the stable, tame, and had no previous experience with secondary positive reinforcement.

  • Elephants were not pregnant.

• Mahouts:

  • Present for safety, but not involved in the training (no interaction).

  • Ensured the safety of trainers and researchers by controlling potential elephant behavior risks.

Research Method and Design

• Controlled Observation: Structured setting with a checklist to observe and record behavior.

• Structured Observation: Use of a predefined behavioral checklist to evaluate performance.

• Design: Repeated Measures Design – Elephants undergo multiple training sessions, and each is tested on different stages of the sequence.

• Independent Variable: The cues/offers made by trainers to prompt behaviors.

• Dependent Variables:

  • Training duration (time from first cue to last cue).

  • Number of cues given by trainers.

  • Success rate of each behavior/sequence.

  • Grading of behavior: 80% success rate required for passing.

Behaviour Grading

• Grading Individual Behaviours: Each behaviour in the trunk wash sequence (e.g., trunk in hand, trunk up, etc.) was graded on success.

  • Success Criterion: 80% correct behaviour in at least 8 out of 10 trials.

  • If a behaviour sequence failed, it was broken into smaller, easier parts and re-tested.

  • Full Sequence Goal: Elephants needed to achieve 80% success in the full five-step trunk wash.

Procedure of the Study

• Training Schedule: Conducted in morning (7:30–10:00 AM) and afternoon (4:00–7:00 PM) sessions.

  • Elephants were chained in their stalls during training.

• Trunk Wash Procedure (5-step sequence):

  1. Trunk here: Elephant places trunk into trainer’s hand (to receive saline).

  2. Trunk up: Elephant lifts trunk and holds fluid inside.

  3. Bucket: Elephant lowers trunk into a bucket.

  4. Blow: Elephant blows fluid into the bucket.

  5. Steady: Elephant holds trunk steady before relaxing.

• Training Techniques:

  • Capturing: Encouraged natural behaviours, like lifting the trunk, and reinforced with bananas.

  • Luring: Used treats to guide elephants to perform behaviours that were not naturally occurring (e.g., stretching trunk out).

  • Shaping: Reinforced small approximations of the final behaviour and gradually required more specific actions.

  • Secondary Reinforcers: Whistle blasts paired with bananas to build association between whistle and food.

• Syringe Desensitization: After trunk-wash behaviour was learned, elephants were gradually introduced to a syringe (aversive stimulus) by using counterconditioning to associate it with bananas.

• Steps for Syringe Exposure:

  • Syringe touched the outside of the trunk.

  • Syringe placed inside the trunk.

  • Gradually increased fluid injection from 1-15ml to 60ml.

Results of the Study

• Training Duration:

  • Elephants required 25-35 training sessions to learn the full trunk wash.

  • Average session length: 12 minutes (range: 10–13 minutes).

  • Total Training Time: 367 minutes on average (range: 257–451 minutes).

• Elephant 5:

  • Did not learn the full trunk wash sequence and was excluded from the final results.

  • Issues: Elephant had an abscess on her foot and impaired vision.

• Learning Progress:

  • Success Rate for Sequences: Increased from 39% after 10 sessions to 89.3% after 35 sessions.

  • Some behaviors took longer to master (e.g., ‘trunk in hand’ required 295 offers on average, while ‘blow into bucket’ took 54 offers).

• Key Behavior Challenges:

  • The "steady trunk" task was the most difficult, with an average success rate of only 70% for that behavior.

  • This was largely due to difficulty with the ‘trunk-down’ task.

Conclusion

• Effectiveness of Secondary Positive Reinforcement:

  • Secondary positive reinforcement can effectively train juvenile elephants to voluntarily participate in a trunk wash, reducing stress during TB testing.

Evaluation of the Study

• Strengths

  • Ethics: Guidelines of consent, psychological harm, & safety were followed as elephants were allowed to refuse participation by walking away.

  • Ecological Validity: Training was conducted in the elephants' natural environment (stables), improving the generalisability of the results to other environments (e.g., zoos, safari parks).

• Weaknesses

  • Internal validity: There could have been subjective observer bias (since observation wasn’t backed up by a video camera or objective proof).

  • Reliability: Since the amount of times Elephant 1 was tested had been reduced, this means standardised procedures were not followed.

Issues & Debates (Individual vs Situational)

• Individual Explanation: Elephant 5's failure to learn might be due to individual factors (older age, physical issues such as abscesses and impaired vision).

• Situational Explanation: Environmental factors (e.g., distractions like a young calf in the stall) may have also contributed to slow learning).

Real Life Applications

• Veterinary Procedures: The study shows that elephants can be trained to voluntarily participate in critical veterinary procedures like trunk washes for TB testing, reducing the need for restraints or sedation.

• Wider Implications: Techniques from this study can be applied to other behaviors like blood sampling and x-rays, helping reduce stress and improve animal welfare in captive settings.