Social learning during human-animal interaction: Effects on broiler chickens' behavior

Description

This study aimed to test whether broiler chickens can learn to trust humans through social learning by observing a conspecific receiving gentle handling. Our research hypothesis was that observer chicks exposed to a demonstrator chick receiving positive human interactions would exhibit reduced fear and increased approach behaviors toward humans. Data were collected from 72 one-day-old male Ross 308 broiler chickens, randomly assigned to three treatments: Gentle Handling, Human Presence, and Control. Each treatment had two pens of 12 chicks each. The experiment spanned from day 1 to day 32 of age. Observer chicks in the Gentle Handling group viewed a demonstrator chick being gently handled by an experimenter twice daily for 15 sessions. The Human Presence group was exposed to the experimenter without interaction, while the Control group received only routine human contact. Behavioral responses toward the experimenter were assessed through a home-pen test on days 31 and 32. Behavioral measures included time spent near the experimenter, latency to approach, and behaviors such as preening, resting, and alertness. Linear mixed models revealed that chicks from the Gentle Handling group approached the experimenter more quickly, spent more time nearby, and displayed less alert behavior than those in the Human Presence and Control groups. Logistic regression indicated a higher probability of entering the proximal zone for Gentle Handling chicks. These results support the hypothesis that broiler chicks can learn socially from conspecifics and that observing positive human-animal interactions reduces fear. The study suggests a practical welfare approach for poultry management through social learning.

Steps to reproduce

1. Animals and Housing: We obtained 72 one-day-old Ross 308 broiler chicks and housed them in six pens (2.25m x 1.20m), with 12 chicks per pen. Chicks had ad libitum access to commercial feed and water. Temperature and lighting were controlled (23L:1D from days 1–3, 20L:4D from days 4–6, and 18L:6D thereafter). 2. Treatments: We randomly assigned two pens per treatment (Gentle Handling, Human Presence, Control) using R software. We conducted 15 sessions (twice daily, 10 minutes each) from days 7 to 30: o Gentle Handling: Observer chicks viewed a demonstrator chick receiving positive interactions (eye contact, soft speech, and stroking) through a wire mesh divider. o Human Presence: The experimenter sat quietly without interaction. o Control: Chicks experienced only routine husbandry. 3. Home-Pen Test (Days 31–32): We marked chicks with animal-safe spray for identification and recorded their behavior for 5 minutes using an overhead camera. Videos were analyzed with BORIS software via continuous focal sampling. We collected data on proximity to the experimenter and behaviors such as latency to approach, preening, and alertness. 4. Data Analysis: We analyzed behavioral data using R software. Linear mixed models were applied for continuous variables, with pen as a random effect. Logistic regression was used for binary outcomes (e.g., entering the proximal zone), also including pen as a random effect. We performed post hoc Tukey's HSD tests for pairwise comparisons and applied transformations to ensure normality where needed.

Institutions

Categories

Funding

Licence