Acoustic markers of negative arousal in lambs: evidence from behavioural and eye thermal profiles

This study demonstrates that lamb vocalizations serve as reliable indicators of negative emotional arousal, with their acoustic properties primarily driven by bodily activation and showing complex, trait-dependent correlations with eye temperature changes during isolation.

The Gist

Imagine you are trying to figure out how a friend is feeling, but they can't speak your language. You might look at their body language (are they pacing nervously?), check their face (are they flushed?), or listen to their voice (is it shaky or loud?).

This study did exactly that, but with lambs. The researchers wanted to know: Can we tell how stressed a lamb is just by listening to its "baa," and does that sound match what its body is doing?

Here is the story of the experiment, broken down into simple parts.

1. The Setup: The "Social Distance" Test

The researchers took 20 young lambs and put them through a two-part test designed to make them feel a little lonely and anxious.

• Phase 1 (The "Fence" Phase): The lamb was put in a small pen, but it could still see, hear, and touch its friends on the other side of the fence. It was like being in a room with your family, but behind a glass wall.
• Phase 2 (The "Isolation" Phase): The rest of the flock was led away to the pasture. The lamb was now completely alone in the pen. No friends, no voices, no touch. This is the "scary" part where the stress levels were supposed to spike.

2. The Three Ways They Checked the Lambs' Moods

To make sure the lambs were actually stressed, the researchers didn't just guess; they used three different "mood meters":

• The Body Meter (Behavior): They watched the lambs like hawk-eyed coaches. Did the lamb stand still? Or was it running in circles, jumping up and down, and banging its head against the pen bars?
  • The Result: In the "Fence" phase, the lambs were chill. In the "Isolation" phase, they went into panic mode, running and jumping like they were trying to escape a fire.
• The Thermometer (Eye Temperature): They used a special thermal camera to look at the lambs' eyes. Think of the eye area like a radiator. When an animal gets stressed, blood rushes to the eyes, making them hotter.
  • The Result: The "radiators" in the smaller lambs got noticeably hotter when they were alone. Interestingly, the larger lambs didn't show this heat spike (maybe their "radiators" were already running hot, or they had a different way of handling stress).
• The Microphone (Vocalizations): They recorded every "baa" the lambs made.

3. The Big Discovery: The "Stressed Lamb" Soundtrack

When the researchers analyzed the recordings, they found that a stressed lamb sounds very different from a relaxed one. It's like the difference between a calm hum and a frantic siren.

• Higher Pitch: The stressed bleats were much higher in pitch (think of a squeaky toy vs. a deep drum).
• More Chaos: The sounds were "noisier" and less musical. A calm bleat is smooth; a stressed bleat sounds like static or chaos.
• Shorter Duration: The calls became shorter and faster. Instead of a long, drawn-out "Baaaaaa," it became a rapid-fire "Baa! Baa! Baa!"

The Analogy: Imagine you are at a party.

• Relaxed: You are chatting with a friend, speaking in a low, smooth voice, taking your time.
• Stressed: Suddenly, the fire alarm goes off. You start shouting, your voice gets higher and shriller, you speak in short, choppy bursts, and your words sound a bit garbled because you are breathing so fast. That is what the lambs sounded like when they were alone.

4. The Twist: It's Not Just About the Sound

The most fascinating part of the study was connecting the dots between the body, the eyes, and the voice.

• Body = Voice: The more the lamb ran and jumped (high bodily activation), the more its voice changed to that high-pitched, chaotic "siren" sound. The body and the voice were in perfect sync.
• Eyes + Body = Duration: The eye temperature alone didn't change the sound much. However, if a lamb had hot eyes AND was running around frantically, its calls got longer. It's as if the body said, "I'm stressed!" and the eyes said, "I'm hot!" and together they made the lamb hold its breath and make a longer noise.
• Size Matters: The smaller lambs were the ones whose eyes got hot and whose voices changed the most. The big, grown-up lambs were a bit more stoic (or maybe just too big to show the same signs).

The Bottom Line

This study proves that lambs have a "stress voice." When they are scared or lonely, they don't just act crazy; they sound crazy. Their voices get higher, noisier, and faster, matching their frantic running and hot eyes.

Why does this matter?
Farmers and scientists can now use these sounds as a "remote control" to check on animals. If they hear the lambs making those high-pitched, chaotic noises, they know the animals are stressed and need help, even if they can't see the animals' eyes or watch them run around. It's like having a built-in alarm system for animal happiness.

Technical Summary

1. Problem Statement

While vocalizations are widely recognized as indicators of emotional arousal in animals, there is a lack of validated studies that simultaneously measure vocal output alongside independent physiological and behavioral indicators of arousal. Specifically, the relationship between the spectro-temporal structure of sheep vocalizations and negative emotional arousal (e.g., stress from social isolation) remains insufficiently understood. Furthermore, it is unclear how physiological proxies, such as eye temperature measured via infrared thermography, correlate with vocal changes, and whether these relationships are modulated by individual traits like body size.

2. Methodology

Experimental Design:

• Subjects: 20 unweaned lambs (4–5 months old) from a commercial flock in Denmark. The sample included 11 females, 7 males, and 2 of unknown sex, representing natural variation in breed and size.
• Procedure: Lambs underwent a two-phase isolation test, each lasting 4 minutes:
  1. Partial Isolation: The lamb was placed in a metal pen within the barn but maintained visual, acoustic, and tactile contact with the flock through a grid.
  2. Full Isolation: The flock was moved to pasture, leaving the lamb alone in the pen with no contact.
• Data Collection:
  • Behavioral: Video recorded and scored using an ethogram. Key metrics included locomotor activity (running, jumping), state changes, and vocalization rates.
  • Physiological: Infrared thermography (Hikmicro G60) captured eye temperature. The "Coefficient of Reaction" (CR) was calculated to measure the relative change in maximum eye temperature from partial to full isolation.
  • Acoustic: High-quality "High Bleats" (open-mouth calls) were recorded. Acoustic features extracted included fundamental frequency ($f_0$), formants (F1–F3), dominant frequency, Wiener entropy (tonality/chaos), and duration.
• Statistical Analysis:
  • Principal Component Analysis (PCA) was used to reduce behavioral and acoustic data into composite scores (behPCs and acPCs).
  • Linear Mixed-Effects Models (LMMs) were employed to test the effects of isolation phase, lamb size (classified as "small" <52cm or "large" >52cm), and their interactions on behavioral, thermal, and acoustic variables.
  • Co-variation models tested how behavioral activation and eye temperature predicted acoustic features.

3. Key Contributions

• Multi-modal Validation: The study provides a rare, simultaneous validation of vocal, behavioral, and physiological markers of negative arousal in sheep.
• Size-Dependent Thermography: It demonstrates that eye temperature responses to stress are not uniform across all individuals but are significantly modulated by body size (biometry).
• Interaction Effects: It reveals that the correlation between physiological arousal (eye temperature) and vocal output is conditional upon the level of behavioral activation.

4. Key Results

Behavioral Response:

• Increased Arousal: Lambs exhibited significantly higher bodily activation during full isolation compared to partial isolation. This was characterized by increased running, jumping, bar head touches, and vocalization rates, alongside decreased time spent in static postures.
• Composite Score: The first behavioral principal component (behPC1), representing bodily activation, increased significantly ($p < 0.001$) in full isolation.

Thermal Response:

• Size Interaction: Eye temperature increased from partial to full isolation, but only in small lambs. Large lambs showed no significant temperature change, suggesting a potential "ceiling effect" where larger animals have higher baseline temperatures or different thermoregulatory dynamics.
• Time Effect: Eye temperature rose progressively over time within the isolation phases, regardless of the phase type.

Acoustic Response:

• Spectro-temporal Shifts: Calls in full isolation were characterized by:
  • Higher Frequencies: Increased mean fundamental frequency ($f_0$), dominant frequency, and formant frequencies.
  • Reduced Tonal Quality: Increased Wiener entropy (more chaotic/noisy calls).
  • Shorter Duration: Calls were significantly shorter in full isolation.
• Size Effect: Small lambs produced calls with higher acoustic scores (indicating higher frequencies and more chaos) compared to large lambs.

Co-variation (Behavioral + Physiological + Acoustic):

• Behavioral Driver: Higher bodily activation (behPC1) was the primary predictor of acoustic changes, specifically driving higher frequencies and reduced tonality.
• Physiological Interaction: Eye peak temperature did not directly predict acoustic features in isolation. However, it interacted with behavioral activation:
  • Call Duration: Call duration increased with eye temperature only in lambs with high behavioral activation.
  • Formants: The second formant mean increased with eye temperature in small lambs but not in large ones.

5. Significance

This study establishes that lamb vocalizations are robust, non-invasive indicators of negative emotional arousal. The findings confirm that as arousal increases, sheep calls become higher-pitched, shorter, and more chaotic. Crucially, the research highlights that biometric factors (body size) and behavioral context are critical when interpreting physiological stress markers.

• Practical Application: The study supports the use of acoustic monitoring and infrared thermography in animal welfare assessments, provided that body size is accounted for.
• Theoretical Insight: It suggests that physiological indicators (like eye temperature) may only correlate with vocal output when the animal is already in a state of high behavioral activation, indicating a complex interplay between the autonomic nervous system and motor output during emotional states.
• Methodological Advance: The integration of PCA and LMMs offers a rigorous framework for disentangling the effects of individual traits and environmental stressors on multi-modal emotional expression.