How can you tell when a horse is feeling stressed?
It’s all in the eyes and the way their eyelids twitch, University of Guelph researchers have discovered.
A horse will blink less and twitch its eyelids more when it’s under mild stress, the research team found – a new finding that could offer handlers a simple, easy-to-spot sign their animal is becoming agitated.
The study, published in the journal Animals, is thought to be the first to reveal the significance of eyelid twitches as an indicator of stress, says Prof. Katrina Merkies, the study’s lead author.
“With humans, we already know our blinking changes when we are under pressure. Some studies have shown we blink more when agitated while others found we blink less.
We wanted to see if horses blink rates change too,” said Merkies, a professor in the Department of Animal Biosciences at the Ontario Agricultural College.
Although many horse handlers can tell when their animals are agitated, it can sometimes be hard to get a good read on a horse’s mood — particularly if the animal has been well trained.
“When we train horses, we specifically teach them to suppress their stress responses because we don’t want horses to react when they are startled or nervous. But even if they’ve learned to suppress their reaction, it doesn’t actually decrease the stress they feel,” she said.
While stress can be measured through heart rate monitors or blood cortisol levels, Merkies and her team wanted a non-invasive measurement, so they decided to test whether a horse’s eyes could offer clues.
They recruited 33 horses of various breeds from three riding lesson facilities in eastern Ontario and exposed them to three mildly stressful scenarios.
In the first, a ball was thrown in front of the horse in an attempt to startle the animal. In the next, the horse was visually separated from its herd for a few minutes.
Finally, the horse’s food was withheld for three minutes at feed time while its herd mates were allowed to begin eating.
The researchers filmed the horses, watching for changes in eye and ear movement, head tilt and general restlessness.
They found that withholding the feed for a few minutes was the most stressful for the horse as indicated by its increased heart rate, restlessness and head movement. Conversely, separation and the startle test evoked little response.
Researchers attempted to startle the horses by throwing a ball in front of them.
Credit: University of Guelph.
“It’s important to remember these were riding school horses, so they were used to being startled and being separated. But the withholding of food was new, so that’s likely why they became stressed,” she said.
When researchers reviewed videos of the horses’ eyes during feed withholding, they noticed the horses blinked less but twitched their upper eyelids more.
On average, the horses’ full blink rate decreased to an average of five blinks per minute during the stress compared to the eight to nine times per minute when relaxed. During the feed restriction, when the horses felt the most stress, their eyelid twitches increased from an average of two twitches per minute to six twitches per minute. There was no increase in eyelid twitches with the other stress tests.
Researchers attempted to startle the horses by throwing a ball in front of them. The image is credited to University of Guelph.
Merkies said she hopes her team’s finding will help horse handlers looking for simple ways to gauge their animals’ moods.
“There’s no one measure that is going to tell us everything, but this is another tool we can add to the toolbox that we can use together to understand our animals better,” she said.
Physiological changes provide indices of stress responses, however, behavioural measures may be easier to determine.
Spontaneous eye blink rate has potential as a non-invasive indicator of stress. Eyelid movements, along with heart rate (HR) and behaviour, from 33 horses were evaluated over four treatments:
(1) control—horse in its normal paddock environment;
(2) feed restriction—feed was withheld at regular feeding time;
(3) separation—horse was removed from visual contact with their paddock mates; and
(4) startle test—a ball was suddenly thrown on the ground in front of the horse. HR data was collected every five s throughout each three min test.
Eyelid movements and behaviours were retrospectively determined from video recordings. A generalized linear mixed model (GLIMMIX) procedure with Sidak’s multiple comparisons of least squares means demonstrated that both full blinks (16 ± 12b vs. 15 ± 15b vs. 13 ± 11b vs. 26 ± 20a full blinks/3 min ± SEM; a,b differ p < 0.006) and half blinks (34 ± 15ab vs. 27 ± 14bc vs. 25 ± 13c vs. 42 ± 22a half blinks/3 min ± SEM; a,b,c differ p < 0.0001) decreased during feed restriction, separation and the startle test compared to the control, respectively.
Eyelid twitches occurred more frequently in feed restriction (p < 0.0001) along with an increased HR (p < 0.0001). This study demonstrates that spontaneous blink rate decreases while eyelid twitches increase when the horse experiences a stressful situation.
Results
On average, horses performed full blinks 8–9 times/min in the absence of any stressors. This rate decreased to 5 blinks/min in the presence of any external stressors. Conversely, eyelid twitches increased from about 2/min in the control situation to 6/min during feed restriction. Full eye blinks occurred more often during control than during any other treatment (F(3,95) = 9.88, p < 0.0001; Figure 2). Half blinks occurred most often during control and feed restriction treatments, and least often during separation or startle test (F(3,95) = 10.65, p < 0.0001; Figure 2). Eyelid twitches were more evident during the feed restriction treatment than during any other treatment (F(3,95) = 9.46, p < 0.0001; Figure 2).
Total number of observations of full eye blinks, half blinks and eyelid twitches (±SD) in horses (n = 33) over a 3 min observation period during control, feed restriction, separation from conspecifics or a startle test. a,b,c differ across treatments p < 0.0001.
Horse heart rate was higher during feed restriction (44 ± 13 beats per minute (bpm)) and lower during separation (37 ± 7 bpm) and the startle test (37 ± 8 bpm) compared to the control (39 ± 8 bpm) (F(3,92) = 306.12, p < 0.0001). There was no effect of facility (p > 0.05) on the behaviours or HR.
The horses’ right ear was forward more often during separation and the startle test (F(3,95) = 8.29, p < 0.0001; Figure 3), whereas it was more often sideways during feed restriction and the control (F(3,95) = 22.53, p < 0.0001). There was no difference among treatments for the percentage of time the horses had their ears back (F(3,95) = 0.82, p > 0.49).
Percentage of time that horses’ (n = 33) ears were forward, sideways or back (±SD) over a 3 min observation period during feed restriction, separation from conspecifics, a startle test or control. a,b,c differ across treatments p < 0.0001.
Horses held their head raised more frequently during feed restriction (F(3,95) = 30.02, p < 0.0001; Figure 4) and held their head low more often during the control treatment and startle test (F(3,95) = 7.15, p = 0.0002).
Percentage of time that horses’ (n = 33) heads were held high, neutral or low (±SD) over a 3 min observation period during feed restriction, separation from conspecifics, a startle test or control. a,b,c differ across treatments p < 0.0001.
Oral behaviours were most evident during the feed restriction, with significantly fewer during separation and the startle test (F(3,95) = 11.42, p < 0.0001). Horses were more often restless during feed restriction than separation or the startle test (F(3,95) = 6.78, p = 0.0003).
Source:
University of Guelph
Media Contacts:
Katrina Merkies – University of Guelph
Image Source:
The image is credited to University of Guelph.
Original Research: Open access
“Eye Blink Rates and Eyelid Twitches as a Non-Invasive Measure of Stress in the Domestic Horses”. Katrina Merkies et al.
Animals doi:10.3390/ani9080562.
