ArticlePDF Available

Abstract and Figures

Emotions are now largely recognised as a core element in animal welfare issues. However, convenient indicators to reliably infer emotions are still needed. As such, the availability of behavioural postures analogous to facial expressions in humans would be extremely valuable for animal studies of emotions. The purpose of this paper is to find out stable expressive postures in sheep and to relate these expressive postures with specific emotional contexts. In an initial experiment, we identified discrete ear postures from a comprehensive approach which integrates all theoretically distinguishable ear postures. Four main ear postures were identified: horizontal ears (P posture); ears risen up (R posture); ears pointed backward (B posture); and asymmetric posture (A posture). In a second experiment, we studied how these ear postures were affected by specific emotional states elicited by exposing sheep to experimental situations in which elementary characteristics (ie suddenness and unfamiliarity, negative contrast and controllability) were manipulated. We found that: i) the horizontal P posture corresponds to a neutral state; ii) sheep point their ears backward (B posture) when they face unfamiliar and unpleasant uncontrollable situations, hence likely to elicit fear; iii) they point their ears up (R posture) when facing similar negative situations but controllable, hence likely to elicit anger; and iv) they expressed the asymmetric A posture in very sudden situations, likely to elicit surprise. By cross-fostering psychological and ethological approaches, we are able to propose an interpretation of ear postures in sheep relative to their emotions.
Content may be subject to copyright.
© 2011 Universities Federation for Animal Welfare
The Old School, Brewhouse Hill, Wheathampstead,
Hertfordshire AL4 8AN, UK
Animal Welfare 2011, 20: 47-56
ISSN 0962-7286
Cognitive sciences to relate ear postures to emotions in sheep
A Boissy*, A Aubert, L Désiré, L Greiveldinger§, E Delvaland I Veissier
INRA UR1213 Herbivores, Site de Theix, F-63122 Saint-Genès-Champanelle, France
Exploitation et Sécurité Routières, CETE Ouest, F-22015 St Brieuc Cedex, France
§UMR UJF CNRS 5525, Acquisition, Fusion d’Information et Réseaux pour la Médecine, Technopole du Biopark, F-74160 Archamps,
* Contact for correspondence and requests for reprints:
Emotions are now largely recognised as a core element in animal welfare issues. However, convenient indicators to reliably infer
emotions are still needed. As such, the availability of behavioural postures analogous to facial expressions in humans would be
extremely valuable for animal studies of emotions. The purpose of this paper is to find out stable expressive postures in sheep
and to relate these expressive postures with specific emotional contexts. In an initial experiment, we identified discrete ear
postures from a comprehensive approach which integrates all theoretically distinguishable ear postures. Four main ear postures
were identified: horizontal ears (P posture); ears risen up (R posture); ears pointed backward (B posture); and asymmetric posture
(A posture). In a second experiment, we studied how these ear postures were affected by specific emotional states elicited by
exposing sheep to experimental situations in which elementary characteristics (ie suddenness and unfamiliarity, negative contrast
and controllability) were manipulated. We found that: i) the horizontal P posture corresponds to a neutral state; ii) sheep point
their ears backward (B posture) when they face unfamiliar and unpleasant uncontrollable situations, hence likely to elicit fear; iii)
they point their ears up (R posture) when facing similar negative situations but controllable, hence likely to elicit anger; and iv)
they expressed the asymmetric A posture in very sudden situations, likely to elicit surprise. By cross-fostering psychological and
ethological approaches, we are able to propose an interpretation of ear postures in sheep relative to their emotions.
Keywords:animal welfare, behaviour, cognition, ear postures, emotions, sheep
The concern for animal welfare stems from the social and
legal acknowledgement that animals are sentient beings,
capable of feeling emotions. But, despite their central role in
welfare, there is still some reluctance to ascribe emotional
life to animals. However, basic (ie Darwinian) emotions, as
fundamental adaptive processes, have a long evolutionary
history and are shared by many species. Moreover, core
components of emotions — physiological, behavioural and
subjective components — have been clearly identified, and
the first two components can be readily measured objec-
tively in animal studies (Dantzer 1988). Finally, recent
development in the study of emotions in animals has bene-
fitted from merging theories and methods from psychology
and ethology (Dantzer 2002; Désiré et al 2002). Adding to
former biological theories of emotions, it is now acknowl-
edged, since the first studies in the field of cognitive-rela-
tional theories, that emotions are the consequence of an
appraisal and are therefore sustained by cognitive processes
(Scherer et al 2001). According to Scherer (1999), an
emotion results from the appraisal by the subject of a trig-
gering situation through the assessment of a limited number
of elementary characteristics (eg the suddenness, unfamil-
iarity and pleasantness of the situation, the ability for the
subject to predict and to control the situation, etc). Recently,
the behavioural and physiological responses of sheep to
controlled variations of the elementary characteristics of the
eliciting situation (eg suddenness, unfamiliarity, etc) were
described in order to set-up the ‘appraisal’ framework of
emotions to animals (Boissy et al 2007a). We showed that
sheep are responsive to most appraisal characteristics
defined in human studies, such as the suddenness and unfa-
miliarity of an event (Désiré et al 2004, 2006), the unpre-
dictability (Greiveldinger et al 2007) and the potential
control of the event (Greiveldinger et al 2009).
Physiological components commonly used to assess
emotions in animals generally provide a quantitative assess-
ment of the emotional activation (ie the arousal or the
intensity of the emotional response) without clearly defining
the exact nature of the emotion (eg the positive or negative
valence of the emotional experience). For example, plasma
concentrations of glucocorticoid can be increased in
Universities Federation for Animal Welfare Science in the Service of Animal Welfare
48 Boissy et al
response to both acute negative emotional stimuli (Mason
1971) or by the expectation of a positive situation, such as
the availability of a sex mate (Colborn et al 1991). The
mobilisation of energy that precedes these situations can lead
to a considerable overlap of responses to aversive and
pleasant stimuli (Dawkins 1983). Likewise, common behav-
ioural measures that generally consist of fixed action
postures, such as startle, offensive or defensive postures,
freezing, or approach, only provide information about the
intensity of the underlying emotion (for a review see Boissy
1998). Several indicators are generally required to differen-
tiate negative and positive emotional valence (Broom 1997;
Dawkins 2001; Désiré et al 2002), and further behavioural
indicators are nevertheless still needed to allow a convenient
and reliable assessment of emotion in sheep.
In humans, expression of emotions has been extensively
studied through changes in facial features. For example, one
of the most popular analysis systems of facial expression in
humans — the FACS (Facial Action Coding System; Ekman
et al 1972) — is based on the description of muscular
contractions (ie action units) of one’s face, the combination
of these action units forming a posture referring to a specific
emotion. Analogous expressive or postural indicators of
emotions in animals are unfortunately still missing
(Berridge 2000; Boissy et al 2007b) but they would be
highly beneficial both to scientific investigation of emotions
in animals and to welfare issues, by providing convenient
tools to understand and solve human-animal related
concerns, for example. So far, few germinal studies have
shown the availability of such parameters. For instance,
facial expressions relating to gustatory sensory pleasures
have been described in rats as in humans (Berridge 2003).
Moreover, it has been recently shown in farm mammals that
individual recognition is partly based on facial features
(cattle: Coulon et al 2009; sheep: Kendrick et al 2001). Such
findings call for further investigation of stable emotion-
specific facial postures in farm mammals. In order to find
out such facial expressions in farm animals, we focalised on
ears’ postures since they are essential for gathering informa-
tion from the environment (Manteuffel 2006).
By contrast with other mammals, such as primates, sheep
have a limited superficial facial muscles’ network and thus
do not appear to have a wide array of facial expressions.
Nevertheless, they are characterised by a high mobility
both of the neck — offering various global head
postures — and of the ears, due to several muscles for
rotating their ears (Nickel et al 1968). Do specific ear
movements occur in particular emotion-eliciting contexts?
Can measuring ear postures in sheep be used to accurately
infer their emotions? So far, very few studies have
measured the ear postures per se in farm mammals. Scarce
reports from the literature suggest that ear postures may be
useful in assessing emotional valence in farm animals. In
cattle, for example, a high occurrence of pendulous ear
postures was used as an indicator of the animals’ positive
rating of their favourite grooming sites (Schmied et al
2008). To our knowledge, only one study has been recently
carried on ear postures in sheep (Reefmann et al 2009). In
that experiment, ear postures were defined a priori and
sheep were exposed to complex situations likely to induce
states of emotions but it was not possible to characterise the
exact nature of the emotion induced.
The present paper focuses on the identification of ear
postures and their possible link with emotional states in
sheep. In an initial experiment, we identified discrete ear
postures from a comprehensive approach which distinguish
all theoretically measurable ear postures without a priori
interpretations. In a second phase, we ran three experiments
in order to explore the possible correspondence between
previously identified discrete ear postures and emotions.
According to the appraisal framework of emotions previ-
ously used in sheep (Désiré et al 2002; Boissy et al 2007a),
ewes were individually exposed to triggering situations
varying in their degree of either suddenness or unfamil-
iarity, and of negative contrast or controllability.
Materials and methods
An initial experiment was conducted to identify discrete
ear postures. In a second experiment, three groups of
sheep have been utilised separately to specifically assess
the relationship between the identified discrete ear
postures and three elementary components of emotions
corresponding to: (i) suddenness and familiarity of a
visual event (Study 1); (ii) negative contrast (ie, the drastic
reduction of a food reward (Study 2); and (iii) control (ie
control of the access to a food reward) (Study 3).
Animals and rearing
Ewes from the Romane breed, aged between six and ten
months, were used; ten for Experiment 1 and 78 for
Experiment 2 (32, 22, 24 in Studies 1, 2, 3). They had been
separated from their dam 24 to 48 h after birth and housed
with other animals of the same age. At the time of the exper-
iments, the ewes were reared in large pens (called home
pens, 3.8 m2per animal) with deep-litter straw bedding,
adjacent to an experimental chamber (Figure 1). They were
provided hay ad libitum and 400 g of food pellets composed
of barley, sugarbeet, wheat, maize, sunflower and soya
(Thivat Nutrition Animale, Cusset, France). The food was
distributed daily at 1700h in the home pen. When the
training procedure started (see below), the ewes received
food pellets during tests and were supplemented with pellets
in their home pen to reach a total provision of 400 g.
Experimental set-up
The experimental chamber was divided into three compart-
ments: a pre-test (1.5 m2), a test arena (3 m2) and a corridor
(1.8 m2). Each compartment was delineated by 1.8-m high
wooden partitions that prevented animals from seeing each
other (Figure 1). Sliding doors permitted access from one
compartment to the next. A device for delivering food
pellets was placed in the test arena at the opposite side to the
entrance. It consisted of a deck 30 cm above floor level with
a central aperture housing an adjustable 15-cm diameter
trough. Food pellets were delivered in the trough via an
© 2011 Universities Federation for Animal Welfare
Facial expressions of emotions in sheep 49
electronic system placed outside the arena and that could be
controlled by the experimenter.
Four video cameras (three Sony SPT-MC128CE and one
Sony DCR-TRV 320E Digital Handycam, Sony
Corporation, Tokyo, Japan) were used to record the behav-
ioural reactions and postures of the lambs around the trough
in the test arena. The cameras were connected to a video
recorder that simultaneously recorded the images from the
four cameras on the same tape (Sony SVT-1000P, Sony
Corporation, Tokyo, Japan) thanks to a quadravision system
(MV25 Multivision Processor, model MX25, Robot
Research, San Diego, California, USA). The cameras were
placed in such a way as to provide top, side and face-on
views of the ewes.
The procedures to train and test the animals have already
been published (Désiré et al 2004, 2006; Greiveldinger et al
2007, 2009, 2010). We will only summarise them briefly in
the present paper.
General training procedure for ewes
To accustom the animals to the experimental chamber, they
were given free access to the chamber from their home pen
for one or two weeks; all access doors were opened and
food pellets were at their disposal in the test arena. Then,
once they accepted to feed in test arena, each ewe was indi-
vidually exposed to the chamber; the access doors were shut
when she moved from one compartment to the next one.
This was repeated on at least five sessions. After this
training phase, each ewe was exposed to test procedures
specific to each study.
Specific procedures for testing ewes
In Experiment 1, a rope-and-pulley device installed behind
the trough of the test arena was utilised to move a white
scarf (0.2 × 0.2 m; length × breadth) from a non-visible
location to 0.2 m above the trough in full sight of the
animal. While the ewe was eating, the scarf was moved by
the experimenter at a speed of 0.88 m s–1. The ewe was left
two minutes more in the arena. The test was repeated once
daily on three consecutive days.
In Experiment 2, Study 1, the rope-and-pulley device was
used, as in Experiment 1, with the exception that a second
object could be used (a flat square made of the same white
textile as the scarf). During training, half the ewes were
familiarised with the scarf and half with the flat square; these
objects always appeared slowly (0.06 cm s–1). The ewes were
trained once daily until they did not step back when object
appeared for three consecutive days. Then, during tests, only
the scarf was presented (it was thus familiar for certain ewes
and not so for others). It was moved down next to the trough
either slowly as previously (0.06 cm s–1) or rapidly
(0.88 m s–1) or. Four test sessions were run.
In Experiment 2, Study 2, the ewes were trained to perform an
operant task (to cross a beam with their muzzle) to get a large
food reward (50 g food pellets) or a small food reward (10 g
food pellets). The photobeam and two photoelectric cells were
placed in a 0.2 × 0.2 × 0.2 m (length × breadth × height) hole
Animal Welfare 2011, 20: 47-56
Figure 1
Experimental set-up (top) and the scarf (bottom) used for the
two experiments (except for Studies 2 and 3).
50 Boissy et al
placed in the partition, 1 m from the left-hand side of the trough.
Each ewe was subjected to one session a day and the session
was terminated when she had eaten four rewards, which took
between 2 and 4 min. After ten days of training the animals were
subjected to three test sessions with the same conditions then for
the next three sessions the ewes trained with a large reward and
received a small one (negative contrast), while the remaining
ewes continued to receive a small reward (no contrast).
In Experiment 2, Study 3, food was always present in the
trough of the test arena but the animals did not have free
access to it. From time-to-time during a session, an air
blower was turned on above the trough and a grid was
moved above the trough to prevent animals from eating.
Half the ewes were trained to perform the same operant task
as in Study 2 (crossing the photobeam) in order to resume
access to the food (air blower is turned off and the grid is
removed). The remaining ewes were yoked to the previous
ones: they received exactly the same access to the food but
without controlling it. Ewes that could control access to the
food and their yoked counterparts were further observed
during four test sessions, once the former had acquired the
operant task.
Identification and detection of ear postures
Video recordings were taken during the following sessions.
Experiment 1 — we selected a total of ten recordings that
allow full vision of the animals; these were taken during the
first, second, or third session where the scarf was presented
to animals and recordings from 30 s before to 30 s after the
appearance of the scarf were analysed.
Experiment 2, Study 1 — recordings from 30 s before to
30 s after the appearance of the scarf during the four test
Experiment 2, Study 2 — recordings from 10 s before to
10 s after delivery of the food reward during the three test
Experiment 2, Study 3 — recordings from 5 s before to 5 s
after the air blower was turned on during the four test
Observer Video Pro (version 4.0.21, Noldus Information
Technology, the Netherlands) was used to record all ear
postures irrespective of the experiment. In Experiment 1,
two independent observers noted all changes in the
position of ears according to two criteria: the position of
the ear in regard to the frontal plane of the head, and the
visibility of the auricle. Within each criterion categories
were exclusive (see Figure 2):
• Position of the ear in the frontal plane — the ears can be
aligned with frontal plane, oriented forward (the top of the ear
is in front of the frontal plane), oriented backward (the top of
the ear is behind the frontal plane), or asymmetric (the two
ears differ in their position in regard to the frontal plane);
• Visibility of the auricle — the auricles can be flat (the
inner and outer sides are not visible, the inner side is
parallel to the floor, the observer can see the full ears),
open (the inner sides are visible by an observer placed in
front of the animal), closed (the inner sides are not visible,
the outer sides are visible at the root of the ear, the ears
cannot be fully seen by the observer), asymmetric (one
inner side visible, one not).
The positions of ears in the frontal plane of the head were
recorded separately from the positions of ears according to
the visibility of the auricle. Table 1 presents the 16 theoret-
ical combinations obtained from the crossing of the
4 × 4 framework (four in the frontal plane and four facial
views for the visibility of the auricle). From the 16 combina-
tions of position of the ears, we identified four discrete ear
postures (see results from Experiment 1). These defined
postures were further encoded in all studies of Experiment 2.
Statistical analysis
We used the software SAS (version 8.1, SAS Institute Inc,
Cary, NC, USA) for all analyses. In Experiment 1, each time
an observer noted a change in ear position, we checked the
position that the other observer noted. We calculated the
frequency and the duration of each of the 256 possible asso-
ciations between the 16 combinations of ear positions from
Observer 1 and those from Observer 2. The consistency
between the two observers encoding the ear postures was
assessed with Cronbach alpha index.
In Experiment 2, the duration of ear postures was
arcsine-transformed and compared between conditions
with a mixed model of variance analysis for repeated
measures. Repetitions were for session number and time
during a session (before vs after the event). The correla-
tion matrix was considered unstructured. The results are
expressed as mean (± SD). The limit of significance was
set at P< 0.05.
Experiment 1 — Identification of discrete ear postures
The consistency between observers encoding the ear
postures was high (Cronbach’s alpha = 0.8645). The obser-
vations were then considered reliable. Figure 3 shows the
frequency and mean duration of the 16 theoretical combina-
tions of ear positions determined from the frontal plane and
visibility of the auricle. Certain combinations were never
observed (ie combinations 5, 9, 10, 12 and 13, such as ears
oriented forward and flat auricles). Some other combina-
tions (ie combinations 3, 4, 7, 8, 14, 15) were displayed for
very brief periods (mean duration<1s).
The remaining combinations were frequently expressed
(combinations 1, 2, 6, 11 and 16), with significant mean
durations (ie > confidence interval of mean). These combina-
tions were therefore considered as distinctive and meaningful
ear postures. In addition, it was decided to regroup postures
1 and 2 (the two ears are either ahead or in the frontal plane
with visible auricle) since they were closely associated during
the longest expressive episodes. Finally, four discrete ear
postures were considered for further analysis:
• ‘Raised ears’ posture (R posture), which correspond to
combinations 1 and 2. The two ears are in the same position
relative to the frontal plane (either ahead or aligned), and
auricles are visible from front view;
© 2011 Universities Federation for Animal Welfare
Facial expressions of emotions in sheep 51
• ‘Ears in the plane’ posture (P posture), which corresponds
to combination 6. The two ears are in the frontal plane and
auricles are concealed from front view’
• ‘Ears backward’ posture (B posture), which corresponds to
combination 11. The two ears are behind the frontal plane
and auricles are concealed from front view; and
• ‘Asymmetrical ears’ posture (A posture), which correspond
to combination 16. The two ears are in two distinct positions
relative to the frontal and visibility of auricles is asymmetrical.
Regarding the combinations displaying for very brief
periods, they could correspond to transitional postures
between two of these four ear postures. Each of the
transitional postures was then considered as belonging
to the same defined posture to which it was associated
most often: combination 3 was then regrouped with
posture R, combinations 4, 8 and 14 were regrouped
with posture A, and combinations 7 and 15 were
regrouped with posture B.
Animal Welfare 2011, 20: 47-56
Figure 2
Position of ears collected from ewes according to two criteria: position of the ears in relation to frontal plane of the head (top) and
orientation of the auricles observed in front of the animal (bottom).
Table 1 Labels of the 16 theoretical combinations between the position of the ears in the frontal plane of the head
and the orientation of the auricles observed in front of the animal.
Front view
pv1 pv2 pv3 pv4
View from above pr1 1, pr1pv1 5, pr1pv2 9, pr1pv3 13, pr1pv4
pr2 2, pr2pv1 6, pr2pv2 10, pr2pv3 14, pr2pv4
pr3 3, pr3pv1 7, pr3pv2 11, pr3pv3 15, pr3pv4
pr4 4, pr4pv1 8, pr2pv2 12, pr4pv3 16, pr4pv4
52 Boissy et al
Experiment 2 — Assessing stable relations between
ear postures and emotional contexts
Study 1 — Suddenness/unfamiliarity
Whatever the session, the ears were mainly in the plane
posture before the appearance of the scarf (21.2 [± 3.2] vs
2.4 [± 2.1] s, 0.7 [± 0.3] and 4.8 [± 4.3] s for ear in the
plane, raised ears, asymmetric ears and ears backward;
F= 18.4, P< 0.001).
The ear postures changed when the scarf appeared. The
change depended on the speed of the presentation and the
object used for training (Table 2). In response to the rapid
appearance of the scarf, the ewes spent more time in the
asymmetric ear posture. Although the time spent in this
posture decreased during subsequent sessions, it was always
significantly higher than the time spent in the other ear
postures, except the posture with the ears in the plane.
In response to the slow appearance of the scarf in the first
session, the ewes that had been trained with the square
spent more time with the ears raised or backward while
the time spent with the ears in the plane was decreased.
From the second to the fourth session, their time spent
with the ears raised decreased whereas their time with the
ears in the plane increased.
© 2011 Universities Federation for Animal Welfare
Table 2 Duration (s) spent in ears’ postures by the ewes exposed to a neutral, a sudden or an unfamiliar event
(Experiment 2, Study 1, n = 32).
Ears in the plane (P) Ears backward (B) Asymmetric ears (A) Ears raised (R) SEM F-value P-value
Session 1
Neutral 20.3a1.0c4.2b3.1bc 2.5 6.43 0.01
Sudden 16.8a1.2d8.1b4.5c2.8 5.23 0.01
Unfamiliar 8.3a5.3c5.0c12.2b1.7 3.52 0.05
Sessions 2-4
Neutral 24.5a0.8c2.1b3.1b3.2 5.56 0.01
Sudden 19.2a0.9d7.1b4.0bc 3.3 3.23 0.05
Unfamiliar 22.0a0.5c3.4b4.2b2.8 3.74 0.05
Figure 3
Distribution of frequency and mean duration of combinations observed for fear postures. Each label indicates a specific combination (cf
Table 1). Combinations that were never observed (ie 0 values) are combinations 5, 9, 10, 12 and 13.
Facial expressions of emotions in sheep 53
The ewes that had been trained with the scarf (ie same
object as for the test) did not show a change of ear posture
from before to after the slow appearance of the scarf. The
time spent with the ears in the plane was higher than the
time spent in the other ear postures (Table 2). This differ-
ence was still significant for the next three sessions.
With the exception of the first session, the ewes spent very
little time with their ears backward after the appearance of
the scarf and this did not vary with treatments.
Study 2 — Negative contrast
The ewes were particularly sensitive to the shift in the
size of the reward (Figure 4). Compared to ewes that had
always received a small reward, those that were shifted
from a large to a small reward (ie negative contrast)
spent more time with their ears asymmetric (F= 4.5;
P< 0.01) or oriented backward (F= 3.2; P< 0.05) at the
time when the reward was delivered. For the next two
sessions, the difference between the two treatments was
significant only regarding the posture with the ears
backward (F= 2.9; P< 0.05).
Study 3 — Control
When the ewes were eating, their ears were mainly in the
plane (10.5 [± 3.5], 1.3 [± 1.1], 2.7 [± 1.3] and 0.9 [± 0.6] s for
the time spent with ears in the plane, ears raised, asymmetric
ears, and ears backward; F= 8.3, P< 0.001). The possibility
to control the situation elicited different ear postures when the
air blow was initiated (Figure 5): the ewes that could interrupt
the air blow raised their ears for longer (F= 4.9; P< 0.01) and
oriented them backwards less often (F= 8.3; P< 0.001) than
the ewes that had no control over the situation.
Animal Welfare 2011, 20: 47-56
Figure 4
Time duration (s) spent in ears postures by the ewes exposed to a negative contrast during (a) the first session and (b) the following two sessions
(Experiment 2, Study 2, n = 22). Half of the ewes that were previously trained with a large reward of food received a small one (negative contrast)
while the other half continued to receive a small reward as during training (no contrast). Four main ear postures were identified: horizontal ears
(pattern P), ears pointed backward (pattern B), asymmetric posture (pattern A) and ears rose up (pattern R). * P< 0.05; ** P< 0.01.
54 Boissy et al
The present study aimed to identify specific emotion-related
ear postures in sheep. First, we objectively identified four
discrete ear postures: ears in the plane (the two ears are in
the frontal plane and auricles are concealed from frontal
view), ears raised (the two ears are either ahead or aligned
and auricles are visible from front view), ears oriented
backward (the two ears are behind the frontal plane and
auricles are concealed from front view), and asymmetric
ears (the two ears are in two distinct positions relative to the
frontal and visibility of auricles is asymmetrical). From the
pool of putative combinations, some combinations were
never observed; this is easily understandable since they
represent anatomically irrelevant postures. For instance,
combination 5 would correspond to ear ahead in the frontal
plane but with auricles facing the ground. Then, we
analysed the occurrences of these four discrete postures in
relation to specific experimental situations defined
according to elementary appraisal components that were
previously shown as being relevant for eliciting emotional
responses in sheep: suddenness, unfamiliarity, negative
contrast, and uncontrollability. We observed that: i) the
posture with the ears in the plane was mainly associated
with neutral situations; ii) the backward posture was associ-
ated with unfamiliar and uncontrollable unpleasant situa-
tions; iii) the posture with raised ears was specifically
displayed in response to an unfamiliar, unpleasant but
controllable situation; and iv) the asymmetric posture is
mainly displayed in sheep exposed to sudden situations.
The two postures representing raised ears (ie forward ears)
and asymmetric ears have already been reported by
Reefmann et al (2009) as increasing during social contexts
eliciting negative emotions. Nevertheless, in our current
experiments, we could link these specific ear postures with
specific emotional experiences by exposing animals to
experimental situations defined according to the cognitive
components sheep use to evaluate their environment (Désiré
et al 2002; Boissy et al 2007a; Veissier et al 2009). In
addition, the comparison between our results and what was
found in humans can help the interpretation of sheep ear
postures. For instance, the facial expression of emotions
triggered by unfamiliarity in humans is described by
lowered eyebrows and raised eyelid (Kaiser & Wehrle
2001). Interestingly, the muscle network implied in
lowering eyebrows in humans would be evolutionarily
related to the muscles controlling the mobility of the ears in
animals (Fridlund 1994). The contraction of homologous
muscles could be therefore responsible for the lowering of
eyebrows in humans and raising of ears in animals. Since
eyebrows play a major role in expressing emotions in
humans (Ekman et al 1972), emotions in farm mammals
may be reflected in their ear postures.
Most of the results reported here were obtained in situations
with a negative valence for the animals. According to Fraser
and Duncan (1998), different evolutionary processes seem
to have selected negative from positive emotions. Negative
emotions are supposed to have evolved in ‘need situations’,
such as a threat to survival or reproductive success, whereas
positive emotions are supposed to have generally evolved in
‘opportunity situations’ where the resulting action may
enhance individual fitness without being essential for it
(Boissy et al 2007b). This may be the reason why positive
© 2011 Universities Federation for Animal Welfare
Figure 5
Time duration (s) spent in ears’ postures by the ewes controlling or not the access of food (Experiment 2, Study 3, n = 26). Ewes that
were previously trained to control throughout the experiment the interruption of air blow and the access of food are compared to
yoked ewes that had no possibility to control the accesses to food. Four main ear postures were identified: horizontal ears (pattern P),
ears pointed backward (pattern B), asymmetric posture (pattern A) and ears rose up (pattern R). ** P< 0.01; *** P < 0.001.
Facial expressions of emotions in sheep 55
emotions are more variable (either inter- or intra-individ-
uals) and versatile, hence more difficult to approach than
negative emotions. Relating to the homology between
muscles involved in lowering eyebrows in humans and in
animals’ ears raising (Fridlund 1994), the previous results
would indicate that some negative emotional experiences
involve rising ears up whereas positive emotional experi-
ences could coincide with non-erect ears (in our study the
plane ear posture). Further research is thus needed on the
positive side of the emotional scale to attempt to identify
specific facial expressions as it is known in humans. For
example, asymmetry of ear postures may differ between
negative and positive emotional states, due to lateralised
behaviour (Quaranta et al 2007) thought to result from
contra-lateral hemispheric brain activity (Wager et al 2003).
From our previous studies and according to the framework
used by Sander et al (2005), sheep appear to have the
potential to feel a wide range of emotions, including fear,
anger, rage and despair, because they use the same appraisal
components involved in such emotions as in humans (Boissy
et al 2007a; Veissier et al 2009). For instance, despair in
humans is triggered by situations which are evaluated as
sudden, unfamiliar, unpredictable, discrepant from expecta-
tions, and uncontrollable, whereas boredom results from an
overly predictable environment, and all these components
have been found to affect emotional responses in sheep.
Since it has been shown in the present study that ear postures
are related to specific appraisal components, these defined
ear postures could represent specific emotional signatures in
sheep. Thus, the backward posture that is associated with
unfamiliar and uncontrollable unpleasant situations would
express fear. The raised ear posture that is displayed in
response to an unfamiliar but controllable unpleasant
situation would characterise anger. Finally, the asymmetric
ear posture that is mainly displayed in response to sudden
situations in relation with a startle response would express
surprise. The assertions described in this study are not so far
from the intuitive knowledge of breeders or other people
who have close relations with animals (and sheep in this
case). The present paper gives support to this intuitive inter-
pretation of animal expressions.
Our interpretation of ear postures in sheep from a quantita-
tive ethogram-based approach could be reinforced by
analysing the concordance with the Qualitative Behaviour
Assessment (Wemesfelder 2001), that is a method based
upon the integration by observers of perceived animal
behaviour expression, using descriptors such as ‘calm’,
‘aggressive’, ‘sociable’ or ‘indifferent’. This method, based
on the body language of animals, has been recently
validated for sheep (Wickham et al 2009). Moreover, the
description of emotion-specific expressions in sheep
provides further information to better understand the social
function of emotions. Indeed, emotions are sustained by
three main processes: physiological (ie somatic and
visceral changes), subjective (ie the mental and representa-
tional content that can be verbally described in humans),
and expressive (ie the visible changes in behaviour,
postures, gestures and expressions). It has been shown that
the expressive component of emotions are fundamental for
many social processes. This has been demonstrated not
only in humans, but also in other primates (Dantzer 1988).
However, little work has been done on other animals, such
as farm species. It has been recently shown that such
animals (eg cattle or sheep) use facial information to
recognise specific individuals (Kendrick et al 2001;
Coulon et al 2009). Hence, such animals are sensitive not
only to the presence vs absence of conspecifics, but also to
subtle physical features of specific social partners. Our
study shows that sheep express specific stable changes in
their facial appearance (ie ear postures). These changes
could provide relevant information to conspecifics about
their environment. Further studies should be done to test
such a hypothesis, ie using conditioning procedures, and to
test whether animals are able not only to recognise
conspecifics’ identity but also to identify their emotional
state throughout their ear postures.
As we already claimed in previous papers, by applying
models developed for humans, we increase the knowledge
on how animals understand their environment and the
likely emotions they can feel. Moreover, such a non-
invasive and convenient method with freely moving
animals should help to better interpret what animals expe-
rience as unpleasant in their housing environments. It
should establish the basis for understanding and then
improving current housing and husbandry conditions from
an animal’s point of view. Such findings from sheep on
how to assess emotional states may be readily transferable
to closely related prey species at the very least.
In conclusion, this study provides some insight into
emotional expressions in sheep. This constitutes a first step
to characterise specific emotion-related facial expressions
in farm animals. Observations of ear postures are a reliable
non-invasive method for assessing the valence of emotional
states in sheep. We found that: i) sheep point their ears
backward when they face unfamiliar, unpleasant, and
uncontrollable situations, hence likely to elicit fear; ii) they
point their ears up when facing a similar negative situation
but controllable, hence likely to elicit anger or at least some
preparation of an active response; and iii) their ears are
more often asymmetric in very sudden situations, likely to
elicit surprise. These findings need to be confirmed in other
situations that might be appraised similarly by sheep but
using different stimuli. It is necessary to extend these first
four expressive postures characterising neutral or negative
emotions (ie fear, anger and surprise) to take into account
positive emotions in animals.
Animal welfare implications
The interpretation of ear postures could be used to assess
farming practices from the viewpoint of animals. It could also
help to understand what a given animal is feeling at a certain
time and the subsequent behaviour it is likely to adopt, eg
flight when the ears are oriented backward or attack when they
are raised up. These specific emotion-related facial expres-
sions could be easily used to enrich the behavioural measure-
ments for an overall assessment of welfare at farm level.
Animal Welfare 2011, 20: 47-56
56 Boissy et al
Berridge KC 2000 Measuring hedonic impact in animals and
infants: microstructure of affective taste reactivity patterns.
Neuroscience Biobehavioural Review 24: 173-198
Berridge KC 2003 Pleasures of the brain: affective neuroscien-
ce. Brain Cognition 52: 106-128
Boissy A 1998 Fear and fearfulness in determining behavior. In:
Grandin T (ed) Genetics and the Behavior of Domestic Animals pp
67-111. Academic Press: San Diego, USA
Boissy A, Arnould C, Chaillou E, Désiré L, Duvaux-
Ponter C, Greiveldinger L, Leterrier C, Richard S,
Roussel S, Saint-Dizier H, Meunier-Salaün MC, Valance D
and Veissier I 2007a Emotions and cognition: a new approach to
animal welfare. Animal Welfare 16: 37-43
Boissy A, Manteuffel G, Jensen MB, Moe RO, Spruijt B,
Keeling LJ, Winckler C, Forkman B, Dimitrov I, Langbein
J, Bakken M, Veissier I and Aubert A 2007b Assessment of
positive emotions in animals to improve their welfare. Physiology
and Behavior 92: 375-397
Broom DM 1997 Welfare evaluation. Applied Animal Behaviour
Science 54: 21-23
Colborn DR, Thompson Jr DL, Roth TL, Capehart JS and
White KL 1991 Responses of cortisol and prolactin to sexual
excitement and stress in stallions and geldings. Journal of Animal
Sciences 69: 2556-2562
Coulon M, Deputte BL, Heyman Y and Baudoin C 2009
Individual recognition in domestic cattle (Bos taurus): evidence from
2D images of heads from different breeds. PLoS ONE 4: e4441
Dantzer R 1988 Les Emotions. Presses Universitaires de France:
Paris, France
Dantzer R 2002 Can farm animal welfare be understood with-
out taking into account the issues of emotion and cognition?
Journal of Animal Science 80: E1-E9
Dawkins MS 1983 La Souffrance Animale. Editions du Point
Vétérinaire: Maisons-Alfort, France
Dawkins MS 2001 How can we recognise and assess good welfare?
In: Broom DM (ed) Coping with Challenge: Welfare in Animals Including
Humans pp 63-76. Dahlem University Press: Berlin, Germany
Désiré L, Boissy A and Veissier I 2002 Emotions in farm ani-
mals: a new approach to animal welfare in applied ethology.
Behavioural Processes 60: 165-180
Désiré L, Veissier I, Després G and Boissy A 2004 On the
way to assess emotions in animals: do lambs evaluate an event
through its suddenness, novelty or unpredictability? Journal of
Comparative Psychology 118: 363-374
Désiré L, Veissier I, Després G, Delval E, Toporenko G
and Boissy A 2006 Appraisal process in sheep: interactive effect
of suddenness and unfamiliarity on cardiac and behavioural
responses. Journal of Comparative Psychology 120: 280-287
Ekman P, Friesen WV and Ellsworth P 1972 Emotion in the
Human Face: Guidelines for Research and a Review of Findings.
Pergamon Press: New York, USA
Fraser D and Duncan IJH 1998 ‘Pleasures’, ‘pains’ and animal
welfare: toward a natural history of affect. Animal Welfare 7: 383-396
Fridlund AJ 1994 Facial expression and the methods of contempo-
rary evolutionary research. In: Fridlund AJ (ed) Human Facial Expression:
an Evolutionary View pp 28-54. US Academic Press: San Diego, USA
Greiveldinger L, Veissier I and Boissy A 2007 Emotional
experiences in sheep: predictability of a sudden event lowers sub-
sequent emotional responses. Physiology and Behavior 92: 675-683
Greiveldinger L, Veissier I and Boissy A 2009 Behavioural
and physiological responses of lambs to controllable versus uncon-
trollable aversive events. Psychoneuroendocrinology 34: 805-814
Greiveldinger L, Veissier I and Boissy A 2010 The ability of
lambs to form expectations and the emotional consequences of a
discrepancy from their expectations. Psychoneuroendocrinology doi:
Kaiser S and Wehrle T 2001 Facial expressions as indicators
of appraisal processes. In: Scherer KR, Schorr A and Johnstone T
(eds) Appraisal Processes in Emotion: Theory, Methods, Research
pp 285-300. Oxford University Press: New York, USA
Kendrick KM, da Costa AP, Leigh AE, Hinton MR and Peirce
JW 2001 Sheep don’t forget a face. Nature 414: 165-166
Manteuffel G 2006 Positive emotions of animals: problems and chan-
ces of scientifically grounded welfare-improvement. Aktuelle Arbeiten zur
artgemäßen Tierhaltung, KTBL-Schrift 448
Mason JW 1971 A re-evaluation of the concept of ‘non-specifi-
city’ in stress theory. Journal of Psychiatry Research 8: 323-333
Nickel R, Schummer A and Seiferle E 1968 Lehrbuch der
Anatomie der Haustiere, Volume 1. Paul Parey Verlag: Berlin, Germany.
[Title translation: Handbook of anatomy of domestic animals]
Quaranta A, Siniscalchi M and Vallortigara G 2007
Asymmetric tail-wagging responses by dogs to different emotive
stimuli. Current Biology 17: 199-201
Reefmann N, Bütikofer Kaszàs F, Wechsler B and Gygax
L 2009 Ear and tail postures as indicators of emotional valence in
sheep. Applied Animal Behaviour Science 118: 199-207
Sander D, Grandjean D and Scherer KR 2005 A systems
approach to appraisal mechanisms in emotion: emotion and brain.
Neural Networks 18: 317-352
Scherer KR 1999 Appraisal theories. In: Dalgleish T and Power
M (eds) Handbook of Cognition and Emotion pp 637-663. Wiley:
Chichester, UK
Scherer KR, Shorr A and Johnstone T 2001 Appraisal
Processes in Emotion: Theory, Methods, Research. Oxford University
Press: Canary, NC, USA
Schmied C, Waiblinger S, Scharl T, Leisch F and Boivin
X 2008 Stroking of different body regions by a human: effects on
behaviour and heart rate of dairy cows. Applied Animal Behavioural
Sciences 09: 25-38
Veissier I, Boissy A, Désiré L and Greiveldinger L 2009
Animals’ emotions: studies in sheep using appraisal theories.
Animal Welfare 18: 347-354
Wager TD, Phan KL, Liberzon I and Taylor SF 2003
Valence, gender, and lateralization of functional brain anatomy in
emotion: a meta-analysis of findings from neuroimaging.
NeuroImage 19: 513-531
Wemelsfelder F 2001 The inside and outside aspects of cons-
ciousness: complementary approaches to the study of animal
emotion. Animal Welfare 10: 129-139
Wickham SL, Collins T, Barnes AL, Beatty DT, Miller D,
Stockman C and Fleming PA 2009 Qualitative behavioural
assessment of sheep naïve and habituated to transport. In:
Hausberger M (ed) Proceedings of the 31st International Ethological
Conference pp 455-456. Rennes University, Rennes, France
© 2011 Universities Federation for Animal Welfare
... Studies in species including pigs have reported an increase in ears being held back during / following negative experiences (e.g. sheep (Bellegarde et al., 2017;Boissy et al., 2011;Guesgen et al., 2016), pigs (Reimert et al., 2013)), whereas ears forward has been shown to indicate heightened vigilance (e.g. sheep (Boissy et al., 2011;Reefmann et al., 2009), dogs (Racca et al., 2012)). ...
... sheep (Bellegarde et al., 2017;Boissy et al., 2011;Guesgen et al., 2016), pigs (Reimert et al., 2013)), whereas ears forward has been shown to indicate heightened vigilance (e.g. sheep (Boissy et al., 2011;Reefmann et al., 2009), dogs (Racca et al., 2012)). In this study, after treatment, the time spent with ears back was reduced and the time with the ears forward was increased suggesting a heightened alertness to the surroundings outwith the observation box (i.e. ...
... increased vigilance towards the (present but not visible) handler (Boissy et al., 2011;Racca et al., 2012;Reefmann et al., 2009) respectively. Potential behavioural evidence of pain was limited to an increase in the time spent investigating the wood shavings after sham-handling that did not occur after tooth clipping or grinding, and a greater occurrence of champing in tooth clipped compared with sham ground piglets. ...
... Both the LGS and the SPFES consider ears that are tense and pointed backwards or downwards as a reliable sign of pain in sheep [54,56]. Ears pointed backwards could, however, as well be a sign of an uncomfortable situation or fear [57]. The SGS describes a slightly different scale with erect ears as a sign of no pain, flattened ears as moderate indication of pain, and hanging ears as severe pain [55]. ...
... The SGS describes a slightly different scale with erect ears as a sign of no pain, flattened ears as moderate indication of pain, and hanging ears as severe pain [55]. Yet, when sheep are being brushed, horizontal and backward ears with only few ear posture changes seem to reflect a neutral or even positive state [31,[57][58][59]. The breed characteristics may also be an important factor in interpreting ear posture, as ear posture may vary between breeds [31]. ...
Full-text available
The value society assigns to animal welfare in agricultural productions is increasing, resulting in ever-enhancing methods to assess the well-being of farm animals. The aim of this study was to review the scientific literature to obtain an overview of the current knowledge on welfare assessments for sheep and to extract animal-based welfare indicators as well as welfare protocols with animal-based indicators. By title and abstract screening, we identified five protocols and 53 potential indicators from 55 references. Three out of the five protocols include animal-based as well as resource-based indicators. All of them were assessed as being practicable on-farm but lacking reliability. Some of the single indicators are endorsed by the literature and widely used in the field like assessment of behaviour, lameness or body condition score. Others (e.g., Faffa Malan Chart FAMACHA©, dag score or pain assessment) are regularly mentioned in the literature, but their reliability and usefulness are still subject of discussion. Several indicators, such as pruritic behaviour, eye condition, lying time or tooth loss are relatively new in the literature and still lack evidence for their validity and usefulness. This literature review serves as a starting point for the development of valid and practicable welfare protocols for sheep.
... Many mammals possess a rich musculature to move their ears [42]. Postures of ears have been investigated in some domestic species, such as horses [21], sheep [43], and dogs [20], in order to assess communicative value or more general expression of emotions. The position of the ears has been shown to be the more important component of the configuration than the position of the tail although the tail-up display is rather conspicuous. ...
... The position of the ears has been shown to be the more important component of the configuration than the position of the tail although the tail-up display is rather conspicuous. In contrast to what was observed in sheep [43], in cats, as in dogs and horses, the movements of ears are symmetrical within dyads. Only the erect ear position in both cats indicated a positive emotion leading to a positive outcome of the interaction. ...
Full-text available
Visual communication involves specific signals. These include the different positions of mobile body elements. We analyzed visual configurations in cats that involve ears and the tail. We aimed at deciphering which features of these configurations were the most important in cats’ interactions with other cats and with humans. We observed a total of 254 cat–cat interactions within a sample of 29 cats, during a total of 100 h of observation scheduled with the “Behavioral dependent onset of sampling” method and using the “All occurences” sampling method. In addition, we sampled 10 interactions between cats and humans. In cat–cat interactions, we noted the positions of ears and tail of both protagonists, as well as the outcome of the interaction, which was either positive/neutral or negative. In a great majority of the 254 interactions sampled, both cats held their tail down. On the contrary, ear position was a critical element in predicting the outcome. When both partners held their ears erect, the outcome was significantly positive, such as rubbing or close proximity. In all other cases of the position of ears in both cats, the outcome was negative, with increased distance of the partners. Although the tail did not seem to play a significant role in visual configurations in cat interactions, the “tail-up” display was important when a cat approached a human being. In the vast majority of cases the cat rubbed itself on a human’s leg(s). Thus, we may conclude that the presence of a human has a specific meaning in the cat’s world, probably as the result of a long period of commensalism. It is important for pet owners to understand the signals that cats use with other cats and with humans in order to promote the welfare of cats.
... Emotions are relatively short-term affective responses (Mendl et al. 2010) triggered by events or stimuli of personal relevance (Gygax 2017). While much evidence indicates that at least mammalian species experience emotional states (e.g., Bennett et al. 2017;Boissy et al. 2011;Caeiro et al. 2017;De Oliveira and Keeling 2018;Dolensek et al. 2020;Hintze et al. 2016), inferring which emotion an animal may be experiencing is challenging (Gähwiler et al. 2020). Triangulating information from different sources, including context and the emotion components physiology, action tendencies, and behavioural expressions (Scherer 2005), can help to infer emotional states in animals (Mills 2017). ...
... Facial expressions are a key to identifying human emotions (see, e.g., Darwin 1872; Ekman and Rosenberg 2005;Matsumoto et al. 2008;Scherer et al. 2013) and have also been examined in animals (e.g., cows (De Oliveira and Keeling 2018;Sandem et al. 2006), pigs (Camerlink et al. 2018), sheep (Boissy et al. 2011;Reefmann et al. 2009); bonobos (Demuru et al. 2015), mice (Defensor et al. 2012), rats (Finlayson et al. 2016), cats (Bennett et al. 2017), and dogs (Bremhorst et al. 2019;Caeiro et al. 2017); for a review on facial expressions of non-human animals, see Descovich et al. 2017). Facial expressions can be considered as reflecting emotional states if they are produced regardless of contextual features whenever a particular emotional state is experienced (see Kraut and Johnston 1979;e.g., in response to emotionally competent stimuli (Caeiro et al. 2017) such as food (Kaminski et al. 2017)). ...
Full-text available
Facial expressions potentially serve as indicators of animal emotions if they are consistently present across situations that (likely) elicit the same emotional state. In a previous study, we used the Dog Facial Action Coding System (DogFACS) to identify facial expressions in dogs associated with conditions presumably eliciting positive anticipation (expectation of a food reward) and frustration (prevention of access to the food). Our first aim here was to identify facial expressions of positive anticipation and frustration in dogs that are context-independent (and thus have potential as emotion indicators) and to distinguish them from expressions that are reward-specific (and thus might relate to a motivational state associated with the expected reward). Therefore, we tested a new sample of 28 dogs with a similar set-up designed to induce positive anticipation (positive condition) and frustration (negative condition) in two reward contexts: food and toys. The previous results were replicated: Ears adductor was associated with the positive condition and Ears flattener, Blink, Lips part, Jaw drop, and Nose lick with the negative condition. Four additional facial actions were also more common in the negative condition. All actions except the Upper lip raiser were independent of reward type. Our second aim was to assess basic measures of diagnostic accuracy for the potential emotion indicators. Ears flattener and Ears downward had relatively high sensitivity but low specificity, whereas the opposite was the case for the other negative correlates. Ears adductor had excellent specificity but low sensitivity. If the identified facial expressions were to be used individually as diagnostic indicators, none would allow consistent correct classifications of the associated emotion. Diagnostic accuracy measures are an essential feature for validity assessments of potential indicators of animal emotion.
... The results indicated that these behaviors are closely associated with negative emotions (Mendl et al., 1997). Numerous studies have shown that tail postures are indicators of behavioral and psychological responses Reefmann et al., 2009;Boissy et al., 2011;Jones and Boissy, 2011;Marcet-Rius et al., 2019). ...
ABSTRACT We evaluated the maternal behavior, physiology, and reproductive performance of both Damin (Min-pig × Large White) and Large White gilts to identify the advantages hybrid sows offer with regard to stress relieve and improvement of the welfare level of sows during late lactation. First-parity Damin gilts (n = 40) and first-parity Large White gilts (n = 40) were farrowed in individual pens. Video surveillance was used to monitor the occurrence of lateral recumbency and compare it to other postures, such as ventral recumbency, defecation, urination, tail posture, sham-chewing, and bar-biting behaviors. Monitoring was conducted from 07:00 to 09:00 h and from 13:00 to 15:00 h on days 3 and 6 of each week from the third to the fifth week post-parturition. In addition, the concentrations of tumor necrosis factor-α, interleukin-6, and salivary α-amylase were assessed. During the fourth week postpartum, Damin gilts showed a higher frequency of postural changes from lateral recumbency to other postures and less ventral recumbency, sham-chewing, and bar-biting behavior compared with Large White gilts. However, no significant differences were found between Damin and Large White gilts with regard to urination, defecation, tail wagging, and “tail low” behaviors. The concentrations of serum interleukin-6, salivary α-amylase, and serum tumor necrosis factor-α were higher in Damin gilts than in Large White gilts during the fifth week postpartum. Damin gilts partly achieve lower stress levels during late lactation and better animal welfare than purebred Large White gilts.
... However, we also found differences that do not point to a higher stress level in barren pigs. Enriched pigs defecated/urinated more often and kept their ears more in a backward position, which could suggest that these pigs had a less positive emotional state than barren pigs during the test (35,82,83). It could be that the Frustration test led to different responses in barren and enriched pigs. ...
Full-text available
Resilience, the capacity of animals to be minimally affected by a disturbance or to rapidly bounce back to the state before the challenge, may be improved by enrichment, but negatively impacted by a high allostatic load from stressful management procedures in pigs. We investigated the combined effects of diverging environmental conditions from weaning and repeated mixing to create high allostatic load on resilience of pigs. Pigs were either exposed to barren housing conditions (B) from weaning onwards or provided with sawdust, extra toys, regular access to a “play arena” and daily positive human contact (E). Half of the pigs were exposed to repeated mixing (RM) and the other half to one mixing only at weaning (minimal mixing, MM). To assess their resilience, the response to and recovery from a lipopolysaccharide (LPS) sickness challenge and a Frustration challenge were studied. In addition, potential long-term resilience indicators, i.e. natural antibodies, hair cortisol and growth were measured. Some indications of more favorable responses to the challenges in E pigs were found, such as lower serum reactive oxygen metabolite (dROM) concentrations and a smaller area under the curve of dROM after LPS injection. In the Frustration challenge, E pigs showed less standing alert, escape behaviors and other negative behaviors, a tendency for a smaller area under the curve of salivary cortisol and a lower plasma cortisol level at 1 h after the challenge. Aggression did not decrease over mixings in RM pigs and was higher in B pigs than in E pigs. Repeated mixing did not seem to reduce resilience. Contrary to expectations, RM pigs showed a higher relative growth than MM pigs during the experiment, especially in the week of the challenges. Barren RM pigs showed a lower plasma cortisol concentration than barren MM pigs after the LPS challenge, which may suggest that those RM pigs responded less detrimentally than MM pigs. Enriched RM pigs showed a higher level of IgM antibodies binding keyhole limpet hemocyanin (KLH) than enriched MM and barren RM pigs, and RM pigs showed a sharper decline in IgG antibodies binding Bovine Serum Albumin (PC-BSA) over time than MM pigs. Hair cortisol concentrations were not affected by enrichment or mixing. To conclude, enrichment did not enhance the speed of recovery from challenges in pigs, although there were indications of reduced stress. Repeated as opposed to single mixing did not seem to aggravate the negative effects of barren housing on resilience and for some parameters even seemed to reduce the negative effects of barren housing.
... By contrast, changes in ear posture seem to be associated with increased negative emotions or decreased positive emotions in pigs (Reimert et al., 2013). The ear postures of sheep (Ovis arie) showed consistent changes in situations involving sudden changes, stimulus familiarity, and controllability (Boissy et al., 2011). Sheep maintained their ears in a horizontal posture in a neutral state, but they pointed their ears backward in unpleasant, uncontrollable, unfamiliar situations (fear-eliciting situations), upward in negative but controllable situations (anger-eliciting situations), and displayed asymmetric ear posture in situations involving unexpected, sudden changes (surprising situations). ...
Mammals respond to an unexpected reward omission or reduction with a variety of behavioral and physiological responses consistent with an aversive emotion traditionally called frustrative nonreward. This review focuses on two aspects of frustrative nonreward, namely (1) the evidence for an aversive emotional state activated by the surprising omission or reduction of a rewarding outcome, and (2) the adaptive value of frustration. Frustrative nonreward has been mainly studied in terms of its mechanisms, across development in rats and across vertebrate species in comparative research. However, its adaptive function remains obscure. Following Domjan's approach to animal learning, this article explores a specific adaptive function hypothesis of frustrative nonreward called the incentive disengagement hypothesis. According to this hypothesis, the adaptive function of frustrative nonreward is to break an attachment to a site, situation, or stimulus that no longer yields appetitive resources (especially food and fluids) to promote the search for rewards in alternative locations. This function is of particular relevance given that mammals are especially vulnerable to reward loss due to their high metabolic rate and the energy demands of their relatively large brain.
... Facial expressions have also been studied as potential indicators of other negative and positive affective states. Most studies compared changes in the face during or immediately after exposure to a presumably positive or negative stimulus to a presumably neutral control situation, focusing on either several parts of the face [e.g., (15,16)] or only aspects, for example, the ears [e.g., (17,18)] or eyes [e.g., (19,20)]. The identified facial expressions are supposed to reflect stimulus-specific short-term affective states, i.e., emotions. ...
Full-text available
Identifying and validating behavioral indicators of mood are important for the assessment of animal welfare. Here, we investigated whether horses' eye wrinkle expression in a presumably neutral situation is a measure of mood as assessed in a cognitive judgment bias task (JBT). To this end, we scored pictures of the left and right eyes of 16 stallions for different aspects of eye wrinkle expression and tested the same individuals on a spatial JBT with active trial initiation. Eye wrinkle expressions were assessed by a qualitative assessment, i.e., the overall assessment of how “worried” horses look, the number of wrinkles, and the angle measured at the intersection of lines drawn through the eyeball and the topmost wrinkle. Correlations between the three eye wrinkle measures and the optimism index as a measure of horses' decisions in the JBT were not statistically significant, but with increasing optimism index, horses tended to be scored as looking less worried (qualitative assessment). We discuss our findings from different perspectives and make suggestions for future research, e.g., by calling for experimental induction of mood and thus greater variation within and/or between individuals and by investigating the interplay between shorter-lasting emotional and longer-lasting mood states to further explore the potential use of the JBT to validate eye wrinkles and other facial or body expressions as indicators of mood.
... Acute responses can be seen in attempts to escape from a situation, distress vocalisations, tachycardia and increased blood cortisol as observed when animals not used to handling are put in a holding cage [17]. More subtle signs can be seen in ear postures, with animals turning their ears back when exposed to a frightening stimulus [18]. When animals are submitted to prolonged stressful situations, their behavioural reactivity can be altered leading to hyperactivity or on the contrary to hyporeactivity (sometimes referred to as 'apathy') and the functioning of the corticotropic axis is modified [19]. ...
Full-text available
Maternal deprivation early in life has been shown to disrupt neonates’ development. Nevertheless, separating the young animals from their dams soon after birth remains a common practice in dairy farm husbandry. This study investigated the effects of different rearing conditions on goat kids’ stress coping abilities. Twenty female kids were raised together with their dams (‘dam-reared’) in a herd composed of other lactating goats and kids, while twenty female kids were separated from their dams three days after birth and reared together with same-age peers (‘artificially-reared’) and visually separated from the lactating herd. All kids shared the same father and two thirds of the kids were twins allocated to each treatment. At one month of age, kids were individually submitted to a series of tests: a novel arena test, a novel goat test, and a novel object test. These tests happened consecutively in this order, and lasted 180 seconds each. The kids’ behaviour was video-recorded and analysed post-hoc by an observer blind to treatments. Five weeks after weaning, the kids were also subjected to human-animal relationship tests. During the three behavioural tests, artificially-reared kids vocalized more (P < 0.001), reared more (P < 0.001), ran more (P = 0.002) and jumped more (P < 0.001) than dam-reared kids, but self-groomed less (P = 0.01) and urinated less (P = 0.05) than dam-reared kids. During the novel goat test and the novel object test, artificially-reared kids gazed less at the novel goat and the novel object (P = 0.02) and initiated contact more quickly (P = 0.05) with the novel goat and the novel object than dam-reared kids. The treatments however did not differ significantly in salivary cortisol response to the tests (P = 0.96). Artificially-reared kids showed significantly less avoidance of humans than dam-reared kids during the human-animal relationship tests after weaning (P < 0.001). The higher intensity of their behavioural reaction showed that artificially-reared kids react to stressful situations more actively than dam-reared kids. The difference between the three tests were only minor, suggesting a general change in the kids’ response to stressful situations rather than a specific change in their social response tested with an unfamiliar adult. Hence, artificial rearing affects goat kids’ behavioural response to challenges, probably maternal deprivation being the main factor.
Full-text available
One of the major topics of applied ethology is the welfare of animals reared by humans. Welfare can be defined as a state of harmony between an individual and its environment. Any marked deviation from this state, if perceived by the individual, results in a welfare deficit due to negative emotional experiences. In humans, verbal language helps to assess emotional experiences. In animals, only behavioural and physiological measurements help to detect emotions. However, how to interpret these responses in terms of emotional experiences remains an open question. The information on the cognitive abilities of farm animals, which are available but scattered, could help the understanding of their emotions. We propose a behavioural approach based on cognitive psychology: emotions can be investigated in farm animals in terms of the individual's appraisal of the situation. This evaluative process depends on: (a) the intrinsic characteristics of the eliciting event (suddenness, novelty, pleasantness); (b) the degree of conflict of that event with the individual's needs or expectations; and (c) the individual's coping possibilities offered by the environment. The result of such an evaluation determines the negative versus positive emotions. We propose an analysis of the emotional repertoire of farm animals in terms of the relationship between the evaluative process of the event on the one hand and the behavioural and physiological responses on the other hand.
Full-text available
This paper reviews the debate that is currently taking place in the field of philosophy of mind on different conceptual models of consciousness. More and more philosophers argue that the explanation of subjective phenomena requires two complementary perspectives of understanding, known as the first- and third-person perspectives. The third-person perspective (ie conventional objectivity) accounts for the physical, functional aspects of consciousness, while the first-person perspective addresses the subjective, experiential aspects of consciousness. It is suggested that each of these conceptual perspectives may facilitate a different type of research in the study of animal emotion. Within the conventional, third-person perspective, a growing enthusiasm for issues of animal consciousness has led to sophisticated physical and cognitive models of animal emotion. The potential of the first-person perspective, however, to provide a basis for models of animal subjective experience has remained largely unexplored. The paper concludes with a brief review of the author's recent experimental work on concepts of animal behavioural expression. The high reliability and repeatability of such concepts indicates that the first-person perspective may provide a valid research perspective in its own right.
Full-text available
Animal welfare concerns stem from recognition of the fact that animals can experience emotions such as pain or joy. Nevertheless, discussion of animal emotions is often considered anthropomorphic, and there is a clear need to use explanatory frameworks to under-stand animals' emotions. We borrowed appraisal theories developed in cognitive psychology to study sheep emotions. Emotions are viewed as the result of how an individual evaluates a triggering situation, following a sequence of checks, including the relevance of the situation (its suddenness, familiarity, predictability, and intrinsic pleasantness), its implications for the individual (including consistency with the individual's expectations), the potential for control, and both internal and external standards. We assumed that if the outcome of checks has an impact on the animal's emotional responses, then animals do not only show emotional responses but also feel emotions. We showed that sheep use similar checks to those used by humans to evaluate their environment, ie suddenness, familiarity, predictability, consistency with expectations, and control. Furthermore, this evaluation affects their emotional responses (behavioural responses, such as startle, ear postures, and cardiac activity). It is concluded that sheep are able to experience emotions such as fear, anger, rage, despair, boredom, disgust and happiness because they use the same checks involved in such emotions as humans. For instance, despair is triggered by situations which are evaluated as sudden, unfamiliar, unpredictable, discrepant from expectations, and uncontrollable, whereas boredom results from an overly predictable environment, and all these checks have been found to affect emotional responses in sheep. These results have implications for animal welfare: although a completely invariable and totally predictable environment should be avoided to prevent boredom, sudden events should probably be minimised, the animals should be offered the possibility to control their environment, and care should be taken to ensure a degree of predictability concerning the various events.
Full-text available
Background: In order to maintain cohesion of groups, social animals need to process social information efficiently. Visual individual recognition, which is distinguished from mere visual discrimination, has been studied in only few mammalian species. In addition, most previous studies used either a small number of subjects or a few various views as test stimuli. Dairy cattle, as a domestic species allow the testing of a good sample size and provide a large variety of test stimuli due to the morphological diversity of breeds. Hence cattle are a suitable model for studying individual visual recognition. This study demonstrates that cattle display visual individual recognition and shows the effect of both familiarity and coat diversity in discrimination. Methodology/principal findings: We tested whether 8 Prim'Holstein heifers could recognize 2D-images of heads of one cow (face, profiles, (3/4) views) from those of other cows. Experiments were based on a simultaneous discrimination paradigm through instrumental conditioning using food rewards. In Experiment 1, all images represented familiar cows (belonging to the same social group) from the Prim'Holstein breed. In Experiments 2, 3 and 4, images were from unfamiliar (unknown) individuals either from the same breed or other breeds. All heifers displayed individual recognition of familiar and unfamiliar individuals from their own breed. Subjects reached criterion sooner when recognizing a familiar individual than when recognizing an unfamiliar one (Exp 1: 3.1+/-0.7 vs. Exp 2: 5.2+/-1.2 sessions; Z = 1.99, N = 8, P = 0.046). In addition almost all subjects recognized unknown individuals from different breeds, however with greater difficulty. Conclusions/significance: Our results demonstrated that cattle have efficient individual recognition based on categorization capacities. Social familiarity improved their performance. The recognition of individuals with very different coat characteristics from the subjects was the most difficult task. These results call for studies exploring the mechanisms involved in face recognition allowing interspecies comparisons, including humans.
To date, most studies on animal emotions have focused on the assessment of negative emotional states, and there is a lack of approaches to characterising positive emotional states. The aim of this investigation was to measure differences in ear and tail postures in sheep exposed to situations likely to induce states of negative, intermediate and positive emotional valence.Nineteen female sheep were observed in emotion-eliciting situations in two experiments. In the home-pen experiment, ear and tail postures were observed during separation from group members (negative situation), during rumination (intermediate), and while feeding on fresh hay (positive situation). In the fodder experiment, individual sheep were conditioned to anticipate the delivery of standard feed. Once familiar with this experimental condition, they were offered either the standard feed (control treatment), unpalatable wooden pellets (negative treatment), or energetically enriched feed mixed with preferred feed items (positive treatment). Ear and tail postures of sheep were recorded during the final 6 min preceding feed delivery (anticipation phase) and for 6 min during feed delivery (feeding phase). Data were analysed using linear mixed-effect models.In the home-pen experiment, sheep separated from group members showed a high number of ear-posture changes and a high proportion of forward ears compared to hay feeding, during which ears were mainly passive. In the fodder experiment, the total number of ear-posture changes was generally high during the anticipation phases, slightly lower during delivery of the wooden pellets, and clearly reduced during the delivery of standard and enriched feed. A higher proportion of passive ear postures occurred when standard feed and enriched feed were offered compared to the delivery of wooden pellets. The proportion of asymmetric and axial ear postures was influenced by the sequence of testing of the different feeding treatments, with a higher proportion of asymmetric and a lower proportion of axial ear postures during the first exposure to either the wooden pellets or the enriched feed. A high proportion of the sheep's tails being raised was only observed during separation from group members.In both experiments, frequent ear-posture changes were most clearly associated with situations inducing negative states, and a high proportion of passive ear postures with situations likely to induce positive emotional states. Unfamiliarity influenced emotional reactions towards a more negative appraisal. A raised tail only appears to occur in specific situations, and was not useful for distinguishing emotional valence. Apart from the need for further validation, observations of ear-posture changes seem to be a promising approach for assessing emotional reactions in sheep.
From observations of intra-specific social grooming in cattle and studies on human stroking in other species, we hypothesised that cows’ reactions to human stroking differ depending on the body regions being stroked. Moreover, we tested, whether cows ‘reactions to stroking change with the animals’ experience of stroking.Sixty dairy cows were stroked in three different body regions, i.e. the withers, W, neck ventral, NV (both licked often in social grooming) and the lateral chest, LC (licked rarely), in a balanced order during 10-min sessions. Behavioural reactions and heart rate during stroking as well as reactions to the human just after stroking were recorded. Two test sessions were carried out with 3 weeks of treatment in-between. During this period, the cows were randomly allocated to four treatment groups: three groups received 5min of daily stroking in either W, NV or LC and the last one (control group) was exposed to simple human presence.During stroking W and NV, cows showed longer neck stretching and ear hanging than during stroking LC (P
The identification of the cognitive processing by which animals evaluate their environment helps to predict situations detrimental to their welfare. Appraisal theories developed in cognitive psychology offer a framework to study such cognitive processing. Here we investigated whether the controllability of an aversive event (an airblast and a sliding grate preventing access to a food reward) affected emotional responses in lambs. The animals could (vs. could not) interrupt the aversive event and thus gain access to food by performing an operant task (placing their muzzle in an aperture). Among lambs trained to perform the operant task, seven learnt it completely and six partially (i.e. they approached their muzzle to the aperture). Each of the 13 lambs that learnt the task completely or partially was paired with a “yoked” partner not taught how to interrupt the aversive event. Behaviour, cortisol and cardiac activity were recorded and the groups were compared with ANOVAs for mixed models. Compared with the lambs unable to interrupt the aversive event, the lambs taught to control it were more inclined to enter and stay in the test arena, and more inclined to eat there. These differences were generally more marked in pairs where the operant task had been fully learnt. An occurrence of the aversive event was followed by a transient backwards-pointing position of the ears and an increased heart rate in all the lambs. These responses were less pronounced in controlling lambs that had completely learnt the operant task. We show that an aversive situation is perceived as less stressful by sheep when they can exert control over it and this effect depends on the degree of control.
Appraisal theories developed in cognitive psychology are used here to attempt to better understand emotional experiences in animals. We investigated whether lambs are able to form expectations and whether their emotional responses are affected by situations discrepant from the expectations they may have formed. Forty-five female lambs were trained to obtain a small or a large amount of food reward by performing an operant task (introducing their muzzle into a hole). Then, half the lambs were shifted to the large or the small reward (i.e. positive or negative shift respectively), while the remaining half continued to get the same amount of reward. Thereafter, the lambs previously submitted to a reward change were shifted back to their initial amount of reward (i.e. successive shifts) while the lambs previously maintained on the same amount of reward were subjected to extinction (no reward, thus a negative shift). Behavior, cortisol levels and cardiac activity were analyzed, and the treatments were compared with ANOVAs for mixed models. When the amount of reward delivered was decreased, the lambs showed more locomotor activity and performed the operant task at a higher frequency but less efficiently, and there was a decrease in the parasympathetic influence on their cardiac activity. These responses were exacerbated when the negative shift followed a positive one. Similar responses were observed under extinction, and these responses were more pronounced when animals were trained with a large amount of reward before extinction. In response to a positive shift, we noticed a decrease in the frequency of the attempted operant task; this occurred only when the positive shift followed a negative one. Variations in plasma cortisol were not consistent with changes in the amount of reward. This study shows that lambs evaluate a reward according to their previous experience with that reward. They are able to form expectations, and a discrepancy from these expectations influences emotional responses, especially in the case of a negative shift. Given the appraisal criteria used by lambs and the matching emotions, we can assume that the emotional response to a negative shift expressed by lambs could reflect the despair caused by frustration.