Cattle discriminate between familiar and unfamiliar conspecifics by using only head visual cues

Abstract

Faces have features characteristic of the identity, age and sex of an individual. In the context of social communication and social recognition in various animal species, facial information is relevant for discriminating between familiar and unfamiliar individuals. Here, we present two experiments aimed at testing the ability of cattle (Bos taurus) to visually discriminate between heads (including face views) of familiar and unfamiliar conspecifics represented as 2D images. In the first experiment, we observed the spontaneous behaviour of heifers when images of familiar and unfamiliar conspecifics were simultaneously presented. Our results show that heifers were more attracted towards the image of a familiar conspecific (i.e., it was chosen first, explored more, and given more attention) than towards the image of an unfamiliar one. In the second experiment, the ability to discriminate between images of familiar and unfamiliar conspecifics was tested using a food-rewarded instrumental conditioning procedure. Eight out of the nine heifers succeeded in discriminating between images of familiar and unfamiliar conspecifics and in generalizing on the first trial to a new pair of images of familiar and unfamiliar conspecifics, suggesting a categorization process of familiar versus unfamiliar conspecifics in cattle. Results of the first experiment and the observation of ear postures during the learning process, which was used as an index of the emotional state, provided information on picture processing in cattle and lead us to conclude that images of conspecifics were treated as representations of real individuals.

Full text

1 23
Animal Cognition
ISSN 1435-9448
Volume 14
Number 2
Anim Cogn (2010) 14:279-290
DOI 10.1007/
s10071-010-0361-6
Cattle discriminate between familiar and
unfamiliar conspecifics by using only head
visual cues

1 23
Your article is protected by copyright and
all rights are held exclusively by Springer-
Verlag. This e-offprint is for personal use only
and shall not be self-archived in electronic
repositories. If you wish to self-archive your
work, please use the accepted author’s
version for posting to your own website or
your institution’s repository. You may further
deposit the accepted author’s version on a
funder’s repository at a funder’s request,
provided it is not made publicly available until
12 months after publication.

ORIGINAL PAPER
Cattle discriminate between familiar and unfamiliar conspecifics
by using only head visual cues
Marjorie Coulon •Claude Baudoin •
Yvan Heyman •Bertrand L. Deputte
Received: 13 June 2010 / Revised: 17 November 2010 / Accepted: 19 November 2010 / Published online: 4 December 2010
ÓSpringer-Verlag 2010
Abstract Faces have features characteristic of the iden-
tity, age and sex of an individual. In the context of social
communication and social recognition in various animal
species, facial information is relevant for discriminating
between familiar and unfamiliar individuals. Here, we
present two experiments aimed at testing the ability of
cattle (Bos taurus) to visually discriminate between heads
(including face views) of familiar and unfamiliar conspe-
cifics represented as 2D images. In the first experiment, we
observed the spontaneous behaviour of heifers when ima-
ges of familiar and unfamiliar conspecifics were simulta-
neously presented. Our results show that heifers were more
attracted towards the image of a familiar conspecific (i.e., it
was chosen first, explored more, and given more attention)
than towards the image of an unfamiliar one. In the second
experiment, the ability to discriminate between images of
familiar and unfamiliar conspecifics was tested using a
food-rewarded instrumental conditioning procedure. Eight
out of the nine heifers succeeded in discriminating between
images of familiar and unfamiliar conspecifics and in
generalizing on the first trial to a new pair of images of
familiar and unfamiliar conspecifics, suggesting a catego-
rization process of familiar versus unfamiliar conspecifics
in cattle. Results of the first experiment and the observation
of ear postures during the learning process, which was used
as an index of the emotional state, provided information on
picture processing in cattle and lead us to conclude that
images of conspecifics were treated as representations of
real individuals.
Keywords Social familiarity Instrumental
conditioning Cattle Spontaneous behavioural choice 
Simultaneous discrimination task 2D images of head
Introduction
Social recognition can be defined as the ability of indi-
viduals to categorize conspecifics into different classes
such as familiar versus unfamiliar, kin versus non kin, and
dominant versus subordinate (Gheusi et al. 1994). Social
recognition has several functions such avoiding inbreeding
and maintaining a dominance hierarchy which in turn
reduces occurrences of social contests. Discrimination
between individuals implies that the subject considers two
conspecifics as different, but it does not necessarily imply
true individual recognition (Zayan 1994; Tibbetts and Dale
2007). In a large group where individuals interact with
many conspecifics, individual recognition may be difficult
and a mode of signalling of familiar versus unfamiliar
individuals may be used. For example, in social insects,
individuals originating from the same nest discriminate
between their nestmates and alien adults or larvae (Isingrini
et al. 1985).
Social recognition involves a process of category for-
mation between categories of familiar conspecifics and of
unfamiliar conspecifics. Discrimination and categorization
reduce the amount of cognitive demand by simplifying the
M. Coulon (&)C. Baudoin B. L. Deputte
Laboratoire d’Ethologie Expe
´rimentale et Compare
´e,
Universite
´Paris 13, 99 Avenue J-B. Cle
´ment,
93430 Villetaneuse, France
e-mail: marjoriecoulon82@yahoo.fr
M. Coulon Y. Heyman
UMR 1198 Biologie du De
´veloppement et Reproduction,
INRA, 78352 Jouy-en-Josas, France
B. L. Deputte
De
´partement d’Ethologie, Ecole Nationale Ve
´te
´rinaire d’Alfort,
94704 Maisons-Alfort, France
123
Anim Cogn (2011) 14:279–290
DOI 10.1007/s10071-010-0361-6
Author's personal copy

complexity of the changing environment of the animals
(Zayan and Vauclair 1998). The discrimination between
familiar and unfamiliar conspecifics may result from a
socialization process, which leads to a direct familiariza-
tion between animals in the group at a very early age
(Deputte 2000). This implies that animals learn the phe-
notypical characteristic of familiar conspecifics, generally
using one dominant sensory modality, such as olfaction
in rodents (Gheusi et al. 1994), audition in pinnipeds
(Charrier et al. 2001) or vision in sheep (Kendrick et al.
1995,1996) and in various primate species (Dasser 1987;
Pokorny and de Waal 2009).
Faces are one of the most relevant visual configurations
for social recognition because they carry information about
age, sex and individual identity of an animal (for example
in nonhuman primates, Parr et al. 2000). The ability to use
information available in faces plays an important role in
discriminating between familiar and unfamiliar conspecif-
ics. Among invertebrates, Tibbetts (2002) demonstrated
that wasps recognize an individual visually using its head
pattern. Recently, Van der Velden et al. (2008) showed that
crayfish recognize an individual previously met during a
fight and that this recognition is based on facial width. In
various vertebrate species, face recognition is the process
most commonly used to achieve visual social recognition.
Experimental studies in birds, sheep and primates have
used 2D photographs of faces as stimuli (reviews in Bovet
and Vauclair 2000; Zayan and Vauclair 1998). In addition,
compared to non-face object discrimination, face discrim-
ination has been shown the most efficient (Farah et al.
1998; Racca et al. 2010) and specific since the presentation
of an upside down face decreases the discrimination per-
formance (inversion effect, e.g., in chimpanzees, Parr and
Heintz 2008; Parr et al. 1999).
The images of faces also elicited specific neurophysio-
logical responses with the temporal cortical cells
responding specifically to faces in several mammalian
species (Tate et al. 2006). However, the evaluation of the
understanding of the picture by the animals has rarely been
investigated in the studies of face discrimination (Fagot
2000). Indeed, animals may either have the ability to pro-
cess the picture as a representation of the real individual or
they may simply mistake the picture with the real indi-
vidual (Fagot et al. 2010; Parron et al. 2008). Observation
of the behavioural responses elicited by the images can
inform us what these pictures mean for the animals. For
example, Parron et al. (2008) showed that pictorially naı
¨ve
baboons and gorillas confused the real object with their
pictures, as they ate the banana picture. In contrast, the
pictorially naı
¨ve chimpanzees may process the picture as
referential stimuli as they did not eat the banana picture.
According to Fagot et al. (2010), the confusion mode is the
first step in pictorially naı
¨ve animals before developing the
ability to establish equivalence between the real individual
and its pictorial representation. The behavioural responses
of pictorially naı
¨ve animals may provide useful informa-
tion on the perceptual and emotional processes involved
during social recognition (Bovet and Vauclair 2000; Houpt
2005).
Cattle are a good model for studying facial discrimina-
tion between familiar and unfamiliar conspecifics. They are
social animals living in stable groups and there is indirect
evidence of discrimination based on familiarity. For
example, when forming a new group, Bouissou and
Andrieu (1978) showed that heifers were less aggressive
towards familiar members than towards unfamiliar ani-
mals. Cattle also easily discriminate among familiar con-
specifics (Hagen and Broom 2003) or among humans
(Taylor and Davis 1998). Cattle appear to use their vision
for discriminating between conspecifics as altering their
vision ability induced an increase in aggressive interactions
(Cummins 1991). In a pioneering study, it was observed
that cows are more attentive to images of familiar animals
(Sato and Yoshikawa 1996). Cattle have good visual acuity
(1/20, Entsu et al. 1992) and colour sensitivity (especially
for high wavelengths, Phillips 2002), although not as good
as humans or sheep have. However, cattle have a large
visual field (330 degrees, Phillips 2002) and good capaci-
ties for visual discrimination learning (Rehka
¨mper and
Go
¨rlach 1997). Cattle display a significant phenotypic
variability between and among breeds. For example, the
Prim’Holstein, a dairy breed, is characterized by individ-
ually distinctive black and white spotted coat patterns and
the Normande breed by individually distinctive brown and
white spotted coat patterns. In a previous study, using an
instrumental conditioning procedure and images of heads
of animals, we showed that heifers visually discriminated
their own species from other animal species (Coulon et al.
2007) and learned to recognize an individual whatever its
degree of familiarity was at the beginning of the testing
procedure (Coulon et al. 2009).
The aim of the present study was to study the discrim-
ination process between 2D images of heads of familiar
and unfamiliar conspecifics, based on a simultaneous dis-
crimination paradigm. We defined as familiar conspecifics
two animals that lived within the same social group and
were familiar to each others since group formation. We
first observed spontaneous behavioural reactions of picto-
rially naı
¨ve heifers to images of familiar and unfamiliar
conspecifics to see how they may understand the images. In
a second experiment, we evaluated the ability of heifers to
visually discriminate between familiar and unfamiliar
conspecifics and to group these two kinds of conspecifics
into two different categories. In order to address these two
issues, we used an instrumental conditioning procedure
associated with a food reward. Demonstrating category
280 Anim Cogn (2011) 14:279–290
123
Author's personal copy

formation requires us to demonstrate a between-category
discrimination and a within-category discrimination for the
two a priori categories (Deputte et al. 2001; Thompson
1995; Vauclair and Fagot 1996; Wasserman et al. 1988).
We showed in a previous study that cattle discriminate
within the category ‘‘familiar conspecifics’’ and within the
category ‘‘unfamiliar conspecifics’’ (Coulon et al. 2009).
Hence, the demonstration of a between-category discrimi-
nation, between familiar and unfamiliar individuals in this
study, is the final step to show that cattle are able to form
visual conspecifics categories. In the second experiment,
we obtained additional information from the observation of
error rates and ear postures during the discrimination tasks
to interpret the observed individual performances in visual
discrimination tests.
Materials and methods
Animals
All subjects were Prim’Holstein heifers born and raised in
the same experimental farm at INRA-UCEA Bressonvil-
liers (France). Before 6 months of age, all animals were
housed in individual stalls under similar conditions in the
same nursery. Afterwards, the animals were housed in a
more open housing system (8 m
2
per animal). Subjects had
unrestricted access to a standard diet (grass silage, hay,
corn straw and minerals) and water. Depending on the
season, there was artificial or natural light between 06:00
a.m. and 07:00 p.m. Each heifer was identified with an I.D.
number printed on an ear tag. In experiment 1, two groups
of heifers were tested: one group of seven heifers aged
between 8 and 12 months, weighing 242 ±6 kg (group
A), and one group of 16 heifers aged between 6 and
13 months, weighing 260 ±12 kg (group B). In the sec-
ond experiment, we tested nine heifers between 10 and
12 months weighing 272 ±14 kg at the beginning of the
study (group C). Each group lived in a different social
group composed of Prim’Holstein heifers of the same age.
Other breeds of cows (e.g., Normande) were also present
on the farm where the experiments were conducted. Thus,
the subjects of our study have seen other breeds of cows
than their own Prim’Holstein breed but without direct
physical contact with them.
Experiment 1: Spontaneous behaviour induced
by images in a simultaneous discrimination task
Stimuli
The stimuli were digitized colour posters (42 950 cm)
representing natural-size heads of heifers, in frontal view,
and of one pony adapted to the natural-size head of heifers.
The set of stimuli consisted of: (1) different images of two
familiar Prim’Holstein heifers living in the same social
group as the individuals tested, (2) five unfamiliar
Prim’Holstein and five Normande heifers living at another
farm and (3) an unfamiliar white and brown pony (similar
coloured patterns as the Normande cows). The proportion of
dark patches relatively to the total head surface area was
similar in the stimuli from the two categories (familiar con-
specifics vs. unfamiliar conspecifics, mean =63 vs. 90%,
ANOVA, F
1,5
=0.23, P=0.65). The original background
of all pictures was replaced by the same uniform back-
ground, a yellow tint mimicking the colour of straw
(D2C48A background, Adobe Photoshop Elements
Ó
).
Procedure
The day before testing, each animal was individually
familiarized for 5 min with the test pen, which was a
rectangular room (3 94 m) adjacent to the free stall where
the subjects were housed. Subjects were tested individually
while remaining in auditory and olfactory contact with
their conspecific group members. Subjects were introduced
into the pen for a 3-min test duration and could see two
images fixed on a metal fence at the back of the pen. The
two images were set at the eye level of subjects and 1.5 m
apart from each other and at 4 m from the pen entrance.
This setting did not allow for the heifer to use their bin-
ocular vision to see the two images at the same time. The
tests were filmed using a video camera (Sony DCR-
TRV80E PAL) facing the subject and placed at its eye
level. The behaviours were coded blindly according to the
side location of the familiar picture (or of the picture of the
Normande heifer) from video recording. The latency from
the start of the test to the time when the subject sniffed or
licked one of the two images was recorded and used as an
indicator of choice. The number of exploratory behaviours
(sniffing, licking that were coded each time that the subject
touched the image with its nose or its tongue) and the
duration of the visual exploration of each image (gaze and
head in the direction of the image with erected ears) were
also recorded. After testing, each heifer was returned to its
social group. No aggressive reaction was observed fol-
lowing their return. Each heifer was tested only once
because we wanted to consider spontaneous reaction of
naı
¨ve heifers towards images.
Discrimination between one image of a familiar individual
and one image of an unfamiliar one (Prim’Holstein breed)
A pair of stimuli (one image of a familiar individual and
one of an unfamiliar one) was randomly selected for each
animal of the group A from the sets of images. For three
Anim Cogn (2011) 14:279–290 281
123
Author's personal copy

subjects, the image of the familiar individual was placed on
their right and for the remaining four heifers on their left
side.
Discrimination between one image of an unfamiliar
Normande heifer and one of an unfamiliar pony
(this discrimination was used as control for helping
in the interpretation of the results)
A pair of stimuli (one image of an unfamiliar Normande
heifer and one image of a pony) was randomly selected for
each animal of the group B. Image of Normande heifer
instead of a Prim’Holstein one was used to test the pref-
erence for their own species (against other species) and not
their preference for the familiar breed. For half of the
subjects, the image of the Normande heifer was placed on
their right and for the other half, it was on their left.
Statistical analysis
Non-parametric statistics were used due to the relatively
small sample size and deviations from the assumption of
normality. In each discrimination task, behavioural reac-
tions to each image were compared using a Wilcoxon
signed-rank test. The test was two-tailed, and the alpha level
was set to 5% throughout the study. All results are presented
in the form Mean ±SE. All analyses were performed with
the statistical package Statistica
Ò
(Version 8, StatSoft).
Experiment 2: Discrimination between images
of familiar versus unfamiliar conspecifics using
an instrumental conditioning procedure
Stimuli
For each phase of the second experiment, we used 20 visual
stimuli. These included representation of heads of six
different heifers viewed from seven various angles (frontal
view, right and left profiles, right and left front views
and right and left back views, Fig. 1). We tested the
discrimination between one set of images from three
heifers and one set of images from three other heifers
(Fig. 1a, b). Each trial consisted in discriminating between
one stimulus from one set and one stimulus from the other
set. In phase 1 and 2, three pictured heifers were familiar to
the subject (living in the same group) and the other three
heifers (from another farm and unrelated to the subjects)
were unfamiliar to the subject. In phase 3, all the stimuli
represented other familiar heifers and in phase 4, all stimuli
represented other unfamiliar heifers. The proportion of
dark patches relatively to the total head area was similar in
stimuli from the two sets in each phase (familiar conspe-
cifics vs. unfamiliar conspecifics—phase 1 & 2—Mean =
76 vs. 62%, ANOVA, F
1,22
=1.65, P=0.21; phase
3—88 vs. 79%, ANOVA, F
1,18
=1.64, P=0.22; phase
4—76 vs. 75%, ANOVA, F
1,18
=0.03, P=0.86). All
subjects (group C) were naı
¨ve to the visual stimuli used in
experiment 2 but they had been familiarized with the
apparatus, pictures and the experimental procedure.
Protocol
We used the same instrumental conditioning apparatus and
procedure as in our previous simultaneous discrimination
studies in heifers using a stimulus?/stimulus-(S?/S-)
paradigm (Coulon et al. 2007,2009,2010). Complemen-
tary studies have been completed to test for a possible
effect of the experimenter and the relevance of stimuli
(Coulon et al. 2009). Subjects were tested individually in a
test pen (6 911 m) adjacent to the free stall. Two spot-
lights of 500 watt were used to illuminate the test pen and
provided a good illumination condition for the experi-
ments. In the test pen, the subject walked to a guillotine
gate at the end of a lane made of rows of straw bundles. At
Fig. 1 Examples of stimuli
used in the experiment 2.
Images of heads from various
views of three conspecific (a)to
be discriminated from images of
heads from various views of
three other conspecifics (b)
282 Anim Cogn (2011) 14:279–290
123
Author's personal copy

the gate, the heifer could see the two images placed at its
eye level. The stimuli were presented at a distance of 2.5 m
from the guillotine gate where subjects were making their
choice. This distance of 2.5 m was chosen according to
Entsu et al.’s (1992) study on visual acuity in cattle. For
each pair of stimuli, one stimulus was consistently asso-
ciated with a reward (S?) and the other was not (S-).
After the heifer had looked at both stimuli, the experi-
menter lifted the gate. The heifer could then walk towards
and pass by one of the 2 images. This image was then
considered as its choice. If the chosen stimulus was S?,it
was a correct choice, and the subject was rewarded with
food for pushing an opaque panel placed beyond the
stimulus. If the chosen stimulus was S-, the choice was
incorrect, and the opaque panel was blocked. To avoid any
olfactory bias, food rewards (blend of oat, wheat and soya
meals; so different from the standard meal) were always
placed behind both panels. The left/right position of the
rewarded stimulus was randomly balanced across trials.
Each subject was tested for two consecutive sessions of 10
trials in the morning. The subjects were just deprived of
food 1 h before testing. At least 48 h and a maximum of
72 h elapsed between each block of two sessions. The
experimental procedure of the second experiment included
four different phases.
Phase 1 The first phase of the second experiment inclu-
ded two steps: first training and then generalization. During
training, heifers had to discriminate between frontal views
of a familiar and an unfamiliar heifer. For each heifer, the
same pair of stimuli was used for every trial. This pair was
drawn from the frontal views of the three familiar con-
specifics and of the three unfamiliar ones. Animals were
trained with either a familiar individual (5 subjects) as S?
or an unfamiliar one as S?(4 subjects). The success cri-
terion was set as at least 80% of correct responses per
session of 10 trials, during two consecutive sessions (80%,
8/10, in two consecutive times, binomial test, P=0.01).
After reaching the learning criterion, animals next pas-
sed the generalization step using the whole set of 20
stimuli, 10 images of familiar heifers and 10 of unfamiliar
ones. The pair of stimuli (familiar vs. unfamiliar head of
heifers) was changed at each trial. For each animal, the
type of positive stimulus (familiar or unfamiliar heifers)
was the same as it was during the training. Each stimulus of
this pair could correspond to one of the various views:
frontal, profile, frontal and back views. The stimuli
were a pair of heads either from the same or a different
view (e.g., familiar conspecific in frontal view vs. unfa-
miliar conspecific in profile view). The criterion of success
was similar to that of the training step. We stopped testing
an animal when it did not reach the criterion after 20 ses-
sions (=200 trials).
Phase 2: Reversal learning In this second phase of
experiment 2, the subjects were required to reverse their
learning. The training and generalization procedure, the
stimuli and the criterion were similar to those of phase 1,
except that images of unfamiliar individuals were now used
as S?for the five animals previously tested with the image
of a familiar individual as S?in phase 1, and the image of
familiar individuals was now S?for the four other animals.
Phase 3: Control phase 1 with images of familiar con-
specifics, control of not identity matching (compared with
phase 1) The phase 3 included only a generalization step
with 20 new images of six other familiar individuals. The
animals were rewarded for choosing the image of three
heifers among the six. The testing procedure and the cri-
terion were the same as in phases 1 and 2.
Phase 4: Control phase 2 with images of unfamiliar con-
specifics, control of discrimination of familiar conspecifics
(compared with phase 3) Phase 4 was similar to phase 3
with images of unfamiliar conspecifics instead of familiar
conspecifics, and the animals were rewarded for choosing
the image of three unfamiliar heifers among six heifers.
One of the nine heifers (Ind. 2) was not tested for this last
phase because this cow became aggressive. Half of
the subjects started with phase 3 and the other half with
phase 4.
Behavioural observation
In various mammals, ear postures are linked to expression
of emotion (horse: Waring 2003; dog: Deputte 2010; Houpt
2005; farmed silver fox: Moe et al. 2006; sheep: Reefmann
et al. 2009). Forward pointing ears generally indicate a low
level of fear and positive anticipation (Moe et al. 2006). In
contrast, backward pointing ears indicate negative antici-
pation and a higher level of fear. During phases 1 and 2 of
the second experiment (generalization step), the position of
ears was noted after the barrier was lifted during each trial.
Four different ear positions were recorded: (1) pointing
forward, (2) in plane, (3) pointing backward and (4)
asymmetric ears. The frequency of each ear configuration
was recorded when heifers moved towards rewarded ima-
ges of either familiar or unfamiliar conspecifics.
Error rate corresponding to the various visual stimuli
In the first phase of the second experiment (generalization
step), we recorded the number of successes and errors for
each of the possible pairs of views (front–front views,
front-profile views, back–front views, etc.) of the two
stimulus individuals (familiar vs. unfamiliar) for each test
subject as well as for each stimulus. We computed the error
Anim Cogn (2011) 14:279–290 283
123
Author's personal copy

rate as the number of errors divided by the number of errors
plus the number of successes for each stimulus.
Statistical analysis
The dependent variable was the number of sessions needed
to reach the criterion. For the same reasons as in experi-
ment 1, non-parametric statistics were used. Performance
of heifers rewarded for images of a familiar heifer was
compared to that of heifers rewarded for images of an
unfamiliar heifer using a Mann–Whitney Utest in both
training and generalization steps. A Kruskal–Wallis test
was used to compare the performance of the heifers
respectively in phases 1 (generalization step), 3 and 4.
A Wilcoxon signed-ranks test was used to compare the
performance of subjects between various tasks of dis-
crimination (training vs. generalization tests, phase 1 vs.
phase 2 (reversal learning), phase 1 vs. phase 3 and phase 3
vs. phase 4) and to compare the frequency of each position
of ears, during the choice of either familiar or unfamiliar
S?. The same test was used to compare the number of
correct trials in the first session of 10 trials between phase 1
and phase 2 and the performances of heifers with the new
pair of stimuli to the situation with a pair of stimuli already
used in the training. We also used the Mann–Whitney
Utest to see whether the subjects failed more often when
the proportion of dark patches differed between the two
stimuli of one pair. All analyses were performed with the
statistical package Statistica
Ò
.
A general linear ANOVA Model was used to compare
the error rate corresponding to the various familiar and
unfamiliar individuals in images and the similarity of view
angles between head visual stimuli. The standardized
residuals did not meet the assumption of normality on a QQ
plot. Therefore, the data were log transformed prior to the
statistical tests. We checked the conditions of homogeneity
of residual variances. The error rates for each view of
similar pairs of stimuli (frontal–frontal views, profile–
profile views, front–front views and back–back
views) and for each stimulus were compared using
ANOVA tests. And we studied the pairwise correlation
between the proportion of dark patches of a stimulus and its
error rate. The ANOVA and correlation analyses were
performed with the statistical package JMP (Version 8.0.,
SAS Institute, Cary, NC).
All tests were two-tailed, and the alpha level was set to
5% all throughout. All means are presented in the form of
Mean ±SE.
Ethical note
Animal care and all procedures were completed in accor-
dance with the authorization 93-031 delivered by the
‘‘Pre
´fecture de la Seine-Saint-Denis, Direction De
´part-
ementale des services ve
´te
´rinaires’’, and the authorization
B91 332 101 of the French Ministry of Agriculture and the
EU directives. The protocol, registered as ‘‘protocol
06-002’’, was approved by the Regional Ethical Committee
of Paris-Sud.
Results
Experiment 1
Discrimination between one image of a familiar individual
and one image of an unfamiliar one
Five of the seven heifers approached the image of the
familiar conspecific and sniffed it first and explored it more
than that of unfamiliar one. Only one heifer chose the
image of the unfamiliar individual first and explored it the
most, and one heifer did not approach images at all.
Overall, the latency to approach the image of the familiar
heifer was significantly shorter than when the image of the
unfamiliar individual was chosen (M
familiar
=44 ±24 s
vs. M
unfamiliar
=87 ±24 s, Wilcoxon signed-ranks test:
T=1, P=0.046, N=7, Fig. 2a). Heifers generally
Fig. 2 Spontaneous behaviours induced by images of familiar and
unfamiliar conspecifics in experiment 1. Behavioural response
(Mean ?SE) of heifers to the presentation of images of either familiar
(white bar) or unfamiliar (black bar)Prim’Holstein heifers in a
spontaneous discrimination task: alatency of approaches, bnumber of
explorations and cduration of observation of images. N=7, *P\0.05
284 Anim Cogn (2011) 14:279–290
123
Author's personal copy

explored significantly more the images of the familiar
individual than those of the unfamiliar one (sniffing and
licking, close exploration: M
familiar
=5.43 ±1.3 vs.
M
unfamiliar
=2±0.93, Wilcoxon signed-ranks test: T=0,
P=0.043, N=7, Fig. 2b). Moreover, heifers spent sig-
nificantly more time observing images of familiar than of
unfamiliar heifers (M
familiar
=5.86 ±2.69 s vs. M
unfamiliar
=
2.14 ±0.77 s, Wilcoxon signed-ranks test: T=0, P=
0.043, N=7, Fig. 2c).
Discrimination between one image of an unfamiliar
Normande heifer and an unfamiliar pony
Four heifers of the 16 tested heifers did not approach the
image of the pony. One heifer did not approach any of the
images. Heifers approached the image of the Normande
heifer significantly sooner than the image of the pony
(M
heifer
=53.62 ±14.11 s vs. M
pony
=105.62 ±16.61 s,
T=21, P=0.027, N =16, Fig. 3a).
All heifers, but one, explored and spent more time
observing the image of the Normande heifer than that of the
pony. The number of explorations (sniffing and licking) was
greater for the image of the heifer than that of the image of
the pony. In addition, the image of the pony was never
licked at (M
heifer
=7.62 ±1.22 vs. M
pony
=2.19 ±0.63,
T=6, P\0.01, N=16, Fig. 3b). The tested heifers spent
significantly more time looking the image of the heifer than
that of the pony (M
heifer
=11.06 ±1.85 s vs. M
pony
=
7.56 ±2s,T=11, P=0.028, N=16, Fig. 3c).
Experiment 2
Phase 1
During the training, all nine heifers discriminated between
images of familiar and unfamiliar heifers regardless of the
image rewarded (Mean =4±0.5 sessions, Fig. 4). Dis-
crimination was completed in an average of 3.8 ±0.8
sessions for S?familiar heifers and 4.75 ±0.48 sessions
for S?unfamiliar heifers (range 2–6 sessions and 4–6,
respectively, according to the heifers; Mann–Whitney
Utest: U=7, N
1
=5, N
2
=4, P=0.46). In the
generalization step, eight of nine animals discriminated
between images of familiar and unfamiliar heifers
(Mean =6.25 ±1.61 sessions, Fig. 4). Criteria were
reached in an average of 7.25 ±2.49 sessions for S?
familiar heifers and 5.25 ±2.27 sessions for S?unfa-
miliar heifers (range 2–14 sessions and 2–12, respectively;
Mann–Whitney Utest: U=5.5, N
1
=N
2
=4, P=0.27).
Only one heifer failed to reach the criterion of success after
20 sessions, with images of familiar heifers as S?. The
performance of this subject has been discarded in the fol-
lowing phases because of failure to reach the criterion in
phase 1.
Fig. 3 Spontaneous behaviours induced by images of an unfamiliar
cow and a pony in the experiment 1. Behavioural response
(Mean ?SE) of heifers to the presentation of images of either an
unfamiliar Normande heifer (white bar) or a white and brown pony
(black bar) in a spontaneous discrimination task: alatency of
approaches, bnumber of explorations and cduration of observation of
images. N=16, *P\0.05, **P\0.01
Fig. 4 Number of sessions (Mean ?SE) necessary to reach criterion
during the training (white bar) and the generalization steps (black
bar) for the four phases of the experiment 2 (Phase 1 discrimination
between familiar and unfamiliar individuals, Phase 2 reversal
learning, Phase 3 discrimination between two groups of familiar
individuals and Phase 4 discrimination between two groups of
unfamiliar individuals). The performance of one subject has been
discarded because of failure to reach the criterion in Phase 1, N=8,
and one subject is missing in the last Phase 4, N=7, *P\0.05
Anim Cogn (2011) 14:279–290 285
123
Author's personal copy

Phase 2: Reversal learning
All eight heifers reached the criterion in the reversal
learning in 4 ±0.71 sessions in the training (range 2–7
sessions) and in 4.5 ±1.03 sessions in the generalization
step (range 2–11 sessions). The performances of heifers did
not differ between phase 1 and phase 2 (Wilcoxon signed-
ranks test: T=10, N=8, P=0.61 for training; T=5.5,
N=8, P=0.08 for generalization).
Individual differences between training
and generalization steps
The individual performances of heifers did not differ
between training and generalization (Wilcoxon signed-
ranks test: in phase 1, T=7.5, N=8, P=0.27; in phase
2, T=14, N=8, P=0.57). However, two of eight
heifers performed better during the generalization step than
in the training step in both phases 1 and 2, and three of the
eight heifers performed better in the generalization step
than in the training step in one of phases 1 or 2.
First trial analysis
The observation of the correct response at the first trial
provided additional information. All heifers tested were
successful in the first trial of the generalization step of the
phase 1 but four of the eight heifers failed in the first trial
of the reversal learning (generalization step, phase 2). The
performances of heifers in the first block (10 first trials), as
observed in the number of session needed to perform, did
not differ between phase 1 and phase 2 (Wilcoxon signed-
ranks test: 8.12 ±0.35 vs. 6.87 ±0.83, T=7.5, N=8,
P=0.14 for training and 6.25 ±0.59 vs. 6.5 ±0.5,
T=8, N=8, P=0.6 for generalization).
Phases 3 and 4: Control phases
All eight heifers reached the criterion in 10.25 ±1.46
sessions (range 5–17 sessions, Fig. 4) in phase 3 and
seven of the seven heifers reached the criterion in
12.57 ±1.50 sessions (range 7–20 sessions, Fig. 4)in
phase 4. The performances of the subjects differed
between phases 1 (generalization step), 3 and 4 (control
phases; Kruskal–Wallis test, V
2
=6, df =2, P=0.049).
Seven of the eight subjects reached the criterion in fewer
sessions when required to discriminate between familiar
and unfamiliar heifers than between two groups of
familiar heifers (phase 1 vs. phase 3, Fig. 4, Wilcoxon
signed-ranks test: T=0, P=0.018, N=7). Six of the
seven subjects reached the criterion in fewer sessions
when required to discriminate between two groups of
familiar heifers than between two groups of unfamiliar
heifers (phase 3 vs. phase 4, Fig. 4, Wilcoxon signed-
ranks test: T=0, P=0.018, N=7).
Behavioural observations
When subjects had to choose images of unfamiliar con-
specifics that were rewarded, they moved their ears back-
wards significantly more often than forward, (19 ±3 vs.
8±2%, Wilcoxon signed-ranks test: T=0, P=0.018).
In contrast, they pointed their ears forward significantly
more often than backwards when choosing and being
rewarded for images of familiar conspecifics (34 ±6 vs.
9±2%, Wilcoxon signed-ranks test: T=2, P=0.025).
The frequency of other ear positions (in the plane and
asymmetric) did not differ across the various discrimina-
tion tasks (S?familiar conspecifics: 30 ±3% in the plane
and 24 ±6% asymmetric; S?unfamiliar conspecifics:
33 ±2% in the plane and 40 ±6% asymmetric).
Error rates corresponding to various views angle of stimuli
In phase 1 of the second experiment, the error rate (ER)
was higher when the two head stimuli (familiar vs. unfa-
miliar conspecifics) displayed the same view angle
than when it was different (ER
same
=0.39 ±0.04 vs.
ER
different
=0.28 ±0.03, ANOVA: F
1,87
=5.73, P=
0.019). However, there was no interaction between the
identity of the familiar or unfamiliar individuals and the
similarity between view angles of the two stimuli
(ANOVA: F
2,87
=0.41 and F
3,87
=0.69, P[0.05). In
addition, the error rate did not differ according to the view
of a pair of stimuli, the view of a stimulus or according to
each stimulus (ANOVAs; ER
facial view
=0.26 ±0.05,
ER
profile view
=0.28 ±0.06, ER
3/4 front view
=0.30 ±0.05,
ER
3/4 back view
=0.37 ±0.04, F
3,32
=0.82, P[0.05;
ER
facial view
=0.32 ±0.03, ER
profile view
=0.30 ±0.03,
ER
3/4 front view
=0.45 ±0.04, ER
3/4 back view
=0.35 ±
0.03, F
3,157
=0.61, P=0.61 and F
19,141
=1.26,
P=0.22, respectively). In the generalization step, the
subjects did not fail more often with new pairs of stimuli
than with the pair of stimuli used during the training step
(ER
training_pair
=0.34 ±0.07 vs. ER
other_pairs
=0.31 ±
0.032, Wilcoxon signed-ranks test: T=19, P=0.68,
N=9) or with pairs that differed in the proportion of dark
patches (difference in the proportion of dark patches
between the two stimuli of a pair =32 ±1% when correct
choices vs. 31 ±1% when incorrect choices, Mann–
Whitney Utest: U=55,334, N
1
=453, N
2
=247,
P=0.81). The error rate of a stimulus was not correlated
with its proportion of dark patches (pairwise correlation,
r=0.08, P=0.72).
286 Anim Cogn (2011) 14:279–290
123
Author's personal copy

Discussion
This study demonstrated that heifers can discriminate
between familiar and unfamiliar conspecifics, using only
visual cues from 2D images of conspecifics’ heads. In the
first experiment with a simultaneous presentation of two
images, many heifers spontaneously approached the images
of familiar heifers more often than those of unfamiliar ones.
Heifers chose the picture of familiar conspecific quickly and
were relying on distant visual information. This result
indicates that images of familiar individuals were highly
relevant for the tested heifers. The sniffing and licking
behaviours, directed towards images of conspecifics likely
suggest that heifers may treat 2D representations of familiar
conspecifics as their real presence. Fagot (2000) proposed
that animals may use different modes for processing pic-
tures: confusion (the picture is perceived as the real object
for the subject), independence (there is no association
between the real objects and their pictures) and equivalence
(the picture is the representation of the real object) modes.
According to these possible modes of picture processing, it
is very likely that, at least for a short time at the beginning of
the short distance visual exploration, in absence of other
sensory information, heifers were processing pictures of
conspecifics under the confusion mode (Fagot 2000).
Rosenfeld and Van Hoesen (1979) also observed a confusion
period in the Rhesus monkey presented with an image of a
conspecific for the first time. Fagot et al. (2010) suggested
that the confusion mode is probably the first step in picto-
rially naı
¨ve animals before developing, through experience
with the stimuli, the ability to establish equivalence between
the real individual and its pictorial representation. In our first
experiment, the heifers were pictorially naı
¨ve and had to
choose pictures of a familiar conspecific as soon as they
entered into the arena. This suggests that heifers used pre-
viously stored visual social information and may treated
images of conspecifics as representations of real individuals.
Results of the control test (Normande vs. pony) are con-
gruent with this interpretation. The subjects, in both cases,
explored more and approached more quickly the more
ecologically relevant picture. The preference for familiarity
(known individuals) has been observed in other species
(review in Bovet and Vauclair 2000). Bouissou et al. (1996)
and Porter and Bouissou (1999) similarly showed that sheep
also use social representation in the discrimination process.
Our findings stand in contrast to studies on poultry by
Dawkins (1996), who showed that hens preferred flock-
mates rather than unfamiliar conspecifics when presented
with live stimuli, but not when presented with images of
conspecifics. The influence of ecological factors has been
also shown in the study of face processing (e.g., dog: Racca
et al. 2010; non human primate: Pascalis et al. 2002;
Neiworth et al. 2007). It has also been shown that the
development and the sex of the subjects or the stimuli could
have an effect on abilities and looking preference (human:
Lewkowicz et al. 2008; non human primate: Zangenehpour
et al. 2009; Lacreuse et al. 2007). Lacreuse et al. (2007)
showed a clear relationship between the hormonal state of
female rhesus monkeys and their preference for looking at
male faces of their own species. In our study, the subjects
and the individuals of the images were all immature female
animals (no bulls in the experimental farm). So, it would be
interesting to study now the effect of the age and the sex of
the subject or the stimulus on looking preference and per-
formance in cattle.
In the second experiment, after being trained to dis-
criminate between an image of a familiar individual and an
image of an unfamiliar individual, heifers were able to
discriminate new images of familiar and unfamiliar con-
specifics. The discrimination between familiar and unfa-
miliar conspecifics was achieved regardless of the category
of stimuli that was rewarded (images of familiar or of
unfamiliar conspecifics). Four heifers of the eight tested
completed the generalization step more rapidly (in fewer
sessions) than in the training step. In addition, all heifers
were successful in the first trial of the generalization step.
This likely suggests that these heifers were really able to
generalize, following Zayan and Vauclair (1998) who
assessed that only the first trial of transfer is relevant to
establish generalization abilities. All the heifers that
reached the criterion of success in the generalization step of
phase 1 also reached the criterion in reversal learning in
phase 2 with four of the eight heifers that were successful
in the first trial of the reversal learning. This result confirms
that the heifers have the ability to discriminate between
familiar and unfamiliar conspecifics. In combination with
our results of a within-category (‘‘familiar conspecifics’’
and ‘‘unfamiliar conspecifics’’) discrimination (Coulon
et al. 2009), this suggests a categorization process of
familiar/unfamiliar in cattle (review in Zayan and Vauclair
1998). This categorization seems to go beyond the physical
similarities of the stimuli of a category and be more related
to conceptual processing (Herrnstein 1990). The control
experiments confirm the results of discrimination between
images of familiar and unfamiliar conspecifics. The heifers
discriminated better between familiar conspecifics than
between unfamiliar individuals (phase 3 vs. phase 4). The
subjects most likely, during daily social interactions, learnt
some multimodal characteristics of their familiar conspe-
cifics and used this information to perform discriminations
in an experimental setting, which presented only cues from
only one sensory modality. Nevertheless, the heads of
conspecifics appear to be sufficient for social discrimina-
tion in cattle. Our results match similar findings in primates
and sheep (review in Tate et al. 2006). For example, sheep
discriminate between images of familiar and unfamiliar
Anim Cogn (2011) 14:279–290 287
123
Author's personal copy

conspecifics and this discrimination is strongly confirmed
by electrophysiological data, which showed that cells in the
temporal cortex responded preferentially to projected
images of faces of familiar conspecifics (Kendrick and
Baldwin 1987; Peirce and Kendrick 2000; Kendrick et al.
2001). The main function of these cells could be to identify
categories of individuals that have similar social signifi-
cance for the subject (review in Kendrick 1998).
In our study, the subjects made more errors when the
heads of the familiar and the unfamiliar individuals were
viewed from the same angle than when they were viewed
from a different angle. Heifers seemed to have more diffi-
culties in discriminating between similar images indepen-
dently of the identity of the familiar and unfamiliar
individuals they had to discriminate between. As the error
rate for a training pair of individuals was similar to the error
rate for another pair of individuals, subjects seemed to base
their discrimination on the familiarity versus unfamiliarity
and not on the identity of the individuals. In addition, the
heifers showed better performances in the discrimination
between familiar and unfamiliar conspecifics (phase 1) than
in the experiment of discrimination between two sets of three
familiar individuals (phase 3). These results suggest that
discrimination is favoured over recognition when comparing
familiar and unfamiliar conspecifics. Individual recognition
is a complex cognitive ability where one organism identifies
another one according to its unique individually distinctive
characteristics (Thom and Hurst 2004; Tibbetts and Dale
2007) and when this organism can form a mental represen-
tation of each property of conspecifics (Zayan 1994). In a
previous study, using the same experimental paradigm with
food reward, we demonstrated that heifers were able to
recognize individuals visually independently of their pre-
vious social experience (Coulon et al. 2009) and indepen-
dently of the angle of view. Heifers may therefore use the
easiest or fastest way to perform the requested task i.e.,
discriminating between familiarity and unfamiliarity com-
pared to recognizing all individuals used as stimuli.
The head view angle of a stimulus (e.g., facial, profile)
or the proportion of dark patches had no effect on the
success of the discrimination. This result contrasts with
those found for sheep where the discrimination between
familiar breeds was easier for images of frontal view of
faces than for profiles (Kendrick et al. 1995). Ferreira et al.
(2004) also showed that sheep can transfer between frontal
and profile views, though with a decrease in performance.
These additional results inform us that the face discrimi-
nation must be based on other features than the view angle
or the coat colour. A natural further step in the study of the
capacity of discrimination would be to find out the salient
features of cow faces that the subjects relied on to make
their discrimination. Several studies showed that the eye
regions play a key role in the face processing in non human
primates (Gothard et al. 2003,2009; Hirata et al. 2010;
Marsh and MacDonald 2008). The use of the morphing
technique or scramble faces (Freedman et al. 2001; Weiss
et al. 2001) could also indicate the cues that the cows are
using to discriminate familiar from unfamiliar conspecifics.
In our study, the performance of heifers cannot be inter-
preted as a discrimination between complex patterns without
social meaning. Heifers rewarded for images of unfamiliar
conspecifics pointed ears backward more frequently (indi-
cating confrontation with unfamiliar stimuli, Lensink et al.
2006) and they showed less forward pointing ears (indicat-
ing less positive expectation, Lensink et al. 2006) when they
directed their attention to unfamiliar images, compared to
heifers rewarded for familiar conspecifics. This indicates
that heifers have different emotional reactions when con-
fronted with these two types of stimuli. Our behavioural
observations are congruent with the existence of a real
ability to discriminate between familiar and unfamiliar
conspecifics, and with our interpretation that heifers asso-
ciate 2D images with real conspecifics.
In conclusion, our study demonstrates that cattle are able
to discriminate visually between familiar and unfamiliar
conspecifics and that images of conspecifics seem to be
treated as representations of real individuals. These results
confirmed and extended our previous results obtained with
a similar procedure (Coulon et al. 2007,2009). The use of
visual cues from the head is sufficient to allow clear dis-
crimination between familiar and unfamiliar conspecifics,
independently of other visual cues. Under natural condi-
tions, cattle use more complex visual information, includ-
ing body shape and size, body postures and movements,
and in particular head position and movements. They also
use acoustic and chemical signals. Our results show that the
discrimination between familiar and unfamiliar conspecif-
ics could occur from a distance and that visual cues from
head are sufficient for discrimination.
Acknowledgments We thank the staff of the experimental farm
UCEA-Bressonvilliers for special care to the animals, especially
Laurent Delatouche and Christophe Richard (INRA UE 298 Unite
´
commune d’expe
´rimentation animale, F-91630 Leudeville). Thanks
to Isabella Bollini for her help in the training of animals. We thank
Prof. Herve
´Abdi, Prof. Patrizia D’Ettorre and Prof. Charles T.
Snowdon for their careful editing on the manuscript. This work was
supported by a grant from the French Ministe
`re de l’Enseignement et
de la Recherche to M. Coulon.
Conflict of interest The authors declare that they have no conflict
of interest.
References
Bouissou MF, Andrieu S (1978) Etablissement des relations pre
´fe
´r-
entielles chez les bovins domestiques. Behaviour 64:148–157
288 Anim Cogn (2011) 14:279–290
123
Author's personal copy

Bouissou MF, Porter RH, Boyle L, Ferreira G (1996) Influence of a
conspecific image of own vs. different breed on fear reactions of
ewes. Behav Process 38:37–44
Bovet D, Vauclair J (2000) Picture recognition in animals and
humans. Behav Brain Res 109:143–165
Charrier I, Mathevon N, Jouventin P (2001) Mother’s voice recog-
nition by seal pups. Nature 412:87
Coulon M, Deputte BL, Heyman Y, Richard C, Delatouche L,
Baudoin C (2007) Visual discrimination by heifers (Bos taurus)
of their own species. J Comp Psychol 121:198–204
Coulon M, Deputte BL, Heyman Y, Baudoin C (2009) Individual
recognition in domestic cattle (Bos taurus): evidence from 2D-
images of heads from different breeds. PLoS ONE 4(2):e4441
Coulon M, Baudoin C, Abdi H, Heyman Y, Deputte BL (2010) Social
behavior and kin discrimination in a mixed group of cloned and
non cloned heifers (Bos taurus). Theriogenology 74:1596–1603
Cummins KA (1991) Olfactory and visual cues. Individual recogni-
tion and social aggression in lactating cows. J Dairy Sci 74:301
Dasser V (1987) Slides of group members as representations of real
animals (Macaca fascicularis). Ethology 76:65–73
Dawkins MS (1996) Distance and social recognition in hens:
implication for the use of photographs as social stimuli.
Behaviour 133:663–680
Deputte BL (2000) Primate socialization revisited: theoretical and
practical issues in social ontogeny. In: Slater PJB, Rosenblatt JS,
Snowdon CT, Roper TJ (eds) Advances in the study of
behaviour. Academic Press, San Diego, pp 99–157
Deputte BL (2010) Communication, perception et expression du
chien. In: Bedossa T, Deputte BL (eds) Comportement et
e
´ducation du chien. Educagri Editions, Dijon, pp 355–424
Deputte BL, Pelletier S, Barbe S (2001) Visual categorization of
natural and abstract items in forest monkeys and humans. Behav
Process 55:51–64
Entsu S, Dohi H, Yamada A (1992) Visual acuity of cattle determined
by the method of discrimination learning. Appl Anim Behav Sci
34:1–10
Fagot J (2000) Picture perception in animals. Psychology Press, Hove
Fagot J, Thompson RKR, Parron C (2010) How to read a picture:
lessons from nonhuman primates. Proc Natl Acad Sci
107:519–520
Farah MJ, Wilson KD, Drain M, Tanaka JN (1998) What is ‘‘special’’
about face perception? Psychol Rev 105:482–498
Ferreira G, Keller M, Saint-Dizier H, Perrin G, Le
´vy F (2004)
Transfer between views of conspecific faces at different ages or
in different orientations by sheep. Behav Process 67:491–499
Freedman DJ, Riesenhuber M, Poggio T, Miller EK (2001) Categor-
ical representation of visual stimuli in the primate prefrontal
cortex. Science 291:312–316
Gheusi G, Bluthe
´RM, Goodall G, Dantzer R (1994) Social and
individual recognition in rodents: methodological aspects and
neurobiological bases. Behav Process 33:59–88
Gothard KM, Erickson CA, Amaral DG (2003) How do rhesus
monkeys (Macaca mulatta) scan faces in a visual paired
comparison task? Anim Cogn 7:25–36
Gothard KM, Brooks KN, Peterson MA (2009) Multiple perceptual
strategies used by macaque monkeys for face recognition. Anim
Cogn 12:155–167
Hagen K, Broom DM (2003) Cattle discriminate between individual
familiar herd members in a learning experiment. Appl Anim
Behav Sci 82:13–28
Herrnstein RJ (1990) Levels of stimulus control: a functional
approach. Cognition 37:133–166
Hirata S, Fuwa K, Sugama K, Kusunoki K, Fujita S (2010) Facial
perception of conspecifics: chimpanzees (Pan troglodytes)
preferentially attend to proper orientation and open eyes. Anim
Cogn 13:679–688
Houpt KA (2005) Domestic animal behavior for veterinarians and
animal scientists. Blackwell, Oxford
Isingrini M, Lenoir A, Jaisson P (1985) Preimaginal learning as a
basis of colony-brood recognition cues in the ant Cataglyphis
cursor. Proc Natl Acad Sci 82:8545–8547
Kendrick KM (1998) Intelligent perception. Appl Anim Behav Sci
57:213–231
Kendrick KM, Baldwin BA (1987) Cells in temporal cortex of
conscious sheep can respond preferentially to the sight of faces.
Science 236:448–450
Kendrick KM, Atkins K, Hinton MR, Broad KD, Fabre-Nys C,
Keverne B (1995) Facial and vocal discrimination in sheep.
Anim Behav 49:1665–1676
Kendrick KM, Atkins K, Hinton MR, Heavens P, Keverne B (1996)
Are faces special for sheep? Evidence from facial and object
discrimination learning tests showing effects of inversion and
social familiarity. Behav Process 38:19–35
Kendrick KM, Leigh AE, Peirce JW (2001) Behavioural and neural
correlates of mental imagery in sheep using face recognition
paradigms. Anim Welf 10:89–101
Lacreuse A, Martin-Malivel J, Lange HS, Herndon JG (2007) Effects
of the menstrual cycle on looking preferences for faces in female
rhesus monkeys. Anim Cogn 10:105–115
Lensink J, Leruste H, Tillie M (2006) L’observation du troupeau
bovin : voir, interpre
´ter, agir. France agricole, Paris
Lewkowicz DJ, Sowinski R, Place S (2008) The decline of cross-
species intersensory perception in human infants: underlying
mechanisms and its developmental persistence. Brain Res
1242:291–302
Marsh HL, MacDonald SE (2008) The use of perceptual features in
categorization by orangutans (Pongo abelli). Anim Cogn
11:569–585
Moe RO, Bakken M, Kittilsen S, Kingsley-Smith H, Spruijt BM
(2006) A note on reward-related behaviour and emotional
expressions in farmed silver foxes (Vulpes vulpes) _Basis for a
novel tool to study animal welfare. Appl Anim Behav Sci
101:362–368
Neiworth JJ, Hassett JM, Sylvester CJ (2007) Face processing in
humans and new world monkeys: the influence of experiential
and ecological factors. Anim Cogn 10:125–134
Parr LA, Heintz M (2008) Discrimination of faces and houses by
rhesus monkeys: the role of stimulus expertise and rotation
angle. Anim Cogn 11:467–474
Parr LA, Winslow JT, Hopkins WD (1999) Is the inversion effect in
rhesus monkeys face-specific? Anim Cogn 2:123–129
Parr LA, Winslow JT, Hopkins WD, de Waal FBM (2000)
Recognizing facial cues: individual discrimination by chimpan-
zees (Pan troglodytes) and rhesus monkeys (Macaca mulatta).
J Comp Psychol 114:47–60
Parron C, Call J, Fagot J (2008) Behavioural responses to photographs by
pictorially naı
¨ve baboons (Papio anubis), gorillas (Gorilla gorilla)
and chimpanzees (Pan troglodytes). Behav Process 78:351–357
Pascalis O, de Haan M, Nelson CA (2002) Is face processing species
specific during the first year of life? Science 296:1321–1323
Peirce JW, Kendrick KM (2000) Configurational coding, familiarity
and the right hemisphere advantage for face recognition in sheep.
Neuropsychologia 38:475–483
Phillips C (2002) Cattle behaviour. Farming Press, Ipswich
Pokorny JJ, de Waal FDM (2009) Monkeys recognize the faces of
group mates in photographs. Proc Natl Acad Sci 106:21539–
21543
Porter RH, Bouissou MF (1999) Discriminative responsiveness by
lambs to visual images of conspecifics. Behav Process 48:101–110
Racca A, Amadei E, Ligout S, Guo K, Meints K, Mills D (2010)
Discrimination of human and dog faces and inversion responses
in domestic dogs (Canis familiaris). Anim Cogn 13:525–533
Anim Cogn (2011) 14:279–290 289
123
Author's personal copy

Reefmann N, Wechsler B, Gygax L (2009) Behavioural and
physiological assessment of positive and negative emotion in
sheep. Anim Behav 78:651–659
Rehka
¨mper G, Go
¨rlach A (1997) Visual discrimination in adult dairy
bulls. J Dairy Sci 80:1613–1621
Rosenfeld SA, Van Hoesen GW (1979) Face recognition in rhesus
monkey. Neuropsychology 17:503–509
Sato S, Yoshikawa C (1996) Estimation of cognition pattern of cattle
by the visual attention to photograph and movie images. Animal
Sci Technol 67:36–42
Tate AJ, Fisher H, Leigh A, Kendrick KM (2006) Behavioural and
neurophysiological evidence for face identity and face emotion
processing in animals. Phys Trans R Soc B 361:2155–2172
Taylor AA, Davis H (1998) Individual humans as discriminative
stimuli for cattle (Bos taurus). Appl Anim Behav Sci 58:13–21
Thom MD, Hurst JL (2004) Individual recognition by scent. Ann Zool
Fenn 41:765–787
Thompson RKR (1995) Natural and relational concepts in animals.
In: Roitblat HL, Meyer JA (eds) Comparative approaches to
cognitive science. MIT Press, Cambridge, pp 175–224
Tibbetts EA (2002) Visual signals of individual identity in the wasp
Polistes fuscatus. Proc R Soc Lond B 269:1423–1428
Tibbetts EA, Dale J (2007) Individual recognition: it is good to be
different. Trends Ecol Evol 22:529–537
Van der Velden J, Zheng Y, Patullo BW, Macmillan DL (2008)
Crayfish recognize the faces of fight opponents. PLoS ONE
3:e1695
Vauclair J, Fagot J (1996) Categorization of alphanumeric characters
by Guinea baboons: within- and between-class stimulus dis-
crimination. Curr Psychol Cogn 15:449–462
Waring GH (2003) Horse behaviour. Noyes Publications, Norwich
Wasserman EA, Kiedinger RE, Bhatt RS (1988) Conceptual behavior
in pigeons: categories, subcategories, and pseudocategories.
J Exp Psychol Anim Behav Process 14:235–246
Weiss DJ, Kralik JD, Hauser MD (2001) Face processing in cotton-
top tamarins (Saguinus oedipus). Anim Cogn 4:191–205
Zangenehpour S, Ghazanfar AA, Lewkowicz DJ, Zatorre RJ (2009)
Heterochrony and cross-species intersensory matching by infant
vervet Monkeys. PLoS ONE 4(1):e4302
Zayan R (1994) Mental representations in the recognition of
conspecific individuals. Behav Process 33:233–246
Zayan R, Vauclair J (1998) Categories as paradigms for comparative
cognition. Behav Process 42:7–99
290 Anim Cogn (2011) 14:279–290
123
Author's personal copy