Ready, set, yellow! color preference of Indian free-ranging dogs

Abstract

Most of the research on color vision related behaviors in dogs has involved training the dogs to perform visual discrimination tasks. We investigated the importance of color to untrained Indian free-ranging dogs (FRDs). Using one-time multi-option choice tests for color preference in 134 adult dogs, we found the dogs to prefer yellow objects over blue or gray ones while there was no preference between blue and gray. We next pitted a yellow object against a gray object that had food. Here, the dogs ignored the food (biscuit or chicken) to approach the yellow object first indicating the color preference to be quite strong. Color preference has previously been investigated in many other animals and has implications for behaviors like mate choice and foraging. Our study provides a new perspective into the ecology of Indian FRDs and might have implications for companion dogs as well, if they too show this preference.

Full text

ORIGINAL PAPER
Animal Cognition (2025) 28:7
https://doi.org/10.1007/s10071-024-01928-9
vertebrates and arthropods (Baden and Osorio 2019; Kelber
et al. 2003; Yilmaz et al. 2022). The presence of color vision
in a species does not automatically suggest the presence of
color preference, but to have color preference, the faculty
of color vision is essential. There are multiple biological
underpinnings for color preference, which may vary from
species to species, while color vision essentially requires
just the presence of cone cells (Kelber et al. 2003). Color
preference has been found in many animals across myriad
taxa. Humphrey (1971 & 1972) investigated color prefer-
ence of Rhesus monkeys and found the monkeys to dem-
onstrate a preference for dierent hues and brightness of
colored elds just like they prefer the content of one picture
over another. Humphrey (1972) suggested interest (towards
the content of stimulus pictures) and pleasure (reaction to
hue and brightness) to be the two drivers of the monkeys’
visual preference. Preference towards colors on an animal’s
body is often linked with mate choice (reviewed by Higham
and Winters 2016). Burley (1982, 1986) found zebra nches
to prefer conspecic body colors (species condence)
in their leg bands, which aected the birds’ mate choice.
Introduction
Color vision and color preference have been subjects of
research for over a century (reviewed by Staples 1931).
Decades of research have conrmed color vision in most
Anindita Bhadra
abhadra@iiserkol.ac.in
1 Department of Biological Sciences, Indian Institute of
Science Education & Research (IISER) Kolkata, Mohanpur,
Nadia, Kolkata 741246, India
2 Indian Institute of Science Education and Research (IISER)
Pune, Dr. Homi Bhabha Road, Pune, Maharashtra
411008, India
3 School of Biological Sciences, National Institute of
Science Education and Research (NISER), Jatni, Khurda,
Bhubaneswar, Odisha 752050, India
4 National Institute of Biomedical Genomics (NIBMG),
Kalyani, West Bengal 741251, India
5 Department of Environmental Science, Calcutta University,
Kolkata, West Bengal 700019, India
Abstract
Most of the research on color vision related behaviors in dogs has involved training the dogs to perform visual discrimina-
tion tasks. We investigated the importance of color to untrained Indian free-ranging dogs (FRDs). Using one-time multi-
option choice tests for color preference in 134 adult dogs, we found the dogs to prefer yellow objects over blue or gray
ones while there was no preference between blue and gray. We next pitted a yellow object against a gray object that had
food. Here, the dogs ignored the food (biscuit or chicken) to approach the yellow object rst indicating the color prefer-
ence to be quite strong. Color preference has previously been investigated in many other animals and has implications
for behaviors like mate choice and foraging. Our study provides a new perspective into the ecology of Indian FRDs and
might have implications for companion dogs as well, if they too show this preference.
Highlights
●Indian free-ranging dogs (FRDs) show preference for the color yellow over blue and gray.
●Indian FRDs show no preference between blue and gray colors.
●Attraction towards a yellow object can be stronger than attraction towards food rewards for Indian FRDs.
Keywords Color preference · Cognition · Behavior · Dog · Free-ranging dogs · FRDs
Received: 1 October 2024 / Revised: 17 December 2024 / Accepted: 19 December 2024
© The Author(s) 2024, corrected publication 2025
Ready, set, yellow! color preference of Indian free-ranging dogs
AnamitraRoy1· AeshaLahiri1· SrijayaNandi1· AayushManchalwar1,2· S.Siddharth1,3· J. V. R.Abishek1·
IndiraBulhan1· ShouvanikSengupta1,4· SandeepKumar1· TushnimChakravarty1,5· AninditaBhadra1
1 3
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
However, Wang et al. (2018) noted that zebra nch color
preference has had poor replication across studies, and hith-
erto observed signicant results might not be generalizable
at a species level.
Ecological valence or the usual role of a particular color
in an animal’s life, has been suggested as another driver
for color preference (Palmer & Schloss, 2010). Several
studies have explored the psychological impact of colors
which are often linked with ecological valence (reviewed
by Elliot 2015). It is postulated that the redness of primate
faces being an indicator of health and mood might have
been a driver for primates developing red-green color vision
(Changizi et al. 2006). Preference of bowerbirds (Ambly-
ornis inornatus) towards colors of decorative objects (they
decorate their bowers with colorful objects - both natural
and human-made) seem to be culturally transmitted, which
means color preference varies across populations of the
same species located in dierent areas, whereas within pop-
ulation the patterns appear quite similar (Diamond 1986). In
summary, color preference is a widespread phenomenon. It
can be both innate or learned, and in some cases, might be
driven by what the color usually signies in an animal’s life.
In dogs (Canis lupus familiaris), color vision was estab-
lished by Neitz et al. (1989) who characterized their cone
cell activities to peak at 429 nm and 555 nm, indicating a
dichromatic color vision (blue-yellow spectrum) similar
to red-green colorblind humans. Further behavioral study
(Siniscalchi et al. 2017), electroretinogram photometry
(Jacobs et al. 1993), and topographical study of retinal cone
cells (Mowat et al. 2008) have solidied our understanding
of dog color vision and it is now possible to digitally simu-
late a dog’s vision (Pongrácz et al. 2017: h t t p s : / / d o g - v i s i o n
. a n d r a s p e t e r . c o m). However, all behavioral research related
to canine color cognition so far has been focused on color
discrimination tasks after a period of training: to mainly
understand the limits of canine color vision (Byosiere 2018,
Kasparson et al. 2013; Tanaka et al. 2000). It was hitherto
unknown if and how untrained dogs are aected by color
cues in their environment. The only available direct research
on this topic is a non-peer-reviewed article on canine color
preference regarding chew toys by Wong (2007) where the
dogs had prior experience with chew toys of unreported col-
ors. Cimarelli et al. (2023), as part of another experiment,
investigated preference of Moroccan FRDs for blue and yel-
low colored objects and found no preference. However, this
experiment was not specically designed to test color pref-
erence, and one of the used objects had white patterns on it.
Thus, a specic color preference experiment is still needed.
The question of usefulness of color cues is especially
important for FRDs, as unlike companion dogs, they are
not under direct human supervision (Serpell 1995), and they
need to make numerous important decisions daily in order
to survive. It is possible that each available cognitive abil-
ity, including color vision, is important to these dogs even
though their sense of smell is highly sensitive (Beaver 2009;
Hayes et al. 2018; Kokocińska-Kusiak et al. 2021). Notably,
olfactory cues are limited by wind directions and the slow
nature of diusion (Fick, 1885). Visual cues thus might be
much more reliable for a quick introductory assessment of
an object. Indian FRDs are primarily scavengers of human-
generated waste (Bhadra et al. 2015; Biswas et al. 2024)
from garbage dumps and bins that are often overwhelm-
ingly smelly (personal observation) and can demand utiliza-
tion of visual cues. Previous work with FRDs suggests that
FRDs use both visual and olfactory cues during scavenging,
though the decision to eat something is mostly guided by
a Rule of Thumb “If it smells like meat, eat it” (Bhadra et
al. 2015; Sarkar et al. 2019). There already is experimen-
tal evidence (Polgár et al., 2015) that companion dogs can
fail to use olfaction for approach-based decision-making if
their distance from the object is too great (3 m). It is not
impossible that reliance on olfaction breaks down in other
contexts as well.
Around the mid-2010s, India saw a craze of “blue bot-
tles” where people were keeping plastic bottles lled with
indigo-dyed water around their houses, in the hope of keep-
ing FRDs away (Jaipuriar 2016). Masih and Bhadra (2019,
unpublished) found that FRDs in and around Kolkata, West
Bengal, did not show any specic response towards the bot-
tles. However, people’s attempt at modifying dog behavior
by use of color acted as a seed towards our exploration of
color preference.
We carried out choice tests with FRDs in India to test if
they have any color preferences. In Experiment 1, we used
yellow, blue, and gray colors as they are the hues visible to
dogs. Under null hypothesis, each color should be chosen a
similar number of times. After nding the dogs’ preferred
color, in Experiment 2, we gave them a choice between their
non-preferred colors to nd out their secondary preference,
if any. We conducted Experiment 3 to ensure only the color,
not the smell of the painted options were dictating the dogs’
preference. Further, to ascertain whether this preference
aects the dogs in their usual lives at all, in Experiment 4
(4a and 4b) we tested the strength of their color preference
by pitting their preferred color against a food reward. If
dogs ignore food and approach their preferred color instead,
they risk losing a vital resource. We hypothesize that the
dogs would ignore their preferred color and approach the
food rst if the amount and type of food is attractive enough
to them.
1 3
7 Page 2 of 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
Materials and methods
Subjects, locations, and time
The experiments were conducted in urban, semi-urban, and
rural locations in Nadia and North 24 Parganas districts, and
the city of Kolkata in West Bengal, India (details available
in Online Resource 1). A total of 458 visibly t adult FRDs
successfully participated in the experiments. The sexes of
the dogs were noted visually based on to their genitals. To
avoid learning bias due to repetitions, each dog was tested
only once, and we did not test dogs that had observed oth-
ers participate in the experiment. We chose new experiment
locations each day to avoid repeating dogs by mistake. All
trials were done in daylight and all the options (bowls)
in one trial were placed equally under direct sunlight, or
equally in shade to ensure they were under the same light-
ing conditions.
General experimental protocol
The general protocol followed the one-time multi-option
choice test (OTMCT) module described by Bhadra and
Bhadra (2014). We used dierent colored terracotta bowls
as the options. Two experimenters conducted each experi-
ment. On dierent days, dierent people acted as experi-
menter 1 and 2. While traveling on the streets of the
aforementioned locations, when an FRD was located,
experimenter 1 approached the dog with the experimen-
tal setup (see below). At about 1.5–2 m from the dog, the
experimenter vocalized positively towards the dog (Bhat-
tacharjee et al. 2017) and placed the setup on the ground in a
way that would keep the options approximately equidistant
from the dog. Then, experimenter 1 slowly backed away a
few steps and assumed a neutral pose. The experimenter
avoided making eye contact with the dog. Experimenter 2
video-recorded the whole experiment for later analysis. A
trial was deemed unsuccessful if the dog did not initiate an
approach within 1 min, or if the dog showed clear disinterest
(e.g., going away from the area, closing its eyes and lazing).
In case multiple dogs were present nearby, a trial was con-
sidered successful if no other dog came near the focal dog
or the setup before the choice was made.
In each trial, the order of the bowls (which colors are
at left, right, or center) was randomized, and all trials hap-
pened with fresh bowls, i.e., the bowls were cleaned with
water after each trial to remove any scent from previous
dogs before being used again in a trial. Whichever bowl a
dog investigated (nose within about 5 cm of a bowl) rst,
was taken as that dog’s choice. AR coded all the videos and
rechecked the data twice more for possible errors.
General experimental setup
The terracotta bowls used in the experiments were all of a
similar make with a diameter of 12.58 ± 0.46 cm, and height
of 4.70 ± 0.23 cm. A total of 42 bowls were used. We col-
ored these bowls using Fevicryl® Acrylic paint. Three sets
of bowls were created with the following paints: golden yel-
low, Prussian blue, and gray (manually created by mixing
black and white). A cardboard platform (70 cm x 17 cm),
made waterproof with brown packing tape, was used as a
base to allow one person to present three equidistant bowls
to a dog. The bowls (yellow, blue, and gray) were placed
approximately 15–16 cm apart on this platform for the
three-choice experiment (Fig. 1) which has been further
detailed about in Experiment 1. We performed three other
two-choice experiments which are discussed in Experi-
ments 2, 3, and 4. In these three experiments, the bowls
were presented approximately 50 cm apart. The cardboard
platform was not used in Experiments 2 and 4 since the two
bowls could be easily presented by hand.
Statistical analysis
All statistical analyses were performed with R (R core team
2022) in RStudio (RStudio 2019). The following packages
were used for analysis and presentation: ggplot2 (Wick-
ham 2016), data.table (Barrett et al. 2023), and DescTools
(Signorell 2023). We performed contingency χ2 tests for
comparing frequency of rst choices between conditions
and goodness of t χ2 tests for comparing observation of one
condition against a null (chance) expectation. Alpha for all
tests was set at 0.05 and it was adjusted according to Bon-
ferroni correction for multiple comparisons. In the graphs,
the whiskers on the bar plots show 95% Wilson’s condence
interval (Wilson 1927) for estimating the population value
of variables from our sample values by assuming a binomial
distribution around our observations. We report Cramér’s V
(Cramér 1946) as the eect sizes of the χ2 tests.
Fig. 1 An FRD approaching the three-choice setup in Experiment 1
that has blue, yellow, and gray (left to right) bowls
1 3
Page 3 of 9 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
Experiment 3: Control for the smell of paint with yellow and
blue bowls, both covered with sieves
Since dierent paints (although the same brand and thus
hopefully the same volatiles) were used to paint the bowls,
it was important to make sure that the dogs were making
their choice based on the visual cue of the paint and not
the olfactory cue. Therefore, we painted two sets (blue and
yellow) of terracotta bowls only on the outside and covered
them with magenta colored plastic sieves (Fig. 3). This way
we made sure all visual cues of the paint were blocked, but
the smell could still permeate through the porous sieve. The
covered bowls were placed on the same cardboard platform
with about 50 cm separation. See Online Resource 6 for
video of the experiment.
We successfully tested 54 dogs (29 females, 25 males).
27 dogs did not make a choice when presented with the setup
Experiment 1: three-choice experiment with yellow, blue,
and gray bowls
For our initial test of color preference, we used yellow,
blue, and gray since these are the three hues visible to
dogs (Fig. 1). This experiment had two conditions. In the
“with_food” condition, each bowl had an equal amount of
food- a 1/4th piece of a Brittania® Marie Gold® biscuit. In
the “no_food” condition, no bowl had any food. See Online
Resource 2 for video of the experiment.
We successfully tested 76 dogs (50 males, 26 females)
for the “with_food” and 58 dogs (27 males and 31 females)
for the “no_food” condition. 45 and 52 dogs respectively
were presented with the setup but they made no choice. A
total of 34 “with_food” and 13 “no_food” trials had to be
discarded because they did not meet all experimental crite-
ria (all bowls not being equidistant to the dogs, interference
from other dogs or humans, experimenter failing to maintain
a neutral posture, experimenter placing the setup too close
to the dog). No dierence was observed for the rst choice
between the two conditions (Data: Online Resource 3, con-
tingency χ2 = 0.219, N = 134, df = 2, p = 0.896), and thus,
the data were pooled for further analysis. Overall, yellow
emerged as the most preferred color (Fig. 2, goodness of t
χ2 = 25.134, N = 134, df = 2, p < 0.001, Cramér’s V = 0.433).
The preferred color was not dierent for male and female
dogs (Data: Online Resource 4, contingency χ2 = 1.938,
df = 2, p = 0.379). The relative positions of the chosen bowls
(left, right, or center) had no signicant eect on the choice
of color (Data: Online Resource 4, contingency χ2 = 3.884,
df = 4, p = 0.422).
Experiment 2: two-choice experiment with blue and gray
bowls
To identify any second preference, a two-choice test was
done between blue and gray. The same blue and gray bowls
as described earlier were used. The bowls were placed about
50 cm away from each other. Since presence of food did
not aect the choice in the previous experiment, all trials in
this experiment were the “with_food” condition only. See
Online Resource 5 for video of the experiment. We success-
fully tested 102 dogs (52 females, 47 males, and 3 dogs of
unknown sex) where blue was chosen 44 times and gray
was chosen 58 times. There was no signicant preference
for any of these colors (goodness of t χ2 = 1.921, df = 1,
p = 0.165). A total of 21 dogs did not make a choice when
presented with the setup, and 19 trials were discarded for
not meeting all experimental criteria.
Fig. 3 The setup for Experiment 3: control for the smell of paint. Blue
and yellow colored bowls are present, but not visible, under the pink
and white sieves. The porous portion (white) of the sieves allow smell
of the paint to pass through
Fig. 2 The bars denote observed frequency of choice in Experiment 1,
converted to percentage. The whiskers show Wilson’s 95% CI around
the observations. Dierent letters atop bars denote signicant dier-
ence at α = 0.05
1 3
7 Page 4 of 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
chosen 11 times while the empty yellow bowl was chosen 41
times (goodness of t χ2 = 17.308, df = 1, p < 0.001, Cramér ’s
V = 0.576). In this condition, 8 dogs made no choice when
presented with the setup and 6 trials were discarded for not
meeting all experimental criteria. When the food was ~ 15 g
of chicken, the empty yellow bowl was chosen 47 times
out of 61 successful trials (28 males, 33 females, dierent
dogs than the biscuit condition), while the gray bowl with
chicken was chosen 14 times (goodness of t χ2 = 17.852,
df = 1, p < 0.001, Cramér’s V = 0.540). Here, 7 dogs made no
choice upon being presented with the setup and 10 trials
were discarded for not meeting all experimental criteria.
There was no signicant dierence between the biscuit
group and the chicken group (contingency χ2 = 4.05*10^-6,
df = 1, p = 0.99), and each of these groups was signicantly
dierent than the control group (contingency X2 tests, con-
trol and biscuit: χ2 = 18.819, df = 1, p < 0.001, Cramér ’s
V = 0.421; control and chicken: χ2 = 18.852, df = 1, p < 0.001,
Cramér’s V = 0.404). With Bonferroni correction, the alpha
for these three tests was at 0.017.
Discussion
Experiment 1 with the three choices revealed a preference
for yellow over blue or gray in the FRDs, rejecting the null
hypothesis of no color preference. We failed to reject the null
hypothesis of there being no preference between blue and
gray in Experiment 2. We had hypothesized the dogs would
ignore their preferred color if there was attractive enough
food associated with a non-preferred color. Experiment 4b
shows that even ~ 15 g pieces of chicken are not enough for
FRDs to ignore nearby yellow objects and thus we fail to
reject this null hypothesis. Further experiments with greater
amount of chicken are needed to test this hypothesis but that
is beyond our current capacity. However, in earlier experi-
ments, FRDs have been found to prefer even trace amounts
of chicken over carbohydrates in choice tests (Bhadra and
Bhadra 2014; Bhadra et al. 2015; Sarkar et al. 2019).
The dogs often explored all the bowls presented in
Experiment 1 (three choice), and in the case of the two-
choice set-up of blue versus gray (Experiment 2), they still
approached the setup. In the strength of preference experi-
ment (Experiments 4a and 4b), they frequently bee-lined to
yellow even when the gray bowl had food in it and the yel-
low one did not. Together, these results lead us to the con-
clusion that the observed preference for yellow is a result of
attraction towards yellow, and not due to repulsion towards
other colors.
In the three-choice experiment (Experiment 1), the pres-
ence of food in the bowls did not aect the preferences while
the dogs were more inclined (as per absolute numbers, the
and 10 trials were discarded for not meeting all experimen-
tal criteria. Blue was chosen 31 times and yellow was cho-
sen 23 times. The observations did not signicantly dier
from chance (goodness of t χ2 = 1.185, df = 1, p = 0.2763).
Experiment 4a: strength of preference control: with two
yellow bowls, one of which has food
The strength of preference experiment was devised to
compare the dogs’ attraction towards yellow, against their
attraction towards food which was associated with gray, a
non-preferred color. The same bowls as before were used
for this experiment, but to ensure the highest visibility of
food, the bowls were upturned, and the food was placed on
top of the upturned bowl (see Online Resources 7 & 8 for
videos of the experiment). While presenting, the distance
between the bowls was about 50 cm.
The piece of biscuit being used in our experiments was
quite small (a quarter of a circle with approximately 3 cm
radius) and while FRDs do eat carbohydrates, they prefer
meat (Bhadra and Bhadra 2014; Bhadra et al. 2015; Sarkar
et al. 2019). Thus, for the experiment to work as intended,
it was important to ensure that the dogs indeed prefer to
choose the biscuit over an empty bowl when the color of the
two bowls are the same. We chose both bowls to be yellow
because the brownish biscuit is most camouaged against
the yellow bowl and so if the dogs can identify and choose
it, they should have no problem choosing it when it is atop
a more contrasting gray bowl. Thus, the two options for the
control trials were both yellow bowls, only one of which
had a piece of biscuit on it. We successfully tested 54 dogs
(20 females, 33 males, and 1 dog of unknown sex). When
presented with the setup, 3 dogs did not make a choice and
6 trials were discarded for not meeting all experimental cri-
teria. The bowls with food (biscuits) were chosen 35 times
while bowls with no food were chosen 19 times (goodness
of t χ2 = 4.740, df = 1, p = 0.029, Cramér’s V = 0.296).
Experiment 4b: strength of preference: “no_food” yellow
bowl vs. “with_food” gray bowl
Here, we kept a piece of food on the gray bowl in a simi-
lar fashion to the control, while the yellow bowl was kept
empty, thereby oering the dogs a choice between food and
their preferred color. There were two conditions: (a) where
the food was a piece of biscuit as mentioned before, and
(b) where the food was a ~ 15 g piece of raw chicken. See
Online Resources 7 and 8 for videos of the experiment. Dif-
ferent sets of dogs participated in Experiment 4a and 4b to
avoid any bias due to learning.
We successfully tested 52 dogs (20 males, 32 females)
with biscuits as food (Fig. 4). The gray bowl with food was
1 3
Page 5 of 9 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
sensing the presence of the biscuit and then ignoring the
amount. Hence, we performed the strength of preference:
control experiment, to make sure the dogs would select a
small piece of biscuit, while the other option had no evident
food. Since the FRDs were discriminating between dierent
amounts of chicken (Banerjee and Bhadra 2019), such an
experiment was not required for chicken.
In the strength of preference experiment with biscuits, we
saw that the dogs overwhelmingly chose the empty yellow
bowl over the gray bowl that had a biscuit. Next, to cre-
ate a higher skew in the reward, we used chicken, a more
attractive food (Bhadra et al. 2015) with the gray bowl.
The ~ 15 g pieces of chicken were the largest that the bowls
could comfortably hold without spillage. We tried to maxi-
mize the amount of chicken given our current setup, hoping
to nd a preference shift towards food instead of color, from
which we may lower the food amount to nd an equilibrium
between color and food. Surprisingly, even with the appar-
ent substantial amount of chicken, the preference for yel-
low dominated. In one trial (Online Resource 8), the focal
dog pawed at and attempted to chew the yellow bowl for
some time, and then proceeded to solicit from experimenter
1, only returning to obtain the chicken from the gray bowl
dierence is not statistically signicant) to participate when
food was involved with the setup. Thus, in Experiments 2
and 3 only “with_food” trials were done.
In Experiment 3, control for paint smell, the amount of
paint was half of what was used in other experiments since
the top of the bowls were not painted to ensure zero vis-
ibility. However, since the participating dogs were not pre-
viously exposed to any of the setups used in any of these
experiments, the task was a simple test of smell preference
and should be unaected by the amount of the paint and its
odor.
Banerjee and Bhadra (2019) previously found Indian
FRDs to prefer higher quantities of chicken based on olfac-
tory cues. The dogs were not fooled when deceptive visual
cues (spreading the pieces of chicken salami apart to show
more counts of pieces, while the other option had the same
amount, but stacked together as one piece) were used to
inate the amount of chicken present. The same prefer-
ence however was absent when the food was biscuits. The
dogs chose randomly between options of more amount of
biscuits and less amount of biscuits. This indicates a dier-
ence in behavior based on preference of food (Bhadra et al.
2015). We needed to know for sure whether the dogs were
Fig. 4 Results of the strength of preference experiments (3.4). The col-
umns denote observed frequency of choice, converted to percentage.
The whiskers show Wilson’s 95% CI around the observations. Dif-
ferent letters atop bars (a or b) denote signicant dierence between
choices under each experiment condition (goodness of t χ2) and the
dierent numbers atop the graphs (1 or 2) denote signicant dierence
between experiment conditions (contingency χ2)
1 3
7 Page 6 of 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
Resource 9 for details) overlapped with respective dog cone
cell activity peaks, and thus the hues were of high contrast
to the dogs. The brightness of the objects could not be con-
trolled due to our dependence on company-manufactured
paints. There are very few and conicting reports regarding
the value of brightness discrimination in dogs (Byosiere et
al. 2018). So, we refrain from commenting on how the dogs
will perceive the relative brightnesses of our experimental
objects. Kasparson et al. (2013) found it easier for compan-
ion dogs to discriminate between objects of dierent hues
over objects of the same hue but dierent brightness. We
conjecture brightness to be more important while looking
for preference among colors that vary more subtly in hue
than in our experiment of high hue contrast. Our objects
being of dierent brightness should not aect the color pref-
erence that we have observed.
Conclusion
Our experiments demonstrate a clear preference for the
color yellow over blue and gray in FRDs of India, at least in
the context of foraging. This preference is so strong that it
supersedes their attraction towards food, whether biscuit or
chicken. This is the rst time that we have observed FRDs
ignore a clear food reward in a choice test. Further experi-
ments can help us understand the ecological advantages, if
any, of this preference and the reasons behind it. Moreover,
comparative studies with companion dogs and wolves can
help to understand the evolutionary trajectory of this prefer-
ence for yellow. The impact of color cues and color prefer-
ence on training can be explored in the future.
Supplementary Information The online version contains
supplementary material available at h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s 1 0 0 7 1 - 0
2 4 - 0 1 9 2 8 - 9.
Acknowledgements We thank Arpan Bhattacharya and Tuhin Shubhra
Pal for their participation in the data collection. AR would like to thank
Rohan Sarkar and Dr. Udipta Chakraborti for their insights regarding
statistical analyses. We would also like to acknowledge insightful and
detailed comments by two anonymous reviewers and the editor, that
helped to substantially improve this manuscript.
Author contributions AR conceived the idea and planned the experi-
ments with inputs from AB. All authors but AB participated in the eld
work to perform the experiments. AR decoded all videos and analyzed
the data. AR wrote the rst draft of the manuscript. AB supervised the
work, acquired funding, edited the manuscript. All authors gave their
consent to the manuscript.
Funding We thank the Department of Biotechnology, India (Project:
BT/HRD-NWB/39/2020-21), and IISER Kolkata, for funding sup-
port. AR and SN were supported by UGC-JRF and DST-INSPIRE JRF
fellowships respectively. SS was supported by IAS summer research
program.
after a few moments of futile solicitation. In the future, to
nd the aforementioned equilibrium, the amount of chicken
may be increased.
We do not yet know what exactly is causing this strong
preference for yellow. The species-condence hypothesis
(Burley 1986) might be a possible reason, as many FRDs
are shades of orange and brown that will appear yellowish
in dog vision (colors simulated by authors using methods
from Pongrácz et al. 2017). But while some dogs might
appear gray, surely no dog is blue. Since we found no pref-
erence between blue and gray, species condence does not
explain the color preference of FRDs. Moreover, species
condence is more applicable in the context of mate choice,
rather than in the context of foraging, and might not be a
suitable hypothesis to explore by object choice tests. The
ecological valence theory (Palmer & Schloss, 2010) links
color preference with a color’s usual ecological role or
value. Research on innate preference of honeybees (Giurfa
et al. 1995) and bumblebees (Raine and Chittka 2007) had
earlier found color preference to be highly correlated with
ower colors and associated nectar rewards. For Indian
FRDs, most of the food available to them is of human ori-
gin (Vanak & Gompper, 2009). Indians often use turmeric
(yellow) and dried chili (red) in their food (personal obser-
vations), both of which will appear yellow to dogs and
even raw meat (pink) and blood (red) will appear yellowish
(Pongrácz et al. 2017). It is important, however, to remem-
ber that while scavenging, a dog that is looking for yellow
food will come across many false positives: yellow, red,
and green human-generated inedible trash. To be accepted,
the ecological valence theory must explain how these false
positive encounters do not end up diminishing any color
preference. It is worth noting that even green grass and trees
will appear yellowish to dogs. Since color preference can be
both learned or innate, a test for innateness of canine color
preference appears to be a prime topic for future research.
Cimarelli et al. (2023), working with FRDs in Morocco, did
not nd any color preference between yellow and blue. This
might indicate our ndings are rather localized and there
is no single underlying mechanism of color preference for
dogs in general. However, it should also be noted that this
particular study (Cimarelli et al. 2023) was not designed
specically to test for color preference.
It could have been that the yellow bowls have a “pop-out
eect” due to high contrast with the other bowls and the sur-
rounding. Experiments presented here cannot rule out this
possibility. Early data from an ongoing experiment suggests
against visual pop-out aecting the dogs’ choice, but it can-
not yet be conrmed and demands detailed investigation.
Our experiments were limited by the availability of vari-
ations in paint color. We nally chose our particular yellow
and blue paints such that their spectral peaks (see Online
1 3
Page 7 of 9 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
Changizi MA, Zhang Q, Shimojo S (2006) Bare skin, blood and the
evolution of primate colour vision. Biol Lett 2(2):217–221. h t t p s
: / / d o i . o r g / 1 0 . 1 0 9 8 / r s b l . 2 0 0 6 . 0 4 4 0
Cimarelli G, Juskaite M, Range F (2023) Free-ranging dogs match a
human’s preference in a foraging task. Curr Zool. h t t p s : / / d o i . o r g
/ 1 0 . 1 0 9 3 / c z / z o a d 0 4 6
R Core Team (2022) R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Aus-
tria. https://www.R-project.org/
Cramér H (1946) Mathematical Methods of Statistics. Princeton:
Princeton University Press, page 282 (Chap. 21. The two-dimen-
sional case) ISBN 0-691-08004-6
Diamond J (1986) Animal art: variation in bower decorating style
among male bowerbirds Amblyornis inornatus. Proc Natl Acad
Sci 83(9):3042–3046. h t t p s : / / d o i . o r g / 1 0 . 1 0 7 3 / p n a s . 8 3 . 9 . 3 0 4 2
Elliot AJ (2015) Color and psychological functioning: a review of
theoretical and empirical work. Front Psychol 6:127893. h t t p s : /
/ d o i . o r g / 1 0 . 3 3 8 9 / f p s y g . 2 0 1 5 . 0 0 3 6 8
Fick A (1855) On liquid diusion. Lond Edinb Dublin Philosophical
Magazine J Sci 10(63):30–39. h t t p s : / / d o i . o r g / 1 0 . 1 0 8 0 / 1 4 7 8 6 4 4 5
5 0 8 6 4 1 9 2 5
Giurfa M, Núñez J, Chittka L, Menzel R (1995) Colour preferences of
ower-naive honeybees. J Comp Physiol A 177:247–259. h t t p s : / /
d o i . o r g / 1 0 . 1 0 0 7 / B F 0 0 1 9 2 4 1 5
Hayes JE, McGreevy PD, Forbes SL, Laing G, Stuetz RM (2018)
Critical review of dog detection and the inuences of physiology,
training, and analytical methodologies. Talanta 185:499–512. h t t
p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j . t a l a n t a . 2 0 1 8 . 0 4 . 0 1 0
Higham JP, Winters S (2016) Color and mate choice in non-human
animals. In A. J. Elliot, & M. D. Fairchild (Eds.), Handbook of
Color Psychology (pp. 502–530). Cambridge University Press. h t
t p s : / / d o i . o r g / 1 0 . 1 0 1 7 / c b o 9 7 8 1 1 0 7 3 3 7 9 3 0 . 0 2 5
Humphrey N (1971) Colour and Brightness preferences in monkeys.
Nature. 229:615–617. h t t p s : / / d o i . o r g / 1 0 . 1 0 3 8 / 2 2 9 6 1 5 a 0
Humphrey NK (1972) Interest and pleasure: two determinants of a
monkey’s visual preferences. Perception 1(4):395–416. h t t p s : / / d o
i . o r g / 1 0 . 1 0 6 8 / p 0 1 0 3 9 5
Jacobs GH, Deegan JF 2nd, Crognale MA, Fenwick JA (1993) Phot-
opigments of dogs and foxes and their implications for canid
vision. Vis Neurosci 10(1):173–180. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 7 / s 0 9
5 2 5 2 3 8 0 0 0 0 3 2 9 1
Jaipuriar V (2016), August 29 Blue ’magic’ danglers to scare dogs. The
Telegraph online. w w w . t e l e g r a p h i n d i a . c o m / j h a r k h a n d / b l u e - m a g i
c - d a n g l e r s - t o - s c a r e - d o g s / c i d / 1 3 3 2 2 9 6 Accessed18 Jan 2024
Kasparson AA, Badridze J, Maximov VV (2013) Colour cues proved
to be more informative for dogs than brightness. Proc Biol Sci
280(1766):20131356. h t t p s : / / d o i . o r g / 1 0 . 1 0 9 8 / r s p b . 2 0 1 3 . 1 3 5 6
Kelber A, Vorobyev M, Osorio D (2003) Animal colour vision–behav-
ioural tests and physiological concepts. Biol Rev Camb Philos
Soc 78(1):81–118. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 7 / s 1 4 6 4 7 9 3 1 0 2 0 0 5 9 8 5
Kokocińska-Kusiak A, Woszczyło M, Zybala M, Maciocha J,
Barłowska K, Dzięcioł M (2021) Canine olfaction: physiology,
Behavior, and possibilities for practical applications. Animals:
Open Access J MDPI 11(8):2463. h t t p s : / / d o i . o r g / 1 0 . 3 3 9 0 / a n i 1 1
0 8 2 4 6 3
Mowat FM, Petersen-Jones SM, Williamson H, Williams DL, Luthert
PJ, Ali RR, Bainbridge JW (2008) Topographical characterization
of cone photoreceptors and the area centralis of the canine retina.
Mol Vis 14:2518–2527. h t t p s : / / w w w . n c b i . n l m . n i h . g o v / p m c / a r t i c l
e s / P M C 2 6 1 0 2 8 8 /
Neitz J, Geist T, Jacobs GH (1989) Color vision in the dog. Vis Neu-
rosci 3(2):119–125. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 7 / s 0 9 5 2 5 2 3 8 0 0 0 0 4 4 3 0
Palmer SE, Schloss KB (2010) An ecological valence theory of human
color preference. PNAS Proceedings of the National Academy of
Sciences of the United States of America, 107(19), 8877–8882. h t
t p s : / / d o i . o r g / 1 0 . 1 0 7 3 / p n a s . 0 9 0 6 1 7 2 1 0 7
Data availability The datasheets and RScripts can be found at: h t t p s : / /
d o i . o r g / 1 0 . 1 7 6 0 5 / O S F . I O / J Z Y 8 W.
Declarations
Competing interests The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License,
which permits any non-commercial use, sharing, distribution and
reproduction in any medium or format, as long as you give appropri-
ate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if you modied the licensed
material. You do not have permission under this licence to share
adapted material derived from this article or parts of it. The images or
other third party material in this article are included in the article’s Cre-
ative Commons licence, unless indicated otherwise in a credit line to
the material. If material is not included in the article’s Creative Com-
mons licence and your intended use is not permitted by statutory regu-
lation or exceeds the permitted use, you will need to obtain permission
directly from the copyright holder. To view a copy of this licence, visit
h t t p : / / c r e a t i v e c o m m o n s . o r g / l i c e n s e s / b y - n c - n d / 4 . 0 /.
References
Baden T, Osorio D (2019) The retinal basis of Vertebrate Color Vision.
Annual Rev Vis Sci 5:177–200. h t t p s : / / d o i . o r g / 1 0 . 1 1 4 6 / a n n u r e v
- v i s i o n - 0 9 1 7 1 8 - 0 1 4 9 2 6
Banerjee A, Bhadra A (2019) The more the Merrier: Dogs can assess
quantities in Food-Choice tasks. Curr Sci. h t t p s : / / d o i . o r g / 1 0 . 1 8 5 2
0 / c s / v 1 1 7 / i 6 / 1 0 9 5 - 1 1 0 0
Barrett T, Dowle M, Srinivasan A (2023) data.table: Extension of
‘data.frame’. R package version 1.14.10. h t t p s : / / C R A N . R - p r o j e
c t . o r g / p a c k a g e = d a t a . t a b l e
Beaver B (2009) Canine behavior: insights and answers (2nd Edition),
Chap. 2, Page 54–57 ISBN 978-1-4160-5419-1
Bhadra A, Bhadra A (2014) Preference for meat is not innate in dogs. J
Ethol 32:15–22. h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s 1 0 1 6 4 - 0 1 3 - 0 3 8 8 - 7
Bhadra A, Bhattacharjee D, Paul M, Singh A, Gade PR, Shrestha P,
Bhadra A (2015) The meat of the matter: a rule of thumb for scav-
enging dogs? Ethol Ecol Evol 28(4):427–440. h t t p s : / / d o i . o r g / 1 0 .
1 0 8 0 / 0 3 9 4 9 3 7 0 . 2 0 1 5 . 1 0 7 6 5 2 6
Bhattacharjee D, Dev Nandi N, Gupta S, Sau S, Sarkar R, Biswas
A, Banerjee A, Babu D, Mehta D, Bhadra A (2017) Free-rang-
ing dogs show age related plasticity in their ability to follow
human pointing. PLoS One. 2017; 12(7):e0180643. h t t p s : / / d o i .
o r g / 1 0 . 1 3 7 1 / j o u r n a l . p o n e . 0 1 8 0 6 4 3. PMID: 28715475; PMCID:
PMC5513426
Biswas S, Bhowmik T, Ghosh K, Roy A, Lahiri A, Sarkar S, and
Bhadra A (2024) Scavengers in the human-dominated landscape:
an experimental study. Phil Trans R Soc B(379):20230179. h t t p s :
/ / d o i . o r g / 1 0 . 1 0 9 8 / r s t b . 2 0 2 3 . 0 1 7 9
Burley N (1986) Comparison of the band-colour preferences of two
species of estrildid nches. Anim Behav 34(6):1732–1741. h t t p s :
/ / d o i . o r g / 1 0 . 1 0 1 6 / S 0 0 0 3 - 3 4 7 2 ( 8 6 ) 8 0 2 6 0 - 3
Burley N, Krantzberg G, Radman P (1982) Inuence of colour-band-
ing on the conspecic preferences of zebra nches. Anim Behav
30(2):444–455. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / S 0 0 0 3 - 3 4 7 2 ( 8 2 ) 8 0 0 5 5 - 9
Byosiere SE, Chouinard PA, Howell TJ, Bennett PC (2018) What do
dogs (Canis familiaris) see? A review of vision in dogs and impli-
cations for cognition research. Psychon Bull Rev 25(5):1798–
1813. h t t p s : / / d o i . o r g / 1 0 . 3 7 5 8 / s 1 3 4 2 3 - 0 1 7 - 1 4 0 4 - 7
1 3
7 Page 8 of 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Animal Cognition (2025) 28:7
Tanaka T, Watanabe T, Eguchi Y, Yoshimoto T (2000) Color discrimi-
nation in dogs. Nihon Chikusan Gakkaiho 71(3):300–304. h t t p s : /
/ d o i . o r g / 1 0 . 2 5 0 8 / c h i k u s a n . 7 1 . 3 0 0
Vanak AT, Gompper M (2009) Dietary Niche Separation between
Sympatric Free-Ranging Domestic Dogs and Indian Foxes in
Central India. J Mammal 90(5);1058–1065 h t t p s : / / d o i . o r g / 1 0 . 1 6
4 4 / 0 9 - M A M M - A - 1 0 7 . 1
Wang D, Forstmeier W, Ihle M, Khadraoui M, Jerónimo S, Martin K,
Kempenaers B (2018) Irreproducible text-book knowledge: the
eects of color bands on zebra nch tness. Evolution 72(4):961–
976. h t t p s : / / w w w . j s t o r . o r g / s t a b l e / 4 8 5 7 6 6 5 0
Wickham H (2016) ggplot2: elegant graphics for data analysis.
Springer-, New York
Wilson EB (1927) Probable inference, the Law of Succession, and sta-
tistical inference. J Am Stat Assoc 22(158):209–212. h t t p s : / / d o i . o
r g / 1 0 . 1 0 8 0 / 0 1 6 2 1 4 5 9 . 1 9 2 7 . 1 0 5 0 2 9 5 3
Wong T (2007) Chew-toy color preference in kenneled dogs. UC
Davis University Writing Program. h t t p s : / / p r i z e d w r i t i n g . u c d a v i
s . e d u / s i t e s / p r i z e d w r i t i n g . u c d a v i s . e d u /  l e s / s i t e w i d e / p a s t i s s u e s / 0 7
% E 2 % 8 0 % 9 3 0 8 % 2 0 W O N G . p d f Accessed 16 Jan 2024
Yilmaz A, Hempel de Ibarra N, Kelber A (2022) High diversity of
arthropod colour vision: from genes to ecology. Philosophical
Trans Royal Soc B: Biol Sci 377(1862). h t t p s : / / d o i . o r g / 1 0 . 1 0 9 8
/ r s t b . 2 0 2 1 . 0 2 7 3
Publisher’s note Springer Nature remains neutral with regard to juris-
dictional claims in published maps and institutional aliations.
Polgár Z, Miklosi A, Gácsi M (2015). Strategies Used by Pet Dogs for
Solving Olfaction-Based Problems at Various Distances. PLoS
ONE. h t t p s : / / d o i . o r g / 1 0 . 1 3 7 1 / j o u r n a l . p o n e . 0 1 3 1 6 1 0
Pongrácz P, Ujvári V, Faragó T, Miklósi Á, Péter A (2017) Do you
see what I see? The dierence between dog and human visual
perception may aect the outcome of experiments. Behav Process
140:53–60. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j . b e p r o c . 2 0 1 7 . 0 4 . 0 0 2
Raine NE, Chittka L (2007) The adaptive signicance of sensory Bias
in a foraging context: Floral Colour preferences in the Bumblebee
Bombus terrestris. PLoS ONE 2(6):e556. h t t p s : / / d o i . o r g / 1 0 . 1 3 7 1
/ j o u r n a l . p o n e . 0 0 0 0 5 5 6
RStudio Team, RStudio (2019) RStudio: Integrated Development for
R. Inc., Boston, MA. http://www.rstudio.com/
Sarkar R, Sau S, Bhadra A (2019) Scavengers can be choosers: a study
on food preference in free-ranging dogs. Appl Anim Behav Sci
216:38–44. h t t p s : / / d o i . o r g / 1 0 . 1 0 1 6 / j . a p p l a n i m . 2 0 1 9 . 0 4 . 0 1 2
Signorell A (2023) DescTools: tools for descriptive statistics. R Pack-
age Version 0.99.48, h t t p s : / / C R A N . R - p r o j e c t . o r g / p a c k a g e = D e s c
T o o l s
Serpell J (1995) From paragon to pariah: Some reections on human
attitudes to dogs. The domestic dog: Its evolution, behaviour and
interactions with people, 245–256
Siniscalchi M, d’Ingeo S, Fornelli S, Quaranta A (2017) Are dogs red-
green colour blind? Royal Soc Open Sci 4(11):170869. h t t p s : / / d o
i . o r g / 1 0 . 1 0 9 8 / r s o s . 1 7 0 8 6 9
Staples R (1931) Color vision and color preference in infancy and
childhood. Psychol Bull 28(4):297–308. h t t p s : / / d o i . o r g / 1 0 . 1 0 3 7
/ h 0 0 7 0 9 7 1
1 3
Page 9 of 9 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com