Appearance matters: artificial marking alters aggression and stress.

Abstract

Artificial marking of animals for identification is frequently employed by researchers in the behavioral, biomedical, agricultural, and environmental sciences. The impact of artificial marking on experimental results is rarely explicitly considered despite evidence demonstrating that changes in phenotypic appearance can modify animal behavior and reproductive success. Here we present evidence that artificial marking of individuals within a social group has frequency-dependent effects on the behavior and physiology of domestic fowl (Gallus gallus domesticus). We demonstrate that when only 20 or 50% of individuals within a group were artificially marked, the marked birds received more aggression and had lesser body mass than the unmarked individuals within the same group. Furthermore, in groups in which only a small proportion of the individuals were marked, we report altered plasma epinephrine and dopamine levels in marked individuals. These effects of marking were imperceptible when all birds in a group were marked. This finding has important implications for animal research because, when only a subset of group members is artificially marked and used for data collection, the results obtained may not be representative of the population.

Full text

Page 1

1939
ENVIRONMENT, WELL-BEING, AND BEHAVIOR
2008 Poultry Science 87:1939–1946
doi:10.3382/ps.2007-00311
Key words: aggression, stress, social behavior, dominance, identification
ABSTRACT Artificial marking of animals for identi-
fication is frequently employed by researchers in the
behavioral, biomedical, agricultural, and environmen-
tal sciences. The impact of artificial marking on experi-
mental results is rarely explicitly considered despite
evidence demonstrating that changes in phenotypic ap-
pearance can modify animal behavior and reproductive
success. Here we present evidence that artificial mark-
ing of individuals within a social group has frequency-
dependent effects on the behavior and physiology of
domestic fowl (Gallus gallus domesticus). We demon-
strate that when only 20 or 50% of individuals within
a group were artificially marked, the marked birds re-
ceived more aggression and had lesser body mass than
the unmarked individuals within the same group. Fur-
thermore, in groups in which only a small proportion of
the individuals were marked, we report altered plasma
epinephrine and dopamine levels in marked individu-
als. These effects of marking were imperceptible when
all birds in a group were marked. This finding has im-
portant implications for animal research because, when
only a subset of group members is artificially marked
and used for data collection, the results obtained may
not be representative of the population.
Appearance Matters: Artificial Marking Alters Aggression and Stress
R. L. Dennis,*1 R. C. Newberry,† H.-W. Cheng,‡ and I. Estevez*2
*Department of Animal and Avian Sciences, University of Maryland, College Park 20742; †Center for the Study
of Animal Well-Being, Department of Animal Sciences and Department of Veterinary and Comparative Anatomy,
Pharmacology and Physiology, Washington State University, Pullman 99164-6520; and ‡Livestock Behavior
Research Unit, USDA-Agricultural Research Service and Animal Sciences Department,
Purdue University, West Lafayette, IN 47907
INTRODUCTION
Artificial marking is a clear manipulation of the
physical appearance of individual animals that can af-
fect behavior and reproductive success. For example,
color and symmetry of leg bands has been shown to
influence mate choice and mate guarding behavior of
wild song birds (Burley, 1988; Johnsen et al., 1997,
2000) and radio-tagging of fledgling Louisiana Water-
thrushes resulted in their removal from the nest and
abandonment by adults (Mattsson et al., 2006). Physi-
cal alterations to the feathers and comb of domestic
fowl (Gallus gallus domesticus) have also been found
to modify behavior and responses from conspecifics, in-
cluding aggressive interactions (Guhl, 1953; Guhl and
Ortman, 1953; Marks et al., 1960; Siegel and Hurst,
1962).
Markings can potentially alter multiple behavioral
and physiological systems in animals. Here we use ag-
gression in the domestic fowl as a model for examining
effects of markings because aggression and dominance
hierarchies have been well studied in chickens with
and without the use of different artificial marking sys-
tems. In groups of domestic fowl with an established
dominance hierarchy, most aggressive interactions are
directed toward subordinates (Guhl, 1953; Candland et
al., 1969; Queiroz and Cromberg, 2006). The formation
of dominance hierarchies and the propensity to show
aggression toward unfamiliar birds, is probably re-
lated to individual recognition (Bradshaw, 1991; Lind-
berg and Nicol, 1996). If so, chickens housed in large
groups may be unable to form dominance relationships
with all other birds in the group because they may be
unable to recognize every other bird individually and,
thus, establish a pair-wise relationship with each bird
(McBride, 1964). This may explain why aggressive dis-
plays are relatively rare in large groups of domestic
fowl (D’Eath and Keeling, 2003). On the other hand,
aggression and dominance have been linked to factors
such as body mass, comb size, and previous experience
of winning or losing (Guhl, 1953; Cloutier and New-
berry, 2000, 2002), and theoretical models demonstrate
that dominance hierarchies can emerge in large groups
of individuals varying in phenotype in the absence
of individual recognition (e.g., Dugatkin and Earley,
Received July 25, 2007.
Accepted April 22, 2008.
1 Current address: Livestock Behavior Research Unit, USDA-Agri-
cultural Research Service and Animal Sciences Department, Purdue
University, West Lafayette, IN 47907.
2 Corresponding author: iestevez@umd.edu
©2008 Poultry Science Association Inc.

Page 2

2004). Pagel and Dawkins (1997) suggested that, in
large groups of domestic fowl, status signals may take
the place of dominance relationships based on indi-
vidual recognition, proposing that identifiable marks,
or status badges, may enable recognition of the social
status of the bird, if not the individual itself. The obser-
vation that runts in large groups appear to be regularly
dominated by other birds suggests that some form of
dominance hierarchy based on phenotype is operating
in these groups.
In previous research on group size (GS) and aggres-
sive behavior in domestic fowl, artificially marked fo-
cal birds delivered fewer, and received more, aggres-
sive acts with increasing GS (Estevez et al., 2003). This
result may have been an artifact of marking a constant
number of focal birds within each group because the
proportion of marked to unmarked birds declined with
increasing GS. This possibility led us to hypothesize
that application of artificial marks to individuals with-
in a group results in frequency-dependent discrepan-
cies in the behavior and physiology of marked vs. un-
marked individuals.
Here, we compare the effects of artificial marking in
groups in which 20, 50, or 100% of the individuals were
marked in both small and large GS (10 and 50 birds
per group, respectively) on the aggression received and
given by marked birds, and the stress parameters of
marked and unmarked birds.
MATERIALS AND METHODS
Birds
Twelve hundred and sixty male 1-d-old domestic
fowl of a rapidly growing (meat-type) white-feathered
strain were randomly assigned to 1 of 21 groups of 10
birds and 21 groups of 50 birds for a density of 2.22 and
11.11 birds/m2, respectively. In each group, 20, 50, or
100% of the birds were marked by placing a mark on
the back of the head using a permanent black marker
and applying a pair of identification tags, yielding 7
replicate groups for each percentage marked within GS.
Identification tags (Cornetto et al., 2002) were made of
white paper disks (3 cm diameter), with a 2-digit black
number printed on both sides that was unique to each
bird. Tags were laminated to maintain their integrity
and affixed to either side of the neck at one day of age
with plastic filaments injected under the skin using the
Swiftack system (Heartland Animal Health Inc., Fair
Play, MO). Unmarked birds had no black mark or iden-
tification tags.
A block design was used, with 7 complete blocks lo-
cated in 2 rooms of one animal facility at the Univer-
sity of Maryland. Food and water were available ad li-
bitum. A low density feed was used to slow growth and
minimize chances of mobility problems. An artificial
lighting program of 14L:10D was followed. Animal care
was provided in adherence with Institutional Animal
Care and Use Committee guidelines. Instances of mor-
tality were low and were corrected for in the statistical
analysis.
Behavioral Observations
Direct behavioral observations were made by the
same observer for 10 min, 5 times in every 2-wk pe-
riod from 3 to 10 wk of age. For all threats, aggressive
pecks directed toward the head, and nonaggressive
pecks directed toward identification tags observed in
each of the groups, the observer recorded the giver and
receiver and whether they were marked or unmarked,
following the behavioral definitions of Estevez et al.
(2003) and Cornetto et al. (2002). Data were collected
using The Observer 3.0 software (Noldus Information
Technology, Wageningen, the Netherlands).
Physical and Physiological Measures
Data were obtained on body mass to the nearest
gram (BM) at 2, 5, and 11 wk of age, fecal corticoster-
one (CORT) at 5 wk of age, and plasma epinephrine
(EP), norepinephrine (NE), and dopamine (DA) con-
centrations at 11 wk of age. Data were collected from
a random sample of 3 marked and 3 unmarked birds
from each group, except the 20% marked groups of GS
10, from which 2 marked birds and 4 unmarked birds
were sampled. To minimize bias and disturbance of the
birds, birds to be sampled for blood, feces, and BM were
gently picked up and removed from the pen by a single
handler who walked very slowly in the pens. To control
for time of day, birds were sampled by experimental
block, with the order of sampling pens within block
balanced across blocks. Birds sampled within each pen
were selected in random order.
For CORT determination, birds were placed indi-
vidually in a large cage where a fecal sample was col-
lected immediately following excretion, placed in a
plastic?bag,?and?maintained?at?−80°C?until?extraction?
of CORT using 90% ethanol. Each sampled bird was re-
turned to its home pen immediately after collection. All
fecal samples were collected between 1000 and 1500
h in a single day. Following extraction (Dehnhard et
al., 2003), CORT concentrations were analyzed using a
commercial 125I CORT radioimmunoassay kit from MP
Biomedicals (Montreal, CA) as described previously by
Cheng et al. (2001). Concentrations of CORT were cal-
culated from a reference curve that ranged from 0.1 to
4.0 ng/mL, and the correlation coefficient was 0.995.
Blood samples were collected from the brachial vein
into heparinized tubes at 11 wk of age following a man-
ual restraint of at least 45 s, and the restraint time
(i.e., time from removal from the pen to completion of
blood collection) was documented for use in the statisti-
cal analysis of catecholamine concentrations. Samples
were centrifuged for 15 min at 3,000 rpm, and plasma
was?removed?and?maintained?at?−80°C?until?extraction?
and HPLC analysis of plasma catecholamines using
the Coulochem II electrochemical detector from ESA
DENNIS ET AL.
1940

Page 3

Biosciences (Chelmsford, MA) as described previously
(Cheng et al., 2001). The concentrations of DA, EP, and
NE were calculated from a reference curve that ranged
from 1.0 to 100 pg/mL, and the correlation coefficient
was 0.9998.
Statistical Analysis
Behavioral data were analyzed in a mixed-model
ANOVA with block as a random effect and age (four
2-wk periods from 3 to 10 wk of age) as a repeated mea-
sure that was fit to the appropriate covariance matrix
for each behavior. Variance partitioning was used to
correct for heterogeneity of variance and log-transfor-
mation was used when appropriate to attain normal-
ity. There were 5 marking treatments (marked birds
from 20, 50, and 100% marked pens, and unmarked
birds from 20 and 50% marked pens). Behaviors both
received and given by marked and unmarked individu-
als were compared across block, marking treatment,
GS, and age period. Per bird measures were analyzed
to correct for differences in GS due to mortalities. Least
square means were determined and contrasts were used
for means comparisons, using the Sidak adjustment to
maintain?an?experimental?α?of?0.05.?Body?mass?and?
CORT were analyzed using a blocked factorial ANO-
VA. Catecholamines were analyzed using a blocked
factorial analysis of covariance (ANCOVA) with nested
analysis; restraint time was the covariate. Because EP
and NE have been reported to increase logarithmically
with the time taken to sample blood (Matt et al., 1997),
using restraint time as a covariate allowed for unbi-
ased comparisons between treatment groups.
RESULTS
Aggression Received
Marked birds in both 20 and 50% pens were found to
receive significantly more pecks than their unmarked
pen mates (Figure 1a) and the marked birds in the pos-
itive control group (100% marked group). No signifi-
cant interactions were found between treatment and
group size for pecks received. Similarly, a significant
treatment effect was found for threats received (Figure
1b), with marked birds in 20% pens receiving signifi-
cantly more threats than their unmarked pen mates or
the birds in the 100% pens, but not significantly more
than marked birds in 50% pens. Frequency of threats
received also tended to be greater for marked than un-
marked birds in 50% pens.
Marked birds received significantly more pecks at
their identification tags in 20 and 50% pens compared
with the birds in 100% pens during period 1 (3 to 4
wk of age) but not at later ages (Figure 1c). Pecks at
tags in 20 and 50% pens were significantly more fre-
quent during period 1 compared with the later periods;
however, this difference across periods was not seen in
100% pens (Figure 1c).
Aggression Given
Not only did marked birds receive more aggression
in groups containing both marked and unmarked birds
but they also delivered less aggression than their un-
marked flock mates. Specifically, in age period 4 (9 to
10 wk of age) and GS 50, marked birds delivered sig-
nificantly fewer aggressive pecks than unmarked flock
mates (ANOVA: marking treatment by GS by age pe-
riod, F12,125 = 2.47, P = 0.006; Pairwise comparisons: t
= 3.24, P = 0.013 for 20 vs. 100% marked groups, and t
Figure 1. Artificial marking of a proportion of the flock alters
aggression received (least squares mean ± SEM per bird in 10 min;
letters show differences at P < 0.05). a) Aggressive pecks received
(marking treatment; F4,111 = 16.83, P < 0.0001, n = 7); b) aggressive
threats received (marking treatment; F4,39 = 5.25, P = 0.0018, n = 7);
and c) pecks at identification tags received (marking treatment by
age period; F12,147 = 9.81, P < 0.0001, n = 7). A,BDifferences between
age periods (3 to 4, 5 to 6, 7 to 8, and 9 to 10 wk) within treatment
group; a,bdifferences between treatment groups within age period.
APPEARANCE MATTERS
1941

Page 4

= 2.97, P = 0.028 for 50 vs. 100% marked groups; Table
1). In age period 1 (3 to 4 wk of age), marked birds in
50% marked groups of GS 10 and 50 also delivered few-
er threats than their unmarked counterparts (marking
treatment by age, F12,77 = 2.07, P = 0.029; Table 1). No
significant differences were detected in giving pecks at
tags.
Body Mass
Main effect contrasts revealed a significant effect of
marking on BM (least square means and SEM) in both
20 and 50% pen treatments and at both GS. Marked
birds were found to have significantly lesser BM than
unmarked penmates at 2 wk (245 ± 7.1 vs. 259 ± 7.5 g;
F1,54 = 7.16, P = 0.010) and 5 wk (1,681 ± 44.2 vs. 1,773
± 43.1 g; F1,54 = 9.82, P = 0.003) of age. Body mass at 11
wk was not significantly affected by marking (marked,
4,999 ± 69.0 g; unmarked, 5,020 ± 62.9 g; F1,52 = 0.02,
P = 0.888).
There was no significant main effect of percentage
marked per group on BM at any age, but there was
a significant interaction between percentage marked
and presence or absence of marking on the bird at 5 wk
of age. Least square means and SEM were as follows:
1,754 ± 28.2, 1,675 ± 28.2, 1,806 ± 28.2, 1,695 ± 33.4,
and 1,742 ± 26.4 g for 5-wk-old birds from 100% pens,
marked in 50% pens, unmarked in 50% pens, marked
in 20% pens, and unmarked in 20% pens, respectively,
where unmarked birds in 50% pens were significantly
heavier than their marked pen mates (t = 3.30, P =
0.003) and, although not significant, marked birds in
100% pens tended to be heavier than marked birds in
50% pens (t = 2.00, P = 0.058).
Catecholamines and CORT
A significant treatment effect was found in plasma
EP concentration. Marked birds in 20% pens had sig-
nificantly lesser EP concentrations compared with
their unmarked counterparts and the positive control
birds in 100% marked groups (Figure 2a). A treat-
ment by GS interaction of plasma DA concentration
was found to be significant. Marked birds in 20% pens
of GS 50 had significantly greater DA concentrations
than their unmarked counterparts (Figure 2b). No sig-
nificant differences were found between treatments in
plasma NE concentration (F4,50 = 2.00, P = 0.108; least
square mean ± SEM, 3.77 ± 0.67 pg/mL) or fecal CORT
metabolite (F4,28.3 = 1.10, P = 0.377; least square mean
± SEM, 2.35 ng/mL ± 0.30) concentration.
DISCUSSION
We have considered 4 possible mechanisms to ex-
plain the increase in aggression toward marked birds:
1) application of novel marks, 2) xenophobia based on
phenotypic dissimilarity, 3) marks perceived as signals
of low or high status, and 4) social challenge to con-
spicuous individuals.
Table 1. Artificial marking alters aggression given1
Proportion of
flock marked
Birds
within flockPeriod 1Period 2 Period 3Period 4
Aggressive pecks, GS 10
20% Marked
Unmarked
Marked
Unmarked
Marked
0.00 ± 0.01A,a
0.04 ± 0.02A,a
0.05 ± 0.02A,a
0.11 ± 0.02B,a
0.03 ± 0.02A,a
0.06 ± 0.02A,a
0.01 ± 0.02A,a
0.03 ± 0.02A,a
0.06 ± 0.02AB,a
0.04 ± 0.02A,a
0.07 ± 0.02A,a
0.02 ± 0.02A,a
0.01 ± 0.02A,a
0.03 ± 0.02A,a
0.01 ± 0.02A,a
0.00 ± 0.02A,a
0.03 ± 0.02A,a
0.01 ± 0.02A,a
0.07 ± 0.02AB,a
0.05 ± 0.02A,a
50%
100%
Aggressive pecks, GS 50
20% Marked
Unmarked
Marked
Unmarked
Marked
0.04 ± 0.01A,a
0.04 ± 0.01A,a
0.04 ± 0.01A,a
0.04 ± 0.01A,a
0.04 ± 0.01A,a
0.01 ± 0.01A,a
0.03 ± 0.01A,a
0.01 ± 0.01A,a
0.03 ± 0.01A,a
0.02 ± 0.01A,a
0.02 ± 0.01A,a
0.04 ± 0.01A,a
0.04 ± 0.01A,a
0.05 ± 0.01AB,a
0.03 ± 0.01A,a
0.05 ± 0.01A,a
0.09 ± 0.01B,b*
0.04 ± 0.01A,a
0.07 ± 0.01B,b*
0.04 ± 0.01A,a
50%
100%
Aggressive threats
20% Marked
Unmarked
Marked
Unmarked
Marked
0.02 ± 0.02A,a
0.05 ± 0.02A,a
0.04 ± 0.02A,a
0.14 ± 0.02B,b*
0.07 ± 0.02A,a
0.04 ± 0.01A,a
0.07 ± 0.01A,ab
0.07 ± 0.01A,ab
0.12 ± 0.01B,b*
0.04 ± 0.01A,a
0.05 ± 0.02A,a
0.06 ± 0.02A,a
0.03 ± 0.02A,a
0.09 ± 0.02AB,a
0.03 ± 0.02A,a
0.05 ± 0.01A,a
0.05 ± 0.01A,a
0.03 ± 0.01A,a
0.07 ± 0.01A,a
0.06 ± 0.01A,a
50%
100%
Pecks at tags
20% Marked
Unmarked
Marked
Unmarked
Marked
0.005 ± 0.003A,a
0.015 ± 0.003B,a
0.035 ± 0.003B,b*
0.027 ± 0.003B,ab
0.016 ± 0.003B,a
0.000 ± 0.003A,a
0.001 ± 0.003A,a
0.008 ± 0.003A,a
0.005 ± 0.003A,a
0.005 ± 0.003AB,a
0.000 ± 0.003A,a
0.000 ± 0.003A,a
0.003 ± 0.003A,a
0.000 ± 0.003A,a
0.000 ± 0.003A,a
0.000 ± 0.003A,a
0.000 ± 0.003A,a
0.001 ± 0.003A,a
0.001 ± 0.003A,a
0.001 ± 0.003A,a
50%
100%
A,BDifferences (P < 0.05) between age periods (3 to 4, 5 to 6, 7 to 8, and 9 to 10 wk) within treatment, with significantly more frequent events
in boldface.
a,bDifferences between marking treatments within age period, with significantly more frequent events marked with an asterisk (*).
1Values indicate least squares means ± SEM of number of acts given per bird in 10 min (n = 7) by marking treatment, group size (GS), and
age period.
DENNIS ET AL.
1942

Page 5

Application of Novel Marks
Novelty can elicit fear and aggression (Marin et al.,
2001), and handling, usually necessary for applica-
tion of marks, can alter stress and behavior (Gariepy
et al., 2002; Queiroz and Cromberg, 2006). Pecks at
identification tags occurred primarily during period 1
and then declined to a low level, suggesting that these
pecks were primarily associated with novelty of the
tags during the initial period after their application and
represented exploratory behavior rather than aggres-
sion. If elevated aggression toward marked birds was
primarily driven by the novelty of artificially applied
marks, or changes in behavior resulting from handling
received during marking, we would have expected the
level of threats and aggressive pecks directed toward
the heads of marked birds to decline with age as ob-
served for pecks at tags. In contrast, aggressive pecks
remained elevated over time in a frequency-dependent
manner according to the proportion of marked birds
in the group. Moreover, the greater frequency of ag-
gressive pecks and threats received by marked birds
in 20 and 50% marked groups compared with 100%
marked groups throughout the study provides evi-
dence that aggression was not triggered simply by the
application of novel marks but was, rather, influenced
by the frequency of the marked phenotype within the
population. Because aggressive behavior is generally
directed toward the head region, and behavior specifi-
cally directed toward the identification tags was short-
lasting, it is plausible that these effects of marking on
aggression were attributable solely to the presence
of the black mark on the back of the head. However,
because we marked birds with both a black mark and
identification tags, we cannot conclude from this study
whether one of these forms of marking alone or both
forms in combination were responsible for the effects
on aggression.
Xenophobia Based
on Phenotypic Dissimilarity
Kin recognition theory suggests that cooperation
and reduced aggression may be controlled by perceived
relatedness (Keller, 1997), which in some species ap-
pears to be based on degree of phenotypic similarity
(Hamilton, 1964a,b; Jaisson, 1991). Although pheno-
typic matching could be accomplished by comparing
one’s own appearance with that of others, mirror stud-
ies suggest that fowl do not recognize their own image
(Panksepp et al., 1980). A more likely mechanism for
phenotypic matching would be growing up with others
of the same phenotype, producing a preference for the
familiar phenotype. All birds were unmarked for a few
hours after hatch, which might have provided sufficient
time to develop a preference for unmarked birds. How-
ever, this would not explain the relatively low aggres-
sion in the 100% marked groups. After marking in the
20 and 50% marked groups, both phenotypes were kept
together, thereby providing no opportunity for birds to
develop exclusive familiarity with members of their
own phenotype. If phenotypic preference was based on
greater familiarity with the most common phenotype in
the population, it would not explain why marked birds
received more aggression than unmarked birds in 50%
marked groups where both phenotypes occurred at
equal frequency. Therefore, xenophobia based on phe-
notypic dissimilarity provides an implausible explana-
tion for our findings.
Marks Perceived as Signals
of Low or High Status
Artificial marks could inadvertently signal low com-
petitive ability similar to the way that dull or soiled
plumage signals reduced fitness associated with par-
asitic infection (Hamilton and Zuk, 1982), injury, or
pathogenic disease. Likewise, a small or artificially
dubbed comb signals low status and attracts aggres-
sion (Marks et al., 1960; Siegel and Hurst, 1962; Clout-
ier and Newberry, 2000). In our study, the artificial
marks could have had a similar effect, especially the
Figure 2. Effects of artificial marking of 20, 50, or 100% of the
flock on mean plasma catecholamine concentrations (±SEM) follow-
ing manual restraint. Letters show differences at P < 0.05. a) Plasma
epinephrine (EP) concentration (marking treatment, F4,121 = 3.43,
P = 0.0108, n = 7); b) plasma dopamine (DA) concentration [group
size (GS) by marking treatment, F4,181 = 6.08, P < 0.0001, n = 7].
A,BDifferences between GS within marking treatment; a,bdifferences
between treatment groups within GS.
APPEARANCE MATTERS
1943

Page 6

black mark on the back of the head, because of its prox-
imity to the socially significant comb. However, our
birds were sexually immature, and signals of status
are usually not evident at such a young age but be-
come prominent at sexual maturity. Even if the marks
were initially perceived as signaling low status, why
would marked birds in the 20 and 50% marked groups
continue to attract elevated aggression over time? We
applied marks to randomly selected birds that, on av-
erage, should have been equally likely to win or lose
fights. Therefore, we might have expected that birds
would learn that marks were not reliable signals of low
status and would no longer direct aggression differen-
tially toward marked birds.
Conversely, the marks may have been perceived as
badges of status. Badges of status may be used as hon-
est signals of dominance in large groups of domestic
fowl in which establishment of dominance through
fighting would be costly and inefficient (Pagel and
Dawkins, 1997). In birds, conspicuously pigmented
feather patches may provide a particularly potent
signal of fighting ability, enabling discrimination of
dominant individuals without risking injury and wast-
ing energy engaging in fights (Senar, 1998; Senar and
Camerino, 1998). If the artificial marks applied to our
birds were perceived by onlookers as representing hon-
est badges of status, we might have expected that birds
bearing them would have received less aggression than
unmarked birds in the same group; in fact, we found
the reverse.
Social Challenge
to Conspicuous Individuals
Instinctive deference to individuals bearing conspic-
uous marks would enable rapid invasion by cheaters
bearing such marks without accompanying fighting
ability, leading to the idea that there are costs associ-
ated with status badge bearing that only honest signal-
ers can withstand (Rohwer, 1982; Senar, 1998). If social
challenge is one such cost (Tibbetts and Dale, 2004),
birds bearing conspicuous artificial marks should at-
tract elevated aggression relative to unmarked birds
in the same flock and the fewer conspicuously marked
birds in a population, the more challenges they should
receive. Our results are consistent with this prediction.
The lowest level of aggression was received by marked
birds in the 100% marked flocks, where the marks were
not conspicuous because they were held by all birds.
Although we did not find a significant difference in ag-
gression received by marked birds in the 20 and 50%
marked groups, the latter did tend to receive levels of
aggression intermediate between those in 20 and 100%
marked groups.
The question remains as to why aggression toward
marked birds would continue over time in a randomly
marked population with a theoretically equal chance
of winning and losing challenges. Estevez et al. (2007)
emphasize the importance of an individual’s ability to
recoup the cost of aggressive encounters in their deci-
sion to engage in these contests as an important factor
in multiple different aggression models. If there is a
social cost associated with badge holding (Rohwer and
Rohwer, 1978; Cloutier and Newberry, 2002; Tibbetts
and Dale, 2004) and, on average, badge holders were
not of greater competitive ability than nonbadge hold-
ers and thus able to overcome this cost, one would ex-
pect that, on average, the condition of marked birds
would deteriorate relative to that of unmarked birds,
thereby increasing their probability of losing encoun-
ters (Cloutier and Newberry, 2000). Given that previ-
ous experience of losing is an important determinant of
the outcome of future social encounters (Hsu and Wolf,
1999; Cloutier and Newberry, 2000; Beacham, 2003),
losing encounters would further cement low status, ex-
plaining why marked birds gave less aggression than
they received.
In accordance with the proposal that marking and
consequent increased aggressive challenge would re-
duce body condition (or fitness), we found that marked
birds had lower BM than unmarked birds in both 20
and 50% marked groups, differences suggestive of
stress (Nicol et al., 1999; Keeling et al., 2003). Low-
er BM in the marked birds may have been related to
energy costs incurred in an attempt to avoid aggres-
sion, lower food intake due to monopolization of food
resources by the more aggressive unmarked birds, or
stress associated with low social status (Senar et al.,
2000). Differences in BM were attenuated by 11 wk of
age, probably because growth slows as birds approach
their mature size allowing smaller birds to catch up in
BM.
Marking affected catecholaminergic reactivity at 11
wk of age. Marked birds in the 20% marked groups
had a depressed EP response and unaltered NE re-
sponse, suggestive of reduced stress-coping ability and
a suppressed fight-or-flight response compared with
their unmarked counterparts or birds in 100% marked
groups (Figure 2a). As stress hormones, both EP and
NE participate in several physiological and behavioral
processes (Kvetnansky and Mikulaj, 1970; Dillon et
al., 1992; Dobrakovova et al., 1993; Matt et al., 1997).
Whereas acute stress is associated with elevated plas-
ma EP (Matt et al., 1997; Wortsman, 2002), prolonged
chronic or chronic intermittent stress regimens have
been observed to lower EP concentrations without
altering NE concentrations in rats (Kvetnansky and
Mikulaj, 1970; Dobrakovova et al., 1993) and humans
(Matthews et al., 2001; Grant et al., 2003). Sensitivity
of the EP system has been shown following pharmaco-
logical and genetic manipulation of aggressiveness in
laying hens without altering NE (Dennis et al., 2006).
Our results are consistent with previous findings indi-
cating a dissociated stress response between the func-
tions of adrenal medullary (indicated by EP) and sym-
pathetic neural (indicated by NE) activity. Suppression
of sympathetic adrenal responsiveness is consistent
DENNIS ET AL.
1944

Page 7

with the reduction in aggressiveness of the marked
birds. We also observed increased plasma DA concen-
tration in marked birds of 20% groups at GS 50 (Fig-
ure 2b), suggestive of increased levels of chronic social
stress (Cheng et al., 2001) as previously seen in chick-
ens subjected to the chronic social stress of high stock-
ing density (10 birds per cage) compared with siblings
maintained at 2 birds per cage (Cheng et al., 2001).
These results support our assertion that effects of ar-
tificial marking were stronger, to the point of produc-
ing endocrinological changes, the lower the proportion
of marked individuals in a group. Although consistent
with the hypothesis that, by 11 wk of age, marked birds
in the 20% marked groups were experiencing chronic
stress as a consequence of receiving repeated aggres-
sion, a definitive explanation cannot be given because
we did not collect blood at earlier ages.
Whereas acute social stress would be expected to
increase CORT (Carere et al., 2003), marking had no
effect on fecal CORT metabolite concentrations in our
study, possibly due to dampening of CORT responsive-
ness in birds exposed to chronic social stress as a con-
sequence of being marked. Consistent with this inter-
pretation, Hester et al. (1996) found that chronic heat
stress in chickens resulted in no significant effect on
circulating CORT concentrations. Fecal CORT metabo-
lite concentrations have been previously validated for
domestic fowl and have been shown to exhibit a delay
of approximately 4 h compared with plasma CORT
levels (Dehnhard et al., 2003). Circadian fluctuation
of CORT is unlikely to provide a plausible explanation
for our results because fecal samples were collected by
experimental block and time was accounted for in the
statistical model. Furthermore, because fecal sampling
provides an average estimate of circulating CORT lev-
els over time, it is relatively insensitive to precise sam-
pling time (Möstl and Palme, 2002).
Conclusion
We conclude that application of artificial identifica-
tion marks to only a proportion of the individuals in
a group can alter aggressive behavior, BM, and endo-
crine parameters in the domestic fowl. Thus, when ar-
tificial marking is required for this species, both the
number and proportion of birds marked within a study
population should be taken into consideration in exper-
imental design and interpretation of results. Further
research will be needed to elucidate the social impact
of form, color, and body location of different types of ar-
tificial marking. The domestic fowl is highly dependent
on both vision and pecking with the beak for sensory
information, and the extent to which marking affects
species less dependent on these sensory modalities is
as yet unknown. Nevertheless, our findings sound a
cautionary note for animal researchers because, when
only a subset of group members is artificially marked
and used for data collection, the results obtained may
not be representative of the population.
ACKNOWLEDGMENTS
We thank L. Douglass (University of Maryland,
College Park) for statistical analysis, the graduate
students and laboratory staff of I. Estevez and H.-W.
Cheng for technical assistance, and the farm crew at
University of Maryland’s Upper Marlboro facility for
animal care. This research was funded by a competi-
tive grant from the Maryland Agriculture Experiment
Station to I. Estevez.
REFERENCES
Beacham, J. L. 2003. Models of dominance hierarchy forma-
tion: Effects of prior experience and intrinsic traits. Be-
havior 140:1275–1303.
Bradshaw, R. H. 1991. Discrimination of group members
by laying hens Gallus domesticus. Behav. Processes
244:143–151.
Burley, N. 1988. Wild zebra finches have band-color prefer-
ences. Anim. Behav. 36:1235–1237.
Candland, D. K., L. D. Taylor, J. M. Leiphart, and S. P. So-
low. 1969. Heart rate aggression and dominance in domes-
tic chickens. J. Comp. Physiol. Psychol. 1:70–76.
Carere, C., T. G. G. Groothuis, E. Mostl, S. Daan, and J. M.
Koolhaas. 2003. Fecal corticosteroids in a territorial bird
selected for different personalities: Daily rhythm and the
response to social stress. Horm. Behav. 43:540–548.
Cheng, H. W., G. Dillworth, P. Singleton, Y. Chen, and W. M.
Muir. 2001. Effects of group selection for productivity and
longevity on concentrations of serotonin, catecholamines,
and corticosterone of laying hens. Poult. Sci. 80:1278–
1285.
Cloutier, S., and R. C. Newberry. 2000. Recent social experi-
ence, body weight and initial patterns of attack predict the
social status attained by unfamiliar hens in a new group.
Behavior 137:705–726.
Cloutier, S., and R. C. Newberry. 2002. Differences in skel-
etal and ornamental traits between laying hen cannibals,
victims and bystanders. Appl. Anim. Behav. Sci. 77:115–
126.
Cornetto, T., I. Estevez, and L. W. Douglass. 2002. Using ar-
tificial cover to reduce aggression and disturbances in do-
mestic fowl. Appl. Anim. Behav. Sci. 75:325–336.
D’Eath, R. B., and L. J. Keeling. 2003. Social discrimina-
tion and aggression by laying hens in large groups: From
peck orders to social tolerance. Appl. Anim. Behav. Sci.
84:197–212.
Dehnhard, M., A. Schreer, O. Krone, K. Jewgenow, M.
Krause, and R. Grossmann. 2003. Measurement of plasma
corticosterone and fecal glucocorticoid metabolites in the
chicken (Gallus domesticus), the great cormorant (Phalac-
rocorax carbo), and the goshawk (Accipiter gentiles). Gen.
Comp. Endocrinol. 131:345–352.
Dennis, R. L., W. M. Muir, and H. W. Cheng. 2006. Effects of
raclopride on aggression and stress in diversely selected
chicken lines. Behav. Brain Res. 175:104–111.
Dillon, J. E., M. H. Raleigh, M. T. McGuire, D. Bergin-Pol-
lack, and A. Yuwiler. 1992. Plasma-catecholamines and
social-behavior in male vervet monkeys (Cercopithecus ae-
thiops sabaeus). Physiol. Behav. 51:973–977.
Dobrakovova, M., R. Kvetnansky, Z. Oprsalova, and D. Je-
zova. 1993. Specificity of the effect of repeated handling on
sympathetic-adrenomedullary and pituitary-adrenocorti-
cal activity in rats. Psychoneuroendocrinology 18:163–
174.
Dugatkin, L. A., and R. L. Earley. 2004. Individual recogni-
tion, dominance hierarchies and winner and loser effects.
Proc. R. Soc. Lond. B. Biol. Sci. 271:1537–1540.
APPEARANCE MATTERS
1945

Page 8

Estevez, I., I. L. Andersen, and E. Naevdal. 2007. Group size,
density and social dynamics in farm animals. Appl. Anim.
Behav. Sci. 103:185–204.
Estevez, I., L. J. Keeling, and R. C. Newberry. 2003. Decreas-
ing aggression with increasing group size in young domes-
tic fowl. Appl. Anim. Behav. Sci. 84:213–218.
Gariepy, J. L., R. M. Rodriguiz, and B. C. Jones. 2002. Han-
dling, genetic and housing effects on the mouse stress sys-
tem, dopamine function and behavior. Pharmacol. Bio-
chem. Behav. 73:7–17.
Grant, I., C. L. McKibbin, M. J. Taylor, P. Mills, J. Dimsdale,
M. Ziegler, and T. Patterson. 2003. In-home respite inter-
vention reduces plasma epinephrine in stressed Alzheimer
caregivers. Am. J. Germ. Psych. 11:62–72.
Guhl, A. M. 1953. Social behavior of the domestic fowl. Tech-
nical Bull. 73:1–43. Kansas Agricultural Experimental
Station, Manhattan.
Guhl, A. M., and L. L. Ortman. 1953. Visual patterns in
the recognition of individuals among chickens. Condor
55:287–298.
Hamilton, W. D. 1964a. The genetic evolution of social behav-
ior. I. J. Theor. Biol. 7:1–17.
Hamilton, W. D. 1964b. The genetic evolution of social behav-
ior. II. J. Theor. Biol. 7:17–52.
Hamilton, W. D., and M. Zuk. 1982. Heritable true fitness and
bright birds a role for parasites? Science 218:384–387.
Hester, P. Y., W. M. Muir, and J. V. Craig. 1996. Group se-
lection for adaptation to multiple-hen cages: Humoral im-
mune response. Poult. Sci. 75:1315–1320.
Hsu, Y., and L. L. Wolf. 1999. The winner and loser effect:
Integrating multiple experiences. Anim. Behav. 57:903–
910.
Jaisson, P. 1991. Kinship and fellowship in ants and social
wasps. Pages 60–93 in Kin Recognition. P. Hepper, ed.
Cambridge University Press, Cambridge, UK.
Johnsen, A., P. Fiske, T. Amundsen, J. T. Lifjeld, and P. A.
Rohde. 2000. Color bands, mate choice and paternity in
the bluethroat. Anim. Behav. 59:111–119.
Johnsen, A., J. T. Lifjeld, and P. A. Rohde. 1997. Colored leg
bands affect male mate-guarding behavior in the blue-
throat. Anim. Behav. 54:121–130.
Keeling, L. J., I. Estevez, R. C. Newberry, and M. G. Correia.
2003. Production-related traits of layers reared in differ-
ent sized flocks: The concept of problematic intermediate
group sizes. Poult. Sci. 82:1393–1396.
Keller, L. 1997. Indiscriminate altruism: Unduly nice par-
ents and siblings. Trends Ecol. Evol. 12:99–103.
Kvetnansky, R., and L. Mikulaj. 1970. Adrenal and urinary
catecholamines in rats during adaptation to repeated im-
mobilization stress. Endocrinology 87:738–743.
Lindberg, A. C., and C. J. Nicol. 1996. Effects of social and
environmental familiarity on group preferences and spac-
ing behavior in laying hens. Appl. Anim. Behav. Sci.
49:109–123.
Marin, R. H., P. Freytes, D. Guzman, and R. B. Jones. 2001.
Effects of an acute stressor on fear and on the social rein-
statement responses of domestic chicks to cagemates and
strangers. Appl. Anim. Behav. Sci. 71:57–66.
Marks, H. L., P. B. Siegel, and C. Y. Kramer. 1960. Effects
of comb and wattle removal on the social organization of
mixed flocks of chickens. Anim. Behav. 8:192–196.
Matt, K. S., M. C. Moore, R. Knapp, and I. T. Moore. 1997.
Sympathetic mediation of stress and aggressive competi-
tion: Plasma catecholamines in free-living male tree liz-
ards. Physiol. Behav. 61:639–647.
Matthews, K. A., B. B. Gump, and J. F. Owens. 2001. Chron-
ic stress influences cardiovascular and neuroendocrine
responses during acute stress and recovery, especially in
men. Health Psychol. 20:403–410.
Mattsson, B. J., J. M. Meyers, and R. J. Cooper. 2006. Detri-
mental impact of radiotransmitters on juvenile Louisiana
Waterthrushes. J. Field Ornithol. 77:173–177.
McBride, G. 1964. Social discrimination + subflock structure
in domestic fowl. Anim. Behav. 12:264–267.
Möstl, E., and R. Palme. 2002. Hormones as indicators of
stress. Domest. Anim. Endocrinol. 23:67–74.
Nicol, C. J., N. G. Gregory, T. G. Knowles, I. D. Parkman, and
L. J. Wilkins. 1999. Differential effects of increased stock-
ing density, mediated by increased flock size, on feather
pecking and aggression in laying hens. Appl. Anim. Be-
hav. Sci. 65:137–152.
Pagel, M., and M. S. Dawkins. 1997. Peck orders and group
size in laying hens: Future contracts for non-aggression.
Behav. Processes 40:13–25.
Panksepp, J., N. J. Bean, P. Bishop, T. Vilberg, and T. L. Sah-
ley. 1980. Opioid blockage and social comfort in chicks.
Pharmacol. Biochem. Behav. 13:673–683.
Queiroz, S. A., and V. U. Cromberg. 2006. Aggressive behav-
ior in the genus Gallus sp. Braz. J. Poult. Sci. 8:1–14.
Rohwer, S., and F. C. Rohwer. 1978. Status signalling in Har-
ris sparrows – Experimental deceptions achieved. Anim.
Behav. 26:1012–1022.
Rohwer, S. A. 1982. The evolution of reliable and unreliable
badges of fighting ability. Am. Zool. 22:531–546.
Senar, J. C. 1998. Plumage coloration as a signal of social
status. Pages 1669–1686 in Proc. 22nd Int. Ornithological
Congr., Durban, South Africa. N. J. Adams and R. H. Slo-
tow, ed. BirdLife South Africa, Johannesburg.
Senar, J. C., and M. Camerino. 1998. Status signalling and
the ability to recognize dominants: An experiment with
siskins (Carduelis spinus). Proc. R. Soc. Lond. B. Biol. Sci.
265:1515–1520.
Senar, J. C., V. Polo, F. Uribe, and M. Gomendio. 2000. Status
signalling, metabolic rate and body mass in the siskin: The
cost of being subordinate. Anim. Behav. 59:103–110.
Siegel, H. S., and D. C. Hurst. 1962. Social interaction among
females in dubbed and undubbed flocks. Poult. Sci.
41:141–145.
Tibbetts, E. A., and J. A. Dale. 2004. A socially enforced sig-
nal of quality in a paper wasp. Nature 432:218–222.
Wortsman, J. 2002. Role of epinephrine in acute stress. En-
docrinol. Metab. Clin. North Am. 31:79–106.
DENNIS ET AL.
1946