Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Pekin ducks are motivated to access their nest site and exhibit
a stress-induced hyperthermia when unable to do so
Lorelle A. Barrett
a,
⁎
,1
,Shane K. Maloney
b
, Dominique Blache
a
a
School of Agriculture & Environment, The University of Western Australia, 35 Stirling Highway, Perth 6009, Western Australia, Australia
b
School of Human Sciences, The University of Western Australia, 35 Stirling Highway, Perth 6009, Western Australia, Australia
abstractarticle info
Article history:
Received 27 April 2020
Received in revised form 25 August 2020
Accepted 28 August 2020
Available online xxxx
Keywords:
Behavioural DEMAND
Frustration
Nesting
Poultry
Stress response
The origins of floor-laying in ducks could be low motivation for a nest, or stress related to difficulties with
accessing a nest (e.g. competition). Using a behavioural demand test, we investigated if increasing the work re-
quired to access their nest impacted ducks’behaviour andtwo indicators of stress:egg corticosterone concentra-
tion and elevation of core body temperature (stress-induced hyperthermia (SIH)). Twelve laying Pekin ducks
previously trained in an operant push-door task were required to use a push-door to access their nest. The
door was loaded with increasing weight (0–160% of individual BW, four nights per workload) and eventually
blocked to prevent nest access. Before testing, temperature data loggers were implanted in the abdomen. Eggs
were collected daily to measure corticosterone concentrations. Behaviour towards the push-door was quantified.
Three birds were excluded from the experiment at an early stage. Five of the nine remaining birds pushed all
workloads up to 160% BW and attempted to pass the blocked door, with another two birds pushing up to 80
and 140% BW. For those that pushed at all workloads (n = 5) the area under the curve (AUC)ofhyperthermia
was larger at workloads of 80% (P< 0.001), 120% (P< 0.01), 140% (P< 0.001), 160% (P< 0.001), and when
the door was blocked (P< 0.001), compared with 0%. On the first night when the door was blocked, all five
birds pushed more at the door, but no attempts were made to push on the following 3 nights, yet the AUC of hy-
perthermia did not differbetween nights 2–4 of the blockeddoor, compared withthe first night that the door was
blocked. Increasing workload and inability to access the nest had no effect on corticosterone in egg albumen. It
was concluded that laying Pekin ducks were motivated to access a nest. Although it was not possibleto differen-
tiate metabolic from psychogenic stress on the first nightthat nest access was denied, we suggest that theoccur-
rence of hyperthermiaon the subsequent nightswas due to SIH resulting from frustration at their inability touse
their preferred nest. Floor-laying therefore is unlikely due solely to low nest-seeking motivation. Egg corticoste-
rone was not a relevant indicator of acute stress. Strategies to improve nest availability (e.g. decreasing compe-
tition) could improve the welfare of commercial ducks.
Crown Copyright © 2020 Published by Elsevier Inc. on behalfofTheAnimalConsortium.Thisisanopenaccessar-
ticle under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Implications
The study is the first to demonstrate that laying Pekin ducks are
highly motivated to access a nest site and likely experience frustration
when theyare unable to use their nest. It appears that, as in other poul-
try species, nests are a highly valued resource for ducks.The floor-laying
behaviour that occurs in commercial duck farms thus cannot be fully
explained by ducks not engaging in nest-seeking behaviour. Further
consideration needs to be given towards adequate provision of appro-
priate nests to laying ducks in large-scale farming operations.
Introduction
On commercial farms, breeding Pekin ducks are provided with nest
boxes on the ground, in which it is hoped they will lay eggs. An industry
issue is that eggs may be laid on the floor instead of in the nest boxes, a
behaviourreferred to as floor-laying. Floor-laid eggs negatively impact a
farm’sproductionefficiency through lost potential income from break-
ages, contamination or decreased hatchability (Appleby et al., 2004);
thus, floor-laying is considered an undesirable behaviour. Factors that
contribute to floor-laying in Pekin ducks have not been well researched.
There is some evidence that competition for nests contributes to floor-
laying (Barrettetal.,2019), and that ducks prefer to lay in nests that al-
ready contain eggs, and nests that are better concealed (Makagon et al.,
2011;Makagon and Mench, 2011). Otherwise, little is known about the
choice that a duck makes about laying site. The expression of nesting be-
haviour in birds is hormonally regulated and occurs approximately24 h
Animal xxx (xxxx) xxx
⁎Corresponding author.
E-mail address: Lorelle.barrett@research.uwa.edu.au (L.A. Barrett).
1
The New Zealand Veterinary Association, PO Box 11,212, Wellington 6142, New
Zealand
ANIMAL-100067; No of Pages 10
https://doi.org/10.1016/j.animal.2020.100067
1751-7311/Crown Copyright © 2020 Pu blished by Elsevi er Inc. on behalf of The Animal Consortium. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Contents lists available at ScienceDirect
Animal
The international journal of animal biosciences
Please cite this article as: L.A. Barrett, S.K. Maloney and D. Blache, Pekin ducks are motivated to access their nest site and exhibit a stress-induced
hyperthermia when u..., Animal, https://doi.org/10.1016/j.animal.2020.100067
after the event that initiates it: ovulation and the subsequent release of
oestrogen and progesterone from post-ovulatory follicles (Appleby
et al., 2004). The performance of nesting behaviour in birds (and other
nest-building species, e.g. pigs) is considered to be a behavioural need
for these animals because the motivation to perform it is driven largely
by these internal rather than external factors (Mason and Burn, 2011).
However, the motivation of commercial Pekin ducks to use nest boxes,
and whether a lack of motivation is associated with floor-laying, has
never been assessed.
Measuring the motivation of an animal to perform a behaviour can
inform us about how they experience their environment. A positive af-
fective state will result whenan animal can express a behaviour that it is
strongly motivated to perform, whereas the inability to express that be-
haviour is likely to result in a negative affective state (Mason and
Bateson, 2009), which is associated with poor welfare outcomes. Thus,
by understanding the motivation of a duck for a nest site, management
strategies that help to minimise floor-laying can be identified. The iden-
tification of such strategies could result inboth improved production ef-
ficiency and welfare outcomes,as birds will have a better opportunity to
express nesting behaviour.
An animal’s level of motivation, and thus need, for a resource can be
assessed using behavioural demand methods that require an animal to
pay an increasing cost for a resource (Dawkins, 1990). Early literature
proposed that behavioural demand curves could be used to determine
the ‘elasticity’of demand for a resource. However, such curves rely on
a rate of change relationship, whereby the amount of resource con-
sumed must co-vary with the increasing ‘price’paid (Mason et al.,
1998). Other measures of demand, such as the maximum price an ani-
mal is willing to pay for a resource, have since been proposed as better
measures for resources that might be considered ‘all or nothing’(Olsson
et al., 2002). A bird’s access to, and use of, a nest site may be considered
an ‘all or nothing’resource. The motivation of chickens to access a nest
site has been well studied using behavioural demand techniques (e.g.
Cooper and Appleby, 1995, 2003), and the lack of a suitable nest site
leads to behavioural indicators of frustration (Mason and Burn, 2011).
Frustration refers to an animal’s underling psychological state (Haskell
et al., 2004) and is recognised as being aversive (Mason and Burn,
2011). Whether the need for a suitable nest is of high importance to lay-
ing Pekin ducks has yet to be determined. However, a behavioural de-
mand method for ducks was recently developed, using a push-door
operant task (Barrett and Blache, 2019) to enable testing of this
question.
Physiological quantification of the stress that is associated with an
animal’s inability to obtain a resource can provide further evidence
that a particular resource might be considered a need. The measure-
ment of glucocorticoid hormones is the most commonly used physio-
logical indicator of stress (Blache et al., 2017), of which corticosterone
is most often used in birds. The use of egg albumen, rather than plasma,
offers a non-invasive means of determining the physiological impact
that a particular stressor has on corticosterone levels (Downing and
Bryden, 2008;Royo et al., 2008). In chickens, the concentration of corti-
costerone in egg albumen has been used to explore the correlation be-
tween pre-laying activity and corticosterone levels (Cronin et al.,
2012) and to investigate the stress response to single vs group housing
(Royo et al., 2008), handling, increased ambient temperatures and mov-
ing hens between cages (Downing and Bryden, 2008). The use of corti-
costerone in egg albumen as an indicator of stress has not previously
been reported in Pekin ducks.
Measuring changes in core body temperature (T
c
)isanothermeans
of remotely quantifying the physiological response to a stressor. The el-
evation of T
c
that results from a stressful event is known as stress-
induced hyperthermia (SIH). Across species, SIH has been recorded in
response to a variety of stressors, such as the handling of pigeons and
eider ducks (Cabanac and Guillemette, 2001;Bittencourt et al., 2015),
the prolonged restraint of Pekin ducks (Gray et al., 2008), open-field
tests in sheep (Pedernera-Romano et al., 2010) and exposure to social
stress in people (Vinkers et al., 2013)orrats(Kataoka et al., 2020).
Stress-induced hyperthermia has also been shown to correlate well
with other physiological stress indicators,such as measures of HPA (Hy-
pothalamic-pituitary-adrenal) axis activity (Blache et al., 2017). As far
as it is possible to determine from the existing literature, T
c
measure-
ments have never been used to determine whether animals exhibit
SIH during a behavioural demand test.
The aims of this study were to: 1) use behavioural demand methods
to assess the motivation of laying Pekin ducks to access an established
nest site and 2) investigate whether ducks exhibit signs of stress or frus-
tration in response to increasing cost, and ultimately the inability to ac-
cess their nest site, quantified through changes in behaviour, egg
corticosterone and T
c
. We hypothesised that ducks would be highly mo-
tivated to gain access to a nest site and so perform increasing amounts
of work until they were physically incapable of doing so, and that a
stress-like response would be exhibited when they were unable to ac-
cess the nest, indicated by changes in behaviour, T
c
, or the concentration
of corticosterone in eggs.
Material and methods
Animals and housing
Twelve female Pekin ducks, previously trained in an operant
push-door task (Barrett and Blache, 2019), were used. The ducks were
approximately 26 weeks of age when the study commenced, having ar-
rived at the research facility when they were 20 weeks old. They were
sourced from a boutique commercial free-range farm and had no prior
nesting experience or exposure to nest boxes. Laying first commenced
at 22 weeks, with all birds in regular lay by 27 weeks of age. Until the
time of surgery for the implantation of data loggers (see below), the
ducks were housed in two outdoor pens (approximate dimensions 12
m × 4 m, see Barrett and Blache, 2019 for full details), with six birds,
two nest boxes and one behavioural demand unit (BDU)ineachpen.
In the 2 weeks prior to surgery, the ducks were habituated to spend
time alone in the BDU, because they were required to be individually
housed in a BDU overnight during the experimental period. Over the
2-week period, each duck was placed in a BDU for gradually longer
times, while the remaining five birds in each pen were free to engage
in their normal activities. Time periods began at 10min on day 1 and in-
creased by 10 min each day for the first week, and by half an hour every
2 days in the second week, ending once each duck had completed 2.5 h
in the BDUs. Food and water were provided in the BDU during this time.
After surgery, the birds had a 2-week recovery period (without con-
tainment in a BDU) in a third pen that was adjacent to the original two
pens, and then four birds were placed into each of the three pens. Each
pen contained four BDUs that were modified from the original design in
Barrett and Blache (2019), so that they consisted of two solid walls and
two steel mesh walls.The design changeswere made to furthermitigate
social stress by allowing the ducks to havevisual, olfactory and physical
contact through the mesh. The BDU was divided into two sections by a
partition containing the push-door, a holding area that contained food
and water, and a ‘reward’area that contained a nest box (hereafter re-
ferred to as the nest area; Fig. 1). Each unit was further protected with
a corrugated iron shelter (approx. 1.2 m height × 1.1 m width × 1.6 m
length) that provided additional weather-proofing.
At all stages, the ducks had free movement throughout their pen
during the day, except when they were undergoing BDU habituation.
They were fed once daily with a standard ration of chicken layer pellets,
as well as being free to forage within the pen, and were provided with
both fresh drinking water and open water troughs for bathing. During
the behavioural demand test, the ducks were individually housed over-
night in their own BDU. The ducks were weighed weekly, with the mass
range being 2.7–4.1 kg at the beginning, and 2.8–4.2 kg at the comple-
tion, of the experimental protocol.
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
2
Core temperature data loggers
Prior to surgical implantation, the data loggers (DST micro-T, Star
ODDI, Iceland) were programmed to record temperature every 5 min.
They were then sterilised by soaking in a chlorhexidine solution for a
minimum of 48 h before surgery.
Two days before surgery, the ducks were brought indoors and indi-
vidually housed (F-suite, Tecniplast, Australia). On the day of surgery,
the ducks were premedicated with butorphanol (1.5 mg/kg; Butorgesic,
Ilium Veterinary Products, Australia) and midazolam (0.2 mg/kg;
Hypnovel®, Roche, Australia) intramuscularly. Pre-operative meloxicam
(0.5 mg/kg; Metacam®, Boehringer, Australia) was administered
subcutaneously. Once the premedicant had taken effect, the duck was
anaesthetised with isoflurane (Isoflo™, Zoetis, Australia) in oxygen,
initially via an anaesthetic face-mask, and then via intubation.
Once the duck was anaesthetised, the surgical site was prepared by
the plucking of feathers, then an iodine surgical scrub followed by a
chlorhexidine/alcohol spray. The site was in the ventral midline, be-
tween the anatomical landmarks of the caudal sternum and the pubic
bones. A sterile drape was then placed over the site, and a small incision
into the body cavity was made. The sterile data logger was placed into
the body cavity, and the incision was then sutured closed in two layers.
Wound glue (Vetbond™; 3 M Animal Care, St Paul, USA) was applied to
the site to reduce the risk of ducks loosening the knots during preening/
grooming activities.
Once surgery was completed, the isoflurane was discontinued and
the duck was extubated when awake. When the bird was able to sit
up on its own, it was returned to a cage and monitored closely for the
next 2 h. The following day, the birds were returned to an outdoor
pen. Meloxicam (0.5 mg/kg/day) was continued for 2 days post-
operatively and the wound sites were checked daily during the recovery
period. All birds recovered uneventfully.
After the experiment, the ducks were humanely euthanised with
pentobarbitone (150 mg/kg; Lethabarb®, Virbac, Australia) and the
data loggers were retrieved. The loggers were then calibrated at 33,
36, 39 and 42 °C, before all data were downloaded. Individual calibra-
tion equations were derived for each logger and applied to all of the
data prior to analysis.
Assessing motivation to access a nest site
After the recovery period, the ducks were held overnight in their
individual BDUs to assess their motivation to access a nest site. The
birds were typically held in the units between 1800 and 0700 h.
For the first 2 weeks, the door between the holding and nest area
was open, to allow the birds to adjust to being in the units overnight
and to establish a nest. To assess the ducks’motivation to access their
nest, a behavioural demand experiment was then conducted, using
the push-door method established by Barrett and Blache (2019).
The ducks were required to push through an increasingly weighted
door to access their nest (Fig. 2). The test began with the door closed
but unweighted (0%), then the door was weighted in 20% increments
of the individual bird’s BW. Each weight was in place for 4 consecu-
tive nights. Failure was classified when a duck did not pass through
Fig. 1. Diagram (A) and photograph (B) of the internal arrangement of each behavioural demand unit used by Pekin ducks.
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
3
the door to access the nest for 3 of the 3 nights at any given weight
level. The intent was to continue testing the birds to their maximum
pushing capability. However, after reviewing video footage at the
level of 160% BW, the decision was made to terminate at that weight,
due to safety and welfare concerns for the birds. The ducks were no-
ticeably struggling to push the door and often vigorously flapped
their wings when making an attempt. The authors were concerned
that any additional weight increase would prevent the ducks from
being able push the door up far enough to pass through, and might
instead result in them getting injured if they were not able to exert
sufficient upwards force to create enough space to pass. Instead,
the door was fixed in place (‘blocked’) so that the ducks could put
their head through the door, see the nest and attempt to push, but
could not pass through. Each bird completed the experiment by ei-
ther failing a weight level or reaching the final night of the experi-
mental schedule.
Each BDU was video recorded every night using a CCTV camera
surveillance system (Techview QV3034; Jaycar, Perth, Australia).
Behavioural data were collected from the video footage using the
software Interact (Interact, version 14.0; Mangold International,
Arnstorf, Germany). The observation period began when the duck
entered the BDU and ended 1 h after the duck passed the door (for
0–160% workloads) or after the last interaction with the door (for
nights when the door was blocked). The behaviours of interest are
outlined in Table 1. The location of each bird (nest area or holding
area) and any eggs laid (nest box, ground nest in nest area, floor of
nest area, ground nest in holding area, floor of holding area) were re-
corded each morning.
Correlation between duck behaviour and stress-induced hyperthermia
From the videofootage, the time at which a duck passed through the
weighted door (for 0–160% workloads) was identified. The T
c
data
within a time window of ±15 min relative to the time the door was
passed were analysed to detect SIH. For the nights when the door was
blocked, it was intended to correlate the time of first attempt to push
through the door with SIH events on all four testing nights. However,
the ducks made no attempts to pass the door after the first night, so
Fig. 2. Photo sequence of a female Pekin duck using thebehavioural demand apparatus to access her nest.
Table 1
Quantified behavioural variables of Pekin ducks during a behavioural demand test to as-
sess their motivation to access a nest site.
Behavioural variable Description
Number of
interactions with
door
Duck looks through the door (head or beak passes
through door), or pecks on the door
Number of attempts
to pass door
Duck places head and neck through the door, shoulders
are engaged with the door and duck is seen to exert effort
and/or door is seen to move
Latency to pass
through door
Time taken between duck’sfirst interaction with the
push-door to when it passes through the door
Latency to nest entry Time taken between duck passing through the push-door
to first entering the nest site (where nest entry is
considered to be both feet placed within the nest box)
Frequency of nest
visits
Number of times duck enters the nest site
Time spent in nest Total percentage of time spent in the nest site during the
hour after passing through the push-door
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
4
the association between SIH and the first attempt at the blocked door on
night 1, and with the first look through the blocked door at the nest on
nights 2–4, was analysed.
For each 5-min point in the day, a smoothed temperature (T
s
)was
calculated by averaging the T
c
for 12 h on either side of each data
point (e.g. T
s
at 0900 h is the average of T
c
from 2100 h the previous
night until 2100 h of the same day). The SD of T
s
was also calculated.
Each original T
c
data point was deemed to represent a SIH when it
was 2.5 × SD higher than the smoothed average for that time.The dura-
tion of SIH was defined as the time during which consecutive T
c
data
points were above 2.5 × SD. The area under the curve (AUC) for a SIH
event was calculated by adding the temperature differentials (T
c
–T
s
)
for the duration of the SIH.
Concentration of corticosterone in egg albumen
A radioimmunoassay was used to measure the concentration of cor-
ticosterone in egg albumen, using a method based on that of Downing
and Bryden (2008). Eggs were collected daily; the albumen was sepa-
rated from the yolk and then frozen at −20 °C until analysis. Eggs col-
lected on a given morning were associated with events two nights
prior to collection, based on the 24 ± h ovulation –oviposition cycle
that is characteristic of poultry species (Johnson, 2000).
Extraction of corticosterone
Albumen samples were thawed and homogenised. A sub-sample of
0.5–0.7 g was placed in a borosilicate glass culture tube (12 × 75 mm;
Rowe Scientific, Perth, WA) along with 500 μl of distilled water and
vortexed for a few seconds. Diethyl ether (4 ml) was added and
vortexed for 10 min. The tubes were then kept at −20 °C for at least
14 h to freeze the aqueous phase. The solvent phase was then poured
into a new culture tube and the diethyl ether was evaporated under
gentle heating (40 °C) and constant air flow. The dried tubes were
then covered and stored at 4 °C until assay.
The CV for the extraction method was 6.5%. The extraction efficiency
was 72% and was calculated using five replicates of three individual
samples of albumen spiked with 24 000 CPM of
125
I corticosterone.
The samples were incubated with the iodinated hormone for 2 days at
4 °C before being processed using the method described below. The
amount of iodinated corticosterone was counted in the extract
reconstituted in 300 μl of phosphate-buffered saline (PBS).
Corticosterone assay
The concentration of corticosterone in the extract was measured with
the Immuchem double antibody corticosterone
125
IRIAkit(MPBiomed-
icals, Orangeburg, NY) using a modified procedure to improve sensitivity.
The assay proceeded as follows: day 1, dried extracts were reconstituted
in 300 μl of PBS (pH 7.5) and vortexed for 5 min. Steroid diluent (75 μl)
and the first antibody (100 μl diluted 4 in 5 in PBS) were added to the
samples and incubated overnight at 4 °C; day 2,
125
Itracer(100μl) was
added to each tube and incubated overnight at 4 °C; day 3, secondary an-
tibody mixture (300 μl) was added to each tube. The mixture consisted of
150 μl of kit precipitant plus 150 μl of donkey anti-rabbit IgG (G4004;
Immundiagnostik, Bensheim, Germany) solution (1/40 in PBS). Samples
were then incubated overnight at 4 °C; day 4, 0.5 ml of 10% PEG solution
was added, and the samples were centrifuged at 3000 G for 25 min. The
supernatant was decanted and the samples were left to dry overnight;
day 5, the radioactivity was counted using a Beckman Gamma counter.
The assay was validated using two QCs (quality control); the limit of de-
tection was 5.2 ng/ml and the coefficients of variation were 3.6% for
15.5 ng/ml and 6.5% for 90.1 ng/ml.
Statistical analysis
Three of the 12 birds were excluded from the experiment: one due
to lameness and two because they established nests in the holding
area rather than the nest area. Of the remaining nine birds, five
completed all workloads (0–160% BW) and attempted to pass the
blocked door. Only data from these five birds were included for statisti-
cal analysis. All data analysis was conducted using R statistical software
(R Development Core Team, 2017).
Assessing the motivation to access the nest
The behavioural data were explored graphically to identify potential
relationships between behaviour and increasing workload or blocking
the door. Where required, the data were log-transformed to normalise
the distribution, before ANOVA was performed. Data for the latency to
enter the nest could not be easily normalised and were not subject to
further analysis. Percentage data underwent arcsine transformation
prior to analysis. In all the analyses described below, the variable Bird
was included as an error term in all final models to account for the
non-independence of data.
For each behaviour, single explanatory variables (door workload,
pen, night) were tested for association with the outcome variable across
all workloads (number of interactions, duration of interactions, number
of attempts, duration of attempts, latency to pass the door, number of
nest visits, percentage of time spent in the nest). Interactions between
variables were also assessed and retained in the final model if signifi-
cant. Where there was a main effect of door workload, Bonferroni
post-hoc testing was conducted to identify pairwise differences.
For the individual nights of the blocked door, the number of interac-
tions with the door and number of attempts to open the door could not
be normalised and were not suitable for non-parametric analysis. A de-
scriptive analysis was thus undertaken. A baseline occurrence (mean +
95% CI) of each behaviour was calculated for each bird using the data
from all four nights of the 0% workload. The number of attempts or in-
teractions that each bird made on each night of the blocked door was
then compared against the bird’s baseline for that behaviour. Values
that lay either above or below the 95% CI were identified, and a contin-
gency table of the number of birds that lay below, within or above their
CI for each behaviour was created.
Stress-induced hyperthermia
A linear mixed effects model was used to analyse differences in the
AUC of hyperthermia between workloads of 0–160% and the blocked
door. A second linear mixed effects model was used to analyse the
AUC of hyperthermia for the individual nights of the blocked door. The
outcome variable (AUC) was regressed against the explanatory vari-
ables of workload, pen and night for comparison of all workloads, and
against night and pen for the blocked door alone. For the model across
all workloads, ‘door’was the only significant explanatory variable in-
cluded in the final model. ‘Night’was the only variable included in the
final model of the blocked door. For both models, ‘bird’was included
as a random effect term, to account for non-independence of the data.
Interactions between explanatory variables were also explored, but
none were significant.Data were log-transformed to normalise distribu-
tions prior to analysis.
Egg corticosterone concentrations
Analysis of variance was used to investigate if door workload, ornot
being able to access the nest (blocked door), affected the concentration
of corticosterone in the egg albumen. Corticosterone concentration was
individually regressed against the explanatory variables; door work-
load, pen, and night, and explanatory variables were included in the
final model if significant. Interactions between explanatory variables
were also tested.
Results
Oviposition site
Before the doors were blocked, the ducks that completed the 160%
workload (n = 5) laid their eggs in the same location on 96.2% of the
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
5
nights (range 94–97.9%, Supplementary Table S1). The ducks that
stopped pushing through the door earlier (n = 4) laid 78% (range 50–
96.7%) in the same location before ceasing to push the door. There
were 31 occurrences when birds pushed through the door but did not
lay an egg (Supplementary Table 1).
Motivation to access the nest
Five birds completed all workloads and attempted to pass the
blocked door. Of the remaining four birds, two established nests in the
nesting area but did not push through the door once it was closed,
and two stopped pushing through the door after completing 80 and
140% BW workloads.
There was a main effect of workload on the number of door interac-
tions; interactionsincreased as workload increased (F
(9, 175)
= 6.60, P<
0.001, Fig. 3). The five ducks had more interactions at 160% than at the
0–100% workloads, and at 120% than at 40 and 60%. There was no differ-
ence in the number of interactions with the blocked door from that at
any other workload.
Across all workloads,there was a main effect of both door andday on
the number of attempts made at the door, and an interaction between
door and day. The five ducks made more attempts at 160% and at the
blocked door than at 0–80% workloads (F
(9, 145)
= 26.97, P<0.001,
Fig. 4). There were more attempts on day 1 of workloads than on days
2–4.
On the first night that the door was blocked, all five birds attempted
to push the door more often than their individual baseline values
(Table 2). No attempts were made to push through the door on nights
2–4 of the blocked door (Table 2). Four of the five birds also interacted
with the door more frequently than their individual baseline on night
1 of the blocked door (Table 2).
Across all workloads,the mean latency to enter thenest after passing
the door was 0.05 ± 0.02 s (mean ± SD). There was no effect of work-
load on the time that the five ducks spent in their nest site. On average,
the ducks spent 94% (SD ± 16%) of the hour after they passed the door
in their nest site. There was no effect of workload on the frequency of
nest visits (mean = 1.16 ± SD 0.62).
Effect of increasing workloads, and inability to access the nest, on corebody
temperature
The AUC of T
c
was larger at workloads of 80% (P< 0.001), 120% (P<
0.01), 140% (P< 0.001), 160% (P< 0.001), and when the door was
blocked (P< 0.001), compared with 0% (Fig. 5). The AUC of hyperther-
mia did not differ for the five ducks between nights 2–4 of the blocked
door, compared with the first night that the door was blocked.
Concentration of corticosterone in egg albumen
The concentration of corticosterone in the egg albumen of the five
ducks was not affected by workload, or the inability to access the nest
site (F
(10, 177)
=1.53,P= 0.13). The mean corticosterone concentration
across all nights was 15.63 (SD = 6.59) ng/ml.
Discussion
The aim of the study was to assess the motivation of laying Pekin
ducks to access their nest and to determine whether they exhibited
frustration in response to a rising cost of nest access or the ultimate in-
ability to access the nest. The hypothesis that ducks are highly moti-
vated to access their nest site was supported by the weight the ducks
were willing to push, and the increased number of interactions and at-
tempts that occurred, both as workload increased and when the door
was blocked. However, the maximum workload that ducks were moti-
vated to push through was not determined. The analysis of T
c
supported
the hypothesis that a stress response, mostlikely due to frustration, was
observed in ducks when they were unable to access their nest, indicat-
ing that access to the nest has value to the ducks. The stress response
was not reflected in changes to concentrations of corticosterone in egg
albumen.
The strong motivation that laying ducks had to access their nest site
was indicated firstly by the willingness of five out of nine birds to push
through a door weighted up to 160% of their own BW. The need to alter
the original study design and impose the blocked door, rather than to
risk bird safety by allowing them to continue working at higher work-
loads, also indicated the motivation level for the nest site. In fact, all
seven birds that established nests in the nesting area and used the
push-door showed willingness to exert physical effort to access their
nest, with the first bird ceasing to push after completing 80% of BW,
and the second bird stopping after completion of stopping 140%. Thus,
all birds that used thepush-door were motivated toaccess their nest, al-
though it is evident that some variation in the levels of motivation
existed between birds.
There are several possible reasons why the two remaining birds
(Birds 6 and 11) that established nests in the nest area did not push
through the door to enter the nest area during the behavioural demand
Fig. 3. Numberof interactions that Pekinducks had with a push-door before passing through it to accesstheir nest over increasing workloads. Numbers presented are the mean (± SEM)
across all fou r nights of eac h workload.
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
6
test: the nest site may not have been perceived as sufficiently motivat-
ing for these birds to pass the door; there may have been a lack of asso-
ciation between operant taskand reward, or an extinction of the learned
operant push-door task had occurred. Low motivation may explain the
behaviour of Bird 6, which showed approximate equivalency for laying
eggs in the nest site and on the ground. Reasons for low motivation for
the nest site could include genetic variation in the expression of nesting
behaviour, or that this individual viewed the available egg-laying sites
as equally suitable or unsuitable. However, Bird 11 showed high nest
conservatism before the door was closed (79% of eggs laid in the same
location), which suggests motivation for the nest. Alternatively, the
birds did not associate the known operant task of pushing a door with
the reward of nest site access, as the differing reward would have
been a new, unlearned paradigm. However, the push-door task does
not appear to ask for an unassociated operant response to the nest re-
ward, as movement towards a desired resource would be part of the
duck’s natural response pattern if sufficiently motivated. Such mis-
match of task to reward has been highlighted in previous studies,
where hens appeared to show low motivation for dust bathing when
Fig. 4. Numberof attempts made by Pekin ducks at a push-doorbefore passing through it to access their nest over increasing workloads. Numbers presented are the mean (±SEM) for all
ducks across all four nights of each workload.
Table 2
Number of interactions and attempts ducks (n = 5) made with a push-doorwhen it was
blocked, thus preventing them from accessing their nest, over four consecutive nights.
Shaded values indicate that the ducks interacted or attempted to push the door more or
less often than their individual baseline value (mean + 95% confidence interval across
all four nights of 0% workload).
Night
Bird Baseline mean (95%CI) 1 2 3 4
Interactions 1 12 (−1–25) 30 4 0 0
44(−6 -13) 50 3 2 5
810(−24–14) 21 2 1 3
99(−19–38) 29 5 4 0
12 5 (0–10) 32 7 3 8
Attempts 1 3 (0–6) 43 0 0 0
41(−1–2) 84 0 0 0
81(0–3) 14 0 0 0
91(1–1) 9 000
12 1 (0–3) 43 0 0 0
Fig. 5. Mean(± SEM) area under thecurve (AUC) of hyperthermia in Pekinducks (n = 5) when theyworked increasingly harderto access a nest site(0–160%, four consecutive nightsper
workload) or when they were unable to access the nest (blocked, four consecutive nights following 160%).
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
7
required to peck keys for access but exhibited high motivation when the
same reward was associated with the breaking of a photobeam demon-
strated much higher motivation (Dawkins and Beardsley, 1986). A more
likely explanation of the ducks’failure to pass the door when first closed
is an extinction of the learned operant push-door task, as the task was
not reinforced between training during the BDU development protocol
(Barrett and Blache, 2019) and the beginning of the behavioural de-
mand test in this study. It is thus recommended for future work that
the operant task is reinforced atregular intervals between different ex-
perimental phases or protocols.
That birds remained interested and motivated to gain nest access
was also evidenced by the greater number of door interactions and at-
tempts at the door at higher workloads. There were more interactions
with the door at 160% compared with lower levels (0–100% BW), and
more attempts to push the door at 160%, and the blocked door, com-
pared with lower levels (0–80% BW). These findings indicate that
ducks were persistent in trying to pass through the door. The increased
number of attempts to push through the blocked door could be behav-
ioural activation, which occurs when animals are frustrated by the
absence of, or inability to procure, a previously accessible reward,
resulting in an ‘up-regulation’of their motivation to perform the previ-
ously successful behaviour (Latham and Mason, 2010;Papini et al.,
2019). Chickens are also persistent in trying to access a nest, with
hens making 75–150 attempts to push through a secured door to a
nest (Follensbee, 1992).
The motivation of ducks for the nest site was further illustrated by
the short latency to enter the nest once they passed through the door,
and the high percentage of time spent in the nest in the hour after
they passed throughthe door. Previouswork with hens has alsodemon-
strated that the latency to enter the nest and the proportion of time
spent in the nest were relevant indicators of the motivation that hens
have for the nest resource (Cooper and Appleby, 2003). It is interesting
to note that several individuals pushed through the door to access the
nest even when no egg was laid, suggesting that factors other than
egg-laying can motivate a duck to seek a nest. Other possible explana-
tions could include the desire for an area thatis comfortable and/or per-
ceived to be safe for them to rest in.
Analysis of the AUC showed that the first occurrence of a significant
hyperthermia occurred at 80% workload. The AUC for 100% was not dif-
ferent to 0%, but a significant difference was seen at 120–160% work-
loads. The interesting decrease at 100% suggests to us that at 80%
workload at least a component of the hyperthermia was not due to
the heat production associated with the physical effort needed to
open the weighted door. It is possible that the birds recognised that in-
creased physical effort was needed to access the nest, leading to a de-
gree of frustration and the resulting hyperthermia associated with
that psychogenic stress.
While we cannot distinguish between the heatof exerciseeffort and
SIH at 120, 140 and 160% workloads, and the first night that the door
was blocked, the hyperthermia that developed when the ducks looked
through the door towards their nest on the second to the fourth nights
when the door was blocked suggests that theducks exhibited a psycho-
genic stress, because there was no exercise effort. Physical exertion
could not have played any role in the elevated T
c
on these nights, as
no bird made any attempt to push through the door. Evidence for SIH
occurring in birds due to psychological stress has been described only
in pigeons (Bittencourt et al., 2015), where hyperthermia was seen
after birds were exposed to the appearance of a researcher, without
any other intervention. Psychological SIH has been identified in mam-
mals. Sheep in an open-field test exhibited SIH as a component of
their fear response (Pedernera-Romano et al., 2010), while humans ex-
hibited SIH in response to the Trier Social Stress Test (Vinkers et al.,
2013). The SIH that we observed in the ducks when they were not
able to access a nest site is likely due to the experience of frustration
and leads to the conclusion that the nest is a valued resource. However,
we cannot compare the magnitude of stress experienced by the ducks
when they were unable to access their nest, relative to other stressors
that also incite SIH. This is due in part to other studies typically use
the absolute change from baseline temperature as their indicator of
SIH (e.g. Cabanac and Guillemette, 2001;Gray et al., 2008), and the gen-
eral lack of studies that have used SIH as a marker of stress in ducks. Of
note, most studies create a stress response by performing a physical in-
tervention on the birds (e.g. handling or restraint), as opposed to a
purely psychological stressor. Comparison with SIH in other bird species
has similar issues. As far as it is possible to establish, this is the first re-
port of SIH in response to frustration due to the thwarting of a highly
regulated behaviour such as nesting.
Taken together, the results of the behavioural observations, laying
locations and the changes in T
c
strongly suggest that the ducks were
motivated to access their nest site and experienced frustration when
they were unable to access it. Frustration in animals is suggested by
their behavioural responses to situations where access to a reward is
prevented. The behaviours are first directed to the source of the reward
blockage, before increases in indicators of agitation, displacement be-
haviours, and in some cases, aggression, are seen (Haskell et al., 2004).
The inability to express normal nesting behaviour is considered the
greatest cause of frustration in conventional-caged hens (Duncan,
2001). Multiple studies have identified changes that are thought to be
associated with frustrated nesting behaviour. When inadequate nests
were provided, or nest access was prevented, hens exhibited more pac-
ingactivity(Cooper and Appleby, 1996; Wood-GushWood Gush, 1972;
Yue and Duncan, 2003). Feeding and preening (often considered dis-
placementbehaviours in poultry) occurred more often in cageswithout
nests than in modified cages and alternative housing systems (Meijsser
and Hughes, 1989;Sherwin and Nicol, 1993).
In accord with the interpretation of changes in behaviours, behav-
ioural demand tests have demonstrated the high value that hens place
on nest access. The work rate of hens for a nest 40 min before egg-
laying was similar to the work rate for food after 4 h confinement, indi-
cating that the two resources (nest and food) were of equivalent value
(Cooper and Appleby, 2003). Based on the high cost that hens were pre-
pared to pay for nest access in a behavioural demand test, Cooper and
Appleby (1996) concluded that hens without a defined nest site may
experience frustration during their pre-laying period. Thus, the amount
of work that ducks were willing to exert and the occurrence of SIH even
when therewas no physical effort made to get to the nest indicates that,
like chickens, laying Pekin ducks place high value on the nest resource.
They have a behavioural need to access the nest and could become frus-
trated if they cannot access a nest site.
Failure to pass through the door on the first night that the door was
blocked significantly impacted the ducks’willingness to attempt to pass
on subsequent nights. A similar finding was reported in chickens: birds
that failed to pass a push-door during training made fewer attempts and
usually gave up without passing the door when re-tested at a lower
weight (Olsson et al., 2002). There are several possible explanations
for the lack of attempts on nights 2–4 of the blocked door. One pos-
sibility is that appetitive extinction of the operant task occurred, due
to frustration at previously being unable to access the nest. This be-
havioural shift is seen when an animal has learned to anticipate the
omission of a reward or the frustration resulting from reward omis-
sion, and so they minimise performance of their previously success-
ful response (Papini et al., 2019). Another explanation is that the lack
of attempts made on nights 2–4 of the blocked door is a form of
learned helplessness. The classic model of learned helplessness
demonstrates that animals unable to control the endpoint of a
stressor show increased passivity and anxiety when subjected to
that stressor in a different environment (Maier and Seligman,
2016). However, it has also been shown that, in some scenarios, an-
imals that have previously had control over the stressor endpoint
show similar behavioural patterns of passivity/anxiety when that
control is removed (Christianson et al., 2008). A final explanation
may be that the ducks recognised the visual differences in the
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
8
blockeddooranddidnotattempttopassonnights2–4 because they
had previously learned that they would be unsuccessful.
The concentration of corticosterone in egg albumen indicated that
this measure was not a sensitive indicator of any acute stress that re-
sulted from either increased physical effort or frustration from being de-
nied nest access. A previous study with chickens also failed to find
changes in albumen corticosterone after handling (Cook et al., 2009).
It was posited that the lack of change in corticosterone in eggs may be
due to a mismatch in the timing of stressor presentation and albumen
deposition. An acute changein corticosteronemay not be detected in al-
bumen unless there was active albumen secretion at or soon after the
increase in serum concentrations due to stress. Birds typically ovulate
30–45 min after oviposition, and the new follicle resides in the magnum
(site of albumen deposition) for 2–3h(Johnson, 2000). The stressor in
the current study (physical exertion, or frustration at being denied
nest access) should, in theory, have occurred within a few hours of albu-
men deposition around the newly ovulated follicle. It is possible then
that any increases of serum corticosterone were not of large enough in
magnitude or duration to be picked up in albumen concentrations.
The decision to block access to the nest after the 160% workload cre-
ated a limitation of the study, but was the appropriate ethical decision.
The consequences of applying an artificial maximum cost are that the
true maximum cost cannot be known from this study, and because of
that the utility of nest access as a future comparator for other resources
becomes limited, unless motivation for those other resources is lower
than the artificial maximum shown here. Although we could not deter-
mine the maximum cost that the ducks would incur to access their nest,
it is clear that motivated birds were willing to exert significant physical
effort for the reward of their nest site. Future designs of the push-door
need to better address how this method can be refined to allow en-
hanced discrimination of differing levels of motivation without risking
bird safety. One possible solution may be to combine the technique of
an electromagnetic cell-controlled door (e.g. Olsson et al., 2002) with
a constant resistance that works on the door after it opens. This would
prevent the door from suddenly giving way once the prescribed force
had been reached, and allow the bird to push through in a controlled
manner. Another method is the spring-loaded door that provides con-
stant resistance during pushing (Kruschwitz et al., 2008).
Retaining visual access to the nest may be considered a limitation, as
it is possible that such cues can influence the motivation for resources
(Warburton and Mason, 2003). The design of the BDUs that were used
in this study limited the opportunity to remove all visual nest cues, as
even if it were possible to reduce nest visibility between the door and
nest in individual units, there remained the possibility of birds being
able to see nests in neighbouring units, due to the steel mesh construc-
tion of one of the outer walls. Future designs of individual units could
take the issue of cue-relianceinto consideration and incorporateoptions
to allow variation in cue exposure.
To further substantiate the presence of frustration when the door
was blocked, quantification of behavioural changes beyond interactions
and attempts to pass the door would have been ideal. Unfortunately,
due to the physical constraints of the housing arrangements, it was
not possible to film the entire area of the BDU. Our primary interest
from the video analysis was the effort that the ducks were willing to
exert on the door, and how they interacted with the nest; thus, the cam-
eras were directed towards the door. However, additional information
from other behavioural indicators of frustration, such as the occurrence
of pacing or displacement activities, may have allowed deeper interpre-
tation of the physiological data. Future studies of a similar nature should
ideally include such indicators.
The final small sample size of five birds in the statistical analysis is an
additional limitation, due to the risk of the analysis having low power
(or a higher chance of Type II error). Consequently, the study findings
should be regarded as preliminary, with further investigations needed
to have heightened confidence that the same changes are reflected in
commercial duck populations.
Conclusion
The study indicated that most laying Pekin ducks were highly
motivated to seek a nest. The nest is a significant resource, and the
inability to use it results in SIH that potentially illustrated the level
of frustration. The concentration of corticosterone in egg albumen
was not a useful indicator of stress. The push-door method for
assessing motivation in ducks was useful, but requires design refine-
ments to better discriminate the levels of motivation at higher work-
loads without risking bird safety. The findings suggest that the floor-
laying behaviour that is seen in commercial flocks is unlikely due
solely to an absence of nest-seeking behaviour. Further research is
therefore needed to determine what other factors might contribute
to floor-laying if low motivation for nests is not one of them. Other
work suggests that competition for nests may contribute (Barrett
et al., 2019), and that floor-eggs may at least in part be explained
by highly motivated birds trying to gain access to nest boxes, but ul-
timately being unable to lay in them.
Supplementary materials
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.animal.2020.100067.
Ethics approval
The study was approved by the University of Western Australia’sAn-
imal Ethics Committee (RA/1/300/1338).
Data and model availability statement
None of the data were deposited in an official repository. Data may
be provided upon request.
Author ORCIDs
LB: 0000–0003–1298-5380; SM: 0000–0002–5878-2266; DB:
0000–0003–3476-3068.
Author contributions
Lorelle Barrett: conceptualisation, methodology, formal analysis, in-
vestigation, data curation, writing –original draft, writing –review and
editing,visualisation, project administration, funding acquisition. Shane
Maloney: conceptualisation, methodology, formal analysis, resources,
writing –review and editing, supervision. Dominique Blache: concep-
tualisation, methodology, validation, formal analysis, resources, writing
–review and editing, supervision, funding acquisition.
Declaration of interest
None.
Acknowledgements
The authors thank Ray Scott of Seer Tech for the construction of the
behavioural demand units.
Financial support statement
This research was funded by a Poultry CRC post-graduate research
scholarship (LB), and conducted within the Poultry CRC, established
and supported under the Australian Government’s Cooperative Re-
search Centres Program.
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
9
References
Appleby, M.C., Mench, J.A., Hughes, B.O., 2004.Poultry behaviour and welfare. CABI, Wal-
lingford, UK.
Barrett, L.A., Blache,D., 2019. Development of a behavioural demand method for use with
Pekin ducks. Applied Animal Behaviour Science 214, 42–49.
Barrett, L.A., Malecki, I.A., Blache, D., 2019. Differences in pre-laying behavior between
floor-laying and nest-laying Pekin ducks. Animals 9, 40.
Bittencourt, M.D.A., Melleu, F.F., Marino-Neto, J., 2015. Stress-induced core temperature
changes in pigeons (Columba livia). Physiology and Behavior 139, 449–458.
Blache, D., Terlouw, C., Maloney, S.K., 2017. Physiology. In: Appleby, M.C., Olsson, I.A.S.,
Galindo, F. (Eds.), Animal welfare, 3rd edition. CABI, Wallingford, UK, pp. 181–212.
Cabanac, A.J., Guillemette, M., 2001. Temperature and heart rate as stress indicators of
handled common eider. Physiology and Behavior 74, 475–479.
Christianson, J.P.,Paul, E.D., Irani, M., Thompson, B.M.,Kubala, K.H., Yirmiya, R., Watkins,L.
R., Maier, S.F., 2008. The role of prior stressor controllability and the dorsal raphé nu-
cleus in sucrose preference and social exploration. Behavioural Brain Research 193,
87–93.
Cook, N.J., Renema, R., Wilkinson, C., Schaefer, A.L., 2009. Comparisons among serum, egg
albumin and yolkconcentrations of corticosterone as biomarkers of basal and stimu-
lated adrenocortical activity of laying hens. British Poultry Science 50, 620–633.
Cooper, J.J., Appleby, M.C., 1995. Nesting behaviour of hens: effects of experience on mo-
tivation. Applied Animal Behaviour Science 42, 283–295.
Cooper, J.J., Appleby, M.C., 1996. Individual variation in prelaying behaviour and the inci-
dence of floor eggs. British Poultry Science 37, 245–253.
Cooper, J.J., Appleby, M.C., 2003. The value of environmental resources to domestic hens:
A comparisonof the work-rate for foodand for nests as function of time. Animal Wel-
fare 12, 39–52.
Cronin,G.M., Barnett, J.L.,Storey, T.H., Thomson,P.C., Hemsworth,P.H., 2012. The relation-
ship betweenpre-laying activity and corticosterone concentrations, and the interpre-
tation for laying hen welfare. Proceedings of the 23rd annual australian poultry
science symposium, 19-22 February 2012, Sydney, Ne w South Wales, Australia,
pp. 168–171.
Dawkins, M.S., 1990. From an animal’s point of view: motivation, fitness, and animalwel-
fare. Behavioral and Brain Sciences 13, 1–61.
Dawkins, M.S., Beardsley, T., 1986. Reinforcing properties of access to litter in hens. Ap-
plied Animal Behaviour Science 15, 351–364.
Downing, J.A., Bryden, W.L., 2008. Determination of corticosterone concentrations in egg
albumen: a non-invasive indicator of stress in laying hens. Physiology and Behavior
95, 381–387.
Duncan, I.J.H., 2001. The pros and cons of cages. World's Poultry Science Journal 57,
386–390.
Follensbee, M., 1992. Quantifying the nesting motivation of domestic hens. MSc, Univer-
sity of Guelph, Guelph, Ontario, Canada.
Gray, D.A.,Maloney, S.K., Kamerman, P.R., 2008. Restraint increases afebrile body temper-
ature but attenuates fever in Pekin ducks (Anas platyrhynchos). American Journal of
Physiology. Regulatory, Integrative and Comparative Physiology 294, R1666–R1671.
Haskell, M.J., Coerse, N.C.A., Taylor, P.A.E., McCorquodale, C., 2004. The effect of previous
experience ove r control of access to food and light on the level of frustratio n-
induced aggression in the domestic hen. Ethology 110, 501–513.
Johnson, A.L., 2000. Reproduction in the Female. In: Whittow, G.C. (Ed.), Sturkie’savian
physiology, 5th edition. Academic Press, San Diego, California, USA, pp. 569–596.
Kataoka, N., Shima, Y., Nakajima, K., Makamura, K., 2020. A central master driver of psy-
chosocial stress responses in the rat. Science 367, 1105–1112.
Kruschwitz, A., Zupan, M., Buchwalder, T., Huber-Eicher, B., 2008. Nest preference of lay-
ing hens (Gallus gallus domesticus) and their motivation to exert themselves to gain
nest access. Applied Animal Behaviour Science 112, 321–330.
Latham, N.,Mason, G., 2010. Frustration and perseveration in stereotypiccaptive animals:
is a taste of enrichment worse than none at all? Behavioural Brain Research 211,
96–104.
Maier, S.F., Seligman, M.E.P., 2016. Learned helplessness at fifty: insights from neurosci-
ence. Psychological Review 123, 349–367.
Makagon, M.M., Mench, J.A., 2011. Floor laying by Pekin ducks: effects of nest box ratio
and design. Poultry Science 90, 1179–1184.
Makagon, M.M., Tucker, C.B., Mench, J.A., 2011. Factors affecting nest choice by Pekin
ducks. Applied Animal Behaviour Science 129, 121–128.
Mason, G., Bateson, M., 2009. Motivation and the organisation of behaviour. In: Jensen, P.
(Ed.), The ethology of domestic animals, 2nd edition. CABI, Wallingford, UK.
Mason, G., Burn, C., 2011. Behavioural restriction. In: Appleby, M.C., Mench, J.A., Olsson, I.
A.S., Hughes, B.O. (Eds.), Animal welfare, 2nd edition. CABI, Wallingford, UK,
pp. 98–119.
Mason, G., Garner, J., McFarland, D., 1998. Assessing animal priorities: future directions.
Animal Behaviour 55, 1082–1083.
Meijsser, F.M., Hughes, B.O., 1989. Comparative analysis of pre-laying behaviour in bat-
tery cages and in three alternative systems. British Poultry Science 30, 747–760.
Olsson, I.A.S., Keeling, L.J., McAdie, T.M., 2002. The push-door for measuring motivation in
hens: an adaptation and a critical discussionof the method. Animal Welfare 11, 1–10.
Papini, M.R., Penagos-Corzo, J.C., Pérez-Acosta, A.M., 2019. Avian emotions: comparative
perspectives on fear and frustration. Frontiers in Psychology 9, 2707.
Pedernera-Romano, C., Ruiz de la Torre, J.L., Badiella, L., Manteca, X., 2010. Effect of per-
phenazine enanthate on open-field test behaviour and stress-induced hyperthermia
in domestic sheep. Pharmacology Biochemistry and Behavior 94, 329–332.
R Development Core Team, 2017. R: A language and environment for statistical comput-
ing. R Foundation for Statistical Computing, Vienna, Austria.
Royo, F., Mayo, S., Carlsson, H.-E., Hau,J., 2008. Egg corticosterone: a noninvasive measure
of stress in egg-laying birds. Journal of Avian Medicine and Surgery 22, 310–314.
Sherwin, C.M., Nicol, C.J., 1993. A descriptive account of the pre-laying behaviour of hens
housed individually in modified cages with nests. Applied Animal Behaviour Science
38, 49–60.
Vinkers, C.H., Penning, R., Hellhammer, J., Verster, J.C., Klaessens, J.H.G.M., Olivier, B.,
Kalkman, C.J., 2013. The effect of stress on core and peripheral body temperature in
humans. Stress 16, 520–530.
Warburton, H., Mason, G., 2003. Is out of sight out of mind? The effects of resource cues
on motivation in mink, Mustela vison.AnimalBehaviour65,755–762.
Wood Gush, D.G.M., 1972. Strain differences in response to sub-optimal stimuli in the
fowl. Animal Behaviour 20, 72–76.
Yue, S., Duncan, I.J.H., 2003. Frustrated nesting behaviour: relation to extra-cuticular shell
calcium and bone st rength in white leghorn hens. British Poultry Science 44,
175–181.
L.A. Barrett, S.K. Maloney and D. Blache Animal xxx (xxxx) xxx
10
