You are on page 1of 9

Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Applied Animal Behaviour Science


journal homepage: www.elsevier.com/locate/applanim

Sow and piglet behaviour in group lactation housing from 7 or 14 days post-
partum
⁎,1
Megan Verdona, , Rebecca. S. Morrisonb, Jean-Loup Raulta,2
a
Animal Welfare Science Centre, Faculty of Veterinary and Agricultural Sciences, University of Melbourne, Victoria, 3010 Australia
b
Rivalea Australia, Corowa, New South Wales, 2646 Australia

A R T I C LE I N FO A B S T R A C T

Keywords: Under natural conditions, sows and piglets abandon the nest between 7 and 14 days postpartum and re-join the
Development group. If given the choice, commercially housed sows begin to nurse piglets communally around this time. The
Group lactation best age at which to group sows and their litters in two-stage group lactation systems has not been determined. A
Nosing total of 112 multiparous sows and their litters were allocated to three lactation housing treatments over four
Nursing
time replicates. All sows farrowed in standard farrowing crates. Group lactation (GL) sows were transferred to
Suckling
Social
group pens (one pen of 5 sows at 8.4 m2/sow and one pen of 7 sows at 8.1 m2/sow, per GL treatment and
replicate) with their litters at 7 (GL7; n = 48 sow and litter units) or 14 (GL14; n = 48 sow and litter units) days
postpartum until weaning. Farrowing crate (FC; n = 16 sow and litter units) sows and their litters remained
within FC in single litters until weaning. In each replicate, two focal sows per GL pen (one high and one low
parity) were video recorded for 1 h post-mixing. The GL focal sows, along with 4 FC sows (two high and two low
parity) and 2 focal piglets per focal sow (one average sized male and one average sized female), were also
recorded from 0700 to 1700 on the day after mixing and 2 days before weaning. Data gathered continuously
from video records included sow aggressive and nursing behaviours while sow and piglet time-budgets were
observed using point sampling at 5-min intervals. Social affiliative nosing was more frequent between GL14 than
GL7 sows (P = 0.01), but aggression did not differ (P > 0.05). There were no other effects of litter age at mixing
on sow or piglet behaviour (P > 0.05). Compared to FC sows, GL sows were more active (P < 0.001), ex-
plorative (P < 0.001) and interactive with piglets (P = 0.03). However, GL sows had fewer successful nursings
(P = 0.002) and terminated more nursings (P = 0.001) than FC sows. The time GL piglets spent suckling
(P < 0.001) and nosing (P = 0.001) sows other than their dam increased over time. Compared to FC piglets, GL
piglets spent more time in the creep (P = 0.05), active when outside of the creep (P = 0.04) and investigating
the pen (P = 0.02), and less time manipulating other piglets (P = 0.005). There do not appear to be behavioural
differences in sows and piglets between mixing into GL at 7 or 14 days postpartum. While GL allows for the
expression of maternal and exploratory behaviours, these benefits come at the cost of disrupted nursing beha-
viour and cross-suckling.

1. Introduction groups during lactation.


Group lactation systems are among the highest scoring on the Baxter
Growing societal concern for the welfare of animals in indoor in- sow/pig welfare index, which assesses how well components of alter-
tensive production systems is challenging the long-standing status quo native farrowing and lactation systems meet the animals’ biological
of confining sows to farrowing crates during parturition and lactation. needs, but further refinements to group systems are required to achieve
In particular, community objections relate to the barren environment acceptable and consistent performance (Baxter et al., 2018). For ex-
associated with farrowing crates, as well as restrictions in social contact ample, high piglet mortality and disrupted nursing behaviour challenge
and opportunities to exercise (Hemsworth, 2018). An alternative that the productivity and ethical acceptability of group lactation systems
addresses these concerns is the housing of sows and their litters in (reviewed by van Nieuwamerongen et al., 2014). Transferring sows and


Corresponding author at: Tasmanian Institute of Agriculture, Faculty of Science, Engineering and Technology, University of Tasmania, Tasmania, 7320 Australia.
E-mail address: megan.verdon@utas.edu.au (M. Verdon).
1
Present address: Tasmanian Institute of Agriculture, Faculty of Science, Engineering and Technology, University of Tasmania, Tasmania, 7320 Australia.
2
Present address: Institute of Animal Husbandry and Animal Welfare, University of Veterinary Medicine, Vienna, 1210 Austria.

https://doi.org/10.1016/j.applanim.2019.03.001
Received 4 October 2018; Received in revised form 15 February 2019; Accepted 4 March 2019
0168-1591/ © 2019 Published by Elsevier B.V.

Please cite this article as: Megan Verdon, Rebecca. S. Morrison and Jean-Loup Rault, Applied Animal Behaviour Science,
https://doi.org/10.1016/j.applanim.2019.03.001
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

their litters to group lactation after an initial period of farrowing crate farrowing crate (FC) housing until weaning, (2) group lactation housing
housing (i.e., ‘two-stage group lactation’ systems) is one available from 7 days postpartum (GL7) until weaning, and (3) group lactation
strategy that could protect piglets when they are at their most vulner- housing from 14 days postpartum (GL14) until weaning (see Section 2.2
able, thus enhancing the ethical and economic viability of group lac- for details). Housing treatments were balanced for sow parity (mean
tation systems for large-scale commercial production. ± s.d., 3.7 ± 2.0), sow weight (mean ± s.d., 275 ± 42.5 kg) and
Under natural conditions, sows and piglets abandon the nest be- litter size (mean ± s.d., 10.5 ± 1.7 piglets/sow), and there were no
tween 7 and 14 days postpartum to reintegrate into the social group treatment differences at selection in piglet weight (per litter) (mean
(Jensen, 1988). This behaviour remains relevant to commercially ± SEM, 2.64 ± 0.16), male to female sex ratio (mean ± SEM,
housed pigs; sows that farrow in pens connected by an outer communal 1.3 ± 0.13), sow back-fat in mm at the P2 site (mean ± SEM,
area begin to nurse piglets outside of the pens between 7 and 14 days 28.8 ± 0.58) or within treatment variation in sow parity (mean ±
post-partum (Arey and Sancha, 1996; Schrey et al., 2018). However, the SEM coefficient of variation, 52 ± 7.6%), sow weight (mean ± SEM
mixing of unfamiliar sows and piglets occurs abruptly in two-stage coefficient of variation, 15 ± 1.8%) and sow back-fat (mean ± SEM
group lactation systems, typically around 14 days postpartum coefficient of variation, 21 ± 1.7%).
(Wattanakul et al., 1997a; Dybjaer et al., 2001; Dybkjær et al., 2003; Pregnant sows were moved from group gestation housing to one of
Verdon et al., 2016). The time sows spend in confinement can be re- 64 FC’s located in the same room of the farrowing and lactation unit at
duced by transferring animals to group lactation at 7 rather than 14 7 days before expected parturition. The bottom bars of the FC (crate
days postpartum. Whether mixing at an earlier stage of lactation offers 2.3 × 0.6 m, total area 2.3 × 1.7 m) operated on a hydraulic ram so
other advantages or challenges, particularly in terms of the ability of that the sides swing in when the sow stands and slowly slide out when
sows and/or piglets to cope with the sudden change in the physical and the sow lies down. Farrowing crates contained a creep area that was
social environment, requires examination. heated using a mat below (1.1 × 0.40 m) and a lamp overhead. Piglets
There are few published research on two-stage group lactation were tail docked and received an iron injection within 24 h of birth, as
systems, and behavioural data from these systems rarely extend beyond per standard commercial practices.
observations of nursing and/or suckling interactions (Wattanakul et al., The four group lactation (GL) pens were located in a single room of
1997a; Dybjaer et al., 2001; Thomsson et al., 2015, 2016). For this the same shed that contained the FC’s. Two of the GL pens had the
experiment we conducted observations on the nursing and aggressive capacity to house 5 sows and their litters at 8.4 m2/sow (6.0 × 7.0 m),
behaviour of sows, and of sow and piglet behavioural timebudgets, while the other two GL pens had the capacity to house 7 sows and their
following transfer from farrowing crates to a commercially-operating litters at 8.1 m2/sow (6.0 × 9.5 m). Each pen had a solid partition with
group lactation system at 7 or 14 days postpartum, in comparison to sloped walls, two heated creep areas with lids and rice hulls for bed-
those that remained in farrowing crates. The hypothesis being tested is ding, with a zone only accessible to piglets that contained creep feed
that the behaviour of sows and piglets is improved when mixed into and a water trough (Fig. 1). From replicates 1–3, the creep area was
group lactation at 7 rather than 14 days postpartum, and in group protected using vertical bars along the entrance (approximately 20 cm
lactation compared to farrowing crates, as indicated by sow-piglet in- distance between bars allowing piglets to pass through) and a hor-
teractions (i.e., nursing behaviour and nosing interactions), social be- izontal anti-crush bar along the bottom of the entrance to the creep
haviour (i.e., piglet play and harmful manipulations, sow aggression) (10–15 cm from ground level). In replicate 4, the creep entrance was
and activity (i.e., exploratory behaviour). modified to better retain heat using a solid wooden barrier that in-
cluded an entry and exit doorway for piglets. The vertical protection
2. Methods bars and horizontal anti-crush bar remained. The flooring of the GL
pens was partially slatted and rice hulls were scatted on the solid por-
All animal procedures were conducted with prior institutional an- tion of the flooring weekly. Sow parity, weight and litter size were
imal ethics approval under the requirement of the New South Wales balanced between GL pens within replicates.
Prevention of Cruelty to Animals Act (1979) in accordance with the Cool drippers over the FC’s as well as over the dunging area of the
National Health and Medical Research Council/Commonwealth GL pens were activated (3 min on and 15 min off) when the internal
Scientific and Industrial Research Organisation/Australian Animal temperature exceeded 26 °C. Creep heat lamps automatically turned off
Commission Australian Code of Practice for the Care and Use of when the temperature exceeded 24 °C. During lactation, sows were fed
Animals for Scientific Purposes (NHMRC, 2013). a standard pelleted lactation diet ad libitum (14.9 MJ DE/kg and 16.2%
crude protein). For sows in FC, this was provided up to three times
2.1. Animals and housing daily, based on individual sow intake. The GL pens were fitted with a
single ad libitum feed hopper with the capacity to feed two sows si-
This experiment was conducted over 4 months in a farrowing and multaneously. Water was supplied ad libitum via one nipple drinker and
lactation unit at a large commercial piggery in southern New South one water trough per group lactation pen. Piglets in all housing treat-
Wales (NSW), Australia. The experimental building had a galvanised ments were offered creep feed from 14 days of age. Animals in FC and
roof and thermostatically controlled blinds that maintained an op- GL were managed by the same stock people, who had previous ex-
timum internal shed temperature of 21◦C. The maximum and minimum perience managing sows and piglets under both housing systems.
mean daily ambient temperatures ( ͦ C) for the region and experi-
mentation period were 27.3 and 12.4, respectively. 2.2. Experimental design and procedures
A total of 1791 piglets and 112 multiparous (parity range 1–8) sows
(Large White × Landrace, PrimeGro™ Genetics, Corowa, NSW) were All sows farrowed in standard FC. At 7 days postpartum (mean ±
studied over 4 time replicates. At approximately 7 days post-partum s.d.: 7.4 ± 1.1 days), the 12 GL7 sows (and their litters) studied per
(mean ± s.d., 7.6 ± 1.4 days), the 28 sow and litter units studied in replicate were transferred from FC to one of the two GL pens (one pen
each replicate were selected from a larger cohort of an average of 39 of 5 sow/litter units and one pen of 7 sow/litter units). Similarly, at 14
sows (range 37–41 sows) that had farrowed within 4 days of each other, days postpartum (mean ± s.d.: 13.6 ± 1.3 days), the 12 GL14 sows
were healthy (not injured, sick or lame) and had approximately 10 (and their litters) studied per replicate were transferred from FC to one
healthy piglets (piglets were not excluded based on size; but piglets that of the remaining two GL pens (one pen of 5 sow/litter units and one pen
were not thriving, sick or lame were cross-fostered to sows outside the of 7 sow/litter units). After mixing, GL7 and GL14 sow and litter units
experiment). Selected sows and their litters were allocated to one of remained in their respective group pens until weaning. The remaining 4
three lactation housing treatments that were applied until weaning: (1) FC sows (and their litters) studied per replicate continued in their FC

2
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

Fig. 1. Layout of group lactation pens. 1Length was 9.5 m for pens of 7 sow/litter units and 7 m for pens of 5 sow/litter units. 2Creep dimensions were 2.0 × 1.2 m for
pens of 7 sow/litter units and 2.1 × 0.85 m for pens of 5 sow/litter units. Position of hack rack in pens of 35 sow/litter units and pens of 47 sow/litter units.
CF = creep feed, WT = water trough, L = heat lamp, ND = nipple drinker.

Table 1 of the creep meaning that attempts to confine piglets to the creeps for
Summary of treatments1, number of focal animals per replicate, experimental 1 h post-mixing of sows were unsuccessful, so piglets and sows were
unit, and days of recording pre-weaning behaviour2. mixed simultaneously. Dybkjær et al. (2003) found that the method of
GL7 GL14 FC3 introduction to group lactation (i.e., individual sows with their litters
sequentially, all sows and piglets simultaneously, all sows followed by
Pen 1 Pen 2 Pen 1 Pen 2 FC7 cohort FC14 cohort all piglets 2 h later) had no effect on sow aggression or piglet mortality.
Researchers monitored animals for 1 h post-mixing in the present ex-
FC 1 FC 2 FC 1 FC 2
periment and interventions to prevent piglet injury or death due to
N focal sows/rep 2 2 2 2 1 1 1 1 crushing in this period were not required. Researchers did not interrupt
N focal piglets/ 4 4 4 4 2 2 2 2 aggressive behaviour between sows.
rep
Experimental GL GL GL GL FC7 cohort FC14 cohort
unit pen pen pen pen
Expression of Per focal sow/ Per focal sow/ Per focal Per focal sow/
2.3. Data collection
data piglet per pen piglet per pen sow/ piglet piglet per crate
per crate Video cameras with built-in infra-red lights (3-Axis dome camera
Behaviour with IR 800 T V lines, QC-8646; Jaycar electronics, Australia) that were
1 hr post-mixing A, N A, N – –
connected to a digital video recorder (TECHview 16 Channel stand-
Day 2 GL7 A, N, BT – A, N, BT –
Day 2 GL14 – A, N, BT – A, N, BT alone DVR, model QV-3039; Techbrands, Australia) were positioned
Pre-weaning A, N, BT A, N, BT A, N, BT A, N, BT above each GL pen and recorded for 1 h on the day of mixing (post-
mixing) commencing after the final pig was added to the pen, and from
1
GL7 and GL14 treatments were mixed into group lactation pens at 7 and 14 0700 to 1700 on the day after mixing (day 2) and the day prior to
days post-partum, respectively, whereas FC treatment remained in farrowing weaning (pre-weaning). One camera captured the entire area of the
crates throughout lactation.
2 smaller pens (i.e., housing 5 sow/litter units) but 2 cameras were re-
Behaviours recorded include A = sow aggression; N = sow nursing; BT = sow
quired to capture the area of the larger pens (i.e., housing 7 sow/litter
and piglet behavioural time budgets.
3
FC7 and FC14 cohorts were observed at the same days at GL7 and GL14 treat- units). Piglets were not recorded while inside the creep of the GL pens.
ments, respectively. At day 2 and pre-weaning, a GoPro camera (GoPro model Hero 3 white
edition, GoPro Inc., San Mateo, CA, USA) was fixed above each FC and
until weaning. The average age at weaning was 25.5 days (range 23–30 also recorded from 0700 to 1700 h.
days). Details on treatments, pens, experimental units and animals are The behaviour of four focal sows (two of high and two low parity;
summarised in Table 1. mean parity ± s.d. of 4.8 ± 1.3 and 1.7 ± 0.6, for high and low
Mixing of GL treatments proceeded as follows. Firstly, piglets were parities, respectively), and two focal piglets per focal sow (one male
transferred from FC to one of the two piglet creep areas located in the and one female of average weight; mean start weight ± s.d. of
allocated GL pen. The exit of the creep into the GL pens was blocked 3.0 ± 0.7 and 2.8 ± 0.5, for males and females, respectively) was
and the lid left open to prevent over-heating. The transfer of sows from studied per treatment and replicate (total of 48 focal sows and 96 focal
their FC to GL commenced after litters had been relocated. Mixing of piglets). Each GL pen contained one focal sow of high parity, one focal
sows and their litters into GL occurred within the hours of 0730 and sow of low parity, and four focal piglets. Of the four FC sows studied per
0930 h and was completed within approximately 30 min. Pens were replicate, one high and one low parity focal sow (as well as the two
mixed sequentially and the allocation of GL treatments to pens alter- focal piglets per sow) were video recorded on the same days as GL7
nated per replicate. Sows could manipulate the timber blocking the exit (labelled FC7 cohort), while the other high and low parity focal sows (as
well as the two focal piglets per sow) were video recorded on the same

3
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

Table 2
Description of sow and piglet behaviours observed around nursing (adapted from Singh et al. (2017) and Fraser (1980)).
Nursing behaviour Description

Nursing bout A bout of nursing where the sow is lying recumbent and at least 75% of the litter in farrowing crates, or 7 piglets in group lactation pens1, gather at udder
and massaging and/or nursing occurred for at least 90 seconds.
Start of nursing bout 75% of piglets in farrowing crates, or 7 piglets in group lactation pens, gather at udder and begin massaging.
End of nursing bout 75% of piglets in farrowing crates have stopped massaging (became still or moved away from the udder), less than 7 piglets in group lactation pens are
massaging the udder, or sow terminates massage/nursing by changing posture (sow terminated nursing bout).
Letdown At approximately the same time for all suckling piglets, udder massaging ceases. The heads of piglets become still, ears flattened, and bodies draw back
from the udder and extend noticeable. This lasts for approximately 10-20 s, after which piglets begin to dart from teat to teat.
Successful nursing A nursing bout in which letdown was observed.
Inter-nursing interval The time between the beginning of one successful nursing bout and the beginning of the next successful nursing bout.

1
Average of 1 sow per 10.6 piglets in group lactation pens.

days as GL14 (labelled FC14 cohort). For clarity, the number of focal data was not collected at day 2 for one of the two sows and her two
animals per experimental unit and the schedule of behavioural ob- focal piglets in the FC7 cohort. Four focal piglets had to be replaced
servations are summarised in Table 1. Different symbols were marked with a similarly sized littermate of the same sex during this experiment:
on the back of focal piglets and sows using stock paint to allow for two GL7 (1 female, 1 male; 1 piglet from a pen with 5 sow/litter units, 1
individual identification and this was re-applied before each recording piglet from a pen with 7 sow/litter units) and one GL14 (female; from a
day. The same focal sows were observed throughout lactation. A simi- pen with 5 sow/litter units) focal piglets died, while one GL7 piglet
larly sized littermate of the same sex replaced focal piglets that died or (female; from a pen with 7 sow/litter units) was removed for lameness.
were removed for injury/illness during lactation (total of 4 focal pig- Sow play behaviour was only recorded once, so was not analysed and
lets). Behaviour data were collected from video records by a single will not be further discussed.
observer. Formal statistical assessments of intra-observer reliability All statistical analyses were carried out using the SPSS statistical
were not conducted, but all pens/crates were re-watched by the same software package (SPSS 22.0, SPSS Inc., Chicago, Illinois, USA). For all
observer for 2 randomly selected hours and there was high agreement variables, observations were made on individual focal sows and focal
with previous categorisation (97.7% in hour 1 and 97.9% in hour 2). piglets and then averaged for each pen or FC cohort (see Table 1).
Variables were assessed for normality using visual methods (quantile-
2.3.1. Nursing behaviour quantile plots and histograms) in combination with Shapiro-Wilkes
Nursing frequency, latency to letdown, inter-nursing interval, total normality tests. Data on inter-nursing interval were y = 1 x transformed
duration of nursing, sow terminated nursing bouts and the success of prior to analysis so that residual variation was homogenous between
nursing bouts were continuously observed for 1 h post-mixing (GL7 and time replicates, pens and treatments. The significance level α was set at
GL14 pens only), and from 0700 to 1700 h on the day after mixing and P ≤ 0.05 for main and interactive effects. When there were significant
the day before weaning (GL7, GL14 pens and FC7, FC14 cohorts). Table 2 interactive effects the LSD test determined where least square means
provides a full description of behaviours recorded around nursing. differed. Non-significant interactions were removed from the model so
that the main effects could be better interpreted. Data for GL7, GL14,
FC7 and FC14 treatments are presented as least square means ± 95%
2.3.2. Sow aggression
confidence intervals. Significant main effects are presented in text as
The frequency of aggressive bouts, and the frequency and duration
least square means ± SEM.
of fights and bullying events were continuously recorded. These in-
The behaviours recorded in the present study were analysed as one
cluded any one or combination of pushes, head knocks and bites (see
of two sets of data. The first set compared sow social behaviour and
Samarakone and Gonyou, 2009). Bouts of aggression were defined as
piglet interactions with sows other than their dams between GL7 and
reciprocal and/or non-reciprocal bouts delivered or received by focal
GL14 treatments. In these comparisons, the aggressive and nursing be-
sows that did not exceed 5 s duration. Fights were defined as bouts of
haviours of sows over 1 h post-mixing into GL were examined using
reciprocal aggression that exceeded 5 s during which both sows in the
general linear mixed models (LMM). These analyses included treatment
fight were delivering aggression. Bullying events were defined as non-
(GL7 or GL14) as a fixed factor and controlled for time replicate and
reciprocal aggression that exceeded 5 s duration. A criteria of 5 s was
group lactation pen as random blocking factors. Sow aggression (re-
used to separate one bout of fighting/bullying from the next.
corded continuously from 0700 to 1700 h as well as a behavioural time
budget), sow-sow nosing interactions (recorded as a behavioural time
2.3.3. Sow and piglet behavioural time budgets budget) and piglet interactions with sows other than their dam (re-
The behavioural time budgets of focal sows and piglets were re- corded as a behavioural time budget) were also analysed using LMM.
corded using instantaneous point sampling at 5-min intervals from Each model included treatment (GL7 or GL14), day (the day after
0700 to 1700 h (121 sample points/day) on the day after mixing and mixing, day 2; day prior to weaning, pre-weaning) and their interaction
the day prior to weaning. General activity, exploratory and social be- as fixed factors. Repeated observations of the same GL pens over days
haviours were recorded, as described in Table 3. Piglets in GL could not were accounted for with a first order autoregressive correlation struc-
be observed while in the creep. To provide a valid comparison between ture, while time replicate and GL pen were included in the model as
FC and GL treatments, piglet behaviour was only recorded when outside random factors.
of the creep. As FC did not have a covered creep area, piglets in FC were All other data (sow nursing behaviour, sow and piglet behavioural
defined as being ‘in the creep’ if at least 70% of the piglet was on the timebudgets [excluding sow to sow and piglet to non-maternal sow
mat underneath the heat lamp. Behavioural time budgets were calcu- interactions] at days 2 and the day prior to weaning) were analysed
lated as a proportion of total observations for individual focal sows and using LMM which included the fixed effects of lactation housing (FC or
focal piglets. GL), litter age at mixing (7 or 14 days), observation day (day 2, pre-
weaning), and their 2- and 3-way interactions. Each analysis accounted
2.4. Statistical analysis for repeated observations of the same GL pens/FC cohorts over days
with a first order autoregressive correlation structure, while time
A technical malfunction in the first replicate meant that behaviour

4
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

Table 3
Description of the behaviour observed in studying the time budget of behaviour (adapted from Singh et al., 2017).
Behaviour Description

1
Inside the creep Group lactation piglet known or reasonably suspected as being inside one of the two group lactation creeps per pen. At least 70% of the
body of farrowing crate piglet lying on the heat mat located underneath the heat lamp. No specific behaviours were recorded when
piglets were in the creep.
Activity (outside of the creep)
Lying/inactive2 Sow/piglet lying or sitting, not engaged in activity.
Walking/standing2 Sow/piglet standing or walking but not otherwise engaged in activity (includes urinating and defecating).
Nursing3 Sow lying recumbent with at least 75% of litter (in farrowing crates) or 7 piglets (in group lactation pens) massaging udder and/or
nursing.
Suckling/udder massage1 Piglet mouth on teat or nose contact to udder with vertical head movements. For piglets in group lactation pens, it was also recorded
whether or not a piglet was suckling its dam or another sow.
Investigatory behaviours
Investigating pen/floor2 (or sow body1) Sow/piglet sniffing, touching, biting, rooting, pawing or rubbing any part of the pen or floor (or sow’s body other than head for
piglets).
Investigating food/water2 Sow/piglet sniffing, touching, pawing, eating or drinking food or water.
Social behaviours
Nosing piglet2 Sow/piglet touching or sniffing any part of the head or nose of a/another piglet.
Nosing sow2 Sow (in group lactation pens)/piglet touching, sniffing or nuzzling any part of the head or nose of another/a sow. For piglets in group
lactation pens, it was also recorded whether the piglet was nosing its dam or another sow.
Aggression2 Sow (in group lactation pens)/piglet pushing, knocking or biting another sow/piglet.
Play2 Sow (in group lactation pen)/piglet shaking head, pivoting, jumping, or running with bouncy or jerky movements.
Manipulative behaviour1 Piglet nibbling, sucking, chewing or rubbing with nose any part of the body of another piglet.

1
Behaviour recorded for piglets only.
2
Behaviour recorded for sows and piglets.
3
Behaviour recorded for sows only.

replicate and GL pen were included in the model as random factors. Table 4
Using this model, differences in the behaviour of sows and piglets in The effects of group lactation housing from 7 (GL7) or 14 (GL14) days post-
GL7, GL14 and FC treatments would be evident through a significant partum on sow aggressive and nursing behaviours for 1 h post-mixing. Data are
(P ≤ 0.05) housing × age at mixing interaction. least square means per focal sow and pen. For data that were transformed prior
to analysis, transformed least square means are presented with raw means in
parenthesis.
3. Results
Inter-sow aggression GL7 GL14 SEP1
A substantial amount of data was produced by the present research.
1 h post-mixing
Data that are most relevant to addressing our research hypothesis have Frequency of aggressive bouts2 4.7 6.1 1.6
been prioritised for presentation in this paper (i.e., main effects of Frequency of fights 1.0 1.2 0.31
housing, and the interactive effects of housing × age, housing × day Total time fighting (s) 30 51 14
and housing × age × day). Least square means for all variables, as well Average fight length (s) 33 43 10
Day 2 post-mixing
as the main and interactive effects of lactation housing, litter age at
Frequency of aggressive bouts2,3 0.89 (8.4) 0.97 (9.5) 0.09
mixing and observation day on these variables, are presented in Frequency of fights 0.06 0.19 0.19
Supplementary Tables S1 to S3. Total time fighting (s) 0.44 2.3 1.55
Average fight length (s)3 0.15 (0.44) 0.23 (2.3) 0.13
Day prior to weaning
3.1. Inter-sow aggression
Frequency of aggressive bouts2,3 1.1 (14) 1.2 (15) 0.09
Frequency of fights 0.38 0.31 0.19
Group lactation treatments (i.e., GL7 and GL14) did not significantly Total time fighting (s) 2.9 2.0 1.55
differ in the frequency of aggressive bouts, the frequency of fights, the Average fight length (s)3 0.22 (2.9) 0.30 (2.0) 0.13
total time fighting or the average fight length in the 1 h following 1
Pooled standard error of the mean.
mixing into GL (Table 4). The GL treatments did not significantly differ 2
Aggressive bouts delivered + received by focal sows.
in the frequency of aggressive bouts or fights, the total time fighting or 3
Log10(Χ + 1) transformed prior to analysis.
the average fight length recorded at day 2 post-mixing and the day prior
to weaning, but the frequency of aggressive bouts was higher pre-
significantly higher inter-nursing interval ( y = 1 x transformed least
weaning than at day 2 (8.9 ± 2.1 vs. 14.5 ± 2.1 bouts per focal sow,
square mean (backtransformed mean) ± SEM were 0.015 (66.7) ± 0.00
F1,15 = 4.5, P = 0.05; Table 4).
vs. 0.018 (55.6) ± 0.001, F1,21 = 18.9; Fig. 2D) than FC sows. De-
scriptive analysis of the data revealed that nursing behaviour remained
3.2. Nursing behaviour severely disrupted 24 h after grouping for a small proportion of GL
sows: one did not nurse successfully at all on day 2, while another
There were no successful nursing bouts in the 1 h period post- nursed successfully only once, although she had 10 unsuccessful at-
mixing, and there was no difference between GL treatments in the tempts to nurse. Nearly ¼ of GL sows averaged fewer than eight suc-
frequency of unsuccessful nursing bouts in this 1-hour period (grand cessful nursing bouts per day, compared with an average of 12.6 suc-
mean of 0.57 ± 0.12 bouts per focal sow; F1,11 = 0.41, P = 0.54). cessful nursing bouts for FC sows.
There was no effect of housing (FC or GL) on nursing frequency There was a significant housing × day interaction on the average
(S1). However, GL sows had significantly fewer successful nursing bouts duration of successful nursing bouts: FC sows had longer successful
(70 ± 3.9% vs. 85 ± 3.8%, F1,9.5 = 18.4; Fig. 2A), spent significantly nursing bouts than GL sows pre-weaning, but there was no difference at
less total time nursing successfully (32.7 ± 2.0 vs. 43.7 ± 2.9 min, day 2 (F1,22 = 7.2; S1). FC7 sows took significantly longer to letdown
F1,21 = 10.1; Fig. 2B), terminated significantly more nursing bouts than GL7 sows, but the same difference was no evident between the
(69 ± 3.0% vs. 49 ± 4.2%, F1,21 = 16.2,; Fig. 2C) and had a

5
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

Fig. 2. The effects of lactation housing (Housing: GL or FC), litter age at mixing (Age: 7 or 14 days postpartum) and observation day (Day: day 2 post-mixing or the
day prior to weaning; x-axis) on variables relating to sow nursing behaviour (A–D) and sow behavioural time budgets (E–H). Data for sows mixed into group lactation
at 7 (GL7; ) or 14 days (GL14; ), and for those that remained in farrowing crates but recorded at the same days as GL7 (FC7; ) or GL14 (FC14; ) treatments are
indicated by separate lines. Least square means ± 95% confidence are presented. Data on inter-nursing interval are y = 1 x back transformed with transformed means
available in supplementary Table S1. Note that the y-axis has been manipulated to enhance visual interpretation of effects and are inconsistent between charts.

GL14 and FC14 sows (Housing × age: F1,16.8 = 4.8; S1). The average spent significantly more time investigating food than FC7, GL7 or GL14
duration of unsuccessful nursing bouts were significantly longer for FC piglets, but at pre-weaning FC7 piglets spent the most time investigating
than GL sows at day 2, but pre-weaning FC sows had shorter un- food (Housing × day × age: F1,20 = 13.2; S2). Lactation housing had
successful nursing bouts (Housing × day: F1,22 = 5.1; S1) and spent no significant effect on the time piglets spent suckling, interacting with
significantly less total time nursing unsuccessfully (Housing × day: the sow, nosing other piglets, playing or delivering aggression (S3). The
F1,21 = 5.7; S1). time GL7 and GL14 piglets spent suckling (3.9 ± 1.0% vs. 9.8 ± 1.0%,
F1,15 = 20.9; Fig. 3F) and nosing (0.6 ± 0.2% vs. 1.1 ± 0.2%, F1,15 =
3.3. Sow behavioural time budgets 13.9; Fig. 3G) sows other than their dam significantly increased from
day 2 to pre-weaning.
FC sows spent significantly more time lying (69 ± 1.8% vs.
62 ± 1.4%, F1,12.3 = 8.5; Fig. 2E) and significantly less time walking/ 4. Discussion
standing (0.9 ± 0.5% vs. 5.3 ± 0.4%, F1,18 = 59.5; Fig. 2F), in-
vestigating the pen (1.3 ± 0.5% vs. 4.4 ± 0.4%, F1,18 = 23.4; Fig. 2G) Mixing sows and their litters into GL at 7 or 14 days postpartum
and interacting with piglets (7.2 ± 0.6% vs. 9.3 ± 0.6%, F1,11.3 = induced a number of differences in the behaviour of sows and piglets in
6.3; Fig. 2H) than GL sows. There were no effects of housing on the time GL compared to FC, but age at grouping had few effects.
sows spent sitting, nursing or investigating food (S2). GL14 sows spent
significantly more time nosing other sows than GL7 sows (0.8 ± 0.1 4.1. Sow-piglet interactions
and 0.2 ± 0.1, F1,14 = 8.7, P = 0.011), but aggressive behaviour (re-
corded as a timebudget) did not significantly differ between the two GL Sows housed in GL pens nosed piglets more frequently than sows
treatments (S2). confined in FC, regardless of mixing at 7 or 14 days postpartum or of
day post-mixing. Other research has found more sow-piglet interactions
3.4. Piglet behavioural time budgets in single-litter loose-housed sow lactation pens than in FC (e.g., Singh
et al., 2017; Bolhuis et al., 2018), and that loose-housing of sows allows
Piglets in GL spent significantly less time lying (18 ± 4.9% vs. for better development and/or expression of other maternal behaviours
29 ± 5.1%, F1,8.8 = 11.6; Fig. 3A) and significantly more time (Arey and Sancha, 1996; Thodberg et al., 2002; Grimberg-Henrici et al.,
walking/standing/sitting (5.9 ± 0.4% vs. 4.5 ± 0.4%, F1,7.6 = 6.3; 2016; Chidgey et al., 2017; Singh et al., 2017). Singh et al. (2017)
Fig. 3B) outside of the creep than those in FC. There were significant suggested that either piglets had greater accessibility to the sow’s head
housing × day interactive effects on the time piglets spent in the creep in loose-housed sow systems or that loose-housed sows were better able
(F1,22 = 4.3; Fig. 3C), investigating the pen (F1,21 = 6.8; Fig. 3D) and to orient themselves towards piglets. That there were no effects of
manipulating other piglets (F1,22 = 9.8; Fig. 3E). While there were no housing treatment on piglet interactions with the sow in the present
effects of housing at day 2 for any of these variables, at pre-weaning GL study (observed as a piglet behavioural time budget) supports the latter
piglets spent more time in the creep and investigating the pen, and less hypothesis, although determining who has initiated the interaction is
time manipulating other piglets, than FC piglets. At day 2, FC14 piglets obviously difficult (Singh et al., 2017).

6
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

Fig. 3. The effects of lactation housing


(Housing: GL or FC), litter age at mixing (Age:
7 or 14 days postpartum) and observation day
(Day: day 2 post-mixing or the day prior to
weaning; x-axis) on variables relating to piglet
behavioural time budgets. Data for sows mixed
into group lactation at 7 (GL7; ) or 14 days
(GL14; ), and for those that remained in far-
rowing crates but recorded at the same days as
GL7 (FC7; ) or GL14 (FC14; ) treatments are
indicated by separate lines. Piglet interactions
with non-maternal sows are only relevant to
GL7 and GL14 treatments (3F and 3G). Least
square means ± 95% confidence are pre-
sented. When there is a significant interactive
effect, means with different superscript letters
a,b
or c,d differ at P ≤ 0.05. Note that the y-axis
has been manipulated to enhance visual inter-
pretation of effects and are inconsistent be-
tween charts.

Research in a range of species demonstrates the relationship be- While attempts to nurse were made in the 1 h post-mixing, no sow in
tween receiving good maternal care, particularly maternal tactile con- the present experiment successfully nursed in this period, and, for a
tact, in the early postnatal period and the expression of good maternal quarter of GL sows, nursing behaviour remained disrupted 24 h after
behaviour as an adult (see review by Curley and Champagne, 2016). In grouping. Further, regardless of day post-mixing, GL sows had a lower
pigs, sows that were reared by a loose-housed dam or have prior ex- proportion of successful nursings, terminated a higher proportion of
perience rearing piglets in a loose-housed sow lactation system display nursings, spent less time nursing successfully and had a longer inter-
better maternal behaviour than sows that are inexperienced with loose- nursing interval during lactation than sows in FC.
housing (Thodberg et al., 2002; Chidgey et al., 2016). Whether the As demonstrated by Wattanakul et al. (1997b), relocation of the sow
maternal behaviour of GL sows further differentiates from those in FC per se impedes piglet’s ability to locate the dam at nursing resulting in
over subsequent generations requires investigation. cross-suckling. Cross-suckling is infrequently or not observed in young
Although the total frequency of nursing bouts and the proportion of pigs living under semi-natural conditions (Newberry and Wood-Gush,
time piglets spent suckling were not affected by treatment in the pre- 1985; Jensen, 1986) but is a known challenge of GL systems (see van
sent study, there were a number of indicators that nursing behaviour Niewamerongen et al., 2014), increasing competition at the udder and
was disrupted in GL. Similarly to previous studies (Wattanakul et al., ultimately disrupting the sows (Pedersen et al., 1998). This disruption
1997a; Dybkjær et al., 2003), the present experiment has found that may be responsible for sows terminating more nursing bouts in GL in
mixing into GL interferes with nursing behaviour for at least for 1 h the present experiment. Indeed, in contrast to the present experiment,
following transfer. An extended observation period is required to con- GL systems that include voluntary access pens for individual sows and
firm the extent of this disruption post-mixing, and to determine whe- their litters report no difference in the frequency of sow terminated
ther the latency to nurse is affected by litter age at mixing into GL. nursing bouts between multi-litter and single-litter housing systems

7
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

(Arey and Sancha, 1996; Bohnenkamp et al., 2013; Thomsson et al., manipulative behaviour may arise when the physical environment does
2016). Achieving high performing GL may require the inclusion of not allow for the expression of highly motivated behaviours like rooting
voluntary access zones where the sow can achieve a level of isolation or suckling (Telkänranta and Edwards, 2018). Indeed, in the present
for herself and her litter (Baxter et al., 2018). experiment, FC piglets spent less time exploring the pen than GL piglets.
No effect of age at mixing on nursing behaviour or cross-suckling On the other hand, aggressive and social affiliative behaviours may be
was found by the present experiment or by others (Thomsson et al., more reliant on the immediate social environment and internal state of
2016). In the present experiment, GL piglets suckled mostly from their the animal.
dam on the day after mixing, but mostly from other dams on the day Arey and Sancha (1996) reported a nine-fold increase in piglet play
prior to weaning. The frequency with which GL piglets interacted with behaviour in GL than in FC, but both the present experiment and van
sows other than their mother was also higher pre-weaning than the day Nieuwamerongen et al. (2015) found no effect of lactation housing in
after mixing. Understanding whether cross-suckling is the consequence the frequency of piglet play. Piglet play has been stimulated by mixing
of reduced maternal bonds or increased piglet hunger, or both, is re- of unfamiliar conspecifics (Donaldson et al., 2002), and there is evi-
quired in order to develop management practices that reduce its in- dence of increased play when piglets are housed in enriched (e.g.,
cidence. substrate, increased space, loose-housed sow) single-litter lactation
In the present experiment, GL sows were confined in FC from 7 days pens (Chaloupková et al., 2007; Oostindjer et al., 2011; Singh et al.,
before parturition until day 7 or 14 of lactation. Confinement during 2017). The frequency with which GL piglets engaged in socially fa-
parturition and early lactation may disrupt the establishment of nursing cilitated behaviours such as play may be underestimated in the present
behaviour and maternal bonding (discussed by Singh et al., 2017). experiment by the fact that a lower proportion of piglets per group
Group-housed lactating sows that are not confined during parturition lactation pen (4 piglets per pen, ˜ 6–8% of piglets) than per farrowing
and the early days of the piglets’ life initiate more, and terminate fewer, crate (2 piglets per FC, ˜ 20% of piglets) were observed. Further, some
nursing bouts than group-housed lactating sows that farrow in crates behaviours, such as butting and shoving, have an ambiguous motiva-
and remain confined until day 11 post-partum (Dybjaer et al., 2001). A tional basis as they are observed during aggressive interactions and
study of 10 commercial farms found that neither nursing frequency nor contact play (Lawrence et al., 2018). Finally, for behavioural events
success, observed in weeks 4 and 6 of lactation, differed between multi- that are generally short in duration or take up a small proportion of the
and single-litter systems when sows are loose-housed during parturition behavioural time budget, Verdon et al. (2016) suggested that con-
and lactation (Šilerová et al., 2006). tinuous observations may be the most reliable sampling technique.

4.2. Sow-sow interactions 5. Conclusion

The results of the present study, and those of Thomsson et al. Mixing sows and their litters into group lactation pens after an in-
(2015), demonstrate that mixing unfamiliar sows at day 7 or 14 of itial period of 7 or 14 days induced behavioural changes in sows and
lactation results in comparable levels of inter-sow aggression in the piglets compared to housing them in farrowing crates during lactation.
hours immediately post-mixing. Sow behavioural time budgets and However, there were few differences between grouping at 7 or 14 days
continuous observations taken in the present experiment show that postpartum on behaviour. In particular, group lactation supports the
inter-sow aggression and fighting continued throughout lactation, and expression of some highly motivated behaviours, such as maternal in-
that sows delivered more aggression pre-weaning than at day 2. GL teractions in sows and exploratory behaviour in piglets, with a reduc-
sows were also more active pre-weaning than at day 2 in the present tion in harmful manipulative behaviour in piglets. These benefits come
experiment, and this may have increased the frequency of encounters at the cost of disrupted nursing behaviour. Research should investigate
between sows that result in aggression. As is the case of group-housed whether cross suckling is motivated by hunger, reduced maternal
gestating sows (see Verdon et al., 2015), the design of the GL pen is bonds, or both, so that interventions can be designed to reduce its in-
likely critical to the regulation of sow aggression. cidence in group lactation pens.
Sows mixed into GL at 14 days post-partum displayed more socially
affiliative nosing interactions with other sows than those mixed at 7 Conflicts of interest
days post-partum. Under natural conditions, sows typically abandon the
nest with their litters between 7 to 14 days post-partum and gradually The authors declare that there are no conflicts of interest.
begin integrating their piglets into their group of familiar sows and
their offspring (Jensen, 1988). By contrast, two-step group lactation Acknowledgements
systems, such as the one utilised in the present experiment, abruptly
mix unfamiliar sows and litters during lactation. The needs of sows This work was supported by Australian Pork Cooperative Research
(e.g., spatial, social, dietary) in the period of nest occupation differs Centre (Pork CRC; grant 1A-113). We are grateful to Rivalea Australia,
from those in later lactation (Baxter et al., 2018). The motivation of and particularly to Brian McLeod, for technical support and use of their
sows to socially interact with peers may be higher at 14 compared to 7 facilities and animals. We would also like to extend our gratitude to the
days postpartum, possibly coinciding with reduced maternal defen- anonymous reviewer for constructive comments on a previous version
siveness as piglets become more autonomous with age. of this manuscript.

4.3. Piglet-piglet interactions Appendix A. Supplementary data

In the present experiment, FC piglets directed more harmful ma- Supplementary material related to this article can be found, in the
nipulative behaviour toward pen mates than those in GL. However, no online version, at doi:https://doi.org/10.1016/j.applanim.2019.03.
effects of treatment on other socio-negative behaviours (i.e., piglet 001.
aggression) or socio-positive behaviours (i.e., affiliative nosing and
play; Lawrence et al., 2018) were found. Other research reports reduced References
harmful manipulations, but comparable aggressive and affiliative
nosing interactions, between piglets in multi-litter (van Arey, D.S., Sancha, E.S., 1996. Behaviour and productivity of sows and piglets in a family
Nieuwamerongen et al., 2015) and single-litter (Singh et al., 2017) system and in farrowing crates. Appl. Anim. Behav. Sci. 50, 135–145.
Baxter, E.M., Andersen, I.L., Edwards, S.A., 2018. Sow welfare in the farrowing crate and
loose-housed sow lactation systems, compared to those in FC. Harmful

8
M. Verdon, et al. Applied Animal Behaviour Science xxx (xxxx) xxx–xxx

alternatives. In: Špinka, M. (Ed.), Ed.), Advances in Pig Welfare. Woodhead enrichment and loose housing of lactating sows on piglet behaviour before and after
Publishing, pp. pp. 27–72. weaning. Appl. Anim. Behav. Sci. 134, 31–41.
Bohnenkamp, A.-L., Traulsen, I., Meyer, C., Müller, K., Krieter, J., 2013. Comparison of Pedersen, L.J., Studnitz, M., Jensen, K.H., Giersing, A., 1998. Suckling behaviour of
growth performance and agonistic interaction in weaned piglets of different weight piglets in relation to accessibility to the sow and the presence of foreign litters. Appl.
classes from farrowing systems with group or single housing. Animal 7, 309–315. Anim. Behav. Sci. 58, 267–279.
Bolhuis, J., Raats-van den Boogaard, A., Hoofs, A., Soede, N., 2018. Effects of loose Samarakone, T.S., Gonyou, H.W., 2009. Domestic pigs alter their social strategy in re-
housing and the provision of alternative nesting material on peri-partum sow beha- sponse to social group size. Appl. Anim. Behav. Sci. 121, 8–15.
viour and piglet survival. Appl. Anim. Behav. Sci. 202, 28–33. Schrey, L., Kemper, N., Fels, M., 2018. Behaviour and skin injuries of sows kept in a novel
Chaloupková, H., Illmann, G., Bartoš, L., Špinka, M., 2007. The effect of pre-weaning group housing system during lactation. J. Appl. Anim. Res. 46, 749–757.
housing on the play and agonistic behaviour of domestic pigs. Appl. Anim. Behav. Sci. Šilerová, J., Špinka, M., Šárová, R., Slámová, K., Algers, B., 2006. A note on differences in
103, 25–34. nursing behaviour on pig farms employing individual and group housing of lactating
Chidgey, K.L., Morel, P.C., Stafford, K.J., Barugh, I.W., 2016. The performance and be- sows. Appl. Anim. Behav. Sci. 101, 167–176.
haviour of gilts and their piglets is influenced by whether they were born and reared Singh, C., Verdon, M., Cronin, G., Hemsworth, P., 2017. The behaviour and welfare of
in farrowing crates or farrowing pens. Livest. Sci. 193, 51–57. sows and piglets in farrowing crates or lactation pens. Animal 11, 1210–1221.
Chidgey, K.L., Morel, P.C., Stafford, K.J., Barugh, I.W., 2017. Sow and piglet behavioral Telkänranta, H., Edwards, S.A., 2018. Lifetime consequences of the early physical and
associations in farrowing pens with temporary crating and in farrowing crates. J. Vet. social environment of piglets. In: Špinka, M. (Ed.), Advances in Pig Welfare.
Behav. Clin. Appl. Res. 20, 91–101. Woodhead Publishing, pp. pp. 101–136.
Curley, J.P., Champagne, F.A., 2016. Influence of maternal care on the developing brain: Thodberg, K., Jensen, K.H., Herskin, M.S., 2002. Nursing behaviour, postpartum activity
mechanisms, temporal dynamics and sensitive periods. Front. Neuroendocrinol. 40, and reactivity in sows: effects of farrowing environment, previous experience and
52–66. temperament. Appl. Anim. Behav. Sci. 77, 53–76.
Donaldson, T.M., Newberry, R.C., Špinka, M., Cloutier, S., 2002. Effects of early play Thomsson, O., Bergqvist, A.-S., Sjunnesson, Y., Eliasson-Selling, L., Lundeheim, N.,
experience on play behaviour of piglets after weaning. Appl. Anim. Behav. Sci. 79, Magnusson, U., 2015. Aggression and cortisol levels in three different group housing
221–231. routines for lactating sows. Acta Vet. Scand. 57, 9.
Dybjaer, L., Olsen, A.N., Møller, F., Jensen, K.H., 2001. Effects of farrowing conditions on Thomsson, O., Sjunnesson, Y., Magnusson, U., Eliasson-Selling, L., Wallenbeck, A.,
behaviour in multi-suckling pens for pigs. Acta Agric. Scand. Sect. A-Anim. Sci. 51, Bergqvist, A.-S., 2016. Consequences for piglet performance of group housing lac-
134–141. tating sows at one, two, or three weeks post-farrowing. PLoS One 11 e0156581.
Dybkjær, L., Olsen, A.N., Møller, F., Jensen, K.H., Giersing, M., 2003. Effects of group size van Nieuwamerongen, S., Bolhuis, J., van der Peet-Schwering, C., Soede, N., 2014. A
during pregnancy and introduction method on behaviour of relevance for piglet review of sow and piglet behaviour and performance in group housing systems for
performance in multi-suckling pens. Acta Agric. Scand. Sect. A-Anim. Sci. 53, 83–91. lactating sows. Animal 8, 448–460.
Fraser, D., 1980. A review of the behavioural mechanism of milk ejection of the domestic van Nieuwamerongen, S., Soede, N., van der Peet-Schwering, C., Kemp, B., Bolhuis, J.,
pig. Appl. Anim. Ethol. 6, 247–255. 2015. Development of piglets raised in a new multi-litter housing system vs. con-
Grimberg-Henrici, C.G., Büttner, K., Meyer, C., Krieter, J., 2016. Does housing influence ventional single-litter housing until 9 weeks of age. J. Anim. Sci. 93, 5442–5454.
maternal behaviour in sows? Appl. Anim. Behav. Sci. 180, 26–34. Verdon, M., Hansen, C.F., Rault, J.-L., Jongman, E., Hansen, L., Plush, K., Hemsworth, P.,
Hemsworth, P., 2018. Key determinants of pig welfare: implications of animal manage- 2015. Effects of group housing on sow welfare: a review. J. Anim. Sci. 93,
ment and housing design on livestock welfare. Anim. Prod. Sci. 58, 1375–1386. 1999–2017.
Jensen, P., 1986. Observations on the maternal behaviour of free-ranging domestic pigs. Verdon, M., Morrison, R.S., Hemsworth, P.H., 2016. Rearing piglets in multi-litter group
Appl. Anim. Behav. Sci. 16, 131–142. lactation systems: effects on piglet aggression and injuries post-weaning. Appl. Anim.
Jensen, P., 1988. Maternal behaviour and mother—young interactions during lactation in Behav. Sci. 183, 35–41.
free-ranging domestic pigs. Appl. Anim. Behav. Sci. 20, 297–308. Wattanakul, W., Sinclair, A., Stewart, A., Edwards, S., English, P., 1997a. Performance
Lawrence, A.B., Newberry, R.C., Špinka, M., 2018. Positive welfare: what does it add to and behaviour of lactating sows and piglets in crate and multisuckling systems: a
the debate over pig welfare? In: Špinka, M. (Ed.), Advances in Pig Welfare. Woodhead study involving European White and Manor Meishan genotypes. Anim. Sci. 64,
Publishing, pp. pp. 415–444. 339–349.
Newberry, R.C., Wood-Gush, D.G., 1985. The suckling behaviour of domestic pigs in a Wattanakul, W., Stewart, A., Edwards, S., English, P., 1997b. Effects of grouping piglets
semi-natural environment. Behaviour 95, 11–25. and changing sow location on suckling behaviour and performance. Appl. Anim.
Oostindjer, M., van den Brand, H., Kemp, B., Bolhuis, J.E., 2011. Effects of environmental Behav. Sci. 55, 21–35.

You might also like