The effect of feed restriction on belly nosing
behaviour in weaned piglets
Andrea Bruni, V. Margaret Quinton, Tina M. Widowski *
Department of Animal and Poultry Science, University of Guelph, Guelph, Ont. N1G2W1, Canada
Accepted 2 March 2007
Available online 16 May 2007
Abstract
Belly nosing is an abnormal behaviour pattern most commonly seen in early-weaned piglets. Because
belly nosing resembles massaging the sow’s udder, it has been suggested that the behaviour is redirected
suckling and may be associated with hunger or feeding motivation. The objective of this study was to
determine the effect of feed restriction on the behaviour of newly weaned piglets. Eight groups of four
piglets were fed ad libitum during weeks 1, 2 and 3 post-weaning (control), while treatment piglets (eight
groups of four) were fed ad libitum during weeks 1 and 3, and feed restricted during week 2. During
restriction, ad libitum intake from the previous day was determined from control groups (voluntary intake of
1.9 times maintenance energy requirement), averaged, and treatment piglets were provided with 65% of this
amount (1.2 times maintenance energy requirement). While both groups continued to gain weight during the
restriction period, weight gain was significantly lower for restricted piglets compared to controls (P < 0.05).
Behaviour was observed on days 5, 7, 9, 12, 14, 17, and 20 post-weaning using a 5-min scan sampling
method for 6 h/day. During the period of restriction, treatment piglets spent less time at the feeder
(5.3 � 0.9%; controls, 8.3 � 0.6%, P < 0.05) and more time nosing and chewing at the pen (4.1 � 1.0%;
controls 1.4 � 0.4%, P < 0.05) compared to controls. Mean percentage of time spent belly nosing was not
statistically different (3.5 � 0.89%; controls 2.1 � 0.53%, P = 0.13), but the variation among piglets was
significantly greater in treatment groups (P < 0.05) during the restriction period. During the post-restriction
period, treatment piglets engaged in significantly more belly nosing behaviour (1.8 � 0.6%; controls,
0.5 � 0.2%, P < 0.05), while nosing and chewing at the pen was similar between the two groups. Feed
restriction stimulated foraging-type behaviour directed at the pen during the period of restriction, and
increased and altered the time course for belly nosing, but it did not elicit belly nosing in all pigs. Individual
differences in response to feed restriction suggest that there is a causal relationship between low feed intake
and belly nosing, but only in piglets with a predisposition for the behaviour.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Piglet behaviour; Belly nosing; Feed restriction; Hunger
www.elsevier.com/locate/applanim
Applied Animal Behaviour Science 110 (2008) 203–215
* Corresponding author. Tel.: +1 519 824 4120x52408; fax: +1 519 836 9873.
E-mail address: [email protected] (T.M. Widowski).
0168-1591/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.applanim.2007.03.016
1. Introduction
Under natural rearing conditions, weaning is a gradual process that involves the transition
from relying solely on sow’s milk to complete reliance of nourishment on solid food while sows
and piglets remain in contact for an extended period of time (Newberry and Wood-Gush, 1985;
Jensen and Recen, 1989; Jensen, 1995). During this process piglets go through three phases:
exclusively suckling, a period of both suckling and independent ingestion as the sow initiates
fewer nursings and terminates more (Jensen and Recen, 1989; Jensen, 1995), and finally,
exclusively independent ingestion. Early weaning in intensive rearing systems involves an abrupt
switch from suckling to feeding, and the absence of the gradual nature of this transition in diet
results in low voluntary feed intake and reduced performance referred to as post-weaning ‘growth
check’ (Pluske et al., 1995; Le Dividich and Seve, 2001). Abrupt and early weaning also results in
the development of abnormal behaviour, such as belly-nosing (see review by Widowski et al.,
2008). Belly nosing follows a consistent time course; it typically develops around 3–5 days after
weaning, peaks around 2 weeks after weaning and then declines (Blackshaw, 1981; Gonyou
et al., 1998; see review by Widowski et al., 2008).
It is well established that both a delay in initiating feeding and the performance of belly nosing
are more prevalent in piglets weaned at an earlier age (Metz and Gonyou, 1990; Gonyou et al.,
1998; Worobec et al., 1999). It has also been recognized that there is a great deal of individual
variation in the amount of time piglets spend belly nosing (Straw and Bartlett, 2001; Li and
Gonyou, 2002). Since belly nosing behaviour clearly resembles the motor patterns involved in
massaging the sow’s udder during the appetitive and post-consummatory phases of nursing, it has
been suggested that it is redirected suckling (Metz and Gonyou, 1990; Dybkjær, 1992; also see
review by Widowski et al., 2008) and might be associated with hunger or feeding motivation
(Gardner et al., 2001; Li and Gonyou, 2002). However, studies that have examined the effect of
different aspects of feeding on the performance of belly nosing behaviour indicate that the
association is not straightforward.
Neither offering a nutritionally complex diet versus a simple diet (Weary et al., 1999) nor
providing piglets with a low quality versus high quality diet (in terms of protein) (Gardner et al.,
2001) affected the performance of belly nosing behaviour, even though piglets had lower intakes
and grew more slowly on the poor quality diets. Additionally, examination of the temporal
association between the performance of belly nosing and feeding, through sequential analysis,
revealed no relationship between these two behaviour patterns (Li and Gonyou, 2002). However,
belly nosing has been demonstrated to be associated with some aspects of ingestive behaviour.
Feeding piglets a mash/liquid diet increases early dry matter intake and results in significantly
less time belly nosing compared to piglets fed a standard pelleted diet (Rau, 2002; Brooks et al.,
2001; Orgeur et al., 2003). Providing piglets water from a bowl type drinker as opposed to a
nipple drinker increased feed intake during the first 48 h after weaning and also reduced the
performance of belly nosing (Torrey and Widowski, 2004).
Belly nosing has also been found to be associated with poor growth rates (Fraser, 1978; Straw
and Bartlett, 2001; Torrey and Widowski, 2006). Several studies have reported a negative
correlation between belly nosing behaviour and feeding; those individuals that perform more
belly nosing spend less time at the feeder (Li and Gonyou, 2002; Bruni, 2004; Torrey and
Widowski, 2006). This raises the question: do piglets that develop belly nosing simply have less
time available for other activities and consequently spend less time eating, or is there a causal
relationship between low feed intake and belly nosing? One way to answer this question is to
determine whether an imposed feed restriction can elicit belly nosing behaviour either by causing
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215204
hunger (the psychological state associated with low nutrient levels) or through reduced growth
(the metabolic consequences of low nutrient status). Therefore, the aim of this study was to
determine the effect imposed feed restriction, on the performance of oral–nasal behaviour in
newly weaned piglets. We hypothesized that if low feed intake is causally related to belly nosing,
piglets subjected to feed restriction would perform more of the behaviour.
2. Materials and methods
The University of Guelph Animal Care Committee approved all procedures of this study.
2.1. Animals and housing
A total of 64 Yorkshire piglets (16 in each of 4 trials), from the Arkell Swine Research facility were used
in this study. For each separate trial, sixteen piglets (with an average weaning weight of 6.40 � 0.08 kg),
from three or four different litters were selected based on weight (between 5.0 and 7.5 kg) and gender
(approximately an equal distribution of gilts and barrows). Piglets were weaned between 18 and 22 days of
age, individually marked, weighed and divided randomly into groups of four.
At weaning, piglets were moved from standard farrowing crates to an on-site nursery room. All groups of
piglets were mixed equally, balancing for weight and gender. The room contained four separate raised deck
pens, each having a dimension of 1.13 m � 1.13 m and a plastic coated expanded metal open mesh floor.
Each pen was equipped with a single nipple drinker, a four-holed stainless steel feeder, and a single heating
pad. Heating pads were controlled manually, monitored daily and adjusted as required to maintain pig
comfort.
Daily ambient temperature averaged approximately 28.36 � 2.6 8C. Throughout the experiment, piglets
were provided with standard phase II feed (University of Guelph, Arkell Feed Mill, Guelph, Ont., Canada;
Table 1) and were given free access to water from bite-style nipple drinkers.
2.2. Experimental design
Two pens of four pigs within each of the four trials were randomly assigned to treatment while the other
two served as controls. Thus, the experiment was a randomized complete block design with subsampling
(Steel et al., 1997), which consisted of two pens per treatment per trial (total of eight replicates). Piglets
assigned to the control group were fed ad libitum for the duration of the experiment (weeks 1, 2 and 3 post-
weaning), while treatment piglets were exposed to a regimen that consisted of a weekly alternation of ad
libitum (weeks 1 and 3) and restricted feeding (week 2 post-weaning). Essentially, the experiment consisted
of three periods that included: the pre-restriction period (days 1–7 of the experiment), the restriction period
(days 8–14) and the post-restriction period (days 15–21).
During the restriction period (second week post-weaning), feed from control groups was weighed back
and intake from the previous day was determined on a daily basis. The amount consumed by each of the
control groups was averaged, and treatment piglets were provided with 65% of this amount. For ethical
reasons, the level of restriction applied in this experiment was chosen based on maintenance energy
requirements for piglets this age such that feeding levels for restricted piglets would be well below voluntary
feed intake, but they would not lose weight.
2.3. Data collection
2.3.1. Feed intake and growth
Piglets were fed at 09:00 h each morning. This daily process consisted of weighing back feed, removing
stale feed, and adding a weighed out amount of fresh feed to feeders. Feed spillage was minimal and was not
accounted for in estimates of intake. Feed intake was determined on a per pen basis.
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215 205
Piglets were individually weighed at weaning and again on days 8, 15 and 21 post-weaning. Weekly
weight gain was determined for each piglet.
2.3.2. Behaviour
Piglets were video taped continuously for a 24-h period on days 5, 7, 9, 12, 14, 15, 17 and 20 post-
weaning. Cameras (Panasonic WV-BP130, BP140 CCTV) were positioned on tripods, which were mounted
to one side of each pen. Cameras were connected to a recording system that consisted of a monitor, a
multiplexer (Simplex Multiplexer) and a time-lapse VCR (Panasonic AG-6730/6040), which recorded at a
speed of 24 h. This allowed all four pens to be viewed at once. For behavioural observations, data was
collected on days 5 and 7 when belly nosing characteristically first develops, days 9, 12 and 14 when belly
nosing usually peaks and days 17 and 20 when belly nosing is most often observed to decline.
Behavioural data were collected using a scan sampling method every 5 min for 6 h per day (Gardner
et al., 2001). Behaviour was sampled during two periods: 06:00–09:00 h, immediately before feeding when
feeders were empty and piglets were presumably hungry, and 14:00–17:00 h, after restricted piglets had
several hours to eat but usually still had feed available in their feeders. Each piglet was individually marked
and its behaviour was recorded in order to obtain an estimate of its time budget. Behaviour was allocated to
one of eight mutually exclusive behavioural categories listed in Table 2.
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215206
Table 1
Nutrient levels and ingredient composition (%) of Piglet Starter II diet
Ingredient Percentage in diet
Corn 32.3
Barley 8.0
Wheat 20.0
Soybean meal 22.0
Fish meal 5.0
Whey 8.0
Fat 2.0
Lysine–HCl, 79% 0.30
Methionine 0.01
Threonine 0.12
Dicalcium phosphate 0.60
Limestone 0.70
Iodized salt 0.30
Vitamin premixa 0.50
Mineral premixb 0.10
Tylosin premixc 0.10
Nutritive value (calculated)
DE (kcal/kg) 3450
Crude protein (%) 19.83
Calcium (%) 0.82
Phosphorus (%) 0.65
Available phosphorus (%) 0.62
Values are presented on an as-fed basis. This diet was available in crumble form and was manufactured at University of
Guelph, Arkell Feed Mill, Guelph, Ont., Canada.a Vitamin premix supplied the following per kg of feed: vitamin A, 10,000 IU; cholecalciferol, 1000 IU; DL-a-
tocopherol acetate, 56 IU; menadione, 2.5 mg; choline, 500 mg; pantothenic acid, 15 mg; riboflavin, 5 mg; folic acid,
2 mg; thiamine, 1.5 mg; pryoxidine, 1.5 mg; biotin, 0.20 mg; vitamin B12, 0.025 mg.b Mineral premix supplied the following per kg feed: Cu, 15 mg; Zn, 104 mg; Fe, 100 mg; Mn 19 mg.c Tylosin premix supplied 22 mg/kg of complete feed.
2.4. Statistical analysis
All statistical analyses were conducted using the SAS statistical software program (SAS Inst., Inc., Cary,
NC). The Proc Mixed procedure was used to determine any differences in feed intake, weight gain and
behavioural means between treatment and control groups over each period. The arcsine of the square root
transformation was used to transform data recorded as proportions when necessary. Linear and quadratic
orthogonal polynomial contrasts across the 3 days in the restriction period and linear orthogonal polynomial
contrasts (differences) across the 2 days in the post-restriction period were analyzed in order to compare
trends across days in the treated and control pens (Kuehl, 1994). Pre-restriction values were used as
covariates for analysis of both the restriction and post-restriction periods in order to eliminate any
differences in behaviour between the two groups before treatment was applied.
Residual plots indicated the variance within pens differed between the treated and control pens for some
of the behavioural traits recorded, particularly belly nosing. Two models were fit to the data set. The
homogenous model considered the within-pen variance to be the same across treatments, while the
heterogeneous model structure considered the within-pen variance different across treatments. The log
likelihood ratio test was employed to determine if the heterogeneous model was a significantly better fit to
the data set, and to examine any differences in variance due to treatment (Steel et al., 1997). The
heterogeneous model was a significantly better fit for belly nosing behaviour, while the homogenous
model was a better fit for all other behaviour patterns recorded.
Behaviour data were averaged over the sampling days for each piglet during week 2 and for all 3 weeks
of the trial combined. Pearsons correlations (PROC CORR) were performed on data from individual piglets
to identify relationships between weight gain and average time spent belly nosing and between time at the
feeder and belly nosing during the restriction period and overall. Data from one piglet in the control group
was removed from the data set due to illness.
3. Results
3.1. Feed intake and growth performance
Average weaning weight was 6.3 � 0.2 kg for control piglets and 6.5 � 0.1 kg for restricted
(P > 0.10). During the first week post-weaning (pre-restriction period) there were no differences
in weight gain (P = 0.87) or feed intake (P = 0.93) between treatment and control groups
(Table 3). During the restriction period, when treatment piglets were provided with only 65% of
ad libitum intake, weight gain of feed-restricted piglets was significantly lower (P = 0.0001) as
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215 207
Table 2
Observed mutually exclusive behaviour patterns and their definitions
Behaviour Definition
Belly nosing A distinctive, rhythmic up-and-down movement of a piglet’s snout against
the belly of another piglet (Fraser, 1978)
Nosing/chewing
pen-mates
Oral–nasal contact with the body of a pen-mate (distinct from belly
nosing behaviour) (Fraser, 1978)
Nosing/chewing pen Oral–nasal contact with any part of the pen (including floor and side bars)
Fighting Pushing, head-thrusting or chasing directed towards a pen-mate
At feeder Head positioned in or above feeder
Drinking Snout in contact with nipple drinker
Resting Lateral and sternal recumbancy when engaged in no other activity
Active Includes sitting, standing, walking and transitional movements between
behaviour categories
compared to control piglets (Table 3). During the post-restriction period, while feed intake was
similar between treatment and control groups (P = 0.08), treatment piglets had a higher weight
gain (P = 0.0013) compared to control piglets (Table 3). Overall, both weight gain and feed
intake showed significant quadratic trends, indicating that, as expected, their change over time
was affected by treatment (P < 0.05).
3.2. Behaviour
During the pre-restriction period, there were no significant differences in any measured
behaviour pattern between treatment and control piglets.
During the restriction period, treatment had a significant effect on several behaviour patterns
during the observation hours. Feed-restricted piglets spent significantly less time at the feeder
(P < 0.01), more time nosing and chewing the pen (P < 0.02), were more active (P < 0.03) and
had a tendency to spend less time resting (P < 0.07) compared to control piglets (Table 4).
However, percentage of time spent belly nosing (P = 0.13), nosing and chewing the ears and tails
of pen-mates (P = 0.22), fighting (P = 0.52) and drinking (P = 0.16) were not different between
the restricted and control pens (Table 4). Although treatment did not significantly affect mean
percentage of time spent belly nosing, treatment did affect the individual variation in the
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215208
Table 3
The effect of feed restriction on weight gain and feed intake (mean � S.E.M.) during each week of the study
Variable Treatment Control
Weight gain (kg/pig)
Pre-restriction 0.48 � 0.14 0.50 � 0.07
Restriction 0.64 � 0.07a 1.80 � 0.14
Post-restriction 3.12 � 1.47a 2.65 � 1.12
Feed intake (kg/pig)
Pre-restriction 0.80 � 0.08 0.70 � 0.04
Restriction 1.17 � 0.05a 1.82 � 0.08
Post-restriction 3.56 � 0.15 3.85 � 0.10
a Means within a row are significantly different (P < 0.05).
Table 4
The effect of feed restriction on the mean of each behaviour pattern during restriction and post-restriction periods
(mean � S.E.M.)
Behaviour Restriction period (mean % of time engaged
in each behaviour pattern)
Post-restriction period (mean % of time
engaged in each behaviour pattern)
Treatment Control P-value Treatment Control P-value
Belly nosing 3.52 � 0.89 2.12 � 0.53 0.13 1.80 � 0.57 0.49 � 0.20 0.003
Nosing/chewing
pen-mates
2.01 � 0.46 1.28 � 0.23 0.22 2.14 � 0.44 1.43 � 0.16 0.31
Nosing/chewing pen 4.11 � 1.06 1.42 � 0.38 0.01 2.48 � 0.65 1.90 � 0.38 0.66
At feeder 5.27 � 0.96 8.34 � 0.66 0.001 9.31 � 0.50 9.29 � 0.83 0.99
Active 19.0 � 2.34 12.67 � 1.16 0.02 14.41 � 1.49 12.53 � 1.07 0.17
Resting 65.47 � 3.84 73.31 � 1.40 0.07 68.56 � 2.73 73.11 � 1.51 0.03
Fighting 0.03 � 0.03 0.06 � 0.04 0.52 0.20 � 0.10 0.08 � 0.05 0.29
Drinking 0.55 � 0.08 0.78 � 0.13 0.16 1.12 � 0.20 1.14 � 0.21 0.26
performance of the behaviour. Percentage of time spent belly nosing ranged from 0 to 13.5% for
control piglets and from 0 to 17.6% for restricted. Two of the 31 control piglets and 3 out of the 32
restricted piglets were never observed belly nosing during the restriction period. The variance of
belly nosing within pens was greater for feed-restricted piglets compared to control piglets (log
likelihood ratio; P < 0.05). There was no evidence that the variation within pens differed
between the treatments for any other behaviour recorded (log likelihood ratio; P > 0.05).
Orthogonal polynomial contrasts across days indicated significantly different quadratic trends in
time spent at the feeder (Fig. 1), nosing and chewing the pen (Fig. 2) and piglet activity (Fig. 3)
but not belly nosing (Fig. 4), and significantly different linear trends for lying (Fig. 5) between
treatment and control piglets (P < 0.05).
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215 209
Fig. 1. Mean (�S.E.M.) percentage of time piglets spent at the feeder (difference between treatments: during restriction
P < 0.05; linear P > 0.05; quadratic P < 0.05; during post-restriction P > 0.05; linear P > 0.05).
Fig. 2. Mean (�S.E.M.) percentage of time piglets spent nosing and chewing the pen (difference between treatments:
during restriction P < 0.05; linear P > 0.05; quadratic P < 0.05; during post-restriction P > 0.05; linear P > 0.05).
During the post-restriction period, when all piglets were once again fed to appetite, treatment
piglets spent significantly more time belly nosing (P < 0.01) and spent less time resting
(P < 0.03) compared to control piglets (Table 4). During this period there were no significant
differences in nosing and chewing the ears and tails of pen-mates, nosing and chewing the pen,
time spent at the feeder, piglet activity, fighting or drinking (Table 4). Orthogonal polynomial
contrasts across days indicted no significant differences in the change in behaviour over time
between treatment and control groups during this period (P > 0.05).
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215210
Fig. 3. Mean (�S.E.M.) percentage of time piglets spent in activity (difference between treatments: during restriction
P < 0.05; linear P > 0.05; quadratic P < 0.05; during post-restriction P > 0.05; linear P > 0.05).
Fig. 4. Mean (�S.E.M.) percentage of time piglets spent belly nosing (difference between treatments: during restriction
P > 0.05; linear P > 0.05; quadratic P > 0.05; during post-restriction P < 0.05; linear P > 0.05; variance analysis during
restriction: log likelihood ratio P < 0.05).
Pearsons correlations indicated no relationship between weaning weight and belly nosing
during the restriction period or overall (Table 5). However, there were several significant negative
correlations between weight gain and belly nosing. Weight gain during week 2 (restriction
period) was significantly negatively correlated with belly nosing in week 2 when data for all
piglets were combined (P < 0.04). Total weight gain (for the entire trial) was negatively
correlated with overall belly nosing for control pigs (P < 0.01) and for all pigs combined
(P < 0.01) and tended to be related for restricted pigs (P < 0.10). Time spent at the feeder was
not correlated with belly nosing when piglets in each treatment were analyzed separately, but
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215 211
Fig. 5. Mean (�S.E.M.) percentage of time piglets spent resting (difference between treatments: during restriction
P > 0.05; linear P > 0.05; quadratic P > 0.05; during post-restriction P < 0.05; linear P > 0.05).
Table 5
Pearsons correlation coefficients between belly nosing and growth and belly nosing and feeding behaviour during the
week of feed restriction (week 2) and for all 3 weeks combined
Treatment Weaning
weight
Weight
gain week 2
Weight gain
overall
At feeder
week 2
At feeder
overall
Control
(N = 31)
Belly nosing
week 2
�0.04,
P = 0.84
�0.22,
P = 0.24
�0.29,
P = 0.11
Belly nosing
overall
�0.03,
P = 0.85
�0.49,
P < 0.01
�0.22,
P = 0.12
Restricted
(N = 32)
Belly nosing
week 2
�0.18,
P = 0.32
�0.25,
P = 0.17
�0.14,
P > 0.45
Belly nosing
overall
�0.23,
P = 0.20
�0.30,
P = 0.09
�0.17,
P > 0.35
All piglets
(N = 63)
Belly nosing
week 2
�0.09,
P > 0.45
�0.26,
P < 0.04
�0.25,
P = 0.05
Belly nosing
overall
�0.12,
P > 0.34
�0.39,
P < 0.002
�0.21,
P = 0.09
there was a negative relationship between time spent at the feeder in week 2 and belly nosing in
week 2 (P = 0.05) and between time at the feeder and belly nosing overall (P = 0.09).
4. Discussion
Our experiment was designed to investigate a causal relationship between low feed intake
and belly nosing behaviour. Our results suggest that belly nosing was affected by feed
restriction, but there were individual differences in the response. Large individual variation in
the amount of time piglets spend belly nosing has been reported in several studies; some
animals are never observed to perform the behaviour, while others belly nose at relatively high
levels (Straw and Bartlett, 2001; Li and Gonyou, 2002). Piglets with a predisposition to belly
nose direct more post-weaning activity at the bodies of pen-mates while those that perform less
belly nosing direct more oral activity at floors and feeders (Torrey and Widowski, 2006). In the
current study, feed restriction appeared to amplify those individual differences in belly nosing.
Although averages for time spent belly nosing were not statistically different, variance within
pen was greater in feed-restricted piglets compared to those fed ad libitum during the period of
feed restriction, and this was the only behaviour pattern for which this effect on variance was
found. It is possible that reduced feed intake (or hunger) may have exacerbated the
performance of belly nosing in those individuals that already had developed this behaviour
during the first week, but did not elicit it in pigs that were not previously exhibiting the
behaviour. This is supported by the observation that three of the 32 piglets in the restricted
treatment (and 2 of the 31 controls) were never observed belly nosing during the observations
in the period of restriction and one of the restricted pigs was never observed to belly nose on
any day.
Some of the increased variation and lack of statistical difference in belly nosing behaviour
during the restriction period may have been due to the increase in activity levels also observed
during that time. Stimulating activity and time spent nosing and chewing the pen may have
reduced the opportunities or number of recipient piglets available for belly nosing, since most
occurrences of this behaviour are targeted at individuals that are less active (Gonyou et al., 1998;
Straw and Bartlett, 2001). The small sample size combined with large variation may also have
contributed to the lack of statistical significance. The relatively small group sizes of only four
piglets per pen could also have resulted in few or no pigs with a predisposition to belly nose being
allotted to a pen simply by chance, however, belly nosing was observed in at least three piglets in
each of the pens.
The young pig has a tremendous biological growth potential but actual growth rate may be
limited by feed intake after weaning (Le Dividich and Seve, 2001). After weaning, piglets are
known to undergo a short period of low voluntary feed and consume amounts below maintenance
energy requirements (Le Dividich and Seve, 2001). In the second week post-weaning, piglets
weaned at approximately 3 weeks of age gradually increase feed intake and reach a plateau in
metabolizable energy intake per unit of body weight between 14 and 21 days after weaning (Le
Dividich and Seve, 2001). The feed restriction period in our study was introduced in the second
week post-weaning after most piglets had presumably established feeding so that reduced feed
intake was imposed rather than voluntary. Control piglets showed the expected increase in
average growth rates and feed intakes during the second week after weaning. Feed restriction
during this time slowed growth rates, also as expected. During the post-restriction period, when
treatment piglets were once again fed to appetite, both groups consumed feed at similar rates but
previously restricted piglets had higher weight gains.
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215212
Individual piglets that exhibit belly nosing typically have poor growth rates (Fraser, 1978;
Straw and Bartlett, 2001; Torrey and Widowski, 2006) and this relationship was also observed in
the control piglets relying on voluntary intake in this study. Overall weight gain only tended to be
related to belly nosing in treatment pigs, probably because even those piglets that did not perform
very much belly nosing grew more slowly. The relationship between time at the feeder and belly
nosing was only significant when data from all piglets were analyzed together, probably because
of the small sample size. Also, there no relationship between time at the feeder and weight gain
(data not shown). Although time spent at the feeder has been shown to be related to feed intake in
piglets weaned at 28 days or older (Dybkjær, personal communication) a lack of correspondence
between time at the feeder and feed intakes has been observed in several studies involving pigs
weaned at younger ages (Gardner et al., 2001; McGlone and Anderson, 2002; Torrey and
Widowski, 2004). The lack of association between weaning weight and belly nosing is consistent
with what has been observed in other studies (Straw and Bartlett, 2001; also see review by
Widowski, et al., this volume).
The time course for belly nosing in the control group was consistent with what has been
reported in the literature (Gonyou et al., 1998; Torrey and Widowski, 2004). Its onset was
obvious by day 7 post-weaning, peaked during the second week, and then declined by the end of
week 3. However, the time course was somewhat different for treatment piglets. Rather than
steadily declining after peak levels on day 9, the percentage of time spent belly nosing remained
elevated into the post-restriction period resulting in significantly higher belly nosing in restricted
pigs compared to controls in the final week of the trial. This suggests that the effects of restricted
feed on belly nosing are not due to hunger per se, but rather through a discrepancy between
growth and growth potential since there is a lag between time of nutrient intake and weight gain
(Prince et al., 1983; Dunshea, 2003).
Feed restriction affected other aspects of the piglets’ behaviour in several ways. Some behaviour
responses appeared to directly correspond to the period of reduced feed availability while others
only appeared over the course of the week. Feed restriction reduced the amount of time piglets spent
at the feeder, and increased the proportion of time piglets engaged in pen-directed nosing and
chewing behaviour throughout the restriction period. Reduction in time spent at the feeder can be
attributed to the fact that feeders were empty about half of the observation time. Once piglets were
given their ration for the day, they consumed it quickly and did not spend any more time than
needed, at the feeder. However, they did spend more time nosing and chewing the pen, likely caused
by hunger, which elevated feeding motivation and in turn, increased the performance of foraging
related behaviour. Day et al. (1995) found that older pigs feed restricted by 20% of ad libitum intake,
showed an increase in time spent rooting substrate. In our study, nosing and chewing the pen could
be considered a form of rooting behaviour, representing appetitive foraging resulting from an
increase in feeding motivation. Since piglets were hungry, they may have been actively searching
the pen for food. An alternative hypothesis for the increase in pen-directed nosing and chewing
behaviour observed in treatment piglets may be increased frustration as a result of prevention from
engaging in feeding behaviour. Under conditions that restrict the performance of normal feeding
behaviour, animals may direct the behaviour towards alternative stimuli (Lawrence et al., 1993).
Lewis (1999) suggested that frustration associated with the inability of pigs to engage in feeding to
satisfy their hunger, rather than hunger itself, leads to increased oral activity. It has been suggested
that belly nosing may be a behavioural indicator of stress (Dybkjær, 1992; also see Widowski et al.,
2008) and it might be argued that the changes in belly nosing observed in restricted pigs in this
experiment could be caused by the ‘stress’ of hunger or reduced feed intake rather than a more
direct effect of nutrient or growth status. However, belly nosing remained significantly higher in
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215 213
restricted pigs during the early post-restriction period when pigs were again eating to appetite and
time at the feeder and nosing the pen were similar for control and treatment pigs.
During the period of restriction, feed-restricted piglets became more active and rested less than
ad libitum fed piglets but these differences only became apparent over time. Examination of diurnal
feeding patterns in swine has indicated that meal patterns typically show a peak in the morning, one
or two subsequent peaks later in the day, and modest intake during the night (Fraser, 1984). In our
experiment, treatment piglets were fed once a day (09:00 h). Observations were conducted during
the 3 h prior to feeding when feeders were usually empty and early in the afternoon when food was
usually available. Piglets may have learned to anticipate feeding time, which altered their activity
patterns over the course of theweek. Similar trends in behaviour were found in a study conducted by
Hammel and Hurnik (1987) who found that meal-fed gilts and cattle rapidly learned the feeding
schedule and were more active (in terms of time spent walking and standing) and spent less time
resting compared to ad libitum fed animals. Day et al. (1995) also found that pigs restricted by 20%
of ad libitum intake spent much less time lying compared to controls but this was observed during
tests for foraging motivation when the pigs were outside of their home pens.
During the period of feed restriction, treatment did not have an effect on nosing and chewing
the ears and tails of pen-mates or fighting. Increased restlessness is often accompanied by an
increase in oral activity directed at pen-mates (Fraser, 1987). However, feed restriction in our
study did not have an influence on nosing and chewing the ears and tails of pen-mates, possibly
because piglets may have been more motivated to engage in foraging related behaviour directed
at the pen, rather than nose and chew their pen-mates. It was surprising that feed restriction did
not have an effect on aggressive behaviour. This may have been because there was little
competition at the feeder since each pen was equipped with a four-placed feeder, allowing all
four pigs to feed simultaneously when feed was available.
In summary, piglets subjected to feed restriction during the second week post-weaning
reduced resting and increased general activity and nosing/chewing the pen compared to piglets
fed ad libitum. Feed restriction increased the variation in and altered the time course for belly
nosing, and increased the occurrence of belly nosing in the week after feed restriction, but it did
not elicit the behaviour in all pigs that were feed restricted. This suggests that there is a causal
relationship between low feed intake and belly nosing, but only in individual piglets with a
predisposition for the behaviour.
Acknowledgements
This research was funded by grants from the Natural Sciences and Engineering Research
Council (NSERC) and the Ontario Ministry of Agriculture and Food. We thank the staff of the
Arkell Swine Research Station for their assistance. Special thanks go to Emily Toth for her many
hours of service in the barn.
References
Blackshaw, J.K., 1981. Some behavioural deviations in weaned domestic pigs: persistent inguinal nose thrusting, and tail
and ear biting. Anim. Prod. 33, 325–332.
Brooks, P.H., Moran, C.A., Beal, J.D., Demeckova, V., Campbell, A., 2001. Liquid feeding for the young piglet. In: Varley,
M.A., Wiseman, J. (Eds.), The Weaner Pig: Nutrition and Management. CABI Publishing, Wallingford, Oxon, UK,
pp. 153–178.
Bruni, A., 2004. Investigations into the effect of cold draft and feed restriction on the behaviour of newly weaned piglets.
M.Sc. Thesis, University of Guelph, Guelph, Ontario.
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215214
Day, J.E.L., Kyriazakis, I., Lawrence, A.B., 1995. The effect of food deprivation on the expression of foraging and
exploratory behaviour in the growing pig. Appl. Anim. Behav. Sci. 42, 193–206.
Dunshea, F.R., 2003. Metabolic and endocrine changes around weaning. In: Pluske, J.R., Le Dividich, J., Verstegan, M.
(Eds.), Weaning the Pig—Concepts and Consequences. Waageningen Academic Publishers, The Netherlands, pp.
61–78.
Dybkjær, L., 1992. The identification of behavioural indicators of ‘stress’ in early weaned pigs. Appl. Anim. Behav. Sci.
35, 135–147.
Fraser, D., 1978. Observations on the behavioural development of suckling and early-weaned piglets during the first six
weeks after birth. Anim. Behav. 26, 22–30.
Fraser, D., 1984. The role of behaviour in swine production: a review of research. Appl. Anim. Ethol. 11, 317–339.
Fraser, D., 1987. Mineral-deficient diets and the pig’s attraction to blood: implications for tail-biting. Can. J. Anim. Sci.
67, 909–918.
Gardner, J.M., de Lange, C.E.M., Widowski, T.M., 2001. Belly nosing in early-weaned pigs is not influenced by diet
quality or the presence of milk in the diet. J. Anim. Sci. 79, 73–80.
Gonyou, H.W., Beltranena, E., Whittington, D.L., Patience, J.F., 1998. Behaviour of pigs weaned at 12 and 21 days of age
from weaning to market. Can. J. Anim. Sci. 78, 517–523.
Hammel, K.L., Hurnik, J.F., 1987. Fixed-interval conditioned feed intake in swine cattle. Can. J. Anim. Sci. 67, 319–325.
Jensen, P., 1995. The weaning process of free-ranging domestic pigs: within- and between-litter variations. Ethology 100,
14–25.
Jensen, P., Recen, B., 1989. When to wean: Observations from free-ranging domestic pigs. Appl. Anim. Behav. Sci. 23,
49–60.
Kuehl, R.O., 1994. Statistical Principles of Research Design and Analysis. Duxbury Press, Belmont, California.
Lawrence, A.B., Terlouw, E.M.C., Kyriazakis, I., 1993. The behavioural effects of undernutrition in confined farm
animals. Proc. Nutr. Soc. 52, 219–229.
Le Dividich, J., Seve, B., 2001. Energy requirements of the young pig. In: Varley, M.A., Wiseman, J. (Eds.), The Weaner
Pig: Nutrition and Management. CABI Publishing, Wallingford, Oxon, UK, pp. 17–44.
Lewis, N., 1999. Frustration of goal-directed behaviour in swine. Appl. Anim. Behav. Sci. 64, 19–29.
Li, Y.Z., Gonyou, H.W., 2002. Analysis of belly nosing and associated behaviour among pigs weaned at 12–14 days of
age. Appl. Anim. Behav. Sci. 77, 285–294.
McGlone, J.J., Anderson, D.L., 2002. Synthetic maternal pheromone stimulates feeding behaviour and weight gain in
weaned pigs. J. Anim. Sci. 80, 3179–3183.
Metz, J.H.M., Gonyou, H.W., 1990. Effect of age and housing conditions on the behavioural and haemolytic reaction of
pigs to weaning. Appl. Anim. Behav. Sci. 27, 299–309.
Newberry, R.C., Wood-Gush, D.G.M., 1985. The suckling behaviour of domestic pigs in a semi-natural environment.
Behaviour 95, 11–25.
Orgeur, P., Rigaud, V., Le Dividich, J., 2003. Liquid feeding to improve welfare and performance of piglets at weaning.
In: Proceedings of the 37th International Congress of the International Society for Applied Ethology. Abano Terme,
Italy, p. 229.
Pluske, J.R., Williams, I.H., Aherne, F.X., 1995. Nutrition of the neonatal pig. In: Varley, M.A. (Ed.), The Neonatal Pig:
Development and Survival. CAB Int, Wallingford, UK, pp. 187–235.
Prince, T.J., Jungst, S.B., Kuhlers, D.L., 1983. Compensatory responses to short-term feed restriction during the growing
period in swine. J. Anim. Sci. 56, 846–851.
Rau, J.A., 2002. Behaviour and performance of early-weaned pigs: effects of trough-anchored blind teats and liquid food.
M.Sc. Thesis, University of Guelph, Canada, pp. 81–83.
Steel, R.G.D., Torrie, J.H., Dickey, D.A., 1997. Analysis of variance: split-plot designs and analysis. In: Principles and
Procedures of Statistics a Biometrical Approach, 3rd ed. McGraw-Hill, Inc., New York, pp. 412–413.
Straw, B.E., Bartlett, P., 2001. Flank or belly nosing in weaned pigs. J. Swine Health. Prod. 9, 19–23.
Torrey, S., Widowski, T.M., 2004. Effect of drinker type and sound stimuli on early-weaned pig performance and
behaviour. J. Anim. Sci. 82, 2105–2114.
Torrey, S., Widowski, T.M., 2006. Is belly nosing redirected sucking behaviour? Appl. Anim. Behav. Sci. 101, 288–304.
Weary, D.M., Appleby, M.C., Fraser, D., 1999. Responses of piglets to early separation from the sow. Appl. Anim. Behav.
Sci. 63, 289–300.
Widowski, T.M., Torrey, S., Bench, C.J., Gonyou, H.W., 2008. Development of ingestive behaviour and the relationship to
belly nosing in early-weaned piglets. Appl. Anim. Behav. Sci. 110, 110–129.
Worobec, E.K., Duncan, I.J.H., Widowski, T.M., 1999. The effects of weaning at 7, 14 and 28 days on pig behaviour. Appl.
Anim. Behav. Sci. 62, 173–182.
A. Bruni et al. / Applied Animal Behaviour Science 110 (2008) 203–215 215