- Email: [email protected]
Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance?
G Model
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
Applied Animal Behaviour Science xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Applied Animal Behaviour Science journal homepage: www.elsevier.com/locate/applanim
Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Camilla Munsterhjelm ∗ , Mari Heinonen, Anna Valros University of Helsinki, Faculty of Veterinary Medicine, Department of Production Animal Medicine, PB 57, FIN-00014 Finland
a r t i c l e
i n f o
Article history: Received 18 April 2016 Received in revised form 18 August 2016 Accepted 25 August 2016 Available online xxx Keywords: Pig Social behaviour Tail biting Tail-in-mouth
a b s t r a c t This study aimed to identify characteristics of pigs performing tail-in-mouth behaviour (TIM; P, n = 34), their recipients (R, n = 23) and neutral penmates (N, n = 31) at two occasions, the first being at weaning (4 weeks of age) before TIM was observed in the pen and the second being at 9 weeks of age when TIM had emerged, but no clinical tail lesions were observed. The groups (n = 22) were formed by siblings, two gilts and two castrates. Behaviour was analysed as 24-h time budgets and continuously sampled during 30 min of the active part of the day. Category (P, R, N) effects were analysed at individual and (directed) dyad level. P was born significantly smaller than R, but the difference had disappeared at 4 weeks. Growth or sex distribution did not differ between categories. Category differences in performed behaviour were evident at 4 weeks of age, when P showed more overall activity and environmental exploration as compared to R, as well as more bouts of tail-nosing than N. Different aspects of behaviour changed in the different categories between 4 and 9 weeks of age. In P social activity increased significantly and went from no preference at 4 weeks to a significant preference for social actions for R over N at 9 weeks. N was socially passive at 9 weeks while receiving more social behaviour than the other categories. These differences in behaviour suggest that the categories represented different phenotypes of pigs. © 2016 Published by Elsevier B.V.
1. Introduction Tail biting is a behavioural disturbance of growing pigs associated with an unsatisfactory environment. It is a significant economical and welfare problem in modern pig production. Although a number of farm-level predisposing factors have been established their effects appear to be interacting heavily with individual characteristics of the pigs, making outbreaks unpredictable (Taylor et al., 2010). The identification of measures able to prevent tail biting requires a better understanding of the initial stages of a tail biting outbreak, especially of the contribution of individual pigs (Edwards, 2006). Gentle tail manipulation or tail-in-mouth behaviour (TIM), causing no obvious injuries to the tail, is considered to be a behavioural precursor to damaging tail biting (Feddes et al., 1993; Feddes and Fraser, 1994). According to the review by Taylor et al. (2010) this two-stage type of tail biting appears to account for the majority of tail biting in commercial pig production and also of the tail-biting cases described in scientific papers.
∗ Corresponding author. E-mail address: camilla.musterhjelm@helsinki.fi (C. Munsterhjelm).
The progression from the pre-injury stage, characterized by TIM, to the injury stage in two-stage tail biting takes place can occur rapidly (Fraser, 1987). TIM may also be performed without subsequent injurious biting (Schrøder-Petersen et al., 2003a). There are diverging views among scientist on the categorization of TIM as normal or abnormal behaviour. TIM has been suggested to be abnormal upon observations that only one or a few individuals in a group have been performing the behaviour (Blackshaw, 1981; Beattie et al., 2005), whereas others have reported lowfrequency TIM performed by all or a majority of pigs in a population and thereby classified it as a normal activity (Feddes et al., 1993; Schrøder-Petersen et al., 2003a). An obvious reason for this discrepancy, in addition to differences in observational strategy, may be variation in TIM frequency and intensity between populations and/or occasions. Schrøder-Petersen et al. (2003a) suggested that TIM behaviour of low frequency and intensity is normal behaviour that may increase to abnormal levels under certain circumstances, to eventually become actual injurious tail biting. TIM is thought to be an expression of the motivation to explore both the environment and in a social context (Schrøder-Petersen et al., 2003a,b; Simonsen, 1995; Newberry and Wood-Gush, 1988). The behaviour has been reported to decrease in response to straw provision, increase with age at least in weaners, and typically be
http://dx.doi.org/10.1016/j.applanim.2016.08.013 0168-1591/© 2016 Published by Elsevier B.V.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
2
performed by females and received by males (Schrøder-Petersen et al., 2003a; Simonsen, 1995). The aim of this study was to identify individual differences in size, growth and behaviour in TIM-performers, TIM-receivers and neutral penmates before TIM behaviour was observed in the pen and when the behaviour had begun or was about to begin in the group. 2. Materials and methods 2.1. Animals and husbandry The animals (n = 88 pigs in 22 groups) were chosen from a larger experiment (252 pigs in 63 groups) investigating different enrichment regimes, conducted over 13 batches in 2004–2005 at the experimental pig farm of MTT Agrifood Research in southern Finland. The animals represented pure Finnish Yorkshire or Finnish Landrace and their crosses with or without ¼ Duroc. Housing and husbandry will be described only shortly, as details are given elsewhere (Munsterhjelm et al., 2009). The animals were housed in pens with partly slatted floors at all times. Sows farrowed and nursed in pens of 5.4 m2 without crates, equipped with a heated creep area. During the first three days litter size was standardized to no more than 12 piglets by cross-fostering onto non-experimental sows. Piglets were subjected to teeth clipping and foreleg taping at one day of age as well as an intramuscular iron injection (1 ml Ursoferran® vet., Serumwerk Bernburg AG, Germany) and castration at day three. Tails were left intact according to Finnish legislation. The piglets were weaned at an average age of 29.5 days (range 26–36). At weaning groups of weight-matched siblings were formed including two gilts and two castrates. In the nursery (4–9 weeks of age) animals were allowed a stocking density of 0.7 m2 per animal in pens equipped with a single-space ad libitum feeder (Groba Diplomat AS, Skælskør, Denmark) and two drinking nipples. Climate was controlled according to age. In the fattening unit pigs were housed with a stocking density of 1.2 m2 per pig and fed from a trough. The animals were provided enrichment according to one of six schemes being investigated in the original experiment. Pens either were (=E) or were not (=0) bedded with wood shavings and chopped straw thinly enough to show the pen floor during suckling, nursery and fattening stages. Accordingly, enrichment regimes were 1) 000 (referring to 0 in the suckling stage, 0 in the nursery and 0 in the fattening stage; n = 8 pigs in 2 groups); 2) E00 (n = 12/3); 3) EE0 (n = 20/5); 4) 00E (n = 8/2); 5) 0EE (n = 24/6) and 6) EEE (n = 16/4). This experimental procedure was approved by the ethical committee of MTT Agrifood Research Finland (permission number SIK 6/04, 2004-08-26). Animals had full artificial lighting between 07:00 and 16:00 h and dimmed lighting enabling night-time videotaping for the remainder of the time. They were fed standard pelleted pig diets without additives formulated to fulfill current Finnish feeding recommendations and slaughtered at 20–22 weeks of age. 2.2. Experimental design TIM was defined as manipulation of the tail irrespective of the reaction by the recipient, but without causing clinical damage. All groups fulfilling three inclusion criteria were selected for the study: 1) no TIM behaviour was observed on video tape during the first day after weaning (age 4 weeks), 2) TIM behaviour was observed on video tape on the last day in the nursery and/or on the first day in the fattening unit (age 9–10 weeks), and 3) no clinical tail wounds were observed from birth to 10 weeks of age. TIM behaviour was defined
as taking a tail in the mouth, regardless of the reaction by the other pig. The animals were categorized as TIM performers (P, n = 34 pigs in 22 groups), TIM receivers (R, n = 23 pigs in 19 groups) and neutral pigs (N, n = 31 pigs in 19 groups). One occasion of TIM was enough to categorize an animal as P or R. P observed to receive TIM (n = 7 pigs in 5 groups) were included in the P category. Characteristics of the experimental groups are summarized in Table 1. The animals were weighed at birth, at weaning at 4 weeks of age, and at the middle and end of the nursery period (age 7 and 9 weeks). Behaviour was monitored with a time-lapse black and white video recording device for 24 h starting at 07:00 h on the first and last day in the nursery (age 4 and 9 weeks), on the first day in the fattening unit (age 10 weeks) and at 14 and 18 weeks of age. The cameras were connected to a video recording device through a multiplexer. The animals were marked individually with permanent drawing ink the day before each monitoring. Social and exploratory behaviours were extracted by continuous focal sampling at 13:00–13:10, 14:00–14:10 and 15:00–15:10 h according to the ethogram given in Table 2. The time slots were chosen in order to represent the most active part of the day according to Beattie (1994). A bout was considered to end if the animal stopped performing a behaviour for two seconds or more. Time budgets were investigated by instantaneous scan sampling with a 10 min interval for the whole 24 h according to the ethogram in Table 3. Continuously sampled behaviours will be referred to as “daytime behaviours” to be distinguished from 24-h time budgets. All behaviours were recorded on individual level. The recipient was recorded in addition to the performer for all social behaviours. 2.3. Statistical analyses The SPSS for Windows statistics package, version 22.0 (IBM Corp., Armonk, NY, USA) was utilized for statistical analyses. The experimental unit was the individual. Growth was expressed as the relative change in body weight over a given time span. The distribution of sex within each category was evaluated against a balanced distribution using the Chi-squared test. Scan sampled behavioural variables were analysed as the number of scans in the behaviour over 24 h. Continuous behavioural variables were analysed as total time in the behaviour and as the number of bouts during 30 min of observation. All performed social behaviours (nosing, tail nosing, ear biting, TIM, fighting, pushing and submissive behaviours) were summarized in the variable “total social behaviours”, and all received social behaviours except “submissive” were summarized in “received total social behaviours”. Being the recipient of a submissive action was, by definition, a response to social activity and thus not considered a true received behaviour. Daytime performed total social behaviour was subjected to dyad-level analysis in order to investigate category-level preferences for social actions. A dyad is a unit including a performer and a receiver. The dyads were considered as directed, meaning that A directing behaviour towards B was a different dyad than B directing behaviour towards A. Hereby the data included nine different dyads (P → R, R → P, P → N, N → P, N → R, R → N, P → P, N → N, R → R), however, the dyad P → P was excluded from all analyses due to a low n. Normality was investigated for all continuous variables using histograms, Q–Q plots, Kolmogorov-Smirnov and Shapiro-Wilk tests; and attempts made to transform the non-normal ones. Category and dyad effects on variables that could not be normalized were determined using the Pair-Wise Related-Samples Wilcoxon Test, or, if the within-pair difference was non-symmetrically distributed, the Sign Test. The group effect was counted for by considering the categories or dyads within a group as the repeated feature. In groups with several animals representing the same category or dyad their median was used in the analyses. The tests were
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
3
Table 1 Characteristics of the groups included in the study. Group1
1 2 3 410 5 6 710 8 9 10 11 12 13 149,10 159,10 16 17 18 1910 20 21 2210
Sex2
Bedding3 in suckling/weaning/ fattening stage
TIM performers
TIM receivers
Neutral pigs
F F F F F F F FBV FBV F FBV FM FBV MBV FBV MM FFM FM FM FM M M M M M M
F F F F FMM M MM
M MM MM MM
M M
FM F MM FM FM
F F M M F F F FM M M
M M F F FM FM FM F FF FF
1
0/1/1 0/1/1 1/0/0 1/1/1 0/0/0 0/1/1 1/1/0 1/1/0 0/1/1 0/1/1 1/0/0 0/0/1 1/1/0 1/1/0 1/0/0 0/1/1 1/1/0 1/1/1 1/1/1 1/0/0 0/0/1 1/1/1
Age (weeks) of continued TIM or tail biting in the group4
17L , 20L 14
14L , 16L 20 20 14, 20 14 14 14 20
14 14
10
TIM was observed on the last day in the nursery (9 weeks of age) in groups without a superscript, TIM was observed on the first day in the fattening unit (10 weeks of age), TIM was observed at both occasions. 2 F = female, M = castrate, BV indicates that the performer also received TIM 3 0 = no bedding 1 = moderate amount of wood shavings and chopped straw. 4)L Indicates that tail wounds were present, the lack of superscript that TIM was detected on video (recording at 14 and 18 weeks) in the absence of tail wounds. 9,10
Table 2 Definitions of behaviours of weaned piglets for continuous sampling.
Table 3 Definitions of behaviours of weaned piglets for scan sampling.
Behaviour
Description
Behaviour
Description
Nosing
Nosing, sniffing or touching penmate with the snout in any area other than the tail, distance from snout to the skin less than 5 cm
Performed behaviours Inactive
Performing no activity
Exploring
Tail-nosing
Nosing, sniffing or touching the tail of penmate with the snout, distance from snout to the tail less than 5 cm
TIMa
Taking penmate’s tail in the mouth irrespectively of the reaction by the receiver
Nosing or manipulating any part of the pen or bedding except for the drinking nipple, or inside the feeder or trough. Observing, nosing or manipulating any object outside the pen with the snout outside the pen.
Ear biting
Taking penmate’s ear in the mouth
Performing social behaviour
Pushing
Pushing penmate with any part of the body in order to displace him, no biting
Any active behaviour involving nosing (distance from snout to skin 0–5 cm), biting, ramming, lifting or otherwise manipulating penmate. Pushing penmate with any part of the body. Mounting with both front hoofs on or resting on sternum on penmate’s back.
Fighting
Mutual pushing parallel or perpendicular, ramming or pushing penmate with the head accompanied by biting, lifting penmate with the snout
Other activity
Any other activity
Missing Received behaviours Receiving social behaviour
Unidentified activity or pig missing Being the recipient of behaviours included in Performing social behaviour irrespective of own actions.
Submissive
Being the recipient of Pushing or Fighting, responding passively or by fleeing, not fighting back
Exploring
Nosing or manipulating any part of the pen or bedding except for the drinking nipple, or inside the feeder or trough. Observing, nosing or manipulating any object outside the pen with the snout outside the pen.
Other behaviour
Any other behaviour
Missing
Unidentified activity or pig missing
a
Tail-in-mouth behaviour.
applied pair-wise due to a large number of groups with one missing category or missing dyads. Dyad-level comparisons were designed to identify category-level preferences for social actions (e.g. B → V was compared to B → C) and to compare reciprocity (e. g. B → V was compared to V → B).
Category effects on normal(ized) variables were analysed using the Linear Mixed Models (LMM) feature in SPSS. Batch- and grouplevel clustering were addressed by allowing random intercept variation of group nested within batch. Category and enrichment status (yes/no) in weaning and/or suckling stage, as appropriate, were included as group-level, and sex as individual-level fixed factors. Factors were removed according to a backward step-wise elimination strategy. Variables with P ≤ 0.1 were included in the model if a good model fit did not require otherwise. A model was accepted if homoscedasticity and sound distribution of residuals were present. If a significant (P < 0.05) category effect was detected pair-wise comparisons were performed using a Bonferroni correction. Subsequently, the variables analysed using LMM were the square root of birth weight reduced by a constant of 4; body weight
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
4
Table 4 Characteristics of TIM performers, receivers and neutral pigs given as an average ± SD.
a
Characteristic
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Birth weight (kg) Weight, 5 weeks (kg) Weight, 7 weeks (kg) Weight, 9 weeks (kg) Females (n) Castrates (n)
1.8 ± 0.3a 10.1 ± 1.8 16.5 ± 4.0 25.5 ± 5.5 19 15
2.0 ± 0.4a 10.0 ± 1.6 15.7 ± 2.7 24.8 ± 4.1 11 12
1.8 ± 0.4 9.8 ± 1.9 15.8 ± 3.5 24.4 ± 5.4 14 17
indicates a significant category effect (P < 0.05) in pair-wise comparisons.
at 4, 7, and 9 weeks, growth from 0 to 4 and 4–7 weeks and the square root of growth from 7 to 9 weeks. LMM was also applied to the following variables describing continuously sampled exploration: the duration at 4 weeks of age, the logarithm of the number of bouts at 4 weeks of age with an added constant of 1, the square root of the duration and the number of bouts at 9 weeks of age. In order to compare the development of behaviour from 4 to 9 weeks of age between the categories and the dyads age effects on behaviour were analysed separately for each category and dyad. All behaviours were anlysed using Pair-Wise RelatedSamples Wilcoxon Tests (PWRSW) with age (9 vs 4 weeks) as the repeated effect, as described above. Results are given as average ± SD for between-pair differences with values for the Z-statistic for PWRSW tests. No statistic exists for the Sign Test. 3. Results 3.1. Sex and performance The distribution of sex did not differ from balanced (X2 = 83.04, P = 0.6, n = 88, Chi-squared test). Birth weight was higher in R than P (LMM for category effect F2,35 = 3.97, P = 0.03, pair-wise Bonferronicorrected comparison F2,44 = 3.96, P = 0.03, Table 4.). Body weight was not affected by category at 4, 7 or 9 weeks of age, nor was growth between succeeding time points. 3.2. Category effects on time budgets At 4 weeks of age R was inactive in a larger number of scans than P (within-pair average difference 3.0 ± 6.1 [SD] scans, Z = −2.25, P = 0.02, n = 19 pairs of pens, Pair-Wise Related-Samples Wilcoxon Test). At 9 weeks of age P was inactive in a larger number of scans than N (3.2 ± 6.6 scans, Z = −2.11, P = 0.04, n = 19). 3.3. Category and dyad effects on behaviour during the active parts of the day at 4 weeks of age Analyses without other notation are carried out using the Pair-Wise Related-Samples Wilcoxon Test, and between-pair differences are given as average ± SD. The time spent in exploration differed between category (LMM F2,29 = 3.77, P = 0.04, Table 6) at 4 weeks of age. Controlling for suckling stage enrichment (F1,19 = 4.52, P = 0.05), P was predicted to explore significantly more (9 min 21 s out of the total 30 min of observation, calculated for a non-bedded suckling pen) than R (7 min 20 s, F2,41 = 3.77, P = 0.046 in pair-wise Bonferroni-corrected comparison). P performed a larger number of bouts of tail-nosing behaviour than N (1.1 ± 5.1, Z = −2.33, P = 0.02, n = 17 pairs of pens). R tended to perform a larger number of bouts of pushing than N (Z = −1.90, P = 0.06, n = 14). Nosing, ear biting, fighting or submissive behaviours did not differ between categories at 4 weeks of age, nor did total social behaviours analysed at dyadic level (Table 7). N received more total social behaviours than P (2.5 ± 4.5 bouts and
18 ± 76 s, Z = −2.10, P = 0.04 for both comparisons, n = 16, Table 8), including more tail nosing (3.6 ± 8.9 s, P = 0.02) and a tendency for more bouts of received nosing (Z = −1.92, P = 0.06). R tended to receive more bouts of pushing than P (Z = −1.75, P = 0.08, n = 17). 3.4. Category effects on behaviour during the active parts of the day at 9 weeks of age Analyses without other notation are carried out using the PairWise Related-Samples Wilcoxon Test. The between-pair difference is given as average ± SD. R performed the least amount of bouts of social behaviours with significant differences both to P (−6.0 ± 8.4 bouts in R as compared to P, Z = −1.97, P = 0.049, n = 16 pairs of pens) and to N (−2.3 ± 3.7, Z = −2.19, P = 0.03, n = 13). The difference in social activity between R and P included tendencies for nosing bouts (Z = −1.92, P = 0.06) and tail-nosing bouts (P = 0.07, Sign test). The difference between R and N included nosing (−2.1 ± 3.3 bouts in R as compared to N, Z = −2.10, P = 0.04 and Z = −1.87, P = 0.06 for duration, n = 13) and tail-nosing (−2.0 ± 4.5 bouts, P = 0.008, Sign test, n = 13). Social activity between P and N differed only for bouts of tail-nosing, which P tended to perform more frequently (Z = −1.85, P = 0.07, n = 16), and submissive behaviours that tended to be performed for a longer time by N than P (Z = −1.75, P = 0.08). Ear biting, fighting and pushing did not differ between categories at 9 weeks og age. R and N did not, by definition, perform any TIM. The difference in TIM activity as both duration and bouts was significant for comparisons of P to R and to N (p = 0.01–0.02, n = 13–16 pairs). In dyad-level analyses on daytime total social behaviour at 9 weeks P directed more attention towards R than towards N (+43 ± 68 s, Z = −2.04, P = 0.04; and 3.0 ± 4.0 bouts, Z = −2.19, P = 0.03, n = 13 pairs of pens). R directed more social behaviours towards P than towards N (0.6 ± 1.5 bouts, Z = −1.97, P = 0.03 and Z = −1.95, P = 0.05 for duration, n = 13). N tended to direct social behaviour for a longer total time towards P than towards R (Z = −1.78, P = 0.08, n = 11). Looking at reciprocity at 9 weeks of age N directed more social behaviour towards R than the other way around (22 ± 35 s, Z = −2.70, P = 0.007 and 1.5 ± 3.6 bouts, Z = −2.20, P = 0.03, n = 12) and P directed more bouts towards R than vice versa (6.1 ± 6.4 bouts, Z = −2.49, P = 0.01, n = 16). N received the least amount of daytime total social behaviours of all categories of pigs with a significant difference as compared to both P (−2.1 ± 4.6 bouts, P = 0.02 and −83 ± 163 s, Z = 2.05, P = 0.04, n = 15 pairs of pens) and R (−3.7 ± 5.7 bouts, Z = −2.10, P = 0.02, n = 16, Table 8.). R tended to receive more bouts of total social behaviour than P (Z = −1.92, P = 0.06, n = 16) and received significantly more TIM (11 ± 17 s and +1.1 ± 1.5 bouts, P = 0.002 for both comparisons). N received less tail sniffing than R (−14 ± 14 s, P = 0.02 and −1.5 ± 1.5 bouts, Z = −2.37, P = 0.02). N was omitted from analyses on TIM as they by definition received none. 3.5. Age effects on behaviour Analyses on age effects without other notation are carried out using the Pair-Wise Related-Samples Wilcoxon Test. The magnitude of change from 4 to 9 weeks of age is given as the within-individual (or unique dyad) average ± SD. The number of 24-h scans in exploration was unaffected by age, whereas many social behaviours changed significantly (Table 5.). The number of inactive scans decreased in all categories (P: −11 ± 26 scans, Z = −2.57, P = 0.01, n = 34 individuals; R: −16 ± 12, Z = −3.33, P = 0.001, n = 23; N: −15 ± 14, Z = −2.97, P = 0.003, n = 31), while other behaviours increased (P: 5 ± 7, Z = −3.12, P = 0.002; R: 5 ± 7, Z = −2.50, P = 0.03, N: 4 ± 7, Z = −2.70, P = 0.007). Performed social behaviour changed only in R (2 ± 4 scans, Z = −2.15, P = 0.03), and received social behaviour only in P (2 ± 3, Z = −3.10, P = 0.002).
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
5
Table 5 24-h time budgets given as the median (min-max) number of scans. Behaviour
Age (weeks)
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Inactive
4 9
111 (78–126)*a 105 (27–116)*a
112 (90–128)*a 100 (33–115)*
110 (86–130)* 102 (27–119)*a
Exploration
4 9
15 (6–40) 12 (4–28)
14 (5–30) 13 (6–28)
13 (7–41) 13 (6–28)
Other
4 9
9 (2–23)* 15 (6–25)*
8 (4–18)* 12 (4–24)*
9 (3–21)* 14 (3–29)*
Performed social
4 9
6 (0–16) 4 (0–14)
4 (0–14)* 7 (1–16)*
5 (1–16) 5 (0–13)
Received social
4 9
1 (0–7)A * 3 (0–12)*
3 (0–7)A 2 (0–8)
2 (0–7) 1 (0–9)
An asterisk (*) indicates a significant within-category age effect (P < 0.05) for a behaviour. Similar lower-case letter superscripts on a row indicate significant (P < 0.05) category effects and upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons.
Table 6 Continuously sampled behaviours of weaned piglets given as median (min-max) for the duration in seconds and for the number of bouts during 30 min of observation during the active part of the day. Behaviour
Age (weeks)
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Exploration, duration
4 9
639 (26–1228)*a 347 (45–840)*
536 (0–928)a 267 (12–837)
462 (73–940) 477 (0–1052)
Exploration, bouts
4 9
16 (1–84) 14 (1–30)
15 (0–62)(*) 10 (1–22)(*)
11 (5–99) 11 (0–35)
Nosing, duration
4 9
36 (0–318) 69 (0–347)
24 (0–208) 23 (0–135)A
29 (0–337)* 33 (0–368)*A
Nosing, bouts
4 9
5 (0–37) 6 (0–22)A
6 (0–24) 4 (0–19)Ab
3 (0–32) 2 (0–14)b
Tail-nosing, duration
4 9
0 (0–57)* 13 (0–55)*
0 (0–100) 0 (0–126)
0 (0–53)* 6 (0–105)*
Tail-nosing, bouts
4 9
0 (0–6)*a 1 (0–9)*AB
0 (0–7) 0 (0–5)Ac
0 (0–3)*a 1 (0–8)*Bc
Ear biting, duration
4 9
0 (0–0)* 0 (0–164)*
0 (0–5) 0 (0–33)
0 (0–6)(*) 0 (0–21)(*)
Ear biting, bouts
4 9
0 (0–0)* 0 (0–4)*
0 (0–1) 0 (0–2)
0 (0–1)(*) 0 (0–3)(*)
TIM1 , duration
4 9
0 (0–0)* 3 (0–64)*ab
0 (0–0) 0 (0–0)a
0 (0–0) 0 (0–0)b
TIM1 , bouts
4 9
0 (0–0)* 1 (0–5)*ab
0 (0–0) 0 (0–0)a
0 (0–0) 0 (0–0)b
Fighting, duration
4 9
0 (0–39) 6 (0–129)a
0 (0–108) 0 (0–70)a
0 (0–86) 0 (0–265)
Fighting, bouts
4 9
0 (0–5)* 1 (0–5)*
0 (0–4) 0 (0–3)
0 (0–7) 0 (0–6)
Pushing, duration
4 9
0 (0–23) 0 (0–72)
0 (0–18) 0 (0–27)
0 (0–74) 0 (0–100)
Pushing, bouts
4 9
0 (0–4) 0 (0–6)
0 (0–5)A 0 (0–3)
0 (0–3)A 0 (0–3)
Submissive, duration
4 9
0 (0–49)(*) 0 (0–4)(*)A
0 (0–1) 0 (0–36)
0 (0–31)(*) 0 (0–19)(*)A
Submissive, bouts
4 9
0 (0–1)(*) 0 (0–1)(*)
0 (0–1) 0 (0–3)
0 (0–2)(*) 0 (0–2)(*)
Total social2 , duration
4 9
49 (0–359)* 159 (0–395)*
38 (0–438) 62 (0–248)
47 (0–260)* 103 (0–514)*
Total social2 , bouts
4 9
6 (0–41) 11 (0–28)a
5 (0–38) 5 (0–20)ab
6 (0–36) 7 (0–32)b
An asterisk * indicates a significant (P < 0.05), an asterisk in brackets (*) a non-significant (0.05 ≤ P < 0.1) age effect within a category and a behaviour. Similar lower-case letter superscripts on a row indicate significant category effects (p < 0.05), upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons. 1 Tail-in-mouth behaviour 2 Total social behaviours include nosing, tail-nosing, ear biting, TIM, fighting, pushing and submissive behaviours.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
6
Table 7 Dyad level social behaviour of weaned piglets given as median (min-max) of the duration in seconds or number of bouts obtained from 30 min of continuous behavioural observation. Unit
Age (weeks)
P → P1 (n = 20/8)2
P→N (n = 31/15)
P→R (n = 30/25)
N→P (n = 32/19)
N→N (n = 22/10)
N→R n = 21/16)
R→P (n = 24/19)
R→N (n = 18/16)
Duration Duration
4 9
28 (0–130) 26 (0–325)
4 9
2 (0–12) 3 (0–8)
22(0–160)(*) 49 (0–218)(*)a 2 (0–15)(*) 6 (0–16)(*)a
8 (0–79)* 50 (0–321)*A 2 (0–8)A 3 (0–15)
8 (0–74)(*) 21 (0–248)(*) 1 (0–12) 2 (0–7)
6 (0–119) 22 (0–192)Ab 2 (0–13) 3 (0–6)b
18 (0–107) 11 (0–227)C 2 (0–13) 2 (0–6)ac
6 (0–254) 9 (0–103)Cb
Bouts Bouts
13 (0–171) 23 (0–161)a 2 (1–13)A 3 (0–6)
1 (0–16) 1 (0–7)bc
1 The arrow stands for direction of behaviour. P = TIM performer, R = TIM receiver, N = neutral pig. 2 n refers to the number of dyads, used in analyses on age effect/the number of pens with the dyad, used in analyses on dyad effect). An asterisk * indicates a significant (P < 0.05), an asterisk in brackets (*) a non-significant (0.05 ≤ P < 0.1) age effect within a category and a behaviour. Similar lower-case letter superscripts on a row indicate significant category effects (p<0.05), upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons.
Table 8 Continuously sampled received behaviours of weaned piglets given as median (min-max) for the duration in seconds or number of bouts during 30 min of observation. Behaviour
Age (weeks)
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Nosed, duration
4 9
18 (0–249) 65 (0–440)
33 (0–271) 55 (0–319)
25 (0–423) 30 (0–374)
Nosed, bouts
4 9
4 (0–22)A 5 (0–18)
3 (0–23) 5 (0–13)
5 (0–38)A 4 (0–12)
Tail nosed, duration
4 9
0 (0–57)a 2 (0–192)
0 (0–60)* 19 (0–56)*a
1 (0–80)a 0 (0–33)a
Tail nosed, bouts
4 9
0 (0–7) 1 (0–11)
0 (0–5)* 2 (0–6)*a
0 (0–6) 1 (0–2)a
Ear bitten, duration
4 9
0 (0–0)* 0 (0–189)*
0 (0–0)(*) 0 (0–25)(*)
0 (0–6)(*) 0 (0–29)(*)
Ear bitten, bouts
4 9
0 (0–0)* 0 (0–1)*
0 (0–0) 0 (0–3)
0 (0–2) 0 (0–2)
Tail bitten, duration
4 9
0 (0–0)(*) 0 (0–20)(*)a
0 (0–0)* 8 (0–67)*a
N/A N/A
Tail bitten, bouts
4 9
0 (0–0)(*) 0 (0–2)(*)a
0 (0–0)* 1 (0–6)*a
N/A N/A
Fighting (recipient1 ), duration
4 9
0 (0–105)* 6 (0–140)*
0 (0–63) 0 (0–42)
0 (0–102) 0 (0–244)
Fighting (recipient1 ), bouts
4 9
0 (0–4)* 1 (0–4)*
0 (0–5) 0 (0–4)
0 (0–8) 0 (0–5)
Pushed, duration
4 9
0 (0–18) 0 (0–111)
0 (0–28) 0 (0–130)
0 (0–74) 0 (0–27)
Pushed, bouts
4 9
0 (0–3)A 0 (0–5)
0 (0–8)A 0 (0–7)
0 (0–4) 0 (0–3)
Submissive (recipient), duration
4 9
0 (0–49) 0 (0–21)
0 (0–24) 0 (0–18)
0 (0–0) 0 (0–39)
Submissive (recipient), bouts
4 9
0 (0–1) 0 (0–1)
0 (0–1) 0 (0–1)
0 (0–1) 0 (0–3)
Total social2 , duration
4 9
40 (0–293)*a 141 (0–653)*a
64 (0–332)* 96 (0–319)*
39 (0–503)*a 78 (0–386)*a
Total social2 , bouts
4 9
5 (0–28)a 9 (0–35)Ab
3 (0–30) 13 (0–23)Ac
5 (0–42)a 7 (0–19)bc
1 Fighting is recorded irrespective of the own reaction, which may be submissive or active fighting. 2 Total social behaviours include nosed, tail nosed, ear bitten, tail bitten, recipient of fighting and pushed (being a recipient of submissive behaviour is excluded). An asterisk * indicates a significant (P < 0.05), an asterisk in brackets. (*) a non-significant (0.05 ≤ P < 0.1) age effect within a category and a behaviour. Similar lower-case letter superscripts on a row indicate significant category effects (p < 0.05), upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons.
Looking at daytime continuously sampled behaviour age (Table 6.) the time spent in exploration decreased in P (−149 ± 467 s, Z = −2.00, P = 0.045, n = 29 individuals), with a similar trend in R for the number of bouts (Z = −1.79, P = 0.08, n = 18). Social behaviour was unaffected in R in contrast to the other categories, which showed an increase in performed total social behaviours (P: 75 ± 151 s, Z = −2.21, P = 0.03; N: 103 ± 150 s, Z = −3.00, P = 0.003,
n = 25). In P social behaviours increased included tail sniffing (9 ± 106 s, Z = −2.09, P = 0.04), TIM (Z = −3.66 and −3.51 for both bouts and duration, respectively; P < 0.001 for both), ear biting (P = 0.02 for bouts and duration) and fighting (0.8 ± 2.0 bouts, Z = −2.09, P = 0.04). In N significant increases were present in sniffing (69 ± 126 s, Z = −2.31, P = 0.02) and tail sniffing (18 ± 33 s, Z = −2.96, P = 0.002 and 1.5 ± 2.4 bouts, Z = −2.99, P = 0.004). Ear
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
biting tended to increase in N (Z = −1.68 and −1.71 for duration and bouts, respectively; P = 0.09 for both). Submissive behaviours tended to decrease in P and N (P = 0.06 for bouts and duration in both categories). Within-dyad age-related changes in daytime total social behaviours were few. A significant increase in the duration was present in N → P (55 ± 73 s, Z = −3.86, P < 0.001, n = 32 dyads, Table 7.), with similar trends in N → N (Z = −1.77, P = 0.08, n = 22) and P → R (Z = −1.80, P = 0.07 for duration and Z = −1.70, P = 0.09 for bouts, n = 25). Age effects for received daytime social behaviours are given in Table 8. The sum of received social behaviours increased in all categories, although the change was significant only in P (92 ± 188 s, Z = −2.24, P = 0.02, n = 29) and N (47 ± 153 s, Z = −2.20, P = 0.03, n = 25), with a trend in R (Z = −1.97, P = 0.06, n = 18). The number of bouts of received social behaviour was unaffected by age. Received social behaviours increasing significantly in P included being ear bitten (9 ± 38 s and 0.4 ± 0.9 bouts, P = 0.03 for both comparisons), receiving TIM (Z = −1.89, P = 0.06 for bouts and Z = −1.82, P = 0.07 for duration) and receiving fighting behaviour (18 ± 49 s, Z = −2.55, P = 0.01 and 0.9 ± 1.8 bouts, Z = −2.29, P = 0.02). In R increases were evident in being tail sniffed (12 ± 20 s, Z = −2.55, P = 0.02 and 1.9 ± 2.5 bouts, Z = −2.45, P = 0.01) and in TIM, (11 ± 16 s, Z = −3.06, P = 0.02 and 1.3 ± 1.5 bouts, Z = −3.11, P = 0.002), with a tendency for an increase in being ear bitten (Z = −1.84 and −1.83 for bouts and for duration, respectively; P = 0.7 for both). In N the time receiving ear biting tended to increase (Z = −1.86, P = 0.06).
4. Discussion 4.1. General characteristics of performers and recipients of TIM and neutral pigs This study aimed to identify characteristics of pigs performing TIM, their recipients and neutral penmates at two occasions, the first being at weaning before TIM was observed in the pen and the second being five weeks later when TIM had emerged or was about to emerge. The results show increased overall activity and exploration in pigs that were observed to perform TIM later on, as well as profound changes in social behaviour between these two occasions. Some studies have identified TIM as a precursor for injurious tail biting (Taylor et al., 2010), however, this was not the case in 20 of the 22 study groups when followed until slaughter at 20–22 weeks of age. If we apply the threshold theory of TIM behaviour proposed by Schrøder-Petersen et al. (2003a) to the animals in our study then TIM was exhibited across a range from normal to abnormally elevated frequency, with the threshold for injurious tail biting reached only in two of our groups. Tail biting prevalence may have stayed low due to relatively low-stress housing and husbandry, including plenty of space, stable four-animal groups of siblings, no road transport and only minor need to compete for resources. The current animals were offered bedding material according to four different regimes, which were controlled for in the statistical analyses. As environmental enrichment is known to affect TIM behaviour (Schrøder-Petersen et al., 2003a), the present results will not reveal effects caused by interactions between enrichment status and category of pigs. Gender has repeatedly been associated with tail-directed behaviours. Schrøder-Petersen et al. (2003a, 2004) reported tendencies for females to perform higher frequencies of TIM and for males to receive the behaviour, which is also true for tail biting (e.g. Kritas and Morrison, 2004; Zonderland et al., 2010). The current results are in line with these findings, although statistical significance was not reached.
7
Body weight or growth did not differ between the categories at 4, 7 or 9 weeks of age in agreement with Schrøder-Petersen et al. (2003b). P were, however, born significantly smaller than R. The importance of this result is uncertain, as birth weight to our knowledge has not previously been reported to associate with TIM or tail biting behaviour (Beattie et al., 2005). A few studies have associated performing tail biting with a low and receiving the behaviour with a high body weight (Beattie et al., 2005; Zonderland et al., 2010). Although a low birth weight is known to predict slow growth and/or low weight at least until slaughter weight (Gondret et al., 2005; Fix et al., 2010), no such effects were found in the present data, perhaps due to the selection process not accepting animals representing the extreme ranges of weight in the litters. It should be noted that the statistical analyses of the current data included a large number of comparisons, increasing the risk for false positive results. The conclusions drawn below regarding intercategory behavioural differences were, fortunately, supported by more than one significant result. 4.2. General activity and exploration of the environment Category differences in performed behaviour were evident in the first day after weaning, suggesting that P, R and N represented different phenotypes of pigs. As compared to R, P showed a smallscale but consistently increased level of overall activity over 24 h, and spent more time exploring the environment during the day. These results appear to agree with Ursinus et al. (2014) showing that pen-level increased levels of exploration or overall activity predicts injurious tail biting weeks or months later. Conflicting result were, however, presented by Statham et al. (2009) showing that overall activity did not predict tail biting outbreak until four days prior. Given that TIM is considered an expression of the motivation for both social and environmental exploration (Schrøder-Petersen et al., 2003a,b; Simonsen, 1995; Newberry and Wood-Gush, 1988), the present results may also suggest that this motivation is elevated in P as compared to R already before TIM was observed. Considering that behaviour at 4 weeks of age was observed during a very stressful time the day after weaning, removal of a number of pen mates and transfer to a new pen, the present results may be interpreted as evidence for P and R representing two extremes of coping strategy. These extremes have been characterized by active and passive behaviour in challenging situations, and are referred to as proactive and reactive, respectively (reviewed by Koolhaas et al., 1999). Proactive copers are thought to have an increased risk for becoming tail biters, and reactive copers for becoming victims (Korte et al., 2009; Zupan et al., 2012). Although no formal coping style testing was performed, the present results would support this theory, if TIM is considered a precursor for tail biting behaviour. 4.3. Social behaviour at weaning before evidence of TIM Daytime social activity did not differ between the categories of pigs at 4 weeks of age, whereas some differences existed in the type of activity. P performed more tail-nosing than N, suggesting a difference in interest for the anogenital or tail area already at young age. Numerical differences in exploration may suggest that the increase in tail-nosing was a by-product of increased environmental exploration in P. Dyad-level analyses showed that social activity was reciprocal and without category-level preferences at 4 weeks of age, whereas the attractiveness as a receiver differed between categories. N received significantly more overall social behaviours and tail-nosing, and non-significantly more nosing than P. The reason for this is unclear. N was not by any measure more active or more passive than the other categories and did thus not provoke more attention or act as stationary object for exploration. N and P may
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
8
thus exhibit different avoidance strategies in response to social approaches by others. 4.4. Social behaviour at 9 weeks of age when TIM behaviour had emerged All categories of pigs became more active when comparing behaviour at the ages of 4 and 9 weeks, which is an expected effect of increasing age (Cox and Cooper, 2001). The observation that different behaviours changed in the different categories does, however, indicate that phenotypic differences existed. Inter-category differences in behavioural development were more clear in daytime analyses than time budgets, which is in line with Larsen et al. (2016) reviewing literature on predicting tail biting to conclude that behavioural differences between pens with and without tail biting in the near future may be more pronounced during the active part of the day. In P a part of the active time was reassigned from environmental exploration to social activity when comparing daytime behaviour at 4 and 9 weeks of age. The increase in social behaviour included significant increases in tail nosing, TIM, ear biting and fighting, all of which were performed to relatively high frequency and/or long duration by P as compared to the other categories at 9 weeks of age. Dyad-level analyses show that P went from no preference for social actions at 4 weeks to a significant preference for R over N at 9 weeks, when P also directed more social actions towards R than vice versa. Similar changes in behaviour have not been described before, although Zonderland et al. (2011) reported preferences in tail-directed behaviours by tail biters during the last six days before the first tail wound in a group. The preference by P for R over N at 9 weeks may have been affected by actions of N. N may have avoided the increased social activity by P, perhaps thereby showing a characteristic of the tail biting – resistant phenotype described by Brunberg et al. (2013). This theory is supported by the low amount of received tail nosing and overall received social behaviour in N at 9 weeks of age. Although R may have attracted attention due to wagging of a subclinically injured tail (Simonsen 1995; Munsterhjelm et al., 2013), the data does not support this as N showed no particular interest for R. N did, in contrast, significantly increase the duration of social behaviour directed towards P from 4 to 9 weeks of age to the degree of a significant preference for P over R. This increase in P-directed behaviour may have been provoked by increased social activity by P, however, the present data does not allow for an investigation on this. The behaviour of R changed to a relatively low degree from 4 to 9 weeks of age. Daytime social activity remained unchanged, in contrast to the other categories showing a significant increase. At 9 weeks of age R was characterized by marked social inactivity, in spite of receiving more social behaviour than the other categories. Both P and N directed more social behaviour towards R than vice versa, leaving R the only category performing less social behaviour than it received. This suggests that a large proportion of performed actions were responses to approaches by others, mainly P, as dyadlevel analyses show that R had a significant preference for P over N as a receiver of its social actions. The observed social inactivity in R may be characteristic of a phenotype predisposed to receive abnormal behaviour by others, as described by Keeling et al. (2004). 5. Conclusions This study shows differences in social and exploratory behaviours in TIM-performers, -receivers and neutral pigs already before the behaviour is observed in the group, to suggest that the categories represent different phenotypes. Emergence of TIM
behaviour coincides with profound changes in social relationships in the group including increased activity in performers, directed mainly towards the receivers; and marked social passivity in receivers.
Acknowledgements The authors are grateful for funding from the Finnish Ministry of Agriculture and Forestry in support of the ANIWHA ERA-Net initiative through the FareWellDock-project, as well as the Welfare and Production of the Pig –project, which also was funded by Snellman Ab and Raisio feed Ltd.
References Beattie, V., Breuer, K., O’Connell, N.E., Sneddon, I.A., Mercer, J.T., Rance, K.A., Sutcliffe, M.E.M., Edwards, S.A., 2005. Factors identifying pigs predisposed to tail biting. Anim. Sci. 80, 307–312. Beattie, V.E., 1994. The Effects of Environmental Enrichment on the Domestic Pig Ph.D. Thesis. The Queen’s University of Belfast, Northern Ireland. Blackshaw, J.K., 1981. Environmental effects on lying behaviour and use of trough space in weaned pigs. Appl. Anim. Ethol. 7, 281–286. Brunberg, E., Jensen, P., Isaksson, A., Keeling, L.J., 2013. Behavioural and brain gene expression profiling in pigs during tail biting outbreaks −evidence of a tail biting resistant phenotype. PLoS One 8, e66513, http://dx.doi.org/10.1371/ journal.pone.0066513. Cox, L.N., Cooper, J.J., 2001. Observations on the pre- and post-weaning behaviour of piglets reared in commercial indoor and outdoor environments. Anim. Sci. 72, 75–86. Edwards, S.A., 2006. Tail biting in pigs: understanding the intractable problem. Vet. J. 171, 367–369. Feddes, J.J.R., Fraser, D., 1994. Non-nutritive chewing by pigs – implications for tail-biting and behavioural enrichment. Trans. Am. Soc. Agric. Eng. 37, 947–950. Feddes, J.J., Fraser, D., Buckley, D.J., Poirier, P., 1993. Electronic sensing of non-destructive chewing by growing pigs. Trans. Am. Soc. Agric. Eng. 36, 955–958. Fix, J.S., Cassady, J.P., Herring, W.O., Holl, J.W., Culbertson, M.S., See, M.T., 2010. Effect of piglet birth weight on body weight growth, backfat, and longissimus muscle area of commercial market swine. Livest. Sci. 127, 51–59. Fraser, D., 1987. Attraction to blood as a factor in tail-biting by pigs. Appl. Anim. Behav. Sci. 17, 1–2. Gondret, F., Lefaucheur, L., Louveau, I., Lebret, B., Pichodo, X., Le Cozler, Y., 2005. Influence of piglet birth weight on postnatal growth performance: tissue lipogenic capacity and muscle histological traits at market weight. Livest. Prod. Sci. 93, 137–146. Keeling, L., Andersson, L., Schütz, K.E., Kerje, S., Fredriksson, R., Carlborg, Ö., Cornwallis, C.K., Pizzari, T., Jensen, P., 2004. Chicken genomics: feather-pecking and victim pigmentation. Nature 431, 645–646. Koolhaas, J.M., Korte, S.M., De Boer, S.F., Van Der Vegt, B.J., Van Reenen, C.G., Hopster, H., De Jong, I.C., Ruis, M.A., Blokhuis, H.J., 1999. Coping styles in animals: current status in behavior and stress-physiology. Neurosci. Biobehav. Rev. 23, 925–935. Korte, S.M., Prins, J., Vinkers, C.H., Olivier, B., 2009. On the origin of allostasis and stress-induced pathology in farm animals: Celebrating Darwin’s legacy. Vet. J. 182, 378–383. Kritas, S.K., Morrison, R.B., 2004. An observational study on tail biting in commercial grower-finisher barns. JSHP 12 (1), 17–22. Larsen, M.L., Andersen, H.M., Pedersen, L.J., 2016. Can tail damage outbreaks in the pig be predicted by behavioural change? Vet. J. 209, 50–56. Munsterhjelm, C., Peltoniemi, O.A., Heinonen, M., Hälli, O., Karhapää, M., Valros, A., 2009. Experience of moderate bedding affects behaviour of growing pigs. Appl. Anim. Behav. Sci. 118, 42–53. Munsterhjelm, C., Simola, O., Keeling, L., Valros, A., Heinonen, M., 2013. Health parameters in tail biters and bitten pigs in a case-control study. Animal 7, 814–821. Newberry, R.C., Wood-Gush, D.G.M., 1988. Development of some behaviour patterns in piglets under semi-natural conditions. Anim. Prod. 46, 103–109. Statham, P., Green, L., Bichard, M., Mendl, M., 2009. Predicting tail-biting from behaviour of pigs prior to outbreaks. Appl. Anim. Behav. Sci. 121 (3), 157–164. Schrøder-Petersen, D.L., Simonsen, H.B., Lawson, L.G., 2003a. Tail-in-mouth behaviour among weaner pigs in relation to age, gender and group composition regarding gender. Acta Agric. Scand. Sect. A: Anim. Sci. 53, 29–34. Schrøder-Petersen, D.L., Simonsen, H.B., Lawson, L.G., 2003b. Ontogeny of tail-in-mouth behaviour in piglets reared in barren or enriched environments. Proceedings of the 37th International Congress of the International Society for Applied Ethology, 115. Schrøder-Petersen, D.L., Heiskanen, T., Ersbøll, A.K., 2004. Tail-in-mouth behaviour in slaughter pigs, in relation to internal factors such as: age, size, gender, and motivational background. Acta Vet. Scand. Sect. A: Anim. Sci. 54, 159–166.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
Simonsen, H.B., 1995. Effect of early rearing environment and tail docking on later behaviour and production in fattening pigs. Acta Agric. Scand. Sect. A: Anim. Sci. 45, 139–144. Taylor, N.R., Main, D.C., Mendl, M., Edwards, S.A., 2010. Tail-biting: a new perspective. Vet. J. 186, 137–147. Ursinus, W.W., Van Reenen, C.G., Reimert, I., Bolhuis, J.E., 2014. Tail biting in pigs: blood serotonin and fearfulness as pieces of the puzzle? PLoS One 9, 1–14.
Zonderland, J.J., Bracke, M.B., den Hartog, L.A., Kemp, B., Spoolder, H.A., 2010. Gender effects on tail damage development in single-or mixed-sex groups of weaned piglets. Livest. Sci. 129 (1), 151–158. Zonderland, J.J., Schepers, F., Bracke, M.B.M., Den Hartog, L.A., Kemp, B., Spoolder, H.A.M., 2011. Characteristics of biter and victim piglets apparent before a tail-biting outbreak. Animal 5 (05), 767–775. Zupan, M., Janczak, A.M., Framstad, T., Zanella, A.J., 2012. The effect of biting tails and having tails bitten in pigs. Phys. Behav. 106, 638–644.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
9
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
Applied Animal Behaviour Science xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Applied Animal Behaviour Science journal homepage: www.elsevier.com/locate/applanim
Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Camilla Munsterhjelm ∗ , Mari Heinonen, Anna Valros University of Helsinki, Faculty of Veterinary Medicine, Department of Production Animal Medicine, PB 57, FIN-00014 Finland
a r t i c l e
i n f o
Article history: Received 18 April 2016 Received in revised form 18 August 2016 Accepted 25 August 2016 Available online xxx Keywords: Pig Social behaviour Tail biting Tail-in-mouth
a b s t r a c t This study aimed to identify characteristics of pigs performing tail-in-mouth behaviour (TIM; P, n = 34), their recipients (R, n = 23) and neutral penmates (N, n = 31) at two occasions, the first being at weaning (4 weeks of age) before TIM was observed in the pen and the second being at 9 weeks of age when TIM had emerged, but no clinical tail lesions were observed. The groups (n = 22) were formed by siblings, two gilts and two castrates. Behaviour was analysed as 24-h time budgets and continuously sampled during 30 min of the active part of the day. Category (P, R, N) effects were analysed at individual and (directed) dyad level. P was born significantly smaller than R, but the difference had disappeared at 4 weeks. Growth or sex distribution did not differ between categories. Category differences in performed behaviour were evident at 4 weeks of age, when P showed more overall activity and environmental exploration as compared to R, as well as more bouts of tail-nosing than N. Different aspects of behaviour changed in the different categories between 4 and 9 weeks of age. In P social activity increased significantly and went from no preference at 4 weeks to a significant preference for social actions for R over N at 9 weeks. N was socially passive at 9 weeks while receiving more social behaviour than the other categories. These differences in behaviour suggest that the categories represented different phenotypes of pigs. © 2016 Published by Elsevier B.V.
1. Introduction Tail biting is a behavioural disturbance of growing pigs associated with an unsatisfactory environment. It is a significant economical and welfare problem in modern pig production. Although a number of farm-level predisposing factors have been established their effects appear to be interacting heavily with individual characteristics of the pigs, making outbreaks unpredictable (Taylor et al., 2010). The identification of measures able to prevent tail biting requires a better understanding of the initial stages of a tail biting outbreak, especially of the contribution of individual pigs (Edwards, 2006). Gentle tail manipulation or tail-in-mouth behaviour (TIM), causing no obvious injuries to the tail, is considered to be a behavioural precursor to damaging tail biting (Feddes et al., 1993; Feddes and Fraser, 1994). According to the review by Taylor et al. (2010) this two-stage type of tail biting appears to account for the majority of tail biting in commercial pig production and also of the tail-biting cases described in scientific papers.
∗ Corresponding author. E-mail address: camilla.musterhjelm@helsinki.fi (C. Munsterhjelm).
The progression from the pre-injury stage, characterized by TIM, to the injury stage in two-stage tail biting takes place can occur rapidly (Fraser, 1987). TIM may also be performed without subsequent injurious biting (Schrøder-Petersen et al., 2003a). There are diverging views among scientist on the categorization of TIM as normal or abnormal behaviour. TIM has been suggested to be abnormal upon observations that only one or a few individuals in a group have been performing the behaviour (Blackshaw, 1981; Beattie et al., 2005), whereas others have reported lowfrequency TIM performed by all or a majority of pigs in a population and thereby classified it as a normal activity (Feddes et al., 1993; Schrøder-Petersen et al., 2003a). An obvious reason for this discrepancy, in addition to differences in observational strategy, may be variation in TIM frequency and intensity between populations and/or occasions. Schrøder-Petersen et al. (2003a) suggested that TIM behaviour of low frequency and intensity is normal behaviour that may increase to abnormal levels under certain circumstances, to eventually become actual injurious tail biting. TIM is thought to be an expression of the motivation to explore both the environment and in a social context (Schrøder-Petersen et al., 2003a,b; Simonsen, 1995; Newberry and Wood-Gush, 1988). The behaviour has been reported to decrease in response to straw provision, increase with age at least in weaners, and typically be
http://dx.doi.org/10.1016/j.applanim.2016.08.013 0168-1591/© 2016 Published by Elsevier B.V.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
2
performed by females and received by males (Schrøder-Petersen et al., 2003a; Simonsen, 1995). The aim of this study was to identify individual differences in size, growth and behaviour in TIM-performers, TIM-receivers and neutral penmates before TIM behaviour was observed in the pen and when the behaviour had begun or was about to begin in the group. 2. Materials and methods 2.1. Animals and husbandry The animals (n = 88 pigs in 22 groups) were chosen from a larger experiment (252 pigs in 63 groups) investigating different enrichment regimes, conducted over 13 batches in 2004–2005 at the experimental pig farm of MTT Agrifood Research in southern Finland. The animals represented pure Finnish Yorkshire or Finnish Landrace and their crosses with or without ¼ Duroc. Housing and husbandry will be described only shortly, as details are given elsewhere (Munsterhjelm et al., 2009). The animals were housed in pens with partly slatted floors at all times. Sows farrowed and nursed in pens of 5.4 m2 without crates, equipped with a heated creep area. During the first three days litter size was standardized to no more than 12 piglets by cross-fostering onto non-experimental sows. Piglets were subjected to teeth clipping and foreleg taping at one day of age as well as an intramuscular iron injection (1 ml Ursoferran® vet., Serumwerk Bernburg AG, Germany) and castration at day three. Tails were left intact according to Finnish legislation. The piglets were weaned at an average age of 29.5 days (range 26–36). At weaning groups of weight-matched siblings were formed including two gilts and two castrates. In the nursery (4–9 weeks of age) animals were allowed a stocking density of 0.7 m2 per animal in pens equipped with a single-space ad libitum feeder (Groba Diplomat AS, Skælskør, Denmark) and two drinking nipples. Climate was controlled according to age. In the fattening unit pigs were housed with a stocking density of 1.2 m2 per pig and fed from a trough. The animals were provided enrichment according to one of six schemes being investigated in the original experiment. Pens either were (=E) or were not (=0) bedded with wood shavings and chopped straw thinly enough to show the pen floor during suckling, nursery and fattening stages. Accordingly, enrichment regimes were 1) 000 (referring to 0 in the suckling stage, 0 in the nursery and 0 in the fattening stage; n = 8 pigs in 2 groups); 2) E00 (n = 12/3); 3) EE0 (n = 20/5); 4) 00E (n = 8/2); 5) 0EE (n = 24/6) and 6) EEE (n = 16/4). This experimental procedure was approved by the ethical committee of MTT Agrifood Research Finland (permission number SIK 6/04, 2004-08-26). Animals had full artificial lighting between 07:00 and 16:00 h and dimmed lighting enabling night-time videotaping for the remainder of the time. They were fed standard pelleted pig diets without additives formulated to fulfill current Finnish feeding recommendations and slaughtered at 20–22 weeks of age. 2.2. Experimental design TIM was defined as manipulation of the tail irrespective of the reaction by the recipient, but without causing clinical damage. All groups fulfilling three inclusion criteria were selected for the study: 1) no TIM behaviour was observed on video tape during the first day after weaning (age 4 weeks), 2) TIM behaviour was observed on video tape on the last day in the nursery and/or on the first day in the fattening unit (age 9–10 weeks), and 3) no clinical tail wounds were observed from birth to 10 weeks of age. TIM behaviour was defined
as taking a tail in the mouth, regardless of the reaction by the other pig. The animals were categorized as TIM performers (P, n = 34 pigs in 22 groups), TIM receivers (R, n = 23 pigs in 19 groups) and neutral pigs (N, n = 31 pigs in 19 groups). One occasion of TIM was enough to categorize an animal as P or R. P observed to receive TIM (n = 7 pigs in 5 groups) were included in the P category. Characteristics of the experimental groups are summarized in Table 1. The animals were weighed at birth, at weaning at 4 weeks of age, and at the middle and end of the nursery period (age 7 and 9 weeks). Behaviour was monitored with a time-lapse black and white video recording device for 24 h starting at 07:00 h on the first and last day in the nursery (age 4 and 9 weeks), on the first day in the fattening unit (age 10 weeks) and at 14 and 18 weeks of age. The cameras were connected to a video recording device through a multiplexer. The animals were marked individually with permanent drawing ink the day before each monitoring. Social and exploratory behaviours were extracted by continuous focal sampling at 13:00–13:10, 14:00–14:10 and 15:00–15:10 h according to the ethogram given in Table 2. The time slots were chosen in order to represent the most active part of the day according to Beattie (1994). A bout was considered to end if the animal stopped performing a behaviour for two seconds or more. Time budgets were investigated by instantaneous scan sampling with a 10 min interval for the whole 24 h according to the ethogram in Table 3. Continuously sampled behaviours will be referred to as “daytime behaviours” to be distinguished from 24-h time budgets. All behaviours were recorded on individual level. The recipient was recorded in addition to the performer for all social behaviours. 2.3. Statistical analyses The SPSS for Windows statistics package, version 22.0 (IBM Corp., Armonk, NY, USA) was utilized for statistical analyses. The experimental unit was the individual. Growth was expressed as the relative change in body weight over a given time span. The distribution of sex within each category was evaluated against a balanced distribution using the Chi-squared test. Scan sampled behavioural variables were analysed as the number of scans in the behaviour over 24 h. Continuous behavioural variables were analysed as total time in the behaviour and as the number of bouts during 30 min of observation. All performed social behaviours (nosing, tail nosing, ear biting, TIM, fighting, pushing and submissive behaviours) were summarized in the variable “total social behaviours”, and all received social behaviours except “submissive” were summarized in “received total social behaviours”. Being the recipient of a submissive action was, by definition, a response to social activity and thus not considered a true received behaviour. Daytime performed total social behaviour was subjected to dyad-level analysis in order to investigate category-level preferences for social actions. A dyad is a unit including a performer and a receiver. The dyads were considered as directed, meaning that A directing behaviour towards B was a different dyad than B directing behaviour towards A. Hereby the data included nine different dyads (P → R, R → P, P → N, N → P, N → R, R → N, P → P, N → N, R → R), however, the dyad P → P was excluded from all analyses due to a low n. Normality was investigated for all continuous variables using histograms, Q–Q plots, Kolmogorov-Smirnov and Shapiro-Wilk tests; and attempts made to transform the non-normal ones. Category and dyad effects on variables that could not be normalized were determined using the Pair-Wise Related-Samples Wilcoxon Test, or, if the within-pair difference was non-symmetrically distributed, the Sign Test. The group effect was counted for by considering the categories or dyads within a group as the repeated feature. In groups with several animals representing the same category or dyad their median was used in the analyses. The tests were
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
3
Table 1 Characteristics of the groups included in the study. Group1
1 2 3 410 5 6 710 8 9 10 11 12 13 149,10 159,10 16 17 18 1910 20 21 2210
Sex2
Bedding3 in suckling/weaning/ fattening stage
TIM performers
TIM receivers
Neutral pigs
F F F F F F F FBV FBV F FBV FM FBV MBV FBV MM FFM FM FM FM M M M M M M
F F F F FMM M MM
M MM MM MM
M M
FM F MM FM FM
F F M M F F F FM M M
M M F F FM FM FM F FF FF
1
0/1/1 0/1/1 1/0/0 1/1/1 0/0/0 0/1/1 1/1/0 1/1/0 0/1/1 0/1/1 1/0/0 0/0/1 1/1/0 1/1/0 1/0/0 0/1/1 1/1/0 1/1/1 1/1/1 1/0/0 0/0/1 1/1/1
Age (weeks) of continued TIM or tail biting in the group4
17L , 20L 14
14L , 16L 20 20 14, 20 14 14 14 20
14 14
10
TIM was observed on the last day in the nursery (9 weeks of age) in groups without a superscript, TIM was observed on the first day in the fattening unit (10 weeks of age), TIM was observed at both occasions. 2 F = female, M = castrate, BV indicates that the performer also received TIM 3 0 = no bedding 1 = moderate amount of wood shavings and chopped straw. 4)L Indicates that tail wounds were present, the lack of superscript that TIM was detected on video (recording at 14 and 18 weeks) in the absence of tail wounds. 9,10
Table 2 Definitions of behaviours of weaned piglets for continuous sampling.
Table 3 Definitions of behaviours of weaned piglets for scan sampling.
Behaviour
Description
Behaviour
Description
Nosing
Nosing, sniffing or touching penmate with the snout in any area other than the tail, distance from snout to the skin less than 5 cm
Performed behaviours Inactive
Performing no activity
Exploring
Tail-nosing
Nosing, sniffing or touching the tail of penmate with the snout, distance from snout to the tail less than 5 cm
TIMa
Taking penmate’s tail in the mouth irrespectively of the reaction by the receiver
Nosing or manipulating any part of the pen or bedding except for the drinking nipple, or inside the feeder or trough. Observing, nosing or manipulating any object outside the pen with the snout outside the pen.
Ear biting
Taking penmate’s ear in the mouth
Performing social behaviour
Pushing
Pushing penmate with any part of the body in order to displace him, no biting
Any active behaviour involving nosing (distance from snout to skin 0–5 cm), biting, ramming, lifting or otherwise manipulating penmate. Pushing penmate with any part of the body. Mounting with both front hoofs on or resting on sternum on penmate’s back.
Fighting
Mutual pushing parallel or perpendicular, ramming or pushing penmate with the head accompanied by biting, lifting penmate with the snout
Other activity
Any other activity
Missing Received behaviours Receiving social behaviour
Unidentified activity or pig missing Being the recipient of behaviours included in Performing social behaviour irrespective of own actions.
Submissive
Being the recipient of Pushing or Fighting, responding passively or by fleeing, not fighting back
Exploring
Nosing or manipulating any part of the pen or bedding except for the drinking nipple, or inside the feeder or trough. Observing, nosing or manipulating any object outside the pen with the snout outside the pen.
Other behaviour
Any other behaviour
Missing
Unidentified activity or pig missing
a
Tail-in-mouth behaviour.
applied pair-wise due to a large number of groups with one missing category or missing dyads. Dyad-level comparisons were designed to identify category-level preferences for social actions (e.g. B → V was compared to B → C) and to compare reciprocity (e. g. B → V was compared to V → B).
Category effects on normal(ized) variables were analysed using the Linear Mixed Models (LMM) feature in SPSS. Batch- and grouplevel clustering were addressed by allowing random intercept variation of group nested within batch. Category and enrichment status (yes/no) in weaning and/or suckling stage, as appropriate, were included as group-level, and sex as individual-level fixed factors. Factors were removed according to a backward step-wise elimination strategy. Variables with P ≤ 0.1 were included in the model if a good model fit did not require otherwise. A model was accepted if homoscedasticity and sound distribution of residuals were present. If a significant (P < 0.05) category effect was detected pair-wise comparisons were performed using a Bonferroni correction. Subsequently, the variables analysed using LMM were the square root of birth weight reduced by a constant of 4; body weight
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
4
Table 4 Characteristics of TIM performers, receivers and neutral pigs given as an average ± SD.
a
Characteristic
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Birth weight (kg) Weight, 5 weeks (kg) Weight, 7 weeks (kg) Weight, 9 weeks (kg) Females (n) Castrates (n)
1.8 ± 0.3a 10.1 ± 1.8 16.5 ± 4.0 25.5 ± 5.5 19 15
2.0 ± 0.4a 10.0 ± 1.6 15.7 ± 2.7 24.8 ± 4.1 11 12
1.8 ± 0.4 9.8 ± 1.9 15.8 ± 3.5 24.4 ± 5.4 14 17
indicates a significant category effect (P < 0.05) in pair-wise comparisons.
at 4, 7, and 9 weeks, growth from 0 to 4 and 4–7 weeks and the square root of growth from 7 to 9 weeks. LMM was also applied to the following variables describing continuously sampled exploration: the duration at 4 weeks of age, the logarithm of the number of bouts at 4 weeks of age with an added constant of 1, the square root of the duration and the number of bouts at 9 weeks of age. In order to compare the development of behaviour from 4 to 9 weeks of age between the categories and the dyads age effects on behaviour were analysed separately for each category and dyad. All behaviours were anlysed using Pair-Wise RelatedSamples Wilcoxon Tests (PWRSW) with age (9 vs 4 weeks) as the repeated effect, as described above. Results are given as average ± SD for between-pair differences with values for the Z-statistic for PWRSW tests. No statistic exists for the Sign Test. 3. Results 3.1. Sex and performance The distribution of sex did not differ from balanced (X2 = 83.04, P = 0.6, n = 88, Chi-squared test). Birth weight was higher in R than P (LMM for category effect F2,35 = 3.97, P = 0.03, pair-wise Bonferronicorrected comparison F2,44 = 3.96, P = 0.03, Table 4.). Body weight was not affected by category at 4, 7 or 9 weeks of age, nor was growth between succeeding time points. 3.2. Category effects on time budgets At 4 weeks of age R was inactive in a larger number of scans than P (within-pair average difference 3.0 ± 6.1 [SD] scans, Z = −2.25, P = 0.02, n = 19 pairs of pens, Pair-Wise Related-Samples Wilcoxon Test). At 9 weeks of age P was inactive in a larger number of scans than N (3.2 ± 6.6 scans, Z = −2.11, P = 0.04, n = 19). 3.3. Category and dyad effects on behaviour during the active parts of the day at 4 weeks of age Analyses without other notation are carried out using the Pair-Wise Related-Samples Wilcoxon Test, and between-pair differences are given as average ± SD. The time spent in exploration differed between category (LMM F2,29 = 3.77, P = 0.04, Table 6) at 4 weeks of age. Controlling for suckling stage enrichment (F1,19 = 4.52, P = 0.05), P was predicted to explore significantly more (9 min 21 s out of the total 30 min of observation, calculated for a non-bedded suckling pen) than R (7 min 20 s, F2,41 = 3.77, P = 0.046 in pair-wise Bonferroni-corrected comparison). P performed a larger number of bouts of tail-nosing behaviour than N (1.1 ± 5.1, Z = −2.33, P = 0.02, n = 17 pairs of pens). R tended to perform a larger number of bouts of pushing than N (Z = −1.90, P = 0.06, n = 14). Nosing, ear biting, fighting or submissive behaviours did not differ between categories at 4 weeks of age, nor did total social behaviours analysed at dyadic level (Table 7). N received more total social behaviours than P (2.5 ± 4.5 bouts and
18 ± 76 s, Z = −2.10, P = 0.04 for both comparisons, n = 16, Table 8), including more tail nosing (3.6 ± 8.9 s, P = 0.02) and a tendency for more bouts of received nosing (Z = −1.92, P = 0.06). R tended to receive more bouts of pushing than P (Z = −1.75, P = 0.08, n = 17). 3.4. Category effects on behaviour during the active parts of the day at 9 weeks of age Analyses without other notation are carried out using the PairWise Related-Samples Wilcoxon Test. The between-pair difference is given as average ± SD. R performed the least amount of bouts of social behaviours with significant differences both to P (−6.0 ± 8.4 bouts in R as compared to P, Z = −1.97, P = 0.049, n = 16 pairs of pens) and to N (−2.3 ± 3.7, Z = −2.19, P = 0.03, n = 13). The difference in social activity between R and P included tendencies for nosing bouts (Z = −1.92, P = 0.06) and tail-nosing bouts (P = 0.07, Sign test). The difference between R and N included nosing (−2.1 ± 3.3 bouts in R as compared to N, Z = −2.10, P = 0.04 and Z = −1.87, P = 0.06 for duration, n = 13) and tail-nosing (−2.0 ± 4.5 bouts, P = 0.008, Sign test, n = 13). Social activity between P and N differed only for bouts of tail-nosing, which P tended to perform more frequently (Z = −1.85, P = 0.07, n = 16), and submissive behaviours that tended to be performed for a longer time by N than P (Z = −1.75, P = 0.08). Ear biting, fighting and pushing did not differ between categories at 9 weeks og age. R and N did not, by definition, perform any TIM. The difference in TIM activity as both duration and bouts was significant for comparisons of P to R and to N (p = 0.01–0.02, n = 13–16 pairs). In dyad-level analyses on daytime total social behaviour at 9 weeks P directed more attention towards R than towards N (+43 ± 68 s, Z = −2.04, P = 0.04; and 3.0 ± 4.0 bouts, Z = −2.19, P = 0.03, n = 13 pairs of pens). R directed more social behaviours towards P than towards N (0.6 ± 1.5 bouts, Z = −1.97, P = 0.03 and Z = −1.95, P = 0.05 for duration, n = 13). N tended to direct social behaviour for a longer total time towards P than towards R (Z = −1.78, P = 0.08, n = 11). Looking at reciprocity at 9 weeks of age N directed more social behaviour towards R than the other way around (22 ± 35 s, Z = −2.70, P = 0.007 and 1.5 ± 3.6 bouts, Z = −2.20, P = 0.03, n = 12) and P directed more bouts towards R than vice versa (6.1 ± 6.4 bouts, Z = −2.49, P = 0.01, n = 16). N received the least amount of daytime total social behaviours of all categories of pigs with a significant difference as compared to both P (−2.1 ± 4.6 bouts, P = 0.02 and −83 ± 163 s, Z = 2.05, P = 0.04, n = 15 pairs of pens) and R (−3.7 ± 5.7 bouts, Z = −2.10, P = 0.02, n = 16, Table 8.). R tended to receive more bouts of total social behaviour than P (Z = −1.92, P = 0.06, n = 16) and received significantly more TIM (11 ± 17 s and +1.1 ± 1.5 bouts, P = 0.002 for both comparisons). N received less tail sniffing than R (−14 ± 14 s, P = 0.02 and −1.5 ± 1.5 bouts, Z = −2.37, P = 0.02). N was omitted from analyses on TIM as they by definition received none. 3.5. Age effects on behaviour Analyses on age effects without other notation are carried out using the Pair-Wise Related-Samples Wilcoxon Test. The magnitude of change from 4 to 9 weeks of age is given as the within-individual (or unique dyad) average ± SD. The number of 24-h scans in exploration was unaffected by age, whereas many social behaviours changed significantly (Table 5.). The number of inactive scans decreased in all categories (P: −11 ± 26 scans, Z = −2.57, P = 0.01, n = 34 individuals; R: −16 ± 12, Z = −3.33, P = 0.001, n = 23; N: −15 ± 14, Z = −2.97, P = 0.003, n = 31), while other behaviours increased (P: 5 ± 7, Z = −3.12, P = 0.002; R: 5 ± 7, Z = −2.50, P = 0.03, N: 4 ± 7, Z = −2.70, P = 0.007). Performed social behaviour changed only in R (2 ± 4 scans, Z = −2.15, P = 0.03), and received social behaviour only in P (2 ± 3, Z = −3.10, P = 0.002).
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
5
Table 5 24-h time budgets given as the median (min-max) number of scans. Behaviour
Age (weeks)
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Inactive
4 9
111 (78–126)*a 105 (27–116)*a
112 (90–128)*a 100 (33–115)*
110 (86–130)* 102 (27–119)*a
Exploration
4 9
15 (6–40) 12 (4–28)
14 (5–30) 13 (6–28)
13 (7–41) 13 (6–28)
Other
4 9
9 (2–23)* 15 (6–25)*
8 (4–18)* 12 (4–24)*
9 (3–21)* 14 (3–29)*
Performed social
4 9
6 (0–16) 4 (0–14)
4 (0–14)* 7 (1–16)*
5 (1–16) 5 (0–13)
Received social
4 9
1 (0–7)A * 3 (0–12)*
3 (0–7)A 2 (0–8)
2 (0–7) 1 (0–9)
An asterisk (*) indicates a significant within-category age effect (P < 0.05) for a behaviour. Similar lower-case letter superscripts on a row indicate significant (P < 0.05) category effects and upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons.
Table 6 Continuously sampled behaviours of weaned piglets given as median (min-max) for the duration in seconds and for the number of bouts during 30 min of observation during the active part of the day. Behaviour
Age (weeks)
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Exploration, duration
4 9
639 (26–1228)*a 347 (45–840)*
536 (0–928)a 267 (12–837)
462 (73–940) 477 (0–1052)
Exploration, bouts
4 9
16 (1–84) 14 (1–30)
15 (0–62)(*) 10 (1–22)(*)
11 (5–99) 11 (0–35)
Nosing, duration
4 9
36 (0–318) 69 (0–347)
24 (0–208) 23 (0–135)A
29 (0–337)* 33 (0–368)*A
Nosing, bouts
4 9
5 (0–37) 6 (0–22)A
6 (0–24) 4 (0–19)Ab
3 (0–32) 2 (0–14)b
Tail-nosing, duration
4 9
0 (0–57)* 13 (0–55)*
0 (0–100) 0 (0–126)
0 (0–53)* 6 (0–105)*
Tail-nosing, bouts
4 9
0 (0–6)*a 1 (0–9)*AB
0 (0–7) 0 (0–5)Ac
0 (0–3)*a 1 (0–8)*Bc
Ear biting, duration
4 9
0 (0–0)* 0 (0–164)*
0 (0–5) 0 (0–33)
0 (0–6)(*) 0 (0–21)(*)
Ear biting, bouts
4 9
0 (0–0)* 0 (0–4)*
0 (0–1) 0 (0–2)
0 (0–1)(*) 0 (0–3)(*)
TIM1 , duration
4 9
0 (0–0)* 3 (0–64)*ab
0 (0–0) 0 (0–0)a
0 (0–0) 0 (0–0)b
TIM1 , bouts
4 9
0 (0–0)* 1 (0–5)*ab
0 (0–0) 0 (0–0)a
0 (0–0) 0 (0–0)b
Fighting, duration
4 9
0 (0–39) 6 (0–129)a
0 (0–108) 0 (0–70)a
0 (0–86) 0 (0–265)
Fighting, bouts
4 9
0 (0–5)* 1 (0–5)*
0 (0–4) 0 (0–3)
0 (0–7) 0 (0–6)
Pushing, duration
4 9
0 (0–23) 0 (0–72)
0 (0–18) 0 (0–27)
0 (0–74) 0 (0–100)
Pushing, bouts
4 9
0 (0–4) 0 (0–6)
0 (0–5)A 0 (0–3)
0 (0–3)A 0 (0–3)
Submissive, duration
4 9
0 (0–49)(*) 0 (0–4)(*)A
0 (0–1) 0 (0–36)
0 (0–31)(*) 0 (0–19)(*)A
Submissive, bouts
4 9
0 (0–1)(*) 0 (0–1)(*)
0 (0–1) 0 (0–3)
0 (0–2)(*) 0 (0–2)(*)
Total social2 , duration
4 9
49 (0–359)* 159 (0–395)*
38 (0–438) 62 (0–248)
47 (0–260)* 103 (0–514)*
Total social2 , bouts
4 9
6 (0–41) 11 (0–28)a
5 (0–38) 5 (0–20)ab
6 (0–36) 7 (0–32)b
An asterisk * indicates a significant (P < 0.05), an asterisk in brackets (*) a non-significant (0.05 ≤ P < 0.1) age effect within a category and a behaviour. Similar lower-case letter superscripts on a row indicate significant category effects (p < 0.05), upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons. 1 Tail-in-mouth behaviour 2 Total social behaviours include nosing, tail-nosing, ear biting, TIM, fighting, pushing and submissive behaviours.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model
ARTICLE IN PRESS
APPLAN-4322; No. of Pages 9
C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
6
Table 7 Dyad level social behaviour of weaned piglets given as median (min-max) of the duration in seconds or number of bouts obtained from 30 min of continuous behavioural observation. Unit
Age (weeks)
P → P1 (n = 20/8)2
P→N (n = 31/15)
P→R (n = 30/25)
N→P (n = 32/19)
N→N (n = 22/10)
N→R n = 21/16)
R→P (n = 24/19)
R→N (n = 18/16)
Duration Duration
4 9
28 (0–130) 26 (0–325)
4 9
2 (0–12) 3 (0–8)
22(0–160)(*) 49 (0–218)(*)a 2 (0–15)(*) 6 (0–16)(*)a
8 (0–79)* 50 (0–321)*A 2 (0–8)A 3 (0–15)
8 (0–74)(*) 21 (0–248)(*) 1 (0–12) 2 (0–7)
6 (0–119) 22 (0–192)Ab 2 (0–13) 3 (0–6)b
18 (0–107) 11 (0–227)C 2 (0–13) 2 (0–6)ac
6 (0–254) 9 (0–103)Cb
Bouts Bouts
13 (0–171) 23 (0–161)a 2 (1–13)A 3 (0–6)
1 (0–16) 1 (0–7)bc
1 The arrow stands for direction of behaviour. P = TIM performer, R = TIM receiver, N = neutral pig. 2 n refers to the number of dyads, used in analyses on age effect/the number of pens with the dyad, used in analyses on dyad effect). An asterisk * indicates a significant (P < 0.05), an asterisk in brackets (*) a non-significant (0.05 ≤ P < 0.1) age effect within a category and a behaviour. Similar lower-case letter superscripts on a row indicate significant category effects (p<0.05), upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons.
Table 8 Continuously sampled received behaviours of weaned piglets given as median (min-max) for the duration in seconds or number of bouts during 30 min of observation. Behaviour
Age (weeks)
TIM performers (P, n = 34)
TIM receivers (R, n = 23)
Neutral pigs (N, n = 31)
Nosed, duration
4 9
18 (0–249) 65 (0–440)
33 (0–271) 55 (0–319)
25 (0–423) 30 (0–374)
Nosed, bouts
4 9
4 (0–22)A 5 (0–18)
3 (0–23) 5 (0–13)
5 (0–38)A 4 (0–12)
Tail nosed, duration
4 9
0 (0–57)a 2 (0–192)
0 (0–60)* 19 (0–56)*a
1 (0–80)a 0 (0–33)a
Tail nosed, bouts
4 9
0 (0–7) 1 (0–11)
0 (0–5)* 2 (0–6)*a
0 (0–6) 1 (0–2)a
Ear bitten, duration
4 9
0 (0–0)* 0 (0–189)*
0 (0–0)(*) 0 (0–25)(*)
0 (0–6)(*) 0 (0–29)(*)
Ear bitten, bouts
4 9
0 (0–0)* 0 (0–1)*
0 (0–0) 0 (0–3)
0 (0–2) 0 (0–2)
Tail bitten, duration
4 9
0 (0–0)(*) 0 (0–20)(*)a
0 (0–0)* 8 (0–67)*a
N/A N/A
Tail bitten, bouts
4 9
0 (0–0)(*) 0 (0–2)(*)a
0 (0–0)* 1 (0–6)*a
N/A N/A
Fighting (recipient1 ), duration
4 9
0 (0–105)* 6 (0–140)*
0 (0–63) 0 (0–42)
0 (0–102) 0 (0–244)
Fighting (recipient1 ), bouts
4 9
0 (0–4)* 1 (0–4)*
0 (0–5) 0 (0–4)
0 (0–8) 0 (0–5)
Pushed, duration
4 9
0 (0–18) 0 (0–111)
0 (0–28) 0 (0–130)
0 (0–74) 0 (0–27)
Pushed, bouts
4 9
0 (0–3)A 0 (0–5)
0 (0–8)A 0 (0–7)
0 (0–4) 0 (0–3)
Submissive (recipient), duration
4 9
0 (0–49) 0 (0–21)
0 (0–24) 0 (0–18)
0 (0–0) 0 (0–39)
Submissive (recipient), bouts
4 9
0 (0–1) 0 (0–1)
0 (0–1) 0 (0–1)
0 (0–1) 0 (0–3)
Total social2 , duration
4 9
40 (0–293)*a 141 (0–653)*a
64 (0–332)* 96 (0–319)*
39 (0–503)*a 78 (0–386)*a
Total social2 , bouts
4 9
5 (0–28)a 9 (0–35)Ab
3 (0–30) 13 (0–23)Ac
5 (0–42)a 7 (0–19)bc
1 Fighting is recorded irrespective of the own reaction, which may be submissive or active fighting. 2 Total social behaviours include nosed, tail nosed, ear bitten, tail bitten, recipient of fighting and pushed (being a recipient of submissive behaviour is excluded). An asterisk * indicates a significant (P < 0.05), an asterisk in brackets. (*) a non-significant (0.05 ≤ P < 0.1) age effect within a category and a behaviour. Similar lower-case letter superscripts on a row indicate significant category effects (p < 0.05), upper-case letters a trend (0.05 ≤ P < 0.1) in pair-wise comparisons.
Looking at daytime continuously sampled behaviour age (Table 6.) the time spent in exploration decreased in P (−149 ± 467 s, Z = −2.00, P = 0.045, n = 29 individuals), with a similar trend in R for the number of bouts (Z = −1.79, P = 0.08, n = 18). Social behaviour was unaffected in R in contrast to the other categories, which showed an increase in performed total social behaviours (P: 75 ± 151 s, Z = −2.21, P = 0.03; N: 103 ± 150 s, Z = −3.00, P = 0.003,
n = 25). In P social behaviours increased included tail sniffing (9 ± 106 s, Z = −2.09, P = 0.04), TIM (Z = −3.66 and −3.51 for both bouts and duration, respectively; P < 0.001 for both), ear biting (P = 0.02 for bouts and duration) and fighting (0.8 ± 2.0 bouts, Z = −2.09, P = 0.04). In N significant increases were present in sniffing (69 ± 126 s, Z = −2.31, P = 0.02) and tail sniffing (18 ± 33 s, Z = −2.96, P = 0.002 and 1.5 ± 2.4 bouts, Z = −2.99, P = 0.004). Ear
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
biting tended to increase in N (Z = −1.68 and −1.71 for duration and bouts, respectively; P = 0.09 for both). Submissive behaviours tended to decrease in P and N (P = 0.06 for bouts and duration in both categories). Within-dyad age-related changes in daytime total social behaviours were few. A significant increase in the duration was present in N → P (55 ± 73 s, Z = −3.86, P < 0.001, n = 32 dyads, Table 7.), with similar trends in N → N (Z = −1.77, P = 0.08, n = 22) and P → R (Z = −1.80, P = 0.07 for duration and Z = −1.70, P = 0.09 for bouts, n = 25). Age effects for received daytime social behaviours are given in Table 8. The sum of received social behaviours increased in all categories, although the change was significant only in P (92 ± 188 s, Z = −2.24, P = 0.02, n = 29) and N (47 ± 153 s, Z = −2.20, P = 0.03, n = 25), with a trend in R (Z = −1.97, P = 0.06, n = 18). The number of bouts of received social behaviour was unaffected by age. Received social behaviours increasing significantly in P included being ear bitten (9 ± 38 s and 0.4 ± 0.9 bouts, P = 0.03 for both comparisons), receiving TIM (Z = −1.89, P = 0.06 for bouts and Z = −1.82, P = 0.07 for duration) and receiving fighting behaviour (18 ± 49 s, Z = −2.55, P = 0.01 and 0.9 ± 1.8 bouts, Z = −2.29, P = 0.02). In R increases were evident in being tail sniffed (12 ± 20 s, Z = −2.55, P = 0.02 and 1.9 ± 2.5 bouts, Z = −2.45, P = 0.01) and in TIM, (11 ± 16 s, Z = −3.06, P = 0.02 and 1.3 ± 1.5 bouts, Z = −3.11, P = 0.002), with a tendency for an increase in being ear bitten (Z = −1.84 and −1.83 for bouts and for duration, respectively; P = 0.7 for both). In N the time receiving ear biting tended to increase (Z = −1.86, P = 0.06).
4. Discussion 4.1. General characteristics of performers and recipients of TIM and neutral pigs This study aimed to identify characteristics of pigs performing TIM, their recipients and neutral penmates at two occasions, the first being at weaning before TIM was observed in the pen and the second being five weeks later when TIM had emerged or was about to emerge. The results show increased overall activity and exploration in pigs that were observed to perform TIM later on, as well as profound changes in social behaviour between these two occasions. Some studies have identified TIM as a precursor for injurious tail biting (Taylor et al., 2010), however, this was not the case in 20 of the 22 study groups when followed until slaughter at 20–22 weeks of age. If we apply the threshold theory of TIM behaviour proposed by Schrøder-Petersen et al. (2003a) to the animals in our study then TIM was exhibited across a range from normal to abnormally elevated frequency, with the threshold for injurious tail biting reached only in two of our groups. Tail biting prevalence may have stayed low due to relatively low-stress housing and husbandry, including plenty of space, stable four-animal groups of siblings, no road transport and only minor need to compete for resources. The current animals were offered bedding material according to four different regimes, which were controlled for in the statistical analyses. As environmental enrichment is known to affect TIM behaviour (Schrøder-Petersen et al., 2003a), the present results will not reveal effects caused by interactions between enrichment status and category of pigs. Gender has repeatedly been associated with tail-directed behaviours. Schrøder-Petersen et al. (2003a, 2004) reported tendencies for females to perform higher frequencies of TIM and for males to receive the behaviour, which is also true for tail biting (e.g. Kritas and Morrison, 2004; Zonderland et al., 2010). The current results are in line with these findings, although statistical significance was not reached.
7
Body weight or growth did not differ between the categories at 4, 7 or 9 weeks of age in agreement with Schrøder-Petersen et al. (2003b). P were, however, born significantly smaller than R. The importance of this result is uncertain, as birth weight to our knowledge has not previously been reported to associate with TIM or tail biting behaviour (Beattie et al., 2005). A few studies have associated performing tail biting with a low and receiving the behaviour with a high body weight (Beattie et al., 2005; Zonderland et al., 2010). Although a low birth weight is known to predict slow growth and/or low weight at least until slaughter weight (Gondret et al., 2005; Fix et al., 2010), no such effects were found in the present data, perhaps due to the selection process not accepting animals representing the extreme ranges of weight in the litters. It should be noted that the statistical analyses of the current data included a large number of comparisons, increasing the risk for false positive results. The conclusions drawn below regarding intercategory behavioural differences were, fortunately, supported by more than one significant result. 4.2. General activity and exploration of the environment Category differences in performed behaviour were evident in the first day after weaning, suggesting that P, R and N represented different phenotypes of pigs. As compared to R, P showed a smallscale but consistently increased level of overall activity over 24 h, and spent more time exploring the environment during the day. These results appear to agree with Ursinus et al. (2014) showing that pen-level increased levels of exploration or overall activity predicts injurious tail biting weeks or months later. Conflicting result were, however, presented by Statham et al. (2009) showing that overall activity did not predict tail biting outbreak until four days prior. Given that TIM is considered an expression of the motivation for both social and environmental exploration (Schrøder-Petersen et al., 2003a,b; Simonsen, 1995; Newberry and Wood-Gush, 1988), the present results may also suggest that this motivation is elevated in P as compared to R already before TIM was observed. Considering that behaviour at 4 weeks of age was observed during a very stressful time the day after weaning, removal of a number of pen mates and transfer to a new pen, the present results may be interpreted as evidence for P and R representing two extremes of coping strategy. These extremes have been characterized by active and passive behaviour in challenging situations, and are referred to as proactive and reactive, respectively (reviewed by Koolhaas et al., 1999). Proactive copers are thought to have an increased risk for becoming tail biters, and reactive copers for becoming victims (Korte et al., 2009; Zupan et al., 2012). Although no formal coping style testing was performed, the present results would support this theory, if TIM is considered a precursor for tail biting behaviour. 4.3. Social behaviour at weaning before evidence of TIM Daytime social activity did not differ between the categories of pigs at 4 weeks of age, whereas some differences existed in the type of activity. P performed more tail-nosing than N, suggesting a difference in interest for the anogenital or tail area already at young age. Numerical differences in exploration may suggest that the increase in tail-nosing was a by-product of increased environmental exploration in P. Dyad-level analyses showed that social activity was reciprocal and without category-level preferences at 4 weeks of age, whereas the attractiveness as a receiver differed between categories. N received significantly more overall social behaviours and tail-nosing, and non-significantly more nosing than P. The reason for this is unclear. N was not by any measure more active or more passive than the other categories and did thus not provoke more attention or act as stationary object for exploration. N and P may
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
8
thus exhibit different avoidance strategies in response to social approaches by others. 4.4. Social behaviour at 9 weeks of age when TIM behaviour had emerged All categories of pigs became more active when comparing behaviour at the ages of 4 and 9 weeks, which is an expected effect of increasing age (Cox and Cooper, 2001). The observation that different behaviours changed in the different categories does, however, indicate that phenotypic differences existed. Inter-category differences in behavioural development were more clear in daytime analyses than time budgets, which is in line with Larsen et al. (2016) reviewing literature on predicting tail biting to conclude that behavioural differences between pens with and without tail biting in the near future may be more pronounced during the active part of the day. In P a part of the active time was reassigned from environmental exploration to social activity when comparing daytime behaviour at 4 and 9 weeks of age. The increase in social behaviour included significant increases in tail nosing, TIM, ear biting and fighting, all of which were performed to relatively high frequency and/or long duration by P as compared to the other categories at 9 weeks of age. Dyad-level analyses show that P went from no preference for social actions at 4 weeks to a significant preference for R over N at 9 weeks, when P also directed more social actions towards R than vice versa. Similar changes in behaviour have not been described before, although Zonderland et al. (2011) reported preferences in tail-directed behaviours by tail biters during the last six days before the first tail wound in a group. The preference by P for R over N at 9 weeks may have been affected by actions of N. N may have avoided the increased social activity by P, perhaps thereby showing a characteristic of the tail biting – resistant phenotype described by Brunberg et al. (2013). This theory is supported by the low amount of received tail nosing and overall received social behaviour in N at 9 weeks of age. Although R may have attracted attention due to wagging of a subclinically injured tail (Simonsen 1995; Munsterhjelm et al., 2013), the data does not support this as N showed no particular interest for R. N did, in contrast, significantly increase the duration of social behaviour directed towards P from 4 to 9 weeks of age to the degree of a significant preference for P over R. This increase in P-directed behaviour may have been provoked by increased social activity by P, however, the present data does not allow for an investigation on this. The behaviour of R changed to a relatively low degree from 4 to 9 weeks of age. Daytime social activity remained unchanged, in contrast to the other categories showing a significant increase. At 9 weeks of age R was characterized by marked social inactivity, in spite of receiving more social behaviour than the other categories. Both P and N directed more social behaviour towards R than vice versa, leaving R the only category performing less social behaviour than it received. This suggests that a large proportion of performed actions were responses to approaches by others, mainly P, as dyadlevel analyses show that R had a significant preference for P over N as a receiver of its social actions. The observed social inactivity in R may be characteristic of a phenotype predisposed to receive abnormal behaviour by others, as described by Keeling et al. (2004). 5. Conclusions This study shows differences in social and exploratory behaviours in TIM-performers, -receivers and neutral pigs already before the behaviour is observed in the group, to suggest that the categories represent different phenotypes. Emergence of TIM
behaviour coincides with profound changes in social relationships in the group including increased activity in performers, directed mainly towards the receivers; and marked social passivity in receivers.
Acknowledgements The authors are grateful for funding from the Finnish Ministry of Agriculture and Forestry in support of the ANIWHA ERA-Net initiative through the FareWellDock-project, as well as the Welfare and Production of the Pig –project, which also was funded by Snellman Ab and Raisio feed Ltd.
References Beattie, V., Breuer, K., O’Connell, N.E., Sneddon, I.A., Mercer, J.T., Rance, K.A., Sutcliffe, M.E.M., Edwards, S.A., 2005. Factors identifying pigs predisposed to tail biting. Anim. Sci. 80, 307–312. Beattie, V.E., 1994. The Effects of Environmental Enrichment on the Domestic Pig Ph.D. Thesis. The Queen’s University of Belfast, Northern Ireland. Blackshaw, J.K., 1981. Environmental effects on lying behaviour and use of trough space in weaned pigs. Appl. Anim. Ethol. 7, 281–286. Brunberg, E., Jensen, P., Isaksson, A., Keeling, L.J., 2013. Behavioural and brain gene expression profiling in pigs during tail biting outbreaks −evidence of a tail biting resistant phenotype. PLoS One 8, e66513, http://dx.doi.org/10.1371/ journal.pone.0066513. Cox, L.N., Cooper, J.J., 2001. Observations on the pre- and post-weaning behaviour of piglets reared in commercial indoor and outdoor environments. Anim. Sci. 72, 75–86. Edwards, S.A., 2006. Tail biting in pigs: understanding the intractable problem. Vet. J. 171, 367–369. Feddes, J.J.R., Fraser, D., 1994. Non-nutritive chewing by pigs – implications for tail-biting and behavioural enrichment. Trans. Am. Soc. Agric. Eng. 37, 947–950. Feddes, J.J., Fraser, D., Buckley, D.J., Poirier, P., 1993. Electronic sensing of non-destructive chewing by growing pigs. Trans. Am. Soc. Agric. Eng. 36, 955–958. Fix, J.S., Cassady, J.P., Herring, W.O., Holl, J.W., Culbertson, M.S., See, M.T., 2010. Effect of piglet birth weight on body weight growth, backfat, and longissimus muscle area of commercial market swine. Livest. Sci. 127, 51–59. Fraser, D., 1987. Attraction to blood as a factor in tail-biting by pigs. Appl. Anim. Behav. Sci. 17, 1–2. Gondret, F., Lefaucheur, L., Louveau, I., Lebret, B., Pichodo, X., Le Cozler, Y., 2005. Influence of piglet birth weight on postnatal growth performance: tissue lipogenic capacity and muscle histological traits at market weight. Livest. Prod. Sci. 93, 137–146. Keeling, L., Andersson, L., Schütz, K.E., Kerje, S., Fredriksson, R., Carlborg, Ö., Cornwallis, C.K., Pizzari, T., Jensen, P., 2004. Chicken genomics: feather-pecking and victim pigmentation. Nature 431, 645–646. Koolhaas, J.M., Korte, S.M., De Boer, S.F., Van Der Vegt, B.J., Van Reenen, C.G., Hopster, H., De Jong, I.C., Ruis, M.A., Blokhuis, H.J., 1999. Coping styles in animals: current status in behavior and stress-physiology. Neurosci. Biobehav. Rev. 23, 925–935. Korte, S.M., Prins, J., Vinkers, C.H., Olivier, B., 2009. On the origin of allostasis and stress-induced pathology in farm animals: Celebrating Darwin’s legacy. Vet. J. 182, 378–383. Kritas, S.K., Morrison, R.B., 2004. An observational study on tail biting in commercial grower-finisher barns. JSHP 12 (1), 17–22. Larsen, M.L., Andersen, H.M., Pedersen, L.J., 2016. Can tail damage outbreaks in the pig be predicted by behavioural change? Vet. J. 209, 50–56. Munsterhjelm, C., Peltoniemi, O.A., Heinonen, M., Hälli, O., Karhapää, M., Valros, A., 2009. Experience of moderate bedding affects behaviour of growing pigs. Appl. Anim. Behav. Sci. 118, 42–53. Munsterhjelm, C., Simola, O., Keeling, L., Valros, A., Heinonen, M., 2013. Health parameters in tail biters and bitten pigs in a case-control study. Animal 7, 814–821. Newberry, R.C., Wood-Gush, D.G.M., 1988. Development of some behaviour patterns in piglets under semi-natural conditions. Anim. Prod. 46, 103–109. Statham, P., Green, L., Bichard, M., Mendl, M., 2009. Predicting tail-biting from behaviour of pigs prior to outbreaks. Appl. Anim. Behav. Sci. 121 (3), 157–164. Schrøder-Petersen, D.L., Simonsen, H.B., Lawson, L.G., 2003a. Tail-in-mouth behaviour among weaner pigs in relation to age, gender and group composition regarding gender. Acta Agric. Scand. Sect. A: Anim. Sci. 53, 29–34. Schrøder-Petersen, D.L., Simonsen, H.B., Lawson, L.G., 2003b. Ontogeny of tail-in-mouth behaviour in piglets reared in barren or enriched environments. Proceedings of the 37th International Congress of the International Society for Applied Ethology, 115. Schrøder-Petersen, D.L., Heiskanen, T., Ersbøll, A.K., 2004. Tail-in-mouth behaviour in slaughter pigs, in relation to internal factors such as: age, size, gender, and motivational background. Acta Vet. Scand. Sect. A: Anim. Sci. 54, 159–166.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
G Model APPLAN-4322; No. of Pages 9
ARTICLE IN PRESS C. Munsterhjelm et al. / Applied Animal Behaviour Science xxx (2016) xxx–xxx
Simonsen, H.B., 1995. Effect of early rearing environment and tail docking on later behaviour and production in fattening pigs. Acta Agric. Scand. Sect. A: Anim. Sci. 45, 139–144. Taylor, N.R., Main, D.C., Mendl, M., Edwards, S.A., 2010. Tail-biting: a new perspective. Vet. J. 186, 137–147. Ursinus, W.W., Van Reenen, C.G., Reimert, I., Bolhuis, J.E., 2014. Tail biting in pigs: blood serotonin and fearfulness as pieces of the puzzle? PLoS One 9, 1–14.
Zonderland, J.J., Bracke, M.B., den Hartog, L.A., Kemp, B., Spoolder, H.A., 2010. Gender effects on tail damage development in single-or mixed-sex groups of weaned piglets. Livest. Sci. 129 (1), 151–158. Zonderland, J.J., Schepers, F., Bracke, M.B.M., Den Hartog, L.A., Kemp, B., Spoolder, H.A.M., 2011. Characteristics of biter and victim piglets apparent before a tail-biting outbreak. Animal 5 (05), 767–775. Zupan, M., Janczak, A.M., Framstad, T., Zanella, A.J., 2012. The effect of biting tails and having tails bitten in pigs. Phys. Behav. 106, 638–644.
Please cite this article in press as: Munsterhjelm, C., et al., Can tail-in-mouth behaviour in weaned piglets be predicted by behaviour and performance? Appl. Anim. Behav. Sci. (2016), http://dx.doi.org/10.1016/j.applanim.2016.08.013
9