The social and feeding behaviour of growing pigs in deep-litter, large group housing systems

Applied Animal Behaviour Science 82 (2003) 173–188

The social and feeding behaviour of growing pigs in deep-litter, large group housing systems R.S. Morrison a,∗ , P.H. Hemsworth a,b , G.M. Cronin b , R.G. Campbell c a

b

Institute of Land and Food Resources, University of Melbourne, Animal Welfare Centre, Parkville, Vic. 3010, Australia Department of Primary Industries, Victorian Institute of Animal Science, Animal Welfare Centre, Werribee, Vic. 3030, Australia c United Feeds, P.O. Box 108, Sheridan, IL 46069, USA Accepted 5 March 2003

Abstract Two studies compared the social and feeding behaviour of Large White × Landrace entire male (non-castrated) growing pigs housed in deep-litter, large group and conventional housing systems. In Study 1, an ethogram of behaviour was developed. The pigs were housed at 1 m2 per pig (200 pigs per pen) and 8.3 pigs per feeding space in the deep-litter system and 0.70 m2 per pig (45 pigs per pen) and 8.5 pigs per feeding space in the conventional system. In Study 2, the social and feeding behaviour of entire male growing pigs were compared. Eight hundred and eighty crossbred entire male pigs were used. There were 200 pigs per pen (1 m2 per pig; 8.3 pigs per feeding space) in the deep-litter system and 20 pigs per pen (0.49 m2 per pig; 10 pigs per feeding space) in the conventional system. When the average pig live weight was 60 kg, five non-focal pigs were removed from each pen resulting in 0.65 m2 per pig (7.5 pigs per feeding space). In both studies, ad libitum feed was provided in double-spaced, wet–dry feeders. In each study there were four replicates, with 10 focal animals per treatment per replicate randomly selected for behaviour observations. All observations were conducted during day light. In Study 1, behaviour was observed when pigs were 21–22 weeks of age and in Study 2 social behaviour observations were conducted from 19 to 22 weeks of age, while feeding behaviour was observed from 20 to 22 weeks of age. Pigs housed in the deep-litter system spent more (P < 0.01) time standing and less (P < 0.05) time sitting and lying, and spent more (P < 0.001) time in locomotion compared to conventionally housed pigs. There was an increase (P < 0.05) in physical pen interactions and a higher (P < 0.001) incidence of social tactile interactions and agonistic (P < 0.001) and sexual behaviours (P < 0.01)

∗ Corresponding author. Present address: West Central Research and Outreach Center, University of Minnesota, State Highway 329, Morris, MN, USA. Tel.: +1-320-589-1711; fax: +1-320-589-4870. E-mail address: [email protected] (R.S. Morrison).

0168-1591/03/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0168-1591(03)00067-4

174

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

in the deep-litter treatment. Such differences in social behaviour may be responsible for differences in growth performance commonly observed in deep-litter systems. Pigs in the deep-litter treatment spent less time (P < 0.05) within 1 m of the feeder, had fewer (P < 0.05) feeding events, had a longer (P < 0.05) duration of feeding and a lower (P < 0.05) frequency of social interactions around the feeder compared to pigs in the conventional treatment. The difficulties in gaining and maintaining access to the feeder in the conventional pens may be responsible for the shorter but more frequent feeding bouts observed. It is possible that the relatively unrestricted feeding that occurred in the deep-litter treatment may have resulted in increased fat deposition and poorer feed conversion efficiency, as observed in the industry setting, because of changes in feeding patterns that affect feed utilisation. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Pig—social behaviour; Feeding behaviour; Deep-litter; Large groups; Group size

1. Introduction Deep-litter, large group housing systems for growing pigs have been developed as an alternative housing system. Large groups of pigs in such alternative housing systems are usually kept on deep-litter between 2 and 8 weeks of age (∼5–20 kg live weight) till slaughter (∼22 weeks of age, live weight ∼110 kg). Conventionally however, growing pigs are housed in a more confined system indoors and with automated ventilation, fully or partially slatted floors and liquid manure handling systems. The number of pigs per pen tends to be in the range of 5–50 pigs with a floor space allowance of a maximum of about 0.7 m2 per pig. In contrast, deep-litter systems are naturally ventilated, have a floor base of deep-litter consisting of rice hulls or straw, accommodate larger group sizes (ranging from 150 to 2000 pigs per pen) and the pigs have a greater space allowance of approximately 1 m2 per pig. Deep-litter systems are cheaper to establish and are perceived as being more “welfarefriendly” for pigs and more environmentally sustainable, as the need for effluent ponds is eliminated and the deep-litter substrate is used as a natural fertiliser. However, recent industry records indicate that pigs grown from 2 weeks of age to slaughter at 22 weeks in deep-litter large group systems show a number of growth performance problems compared to conventionally housed pigs (Connor, 1995; Payne, 1997; Brumm, 1999; Honeyman et al., 1999; Payne et al., 2000). For instance, pigs in the former systems are 10% less efficient in converting feed provided to live weight gain (feed:gain) and may be 1–2 mm greater backfat. Deep-litter systems are disparate in the type and amount of bedding provided, the number of pigs per group, floor space allowance per pig, method of providing feed and water and the construction materials used to build the deep-litter, large group system. Few rigorous comparisons have been conducted between the various modifications of deep-litter systems or between deep-litter, large group and conventional housing systems. Nevertheless, while many key factors in addition to the basic features of the system will vary, such as climate, disease status and location, it is possible that the apparent growth performance differences between deep-litter, large group and conventional housing systems are largely behavioural and possibly stress related.

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

175

The scientific literature on comparative effects of deep-litter and large group sizes on social and feeding behaviour and stress physiology of growing pigs is very limited. There is a plethora of information on the effects of environmental enrichment on the social behaviour and growth performance of growing pigs (Wood-Gush and Vestergaard, 1989a,b; Beattie et al., 1995, 1996, 2000a,b; Peterson et al., 1995; O’Connell and Beattie, 1999), however the majority of research has been conducted in conventional systems with smaller groups of pigs. Furthermore, little is known of the social behaviour of pigs in large groups in deep-litter systems. Aggressive behaviours may be higher in larger groups of pigs, as individual animal recognition may not be functional (Spoolder et al., 1999). On the other hand, aggression may be reduced on deep-litter pigs with the attention of the pigs redirected from pen mates to the litter. Furthermore, the greater availability of resources in these large groups may reduce the need for aggression (Hemsworth and Barnett, 2001), or the pigs may be socially tolerant in these large groups (Hughes et al., 1997). Factors that may affect the feeding and social behaviour and performance of pigs in deep-litter systems, such as the provision of litter, increased group sizes and space allowance, require attention. To compare the behaviour of growing pigs in deep-litter, large group systems to that of pigs in conventional systems, it is necessary to conduct controlled studies in which inputs unrelated to the system, such as locality, feed, feeder type and genetics are similar. Therefore, the aim of the present studies was to examine the social and feeding behaviour of entire male growing pigs in deep-litter, large group housing systems compared to pigs in conventional housing systems. An ethogram of the behaviour of growing pigs in deep-litter and conventional housing systems was developed in Study 1, to enable a comparative study of social and feeding behaviour in Study 2, while controlling for locality, feed, feeder type and genotype.

2. Materials and methods 2.1. Animals and housing treatments The care and experimental use of pigs in these experiments conformed to the requirements of the Australian Prevention of Cruelty to Animals Act 1986 and the NHMRC “Australian code of practice for the care and use of animals for scientific purposes”. These studies were conducted in a deep-litter, large group housing system and a conventional housing system at a large commercial piggery in Corowa, NSW, Australia (36◦ S latitude, 146.5◦ E longitude). The two housing treatments were located approximately 3 km apart. 2.1.1. Study 1 Nine hundred and eighty crossbred (Large White×Landrace) entire male (non-castrated) growing pigs were used in this study. The pigs in both housing treatments came from a common source in which conventional farrowing crates were used. The pigs were introduced to either the deep-litter, large group or conventional housing systems at 8 weeks of age and remained in treatment until slaughter at 22 weeks of age. Behaviour observations were conducted when the pigs were 21 weeks old in both housing treatments. Commercial male finisher diet (13.7 MJ/kg DE, 14.7% crude protein and 0.5 g/(MJ kg) DE available lysine) in

176

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

a pelleted form was provided ad libitum in double-spaced wet–dry feeders in both housing treatments during the behaviour observation period. The study was conducted during autumn when the average daily temperatures ranged from a minimum of 10 ◦ C to a maximum of 21 ◦ C. Both buildings were open-sided with natural ventilation. Blinds positioned on the four sides of each building were automatically controlled with the use of a thermostat. The two housing treatments studied were: • Deep-litter, large group system: four pens measuring 20 m × 10 m, housing 200 pigs per pen at 1 m2 floor space per pig, were located within the building. There were 8.3 pigs per feeding space. Rice hull bedding was provided at a rate of 0.7 kg per pig per day and was added before the pigs were introduced to the deep-litter treatment at 8 weeks of age. The floor base under the rice hulls was graded earth with a cement render. Bedding was not replenished during the study. • Conventional housing system: four pens measuring 4.5 m × 7 m and housing 45 pigs per pen with 0.7 m2 floor space per pig were used. There were 8.5 pigs per feeding space. The concrete pen floor was two-third solid and one-third slatted. 2.1.2. Study 2 Eight hundred and eighty crossbred (Large White × Landrace) entire male growing pigs were used. The source and previous management (in terms of genotype, nutrition and health) of the pigs was identical to that in Study 1. The study was conducted during winter when the average daily temperatures ranged from a minimum of 5 ◦ C to a maximum of 12 ◦ C. The two housing treatments were: • Deep-litter, large group system: this was identical to the system used in Study 1. • Conventional housing system: four pens measuring 2.7 m × 3.6 m were used in the study. Twenty pigs were housed in each pen at the start of the behaviour observations (at 18 weeks of age), with a space allowance of 0.49 m2 per pig and 10 pigs per feeding space. When average pig live weight reached 60 kg, five non-focal pigs were removed from each pen to provide a space allowance of 0.65 m2 per pig and 7.5 pigs per feeding space, that is, the acceptable space allowance for pigs of this weight. This process simulated moving from the grower to finisher pens in the conventional system. The environmental conditions and flooring type were similar to those in Study 1. 2.2. Behaviour observations 2.2.1. Study 1 Focal-animal sampling, as described by Jensen et al. (1986), was used to observe behaviour. Ten pigs per pen were randomly selected as focal animals at the start of the study. Focal pigs were identified by a large coloured ear tag in each ear and coloured spray marking (Dulux® Quick dry SpraypakTM paint, East Melbourne, Australia) which was applied the day before observations. Each pen was observed four times over a 2-week period. The sequence in which pens were observed was randomised, with one deep-litter, large group pen and one conventional pen observed each day. Every observation day was divided into two sessions; 2 h in the morning (from 07:30 h) and 2 h in the afternoon (from 15:00 h).

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

177

During each observation session each focal pig was randomly selected and continuously observed for 6 min using a portable video camera. 2.2.2. Study 2 2.2.2.1. Activity and social behaviour observations. The social behaviour of the pigs was recorded by the main author using the direct observation technique via “The Observer® ” (Noldus Information Technology, 1995) package on a portable computer. Observation platforms were erected 3 m above the deep-litter and conventional pens to reduce distractions to the pigs. Focal-animal sampling, and means of identification were identical to Study 1. Each pen was observed eight times over a 4-week period. The sequence in which pens were observed was randomised. Every observation day was divided into two, 2 h sessions identical to Study 1 and during each observation session each focal pig was continuously observed for 6 min in a random sequence. Social behaviour was measured using the ethogram developed in Study 1. Postures were recorded as the percentage of observation time spent standing, sitting or lying. In addition, the time spent by pigs in the standing posture was used as an estimate of total activity. The activity and social behaviours were pooled into the following categories: idle, locomotory (walking, frolicking and trotting), non-social physical pen interactions (jumping, nose pen fixtures, nosing concrete, rooting, rubbing and rolling), agonistic behaviours (parallel pressing, head-to-head knocks and levering—performed and received), sexual behaviours (mounting—performed and received) and social tactile interactions (anal nosing, nose-to-body, nose-to-nose and pushing—performed and received). Means (±S.E.D.) for the duration of the social behaviour were expressed as percentage of observation time and the frequency of behaviours were expressed as number of occurrences in 6 min using a bout criterion interval of 5 s (Jensen et al., 1986). Feeding behaviour was recorded using time-lapse video with cameras positioned above the feeding areas of both the deep-litter and conventional housing treatments. Each pen was video recorded once per week, for 2 weeks prior to slaughter at 23 weeks of age. Recordings were made during day light hours (between about 07:30 h and 17:30 h). The video-taped data were viewed and feeding behaviour was measured using the following parameters: • Total time within 1 m of feeder (s): the total time spent within 1 m of the feeder. • Total time feeding (s): the total time the pig’s head was in the feeder (assumed to be feeding). • Number of feeding events: the number of times a pig put its head in the feeder (assumed to be feeding). • Average duration of feeding events (s): the total time feeding divided by the number of feeding events. • Number of social interactions within 1 m of the feeder: number of social tactile interactions performed and received (i.e. pushing, head-to-head knocks, anal nosing, levering, mounting, nose-to-body, nose-to-nose, parallel pressing and pushing) within 1 m of the feeder.

178

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

Feeder occupancy during day light was measured by counting the number of pigs (including non-focal pigs) with their head in the feeder at 10 min intervals during the feeding observation period. 2.3. Video data analysis and statistical analysis In Study 1, descriptive terms were chosen to describe the behaviours observed. Ethograms of growing pigs by McGlone (1985), Stolba and Wood-Gush (1989) and Andresen and Redbo (1999) and group-housed dry sows by Jensen (1980, 1982) were used to assist in analysis and descriptions of the behaviour. In Study 2, the behaviour frequency counts and duration data were analysed by one-way analysis of variance using only the treatment effect, using Minitab (Minitab Inc., 1996, Release 11 for WindowsTM ) statistical software. The experimental unit was the pen group.

3. Results 3.1. Study 1 The following ethogram was defined for entire male growing pigs in the deep-litter, large group system and the conventional system. 3.1.1. Postures • Standing: the pig is upright on all four legs. The pig can be stationary or moving about. • Sitting: the pig is in an upright position, with its back legs bent such that the pig supports its weight on its hind quarters and has its’ fore legs straight. • Lying: the pig is recumbent on its belly (sternal recumbency) or side (lateral recumbency). 3.1.2. Behavioural states/events • Locomotory ◦ Walking: a slow regular symmetrical gait where the left legs perform the same movement as the right, but half a stride later, in which two of the four legs support the pig at any one time. The sequence of leg movements is left front, right hind, right front and left hind as described by Fraser and Broom (1998). ◦ Trotting: a symmetrical gait of medium speed in which the pig is supported by alternating diagonal pairs of legs. The sequence of leg movements is left front and right hind, then right front and left hind as described by Fraser and Broom (1998). ◦ Frolicking: the pig gallops along, sometimes travelling in a circular motion, moving around the body axis. This behaviour only occurs for a short period of time (deep-litter system only). • Maintenance ◦ Drinking: the pig stands with its mouth over the drinker nozzle, assumed to be drinking. ◦ Feeding: the pig stands with its head in the feeder, assumed to be feeding.

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

179

◦ Urination: the pig stands stationary, urinating (Fraser and Broom, 1998). ◦ Defecation: the pig stands stationary, defecating (Fraser and Broom, 1998). 3.1.3. Tactile interactions • Social tactile interactions: the focal animal can either be the performer or receiver of listed social interactions. ◦ Anal nosing: the nose of a pig approaches within 5 cm of the anus of another pig and makes physical contact. ◦ Nose-to-body: the nose of a pig approaches within 5 cm of any part of the body behind the ears, apart from the anal region, and makes physical contact. ◦ Nose-to-nose: the nose of a pig approaches and has physical contact with the nose, head or ears of another pig. ◦ Pushing: pushing through a group of other pigs, without apparent interest in making social contact, often to get to the feeder. • Agonistic behaviour: the focal animal can either be the performer or receiver of the following social interactions. ◦ Parallel pressing: two pigs stand side-by-side, pressing against each other’s shoulder, with one throwing its head against the head or the neck of the other pig. Pigs tend to champ their jaws causing saliva to froth. ◦ Head-to-head knock: a rapid thrust upward or sideways with the head or snout against the neck, head or ears of another pig. ◦ Levering: the pig puts its snout under the body of another pig (from behind or the side) and lifts the pig into the air. • Sexual behaviour: the focal animal can either be the performer or receiver of the following social interactions. ◦ Mounting: the pig places both front legs over the front or back of a pig. While mounted on the pig it may thrust its hips, sometimes achieving an anal intromission with its penis and ejaculating either in the pig’s rectum or on the pig’s back. • Non-social physical pen interactions ◦ Jumping: the pig jumps down from the concrete ledge onto the bedding substrate (deep-litter system only). ◦ Nose pen fixtures: the pig’s snout approaches within 5 cm and touches a part of the pen other than the floor. ◦ Nosing concrete: the snout of a pig approaches within 5 cm of the concrete in the pen, and has physical contact with the concrete. ◦ Rooting: the pig uses its snout to burrow into the deep-litter bedding (deep-litter system only). ◦ Rubbing: the pig rubs the side of its body along a pen fixture. ◦ Rolling: the pig lies down and moves sideways (deep-litter system only). 3.1.4. Other • Idle: the pig is not performing the behavioural repertoire mentioned in the current ethogram. The pig may be standing whilst idle.

180

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

Table 1 Postures and main behaviours of entire male growing pigs observed away from the feeder in Study 2 Deep-litter, large group

Conventional

S.E.D.

Postures Standing Sitting Lying

44.10 c 1.60 a 54.30 a

25.10 d 6.50 b 68.40 b

4.48 1.81 5.05

Behaviours Idle Locomotory behaviour (%) Physical pen interactions (frequency) Social tactile interactions (frequency) Agonistic behaviour (%) Sexual behaviour (%)

54.30 a 10.40 e 9.40 a 8.30 e 4.30 e 1.80 c

68.40 b 2.00 f 4.60 b 3.70 f 0.70 f 0.60 d

5.06 0.76 1.52 0.60 0.48 0.30

Values shown are the mean (±S.E.D.) duration (% of observation time) and frequency (times per 6 min per pig) of occurrence for the deep-litter, large group and conventional housing treatments. Within rows, means with different letters (a, b, c, d, e, f) are significantly different at P < 0.05, P < 0.01 and P < 0.001, respectively.

3.2. Study 2 There were significant differences in the time pigs occupied the three postures (Table 1). Pigs housed in the deep-litter, large group system spent more time (P < 0.01) standing and, probably as a consequence, less (P < 0.05) time sitting and lying, compared to the conventionally housed pigs. Activity levels, based on time spent standing and locomotive and social behaviour of pigs, also differed between the treatments (Table 1). Pigs in the deep-litter, large group treatment spent more (P < 0.001) time in locomotion than pigs in the conventional treatment and there was a higher frequency of exploratory behaviours recorded for pigs, as shown by an increase (P < 0.05) in non-social physical pen interactions (Table 1). Pigs housed in the deep-litter system were also recorded to perform higher incidences of social tactile interactions away from the feeding area (P < 0.001), and agonistic (P < 0.001) and sexual behaviours (P < 0.01).

Table 2 Mean (±S.E.D.) feeding behaviour of entire male growing pigs housed in deep-litter, large group and conventional housing treatments in Study 2

Total time within 1 m of feeder spaces (s/10 h) Total time feeding (s/10 h) Number of feeding events Average duration of feeding events (s/10 h) Number of social interactions within 1 m of the feeder Average % of feeders occupied during day light 07:30 h to 17:30 h

Deep-litter, large group

Conventional

S.E.D.

3776 a 1554 30.5 a 52.4 a 17.0 a 43.28

6483 b 1724 52.9 b 33.3 b 37.8 b 53.77

759.9 198.5 4.32 6.96 8.19 4.53

Within rows, means with different letters (a, b) are significantly different at P < 0.05.

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

181

Fig. 1. Daily feeding pattern in Study 2. Percentage of feeders occupied throughout the day in the deep-litter, large group (䉬) and conventional (䊏) housing treatments.

Pigs in the deep-litter, large group treatment spent less time (P < 0.05) within 1 m of the feeder, had fewer (P < 0.05) feeding events and a longer (P < 0.05) average duration of feeding events compared to pigs in the conventional treatment (Table 2). Pigs in the deep-litter, large group treatment also had a lower (P < 0.05) frequency of social interactions around the feeder than those in the conventional treatment. There was a tendency (P = 0.06) for the average percentage of feeders occupied during day light to be lower in the deep-litter, large group than conventional treatment (Fig. 1).

4. Discussion The behaviours described in the ethogram from Study 1 were identified as relevant behaviours to study differences in the social and feeding behaviour of entire male growing pigs in deep-litter, large group systems and conventional systems in Study 2. The behaviours defined in the ethogram are similar to behaviours observed by Kelley et al. (1980), McGlone (1985) and Stolba and Wood-Gush (1989) in growing pigs and by Jensen (1980, 1982) in group-housed dry sows. Social tactile interactions are important in maintaining social organisation in pigs. Hafez (1975) described pigs as ‘contact’ animals, that is, animals which tend to rest in contact with each other. Social tactile interactions such as nose-to-body, nose-to-nose and anal nosing have been described in group-housed dry sows by Jensen (1980, 1982) and were also observed in Study 2 in growing pigs. Jensen (1980) concluded that these social tactile interactions are involved in individual animal recognition, which is important when group-housed animals need to establish effective competitive and cooperative social relationships (Stricklin and Mench, 1987). A common estimate of the total number of group members that can be recognised by an individual is 20–30 pigs (Fraser and Broom, 1998). It is still uncertain how the mechanism of individual animal recognition operates in pigs, although it is evident that different types of recognition exist (Fraser and Broom, 1998).

182

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

According to Hart (1985), pigs identify other individuals within a group using auditory, olfactory, tactile and visual cues. The pigs in the deep-litter, large group treatment in Study 2 had a higher frequency of social tactile interactions (anal nosing, nose-to-body, nose-to-nose and pushing) and spent significantly more time performing agonistic behaviour (parallel pressing, head-to-head knocks and levering), compared to their conventionally housed counterparts. These results are not in agreement with Lay et al. (2000) who found a higher incidence of aberrant behaviours in conventionally housed pigs compared to pigs raised in a deep-litter, large group system. Agonistic behaviour includes behaviours that occur in response to a conflict and include offence, defence, submissive or escape components which may involve physical contact (e.g. biting or pushing) or no contact at all (e.g. body postures or facial gestures) between individuals (Petherick and Blackshaw, 1987). Thus, agonistic behaviour refers to species-typical threats and submissive behaviours, as well as conflict resolution through fighting. Aggressive behaviour is often used interchangeably with agonistic behaviour and in the strict definition refers to the attack and actual fighting (Hart, 1985; Fraser and Broom, 1998). Parallel pressing, head-to-head knock and levering are defined as aggressive behaviours and have been reported by McBride et al. (1964), Beilharz and Cox (1967), Hafez and Signoret (1969), Signoret et al. (1975), Jensen (1980) and Fraser and Broom (1998). Agonistic behaviour in entire males has been described by Hafez and Signoret (1969) and these authors considered agonistic behaviour to be part of the establishment of the social hierarchy. Pigs in the deep-litter, large group housing treatment in Study 2 also spent significantly more time in locomotory behaviour (trotting, walking and frolicking). This may have contributed to the increase in pig social tactile interactions in the deep-litter system as being more active, pigs in the deep-litter system are more likely to frequently encounter and interact with other pigs. McKinnon et al. (1989) studied weaner pigs housed with or without straw in groups of 13 pigs, and found that while pigs were more active when provided with straw there were more social tactile interactions without straw. Another explanation for the increased tactile social interactions observed in the deep-litter, large group treatment in Study 2 may be associated with problems with the social structure of the group. The social stability of a group of pigs relies on the development of a stable dominance hierarchy, which is usually established after mixing and results in reduced aggression within a few hours (Spoolder et al., 1999). However, the very large group size in the deep-litter treatment may have affected the establishment of a stable dominance hierarchy. In order for a group of animals to establish a dominance hierarchy, it is necessary for the animals to promptly identify each other (Stricklin and Mench, 1987). It has been hypothesised that group size is positively correlated with a failure to resolve rank disputes without aggression (Al-Rawi and Craig, 1975) because of recognition problems in large groups. Spoolder et al. (1999) studied groups of 20, 40 and 80 pigs found that larger groups had significantly more skin lesions. Skin lesions were not measured in Study 2, although industry evidence suggests that large groups of pigs housed in deep-litter systems have more carcass damage than conventionally housed pigs (Payne et al., 2000). In larger groups, individuals are more likely to encounter other pigs which they do not recognise or with which they have to reconfirm their relative rankings, resulting in aggression and more carcass bruising (Spoolder et al., 1999). Studies conducted on laying hens (Bilcik and

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

183

Keeling, 2000) have shown that the highest levels of feather pecking and aggressive attacks were in the largest group of hens. Fraser and Rushen (1987) reviewed the literature on aggression in pigs and stated that increases in group size and a reduction in floor space were associated with increased aggression. However, the majority of studies on pigs involved group sizes of 60 or less, which is substantially less than the group size in the current studies. On the other hand, Pagel and Dawkins (1997), Hughes et al. (1997), Estevez (1998) and Nicol et al. (1999) showed that aggressive behaviour declined as group size increased. The authors suggested that the greater availability of resources such as total free space, availability of feeding places and preferred lying areas may eliminate the need for a dominance hierarchy, which functions to control aggression when resources are limited (Hemsworth and Barnett, 2001). Perhaps animals in large groups are more socially tolerant (Hughes et al., 1997) and abandon all attempts to establish social hierarchies (Pagel and Dawkins, 1997). The studies conducted by Spoolder et al. (1999) on the effects of group size utilised pens without deep-litter or straw. The authors suggested that perhaps social tactile interactions would be less if straw was present, as the provision of straw has been suggested to reduce social tactile interactions (Fraser et al., 1991). The effects of both large group size and deep-litter in large group housing systems, on aggression require further study. The results from Study 2 suggest that there may be a degree of social instability within larger groups of pigs on deep-litter, perhaps through an inability of individuals to recognise other pigs. Even though the pigs were housed in an enriched environment, they spent considerably more time interacting with other pigs, as well as with the physical features of the pen, compared to the conventional treatment. Social stress, as a consequence of unstable social hierarchies in large groups, may contribute to the growth performance problems commonly observed in deep-litter, large group housing systems. Study 2 showed that pigs in deep-litter, large group systems engage in more social tactile interactions. Neuroendocrine responses such as elevated concentrations of catecholamines (adrenaline), adrenocorticotrophic hormone (ACTH) and corticosteroids are known to have effects on growth performance (Bartov et al., 1980; Siegel and van Kampen, 1984; Dubeski et al., 1999). Little is known on the stress physiology of pigs in deep-litter, large group systems. Mounting is described as a sexual behaviour and visual cues initiate this behaviour. While it is not unusual for entire males reared in groups to form stable homosexual relationships (Signoret et al., 1975; Fraser and Broom, 1998), high levels of sexual behaviour amongst 21-week-old entire males housed in groups of 15 in a conventional system have been implicated in reduced feeding behaviour, feed intake and growth (Cronin et al., 2003). The latter authors found that although pigs performed more than seven mounts per 24 h on average, within the groups about one-half of the pigs were responsible for the majority of mounts, while the other half either did not mount (∼30% of boars) or performed fewer than six mounts per 24 h (∼20% of boars). Cronin et al. (2003) also indicated that mounting was usually associated with agonistic behaviour. In Study 2, mounting was observed in both treatments but at three times the occurrence in the deep-litter, large group treatment. There are no other reports of the incidence of sexual behaviour by entire male growing pigs in alternative, large group housing systems, nor the potential that this level of sexual activity may have on pig production in large groups. Thus, the increased level of sexual behaviour by

184

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

entire boars housed in deep-litter, large group systems in the present studies may also have contributed to the lower incidence of feeding behaviour recorded compared to conventional systems. Social grooming rarely occurs among pigs and was not observed in Study 2, however social grooming was observed by Krosniunas (1979) and Stolba and Wood-Gush (1989). Pigs in a semi-natural environment formed small subgroups of the female with her offspring or other females and performed social grooming. It was concluded that this behaviour may be involved in maintaining social bonds within the social organisation. Pigs will rub (rubbing and rolling) against pen fixtures or upright objects and this may be considered part of their grooming behaviour (Signoret et al., 1975). Frequent interactions with the physical pen such as rooting, jumping, nosing pen fixtures and nosing concrete were observed in Study 2. Pigs are highly motivated to perform rooting behaviour and it is an obvious feature of ingestive behaviour. In unenriched environments with little complexity, pigs may redirect exploratory behaviour towards pen mates (Peterson et al., 1995), with potential adverse consequences of tail-biting and cannibalism (Ewbank, 1973; Beattie et al., 1995; Fraser and Broom, 1998). The effects of environmental enrichment on behaviour have been studied by Wood-Gush and Vestergaard (1989a,b) and Beattie et al. (1995, 1996, 2000a,b) who showed that environmental enrichment increased the incidence of non-social pen tactile interactions, as observed in Study 2. Locomotory behaviour is rather limited in pigs. While pigs are not well adapted for running per se, walking or trotting gaits are used for traversing large distances and galloping is used for flight (Signoret et al., 1975; Fraser and Broom, 1998). ‘Frolicking’, which was observed in the deep-litter, large group treatment in Study 2, has not been described by other authors, although ‘running and scampering’ which were described by Fagen (1981) as ‘play behaviours’, may be similar to frolicking. Pigs in the deep-litter, large group treatment in Study 2 spent approximately 75% more time standing and performing locomotory behaviour and performed over 113% more exploratory behaviours than pigs in the conventional housing treatment. Our results for the conventional treatment are similar to previous observations. For example, by Dinusson (1965) who showed that pigs kept in conventional housing systems and fed concentrated diets may rest (lying) for up to 80% of the time. Conventional treatment pigs have long periods of inactivity each day in which they are idle (Fraser and Broom, 1998). In Study 2, the conventional treatment pigs spent approximately 75% of their time lying or sitting, and thus ‘resting’. Lyons et al. (1995) studied the behaviour of entire male pigs in groups of 15 in deep-straw and conventional slatted pens and found that pigs spent more time standing (18% compared to 16%) in the deep-straw treatments over a 24 h period and straw-directed behaviour occupied approximately 26% of the time. Pigs in a semi-natural environment were reported by Stolba and Wood-Gush (1989) to spend up to half of the day light hours foraging and rooting. However, pigs in the deep-litter, large group treatment in the present study spent markedly less time foraging and rooting since only 44% of time during day light hours was spent standing. Differences in nutrition, group size and composition, climate and size of the enclosure and features in the environment will influence time spent foraging and rooting. Maintenance behaviours observed in the present study included feeding, drinking, defecation and urination. Pigs are omnivorous and domesticated pigs on pasture will spend

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

185

6–7 h per day foraging (Hafez, 1975). However, pigs in conventional housing systems on concentrated diets may only spend 15 min per day feeding (Signoret et al., 1975). Pigs usually drink two to three times as much water by weight as the dry feed they are eating each day. However, feed restriction (Yang et al., 1981), high ambient temperatures and poor water quality may increase this ratio by up to six times (Gardner et al., 1990). Pigs will avoid defecating and urinating near their feeding area and thus when given the opportunity, specific sites within the pen are chosen for urination and defecation by groups of pigs (Whatson, 1978). This organised group eliminative behaviour is learnt throughout infancy (Fraser and Broom, 1998). In the present study we did not record the location where pigs performed elimination behaviour. Feeding behaviour was affected by housing system in Study 2. Pigs in the deep-litter, large group treatment spent less time near the feeder, had fewer feeding events and feeding events of longer duration, than pigs in the conventional treatment. Pigs in the deep-litter, large group treatment also had a lower frequency of social interactions around the feeder, which may have caused less interruption during feeding and led to the longer mean duration of feeding events. de Haer and Merks (1992), de Haer et al. (1992), de Haer and de Vries (1993), Nielsen and Lawrence (1993) and Nielsen et al. (1996a,b) studied the feeding behaviour of individually and group-housed pigs, and found significant differences in feeding behaviour: group-housed pigs ate less frequently and had larger meals compared to individually housed pigs. The authors suggested that this difference in feeding behaviour was caused by an increased level of social interaction in pigs in groups. Furthermore, Nielsen and Lawrence (1993) found that the number of visits to the feeder decreased and duration of feeding increased as group size increased, however the maximum group size was only 20 pigs and the feeding space allowance was not increased with increasing group size. There were large differences in the pen geometry between the deep-litter, large group and conventional pens in the current study. Pigs had more pen space and ultimately more total effective space, in the deep-litter pens. This may result in pigs using space away from the feeder for non-feeding behaviours. On the other hand, pigs in the conventional pens have less pen space and thus they may use the pen space within 1 m of the feeder for non-feeding behaviours such as lying down or social tactile interactions. It is these differences in pen geometry that may contribute to the changes in the pig’s feeding behaviour. Difficulties in gaining and maintaining access to a feeder space in the conventional pens, due to limited pen space, may be responsible for the shorter but more frequent feeding events observed in these pigs. It is possible that the relatively unrestricted feeding that occurred in deep-litter, large group systems, perhaps as a result of fewer tactile interactions, may result in the reported increased fat deposition and poorer feed conversion in alternative systems. The scientific literature suggests that small, frequent feeding events throughout the day are more efficient for growth performance than larger less frequent feeding events. Large meals, eaten infrequently, lead to increased body fat deposition, decreased body protein and water, increased urinary nitrogen excretion and higher food to gain ratio than frequent small meals (Cohn et al., 1962). The results of these nutritional studies and the observations on feeding behaviour in Study 2 suggest that the feeding behaviour of pigs in deep-litter, large group systems may also contribute to the reported poorer growth performance observed in the industry setting.

186

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

5. Conclusions In conclusion, Study 1 defined the behavioural repertoire of entire male growing pigs in deep-litter, large group housing and conventional group housing systems. It is these behaviours that need to be observed in order to quantify the effects of social and feeding behaviour on the growth performance of entire male growing pigs in deep-litter, large group housing systems. Furthermore, the results from Study 2 showed that there were differences in the social and feeding behaviour of entire male pigs in deep-litter, large group and the conventional housing system. These behavioural differences may influence the growth performance of pigs in deep-litter, large group housing systems. Obviously further research on social and feeding behaviour, measured simultaneously with growth performance data, is required to examine the possibility that changes in behaviour will affect growth performance of pigs in deep-litter, large group housing systems.

Acknowledgements We wish to acknowledge financial support from the National Pork Industry Development Program and QAF Meat Industries Ltd. The technical support from J. Boyce, E. Leeson, D. Harrison and other members of the Research and Development Unit at QAF Meat Industries and the Victorian Institute of Animal Science, Department of Primary Industries, Animal Welfare Centre is also gratefully acknowledged.

References Al-Rawi, B., Craig, J.V., 1975. Agonistic behaviour of caged chickens related to group size and area per bird. Appl. Anim. Ethol. 2, 69–80. Andresen, N., Redbo, I., 1999. Foraging behaviour of growing pigs on grassland in relation to stocking rate and feed crude protein level. Appl. Anim. Behav. Sci. 62, 183–197. Bartov, I., Jensen, L.S., Veltmann, J.R., 1980. Effect of corticosterone and prolactin on fattening in broiler chickens. Poult. Sci. 59, 1328–1334. Beattie, V.E., Walker, N., Sneddon, I.A., 1995. Effects of environmental enrichment on behaviour and productivity of growing pigs. Anim. Welf. 4, 207–220. Beattie, V.E., Walker, N., Sneddon, I.A., 1996. An investigation of the effect of environmental enrichment and space allowance on the behaviour and production of growing pigs. Appl. Anim. Behav. Sci. 48, 151–158. Beattie, V.E., O’Connell, N.E., Kilpatrick, D.J., Moss, B.W., 2000a. Influence of environmental enrichment on welfare-related behavioral and physiological parameters in growing pigs. J. Anim. Sci. 70, 443–450. Beattie, V.E., O’Connell, N.E., Moss, B.W., 2000b. Influence of environmental enrichment on the behaviour, performance and meat quality of domestic pigs. Livest. Prod. Sci. 65, 71–79. Beilharz, R.G., Cox, D.F., 1967. Social dominance in swine. Anim. Behav. 15, 117–122. Bilcik, B., Keeling, L.J., 2000. Relationship between feather pecking and ground pecking in laying hens and the effect of group size. Appl. Anim. Behav. Sci. 681, 55–66. Brumm, M., 1999. Managing market pigs in hoop structures. Pork Industry Handbook. Iowa State University, Iowa. Cohn, C., Joseph, D., Allweiss, M.D., 1962. Nutritional effects of feeding frequency. Am. J. Clin. Nutr. 11, 356–361. Connor, M.L., 1995. Update on alternative housing systems for pigs. Manitoba Swine Sem. Proc. 9, 80–85.

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

187

Cronin, G.M., Dunshea, F.R., Butler, K.L., McCauley, I., Barnett, J.L., Hemsworth, P.H., 2003. The effects of immuno- and surgical-castration on the behaviour and consequently growth of group-housed, male finisher pigs. Appl. Anim. Behav. Sci. 81, 111–126. Dinusson, W.E., 1965. A day in the life of a pig. Feedstuffs 37, 5. Dubeski, P.L., Schaefer, A.L., Cook, N.J., 1999. Stimulation of chronic stress using ACTH infusion. Banff Pork Seminar, Banff. de Haer, L.C.M., de Vries, A.G., 1993. Feed intake patterns of, and feed digestibility in growing pigs housed individually or in groups. Livest. Prod. Sci. 33, 277–292. de Haer, L.C.M., Merks, J.W.M., 1992. Patterns of daily food intake in growing pigs. Anim. Prod. 54, 95–104. de Haer, L.C.M., Merks, J.W.M., Kooper, H.G., Buiting, G.A.J., Van Hattum, J.A., 1992. A note on the IVOG® -station, a feeding station to record the individual food intake of group-housed growing pigs. Anim. Prod. 54, 160–162. Estevez, I., 1998. Broiler chickens, a tolerant social system? Etologia 5, 19–29. Ewbank, R., 1973. Abnormal behaviour and pig production. An unsuccessful attempt to induce tailbiting by feeding a high energy low fibre vegetable protein ration. Br. Vet. J. 129, 366–369. Fagen, R., 1981. Animal Play Behaviour. Oxford University Press, New York. Fraser, A.F., Broom, D.M., 1998. Farm Animal Behaviour and Welfare. Bailliere, Tindall and Cassell, London. Fraser, D., Rushen, J., 1987. Aggressive behaviour. In: Price, E.O. (Ed.), Veterinary Clinics of North America: Farm Animal Behavior. Saunders, Philadelphia, pp. 285–305. Fraser, D., Phillips, P., Thompson, B.K., Tennessen, T., 1991. Effect of straw on the behaviour of growing pigs. Appl. Anim. Behav. Sci. 30, 307–318. Gardner, J.A.A., Dunkin, A.C., Lloyd, L.C., 1990. Pig Production in Australia. Butterworths, Sydney. Hafez, E.S.E., 1975. The Behaviour of Domestic Animals, third ed. Williams & Wilkins, Baltimore. Hafez, E.S.E., Signoret, J.P., 1969. The behaviour of swine. In: Hafez, E.S.E. (Ed.), The Behaviour of Domestic Animals, second ed. Bailliere, Tindall and Cassell, London. Hart, B.L., 1985. The Behavior of Domestic Animals. Freeman, New York, USA. Hemsworth, P.H., Barnett, J.L., 2001. The importance of animal comfort for animal production in intensive grassland systems. In: Proceedings of the 19th International Grassland Congress, Sao Paulo, Brazil, pp. 425–433. Honeyman, M.S., Lay, D.C., Jedlicka, L., Mugge, P., Kliebenstein, J., Duffy, M., 1999. Research and demonstration updates, ISU Rhodes Research Farm and PFI on-farm cooperator data. Swine System Options for Iowa, Leopold Center for Sustainable Agriculture, Iowa State University, Ames, IA. Hughes, B.O., Carmichael, N.L., Walker, A.W., Grigor, P.N., 1997. Low incidence of aggression in large flocks of laying hens. Appl. Anim. Behav. Sci. 154, 215–234. Jensen, P., 1980. An ethogram of social interaction patterns in group-housed dry sows. Appl. Anim. Ethol. 6, 341–350. Jensen, P., 1982. An analysis of agonistic interaction patterns in group-housed dry-sows-aggressive regulation through an avoidance order. Appl. Anim. Ethol. 9, 47. Jensen, P., Algers, B., Ekesbo, I., 1986. Methods of Sampling and Analysis of Data in Farm Animal Ethology. Birkhauser, Germany, 86 pp. Kelley, K.W., McGlone, J.J., Gaskins, C.T., 1980. Porcine aggression, measurement and effects of crowding and fasting. J. Anim. Sci. 50, 336–341. Krosniunas, E.H., 1979. Social facilitation and foraging behavior of the feral pig Sus scrofa on Santa-Cruz Island, CA. MA Thesis. Zoology Department, University of California, Davis, CA. Lay, D.C., Haussmann, M.F., Daniels, M.J., 2000. Hoop housing for feeder pigs offers a welfare-friendly environment compared to a non-bedded confinement system. J. Appl. Anim. Welf. Sci. 31, 33–48. Lyons, C.A.P., Bruce, J.M., Fowler, V.R., English, P.R., 1995. A comparison of productivity and welfare of growing pigs in four intensive systems. Livest. Prod. Sci. 43, 265–274. Minitab Inc., 1996. MINITAB Release 11 for WindowsTM . Pennsylvania, USA. McGlone, J.J., 1985. A quantitative ethogram of aggressive and submissive behavior in recently regrouped pigs. J. Anim. Sci. 61, 559–565. McBride, G., James, J.W., Hodgens, N., 1964. Social behaviour of domestic animals. IV. Growing pigs. Anim. Prod. 6, 129–139.

188

R.S. Morrison et al. / Applied Animal Behaviour Science 82 (2003) 173–188

McKinnon, A.J., Edwards, S.A., Stephens, D.B., Walters, D.E., 1989. Behaviour of groups of weaner pigs in three different housing systems. Br. Vet. J.145, 367–372. Nicol, C.J., Knowles, N.G., Parkman, T.G., Wilkins, L.J., 1999. Differential effects of increased stocking density, mediated by increased flock size, on feather pecking and aggression in laying hens. Appl. Anim. Behav. Sci. 65, 137–152. Nielsen, B.L., Lawrence, A.B., 1993. The effect of group size on the behaviour and performance of growing pigs using computerised single-space feeders. Pig News Inf. 143, 127N–129N. Nielsen, B.L., Lawrence, A.B., Whittemore, C.T., 1996a. Effect of individual housing on the feeding behaviour of previously group housed growing pigs. Appl. Anim. Behav. Sci. 47, 149–161. Nielsen, B.L., Lawrence, A.B., Whittemore, C.T., 1996b. Feeding behaviour of growing pigs using single or multi-space feeders. Appl. Anim. Behav. Sci. 47, 235–246. Noldus Information Technology, 1995. The Observer® , Base Package for WindowsTM . Reference manual, Version 3.0 Edition. Wageningen, The Netherlands. O’Connell, N.E., Beattie, V.E., 1999. Influence of environmental enrichment on aggressive behaviour and dominance relationships in growing pigs. Anim. Welf. 8, 269–279. Pagel, M., Dawkins, M.S., 1997. Peck orders and group sizes in laying hens, ‘future contracts’ for non aggression. Behav. Proc. 40, 13–25. Payne, H., 1997. Low cost, straw bedded, alternative housing systems for grower/finisher pigs. A final report for the Pig Research and Development Corporation. Pig Research and Development Corporation, Canberra, Australia. Payne, H., Mullan, B., Trezona, M., 2000. Review of alternative housing systems for pigs. A final report for the Pig Research and Development Corporation. Pig Research and Development Corporation, Canberra, Australia. Peterson, V., Simonsen, H.B., Lawson, L.G., 1995. The effect of environmental stimulation on the development of behaviour in pigs. Appl. Anim. Behav. Sci. 45, 215–224. Petherick, C.J., Blackshaw, J.K., 1987. A review of the factors influencing the aggressive and agonistic behaviour of the domestic pig. Aust. J. Exp. Agric. 27, 605–611. Siegel, H.S., van Kampen, M., 1984. Energy relationships in growing chickens given daily injections of corticosterone. Br. Poult. Sci. 25, 477–485. Signoret, J.P., Baldwin, B.A., Fraser, D., Hafez, E.S.E., 1975. The Behaviour of Swine. In: Hafez, E.S.E. (Ed.), The Behaviour of Domestic Animals. Bailliere, Tindall and Cassell, London. Spoolder, H.A.M., Edwards, S.A., Corning, S., 1999. Effects of group size and feeder space allowance on welfare in finishing pigs. Anim. Sci. 69, 481–489. Stolba, A., Wood-Gush, D.G.M., 1989. The behaviour of pigs in a semi-natural environment. Anim. Prod. 48, 419–425. Stricklin, W.R., Mench, J.A., 1987. Social organization. In: Price, E.O. (Ed.), Veterinary Clinics of North America: Farm Animal Behavior. Saunders, Philadelphia, pp. 307–322. Whatson, T.S., 1978. The development of dunging preference in piglets. Appl. Anim. Ethol. 4, 293. Wood-Gush, D.G.M., Vestergaard, K., 1989a. Exploratory behaviour and the welfare of intensively kept animals. J. Agric. Ethic. 2, 161–169. Wood-Gush, D.G.M., Vestergaard, K., 1989b. The enrichment of a bare environment for animals in confined conditions. Appl. Anim. Ethic. 10, 209–217. Yang, T.S., Howard, B., Macfarlane, W.V., 1981. Effects of food on drinking behaviour of growing pigs. Appl. Anim. Ethol. 7, 259–270.