Piglet crushing remains a problem in crated sows and it is even more of a problem in loose housing systems. Most piglets are crushed when the sow moves from standing to lying or when she rolls over while already lying. This paper reviews lying down and rolling behaviours in sows and their association with piglet crushing in loose housing environments. Only little is known of the rolling behaviour, but the frequency of rolling may be affected by floor properties and fast rolling seems to be more riskful for the piglets than slower rolling. There is substantially more knowledge of the lying down behaviour. The lying down sequence is commonly divided into five stages. The speed of performance of these stages may be related to the risk of piglet crushing, e.g. flopping (the hindquarter drops quickly to the floor) is the most dangerous way the sow can lie down. In contrast, lying down by leaning against a surface rather than lying down in the open reduces the risk of piglet crushing, as does the performance of pre-lying behaviour and adequate sow responsiveness to piglets. The factors that may affect the speed and way of lying down (domestication, breed, muscular control, previous and current housing), pre-lying behaviour (piglet stimuli and space) and responsiveness to piglets (piglet stimuli, timing in relation to parturition, space and bedding) are discussed. The presented knowledge could be used in design and management of farrowing pens in order to reduce piglet crushing. However, several areas still require investigation. The aim of these investigations should be to encourage sows to use support when lying down, but also to make the few very riskful unsupported lying events as well as rolling safer for the piglets. This may be achieved though, e.g. increased knowledge of the motivational context of the various ways of lying down and rolling, development of attractive support features that sows will willingly use and determination of space requirements for the entire lying down and rolling complexes (i.e. pre-lying/-rolling, actual lying/rolling and responsiveness to overlain piglets).

The aim of the present study was to identify characteristics of sow behaviour and parturition related to early piglet mortality in loose-housed farrowing sows. Data from 152 farrowings that originated from three different herds with loose-housed sows during parturition were used. Graphical chain models were used to model the relationships between perinatal behaviour, periparturient individual conditions (time of day of parturition, rectal temperature 1-3 days postpartum) and causes of early piglet mortality. Modelling was based on the correlation between variables within herd and fat-rowing batch.The analysis showed that different causes of mortality were linked to different behavioural variables during the periparturient period and that they grouped into three independent categories. The first category was associated with stillbirth and death due to other causes. Stillbirth was positively related to the variation of the inter-birth interval and negatively related to the percentage of piglets that suckled during the first 8 h after birth of first piglet (BFP). Death due to other causes was negatively related to the suckling activity during the post-partal period (9-24 h after BFP). The second category was associated with piglet crushing, which was positively related to much lateral lying the last 4 h before BFP. Finally, the third category was associated with death due to lack of colostrums ingestion of colostrum, which was linked to the time of parturition and sow rectal temperature on days I and 2 after farrowing. Sows that gave birth during the morning compared to evening/night had more dead piglets due to lack of colostrums ingestion. Death due to lack of colostrums ingestion was also higher in sows with high rectal temperature on the day after BFR The results further showed that litter size not only influenced mortality but also behavioural variables. High litter size was related to high nest building activity before BFP, low but more variable inter-birth intervals, and much lateral lying after BFP. The study points towards several areas of interest for further research that can help understand early piglet mortality. These include causes of variation in the progress of parturition, causes of individual differences in sow activity and responsiveness to piglets, and early management of farrowing. In addition, the study suggests possible indicators of problematic farrowings rendering liveborn piglets at risk. These include prepartum lateral lying in the sow, low suckling activity and rate of stillbirth. (c) 2005 Elsevier B.V. All rights reserved.

The aim of this study was to determine whether provision of nest materials in a form and amount that could be handled in a commercial Danish pig farm with modified Schmid pens would affect the environment for sows and piglets to an extent that could be detected in sow behaviour and parameters relevant both for production and piglet welfare, i.e. piglet mortality and growth. Three-hundred twenty-six farrowings were included in the study and the sows in the treatment group were provided with 2.5kg long-stemmed barley straw from 2 days prior to expected farrowing until 4 days after farrowing in addition to the 4kg chopped straw that was given both at treatment (n=163) and control farrowings (n=163). The median mortality rates (and interquartile ranges) due to still-birth were 6.7 (0–17.6) and 6.3 (0–13.3)% for the treatment and control group, respectively. The median mortality rates within the first 3 weeks after birth for live-born piglets that remained with their biological mother, i.e. piglets that were not cross fostered, were 18.2 (6.3–30.8)% and 20.0 (7.7–30.8)% for the treatment and control group, respectively. On average, the live-born piglets that had remained by their biological mother died 0.76±0.05 days after birth in the treatment group and 0.74±0.06 days after birth in the control group. The primary mortality causes were still-birth and piglet crushing, but 72% of all the live-born piglet which died had also not received colostrum. The mortality rates, the distribution on specific mortality causes and the time of death did not differ between treatments. The results indicate that 2.5kg long-stemmed straw in addition to the chopped straw provided to both groups was insufficient to achieve the positive effect on maternal behaviour suggested possible in previous more fundamental studies of provision of nest materials to sows. This conclusion was supported by behavioural observation on a randomly selected subset sample (27 treatment farrowings and 20 control farrowings). The observations revealed no significant differences between treatment and control group in frequency of nest building or lying in lateral recumbency without changing posture from 4h prepartum until 6h postpartum. There was also no difference in the inter birth intervals or percentage of the litter that was suckling during the first 8h after birth of the first piglet. The results emphasise the need for further studies of the amounts and types of nest materials that should be given in order to fulfil the intention of the EU requirement for nesting materials to periparturient sows, and to obtain the effects on behaviour and production that have been suggested possible by previous more fundamental studies on nesting materials to loose housed sows.

The Schmid farrowing pen offers sows some degree of isolation and the opportunity to carry out nest-building behaviour. In order to evaluate the welfare of sows in this environment relative to that of sows in farrowing crates, and in order to give a detailed description of nest-building in an indoor environment eight 2nd to 8th parity sows were studied at two consecutive farrowings in a crossover design experiment. Seventeen behaviours (nest-building, postures, oral/nasal stereotypies, straw chewing and fighting the environment) were recorded and heart rate was measured by telemetry. In the Schmid pens the sows performed quantitatively more nest-building behaviour (P=0.004), which was more varied and less fragmented as measured by postural changes (P<0.001). In addition, the sows had lower heart rates during the last hour preceding farrowing (P=0.02) and performed fewer oral/nasal stereotypies in the Schmid pens than in the crates (P=0.04). Finally, the nest-building phase also tended to be shorter in Schmid pens (P=0.056). A detailed description of the behaviours involved in nest-building in sows housed indoors was achieved. The results indicate that nest-building was impaired in the crates compared to in the Schmid pens. Nest-building constitutes a behavioural need for sows, and we therefore conclude that sow welfare was improved in the Schmid pens compared to in the crates.

In order to elucidate whether feed-back from a farrowing nest affects the sows’ activity level and responsiveness to piglets during the first 24h after the onset of parturition, 18 primiparous Landrace × Yorkshire sows were given the opportunity to build a farrowing nest of peat, straw and branches in Schmid pens. Eight treatment sows then had their nest removed 8–10h after the onset of nest-building and again every 4h until farrowing began, whereas 10 control sows were allowed to keep the nest. During the first 24h after birth of the first piglet the behaviour of the sows and piglets was observed and heart rate of the sows monitored using a transmitter belt and watch receiver. The frequency of postural changes and a maternal responsiveness index was calculated. In addition, time from birth of each piglet until it suckled for the first time was calculated. In the treatment group there was a constantly higher level of maternal responsiveness (P<0.0001), but the frequency of postural changes did not differ from that of control sows. The timing of maternal responsiveness was not affected by treatment, but in both treatment groups maternal responsiveness was significantly higher during the first 2h after birth of the first piglet than in the following 6h (P<0.0001). Hereafter (8–24h postpartum) the maternal responsiveness significantly increased (P=0.05). The frequency of postural changes also was higher during the first 2h (P<0.0001) than in the following 6h after which it increased again (P=0.05). Heart rate gradually declined over the first 8h after birth of the first piglet (P<0.0001) after which it stayed level. Piglets from treatment sows took significantly longer to suckle for the first time compared to piglets from control sows (P<0.05). The decrease in activity and maternal responsiveness shortly after parturition has begun is likely to be advantageous for the piglets as it reduces the risk of piglet crushing while at the same time giving early access to the udder and to warmth from the sow. The results emphasise the importance of a proper farrowing environment and the opportunity to construct a nest, particularly in loose housed sows where the survival and vigorousness of the piglets depend greatly on the behaviour of the sow because the piglets are not protected by pen features as is the case in farrowing crates.

The influence of social subordination on sexual motivation during oestrus was studied using 36 sows of which 24 treatment sows were housed in pairs and 12 control sows were housed individually in 12 and 6m2 pens, respectively. Video recordings were made from 07:00h to 19:00h during the first 2 days after grouping, which took place 3 days after weaning of the piglets. Based on the aggressive interactions between the pair-housed sows, their rank was determined. From day 4 after weaning, a test for sexual proceptive behaviour was carried out twice daily and back-pressure test was carried out four times daily in order to detect standing oestrus. When standing oestrus had occurred, transrectal ultrasonographical scans were also carried out in order to determine if ovulation took place. The proceptivity test took place in a T-maze with a 2m×10m runway ending in two 1.5m×1.5m goal boxes each adjacent to a stimulus compartment. One compartment contained an adult sexually experienced boar and the other was empty. Latency to and duration of time spent close to the boar and time spent presenting were recorded during the 10-min test period. On the first day that standing oestrus had been detected, a test for sexual receptivity was also carried out by introducing the sow to a mature boar in his home pen (9m2). Sexual- and fear-related behaviour of sow and boar were recorded until mating was terminated or the sow had spent 5min in the pen without mating being initiated. During oestrus the proceptivity test showed a significant increase (P<0.05) in the time spent standing close to the boar and in presenting for single-housed sows and for pair-housed dominant sows, but not for subordinate sows. During oestrus subordinate sows spent significantly less time standing close to the boar than the dominant sows (P=0.01) and the same tended to be the case for presenting (P=0.07). In the receptivity test more subordinate sows than dominant sows fled (40% versus 0%, P=0.001) and more subordinate sows than dominant sows squealed (58% versus 15%, P=0.02) as a response to boar stimulation. In both tests, the single-housed sows differed neither from the dominant nor the subordinate sows. There was however no difference between the groups in the weaning to oestrous interval, duration of oestrus and number of piglet born. In addition, all the sows ovulated. The results indicate that social subordination can have significant consequences for sexual motivation in sows. Subordinate sows showed fear-related behaviour in response to boar stimulation even when they were in standing oestrus. Thus, both heat detection and mating may be impaired in subordinate sows. The results emphasise the importance to alleviate the social stress experienced by subordinates as well as the need for stock people to pay special attention to these animals when they are to be mated or inseminated.

To determine the effects of feed-back from a farrowing nest on periparturient behaviour, heart rate and hormones, 20 gilts housed in Schmid pens were permitted to build a nest of peat, straw and branches. Ten gilts then had their nest removed (NR) 10 h after the onset of nest-building and again every 4h until parturition, whereas 10 gilts were given sham removals (C). Based on video recordings, gilt periparturient postures and nest-building behaviour were observed from 7 h before until 6 h after the birth of the first piglet. Maternal heart rate was measured using a transmitter belt and a watch receiver and the heart rate during the last 7 h prepartum analysed. Maternal blood samples were taken via a jugular catheter every 20 min over the last 7 h prepartum and later analysed for plasma concentrations of cortisol and oxytocin.Treatment did not affect the quantity or timing of nest-building and postural behaviour prior to or during parturition. However, treatment significantly affected changes over time in heart rate and plasma cortisol as parturition approached. Heart rate (P = 0.008) and plasma cortisol (P = 0.002) increased for NR gilts, but remained constant for C gilts, although nest-building activity decreased during this period in both groups (P = 0.006). Plasma oxytocin was unaffected by treatment. The prepartum increases in plasma cortisol and heart rate in NR gilts in the absence of greater activity suggest that nest removal was stressful. The observed treatment effects could have resulted from gilts perceiving environmental novelty following nest disruption, a specific lack of feed-back from the nest or disturbance from nest removal activity per se. The results emphasise the importance of the farrowing environment and the opportunity to construct a nest on sow welfare. (C) 2003 Elsevier B.V. All rights reserved.

In EU organic pig production it is required that the piglets are weaned no earlier than 6 weeks of age. During the lactation period, a gradual weaning process occurs. Outdoor and organic pig production differ markedly from semi-natural and indoor environments, where the gradual weaning process has been studied in detail. The differences include the sows' opportunity for foraging and the feed availability, and compared to indoor pigs also a greater incidence of gastro-intestinal parasitic infections, which may interfere with feed consumption and utilisation of nutrients.In order to describe the gradual weaning process in an outdoor environment and investigate the possible effects of gastro-intestinal nematode infections on the behaviour during the period relevant for organic pig production, we studied the behaviour of 16 sows and their litters kept in separate paddocks with farrowing huts on days 16, 23, 30 and 44 postpartum and their growth rate and feed consumption during the same period. Half the sows and half of each litter were experimentally infected with Ascaris suum, Oesophagostomum dentatum and Trichuris suis. A gradual weaning process took place. The frequency of nursings declined, the piglets were increasingly responsible for initiating nursings, they increasingly visited a piglet feed trough and ate more solid feed, as lactation progressed (all P less than or equal to 0.001) The sow terminated 100% of the nursings and neither this nor the duration of the post-massage phase changed during the experimental period. Other behaviours usually considered important in the gradual weaning process were unexpectedly rare, i.e. piglets missing at milk let-down (0% of the litter), piglet massaging the udder between nursings (0.7% of scans), non-nutritive nursings (6.3% of nursings) and nursing in a standing position (one sow). There were only negligible effects of nematode infection on nursing behaviours, behaviours performed between nursings, body weight, weight gain and feed consumption in sows or piglets. The results indicate that infected sows did not have to restrict their allocation of resources to piglets further when infected than that which is part of the normal gradual weaning process. However, the infected sows spent more time in the huts (42% versus 29% of scans, P = 0.02) perhaps to preserve energy, whereas their piglets probably did so (P = 0.04) because they wanted to stay close to their mothers. (C) 2003 Elsevier Science B.V All rights reserved.

In a 2×2 factorial experiment, the effects of gestation and farrowing housing on (1) periparturient behaviour and circulating prolactin, prostaglandin F2α (PGF2α) and oxytocin in gilts with access to peat, straw and branches, and (2) correlational relationships between the periparturient behaviour and hormones were studied. The treatments consisted of housing in stalls or pens from mating to day 110 of gestation followed by farrowing crates or pens until after parturition. Landrace×Yorkshire gilts were observed from video recordings (n=25) from 20h prepartum and blood sampled via jugular catheters (n=16) from 24h prepartum until 2h after the birth of the first piglet.

The aim of the study was to investigate the effects of a number of environmental, behavioural and biological factors on passive immunization of piglets as assessed by transfer of porcine parvovirus (PPV) antibodies (ab) from the colostrum of PPV vaccinated mothers to the serum of the piglets. Twenty primiparous sows were housed in pens with peat, straw and branches for nest building. Half the sows were prevented from achieving feedback from a completed farrowing nest by repeated removal of the nest from 10 to 12 h after nest building had begun, whereas the other half kept their nests. Sow serum PPV-ab titres were positively related to colostrum PPV-ab titres at birth of the first piglet (BFP) (P < 0.001). Litter average piglet PPV-ab titre was positively related to both sow serum and colostrum PPV-ab titres (both P < 0.001). In addition, in the individual piglets, PPV-ab titres were reduced as time from BFP to birth and time from birth to first sucking increased and time spent sucking decreased (all P < 0.01). There were no effects of treatment, time spent in lateral recumbency by the sow, number of times the sow stood or piglet weight on day 1 on piglet serum PPV-ab titres. Preventing prolonged farrowing, while at the same time ensuring the piglets access to the udder, is important for transfer of maternal immunity. Measurements of specific antibodies in sow serum during the periparturient period and in piglet serum at 28 days of age may provide a practical tool for evaluating transfer of maternal immunity from sow to piglets.

Restraint and venepuncture affect blood parameters and are stressful for pigs. Therefore, when experiments require frequent blood sampling, catheterization is the best option, both ethically and scientifically. A simplified non-surgical method for catheterization of the vena cava cranialis in sows is described. The vena cava is catheterized via the vena jugularis externa, which is punctured while the sow is restrained by nose snaring. The procedure can be carried out by two experimenters in 5 min, it requires little equipment and can be performed under farm or very simple experimental conditions. Results from 89 catheterizations are reported. Seventy-three of the catheters remained functional for 5 days or more. There was a tendency for a higher success rate in dry sows than in highly pregnant sows. The only behavioural effect observed after catheterization was a short-lasting increase in motivation for wallowing in some sows. No illness that could be attributed to the catheterization occurred.

The success of pens which provide pigs with a bedded resting area and an area for feeding, elimination and other activities depends on pigs' ability to show discrimination in elimination behaviour and thereby keep the resting area dry and clean. The eliminative behaviour of 16 gilts was studied during the last day of gestation when motivation for prepartum nest-building was high. During gestation the gilts had been housed in stalls (n=8) or in pens (n=8). Before parturition the animals were moved to farrowing pens with a peat and straw bedded resting area and a solid floor activity area. The gilts urinated and defecated significantly more in the activity area than in the resting area (P<0.001). However, significantly more gilts which had previously been stalled defecated in the resting area than gilts which had never been confined (P<0.05). Thus, previous housing experience may influence patterns of elimination in pigs.

Sows farrowing in a semi-natural environment terminate nest building 1-7 h prior to parturition after having built a nest for which a variety of materials are used. No nest-building behaviour occurs during parturition and the sows remain lying in the nest throughout most of the farrowing. In contrast, many intensively housed sows are restless during farrowing. To investigate whether gilts housed indoors would use branches for nest building and whether access to branches would affect the termination of nest building and parturient behaviour, we studied gilts housed individually in pens designed to stimulate natural nest building. The control group (n=21) had unlimited access to straw and the experimental group (n=21) had unlimited access to straw and branches. During nest building all the gilts used straw and all the experimental gilts also used branches. In the experimental group the interval from termination of nest building to birth of the first piglet (BFP) was significantly longer than in the control group (132 versus 58 min, P=0.04). In the experimental group, nest-building behaviour was also performed by fewer individuals during the interval from BFP until 2 h after than in the control group (38% versus 71% of the gilts, P=0.03). Gilts that performed nest building during this interval carried out more postural changes (P<0.001) and spent less time in lateral recumbency (P=0.001) than gilts which did not perform nest building. On average, gilts that performed nest building behaviour after BFP (n=26) spent 54% of the first 2 h of parturition in lateral recumbency and carried out 16 postural changes. Gilts that did not perform nest building behaviour during this interval (n=16) spent 85% of the time in lateral recumbency and carried out five postural changes. In 10 gilts that were selected randomly from the experimental group nest building was studied in more detail. In these gilts nest building peaked between 17 and 6 h prepartum. There was no difference in amount of behaviour directed towards straw and amount of behaviour directed towards branches.The results indicate that the termination of nest building in sows is under environmental feedback control. When only straw was provided the nests did not have much of a lasting structure. However, when gilts had access to straw and branches more structured and functional nests could be built. These nests may have been more effective in reducing the motivation for nest building prior to the onset of parturition.

In Lorenz's 'psychohydraulic model', behaviour is regulated by performance: the motivation to perform a behaviour builds up with time and can be reduced only by performance itself. However, a convincing example of Lorenzian regulation has been lacking. We studied dustbathing in featherless and feathered chicks, Gallus gallus domesticus, the latter trained to dustbathe on glass with sand underneath. In both cases the function, that is, plumage cleaning, was eliminated. In both featherless and feathered chicks dustbathing increased with time since the last performance as found in normal dustbathing, and there was a compensatory adjustment when the amount of previous dustbathing was artificially reduced. We suggest that the amount of dustbathing performed at any one time may be controlled by the cooperative action of a deprivation or timer mechanism and a counter mechanism. The timer mechanism responds to the time since the last performance, whereas the counter mechanism records the number of elements during that last dustbathing. Alternatively, there may be a single mechanism that continually records the deficit in the amount of dustbathing performed. Lorenz's model may be valid not only for dustbathing in fowl but also for other comfort behaviours. Our results suggest the existence of behavioural needs that can be satisfied only by performance. Copyright 1999 The Association for the Study of Animal Behaviour.