Over the past few decades, in order to optimise sow fertility research in the field of animal reproduction has focused mainly on ovulation and pregnancy rates. During each farrowing, however, an average of one piglet is lost due to problems in the process of parturition, and an additional piglet is lost within a few days after birth (Edwards, 2002). Therefore, focusing research on farrowing is not only economically important but it can improve the health and welfare of the dam and her offspring. Problems during and shortly after parturition can seriously compromise animal health and can have an effect on the production economy of the pig farms. In modern pig herds, the lack of surveillance at farrowing is a current problem. Increased intensity of pig production has inevitably raised challenges during farrowing. Large litter sizes may result in delivery problems often associated with dystocia (Smith, 1997). Also, in large litters some of the piglets are more likely to be weaker and to require assistance soon after birth. Parturition should therefore be closely supervised in order to minimise losses due to problems with sows or piglets. Some studies have demonstrated that human supervision at farrowing can halve the piglet perinatal mortality (Holyoake et al., 1995; White et al., 1996). However, constant human supervision of all the farrowing sows has not been considered feasible, due to the high cost of labour and the working hours required. Currently, research is being carried out in the field of automatic growth control of pigs (Parsons et al., 2007). However, systems for automatically detecting farrowing do not exist today. Prediction of the onset of farrowing with a degree of accuracy could improve herd management and consequently improve the supervision of parturition. In sows, certain physiological and behavioural parameters are clear signals of impending parturition. Some of these parameters have been studied around farrowing, including the rise in body temperature (King et al., 1972; Elmore et al., 1979), the rise in respiratory rate (Hendrix et al., 1978) and behavioural changes (Hartsock and Barczewski, 1997; Bradshaw and Broom, 1999). All these changes occur 24–36h before farrowing starts. Thus, monitoring the variation in these parameters during the last days of pregnancy could likely serve to predict the onset of farrowing. One of the clearest signals of approaching farrowing is the increased activity due to nest building behaviour. Hartsock and Barczewski (1997) found that rooting, pawing, turning and walking behaviour in sows kept in pens or crates begins to increase 24h prior to farrowing. Therefore, the aim of our study was to test different movement sensors to measure changes in the farrowing-related activity of crated sows. These sensors could also be used to develop an automatic alarm system to help farmers to predict the onset of parturition. This would allow for more effective work organisation, reduce losses due to farrowing complications and weak piglets, and thus improve animal welfare. On the 764 farms included in the national litter-recording scheme in Finland, total piglet mortality averaged 22.4% (the mortality of liveborn piglets was 13.9%) in 2001 (Sternberg, 2002). For sows, 2768 treatment cases for farrowing problems on 942 farms included in the health recording scheme were recorded that same year (Rautala, 2002). Moreover, 15.4% of all veterinary treatments in the herds participating in the scheme were due to post partum dysgalactia syndrome (PPDS).

For more information the full article can be found at  http://www.sciencedirect.com/science/journal/15375110