The aims of the experiment were to determine the incidence and severity of piglet- and stockperson-directed aggression in two different farrowing systems and to determine whether either, could be predicted using behavioural and physiological measures recorded previously during a standard human approach test. Sixty-two gilts, permanently housed in groups of 4-5, were approach tested at 6-8 weeks prior to parturition. Gilts were moved to a handling area, fitted with a heart rate monitor and moved to a square test arena. After 2 min familiarisation, an unfamiliar human entered the pen and stood for 3 min against one wall. Then, the human approached the gilt and touched, her: snout. Behaviour and heart rate were recorded continuously. Social rank was determined for each gilt using pair-wise food competition tests. At farrowing, gilts were randomly assigned to either open pens or conventional crates. At parturition, litter size and piglet weights were recorded and causes of any deaths ascertained, with emphasis on those piglets dying due to savaging. The aggression directed at the stockperson during piglet weighing at birth, 7 days and 14 days of age was scored on a scale of 1-5 (non-aggressive to extremely aggressive). Five gilts savaged their piglets and five other gilts were particularly aggressive towards the stockperson. Savaging of piglets and levels of stockperson-directed aggression were higher in the open pen system than in farrowing crates. Gilts that savaged piglets were more likely to show 'shy' behaviour during the human approach test, whereas gilts that were highly aggressive towards the stockperson were more likely to show 'bold' behaviour in the same test. Stockperson-directed aggression was displayed consistently, both within the same lactation and between subsequent lactations. Stockperson-directed aggression was unrelated to piglet survival but did confer advantages in piglet growth rate between birth and 7 days of age. (C) 2002 Elsevier Science B.V. All rights reserved.
