Behaviour of Young Growing Pigs in a Resident‐Intruder Test Designed to Measure Aggressiveness

Abstract A total of 125 growing pigs (47 days old) were tested for aggressive responses on two occasions using a resident‐intruder (R‐I) design. Our aims were twofold: (1) to attempt to replicate earlier work on pigs showing that resident aggression is a consistent individual characteristic, unaffected by weight or sex of the resident or intruder and (2) to develop behavioural measures to characterise the wide range of aggressive responses in the test. Resident pigs, housed since birth with littermates, were placed individually in a divided‐off portion of their home pen, and a smaller, unfamiliar intruder (approximately 66% of the resident's weight) was introduced. The test ended 5 min after the first investigation of the intruder by the resident or when one of the pigs began to attack the other (by delivering a sudden, rapid series of bites). On days 1 and 2, 33.6% and 43.2% of tests, respectively, ended in an attack by a resident. Intruder attacks were rare. Pigs were consistent in whether they attacked or not over the two tests, although attack latencies for pigs attacking in both tests were not correlated. Females were more likely to attack and attacked more quickly than males on the first test day but not in the second test or overall. Intruder sex had some effect on the test outcome (males were attacked more rapidly in the second test only). Resident and intruder weight had no effect. Aggressive pigs (meaning pigs that attacked vs. pigs that did not and fast‐attacking pigs vs. slow‐attacking pigs) showed a number of differences in behaviour during the R‐I test: (1) they took longer to make initial contact with the intruder in their first test; (2) they showed a higher frequency of aggressive acts (single head knocks, bites, and shoves); (3) they spent a greater proportion of the test engaging in social contact with the intruder rather than non‐social behaviours; (4) their social behaviour involved more postures directed toward the head as opposed to flank‐ or rear‐directed postures or re‐establishing social contact; and (5) they showed closer physical contact with intruders during social encounters, as characterised by their lower head positions. Some of these behavioural measures could be used to improve the measuring power of the test in the future. Improved behavioural measures would enable aggressiveness scoring among pigs that did not attack instead of classifying them all together as “non‐attacking.” Aggr. Behav. 28:401–415, 2002. © 2002 Wiley‐Liss, Inc.