Compared with small groups, housing in large groups offers the pig more tot
al available space, resulting potentially in an increased degree of control
over its (micro) environment. For the producer, large groups require fewer
pen divisions and offer more possibilities for the sharing of resources su
ch as feeders and drinkers. However, whilst large groups may offer benefits
to higher ranking animals in the group, there may be serious disadvantages
for those further down the social hierarchy, who also need to compete for
access to resources. This study investigated the interactive effects on wel
fare of food availability (one single space hopper per 20 or per 10 pigs) a
nd group size (20, 40 or 80 pigs per pen), at constant stocking density (0.
55 m(2) per pig) in part-slatted pens. Groups provided with two feeding spa
ces per 20 pigs were less active than,groups with one feeding space per 20
pigs. The number of aggressive interactions per pig at the food trough was
not affected by group size but decreased with number of feeder spaces per 2
0 pigs. The number of skin lesions increased with group size. Average daily
gain in the first half of the finishing period was negatively influenced b
y group size and positively by number of feeding spaces. No effect on weigh
t gain was found subsequently. Within-group variation in growth was not aff
ected by group size or number of feeder spaces. No differences between trea
tments were found in the number of pigs removed for health reasons. Interac
tive effects of the two treatments were found on some behaviours but not on
any of the performance variables measured. It is concluded that, from a we
lfare point of view, the number of pigs per feeder spare should be lower th
an 20, although performance levels appear acceptable at 20 pigs per feeder.
Further research will have to identify whether the effects of group size o
n general aggression is common to all finishing pig systems, or whether the
presence of straw can serve as a mitigating factor.