In many sexually dimorphic mammals, adults tend to form single-sex groups (
'social segregation'). It has been assumed that social segregation is simpl
y a by-product of sex differences in habitat use ('habitat segregation'). I
tested this assumption on red deer, Cervus elaphus, and feral soay sheep,
Ovis aries, using data on group Composition, habitat use and space use coll
ected on the Scottish islands of Rum (1974-1993) and Hirta (1985-1994), res
pectively. If social segregation had been a by-product of habitat segregati
on, then (1) social segregation should have been influenced by the same env
ironmental parameters that influence habitat segregation and (2) degree of
social segregation should have equalled (and in no case been larger than) d
egree of habitat segregation. However, I found that weather parameters that
influence habitat segregation did not influence social segregation in red
deer and that degree of social segregation was significantly larger than de
gree of habitat segregation in both species. I conclude that social segrega
tion is not a by-product of habitat segregation in either species, and disc
uss the implications of this finding. (C) 1999 The Association for the Stud
y of Animal Behaviour.