The effects novel predator and nonpredator odors on risk assessment an
d nonagonistic behaviors were investigated in a series of 4 experiment
s. During initial exposure, a synthetic predator odor proved to be no
more effective than sheeps' wool in evoking defensive behaviors among
either naive or previously defeated mice. When the mice were tested fo
llowing 60-min habituation to the odors, the predator odor consistentl
y elicited stronger defensiveness than sheeps' wool. Substitution of a
pungent nonpredator odor (citronella) for that of sheeps' wool in Exp
eriment 4, however, elevated levels of risk assessment among habituate
d mice to those evoked by the predator stimulus. The use of prior habi
tuation trials seems to provide a sensitive procedure for assessing di
fferential responsiveness to odors and should be useful for anxiogenic
, but not anxiolytic, experimental manipulations. The data further sug
gest that odor novelty and intensity are important contributors to hei
ghtened levels of antipredator defense. (C) 1997 Elsevier Science Inc.