Gerbils monitor the behaviou=
r of
conspecifics to assess risk.
Colin Ellard, Department of
Psychology
![](3D"slide0001_image002.gif")
Introduction
In previous studies, we have shown that gerbils respond to predatory threat fle=
xibly,
depending on the presence and location of conspecifics. When given a choice of refuges, threatened g=
erbils
will flee to the one that is farthest from the location of the threat. Gerbils tested in pairs or in l=
arger
groups are more likely to choose the closest refuge regardless of the location of the threat. In the current study, we set out =
to determine whether gerbils al=
so use
the behaviour of conspecifics to assess risk.
General Method
Gerbils were placed into a round chamber (180 cm diameter) with a high partition
through its centre. The partition was opaque so that on=
ly
about half of the testing room was visible from either side of the chamber. The bottom 10 cm of the partition=
was
left open
and screened with wire mesh so that when two gerbils were placed into the ch=
amber,
they could observe
and listen to one another easily.
On a test trial, a visual stimulus (a black cardboard rectangle suspended from a wire track) was
positioned in such a way that it was only visible from one side of the <=
/span>chamber. The stimulus was released remotel=
y from
an adjacent=
room
and the responses of gerbils to the movement of the stimulus was observed and recorded. Gerbils were either tested in pai=
rs
with one animal
on either side of the partition or individually on either the side that contained =
or did
not contain the stimulus.
Experiment 1
If gerbils use the behaviour of conspecifics to assess the=
level of risk, then one would p=
redict
that the frequency with which they respond to a mild stimulus (I.e. a stimulu=
s which would sometimes cause a solitary gerbil to fle=
e and sometimes not) would depend on the frequ=
ency
with which
observed gerbils responded to the stimulus. In this experiment, the behaviour =
of
observed gerbils was manipulated by habituating them to the presentation of the stimulus. “Confederate” gerbils=
were
habituated by exposing them repeatedly to the ‘threatening’
stimulus until
all responding ceased. Habituated gerbils were placed on the unshielded side o=
f the
chamber. The <=
span
style=3D'position:absolute;top:41.2%;left:36.1%;width:30.86%;height:1.68%'=
>frequency of response of gerbil=
s on
the shielded side of the chamber that were paired with habituated gerbils was <=
/span>compared with the frequencies of
responses of those tested alone (again, on the shielded side of the chamber) =
and those tested while paired w=
ith
naïve gerbils.
Experiment 2
![](3D"slide0001_image005.jpg")
Results
The figure shows that solitary gerbils on the shielded side of the chamber responded t=
o the
stimulus on 70%
of trials, suggesting that they could hear it move along the track. Gerbils paired with a naïve
partner showed
about the same level of responsiveness.&n=
bsp;
Gerbils
paired with habituated partners showed much lower levels of responding to t=
he
stimulus (15%). The <=
/span>figure is based on data from th=
ree
groups of gerbils, n=3D25/group.
Freezing (more than 5 seconds of immobility) and fleeing were bo=
th
scored as a response. Th=
is
observation suggests that gerbils observe responses of conspecifics to help assess ris=
k.
The results of the first experiment suggest that gerbils <=
/span>on the shielded side of the cha=
mber
assessed risk by observing the behaviour of the unshielded gerbils. What aspect of the behaviour of the
habituated gerbil was responsible for the diminution of responding in its partner? One possibility is that habituated
gerbils behaved
differently from naïve gerbils from the beginning of the trial, long be=
fore
the stimulus appears and that these differences (whatever they might be) are responsible for the lack of
responding in the test gerbils. A second possibility is that it=
is
the behaviour of the habituated gerbils at stimulus onset that is responsible <=
/span>for the effect. In order to distinguish between t=
hese
two possibilities,
I repeated experiment 1 with one change: the unshielded side of the cham=
ber
contained a partition so that the shielded gerbil could not always see its partner. Trials in which the habituated ge=
rbil
was visible at stimulus onset were compared with those in which the shielded gerbil could not see t=
he
habituated gerbil. Each gerbil received one trial =
with a
partner and one trial alone, all on the shielded side of the chamber.
The figure shows that the presence of a habituated =
partner only decreased respondi=
ng
when the partner was visible at the moment of stimulus onset.=
Conclusion
The experiments reported here suggest that when gerbils are confronted with mild
threats, they use the behaviour of others to help assess the threat and that at =
least a part of this assessment
depends on whether other gerbils are visible when the threat occurs.
![](3D"slide0001_image012.gif")
This research was supported by a=
grant
from the Natural Sciences and Engineering
Research Council of Canada