Introduction
It has been well documented that predators have both
consumptive and non-consumptive effects on prey animals. Nonconsumptive predator effects can cause prey populations to
change their foraging behavior through a shift to safer habitat
with reduced food quality, a decrease in overall food intake, or
by minimizing foraging effort and time.
With the natural recolonization of wolves (Canis lupus) into
Washington State, it is important to know how their prey species
alter their behavior in response to the new predation threat. In
Northeastern Washington, wolves prey primarily on white-tailed
deer (Odocoileus virginianus) and mule deer (Odocoileus
hemionus). Bimodal activity patterns peak around sunrise and
sunset and the activity patterns of wolves preying on them was
predicted to also be bimodal. Deer may increase vigilance during
these time periods when most depredations occur.
Using video data from deer deployed with camera collars in
areas of wolf-presence (WP) and wolf-absence (WA), my study
was intended to answer two questions:
Deer Vigilance in Response to Wolf
Presence
By: Lyndsey Smith
School of Environmental and Forest Science, Autumn 2015
Faculty Advisors Aaron Wirsing and Apryle Craig
Camera view of deer
foraging
Figure 2: Roughly parallel lines suggest no interaction
between wolf status and time of day
Camera view of deer
being vigilant
1) Does living in areas of wolf presence cause deer to increase
their total time spent vigilant when compared to deer in areas
with no wolves?
2) Do deer dedicate more time to vigilance during hours when
predators are thought to be more active (crepuscular time
periods)?
Results
❏
❏
❏
❏
Figure 1: Study area in north-central
Washington state
Methods
Reviewed videos collected from 13 camera-collared deer in
wolf-present (7) and wolf-absent (6) areas of north-central
Washington (Figure 1) during winter months of 2015. Videos
were recorded in 10-second bursts of anywhere from 10 to 30
minute time periods throughout the day. The number of days
captured on camera varies with each individual for various
reasons (including early release of camera collar or
malfunctioning of equipment).
A total of 940 videos were reviewed; 390 videos from wolfpresent areas; 252 were male, 139 were female. 550 videos from
wolf-absent areas; 73 were male, 477 were female. 543 videos
were of mule deer, 397 videos were of white-tail deer.
Time spent vigilant was recorded to the 1/10th of a second
during dawn (0730-900), mid-day (11-1330) and dusk (15301700).
Using the statistical program R, a Two-Way ANOVA with Tukey’s
HSD was used to compare total time spent vigilant between
wolf-present and wolf-absent areas, and temporally within each
area.
❏
❏
Two variables of vigilance behavior measured:
presence or absence of wolves and time of day.
Two-factor ANOVA showed significant effect of
wolf presence (F=10.790, p<0.01) and time of
day (F=4.497, p<0.01) on the amount of time (in
seconds) deer spent vigilant
Interaction plot (Figure 2) shows no interaction
between wolf-status and time of day
Tukey’s HSD found a significant difference
between the AM (p<0.01, 95% CI [0.1959,
1.7645]) and MID (p>0.01, 95% CI [-1.2485,
0.4496]) time periods.
Average time spent vigilant was higher in all
categories (AM, MID, PM) of wolf-present areas
than in wolf-absent areas (Figure 3; Table 1).
Overall, a higher average time spent vigilant was
found in wolf-present areas (Table 1).
Wolf
Status
AM
MID
PM
Overall
WP
3.65 sec
SE± 0.34
5.09 sec
SE± 0.44
4.10 sec
SE± 0.40
4.19 sec
SE± 0.39
WA
2.90 sec
SE± 0.25
3.54 sec
SE± 0.34
3.53 sec
SE± 0.40
3.26 sec
SE± 0.31
Table 1: Average time (in seconds) spent vigilant
across wolf status and time of day
Figure 3: Average time (with SE bars) spent vigilant
during each 10-sec video across wolf status and time
of day
Discussion
The results showed a statistical difference between time deer
spent vigilant in wolf-present areas and wolf-absent areas. As
predicted, a higher amount of time was spent vigilant in areas of
wolf-presence.
Contrary to the hypothesis that deer would be more vigilant
during crepuscular time periods, deer were found to be more
vigilant during the mid-day time period in both wolf-present areas
and wolf-absent areas. This may be because deer are often bedded
down in the middle of the day, and are less likely to be foraging in
a group where a sense of security may be higher.
Results may differ if vigilance was compared to wolf use on a
continuous scale as opposed to the binary categories of
presence/absence that was used for analysis. Seasonal shifts in
habitat use and home-range size may influence deer behavior.
Additional evaluation of the influence sex, species and time of year
have on vigilance rates would be potential future research
projects.
There is data available from 2012-2015. Evaluating data across
multiple years could also provide a broader view of deer behavior
.