Research Article
Contribution of Dogs to White-Tailed
Deer Hunting Success
CHRISTOPHER GODWIN,
1 Environmental and Life Sciences Graduate Program, Trent University, 1600 West Bank Drive, Peterborough, ON,
Canada K9J 7B8
JAMES A. SCHAEFER,
2 Biology Department, Trent University, 1600 West Bank Drive, Peterborough, ON, Canada K9J 7B8
BRENT R. PATTERSON,
Wildlife Research and Development Section, Ontario Ministry of Natural Resources, 2140 East Bank Drive,
Peterborough, ON, Canada K9J 7B8
BRUCE A. POND,
Wildlife Research and Development Section, Ontario Ministry of Natural Resources, 2140 East Bank Drive, Peterborough, ON,
Canada K9J 7B8
ABSTRACT
Dogs (Canis familiaris) are used in hunting white-tailed deer (Odocoileus virginianus) in
10 North American jurisdictions. Although the practice is longstanding and controversial, the effects of
dogs on the outcome of the hunt have rarely been studied. We evaluated the influence of dogs on recreational
hunting of white-tailed deer based on long-term data from southeastern Ontario, Canada. Over 25 years,
annual surveys of hunters were used to collect data on hunting effort and deer harvest from approximately
85 camps, roughly half of which had dogs. We investigated the relationship between harvest and 3 treatments
(i.e., 0 dogs, 1 dog, and
2 dogs in camp), interactions with weather and deer density, and effects of
neighboring camps. Dogs enhanced hunter success. We found no difference in deer encounter rates but, per
unit effort, camps with
2 dogs harvested 0.013 (26%) more deer per hunter-day, missed 0.010 (23%) more
deer per hunter-day, and wounded 0.002 (40%) more deer per hunter-day than camps without dogs.
Conversely, camps without dogs saw, without shooting at, 0.033 (23%) more deer per hunter-day than
camps with
2 dogs. These results are consistent with the idea that hunters with dogs are less selective.
Hunters with dogs harvested more fawns per unit effort, but we found no difference in the harvest rate of
older female deer. More precipitation, greater wind speed, lower temperatures and greater deer density
improved harvest success but had no differential effect among dog treatments. Hunter success at camps with
2 dogs was less when neighboring camps also had 2 dogs. Because antlerless deer quotas are the principal
means to control populations, increasing use of hunting dogs is unlikely to have substantial effects in
managing overabundant deer.
2012 The Wildlife Society.
KEY WORDS
dogs, harvest, hunting, neighborhood effects, Odocoileus virginianus, weather, white-tailed deer,
wounding loss.
Overabundant populations of white-tailed deer (
Odocoileus
virginianus
) have major ecological and economic effects. In
North America, the annual cost of vehicle collisions and
damage to crops and forests is in the billions of dollars
(Conover 1997). At high density, deer can limit plant regeneration
and alter forest ecosystems (Campbell et al. 2004,
Coˆte´ et al. 2004). In many cases, sport hunting represents the
most effective and cost-efficient tool for controlling deer
abundance (Giles and Findlay 2004).
Dogs (
Canis familiaris) are permitted in white-tailed deer
hunting in 10 jurisdictions in North America, including
Ontario. The practice is longstanding but increasingly an
issue with the public and has generated continual debate
on the ethics of dog use, treatment of dogs, and trespass on
private lands (Lawrence 1993, Rabb 2010, Hansen 2011).
Although the subject of many popular articles, relationships
among dogs, hunter success, and deer biology have been the
focus of few studies. Some suggest little or no physiological
or demographic effects on deer (Progulske and Baskett 1957,
Marchinton et al. 1970, Gavitt et al. 1974, Gipson and
Sealander 1977); others have noted deer injuries and mortality
(Corbett et al. 1971, Nichols and Whitehead 1978) and
elevated stress hormones in deer populations hunted with
dogs (Bateson and Bradshaw 1997). Novak et al. (1991)
reported hunters with dogs were less selective and more
successful than still hunters. Hunting success, however,
may be influenced by many factors such as weather (Fobes
1945, Hansen et al. 1986), tactics of hunters in neighboring
areas (Milner-Gulland et al. 2004), and their interactions
(Perry and Giles 1971).
We used long-term observations from southeastern
Ontario to investigate the effects of hunting dogs on
white-tailed deer hunter success. Over 25 years, annual
data were collected on hunter effort and deer harvest from
Received: 24 January 2012; Accepted: 30 July 2012
1
Present address: Ontario Ministry of Agriculture, Food and Rural
Affairs, 581 Huron Street, Stratford, ON, Canada N5A 5T8.
2
E-mail: jschaefer@trentu.ca
The Journal of Wildlife Management; DOI: 10.1002/jwmg.474
Godwin et al.
Dogs and Deer Hunting Success 1
approximately 85 camps, roughly half of which had dogs. We
tested whether dogs affected the likelihood of harvesting,
wounding, missing and encountering deer, as well as age–sex
composition of the harvest. We also explored the interactions
of dogs with weather and deer density. Finally, because dogs
may influence the outcome of other hunters, we investigated
the potential spill-over effect of neighboring camps.
STUDY AREA
The Canonto Study Area (CSA; Fig. 1) covered 230 km
2
in southeastern Ontario (45
8080N, 768500W) along the
Canadian Shield; 88% of the CSA was Crown land.
Elevation ranged from 195 m to 396 m above sea level.
The landcover consisted mainly of hardwood forest with
some coniferous patches, interspersed with numerous water
bodies (Fryxell et al. 1991, Godwin 2010). Vehicle access was
primarily via a main road running through the center of the
CSA, a hydroelectric transmission line, and secondary roads
in the northern and southern limits (Fig. 1). The roads were
used by the public and provided access to permanent hunting
camps.
The firearms season for white-tailed deer was in November.
Before 1985, the season duration was 1-week and 2 weeks
thereafter. Party hunting—where hunters were allowed to
share game seals—was not allowed for antlerless deer in
1981–1991, although it likely remained common practice
in years when not permitted. The traditional method of
hunting deer with dogs was to place hunters on watch,
usually along roads or trails, while dogs and their handlers
moved to the opposite end of the area. The dogs were then
either released or kept on leash while handlers progressed
through the forest. Once dogs and handlers reached the line
of hunters on watch, dogs were collected. Most dogs let offleash
carried collars with return addresses, some with radio
transmitters, to prevent them from being lost.
The majority of hunting in our study area was from permanent
hunting camps (annual average
¼ 76 camps; range:
62–80). The permanent camps were relatively uniform in
their distribution, typically 1 km from the nearest neighbor
(Fig. 1). Each had a traditional hunting area with little
overlap among camps. In addition, on average 16 tent and
casual hunt camps (range: 12–19) occurred in the area in
1980–1986, which declined to an average of 7 such camps in
1988–2004 (range: 3–13). Not all camps hunted every year.
METHODS
Data Collection
Each year, 1980–2004, harvest information was collected
from an average of 85 hunting camps (range: 71–99).
Of those camps, approximately 40% had 0 dogs, 12% had
1 dog, and 48% had
2 dogs (max. 9 dogs). Data were
gathered, 1980–1986, from check station interviews and
thereafter from questionnaires mailed to the lead member
of each camp. Hunters were also individually canvassed by
Ministry of Natural Resources personnel each year. The
interviews and questionnaires requested information on
hunting activities for each day of the hunt: numbers of
hunters, dogs, antlerless permits, full-day (
4 hr) hunters,
half-day (
<4 hr) hunters, bucks killed, does killed, and deer
seen, missed, and wounded. In 1987, no information on the
number of dogs in camp was collected; we therefore excluded
1987 from analysis. We calculated annual hunter effort per
camp by summing full hunter-days (1.0) and half hunterdays
(0.5) across the season. We based effort for antlerless
deer on the number of permit-days per season, which we
estimated as the number of permits multiplied by the number
of days a camp operated. Despite the possibility of overestimation
of effort bias due to early filling of permits, and
underestimation due to party hunting, we believe that this
represented a better indication of effort than hunter-days for
the antlerless deer harvest. Hunters were asked to fill out the
forms daily during the hunt. The overall return rate of the
mailed surveys was 83.8%, an excellent response rate for mail
survey, indicating minimal non-response bias (Filion 1978).
Data Analysis
During 1980–1986, hunters were asked at check stations to
indicate their daily dog usage (as full day, half-day, or not
used). From this, we computed dog use over the hunting
season each year. On average, camps with 1 dog used it 58%
of the time spent hunting each season; camps with
2 dogs
used them 87–100% of hunting time (Godwin 2010).
Therefore, we identified 3 treatments: 0 dogs, 1 dog, and
2 dogs per camp. The clear relationship between number of
dogs in camp and dog use enabled us to use mail survey
results (1988–2004), which reported only the number of dogs
in camp during the hunt.
We identified 5 response variables, each computed annually
based on the total number of deer divided by total hunter
Figure 1.
The Canonto Study Area (CSA) in southeastern Ontario,
Canada. Roads (gray lines) and hunting camps (pentagons) are depicted.
2 The Journal of Wildlife Management
effort: kill per unit effort, number of deer harvested; seen per
unit effort, number of deer seen (but not shot at); missed per
unit effort, number of deer shot at and missed; wounded
per unit effort, number of deer wounded but not recovered;
and encountered per unit effort, the sum of kill, seen, missed,
and wounded per unit effort. All treatments followed a
normal distribution with no obvious outliers after visual
inspection and they passed Levene’s test for homogeneity
of variances (
P > 0.05) unless otherwise noted. We used
Statistica version 7 (Statsoft Inc., Tulsa, OK) in our analyses.
Kill per unit effort is a common metric for assessing harvest
success, but can be confounded if harvest success varies
systematically with effort (Lancia et al. 1996, Van Deelen
and Etter 2003, Giles and Findlay 2004). Indeed, we found
an accelerating relationship between total harvest (
K) and
effort (
E) where K ¼ 0.014 E1.182 (95% CI of the exponent:
1.005, 1.359). Moreover, effort varied among treatments
(Fig. 2), a discrepancy driven more by the number of
hunters per camp (
r ¼ 0.689) than the number of days
hunted (
r ¼ 0.279). Any differences in hunter success
among treatments, therefore, may have been confounded
by the number of hunters per camp (Fig. 2). To test for
this possibility, we controlled for camp size. We matched
camps into equal numbers of hunters (2–18 hunters per
camp) and calculated the mean difference in kill per unit
effort (and 95% CI) between treatments each year for each
matched group, under the null hypothesis of 0 difference.
Each year, on average, we were able to include 54, 30, and 29
camps in contrasts between
2 dogs and 0 dogs, 2 dogs and
1 dog, and 0 dogs and 1 dog, respectively. Some camps were
used in
1 contrast. We repeated this matching procedure
for the other response variables (seen, missed, wounded,
and encountered per unit effort) to investigate further
the potentially confounding influence of effort on harvest
success.
We used a mixed model analysis of variance (ANOVA) to
investigate the effect of dog use on dependent variables (kill,
missed, and encountered per unit effort) that satisfied the
assumptions for a parametric test. Treatment served as the
fixed variable and year as a random variable. We applied
Tukey’s Honestly Significant Difference (HSD) for post
hoc comparisons. Seen per unit effort and wounded
per unit effort failed Levene’s test, so we performed a
non-parametric Friedman 2-way ANOVA. We used the
Mann–Whitney
U-test with a Bonferroni-corrected a of
0.017 (i.e., 0.05/3 tests) as post hoc tests.
We investigated hunter success on the basis of 4 age–sex
classes of deer: adult bucks (
>1-yr-old males), adult does
(
>1-yr-old females), antlerless deer (female deer and male
fawns), and fawns. We computed each metric as the total
number killed per hunter- (or permit-) day each hunting
season. For these 4 response variables, we used the same
approach—a mixed-model ANOVA (for adult bucks
and adult does) and a Friedman 2-way ANOVA test (for
antlerless deer, which failed the Levene’s test).
We obtained observations from the closest meteorological
station with complete data in 1980–2004, Ottawa
International Airport approximately 100 km northeast of
the CSA (Environment Canada, unpublished data). We
used daily meteorological variables as predictors: maximum
temperature, minimum temperature, mean temperature, total
snowfall, total rainfall, total precipitation, snow on the
ground, and speed of maximum wind gust. Although winds
appear to have little effect on deer activity (Webb et al. 2010),
given the potential impact of strong winds on the ability
of hunters to detect deer, we predicted that the number
of days with strong winds (i.e.,
>31 km/hr) would be negatively
correlated with hunting success. We summed the
numbers of days per hunting season with wind gusts
(
>31 km/hr), snowfall, rainfall, snow on the ground, and
total precipitation per season. We calculated the average
maximum, minimum, and mean temperatures during each
season for each year.
We used principal component analysis (PCA) to collapse 8
weather variables, which exhibited substantial correlations,
into fewer synthetic variables. Based on the proportion of
variance explained (86%), we retained the first 3 PCA axes
(Table 1). The first component (PC1) consisted mainly of
average maximum, minimum, and mean temperature at the
negative end, and total precipitation and days with rainfall at
the positive end. The second component (PC2) consisted
mainly of days with gusts
>31 km/hr and days with snowfall
with a strong negative loading. The third component was
reflective of the number of days with snow on the ground.
Since days with snow on the ground was the only heavily
Figure 2.
Annual harvest of white-tailed deer in relation to annual hunting
effort, classified by number of dogs in camp, Canonto Study Area, Ontario,
1980–2004.
Table 1.
Eigenvectors and percent variance explained for 8 weather variables
on the first 3 axes (PC1, PC2, and PC3) from principal component analysis.
Data from Ottawa International Airport, 1980–2004.
Variable PC1 PC2 PC3
Percent variance explained 51.58 22.29 12.07
Days with snow on the ground 0.320 0.084 0.689
Days with snowfall 0.176
0.624 0.331
Days of rainfall 0.412 0.037 0.385
Total precipitation 0.353
0.438 0.121
Average max. temperature
0.428 0.204 0.280
Average min. temperature
0.410 0.259 0.289
Average mean temperature
0.441 0.245 0.299
Days with wind gusts
>31 km/hr 0.162 0.491 0.042
Godwin et al.
Dogs and Deer Hunting Success 3
weighted parameter, it was used in place of the third principal
component. We performed an analysis of covariance
(ANCOVA) with PC1, PC2, and days with snow on the
ground as covariates, kill per unit effort as the dependent
variable, and 3 dog treatments as predictor variables. To
test for interactions, we tested for heterogeneity of slopes
among treatments for the 3 covariates (Quinn and Keough
2002).
We followed Giles and Findlay (2004) by computing an
index of deer density from the residuals of the log–log plot of
number of deer encountered and hunter effort. We conducted
an ANCOVA with deer density as a covariate, kill per
unit effort as the dependent variable, and 3 dog treatments as
predictor variables. We tested for heterogeneity of slopes
among treatments as indicative of an interaction between
deer density and dog use.
We investigated the potential effect of neighboring camps
by assessing harvest success in relation to the number and
type of camps in the vicinity. Because we expected hunters to
exert a greater effect in close proximity, for each camp we
weighted linearly and negatively each adjacent camp by its
distance,2 km, to the focal camp (Fig. 1). For each year, we
fitted a regression between kill per unit effort and summed,
weighted number of neighboring camps, and then calculated
the mean slope and 95% confidence intervals across years.
We completed this separately for each treatment. We tested
for deviation from the null hypothesis of 0 slope. Because the
treatment effect was most pronounced between 0 dogs and
2 dogs (see Results Section), we focused on these 2 groups.
RESULTS
When we controlled for the number of hunters per camp, we
found results consistent with other analyses. The mean
difference in kill per unit effort and wounded per unit effort
between 0 dogs and
2 dogs was significantly greater than 0,
indicating greater harvest success and wounding rate of the
2 dog treatment (Table 2). In contrast, the mean differences
in seen per unit effort, missed per unit effort, and
encountered per unit effort between treatments were not
significantly different than 0 (Table 2). These results largely
corroborated those from the ANOVA where camp size was
not controlled for; of 15 contrasts, 11 were consistent with
the results from the ANOVA. Moreover, in no instance did a
monotonic relationship appear to exist between the magnitude
of the contrast and camp size (Godwin 2010). These
outcomes discount the potential confounding effect of camp
size on our results.
Based on the mixed modelANOVA, dogs had a substantial
effect on the outcome of the hunt. Encounter rates with deer
(encountered per unit effort) did not differ with respect to
dogs, but kill per unit effort varied significantly between
treatments (Table 3); hunters in camps with
2 dogs killed
0.013 (26%) more deer per hunter-day compared to those
with 0 dogs. Camps with
2 dogs also missed 0.010 (23%)
more deer per hunter-day and wounded 0.002 (40%) more
deer per hunter-day than camps without dogs. In contrast,
camps without dogs saw, without shooting, 0.033 (23%)
more deer per hunter-day than camps with
2 dogs
(Table 3).
The number of adult bucks killed per hunter-day differed
among treatments (Table 4). Hunters using
2 dogs killed
17% more adult bucks and 54% more fawns per hunter-day
compared to the 0 dog treatment, but we found no significant
treatment effect on the harvest success of adult does or
antlerless deer (Table 4).
Weather influenced hunting outcomes. We found a positive
relationship between kill per unit effort and PC1, indicating
greater success with less than average maximum,
Table 2.
Mean differences in white-tailed deer hunting success between dog treatments, Canonto Study Area, Ontario, 1980–2004, after controlling for
number of hunters per camp. Numbers in parentheses represent 95% confidence limits;
n represents number of camp sizes.
Dependent variable
Contrasts
2 dogs minus 0 dogs 2 dogs minus 1 dog 0 dogs minus 1 dog
Deer killed per unit effort 0.030 (0.017, 0.042)
; n ¼ 17 0.032 (0.030, 0.093); n ¼ 13 0.008 (0.029, 0.012); n ¼ 13
Deer seen per unit effort 0.108 (
0.146, 0.362); n ¼ 18 0.085 (0.133, 0.303); n ¼ 12 0.022 (0.068, 0.023); n ¼ 12
Deer missed per unit effort 0.003 (
0.011, 0.017); n ¼ 18 0.010 (0.002, 0.022); n ¼ 12 0.005 (0.009, 0.020); n ¼ 12
Deer wounded per unit effort 0.003 (0.001, 0.005)
; n ¼ 18 0.001 (0.005, 0.004); n ¼ 12 0.002 (0.006, 0.002); n ¼ 12
Deer encountered per unit effort 0.131 (
0.110, 0.373); n ¼ 18 0.126 (0.150, 0.403); n ¼ 12 0.027 (0.095, 0.041); n ¼ 12
P < 0.05.
Table 3.
Effects of hunting dogs on hunting success for white-tailed deer, Canonto Study Area, Ontario, 1980–2004. Superscripts (A, B) indicate no significant
difference between treatment means in post hoc tests.
Dependent variable Statistic
a df P
Treatment
0 dogs 1 dog
2 dogs
Deer killed per unit effort
F ¼ 11.86 2, 23 <0.001 0.051A 0.052A 0.064
Deer seen per unit effort
x2 ¼ 8.083 2 0.018 0.176A 0.164AB 0.143B
Deer missed per unit effort
F ¼ 5.34 2, 23 0.012 0.044A 0.046A 0.054
Deer wounded per unit effort
x2 ¼ 7.00 2 0.030 0.005A 0.006AB 0.007B
Deer encountered per unit effort
F ¼ 0.28 2, 23 0.758 0.275 0.267 0.267
a
F from mixed model analysis of variance (ANOVA); x2 from Friedman 2-way ANOVA.
4 The Journal of Wildlife Management
minimum, and mean temperatures; more precipitation;
and more days with rainfall (
r2 ¼ 0.16, P < 0.001), as
well as a negative relationship between kill per unit effort
and PC2, pointing to less success with fewer days with
gusts
>31 km/hr and fewer days with snowfall (r2 ¼ 0.06,
P
¼ 0.033). The number of days with snow on the ground
had no effect. We found no evidence of interactions. The
slopes of PC1 (
P ¼ 0.48), PC2 (P ¼ 0.97), and days with
snow on the ground (
P ¼ 0.54) with kill per unit effort were
homogenous among treatments.
A relatively larger deer population translated into greater
hunting success. We found a positive relationship between
deer density and kill per unit effort (
r2 ¼ 0.48, P < 0.001),
but no differential effect of density on kill per unit effort
among dog treatments. The slope of density and kill per unit
effort was homogenous among treatments (
P ¼ 0.80).
Hunting success was reduced in the vicinity of other camps,
but only when those camps had
2 dogs. The kill per unit
effort of focal camps with
2 dogs tended to decline with a
greater number of neighboring camps with
2 dogs
(Table 5). The slope of this relationship was significantly
less than 0. None of the other treatment combinations
showed a significant effect on the kill per unit effort from
neighboring camps (Table 5).
DISCUSSION
Dogs enhanced hunting success. In our study, with
2 dogs
in camp, hunters harvested more deer per unit effort, irrespective
of the number of hunters in camp, weather, deer
abundance, and number of neighboring camps, unless they
too used dogs. Improved hunting success from dogs has been
a popular notion, albeit the topic of few studies. In South
Carolina, deer were 2.4 times more likely to be killed in a
dog-hunted area compared to a non-dog hunted area (Novak
et al. 1991). Similarly, Scribner et al. (1985) found hunters
with dogs harvested more deer per unit effort compared to
non-dog, stand hunters. When camps switched from stand
hunting to using dogs, their success increased to the same
level as camps traditionally using dogs. These results are
consistent with the 26% increase in harvest rate (0.013
more deer per hunter-day) for camps with
2 dogs in the
CSA (Table 3).
Compared to dog hunters, stand hunters often have greater
opportunities to watch undisturbed deer until within range;
they can generally take their time to shoot (Novak et al. 1991,
Martinez et al. 2005). Indeed, CSA hunters with
2 dogs
fired at and missed significantly more deer, although encounter
rates with deer did not differ. This situation may also
account for the increase in wounding rates by dog hunters—a
surprising result given that dogs are often justified by their
utility in finding wounded deer (Jeanneney 1977). In our
study, any potential improvement in the recovery of wounded
deer with dogs apparently was insufficient in reducing
wounding rates. To the contrary, we estimate that hunting
with dogs (
2 dogs per camp) resulted in approximately 4
more deer wounded per season compared to a complete
absence of dogs in the CSA. Overall, these findings suggest
that hunters with dogs are less selective than those without
dogs. Indeed, stand hunters tend to display selectively for
older or heavier deer (Novak et al. 1991, Martinez et al.
2005). Dogs also affected the composition of the harvest.
Hunters with
2 dogs harvested significantly more adult
bucks and more fawns per hunter-day, but not antlerless
deer, compared to camps without dogs (Table 4).
Our study is not the first to question whether catch per unit
effort can serve as a reliable index of harvest success (Lancia
et al. 1996, Schmidt et al. 2005). In the CSA, because greater
effort was associated with dogs in camp as well as disproportionately
greater harvest (Fig. 2), our results were potentially
confounded. Camp size, the major contributor to effort,
varied systematically with dog treatments. Nevertheless, after
controlling for the number of hunters per camp, we found
broadly similar findings to those from the ANOVA;
i.e., significant treatment effects (especially
2 dogs) for
deer killed, wounded, and encountered. These are, arguably,
the most biologically relevant parameters in deer management
(Jeanneney 1977, Giles and Findlay 2004). We found a
discrepancy, nevertheless, in the magnitude of difference in
kill per unit effort between
2 dogs and 0 dogs, 0.030 deer
per hunter-day when camp size was controlled for (Table 2)
versus 0.013 deer per hunter-day from the ANOVA
(Table 3). This is likely attributable to differences in
the experimental unit when computing these averages. We
Table 4.
Effects of hunting dogs on the age–sex composition of harvest of white-tailed deer, Canonto Study Area, Ontario, 1980–2004. Superscripts (A, B)
indicate no significant difference between treatment means in post hoc tests.
Dependent variable Statistic
a df P
Treatment
0 dogs 1 dog
2 dogs
Adult bucks killed per hunter-day
F ¼ 5.80 2, 23 0.009 0.024A 0.022A 0.028
Adult does killed per permit-day
F ¼ 0.784 2, 23 0.468 0.063 0.062 0.055
Antlerless deer killed per permit-day
x2 ¼ 0.067 2 0.967 0.088 0.092 0.092
Fawns killed per permit-day
F ¼ 9.44 2, 23 0.001 0.024A 0.031AB 0.037B
a
F computed from mixed model analysis of variance (ANOVA); x2 from Friedman 2-way ANOVA.
Table 5.
Mean slope of number of white-tailed deer killed per unit effort
versus number of neighboring camps 2 km, inversely weighted by distance,
Canonto Study Area, Ontario, 1980–2004. Focal camps and neighboring
camps were classified by treatment (
2 dogs or 0 dogs). Numbers in
parentheses represent upper and lower 95% confidence limits.
Focal camp
Neighboring camp(s)
2 dogs 0 dogs
2 dogs 0.025 (0.047, 0.002) 0.013 (0.052, 0.025)
0 dogs
0.008 (0.030, 0.013) 0.012 (0.011, 0.035)
Godwin et al.
Dogs and Deer Hunting Success 5
suggest the value based on the whole dataset (0.013) is more
accurate. Overall, these results imply that effort alone does
not account for the apparent differences among treatments.
We anticipated dogs would impart greater hunting success
when weather was unfavorable and deer were less abundant.
Lowry and McArthur (1978), for instance, suggested dogs
were more effective in pursuing deer when the ground
was covered with snow, as deep snow may impede deer.
Fobes (1945) also noted that white-tailed deer harvest was
weather-dependent (i.e., increased total harvest in warm, wet
conditions). Our results differed somewhat from expectation.
Although we uncovered a positive effect of cold and wet
conditions on hunter success, we also found no interaction
with dog treatments, even though scenting by dogs can
increase in warm and wet conditions (Styrotuck 1972).
Hansen et al. (1986) noted snowfall and high wind speeds
created unfavorable hunting conditions, and were correlated
negatively with harvest in Illinois. In contrast, snowfall and
wind had a positive relationship with kill per unit effort in
the CSA.
Not surprisingly, harvest success tended to increase with
deer density (Hansen et al. 1986). Although some have
suggested the effectiveness of dogs may be inversely density
dependent (Perry and Giles 1971), we found no interaction
of the impact of dogs with deer abundance. Hunter-deer
encounters are also the basis for tracking population trends
(Giles and Findlay 2004). In our study, dogs appeared to
have no differential effect in deer encounter rates (Table 3),
which lends confidence in this population index.
Camps with dogs diminished the hunting success of their
neighbors. Where hunting camps had
2 dogs, kill per unit
effort of camps within 2 km was reduced if they too used
dogs (Table 5). Foster et al. (1997) suggested deer hunting
was dependent on hunter densities. At low hunter densities,
hunters are too scattered to encourage deer movement,
whereas at high densities, hunter interference reduces the
per capita harvest. Spillover effects have been documented in
Scotland, too, where unhunted estates can act as sources,
allowing for migration of deer into heavily hunted estates
(Milner-Gulland et al. 2004). At a broader scale, however,
Giles and Findlay (2004) found little evidence of hunter
interference between wildlife management units in central
Ontario.
MANAGEMENT IMPLICATIONS
Hunting remains the principal tool for controlling whitetailed
deer populations and their adverse ecological and
economic effects. Our study shows that dogs affect the
outcome of the recreational hunt; 0.013 (26%) more deer
were harvested per hunter-day in camps with
2 dogs,
although dogs did not enhance the harvest rates of antlerless
deer nor adult does. Because antlerless deer quotas represent
the primary means to controlling populations, increasing the
use of hunting dogs is likely to have marginal effects in
managing overabundant deer. Camps with dogs in close
proximity (2 km) may negatively influence one another’s
success. Increasing the dispersion of hunters with dogs,
therefore, is likely to enhance hunter success and satisfaction.
ACKNOWLEDGMENTS
We thank the numerous CSA hunters for their collaboration,
L.S. Duquette for statistical guidance, and 2 anonymous
reviewers for their ideas and suggestions. We are
grateful to P.C. Smith who provided insightful comments
on the manuscript and whose longstanding work in the
CSA (1958–2004) formed the foundation of this paper.
This project was supported by Ontario Ministry of Natural
Resources and a Natural Sciences and Engineering Research
Council of Canada Discovery Grant to J.A.S.
LITERATURE CITED
Bateson, P., and E. L. Bradshaw. 1997. Physiological effects of hunting red
deer (
Cervus elaphus). Proceedings of the Royal Society of London Series
B-Biological Sciences 264:1707–1714.
Campbell, T. A., B. R. Laseter, W. M. Ford, and K. V. Miller. 2004.
Feasibility of localized management to control white-tailed deer in forest
regeneration areas. Wildlife Society Bulletin 32:1124–1131.
Conover, M. R. 1997. Monetary and intangible valuation of deer in the
United States. Wildlife Society Bulletin 25:298–305.
Corbett, R. L., R. L. Marchinton, and C. E. Hill. 1971. Preliminary study of
the effects of dogs on radio-equipped deer in a mountainous habitat.
Proceedings of the Annual Conference of Southeastern Association of
Fish and Wildlife Agencies 25:69–77.
Coˆte´, S. D., T. P. Rooney, J.-P. Tremblay, C. Dussault, and D. M. Waller.
2004. Ecological impacts of deer overabundance. Annual Review of
Ecology Evolution and Systematics 35:113–147.
Filion, F. L. 1978. Increasing the effectiveness of mail surveys. Wildlife
Society Bulletin 6:135–141.
Fobes, C. B. 1945. Weather and the kill of white-tailed deer in Maine.
Journal of Wildlife Management 9:76–78.
Foster, J. R., J. L. Roseberry, and A. Woolf. 1997. Factors influencing
efficiency of white-tailed deer harvest in Illinois. Journal of Wildlife
Management 61:1091–1097.
Fryxell, J. M., D. J. T. Hussell, A. B. Lambert, and P. C. Smith. 1991. Time
lags and population fluctuations in white-tailed deer. Journal of Wildlife
Management 55:377–385.
Gavitt, J. D., R. L. Downing, and B. S. McGinnes. 1974. Effect of dogs on
deer reproduction in Virginia. Proceedings of the Annual Conference of
Southeastern Association of Fish and Wildlife Agencies 28:532–539.
Giles, B. G., and C. S. Findlay. 2004. Effectiveness of a selective harvest
system in regulating deer populations in Ontario. Journal of Wildlife
Management 68:266–277.
Gipson, P. S., and J. A. Sealander. 1977. Ecological relationship of whitetailed
deer and dogs in Arkansas. Pages 3–16
in Proceedings of the 1975
Predator Symposium. Montana Forest and Conservation Experiment
Station, Missoula, Montana, USA.
Godwin, C. 2010. The contribution of dogs to hunter success in white-tailed
deer (
Odocoileus virginianus) harvest in south-eastern Ontario. Thesis,
Trent University, Peterborough, Ontario, Canada.
Hansen, L. 2011. Extensive management. Pages 409–451
in D. G. Hewitt,
editor. Biology and management of white-tailed deer. CRC Press, Boca
Raton, Florida, USA.
Hansen, L. P., C. M. Nixon, and F. Loomis. 1986. Factors affecting daily
and annual harvest of white-tailed deer in Illinois. Wildlife Society
Bulletin 14:368–376.
Jeanneney, J. 1977. The use of leashed trailing dogs for tracking wounded
deer. Transactions of the Northeast Fish and Wildlife Conference 34:143–
151.
Lancia, R. A., J. W. Bishir, M. C. Conner, and C. S. Rosenberry. 1996. Use
of catch-effort to estimate population size. Wildlife Society Bulletin
24:731–737.
Lawrence, R. 1993. Hunting deer and moose with dogs. Animals’ Voice
Winter:5–8.
Lowry, D. A., and K. L. McArthur. 1978. Domestic dogs as predators on
deer. Wildlife Society Bulletin 6:38–39.
Marchinton, R. L., A. S. Johnson, J. R. Sweeney, and J. M. Sweeney. 1970.
Legal hunting of white-tailed deer with dogs: biology, sociology and
6 The Journal of Wildlife Management
management. Proceedings of the Annual Conference of Southeastern
Association of Fish and Wildlife Agencies 24:74–89.
Martinez, M., C. Rodriguez-Vigal, O. R. Jones, T. Coulson, and A. San
Miguel. 2005. Different hunting strategies select for different weights in
red deer. Biology Letters 1:353–356.
Milner-Gulland, E. J., T. Coulson, and T. H. Clutton-Brock. 2004. Sex
differences and data quality as determinants of income from hunting red
deer
Cervus elaphus. Wildlife Biology 10:187–201.
Nichols, R. G., and C. J. Whitehead, Jr. 1978. The effects of dog harassment
on relocated white-tailed deer. Proceedings of the Annual
Conference of Southeastern Association of Fish and Wildlife Agencies
32:195–201.
Novak, J. M., K. T. Scribner, W. D. Dupont, and M. H. Smith. 1991.
Catch-effort estimation of white-tailed deer population size. Journal of
Wildlife Management 55:31–38.
Perry, M. C., and R. H. Giles, Jr. 1971. Studies of deer-related dog activity
in Virginia. Proceedings of the Annual Conference South Eastern
Association of Game and Fish Commissions 24:64–73.
Progulske, D., and T. S. Baskett. 1957. Mobility of Missouri deer and
their harassment by dogs. Journal of Wildlife Management 22:184–
192.
Quinn, G. P., and M. J. Keough. 2002. Experimental design and data
analysis for biologists. Cambridge University Press, United Kingdom.
Rabb, J. 2010. The dog-hunting debate. North American Whitetail.
<
http://www.northamericanwhitetail.com/2010/09/22/deermanagement_
naw_debate_1009/2/
>. Accessed 8 Aug 2012.
Schmidt, J. I., J. M. V. Hoef, J. A. K. Maier, and R. T. Bowyer. 2005. Catch
per unit effort for moose: a new approach using Weibull regression.
Journal of Wildlife Management 69:1112–1124.
Scribner, K. T., M. C. Wooten, M. H. Smith, and P. E. Johns. 1985.
Demographic and genetic characteristics of white-tailed deer populations
subjected to still or dog hunting. Pages 197–212
in Game Harvest
Management. Caesar Kleberg Wildlife Research Institute, Kingsville,
Texas, USA.
Styrotuck, W. G. 1972. Scent and the scenting dog. Arner, Rome, New
York, USA.
Van Deelen, T. R., and D. R. Etter. 2003. Effort and the functional response
of deer hunters. Human Dimensions of Wildlife 8:97–108.
Webb, S. L., K. L. Gee, B. K. Strickland, S. Demarais, and R. W. DeYoung.
2010. Measuring fine-scale white-tailed deer movements and environmental
influences using GPS collars. International Journal of Ecology,
vol. 2010, Article ID 459610, 12 pages. DOI: 10.1155/2010/459610
<
http://www.hindawi.com/journals/ijeco/2010/459610/cta/>.
Associate Editor: David Euler.
Godwin et al.
Dogs and Deer Hunting Success 7