Site by Bennett Web & Design Co.
2011 NORTHEASTERN NATURALIST 18(4):395–410
Fall–Winter Survival of Ruffed Grouse in New York State
Megan M. Skrip1,2,*, William F. Porter1,3, Bryan L. Swift4,
and Michael V. Schiavone4
Abstract - In New York, Bonasa umbellus (Ruffed Grouse) abundance has declined
since the 1960s, presumably due to forest maturation. Wildlife managers expressed concern
that hunting may contribute to the population decline as habitat quality decreases.
We monitored fall–winter survival of 169 radio-marked Ruffed Grouse at 2 study areas
in New York differing in forest age and composition. Fewer than 11% of radio-marked
birds were harvested, and seasonal survival was similar at the 2 study areas in both study
years (0.38 and 0.51, 2007–2008; 0.48 and 0.48, 2008–2009). Predation, particularly by
raptors, was the largest source of mortality, but locations of predation events were not associated
with forest age or configuration within 300 m. We found no evidence to support
a reduction in harvest limits, although our harvest estimates may have been biased low.
Sustainable wildlife management depends on reexamining existing management
policies in light of changing habitat conditions. As habitat quality for
Bonasa umbellus L. (Ruffed Grouse) declines throughout the Northeast due
to forest maturation (DeGraaf and Yamasaki 2003, Dessecker and McAuley
2001), wildlife managers express concern that harvest regulations may require
revision. Ruffed Grouse have long been popular quarry for sportsmen in North
America (Edminster 1954), and the impact of harvest on populations has been
the subject of many studies (e.g., Clark 2000, DeStefano and Rusch 1986, Devers
et al. 2007, Dorney and Kabat 1960, Edminster 1937, Fischer and Keith
1974, Palmer and Bennett 1963, Small et al. 1991). However, the combined
contribution of harvest and declining habitat quality to mortality, particularly
in declining Ruffed Grouse populations, has not been addressed outside of the
Appalachians. Ruffed Grouse could be at greater risk of harvest in areas with
increasingly inadequate cover and, thus, it is important to assess the effect of
harvest on declining populations as habitat conditions change at broad scales.
Continent-wide, Ruffed Grouse populations have declined 54% since the
mid-20th century as high-quality early-successional habitat has disappeared
(National Audubon Society, Inc. 2008). In New York State alone, Ruffed Grouse
have declined 5% per year 1966–2007, and as much as 16% per year since 1980
(Post 2008, Sauer et al. 2008), as forests regenerating from abandoned farmlands
1Department of Environmental and Forest Biology, State University of New York College
of Environmental Science and Forestry, 1 Forestry Drive, Syracuse, NY 13210.2Current
address - Department of Natural Resources Science, 105 Coastal Institute in Kingston, University
of Rhode Island, 1 Greenhouse Road, Kingston, RI 02881. 3Current address - Department
of Fisheries and Wildlife, 13 Natural Resources Building, Michigan State University,
East Lansing, MI 48824. 4New York State Department of Environmental Conservation, 625
Broadway, Albany, NY 12233. *Corresponding author - firstname.lastname@example.org.
396 Northeastern Naturalist Vol. 18, No. 4
have matured. Despite changes in population size, Ruffed Grouse remain the
second most popular game bird in New York (after Meleagris gallopavo L. [Wild
Turkey]; New York State Department of Environmental Conservation, Albany,
NY, unpubl. data). The last assessment of Ruffed Grouse survival and harvest
mortality in New York was performed over a half-century ago (i.e., Bump et al.
1947). Our investigation responded to concerns that harvest may contribute to the
decline of Ruffed Grouse in the state, as New York has one of the longest hunting
seasons in the Northeast.
Young, early-successional forest with high stem densities is ideal habitat for
Ruffed Grouse (Thompson and Dessecker 1997). We proposed that landscapes
consisting of large patches of mature forest may necessitate larger home range
sizes and consequently favor human and predator encounters, contributing to
higher mortality. Conversely, high availability of early-successional habitat may
improve survival by providing better cover and food resources, thereby decreasing
energy demands, activity times, and predator encounters (Endrulat et al. 2005,
Fearer and Stauffer 2004, Hewitt and Kirkpatrick 1997, Whitaker et al. 2007, Yoder
et al. 2004). Within home ranges, dense vertical cover of saplings and shrubs
is essential for shielding Ruffed Grouse from predators (Dessecker and McAuley
2001, Thompson and Dessecker 1997), including hunters.
We examined Ruffed Grouse survival through 2 hunting seasons in 2 areas
of New York State with different forest age and composition, and we assessed
the contribution of harvest to overwinter mortality. We predicted that fall–winter
survival would be higher, and harvest rate lower, in the study area dominated by
young, early successional forest, i.e., that Ruffed Grouse mortality is lower in
higher-quality habitat. We expected that mortality locations of Ruffed Grouse
would occur more frequently in older forest than was available at random. If our
predictions were correct, modifications to harvest limits or forest management in
some areas of the state might be considered by management agencies.
Field Site Description
The two field sites chosen for this 2007–2008 and 2008–2009 investigation
were 1943 ha of the Fort Drum Military Installation in Jefferson County, NY, and
the 1859-ha Partridge Run Wildlife Management Area in Albany County, NY. Fort
Drum (44°3'N, 75°33'W) is <80 km east of Lake Ontario in New York’s Western
Adirondack Transition zone adjacent the Eastern Ontario plain (NYSDEC Habitat
Inventory Unit 1990) in the Laurentian Mixed Forest Province (Bailey 1995).
Partridge Run (42°34'N, 74°11'W) is approximately 32 km west-southwest of
Albany in the Helderberg Highlands region of New York’s Appalachian Plateau
(NYSDEC Habitat Inventory Unit 1990), with vegetation typical of the Eastern
Broadleaf Forest Province (Bailey 1995). Ruffed Grouse hunting seasons opened
on 20 September at Fort Drum and 1 October at Partridge Run and closed at both
areas on the last day in February both study years. Both areas were popular sites
for small- and big-game hunting (C. Dobony, Fort Drum Fish and Wildlife Management
Program, Fort Drum, NY, and B. Swift, New York State Department of
Environmental Conservation, Albany, NY, pers. comm.) and represented areas
with some of the highest grouse hunting pressure in the state.
2011 M.M. Skrip, W.F. Porter, B.L. Swift, and M.V. Schiavone 397
Fort Drum and Partridge Run varied in tree species composition and stand size
class distribution. Fort Drum was dominated by second-growth forests regenerated
from farmland abandoned in the 1940s (Dobony and Rainbolt 2008). Forest
there was predominantly early and mid-successional (Populus grandidentata
Michx. [Bigtooth Aspen], Populus tremuloides Michx. [Quaking Aspen], Betula
spp. [birch], Prunus spp. [cherry], Acer spp. [maple], and Pinus spp. [pine]),
with some mature plantations of Pinus sylvestris L. (Scots Pine) and P. resinosa
Soland. (Red Pine). At Partridge Run, forests regenerated from marginal agricultural
lands abandoned in the 1930s in the Helderberg Highlands (Moser et
al. 2001). Our study area there consisted mainly of mature mixed hardwoods
(maple, birch, cherry, Fraxinus spp. [ash], Quercus spp. [oak], and Fagus grandifolia
Ehrh. [American Beech]) and some mature conifer plantations (including
pine and Picea abies L. [Norway spruce]), with scattered abandoned Malus spp.
Fort Drum had nearly twice as much seedling-sapling forest (2–12 cm diameter
at breast height [DBH]) than Partridge Run (14% of forest at Fort Drum vs.
7.2% at Partridge Run; C. Dobony, unpubl. data; Moser et al. 2001). The majority
of forest at each area was pole timber (13–28 cm DBH; 58% at Fort Drum, and
82.3% at Partridge Run), with a saw timber (>28 cm DBH) component (28% of
Fort Drum forest, and 10.4% of Partridge Run forest) (C. Dobony, unpubl. data;
Moser et al. 2001).
We captured Ruffed Grouse at Fort Drum and Partridge Run in autumn 2007
and 2008 using modified lily-pad traps (Backs et al. 1985, Gullion 1965, Hunyadi
1984). At Fort Drum, birds were captured 11 September–5 October 2007
and 4 September–6 November 2008. At Partridge Run, birds were captured 6–28
September 2007 and 2–18 September 2008. We checked traps twice a day, once
in the morning and once in late afternoon to evening. We fit Ruffed Grouse with
an anodized aluminum butt-end band and a 5.6-g (2007 only) or 10.7-g (2007
and 2008) necklace-style VHF radio-transmitter (Advanced Telemetry Systems,
Isanti, MN) with either an 8-hr (2007 only) or 4-hr (2008 only) mortality switch
(Devers et al. 2007). Bands and radio-transmitters were labeled with toll-free
telephone numbers to encourage hunters to report harvested birds. All capture
and handling procedures were conducted under the approval of the Institutional
Animal Care and Use Committee of the State University of New York College of
Environmental Science and Forestry (IACUC Protocol No. 2007-7).
We determined the sex of captured Ruffed Grouse by counting the number of
spots on rump feathers and observing the continuity of the sub-terminal tail band
(Bump et al. 1947, Roussel and Ouellet 1975, Servello and Kirkpatrick 1986). We
differentiated adult (after-hatching-year) and juvenile (hatching-year) birds by
the curvature of the ninth and tenth primary feathers and the presence or absence
of primary feather sheathing (Dorney and Kabat 1960). We knew of no problems
with these criteria in New York State.
398 Northeastern Naturalist Vol. 18, No. 4
We monitored the status of radio-marked Ruffed Grouse via radio-telemetry
2–3 times a week from the time of capture through the end of the hunting season
(29 February 2008 and 28 February 2009). Upon detection of a mortality signal,
we retrieved the transmitter and remains, recorded the location with a GPS
unit, and determined cause of death via standard field sign (Bumann 2002). We
classified mortality sources as harvest, avian predation, mammalian predation,
unknown predation, other, or unknown. Harvests were reported by hunters or
determined via field sign. We assigned the date of mortality to midpoint between
the last live signal detected and the first mortality signal (Devers et al. 2007).
We right-censored all birds with which we lost contact (dropped collar, transmitter
failure, or signal loss) or which survived the analysis interval (Pollack et al.
1989a, b). For censored birds with unknown fates, we assigned the date of censor
to the day after the last date of detection (Devers et al. 2007). For all birds still
alive at the end of the hunting season, we assigned the censor date as the last day
of the respective hunting season. Two birds lost from the 2007–2008 season due
to battery failure were re-captured autumn 2008 and re-admitted to the study at
that time with new functioning transmitters. Because their fates were known,
these two birds were censored from the 2007–2008 study interval at the end of the
hunting season rather than upon transmitter failure. We excluded from analysis
any Ruffed Grouse that died less than 1 week post-capture, to control for capturerelated
transmitter effects (Clark 2000, Devers et al. 2007). In total, 169 Ruffed
Grouse were included in survival analyses.
We estimated fall–winter survival through the hunting season for study areas,
sexes, and age classes via the Kaplan-Meier product-limit estimator in SAS
(Kaplan and Meier 1958, SAS Institute Inc. 1999), using code from White and
Garrott (1990) to allow for staggered entry (Pollack et al. 1989a, b; Roberts
1993). We compared survival distributions using the log-rank test. We used a
2-tailed Z statistic to test for differences in survival rate at the end of the hunting
season, with a significance level of α = 0.05 (Pollack et al. 1989a, Roberts
1993). We compared survival distributions within study areas between years after
Migoya and Baldasarre (1995). We used 2 September (the earliest capture date)
as an interval start point (White and Garrott 1990) to compare year-to-year and
study-site variation. For comparisons between sexes and age groups, individuals
were pooled across study areas.
We performed a stratified, cause-specific risk analysis using Cox regression
in SAS (PROC PHREG) with data duplication (Heisey and Patterson 2006,
Lunn and McNeil 1995). Lunn and McNeil (1995:527–8) describe an approach
to fitting a stratified proportional hazards model in which the stratification
variable is risk type (δ), allowing the baseline hazard to vary by risk type. The
covariates are coded as the pairs δx1, δx2 and (1 – δ)x1, (1 – δ)x2, replacing main
effects and interactions with fully specified interactions. This model yields coefficients
as if fitting a separate proportional hazard model for each risk type.
We fit a full model with effects of age (0 = adult, 1 = juvenile), sex (0 = male,
2011 M.M. Skrip, W.F. Porter, B.L. Swift, and M.V. Schiavone 399
1 = female), and study area (0 = Fort Drum, 1 = Partridge Run) on the risk
of harvest vs. non-harvest mortality (i.e., on harvest hazard vs. non-harvest
hazard). Non-harvest mortality consisted mainly of predation, with a small
component of non-predation and unknown cause. We pooled all data from both
2007–2008 and 2008–2009 seasons and used a significance level of α = 0.05.
To compare overall landscape composition and arrangement between study
areas, we used location information to estimate the total area used by study birds.
For each study area, we overlaid recovery and trap locations from the 2008–2009
season on a 2001 National Land Cover Data (NLCD) raster in ArcMap (ESRI
2006) and determined the straight-line distance (m) between trap site and recovery
site for each bird captured (Thomas and Taylor 2006). Recovery location data
from the 2007–2008 season were incomplete. We found 1.5 km to be the average
distance from trap to recovery (range 79 m to 7.7 km) and, therefore, generated
1.5-km-radius buffers around each recovery and trap coordinate, dissolving
those buffers at each site to form study area boundaries. The dissolve yielded a
Fort Drum analysis area of 6133 ha and Partridge Run area of 5728 ha. We used
FRAGSTATS (McGarigal et al. 2002) to calculate the following metrics in each
area: percentage of each land cover class (PLAND), patch density (PD), edge
density (ED), contrast-weighted edge density (CWED), mean patch shape index
(SHAPE), interspersion and juxtaposition index (IJI), and patch richness (i.e.,
number of different landscape classes present; PR).
To compare landscape characteristics between mortality locations and random
locations within study areas, we used ArcMap to cast 50 random points within
each study area boundary and compared those points to predation locations of
grouse during the 2008–2009 field season (Fort Drum, n = 18; Partridge Run,
n = 15). We buffered each predation and random point by a 300-m-radius (28.44-
ha) circle to approximate the average home range size of a Ruffed Grouse around
each location (Fearer 1999). We used FragStatsBatch (Mitchell 2005) to calculate
metrics within each buffer (Brown and Litvaitus 1995, Fearer 1999, Thogmartin
and Schaeffer 2000) and used all of the metrics evaluated in the study area
comparison except for percentage of each land-cover class. To avoid multicollinearity
among PLAND variables, we included only percent shrub (PLAND for
shrub), percent forest (pooled PLAND for deciduous, mixed, and evergreen forest
and woody wetlands), proportion evergreen (PLAND for evergreen divided
by pooled forest PLAND), and proportion deciduous. We performed a correlation
analysis using PROC CORR in SAS to compare landscape metrics and eliminated
the less biologically meaningful variable of pairs with ρ ≥ 0.7 to further
avoid multicollinearity. We used PROC LOGISTIC in SAS (stepwise selection,
α = 0.05) to develop models differentiating predation locations and random locations
(Henner et al. 2004, Kunkel and Pletscher 2000, Thogmartin and Schaeffer
2000). We used area under the curve values from receiver operating characteristic
(ROC) curves to evaluate model discriminatory performance (Fielding and Bell
1997, Pearce and Ferrier 2000).
400 Northeastern Naturalist Vol. 18, No. 4
At Fort Drum, an additional forest inventory dataset managed by installation
personnel allowed comparison of forest stand size class between predation and
random points (C. Dobony, unpubl. data). For each mortality location (n = 16)
and random location (n = 34) for which this information was available, we used
the Fort Drum coverage to determine size class (seedling-sapling, pole, saw timber)
of the patch containing each point. We used Fisher’s exact test to test whether
size-class frequency differed between predation and random locations (McDonald
2009). We expected that a greater frequency of predation locations would be in
saw timber patches than in random locations, and a lower frequency would be in
seedling-sapling patches than in random locations.
Mortality sources and their temporal distribution
Over both study areas and years, 48–58% of 37–50 radio-marked Ruffed
Grouse were confirmed mortalities through the hunting season, due to several
mortality sources. Predation, particularly by raptors, was the largest source of
mortality (Table 1). At each study area, 39% (Partridge Run) and 55% (Fort
Drum) of all mortalities were confirmed predation events in 2007–2008, although
these values are likely biased low because unknown determinations may have
been predation. In 2007–2008, interference by scavengers and weather conditions
complicated mortality diagnoses, contributing to an elevated number of
unknowns. In 2008–2009, 91% of mortalities at Fort Drum and 62% at Partridge
Run were due to predation, . Hunting represented 0–22% of all mortalities across
study areas and years and was the mortality source of 0–11% of all captured birds.
At both study areas, several birds each year lost contact and were right-censored:
3 in 2007–2008 and 4 in 2008–2009 at Partridge Run, and 5 in 2007–2008 and 1
in 2008–2009 at Fort Drum.
Table 1. Mortality sources of radio-marked Ruffed Grouse at Fort Drum and Partridge Run, NY,
2007–2008 and 2008–2009. In 2007–2008, some “unknown” determinations were likely predation
but not recorded as such; in 2008–2009, “unknown” indicates inaccessibility of the remains.
Fort Drum Partridge Run
2007–2008 2008–2009 2007–2008 2008–2009
No. of birds in study 38 44 37 50
Total confirmed mortalities 22 21 18 26
Total harvest _3 _0 _4 _3
Total non-harvest 19 21 14 23
Raptor _6 13 _2 10
Mammal _2 _5 _1 _2
Unknown predator _4 _1 _4 _4
Unknown _7 _1 _7 _3
OtherA _1 _4
Lost transmitter or signal _5 _1 _3 _4
Survived hunting season 11 22 16 20
AOne Ruffed Grouse was sealed in a snow roost by an ice storm at Fort Drum. At Partridge Run, 3
were killed by vehicles, and 1 whole carcass was found without evidence of predation.
2011 M.M. Skrip, W.F. Porter, B.L. Swift, and M.V. Schiavone 401
Harvest mortality was low and similar between study areas in 2007–2008, but
displayed different temporal patterns (Fig. 1). Three birds were harvested at Fort
Drum and 4 at Partridge Run in 2007. We documented no harvest of radio-marked
Ruffed Grouse at Fort Drum and 3 at Partridge Run during the 2008–2009 hunting
season. At Fort Drum in 2007–2008 and Partridge Run in 2008–2009, monthly
mortality was highest in October and late winter; in 2007–2008 at Partridge Run
and 2008–2009 at Fort Drum, mortality peaked in November (Fig. 1).
At Fort Drum, 1 adult female (of 6 marked), 1 adult male (of 7 marked), and
1 juvenile female (of 14 marked) were harvested in the 2007–2008 season. At
Partridge Run, across both seasons, 2 adult females (of 10), 2 adult males (of 19),
and 3 juvenile females (of 26) were harvested. In total, 14% of 42 adults and 4%
of 93 juveniles captured were harvested; 13% of 56 females and 4% of 76 males
were harvested. No radio-marked juvenile males (of 53 total) were harvested at
either site, although 29 were mortalities.
Figure 1. Monthly mortality of radio-marked Ruffed Grouse by source, Fort Drum and
Partridge Run, NY, 2007–2008 and 2008–2009.
402 Northeastern Naturalist Vol. 18, No. 4
Survival analysis using the Kaplan-Meier estimator
Survival curves through the hunting season did not differ between Partridge
Run and Fort Drum (Figs. 2, 3) for both 2007–2008 (log-rank, P = 0.99) and
2008–2009 (log-rank, P = 0.09). Male and female survival curves (data pooled
across study areas) did not differ in the 2007–2008 season (log-rank, P = 0.45),
but did in the 2008–2009 season (log-rank, P = 0.01). Survival curves differed
between adults and juveniles (data pooled across study areas) during both 2007–
2008 (log-rank, P < 0.01) and 2008–2009 seasons (log-rank, P < 0.01). Seasonal
survival curves did not differ at Fort Drum between 2007–2008 and 2008–2009
(log-rank, P = 0.65), but did so at Partridge Run (log-rank, P = 0.01). Unlike survival
curves, final survival probabilities at end of hunting seasons did not differ
between any comparison classes (Z-test, P > 0.05; Table 2).
Competing risks analysis using the Cox model
The harvest hazard of females in our study was 3.62 times the male harvest
hazard, and the harvest hazard of juveniles was 25% of the adult harvest hazard
Figure 2. Kaplan-Meier survival probability distributions (solid lines) and point estimate
95% confidence intervals (dashed lines) for Fort Drum (black) and Partridge Run (gray),
NY, 2007–2008 (log-rank test, P = 0.99).
2011 M.M. Skrip, W.F. Porter, B.L. Swift, and M.V. Schiavone 403
(Table 3). The harvest hazard of Partridge Run birds did not differ from that of
Fort Drum birds (Table 3). None of the demographic variables were significant
predictors of predation risk (P > 0.05; Table 3).
Figure 3. Kaplan-Meier survival probability distributions (solid lines) and point estimate
95% confidence intervals (dashed lines) for Fort Drum (black) and Partridge Run (gray),
NY, 2008–2009 (log-rank test, P = 0.09).
Table 2. Kaplan-Meier final survival probabilities (S) at end of February, and 95% upper and
lower confidence limits, for Ruffed Grouse at Fort Drum and Partridge Run, NY, September–
February 2007–2008 and 2008–2009. Males and females, and adults and juveniles, were pooled
across study areas.
n LCL S UCL n LCL S UCL
Fort Drum 38 0.22 0.38 0.54 44 0.33 0.48 0.63
Partridge Run 37 0.35 0.51 0.67 50 0.34 0.48 0.62
Males 35 0.25 0.41 0.58 54 0.27 0.40 0.54
Females 39 0.34 0.50 0.65 37 0.41 0.57 0.72
Adults 24 0.16 0.36 0.56 29 0.26 0.44 0.62
Juveniles 50 0.35 0.49 0.63 61 0.39 0.51 0.63
404 Northeastern Naturalist Vol. 18, No. 4
Despite differences in local forest composition and structure between areas,
Fort Drum and Partridge Run exhibited similar overall landscape-scale configuration.
No variables were significant predictors of mortality location versus random
location at either Fort Drum or Partridge Run (models not shown). When study
areas were pooled, the final model indicated that predation locations had more
forest and greater patch richness than random locations (Table 4). The classification
accuracy of the model was low (area under the ROC curve = 0.690), and it
was more effective at predicting random locations than predation locations. The
true probability that a location was a predation site was 33%, but at a probability
level of 0.33, only 42% of predation locations and 82% of random locations were
predicted correctly by the model.
Analysis of the Fort Drum forest inventory dataset revealed no difference
in tree size-class frequency between mortality and random locations. While
a greater proportion of random (vs. mortality) points occurred in sapling
patches, and a greater proportion of mortality (vs. random) points occurred in
sawtimber patches, this difference was not statistically meaningful (Fisher’s
exact test, P = 0.51).
Our objective was to compare the contribution of harvest to Ruffed Grouse
mortality during the hunting season in different forest types in New York State.
We expected harvest to be lower and survival higher in the study area with more
Table 3. Parameter estimates from a competing-risks model assessing the hazard of harvest versus
non-harvest death of Ruffed Grouse based on study area, sex, and age, New York, 2007–2009.
Risk type Variable DF estimate Std. error Chi-square P Hazard ratio
Study area 1 0.54 0.51 1.10 0.29 1.709
Sex 1 1.29 0.54 5.60 0.02 3.619
Age 1 -1.38 0.50 7.50 less than 0.01_ 0.252
Study area 1 -0.32 0.20 2.71 0.10 0.724
Sex 1 -0.39 0.20 3.76 0.05 0.676
Age 1 0.11 0.21 0.27 0.60 1.116
Table 4. Parameter estimates (estimated value of regression coefficients) from a stepwise logistic
regression model comparing random and predation locations of radio-marked Ruffed Grouse with
study areas pooled, New York, 2007–2009.
Parameter DF Parameter estimate Std. error Wald chi-square P
Intercept 1 -7.85 2.76 8.11 less than 0.01
Percent forest 1 0.04 0.02 4.24 __0.04_
Patch richness 1 0.51 0.19 7.40 less than 0.01
2011 M.M. Skrip, W.F. Porter, B.L. Swift, and M.V. Schiavone 405
early-successional forest. Despite differences in seral stage and tree size distribution
at Partridge Run and Fort Drum, overwinter survival of Ruffed Grouse
did not differ between study areas. Harvest levels were lower than expected, and
predation events were not more likely to occur in forest stands of a particular age
or landscape configuration than stands at random locations.
Overall harvest and survival rates we observed were comparable to those from
decades ago in New York (Bump et al. 1947) and from cyclical grouse populations
elsewhere (Davies and Bergerud 1988, Edminster 1954, Rusch and Keith
1971). If harvest rates are similar elsewhere in the state, the level of take we observed
suggests that hunting is likely not the driver of grouse population decline
in New York (Ellison 1991). Other studies have found that the high background
mortality rate of grouse can allow for compensatory survival even at higher rates
of harvest (Ellison 1991), although additive harvest mortality was documented
in Wisconsin when harvest accounted for 28% of total mortality during hunting
seasons (Small et al. 1991). These authors suggest that a regional grouse population
may decline when high hunting mortality combines with low immigration
caused by landscape fragmentation (Small et al. 1991).
The last empirical assessment of Ruffed Grouse harvest rate in New York
was performed by Bump et al. (1947) in the 1930s. These authors concluded that
17% of New York’s autumn Ruffed Grouse population was harvested each year
(including crippling losses) and that this was not detrimental to the overall population
based on continued abundance of breeding stock (Bump et al. 1947:379).
Bump et al. (1947:538) further suggest that harvesting up to 20% of the preseason
population is not completely additive, citing a three-year study in which harvest
losses were greater than overall differences in grouse mortality between a hunted
and unhunted area. Seasonal survival ranged from 34% to 61% across areas and
years (Bump et al. 1947:538), encompassing the range of survival estimates
from our study. At our study areas, we found that ≤11% of radio-marked Ruffed
Grouse were harvested per year, and that seasonal survival rates were similar in
both areas and years despite variability in harvest.
The potential exists, however, that some of the harvest rates found in our study
were biased low. At Fort Drum in 2008–2009, no radio-marked Ruffed Grouse
were taken by hunters, but poor capture success resulted in a small sample size early
in the hunting season, when most hunting activity likely occurred. Across study
areas and years, 4 radio signals were lost for unknown reasons and could represent
unreported harvest or emigration from the study area or technical failure.
The current hunting season in New York (5–5.5 months) is longer than that
during Bump et al.’s (1947) investigation (3 weeks to 1.5 months), and bag limits
(4 Ruffed Grouse/day, no season limit) are greater than in the 1930s (3/day,
maximum 15/season). Nevertheless, our harvest rates fell within the range of
rates estimated 8 decades ago (9.7–15.6% from check stations and 4–28% from
recorded harvest per county; Bump et al. 1947:373–375). Across the state, the
current average number of Ruffed Grouse harvested per hunter per season (2.8;
estimated over 5 years of Cooperator Ruffed Grouse Hunting Log data, New York
State Department of Environmental Conservation, Albany, NY, unpubl. data) is
406 Northeastern Naturalist Vol. 18, No. 4
similar to the average estimate over 15 years of data in the 1920s and 1930s (2.9;
Bump et al. 1947:372).
Harvest rates in New York appear to be generally at the low end of those
reported elsewhere: 5–40% in Wisconsin (Dorney and Kabat 1960), 9% in Alberta,
Canada (Rusch and Keith 1971), and 10–30% in Michigan (Clark 2000,
Palmer and Bennett 1963). The contribution of harvest to overall mortality at
our study areas (0–22% of known deaths) was within the range of estimates
from other contemporary studies of radio-marked Ruffed Grouse: 12–35%
in Michigan (Clark 2000) and 12% in the Appalachian Cooperative Grouse
Research Project (Devers et al. 2007). The final fall–winter survival probabilities
we observed (0.38–200.51) are also comparable to those estimated during
banding studies of grouse populations elsewhere (Davies and Bergerud 1988,
Edminster 1954, Rusch and Keith 1971).
Results of the competing risks analysis suggested that adults and females in
our study areas were at greater risk of harvest than juveniles and males. These
results are inconsistent with most reports in the literature (e.g., see Clark 2000,
Dorney and Kabat 1960, and Small et al. 1991). The low harvest we observed
and the timing of trapping at Fort Drum may have resulted in a spurious trend. At
Partridge Run, all radio-marked birds were captured before the hunting season,
but at Fort Drum, low capture success in 2008 necessitated trapping through October.
Birds radio-marked in October were possibly less vulnerable to harvest (or
mortality) than birds that went uncaptured and died earlier in the season. Most of
these late-season captured birds were juveniles, and more juveniles in the overall
sample were males, so results of our analyses may reflect a sampling bias. Consequently,
our finding that adults and females were most vulnerable to harvest is
open to question.
Our landscape analyses indicated that predation events at Fort Drum and
Partridge Run occurred in sites with more forest and greater patch richness than
random sites; i.e., Ruffed Grouse were killed in areas more diverse (in terms of
number of patch types present) and more forested than random areas. These attributes,
however, are typically noted in general as preferred by Ruffed Grouse for
inclusion in home ranges (e.g., Fearer and Stauffer 2004). A comparison between
live-grouse locations and mortality locations would clarify our findings. At Fort
Drum, where recent forest surveys allowed us to examine the vicinity of mortality
events on a finer scale, we found no evidence that predation events occurred
more often in large-diameter stands than would be expected at random.
Consequences for management
The results of this observational study suggest that, if representative of New
York State, overwinter mortality rates of Ruffed Grouse have not changed considerably
over 60 years of forest succession. More replication across a gradient of
habitat quality is required to quantify the relationship between seasonal survival
and forest age and condition. Yet, despite some potential bias in sampling, our
evidence suggests that harvest likely does not drive overwinter survival, and that
tightening hunting regulations would not curb further population decline. As in
2011 M.M. Skrip, W.F. Porter, B.L. Swift, and M.V. Schiavone 407
the early decades of Ruffed Grouse research in New York, harvest mortality in our
study was not a dominant factor in estimated Ruffed Grouse fall–winter mortality.
We, therefore, recommend focusing further investigation towards other factors
that may limit grouse abundance, e.g., production. Early-successional habitat may
be most important for production, nest success, and brood survival. Devers et al.
(2007) demonstrated that spring production drives population growth rates more
than overwinter mortality for Ruffed Grouse in the Appalachians. We expect the
case is true for New York Ruffed Grouse as well.
Financial support for this study was provided by the New York State Department
of Environmental Conservation (NYSDEC), including funding from the Federal Aid
in Wildlife Restoration Grant WE-173-G. Field support, study-site access, and guidance
for planning and conducting this work were provided by numerous NYSDEC and
Department of Defense-Fort Drum Military Installation personnel, ESF employees,
and volunteers. J. Frair, L. Zhang, D. Keppie, and two anonymous reviewers provided
helpful comments on earlier versions of the manuscript.
Backs, S.E., P.D. Major, C.H. Eisfelder, and J.S. Thompson. 1985. Techniques for trapping
and restocking Ruffed Grouse in Indiana. Indiana Department of Natural Resources.
Indianapolis, IN. 13 pp.
Bailey, R.G. 1995. Description of the ecoregions of the United States. Available online at
http://www.fs.fed.us/land/ecosysmgmt/index.html. Accessed 1 October 2009.
Brown, A.L., and J.A. Litvaitus. 1995. Habitat features associated with predation of
New England Cottontails: What scale is appropriate? Canadian Journal of Zoology
Bumann, G.B. 2002. Factors influencing predation on Ruffed Grouse in the Appalachians.
M.Sc. Thesis. Virginia Polytechnic Institute and State University, Blacksburg,
VA. 140 pp.
Bump, G., R.W. Darrow, F.C. Edminster, and W.F. Crissey. 1947. The Ruffed Grouse:
Life History, Propagation, Management. New York State Conservation Department,
Albany, NY. 915 pp.
Clark, M.E. 2000. Survival, fall movements, and habitat use of hunted and non-hunted
Ruffed Grouse in northern Michigan. Ph.D. Dissertation. Michigan State University,
East Lansing, MI. 174 pp.
Davies, R.G., and A.T. Bergerud. 1988. Demography and behavior of Ruffed Grouse in
British Columbia. Pp. 78–121, In A.T. Bergerud, and M.W. Gratson (Eds.). Adaptive
Strategies and Population Ecology of Northern Grouse. Volume II: Theory and Synthesis.
University of Minnesota Press, Minneapolis, MN. 408 pp.
DeGraaf, R.M., and M. Yamasaki. 2003. Options for managing early-successional forest
and shrubland bird habitats in the northeastern United States. Forest Ecology and
Dessecker, D.R., and D.G. McAuley. 2001. Importance of early-successional habitat to
Ruffed Grouse and American Woodcock. Wildlife Society Bulletin 29:456–465.
DeStefano, S., and D.H. Rusch. 1986. Harvest rates of Ruffed Grouse in northeastern
Wisconsin. Journal of Wildlife Management 50(3):361–367.
408 Northeastern Naturalist Vol. 18, No. 4
Devers, P.K., D.F. Stauffer, G.W. Norman, D.E. Steffen, D.M. Whitaker, J.D. Sole, T.J.
Allen, S.L. Bittner, D.A. Buehler, J.W. Edwards, D.E. Figert, S.T. Friedhoff, W.W.
Giuliano, C.A. Harper, W.K. Igo, R.L. Kirkpatrick, M.H. Seamster, H.A. Spiker, Jr.,
D.A. Swanson, and B.C. Tefft. 2007. Ruffed Grouse population ecology in the Appalachian
region. Wildlife Monographs 168:1–36.
Dobony, C.A., and R. Rainbolt. 2008. American Woodcock on Fort Drum Military Installation,
New York. Northeastern Naturalist 15(2):241–248.
Dorney, R.S., and C. Kabat. 1960. Relation of weather, parasitic disease, and hunting to
Wisconsin Ruffed Grouse populations. Wisconsin Conservation Department Technical
Bulletin 20. Wisconsin Conservation Department, Madison, WI. 64 pp.
Edminster, F.C. 1937. An analysis of the value of refuges for cyclic game species. Journal
of Wildlife Management 1(1–2):37–41.
Edminster, F.C. 1954. The Ruffed Grouse. Pp. 195–242, In American Game Birds of
Field and Forest: Their Habits, Ecology and Management. Charles Scribner’s Sons,
New York, NY. 490 pp.
Ellison, L.N. 1991. Shooting and compensatory mortality in tetraonids. Ornis Scandinavica
Endrulat, E.G., S.R. McWilliams, and B.C. Tefft. 2005. Habitat selection and home-range
size of Ruffed Grouse in Rhode Island. Northeastern Naturalist 12:411–424.
ESRI. 2006. ArcMap Version 9.2. ESRI, Redlands, CA.
Fearer, T.M. 1999. Relationship of Ruffed Grouse home-range size and movement to
landscape characteristics in southwestern Virginia. M.Sc. Thesis. Virginia Polytechnic
Institute and State University, Blacksburg, VA. 84 pp.
Fearer, T.M., and D.F. Stauffer. 2004. Relationship of Ruffed Grouse, Bonasa umbellus, to
landscape characteristics in southwest Virginia, USA. Wildlife Biology 10(2):81–89.
Fielding, A.H., and J.F. Bell. 1997. A review of methods for the assessment of prediction
errors in conservation presence/absence models. Environmental Conservation
Fischer, C.A., and L.B. Keith. 1974. Population responses of central Alberta Ruffed
Grouse to hunting. Journal of Wildlife Management 35:585–600.
Gullion, G.W. 1965. Improvements in methods for trapping and marking Ruffed Grouse.
Journal of Wildlife Management 29:109–116.
Heisey, D.M., and B.R. Patterson. 2006. A review of methods to estimate causespecifi
c mortality in presence of competing risks. Journal of Wildlife Management
Henner, C.M., M.J. Chamberlain, B.D. Leopold, and L. Wes Burger, Jr. 2004. A multiresolution
assessment of Raccoon den selection. Journal of Wildlife Management
Hewitt, D.G., and R.L. Kirkpatrick. 1997. Daily activity times of Ruffed Grouse in southwestern
Virginia. Journal of Field Ornithology 68(3):413–420.
Hunyadi, B.W. 1984. Ruffed Grouse restoration in Missouri. Pp. 21–35, in W.L. Robinson
(Ed.). Ruffed Grouse Management: State of the Art in the Early 1980s. North
Central Section of the Wildlife Society, Bethesda, MD. 181 pp.
Kaplan, E.L., and P. Meier. 1958. Nonparametric estimation from incomplete observations.
Journal of the American Statistical Association 53(282):457–481.
Kunkel, K.E., and D.H. Pletscher. 2000. Habitat factors affecting vulnerability of Moose
to predation by Wolves in southeastern British Columbia. Canadian Journal of Zoology
2011 M.M. Skrip, W.F. Porter, B.L. Swift, and M.V. Schiavone 409
Lunn, M., and D. McNeil. 1995. Applying Cox regression to competing risks. Biometrics
McDonald, J.H. 2009. Fisher’s exact test. Pp. 70–75, In Handbook of Biological Statistics,
Second Edition. Sparky House Publishing, Baltimore, MD. 319 pp. Available
online at http://udel.edu/~mcdonald/statfishers.html. Accessed 1 October 2009.
McGarigal, K., S.A. Cushman, M.C. Neel, and E. Ene. 2002. FRAGSTATS: Spatial Pattern
Analysis Program for Categorical Maps. Computer software program produced
by the authors at the University of Massachusetts, Amherst. Available online at www.
umass.edu/landeco/research/fragstats/fragstats.html. Accessed 1 December 2007.
Migoya, R., and G.A. Baldassarre. 1995. Winter survival of female Northern Pintails in
Sinaloa, Mexico. Journal of Wildlife Management 59(1):16–22.
Mitchell, B.R. 2005. FragStatsBatch for ArcGIS 9. Computer software. Available online
at http://arcscripts.esri.com/details.asp?dbid=13995. Accessed 26 March 2008.
Moser, J., J. Herron, B. Schongar, N. McBride, E. Haslun, J. McDermott, N. Channing,
P. McLaren, B. Henry, D. Martin, T. Folts, and P. Haskins. 2001. Helderbergs Unit
management plan: A management unit consisting of three state forests and three wildlife
management areas in western and southeastern Albany County. New York State
Department of Environmental Conservation, Albany, NY. 73 pp. Available online
at http://www.dec.ny.gov/docs/lands_forests_pdf/helderbergsump.pdf. Accessed 1
National Audubon Society, Inc. 2008. State of the birds. Common birds in decline:
Ruffed grouse. Available online at http://stateofthebirds.audubon.org/cbid/profile.
php?id=20. Accessed 15 April 2009.
NYSDEC Habitat Inventory Unit. 1990. Ecological Zones—New York State. New York
State Department of Environmental Conservation (NYSDEC), Albany, NY. Vector
digital data. Available online at http://www.nysgis.state.ny.us/gisdata/inventories/
details.cfm?DSID=1131. Accessed 19 October 2009.
Palmer, W.L., and C.L. Bennett, Jr. 1963. Relation of season length to hunting harvest of
Ruffed Grouse. The Journal of Wildlife Management 27(4):634–639.
Pearce, J., and S. Ferrier. 2000. Evaluating the predictive performance of habitat models
developed using logistic regression. Ecological Modelling 133:225–245.
Pollack, K.H., S.R. Winterstein, C.M. Bunck, and P.D. Curtis. 1989a. Survival analysis
in telemetry studies: The staggered entry design. Journal of Wildlife Management
Pollack, K.H., S.R. Winterstein, and M.J. Conroy. 1989b. Estimation and analysis of
survival distributions for radio-tagged animals. Biometrics 45(1):99–109.
Post, T. 2008. Ruffed Grouse, Bonasa umbellus. Pp. 138–139, In K.J. McGowan and K.
Corwin (Eds.). Second Atlas of Breeding Birds in New York State, Cornell University
Press, Ithaca, NY. 688 pp.
Roberts, S.D. 1993. Survival and reproduction of Wild Turkeys in south-central New
York. M.Sc. Thesis. State University of New York College of Environmental Science
and Forestry, Syracuse, NY. 81 pp.
Roussel, Y.E., and R. Ouellet. 1975. A new criterion for sexing Quebec Ruffed Grouse.
Journal of Wildlife Management 39:443–445.
Rusch, D.H., and L.B. Keith. 1971. Seasonal and annual trends in numbers of Alberta
Ruffed Grouse. Journal of Wildlife Management 35:803–822.
SAS Institute Inc. 1999. SAS/STAT ® User’s Guide, Version 8. Cary, NC. 1464 pp.
Sauer, J.R., J.E. Hines, and J. Fallon. 2008. The North American Breeding Bird Survey:
Results and analysis 1966–2007. Version 5.15.2008. USGS Patuxent Wildlife Research
Center, Laurel, MD.
410 Northeastern Naturalist Vol. 18, No. 4
Servello, F.A., and R.L. Kirkpatrick. 1986. Sexing Ruffed Grouse in the Southeast using
feather criteria. Wildlife Society Bulletin 14:280–282.
Small, R.J., J.C. Holzwart, and D.H. Rusch. 1991. Predation and hunting mortality of
Ruffed Grouse in central Wisconsin. Journal of Wildlife Management 55:512–520.
Thogmartin, W.E., and B.A. Schaeffer. 2000. Landscape attributes associated with mortality
events of Wild Turkeys in Arkansas. Wildlife Society Bulletin 28(4):865–874.
Thomas, D.L., and E.J. Taylor. 2006. Study designs and tests for comparing resource use
and availability II. Journal of Wildlife Management 70(2):324–336.
Thompson, F.R., III, and D.R. Dessecker. 1997. Management of early-successional
communities in central hardwood forests: With special emphasis on the ecology and
management of oaks, Ruffed Grouse, and forest songbirds. General Technical Report
NC-195. US Department of Agriculture, Forest Service, North Central Forest Experiment
Station, St. Paul, MN. 33 pp.
Whitaker, D.M., D.F. Stauffer, G.W. Norman, P.K. Devers, J. Edwards, W.M. Giuliano,
C. Harper, W. Igo, J. Sole, H. Spiker, and B. Tefft. 2007. Factors associated with
variation in home-range size of Appalachian Ruffed Grouse (Bonasa umbellus). Auk
White, G.C., and R.A. Garrott. 1990. Analysis of Wildlife Radio-Tracking Data. Academic
Press, London, UK. 383 pp.
Yoder, J.M., E.A. Marschall, and D.A. Swanson. 2004. The cost of dispersal: Predation
as a function of movement and site familiarity in Ruffed Grouse. Behavioral Ecology