Seasonal migrations of black bears (Ursus americanus): Causes and consequences

Abstract

American black bears frequently abandon their home ranges in late summer and move to feeding areas to fatten themselves for
hibernation. We examined seasonal movements of 206 radio-collared bears in north-central Minnesota during 1981–1990. We exploited
the variability in this long-term data set to test tradeoffs for animals leaving their home range. Late summer movements were
common for both sexes and all ages (39% of females, 44% of males), but were variable from year-to-year in prevalence, timing,
and destination. Bears typically left their summer home ranges in August and returned ~6weeks later in September or October.
Most traveled southward, where acorns were more plentiful (median = 10km for females, 26km for males; maximum = 168km).
These facultative migrations were most common when rich resources were available outside home ranges. Bears were least apt
to leave when foods were scarce in their home range, possibly sensing a risk of migrating during a widespread food failure.
Among females, those whose body mass was close to a reproductive threshold were most prone to migrate. Migrating bears were
less likely to be killed by hunters, suggesting that they were especially vigilant.

KeywordsCost/benefit trade-offs–Food abundance–Hunting mortality–Oak mast–Reproductive threshold–Seasonal movements

Full text

ORIGINAL PAPER
Seasonal migrations of black bears (Ursus americanus):
causes and consequences
Karen V. Noyce &David L. Garshelis
Received: 16 September 2010 / Revised: 1 October 2010 / Accepted: 6 October 2010
#Springer-Verlag 2010
Abstract American black bears frequently abandon their
home ranges in late summer and move to feeding areas to
fatten themselves for hibernation. We examined seasonal
movements of 206 radio-collared bears in north-central
Minnesota during 1981–1990. We exploited the variability
in this long-term data set to test tradeoffs for animals
leaving their home range. Late summer movements were
common for both sexes and all ages (39% of females, 44%
of males), but were variable from year-to-year in preva-
lence, timing, and destination. Bears typically left their
summer home ranges in August and returned ~6 weeks later
in September or October. Most traveled southward, where
acorns were more plentiful (median = 10 km for females,
26 km for males; maximum = 168 km). These facultative
migrations were most common when rich resources were
available outside home ranges. Bears were least apt to leave
when foods were scarce in their home range, possibly
sensing a risk of migrating during a widespread food
failure. Among females, those whose body mass was close
to a reproductive threshold were most prone to migrate.
Migrating bears were less likely to be killed by hunters,
suggesting that they were especially vigilant.
Keywords Cost/benefit trade-offs .Food abundance .
Hunting mortality .Oak mast .Reproductive threshold .
Seasonal movements
Introduction
Animals that rely on geographically shifting resources can
either move with those resources or stay put and withstand
periods of scarcity. Consistent seasonal changes in resource
distribution give rise to the predictable migratory patterns
found in a broad range of species (Dingle 1996). More
variable conditions favor behavioral flexibility that enables
animals to adjust movements from year to year in
accordance with resource availability (Newton 2006).
Animals that travel outside their familiar territory in an
attempt to find better food sources are subject to heightened
risk. Travel itself is energetically costly (Alerstam et al. 2003)
and extra time and energy may be required to locate food
and navigate in novel surroundings. Lack of familiarity with
local conditions may increase exposure to harm (Nicholson
et al. 1997) and the need for extra vigilance can reduce
foraging efficiency (Brown et al. 1999). In travelling to
unfamiliar places, animals ultimately risk failing to find
better foods than those they left behind (John and Roskell
1985). To be adaptive, travel must provide sufficient benefit
to offset these costs and enhance reproduction or survival
(Brönmark et al. 2008).
American black bears are generally not recognized as
typical “round-trip”migrators (Dingle 1996), yet there are
many accounts of their long-distance movements to
concentrated food sources in late summer and fall (up to
200 km; Rogers 1987a). These travels occur during a time
of hyperphagia, when bears extend daily foraging time
(Garshelis and Pelton 1980; Lariviere et al. 1994)and
increase caloric consumption (Hashimoto and Yasutake
1999; Hilderbrand et al. 1999) in preparation for hiberna-
tion. The specific foods they seek vary geographically, but
most are fruits or nuts that exhibit pronounced stochas-
ticity in annual yield (Koenig and Knops 2000;McShea
Communicated by M. Festa-Bianchet
K. V. Noyce (*):D. L. Garshelis
Minnesota Department of Natural Resources,
Forest Wildlife Populations and Research Group,
1201 East Highway 2,
Grand Rapids, MN 55744, USA
e-mail: karen.noyce@state.mn.us
Behav Ecol Sociobiol
DOI 10.1007/s00265-010-1086-x

and Schwede 1993;NoyceandCoy1990). Variability in
the prevalence and geographic extent of seasonal travel,
both within and among bear populations, has confounded
the development of a cohesive explanation of this
behavior.
The sporadic late-summer movements of bears resemble
the “partial,”“facultative,”and sometimes “irruptive”
migrations described for other vertebrate species (Newton
2006; Olsson et al. 2006; White et al. 2007). As in other
facultative migrators (Newton 2006), food scarcity is
generally thought to trigger such movements (Hellgren et
al. 2005; Schorger 1946). However, contradictory observa-
tions have also been reported (Kasbohm et al. 1998; Rogers
1987a; Schooley et al. 1994).
As for a host of migratory species (Dingle 1996), the
potential for gain or loss in undertaking seasonal travel will
differ among individuals by sex, age, size, reproductive
status, and local habitat conditions. Bears also are con-
strained in that, instead of moving to better food conditions
and remaining there during the period of food scarcity, their
movements are directed at finding sufficient food to store as
fat to survive an extended winter fast (hibernation).
Females that find rich food supplies in the fall may aid
the growth and survival of cubs traveling with them or yet
unborn (Noyce and Garshelis 1994). Males that increase
growth in the fall may enhance the likelihood of siring cubs
the following spring (Kovach and Powell 2003; Costello et
al. 2009). However, animals that move into unfamiliar areas
may suffer greater mortality (Hellgren et al. 2005; Pelton
1989; Schwartz and Franzmann 1992), so this must be
weighed against the potential benefits.
Here, we document the late-summer movements of
Minnesota black bears over a 10-year period and exploit
the variability we observed to test the following hypothe-
ses: (1) Individuals leave their summer ranges if the
likelihood is high that they will find better food elsewhere,
(2) individuals are most likely to travel in search of better
foods if their body mass is close to a reproductive or
survival threshold, and (3) travel behavior reflects decisions
that minimize risk. We investigate the influence of sex, age,
body size, natural food abundance, and habitat, and
interpret results from a risk–benefit perspective.
Materials and methods
Study site
The study area encompassed 360 km
2
in the Chippewa
National Forest (CNF) and adjoining George Washington
State Forest (47°30′N, 93°30′W) in north-central Minnesota.
The terrain was >95% forested and shaped by Pleistocene
glaciations. A distinct ecotone divided the hilly St. Louis
Moraines (SM) landscape of mixed uplands, lowlands, and
lakes, from the flat lowland landscape of the Chippewa
Plains (CP) glacial outwash (Minnesota Department of
Natural Resources 2003). Female bears living near this
ecotone exhibited a strong fidelity to either SM or CP during
most of the year, whereas males traversed larger, more
heterogeneous areas. Upland forests, comprising about 67%
of the study area, were an aspen-dominated mix (Populus
tremuloides,Betula papyrifera,andAbies balsamea), with
scattered pine (Pinus resinosa and Pinus strobus) and
hardwoods (Acer saccharum,Tilia america na,andQuercus
rubra). Lowland forests contained mainly black spruce
(Picea mariana), tamarack (Larix laricina), northern white
cedar (Thuja occidentalis), and black ash (Fraxinus nigra).
Principal fruits available to bears during July–August
included wild sarsaparilla (Aralia nudicaulis), blueberries
(Vaccinium spp.), raspberries (Rubus idaeus), Juneberries
(Amelanchier spp.), and cherries (Prunus virginiana and
Prunus pennsylvanica). Dogwood berries (Cornus spp.),
hazelnuts (Corylus cornuta), and acorns, primarily of red
oak (Q. rubra), were the main foods during late August–
September. Berry and nut production was highly variable
from year to year (Noyce and Coy 1990), which was
reflected in bears' diets (Garshelis and Noyce 2008).
Lowland habitats produced less berry and nut biomass than
upland habitats (Noyce and Coy 1990).
The study area supported timber production, seasonal
homes, and forest- and lake-centered recreation. Two open
landfills were used regularly by bears until 1986, when they
were converted to secure dumpsters. Bear hunting was legal
from September 1 to mid-October, and most hunters
attracted bears with bait. Bear density was ca. 20 bears/
100 km
2
(Garshelis and Noyce 2008) and harvest density
was among the highest in the state (on average, >6 bears
killed/100 km
2
).
Delineation of home range and seasonal movements
Bears were captured in baited barrel traps or Aldrich foot
snares and fitted with VHF radio-collars (Telonics, Mesa, AZ,
USA) during May–July, 1981–1989. Radio-collared bears
were handled annually in their winter dens, at which time
yearlings (denned with their mothers) were fitted with radio-
collars. We immobilized bears with ketamine hydrochloride
(11–13 mg/kg) and xylazine (0.6–0.7 mg/kg), or premixed
tiletamine hydrochloride and zolazepam (Telazol®, Elkins-
Sinn, Cherry Hill, NJ, USA, 3.9–5.3 mg/kg). At first handling,
we extracted a first upper premolar for estimating age and
discerning reproductive history (Coy and Garshelis 1992;
Willey 1974). Bears were weighed with hanging spring
scales and body measurements were recorded to the nearest
centimeter. Handling procedures were in accordance with
companion work that was approved by the Institutional
Behav Ecol Sociobiol

Animal Care and Use Committee of the University of
Minnesota. Radio-collared bears were located from fixed-
wing aircraft during daylight hours at intervals of 3–5days
in 1981–1982, 6–9daysin1983–1984, and weekly or bi-
weekly during 1985–1990.
We defined a bear's summer home range (or simply
home range or home) as the area where it centered its
activities for most of the year, including, as a minimum, the
mid-May to mid-July breeding season. Departures from the
home range were easily identified by a sudden, distinct
traverse to a new area. We delimited the summer home
range using a minimum convex polygon (MCP) encom-
passing all telemetry locations within 2 km (females) or
5 km (males) of at least one other point in the cluster.
Beyond those distances, locations were considered move-
ments outside the summer range (Fig. 1a). Using these
criteria, nearly all points were included within the home
range during early summer, when bears traveled little, but
these thresholds were exceeded at other times of the year,
enabling us to detect movements to new areas. To delineate
MCPs, we pooled locations across years for individual
bears, unless home range shifts between years were evident.
We measured movements outside the home range along a
perpendicular line from the perimeter of the MCP to the
most distant location (Fig. 1b). Direction of travel was
measured to the same point from the geographic center of
the MCP.
We classified movements as either seasonal forays
(moves from which bears returned) or dispersal (perma-
nent departures of bears known to be born in the study
area). We separated seasonal forays by the time of year
that they occurred: (1) early season, commencing April–
June, with return usually before mid-July; (2) late season,
initiated after 1 July, with return typically after 1
September; or (3) overwinter, to dens outside the summer
home range, with return in the spring. We focus this
paper primarily on the characteristics and causes of late-
season forays.
We could not precisely define the timing and duration
of seasonal forays because individuals often left their
home range abruptly and traveled quickly, so we often
could not find them for a week or more after they left.
Thus, instead of estimating departure dates, we used the
date of each bear's last known location within its summer
range. This way, bears that traveled farther did not appear
to leave later simply as an artifact of the difficulty in
finding them. Similarly, for their returns, we used the
date that we detected them back in their summer range.
Since we routinely monitored all radio-collar frequencies
in the central study area, we were apt to note the
disappearance of bears shortly after their leaving and
likewise hear their radio signals shortly after they arrived
back from forays.
Assessment of foods
We investigated the relationship between the abundance of
natural foods (fruits and nuts) and year-to-year differences in
the late summer movements of bears. Natural resources
personnel across the bear range in Minnesota provided annual
ratings of fruit production, on a 0–4 scale (2 = “average”), for
each of 14 different food types (Noyce and Garshelis 1997).
We used surveys conducted within ≈40 km of the study area
in conjunction with our ratings in the study area to
characterize local food abundance. To corroborate indices,
we quantified fruits and hazelnuts in 45–100 forest stands
during 1984–1989 (Noyce and Coy 1990). Acorns were
counted separately, sampling ten trees each year (1982–
1990) in six red oak and 1–2buroak(Quercus macrocarpa)
stands (Whitehead 1969).
Fig. 1 Delineation of MCP home ranges and method for defining and measuring movements by black bears outside their summer range
Behav Ecol Sociobiol

Data analysis
We used chi-square analysis to test for differences in the
frequency of seasonal movements outside the summer
home range by sex, age (1, 2, 3, and >4 years old [adult]),
and reproductive class. We compared mean duration and
median distance of early vs late-season travel using two-
sample ttests and Kruskal–Wallace nonparametric AOV.
We grouped bears by azimuth of travel (12 groups, each
spanning 30° of arc) and used chi-square to compare travel
orientation of males and females and chi-square goodness-
of-fit to test for deviation from random.
We modeled the effects of sex, age, and abundance of
key bear foods on the likelihood, timing, distance, and
duration of late season movement using logistic and
linear regression, and compared models with Akaike's
information criterion (AIC
c
). For food covariates, we used
the annual local (within ≈40 km) productivity indices of
oak, hazel, chokecherry, and sarsaparilla, raspberry, and
chokecherry, summed (three important summer foods). To
allow for non-ordinal effects, we represented age as two
binary (0,1) parameters that separately identified 2-year-
olds and 3+-year-olds, with yearlings as the reference
group. Collinearity of parameters was acceptable (variance
inflation factors for main covariates ≤3.0). We posed two a
priori models, one including covariates for food abun-
dance (with interactions) and one without. We sequentially
eliminated the least important covariate, identified by
minimal absolute value of b/SE (Arnold 2010), continuing
until elimination of additional covariates increased AIC
c
by ≥2.0. When a covariate was identified as unimportant,
we first eliminated interaction terms involving that
covariate, only then eliminating the main covariate, if still
warranted. If two covariates had equivalent scores, we
modeled each elimination separately, then continued
eliminations from both of these models. Tables present a
priori models, plus the three or four best reduced models
so derived, recognizing that there is debate regarding the
best method for defining a set of candidate models (Arnold
2010; Symonds and Moussalli 2010). For females, we
modeled the effect of body mass and home range
landscape (CP or SM) on the likelihood of late-season
travel. We categorized body mass as: <25 kg, 25−34.9 kg,
35−49.9 kg, 50−69.9 kg, and ≥70 kg. In one model, to
allow for non-ordinal effects, the higher categories were
represented as four separate binomial parameters, with
bears weighing <25 kg comprising the reference group.
Each year that an individual bear was monitored consti-
tuted one record. Though we recognize that multiple
records for individuals may not be fully independent, the
degree of behavioral variation that we observed within
individuals from year to year indicated that the effect of
the individual was minor relative to other covariates.
We compared daily survival of bears that remained
within their home range vs those that traveled outside
during the first 2 weeks of fall bear hunting, using the
Gehan-Wilcoxon two-sample survival test (Statistix 9,
2008, Analytical Software, Tallahassee, FL, USA). Hunter
kills and locations were ascertained through mandatory
hunter reporting.
Results
During 1981–1990, we tracked the movements of 206
individual bears (82 females, 124 males), aged 1–19 years
old. Individuals were monitored 1–10 years each, totaling
540 bear-years (297 female, 243 male); 62 females and 65
males were followed multiple years. Travel outside the
summer home range was common for both sexes, occurring
in 43% (F) and 70% (M) of bear-years monitored (Table 1).
Two-year-olds of both sexes were the most likely to travel
(females: χ
2
=13.5, df=4, P=0.004; males: χ
2
=13.9, df=3,
P=0.003). All males born in the study area either dispersed
(n=27), died, or were otherwise lost from the study by age
4, whereas only one of 42 females dispersed. Dispersal
occurred during all non-hibernating months (April–Novem-
ber) with nomadic movements spanning 5 days to months,
sometimes >1 year.
Early-season forays (April–June) were made by 8% of
bears, but were most common for 3-year-old females and 2-
year-old males (Table 1; females: χ
2
=15.3, df=4, P=
0.004; males: χ
2
=14.4, df=3, P=0.002). Median move-
ment was 3 km for females and 9 km for males, about half
the typical length of the long axis of the average home
range; maximum distances were 41 km and 83 km,
respectively. Early-season forays showed no predominant
orientation (χ
2
=4.3, df=11, P=0.96). Males typically
stayed away longer than females (mean: 27 vs 17 days,
respectively). Of eight subadult (preparous) females that
made early-season forays, seven were of breeding age and
size; three were later confirmed to have been in estrus that
spring. Of five adult females making early-season forays,
four (three with cubs, one with yearlings) traveled to active
garbage dumps.
Late-season forays were more common than early-
season forays (Table 1). Of bears that we tracked for ≥3
consecutive years, 87% made at least one late-season foray
(41 of 47 females, 24 of 28 males; median years tracked:
4.5 for females, 4.0 for males). However, only 10 of 35
females and two of 15 males tracked ≥4 years made late
season forays in 3 consecutive years. The proportion of
bears that moved varied from year to year (range: 25–64%
of juveniles, 3–87% of adults). For females, this appeared
unrelated to whether they had cubs (Table 1). Late-season
forays were longer in distance (median = 10 km for females
Behav Ecol Sociobiol

[χ
2
=10.3, df=1, P=0.0013], 26 km for males [χ
2
=16.08,
df=1, P=0.0001]) and in duration (mean = 39 days for
females [t=3.9, d.f=131, P=0.0002]), 46 days for males
[t=3.8, df= 111, P=0.0002]) than early-season travels.
Movement was strongly directional (females: χ
2
=67.0,
df=11, P<0.0001; males: χ
2
=88.8, df=11, P=0.0001),
tending to the south and southwest (Fig. 2). Direction of
movement differed for males and females, suggesting
segregation in fall feeding areas (χ
2
=29.4, df=11,
P=0.002). Though departure from home ranges occurred
from early July to late October, most bears left during late
July and August, with 35% departing between July 29 and
August 11. Most returned home in September or early
October, but 25% of males on forays moved up to 144 km
(median = 49 km, n=47) directly from late-season foraging
areas to den sites outside their summer home range (usually
to the north) and did not return home until spring (Table 1).
Late-season and overwinter travels easily fit the definition
of migration in being: (1) highly seasonal, directional, and
outside the home range; (2) conducted by many individu-
als; (3) resulting in a redistribution of the population; and
(4) temporary, from which individuals eventually returned
home (Dingle 1996, Dingle and Drake 2007).
Regression modeling indicated that food availability
influenced migration behavior; in all cases, addition of
food covariates markedly improved model fits (Tables 2A
and 3, all models II vs III; Table 2B, models II and III vs
IV). The best-supported reduced models (Ia–d) all includ-
ed ≥3 food covariates, along with various sex × food and
age × food interactions. Oak (acorn) abundance had the
largest effect on probability of travel: for example, when
we set sex and age at “0”(=yearling male) and held other
food indices constant, model Ia (Table 2, A) predicted that
an increase from 1 to 2 in the acorn index added 0.14 to
the calculated probability of migration. The same increase
in hazel or summer food index (holding others constant)
added only 0.01–0.02. The influence of acorn abundance
was greatest for adults (age × oak, model Id, Table 2,A);
with other foods constant, an increase from 1 to 3 in the
oak index (representing a change from “below average”to
Fig. 2 Direction of late-season forays made by male and female black
bears in north-central Minnesota. Arrows represent 30° increments in
travel orientation and the length of each arrow represents the percent
of forays oriented in that direction. Travel orientation differed between
males and females and differed from random for both sexes
Table 1 Percent of radio-collared bears of each sex and age that moved outside their regular summer home range in a year, north-central
Minnesota, USA, 1981–1990
Sex–age of bear Number Early season (April–Jun) % Late season (Jul–Sep) % Overwinter % Dispersal % All types %
Females:
1 year old 46 2 29 4 0 31
a
2 years old 36 0 64 8 3 69
3 years old 39 15 31 8 3 41
Adult with cubs 78
b
5424049
Adult no cubs 98
b
2341036
All 297
b
4394 43
Males:
1 year old 60 10 42 5 17
c
61
2 years old 53 23 33 17 48
c
90
3 years old 39 13 56 23 5
c
69
Adult 91
b
847271
cd
65
All 243
b
12 44 20 70
a
Some bears made more than one type of movement in a year; thus, columns do not sum to this value
b
Bears monitored for >1 year provide a data record for each full year monitored, thus some individuals account for >1 record in the adult age class
c
Although all males born on the study area dispersed, percentages of males shown here dispersing at each age do not sum to 100% because many bears
included had already dispersed from elsewhere into the study area
d
The oldest male to disperse left his natal range at 4 years old
Behav Ecol Sociobiol

“above average”) added a substantial 0.39 to the estimated
migration probability for adults, but only 0.13 for year-
lings. In 1988, when acorns were the most plentiful, 26 of
29 adults migrated vs only one of 28 in 1985, when acorns
(and most other bear foods) failed (Fig. 3a). However,
bears that migrated during poor acorn years often traveled
far. A single adult male that migrated in 1985, and three of
six that did so in 1990 (another poor year), made four of
the five longest moves recorded during this study (114–
168 km, one-way, straight-line).
Model Id (Table 2, A) suggested that chokecherry
abundance was particularly significant for young females; a
change from 1 to 3 in the index added 0.11–0.16 to
migration probability for females <3 years old, but had
negligible effect on other bears (sex × chokecherry and age ×
chokecherry ). The same change in hazel added 0.17 to the
migration probability for females, but not males (Table 2,A,
model Id, sex × hazel). All models (Table 2, A), indicated
that 2-year-old females and adult males were most likely to
migrate and yearling females least likely (sex × age). Two-
year-old males exhibited less migration than other males
because many were dispersing.
Among females (Table 2,B,Ia–d), the odds of migrating
were 2.1–2.4 times higher for those living in lowland
landscapes than for those in uplands. Model Id suggested
an age × weight interaction wherein heavy yearlings (25–
34.9 kg) were more likely to migrate than lightweight
yearlings (<25 kg), but in older bears (all ≥35 kg),
lightweight individuals (35–49.9 kg) were more likely to
move.
Differences in the timing, distance, and duration of late-
season travels were also influenced by sex, age, and food
abundance (Table 3). Males 1–2 years old generally left
home earlier, traveled farther, and stayed away longer than
older bears (Tables 3and 4, all models, sex and age
effects); 1 and 2-year-old females were away the shortest
time. High hazelnut abundance prompted females to move
sooner than usual, but not males (Table 3, A, models Ia–Id,
sex × hazel interaction). Bears left sooner, traveled farther,
and stayed away longer when acorns were plentiful (Table 3,
all models I and II, oak effect) and they tended to leave
later when chokecherry was abundant (Table 3, A, models
Ia–Ic, chokecherry effect). In 1988, a year with outstanding
acorn production and good summer berry production (albeit
cut short by a July drought), bears left home ranges
particularly early (Table 4). Conversely, in 1983, when
berry production (especially chokecherry) was the best in
this study, most bears delayed their migrations by 3–
6 weeks compared to 1988. We discerned no effect of
covariates on migration return dates.
Females living near the interface of the CP lowlands and
the SM uplands provided an enlightening example of
Table 2 Akaike's information criterion (AIC
c
) model selection for logistic regressions of probability of late-summer migration by black bears in
north-central Minnesota as a function of sex, age, natural food availability, home range habitat, and body mass
Sample Model
a
Model covariates
b
Deviance KAIC
c
ΔAIC
c
A. All bears
(n=542)
Ia SX + A(2) + O + S + H + SX×A(2) + SX×H + SX×S 657.2 11 679.7 0.0
Ib SX + A(2) + O + S + Ch + H + SX×A(2) + SX×H 657.2 11 679.7 0.0
Ic SX + A(2) + O + S + Ch + SX×A(2) + SX×O 659.73 10 680.1 +0.4
Id SX + A(2) + O + S + Ch + H + SX×A(2) + SX×Ch + SX×S + SX×H + A(2)×O + A
(2)×Ch
645.9 17 681.1 +1.4
II SX + A(2) + O + S + Ch + H + SX×A(2) + SX×O + SX×Ch + SX×S + SX×H + A
(2)×O + A(2)×H + A(2)×C + A(2)×S
642.5 22 688.4 +8.7
III SX + A(2) + SX×A(2) 720.5 6 726.7 +47.0
B. Females
c
(n=185)
Ia A(2) + L + WT + O + S + Ch + H + A(2)×Ch 195.2 11 218.7 0.0
Ib A(2) + L + O + S + Ch + H + A(2)×Ch 198.6 10 219.6 +0.9
Ic A(2) + L + WT + O + S + Ch + H 200.9 9 219.9 +1.2
Id A(2) + L + WT + O + S + Ch + H + A(2)×Ch + A(2)×WT 192.7 13 220.8 +2.1
II A(2) + L + WT + A(2)×WT + O + S + Ch + H + A(2)×O + A(2)×S + A(2)×Ch + A
(2)×H
191.5 19 231.0 +12.3
III A(2) + L + O + S + Ch + H + A(2)×O + A(2)×S + A(2)×Ch + A(2)×H 197.2 16 232.4 +13.7
IV A(2) + L 237.6 4 245.8 +28.4
a
Models Ia–d represent the best-fitting models, based on AIC
c
derived by reduction from a priori models II–IV. A priori models are included to highlight
the importance of food abundance in explaining likelihood of migration
b
Model covariates: SX sex, A(2) age represented as two binomial parameters, LLandscape (SM uplands or CP lowlands), Ooak production index, S
summed fruit production indices of wild sarsaparilla, raspberry, and chokecherry (important mid-summer bear foods, chosen based on prevalence in scats),
Ch chokecherry production index, Hhazel production index
c
Model B includes covariates for landscape type (only females showed landscape fidelity in their summer home ranges) and body mass, standardized by
date
Behav Ecol Sociobiol

within-population variation in migration behavior. Both
upland and lowland females exhibited strong fidelity to
their home landscape during spring and early summer.
However, in late summer, lowland females commonly (30
of 60 bear years sampled) left their home range and most
often (87% of the time, excluding 1988) traveled to the
Suomi Hills, with the highest density of red oaks in the
study area. Only one male was located in Suomi Hills
during that time of year. Males tended to travel farther
south, to areas only occasionally visited by radio-collared
females. Females that lived in the Suomi Hills, however,
rarely left home (Fig. 4a, b).
In 1988, when acorn production (particularly bur oak)
was exceptionally high (index 3.3 vs mean 1.9), almost all
collared adult females, including all but one from Suomi
Hills, left their summer range (Fig. 4c, d) and traveled up to
60 km (mean 20 km) to late-summer feeding areas. That
year, females bypassed the Suomi Hills and continued
farther south and southwest to novel destinations where bur
oak (Q. macrocarpa), which was rare in the study area, was
more common. Though we were generally unable to detect
differences in body mass, growth, or reproduction that
could be attributed to migration, eight lactating females that
migrated with their cubs in 1988 lost less weight between
winters (x¼8:5kg, range 4–16 kg; x¼11:5% of body
mass lost, range 5–22%) than the sole lactating female that
did not migrate (20 kg=24% of body mass).
Contrary to expectation, migrating bears were no more
vulnerable to hunters than non-migrators. In fact, during the
first week of the hunting season (1–7 Sept), when most
bears were killed, hunting mortality was higher for females
that remained at home than for those that migrated (Cox’sF
test: F
(52,2)
=0.23, P=0.03). This was also true for males,
but not significantly so (F
(46,20)
=0.61, P=0.17; Fig. 5).
Hunting mortality during later weeks of the hunting season
was very low and not discernibly different for migrators
and non-migrators. Non-hunting mortality also was low for
migrating bears. In 10 years, 13 non-migrating bears were
shot illegally during 15 July–15 October, or as a result of
nuisance activity, and one was hit by a car. No migrating
bears were shot other than by legal hunters during that time
and two were hit by cars on major highways.
Discussion
Most modern studies have not recognized bears as
migrators, because their movements do not occur en masse
and are not readily observed, but some historic accounts,
drawing from traditional and anecdotal observations,
Table 3 Akaike's information criterion (AIC
c
) selection for linear regression models of departure dates, duration, and distance of seasonal black
bear migrations, as a function of sex, age, and natural food availability in north-central Minnesota
Response Model
a
Model covariates
b
KAIC
c
ΔAIC
c
Adj R
2
A. Departure
date
Ia SX + A(2) + O + Ch + H + SX×H + A(2)×H + A(2)×Ch 12 1,339.2 0.0 0.18
Ib SX + A(2) + O + S + Ch + H + SX×S + SX×H + A(2)×H + A(2)×Ch 14 1,340.3 +1.1 0.19
Ic SX + A(2) + O + S + Ch + H + SX×O + SX×S + SX×H + A(2)×H + A(2)×Ch 15 1,341.9 +2.7 0.18
II SX + A(2) + O + S + Ch + H + SX×A(2) + SX×O + SX×Ch + SX×S + SX*H + A(2)×O +
A(2)×H + A(2)×Ch + A(2)×S
22 1,353.4 +13.2 0.18
III SX + A(2) + SX×A(2) 6 1,372 +32.8 0.02
B. Duration Ia SX + O + Ch + H + SX×O + SX×Ch + SX×H 8 1,094.8 0.0 0.16
Ib SX + A(2) + O + Ch + H + SX×O + SX×Ch + SX×H + A(2)×Ch 12 1,098.3 +3.5 0.17
Ic SX + A(2) + O + Ch + H + SX×O + SX×Ch + SX×H + A(2)×O, A(2)×Ch 14 1,098.7 +3.9 0.18
II SX + A(2) + O + S + Ch + H + SX×A(2) + SX×O + SX×Ch + SX×S + SX*H + A(2)×O +
A(2)×H + A(2)×Ch + A(2)×S
22 1,115.6 +21.8 0.16
III SX + A(2) + SX×A(2) 6 1,123 +28.2 0.01
C. Log
(distance)
c
Ia SX + A(2) + O + H + S + SX×O + SX×S + SX×H 10 −436.9 0.0 0.38
Ib SX + O + S + H + SX×O + SX×S + SX×H 8 −435.1 +1.8 0.37
Ic SX + A(2) + O + Ch + H + S + SX×O + SX×S + SX×H 11 −435.0 +1.9 0.38
II SX + A(2) + O + S + Ch + H + SX×A(2) + SX×O + SX×Ch + SX×S + SX×H + A(2)×O + A
(2)×H + A(2)×Ch + A(2)×S
22 −416.4 +20.4 0.36
III SX + A(2) + SX×A(2) 6 −395.2 +39.8 0.23
a
Models Ia–c represent the best-fitting models, based on AIC
c
derived by reduction from a priori models II and III. A priori models are included to
highlight the importance of food abundance in explaining migration characteristics
b
Model covariates: SX sex, A(2) age represented as two binomial parameters, Ooak production index, Ssummed production indices of wild sarsaparilla,
raspberry, and chokecherry, important mid-summer bear foods, chosen based on prevalence in scats, Ch chokecherry production index, Hhazel production
index
c
Distance values were highly skewed, so were log-transformed to achieve normal distribution
Behav Ecol Sociobiol

recognized the migratory nature of these seasonal move-
ments (Kudaktin and Chestin 1993; Schorger 1946). In
areas with high topographic relief, black bears may not
travel seasonally outside their summer range (Amstrup and
Beecham 1976). In places where they do, they generally do
not move as far as the bears in Minnesota (Garshelis and
Pelton 1980,2–18 km; Hellgren and Vaughan 1990,≈5 km;
Beck 1991,8–23 km). The flat Minnesota landscape offers
less diversity in microclimate and habitat over short
distances than more mountainous terrain, prompting longer
travels to find key resources. These extensive late-season
movements clearly fit within current paradigms for migra-
tion in being “straightened-out”travels that took place on a
consistent temporal schedule and shifted animals among
habitat zones (Dingle 1996). Migration was “partial,”in
that not all animals participated, and “facultative,”in that it
did not occur every year (Dingle and Drake 2007). In
contrast, early-season forays were not coordinated in
timing, direction, or destination, and may have been
exploratory in nature, possibly aimed at assessing or
enhancing breeding opportunities or as a precursor to
dispersal (Klenner 1987; Lee and Vaughan 2003; Schwartz
and Franzmann 1992). We focus discussion hereafter on
our three hypotheses regarding variability in late summer
migration patterns among individuals and year to year.
Hypothesis 1: Individuals leave their summer range in late
summer if the likelihood is high that they will find better
resources elsewhere Migration has generally been viewed
as an adaptive response to adversity (Dingle 1996). This
characterization fits the fall migration of temperate-nesting
songbirds that flee northern latitudes as winter approaches
and migrations of ungulates to wintering areas. Migrations,
however, also include the movements of animals to areas
with plentiful resources, such as the return of insects, birds
and whales to northern latitudes in the spring. Likewise,
some bear migrations have occurred in response to local
food shortages: prime examples include the exodus of bears
from Big Bend National Park during a year of extreme
drought (Hellgren et al. 2005) and several similar cases
elsewhere during extreme food failures (Pelton 1989;
Schorger 1946). However, the salient feature of most bear
migrations seems to be their orientation toward concen-
Fig. 3 Relationship between regional food abundance indices in
north-central Minnesota and the proportion of aadult and bjuvenile
black bears, by sex, that made late-season migrations in north-central
Minnesota, 1981–1990
Parameter Sex/age Year
1983
a
1988
b
Other years
Last date home (mean) F 1–2 years 11 Sep 2 Aug 21 Aug
F 3+years 20 Aug 3 Aug 16 Aug
M1–2 years 6 Aug 28 Jul 10 Aug
M 3+years 25 Aug 3 Aug 13 Aug
Duration in days (mean) F 1–2 years 19 57 31
F 3+years 27 51 36
M1–2 years 20 63 46
M 3+years 28 50 41
Distance in km (median) F 1–2 years 8 24 5
F 3+years 7 20 8
M1–2 years 27 33 23
M 3+years 8 22 18
Table 4 Departure dates, dura-
tion of travel, and distance
traveled for late-season bear
migrations in north-central
Minnesota during years with
disparate food conditions,
1981–1990
a
Highest chokecherry abundance
during the study
b
Highest oak abundance during
the study, particularly bur oak
Behav Ecol Sociobiol

trations of preferred foods (Garshelis and Pelton 1981;
Hellgren and Vaughan 1990; Schwartz and Franzmann
1991). Bears in this study migrated primarily south and
southwest, along an increasing food gradient (Noyce,
unpublished data). Statewide food surveys indicated that
oak trees, particularly bur and white oak (Quercus alba),
were more abundant in the hardwood forests that were
common south of our study area. Bears and many species
prefer these acorns to those of red oak, presumably due to
their lower tannin content (Kirkpatrick and Pekins 2002).
As large animals that feed on scattered, small food items,
bears must employ an “energy-maximizing”strategy
(Welch et al. 1997), focusing on food patches that provide
the highest return of calories. Moreover, fruits and nuts,
which are favored bear foods, are masting species, meaning
that they produce small-to-moderate crops most years, but
occasionally a massive over-abundance that is often
synchronized over large distances (Koenig and Knops
2000). Despite general synchrony, dispersion and local
density of these plants are highly variable. Mast abundance
within an animal's home range may thus be a signal that hot
spots may exist elsewhere, whereas mast failure at home
suggests there may be little to gain by leaving.
The chance of finding better foods elsewhere is also a
function of mobility and the types and quality of habitat
that an animal is likely to encounter (Sabine et al. 2002). In
our study, lowland female bears frequently moved south to
the nearby Suomi Hills to seek red oak acorns. The risks
entailed in this short move were minimal, for even if acorns
Fig. 4 Contrasting migration
patterns of female black bears
inhabiting the Chippewa Plains
lowlands (a,c) and those in the
adjacent Suomi Hills uplands (b,
d) in north-central Minnesota.
Each solid symbol denotes a
different bear on its summer
home range. Open symbols de-
pict destinations of each migra-
tion event, so bears that
migrated multiple years show
multiple destinations. Typically
(1981–1990, excluding 1988)
about 50% of lowland females
migrated, usually to the nearby
Suomi Hills (a), whereas Suomi
Hills residents rarely moved (b).
However, in 1988, nearly all
females left their home ranges
and bypassed the Suomi Hills to
travel southwest of the study
area to stands of bur oak and
agricultural fields (c,d)
Fig. 5 Cumulative survival of radio-collared black bears that
remained on their summer home range during the first 2 weeks of
Minnesota's bear hunting season (1–14 September) vs survival of
bears on seasonal migrations, 1981–1990
Behav Ecol Sociobiol

were scarce, the upland forest habitat was still likely to
provide better fall food than their lowland home ranges
(Noyce and Coy 1990). In years of food failure, however,
even these bears, apparently cueing on local scarcity, chose
not to move and instead subsisted on less preferred foods,
such as vegetation and insects (Garshelis and Noyce 2008).
Suomi Hills females, who lived in the best local habitat,
had less to gain from migrating, and finding better foods
would likely require traveling much longer distances.
However, in 1988, an exceptional crop of bur oak acorns,
which were uncommon locally but more available south
and west of the study area, was sufficient to entice these
bears and others from across the study area to forego a
reasonable local crop of red oak acorns and travel unusually
long distances to take advantage of the bounty.
Migrators that move in response to food scarcity are
often highly mobile specialists that must leave their home
area when key foods fail (Newton 2006; Fox et al 2009).
Black bears can move long distances, but not enough to
escape the winter dearth of foods. Their strategy is to eat as
much as possible, and store sufficient energy in the form of
fat to sustain them through an extended hibernation fast. To
maximize weight gain, bears are drawn to food abundance
and thus make long distance migrations when they have
reason to believe that rich food sources exist on the
landscape; they may also be forced to migrate in cases of
extreme food scarcity, although that was never the case on
our study site. How bears and other migrators make
decisions about whether, where and when to migrate in
search of foods that are beyond their normal range of
familiarity and, at least initially, beyond their sensory
detection, apparently relying on environmental and social
cues, remains an intriguing question (Kenney et al 2001).
Hypothesis 2: Individuals are most likely to travel in search
of better foods if their body mass is close to a reproductive
or survival threshold Because migration is risky and
potentially energetically costly, we presumed that animals
with the most to gain (e.g., earlier attainment of sexual
maturity) would also be most inclined to take such a risk.
Previously, Noyce and Garshelis (1994) identified three
significant body mass thresholds (measured in late winter)
for female black bears: (1) those weighing <41 kg never
produced cubs the following year; (2) below 65 kg, but not
above, maternal body mass was positively related to
fecundity and to growth and survival of cubs; (3) further
positive effects of body mass on reproduction were not
apparent until bears reached about 90 kg. Average late-
winter mass of female bears in the CNF was 35 kg at
2 years old and 46 kg at 3 years old (Noyce and Garshelis
1998). Few individuals reached 41 kg by age 2, but most
could attain it by age 3, with ample foods as 2-year-olds.
Our findings here that 2- and 3+-year-olds weighing 35–
50 kg were the most likely to migrate supports our
hypothesis, as those bears were close to the 41-kg weight
threshold. Moreover, particularly heavy yearlings could
reach 41 kg in a year, whereas smaller bears could not, so
the greater propensity of heavy yearlings to migrate also fits
this hypothesis. In lowland habitats, body mass of first-time
mothers averaged 58.4 kg (95% CI: 55.5–61.3) and
multiparous females 68.7 kg (95% CI: 64.9–72.5), hence,
close to the 65-kg threshold for increased litter size and
survival. Upland females averaged 65.5 kg at first birth
(95% CI: 62.6–68.5) and 90.6 kg (95% CI: 87.3–94.0) after
that (Garshelis and Noyce 2008). Accordingly, lowland
adult females migrated at twice the rate of upland females
and upland adult females migrated only when unusual bur
oak abundance put them within closer reach of the higher
(90 kg) weight threshold. Males should be less tied to
specific body mass thresholds. Instead, they should strive
for ever-greater mass throughout their life, given that only a
relative few of the largest males in an area have opportunity
to sire cubs (Kovach and Powell 2003; Costello et al.
2009). This may explain the somewhat greater propensity
for males in general to migrate.
Hypothesis 3: Travel behavior reflects decisions that
minimize risk Risks faced by animals that migrate include
failing to find sufficient food (to offset the cost of travel)
and encountering increased threats of mortality due to the
unfamiliar surroundings (Nicholson et al 1997). Our data
suggest that bears acted cautiously in both their choice to
migrate and their behavior during migration. Most mini-
mized the risk of not finding adequate nutrition by staying
home when regional food abundance was poor. As in other
species (Sæther and Andersen 1990; Kohlmann and
Risenhoover 1994), increased food abundance reduced this
risk, encouraged greater movement, and perhaps enabled
greater food selectivity. When chokecherry, a favored mid-
summer food, was exceptionally plentiful, as in 1983, bears
took advantage of this by delaying departure and shortening
the duration of their late-season migrations (also noted by
Garshelis and Pelton 1981). But it was primarily oaks,
which can create some of the highest caloric densities on
the landscape (Inman and Pelton 2002), that drew bears
away from home. These trees seem to have a keystone
effect on bears and other species across large areas of North
America (Pelton 1989; Vaughan 2002).
Potential risks during migration are, for most species,
greatest for juveniles. Juvenile bears are the most vulner-
able to predation and cannibalism (Garshelis 1994; Samson
and Huot 1998), undernutrition (Noyce and Garshelis
1994), and possibly spatial disorientation (Landriault et al.
2006). Accordingly, yearling females, the smallest bears,
were the least likely to migrate. Those that did migrate
traveled shorter distances than older bears and appeared to
Behav Ecol Sociobiol

cue their movements on different foods (especially
cherries). Small-bodied bears can achieve maximal weight
gains while feeding on small fruits, whereas adults require
more calorie-dense foods, like nuts (Welch et al. 1997).
Cherries were dispersed in small patches across habitat
types and fruit sometimes persisted late into the summer,
providing small bears with excellent forage without
necessitating extensive travel or exposure to larger animals
that may have congregated at richer feeding sites.
The low hunting mortality of bears that were outside
their home ranges during the hunting season suggested a
reticence to visit hunters' baits when bears were in
unfamiliar settings. Similarly, Brown and Alkon (1990)
found that porcupines (Erethizon dorsatum) adopted spe-
cific vigilance behaviors in particularly risky habitat
conditions, and thereby greatly reduced mortality risk.
Also, it has been observed that many species of migrating
animals tend to move directly toward a destination and
bypass food resources that normally would be of interest
(Dingle 1996; Dingle and Drake 2007). Rogers (1989)
found that radio-collared black bears traveling home from
forays followed straight-line trajectories, often at night,
forgoing trails and foraging. Studies that reported increased
mortality for bears that left their summer range regarded
dispersing males (Elowe and Dodge 1989; Schwartz and
Franzmann 1992), whose behavior appears to be quite
different than seasonal migrants.
From an evolutionary standpoint, animals would not
migrate if it were not adaptive, but that is not to say that
each individual, faced with the annual choice of whether to
do so, chooses correctly. The inherent conundrum of
measuring whether an individual is better off than it would
have been had it behaved differently is self-evident.
Comparisons of outcomes for conspecifics that did and
did not migrate are compromised by the unique and
complex suite of factors that influence each individual's
behavior each year. Longitudinal tracking of individuals
through multiple years presents similar problems, as age,
reproductive status, and habitat conditions change yearly. In
our study, opportunities for such comparisons were further
limited by the rapid turnover of individuals due to hunting.
In human-dominated landscapes, long-term evolutionary
advantages of a behavior like migration could be overrid-
den by threats introduced by humans. Our findings suggest
that vigilance behaviors evolved in bears during travel may
protect them from some manmade dangers (hunters, other
bears at hunters' baits) but perhaps not all (cars). Factors
including year-specific resource distribution and each
bear’s nutritional condition, physical stature, social domi-
nance, age, experience, and personality, all likely inform its
decision about whether to travel, and if so, when and where
to go. Recent theoretical and empirical evidence support the
idea that multiple behavioral strategies may be successful
within populations living where environmental conditions
vary from year to year (Dingemanse and Réale 2005;
Kaitala et al. 1993; Nicholson et al. 1997).
Crucial to successful seasonal migrations are high
mobility and well-developed navigational ability, character-
istics that are evident in bears (Landriault et al. 2006;
Rogers 1987b; Sauer et al. 1969). A remarkable physio-
logic lability enables bears to derive maximum benefit from
periodic nutritional bursts like masting, insect outbreaks,
and fish spawns: even young bears whose growth has been
severely curtailed in a year of food shortage can fully
rebound with abundant resources the following year (Noyce
and Garshelis unpublished data). We suggest that in bears,
high intelligence, in tandem with migratory flexibility,
complement these characteristics to create a population of
individuals poised to take maximum advantage of a highly
stochastic environment.
Acknowledgements This project was initiated and supported by the
Minnesota Department of Natural Resources as part of a long-term
research project on the population dynamics of black bears in
Minnesota. We are grateful for the assistance of many DNR biologists,
foresters, and other personnel, as well as the pilots, student interns and
volunteers who assisted with radio-telemetry, trapping and handling
bears, conducting food surveys, and maintaining records. In particular,
we thank P. Coy, P. Harris, J. Young, D. Clapp, B. Sampson, K.
Soring, M. Gallagher, and T. Lizotte. K. Kerr and G. Matson sectioned
teeth for age determination. We thank J. Fieberg for statistical advice
and E. Hellgren for helpful comments on the manuscript.
References
Alerstam T, Hedenström A, Åkesson S (2003) Long-distance
migration: evolution and determinants. Oikos 103:247–260
Amstrup SC, Beecham JJ (1976) Activity patterns of radio-collared
black bears in Idaho. J Wildl Manage 40:340–348
Arnold TW (2010) Uninformative parameters and model selection using
Akaike’s information criterion. J Wildl Manage 74:1175–1178
Beck, TDI (1991) Black bears of west-central Colorado. Colorado Div
Wildl Tech Pub No 39
Brönmark C, Skov C, Brodersen J, Nilsson PA, Hansson L (2008)
Seasonal migration determined by a trade-off between predator
avoidance and growth. PLoS ONE 3:e1957. doi:10.1371/journal.
pone.0001957
Brown JS, Alkon PU (1990) Testing values of crested porcupine
habitats by experimental food patches. Oecologia 83:512–518
Brown JS, Laundré JW, Gurung M (1999) The ecology of fear:
optimal foraging, game theory, and trophic interactions. J
Mammal 80:385–399
Costello C, Creel SR, Kalinowski ST, Vu NV, Quigley HB (2009)
Determinants of male reproductive success in American black
bears. Behav Ecol Sociobiol 64:125–134
Coy PC, Garshelis DG (1992) Reconstructing reproductive histories
of black bears from incremental layering in dental cementum.
Can J Zool 70:2150–2160
Dingemanse NJ, Réale D (2005) Natural selection and animal
personality. Behav 142:1165–1190
Behav Ecol Sociobiol

Dingle H (1996) Migration: the biology of life on the move. Oxford
University Press, Cambridge
Dingle H, Drake VA (2007) What is migration? BioSci 57:113–121
Elowe KD, Dodge WE (1989) Factors affecting black bear reproduc-
tive success and cub survival. J Wildl Manage 53:962–968
Fox AD, Kobro S, Lehikoinen A, Lyngs P, Vaisanen R (2009)
Northern bullfinch Pyrrhula p pyrrhula irruptive behavior linked
to rowanberry Sorbus aucuparia abundance. Ornis Fenn 86:51–
60
Garshelis DL (1994) Density-dependent population regulation of
black bears. In: Taylor M (ed) Density-dependent population
regulation of black, brown, and polar bears. Int Conf Bear Res
Manage Monog Ser 3. pp 3–14
Garshelis DL, Noyce KV (2008) Seeing the world through the nose of
a bear—diversity of foods fosters behavioral and demographic
stability. In: Fulbright TE, Hewitt DG (eds) Wildlife science:
linking ecological theory and management applications. CRC
Press, Boca Raton
Garshelis DL, Pelton MR (1980) Activity of black bears in the Great
Smoky Mountains National Park. J Mammal 61:8–19
Garshelis DL, Pelton MR (1981) Movements of black bears in the
Great Smoky Mountains National Park. J Wildl Manage 45:912–
925
Hashimoto Y, Yasutake A (1999) Seasonal changes in body weight
of female Asiatic black bears under captivity. Mamm Study
24:1–6
Hellgren EC, Vaughan MR (1990) Range dynamics of black bears in
Great Dismal Swamp, Virginia-North Carolina. Proc Annual
Conf SE Assoc Fish Wildl Agencies 44:268–278
Hellgren EC, Onorato DP, Skiles JR (2005) Dynamics of a black bear
population within a desert metapopulation. Biol Conserv
122:131–140
Hilderbrand GV, Jenkins SG, Schwartz CC, Hanley TA, Robbins CT
(1999) Effect of seasonal differences in dietary meat intake on
changes in body mass and composition in wild and captive
brown bears. Can J Zool 77:1623–1630
Inman RM, Pelton MR (2002) Energetic production by soft and hard
mast foods of American black bears in the Smoky Mountains.
Ursus 13:57–68
John AWG, Roskell J (1985) Jay movements in autumn 1983. Br
Birds 78:611–637
Kaitala A, Kaitala V, Lundberg P (1993) A theory of partial migration.
Am Nat 142:59–81
Kasbohm JW, Vaughan MR, Kraus JG (1998) Black bear home range
dynamics and movement patterns during a gypsy moth infesta-
tion. Ursus 10:259–267
Kenney RD, Mayo CA, Winn HE (2001) Migration and foraging
strategies at varying spatial scales in western North Atlantic right
whales: a review of hypotheses. J Cetacean Res Manage spec
issue 2:251–260
Kirkpatrick RL, Pekins PJ (2002) Nutritional value of acorns for
wildlife. In: McShea WJ, Healy WM (eds) Oak forest ecosys-
tems: ecology and management for wildlife. Johns Hopkins
University Press, Baltimore, pp 173–181
Klenner W (1987) Seasonal movements and home range utilization
patterns of the Black Bear, Ursus americanus, in western
Manitoba. Can Field Nat 101:558–568
Koenig WD, Knops JMH (2000) Patterns of annual seed production
by northern hemisphere trees: a global perspective. Am Nat
155:59–69
Kohlmann SG, Risenhoover KL (1994) Spatial and behavioural
response of white-tailed deer to forage depletion. Can J Zool
72:506–513
Kovach AI, Powell RA (2003) Effects of body size on male mating
tactics and paternity in black bears, Ursus americanus. Can J
Zool 81:1257–1268
Kudaktin AN, Chestin IE (1993) The Caucasus. In: Vaisfeld MA,
Chestin IE (eds) Game animals of Russia and adjacent countries
and their environment: bears. Nauka, Moscow, pp 136–169
Landriault J, Hall MN, Hamr J, Mallory FF (2006) Long-range
homing by an adult female black bear, Ursus americanus. Can
Field Nat 120:57–60
Lariviere S, Huot J, Samson C (1994) Daily activity patterns of female
black bears in a northern mixed-forest environment. J Mammal
75:613–620
Lee DJ, Vaughan MR (2003) Dispersal movements by subadult
American black bears in Virginia. Ursus 14:162–170
McShea WJ, Schwede G (1993) Variable acorn crops: responses of white-
tailed deer and other mast consumers. J Mammal 74:999–1006
Minnesota Department of Natural Resources (2003) Field guide to the
native plant communities of Minnesota: the Laurentian mixed
forest province. Ecological land classification program, Minne-
sota County Biological survey and Natural Heritage and
Nongame Research Program. Minnesota Dept Nat Resources
St. Paul
Newton I (2006) Advances in the study of irruptive migration. Ardea
94:433–460
Nicholson MC, Bowyer RT, Kie JG (1997) Habitat selection and
survival of mule deer: tradeoffs associated with migration. J
Mammal 789:483–504
Noyce KV, Coy PL (1990) Abundance and productivity of bear food
species in different forest types of northcentral Minnesota. Int
Conf Bear Res Manage 8:169–181
Noyce KV, Garshelis DL (1994) Body size and blood characteristics
as indicators of condition and reproductive performance in black
bears. Int Conf Bear Res Manage 9:481–496
Noyce KV, Garshelis DL (1997) Influence of natural food abundance on
black bear harvests in Minnesota. J Wildl Manage 61:1067–1074
Noyce KV, Garshelis DL (1998) Spring weight changes in black bears
in northcentral Minnesota: the negative foraging period revisited.
Ursus 10:521–531
Olsson IC, Greenberg LA, Bergman E, Wysujack K (2006)
Environmentally induced migration: the importance of food.
Ecol Lett 9:645–651
Pelton MR (1989) The impacts of oak mast on black bears in the
southern Appalachians. In: McGee CH (ed) Proceedings of the
workshop on Southern Appalachian mast management. Univer-
sity of Tennessee, Knoxville, pp 7–11
Rogers LL (1987a) Effects of food supply and kinship on social
behavior, movements, and population growth of black bears in
northeastern Minnesota. Wildl Monog 97
Rogers LL (1987b) Navigation by adult black bears. J Mammal
68:185–188
Rogers LL (1989) Home, sweet-smelling home. Nat Hist 9(89):61–
67
Sabine DL, Morrison SF, Whitlaw HA, Ballard WB, Forbes GJ
(2002) Migration behavior of white-tailed deer under varying
winter climate regimes in New Brunswick. J Wildl Manage
66:718–728
Sæther BE, Andersen R (1990) Resource limitation in a generalist
herbivore, the moose Alces alces: ecological constraints on
behavioural decisions. Can J Zool 68:993–999
Samson C, Huot J (1998) Movements of female black bears in relation
to landscape vegetation type in southern Quebec. J Wildl Manage
62:718–727
Sauer PR, Free SL, Browne SD (1969) Movement of tagged black
bears in the Adirondacks. NY Fish Game J 16:205–223
Schooley RL, McLaughlin CR, Krohn WB, Matula GJ (1994)
Spatiotemporal patterns of macrohabitat use by female black
bears during fall. Int Conf Bear Res Manage 9:339–348
Schorger AW (1946) Influx of bears into St. Louis County, Minnesota.
J Mammal 27:177
Behav Ecol Sociobiol

Schwartz CC, Franzmann AW (1991) Interrelationship of black bears
to moose and forest succession in the northern coniferous forest.
Wildl Monog 113
Schwartz CC, Franzmann AW (1992) Dispersal and survival of subadult
black bears from the Kenai Peninsula, Alaska. J Wildl Manage
56:426–431
Symonds MRE, Moussalli A (2010) A brief guide to model selection,
multimodal inference, and model averaging in behavioural ecology
using Akaike’s information criterion. Behav Ecol Sociobiol.
doi:10.1007/s00265-010-1037-6
Vaughan M (2002) Oak trees, acorns, and bears. In: McShea WJ,
Healy WM (eds) Oak forest ecosystems: ecology and manage-
ment for wildlife. The Johns Hopkins University Press, Balti-
more, pp 224–240
Welch CA, Keay J, Kendall KC, Robbins CT (1997) Constraints on
frugivory by bears. Ecology 78:1105–1119
White PJ, Davis TL, Barnowe-Meyer KK, Crabtree RL, Garrott RA
(2007) Partial migration and philopatry of Yellowstone pronghorn.
Biol Conserv 135:502–510
Whitehead CJ (1969) Oak mast yields on wildlife management
areas in Tennessee. Tennessee Game and Fish Commission,
Nashville
Willey CH (1974) Aging black bears from first premolar tooth
sections. J Wildl Manage 38:97–100
Behav Ecol Sociobiol