-
https://cpw.cvlcollections.org/files/original/b54e2247d1a5cc894831c03d4978e6ee.pdf
962c2d1066f6d1f06804102cc7e67f40
PDF Text
Text
Dan Prenzlow, Director, Colorado Parks and Wildlife • Parks and Wildlife Commission: Marvin McDaniel, Chair • Carrie Besnette Hauser, Vice-Chair
Marie Haskett, Secretary • Taishya Adams • Betsy Blecha • Charles Garcia • Dallas May • Duke Phillips, IV • Luke B. Schafer • James Jay Tutchton • Eden Vardy
�Received: 4 March 2017
|
Accepted: 25 January 2018
DOI: 10.1111/1365-2656.12810
RESEARCH ARTICLE
Spatiotemporal heterogeneity in prey abundance and
vulnerability shapes the foraging tactics of an omnivore
Nathaniel D. Rayl1
| Guillaume Bastille-Rousseau2
Matthew A. Mumma4
| John F. Organ1,3 |
| Shane P. Mahoney5 | Colleen E. Soulliere5 |
Keith P. Lewis5 | Robert D. Otto5,6 | Dennis L. Murray2 | Lisette P. Waits4 |
Todd K. Fuller1
1
Department of Environmental Conservation, University of Massachusetts, Amherst, MA, USA; 2Environmental and Life Sciences Graduate Program, Trent
University, Peterborough, Ontario, Canada; 3U.S. Geological Survey, Cooperative Fish and Wildlife Research Units, Reston, VA, USA; 4Department of Fish
and Wildlife Sciences, College of Natural Resources, University of Idaho, Moscow, ID, USA; 5Department of Environment and Conservation, Government of
Newfoundland and Labrador, St. John’s, Newfoundland and Labrador, Canada and 6Department of Environment and Conservation, Institute for Biodiversity,
Ecosystem Science, and Sustainability, Government of Newfoundland and Labrador, Corner Brook, Newfoundland and Labrador, Canada
Correspondence
Nathaniel D. Rayl
Email: nathanielrayl@gmail.com
Present addresses
Nathaniel D. Rayl, U.S. Geological
Survey, Northern Rocky Mountain Science
Center, Bozeman, MT 59715, USA
Abstract
1. Prey abundance and prey vulnerability vary across space and time, but we know
little about how they mediate predator–prey interactions and predator foraging
tactics. To evaluate the interplay between prey abundance, prey vulnerability and
predator space use, we examined patterns of black bear (Ursus americanus) preda-
Guillaume Bastille-Rousseau, Department
of Fish, Wildlife and Conservation
Biology, Colorado State University, Fort
Collins, CO 80523, USA
tion of caribou (Rangifer tarandus) neonates in Newfoundland, Canada using data
Matthew A. Mumma, Ecosystem Science
& Management Program, University of
Northern British Columbia, Prince George,
British Columbia V2N 4Z9, Canada
2. During the caribou calving season, we predicted that landscape features would
Shane P. Mahoney, Conservation Visions
Inc., St. John’s, Newfoundland and Labrador
A1C 5W4, Canada
Keith P. Lewis, Fisheries and Oceans Canada,
St. John’s, Newfoundland and Labrador A1C
5X1, Canada
Robert D. Otto, Atlantic Salmon Federation,
St. Andrews, New Brunswick E5B 3S8,
Canada
Funding information
Department of Environment and
Conservation, Government of
Newfoundland and Labrador; Safari Club
International Foundation; University of
Massachusetts, Amherst
Handling Editor: Anne Loison
from 317 collared individuals (9 bears, 34 adult female caribou, 274 caribou
calves).
influence calf vulnerability to bear predation, and that bears would actively hunt
calves by selecting areas associated with increased calf vulnerability. Further, we
hypothesized that bears would dynamically adjust their foraging tactics in response to spatiotemporal changes in calf abundance and vulnerability (collectively, calf availability). Accordingly, we expected bears to actively hunt calves
when they were most abundant and vulnerable, but switch to foraging on other
resources as calf availability declined.
3. As predicted, landscape heterogeneity influenced risk of mortality, and bears displayed the strongest selection for areas where they were most likely to kill calves,
which suggested they were actively hunting caribou. Initially, the per-capita rate
at which bears killed calves followed a type-I functional response, but as the calving season progressed and calf vulnerability declined, kill rates dissociated from
calf abundance. In support of our hypothesis, bears adjusted their foraging tactics
when they were less efficient at catching calves, highlighting the influence that
predation phenology may have on predator space use. Contrary to our expectations, however, bears appeared to continue to hunt caribou as calf availability
874
|
© 2018 The Authors. Journal of Animal Ecology
© 2018 British Ecological Society
wileyonlinelibrary.com/journal/jane�
J Anim Ecol. 2018;87:874–887.
�Journal of Animal Ecology
RAYL et al.
|
875
declined, but switched from a tactic of selecting areas of increased calf vulnerability to a tactic that maximized encounter rates with calves.
4. Our results reveal that generalist predators can dynamically adjust their foraging
tactics over short time-scales in response to changing prey abundance and vulnerability. Further, they demonstrate the utility of integrating temporal dynamics of
prey availability into investigations of predator–prey interactions, and move towards a mechanistic understanding of the dynamic foraging tactics of a large
omnivore.
KEYWORDS
Black bear (Ursus americanus), caribou (Rangifer tarandus) calves, cause-specific survival
analysis, foraging tactics, kill rates, predation risk, trophic interaction, ungulate
1 | I NTRO D U C TI O N
how predators might adjust their space-use patterns in response to
this spatiotemporal heterogeneity.
Predator–prey interactions frequently structure ecological commu-
When evaluating how predators may alter their foraging patterns
nities (Holt, 1977; Schmitz, 1998), alter population dynamics (Krebs
in response to spatiotemporal variation in prey abundance and prey
et al., 1995) and cause significant evolutionary change (Young,
vulnerability, there are two primary hunting modes of predators to
Brodie, & Brodie, 2004). To develop a comprehensive understanding
consider (Schmitz, 2008). Some predators, typically cursorial spe-
of these interactions, and thus of ecosystem function, it is neces-
cies, may attempt to maximize their encounters with prey by se-
sary to examine spatiotemporal dynamics of predation (Lima & Dill,
lecting areas with the highest probability of encountering their prey
1990). Significant work has been done to quantify the spatial dy-
(Murray, Boutin, & O’Donoghue, 1994). Other predators, typically
namics of predation from the perspective of prey species (e.g. model
stalking species, may try to maximize their success rate by selecting
predation risk, Hebblewhite, Merrill, & McDonald, 2005; Schmitz,
areas where they are more likely to kill their prey (Hopcraft et al.,
1998, 2008), and increasingly, to consider the temporal dynamics
2005). Predators may also use some combination of these two for-
of predation risk (Basille et al., 2015; Latombe, Fortin, & Parrott,
aging tactics to hunt their prey (Fuller, Harrison, & Vashon, 2007).
2014). Yet, few studies have examined the spatiotemporal context of
The extent to which predators exhibit facultative switching between
predator–prey interactions from the perspective of predators (Lima,
these tactics remains largely unexplored.
2002), especially in large vertebrate communities (but see Bastille-
In many systems across North America, black (Ursus americanus)
Rousseau, Fortin, Dussault, Courtois, & Ouellet, 2011; Hopcraft,
and brown bears (Ursus arctos) are significant predators of ungulate
Sinclair, & Packer, 2005).
neonates (Zager & Beecham, 2006). Bear predation of ungulate
In classic models of predator–prey dynamics, the predator’s
calves typically has a unique temporal signature, with early and in-
perspective of predation has been characterized by the per-capita
tense predation that quickly tapers off as calves grow large enough
rate at which predators kill prey, which has been modelled as a func-
to outrun bears (Zager & Beecham, 2006). As opportunistic omni-
tion of prey density (the functional response; Holling, 1959, 1965;
vores, bears frequently forage across multiple trophic levels, making
Solomon, 1949). Changing vulnerability to predation, however, may
trade-offs between foraging on abundant and predictable, but lower
also greatly influence the rate at which predators kill prey (Tilman,
quality vegetation, and rarer and less predictable, but high-quality
1978). Thus, from a predator’s perspective, the availability of prey is
animal matter (Orians et al., 1997). Thus, bear predation of ungulate
a function of prey abundance and prey vulnerability, both of which
calves may result from bears actively searching for calves, or from
vary in space and time. Spatiotemporal heterogeneity in the abun-
bears incidentally encountering calves while foraging for other re-
dance of prey may emerge because of variation in the timing and
sources (Bastille-Rousseau et al., 2011). The temporal dynamics and
location of births (Rayl et al., 2014), the distribution of resources
consumer characteristics of bear-ungulate calf interactions offer a
(Clutton-Brock & Harvey, 1978) or climatic conditions (Coulson,
unique opportunity to investigate a predator’s perspective of preda-
Milner-Gulland, & Clutton-Brock, 2000). Spatial and temporal pat-
tor–prey interactions and to evaluate the influence of prey availabil-
terning in the vulnerability of prey may arise because of variation in
ity on predator foraging tactics.
landscape features (Kauffman et al., 2007), prey age (Adams, Singer,
In the last two decades, caribou (Rangifer tarandus) in
& Dale, 1995), prey body condition (Wirsing, Steury, & Murray, 2002)
Newfoundland have declined from c. 94,000 to c. 32,000 individuals,
or predator–prey encounter rates (Holling, 1959). To date, there is
with calf predation by black bears and coyotes (Canis latrans) identi-
limited basis from either theoretical or empirical studies to predict
fied as a major proximate mechanism of the decline (Bastille-Rousseau
�876
|
Journal of Animal Ecology
RAYL et al.
et al., 2016; Mahoney et al., 2016). In this system, bears are the most
characterize the landcover of the herd’s range, we used a series
significant predator of caribou calves. Characteristically, bear preda-
of Landsat 7 scenes (30-m resolution) that were initially classified
tion of calves on the island is temporally constrained, with most pre-
into nine landcover types that we collapsed into six types: (1) co-
dation occurring during June (Mahoney et al., 2016) when new-born
nifer forest, which subsumed two rare landcover types (decidu-
calves are aggregated on calving grounds (Rayl et al., 2014). Here, we
ous forest, mixed forest), (2) conifer scrub (stunted conifer forest),
examine the spatiotemporal dynamics of this predation, focusing pri-
(3) wetlands (fens and bogs), (4) barrens (barrens, rocky and other
marily on the bear’s perspective of these predator–prey interactions.
open habitats), (5) heathland (lichen and heathland habitats), and
First, we evaluated whether landscape heterogeneity mediated
(6) water (lakes, ponds, rivers and streams). We did not consider
mortality risk of calves to predation by bears. Elsewhere, land-
anthropogenic landcover, an extremely rare landcover type, in our
scape features influence patterns of predation on adult ungulates
analyses. We calculated topographic variables from a digital eleva-
(Hebblewhite et al., 2005; Hopcraft et al., 2005; Kauffman et al.,
tion model (25-m resolution).
2007). We hypothesized (H1) they would similarly influence predation risk for juvenile ungulates. As an ambush predator, we predicted that bears would be most successful at killing calves in areas
2.2 | Animal collaring and monitoring
with greater stalking cover (Schmitz, 2008). Second, we investigated
We captured black bears >2 years of age by darting them from a
whether calf predation resulted from bears actively searching for
helicopter during May to October 2008–2012 and fitted them with
calves or from bears incidentally encountering them while foraging
Global Positioning System (GPS) collars programmed to acquire a lo-
for other resources. Although prior studies examining black bear-
cation every 1 or 2 hr (Advanced Telemetry Systems [ATS], Isanti,
caribou calf interactions concluded that most bears were not ac-
MN, USA; Lotek Wireless Inc., New Market, ON, Canada). We cap-
tively hunting calves (Bastille-Rousseau et al., 2011; Latham, Latham,
tured adult female caribou in the same way during November to May
& Boyce, 2011), we hypothesized (H2) that bears in Newfoundland
2006–2012 and fitted them with GPS collars programmed to acquire
would actively search for calves because of the high density of calves
a location every 2 hr (Lotek Wireless Inc.). We located caribou neo-
in this system (e.g., density of caribou calves in Bastille-Rousseau
nates <5 days old from helicopters, captured them on foot during
et al., 2011 = 0.47 calves/100 km2 vs. 546 calves/100 km2 in our
late May to early June 2003–2013, and fitted them with very high
system; N.D. Rayl, unpubl. data). Further, we expected the hunting
frequency collars with motion-sensitive transmitters (ATS; Lotek
mode of bears to mirror that of other ambush predators (Hopcraft
Wireless Inc.; Sirtrack, Havelock North, New Zealand; Telemetry
et al., 2005); therefore, we predicted bears would select habitat
Solutions, Concord, California). We monitored calves daily during
features associated with increased vulnerability of calves that maxi-
the first week post-capture, every 2–4 days thereafter until July,
mized their predation success. Third, we evaluated whether changes
every 5–10 days during July, and thenceforth every 2–4 weeks.
in the availability of caribou calves influenced the foraging tactics
When we detected a mortality signal, we conducted a systematic
of black bears. We hypothesized (H3) that bears, as opportunistic
field investigation to determine the cause of death. Additionally, we
omnivores, would alter their foraging patterns in response to chang-
verified many field assessments using laboratory necropsy of col-
ing abundance and vulnerability (collectively, calf availability) of calf
lected remains or DNA analysis to identify the cause of death (fur-
prey. Accordingly, we predicted that bears would actively hunt car-
ther details in Mahoney et al., 2016; Mumma, Soulliere, Mahoney, &
ibou calves when they were most available, but switch to foraging
Waits, 2014).
on other resources as calf availability declined. Our novel approach
Previously, we determined that most female-calf caribou pairs
reveals new insights about the complex decisions predators make
migrated from Middle Ridge North by 30 June (Rayl et al., 2014),
when searching for prey and moves towards a mechanistic under-
and that some black bears visited the calving grounds when calves
standing of the dynamic foraging tactics of a generalist predator in
were present, whereas others did not (Rayl et al., 2015). Here, we
a system where it is strongly influencing prey population dynamics.
classified 27 May, the earliest recorded calf capture, to 30 June as
the calving season. We included caribou that calved in Middle Ridge
North and bears that visited the calving grounds during the calving
2 | M ATE R I A L S A N D M E TH O DS
season in our analyses. From these datasets, we removed individual-
years with <15 days of GPS data in a calving season. Additionally, we
2.1 | Study area
removed portions of two adult female caribou datasets that were
2
The island of Newfoundland, Canada (108,860 km ), is a mixture
located in an area south of Middle Ridge North that lacked Landsat
of bogs, heaths, barrens, and coniferous and mixed forests. We
coverage. Our final calving season dataset included nine black bears
studied caribou and black bears in the range of the largest caribou
monitored from 2008 to 2013 (3 F, 6 M, 17 bear-years, 5,648 loca-
herd in Newfoundland, the Middle Ridge herd (c. 10,000 individu-
tions), 34 adult female caribou monitored from 2009 to 2013 (91
als). This herd’s range was almost entirely roadless, with human
caribou-years, 47,088 locations) and 274 caribou calves monitored
settlements confined to the coast. Over 95% of females from
from 2003 to 2013 (119 F, 150 M, 5 unknown; Figure 1). Bear and
this herd calved in a calving ground called Middle Ridge North
caribou capture and handling procedures conformed to guide-
(867 km2; Fifield, Lewis, & Gullage, 2013; Rayl et al., 2014). To
lines established by the American Society of Mammalogists (Sikes,
�Journal of Animal Ecology
RAYL et al.
|
877
F I G U R E 1 Calving grounds of
the Middle Ridge caribou herd, with
calving season locations of kill sites
of caribou calves killed by black bears
(n = 61), Global Positioning System (GPS)
locations of black bears (n = 9), and GPS
locations of adult female caribou (n = 34),
Newfoundland, Canada, 2003–2013
Gannon, & A.S. of Mammalogists, 2011), and bear captures were
vegetation abundance of graminoids and forbs as potential food
approved by the University of Massachusetts Amherst Institutional
items for black bears in the spring (Boileau, Crête, & Huot, 1994;
Animal Care and Use Committee (Protocol 2009-0 047).
Zieminski, 2016). This model combined a spatial model based on the
average vegetation biomass in each landcover type with a tempo-
2.3 | Evaluating the foraging tactics of bears
Black bears may prey on caribou calves when they are actively
ral model of vegetation growth based on the normalized-difference
vegetation index (NDVI) and field vegetation surveys of 173 plots
during 2011–2012 (see Appendix S1).
searching for calves or when they incidentally encounter them
while foraging for other resources. Bears may search for calves by
selecting areas where they are most successful at killing calves, by
2.5 | Occurrence of ant colonies
selecting areas where they are most likely to encounter calves, or by
We estimated the occurrence of ant colonies by systematically
using some combination of these two foraging tactics. The primary
investigating all stumps, woody debris, rocks and soil mounds
food resources for bears during the calving season in the range of
within a 10-m radius of the centre of 140 of our vegetation plots
the Middle Ridge herd, as revealed by scat analyses, were vegeta-
in 2012 and recording the number of ant colonies we found
tion, ants (family Formicidae), and caribou (in order from most to
(Noyce, Kannowski, & Riggs, 1997). We then created an index of
least common; Zieminski, 2016). Therefore, to evaluate the foraging
the relative abundance of ant colonies per landcover type by di-
tactics of bears that led to calf predation, we estimated the occur-
viding the average number of ant colonies in each landcover type
rence of vegetation, ant colonies, caribou calves and kill sites (we
by the maximum average ant colony value for the five terrestrial
assumed that kill sites were indicative of areas where bears were
landcover types. We assigned water landcover an ant colony index
most successful at killing calves, but they may also represent areas of
value of 0.
greater calf detectability or density). We then assessed the connection between these occurrence variables and the resource selection
patterns of bears to determine their foraging tactics, and evaluated
2.6 | Occurrence of caribou calves
whether or not changes in the availability of calves influenced these
We used the occurrence of adult female caribou as a surrogate
foraging tactics (Figure 2).
for the occurrence of caribou calves because we did not collect
accurate or precise location data for calves prior to their death.
2.4 | Occurrence of vegetation
Because caribou calves are followers (Lent, 1974), and female-
calf herds in Newfoundland were highly aggregated during the
We estimated the occurrence of vegetation by developing a spa-
calving season (Bergerud, 1974; Rayl et al., 2014), GPS locations
tiotemporal model of vegetation biomass for black bears based
of female caribou served as an appropriate proxy for calf loca-
on an approach used for ungulates (Bastille-Rousseau et al., 2015;
tions. We developed a resource selection function (RSF; Manly,
Hebblewhite, Merrill, & McDermid, 2008). We considered the green
McDonald, Thomas, McDonald, & Erickson, 2002) to estimate
�878
|
Journal of Animal Ecology
RAYL et al.
F I G U R E 2 Flow chart decomposing the analysis of black bear foraging tactics into its constituent parts. Where applicable, parenthetical
contents refer to relevant hypotheses (H1-H3) and equations (eqn) from the text and appendices (App.)
the relative probability of occurrence of caribou calves using a
GLMM with a binomial distribution, logit link and individual-year
2.7 | Occurrence of kill sites
as the random intercept (Gillies et al., 2006). We compared the
We developed an RSF to estimate the relative probability of oc-
landscape attributes of used locations of adult female caribou
currence of kill sites using a GLM with a binomial distribution and
(coded as 1) with an equal number of randomly sampled available
logit link (Hebblewhite et al., 2005). This model allowed us to test
locations (coded as 0) drawn from within the 100% minimum con-
our hypothesis (H1) that landscape heterogeneity mediated vulner-
vex polygons (MCP) of each caribou in each calving season. We
ability of caribou calves to predation by black bears. As above, GPS
considered slope, aspect (included using dummy variable coding
locations of female caribou served as an appropriate proxy for calf
with five categories: north [315–359 and 0–44; reference cat-
locations. Therefore, we compared the landscape attributes of black
egory], south [135–224], east [45–134], west [225–314], and flat
bear kill sites (coded as 1) of collared (n = 57) and uncollared calves
[slope = 0.0]), elevation, landcover type (included using dummy
we opportunistically encountered (n = 4) with subsampled locations
variable coding with “wetlands” as the reference category), and
of adult female caribou (n = 34; coded as 0) during the calving sea-
two-w ay interactions between each landcover type (except
son. We used this approach because we did not collar the calves of
“water”) and the proportion of that landcover type within a 5-k m
outfitted females, so we did not have paired data for caribou calves
radius (to account for a functional response in habitat selection;
and adult females. To maintain equal sampling among individuals, we
Moreau, Fortin, Couturier, & Duchesne, 2012; Mysterud & Ims,
randomly subsampled the telemetry locations of adult female cari-
1998) as potential explanatory variables. Prior to building candi-
bou by sampling an identical number of locations equal to the total
date models, we conducted univariate logistic regression analy-
number of locations in the smallest individual dataset (408 locations).
sis, and excluded explanatory variables where p > .25 on a Wald
We considered slope, elevation, local-scale landcover and landscape-
statistic from our candidate models (Hosmer & Lemeshow, 2000).
scale landcover as potential explanatory variables. Because calves
We then developed a set of candidate RSFs (see Appendix S2:
were likely killed after a chase and carcasses may have been moved
by predators, we characterized the local-scale landcover at three dif-
Table S1), which took the form:
(
w(x) = exp β1 x1 + ⋯ + βu xuij + βu5k x(u5k)ij + ⋯ + βu xu × βu5k x(u5k)ij + γ0j
)
(1)
ferent potentially explanatory scales. We calculated the proportion of
each landcover type within a 100-, 250-, and 500-m radius buffer of
where w(x) represented the RSF scores, βu was the selection coef-
the kill site. We characterized the landscape-scale landcover as the
ficient for explanatory variable xu for the ith observation and the jth
proportion of each landcover type within a 5-km radius buffer of the
individual-year, βu5k was the selection coefficient for the proportion
kill site. Prior to building candidate models, we conducted univari-
of the landcover type x(u5k) within a 5-km buffer, and γ0j was the ran-
ate logistic regression analysis, and excluded explanatory variables
dom intercept for the jth individual-year.
where p > .25 on a Wald statistic from our candidate models (Hosmer
�Journal of Animal Ecology
RAYL et al.
& Lemeshow, 2000). Our candidate RSFs (see Appendix S2: Table S2)
took the form:
|
879
on ordinal day i, and H was the CIF value for day t − (i − 1) or t − i and
cause=j. We calculated Bi by dividing previous estimates of the pro-
w(x) = exp (β1 x1 + β2 x2 + ⋯ + βu xu )
(2)
where w(x) represented the RSF scores and βu was the selection coefficient for explanatory variable xu.
portion of calves born during 2-day periods in half (see Appendix S2:
Table S3; Bergerud, 1975). Although our source for the timing of
calving was dated, anecdotal evidence (timing of observations of
first-born and last-born calves of the year, field estimates of the peak
of calving) from 2003 to 2013 suggested no appreciable change in
2.8 | Availability of caribou calves (predation
phenology)
timing. We calculated the daily cumulative number of calves killed
by each cause (Qt,j) as:
⎧∑
n
⎪ (C × Bi × Ht−(i−1),j ) n ≤ t
⎪ i=1
Qt,j = ⎨ t
⎪ ∑ (C × B × H
i
t−(i−1),j ) n > t
⎪ i=1
⎩
From a black bear’s perspective, the availability of caribou calves is
a function of both the abundance of calves and their vulnerability to
bear predation. We developed an approach to estimate the average
daily number of calves killed by bears. This integrated metric of predation phenology incorporated changes in the abundance and vulnerability of caribou calves throughout the calving season and allowed
(5)
and the daily number of calves that remained alive (At) as:
us to investigate whether changes in this metric affected the foraging
�
�
n
3
⎧∑
∑
⎪
C × Bi − Qt,j
n≤t
⎪ i=1
j=1
�
At = ⎨ �
t
3
⎪ ∑ C×B − ∑ Q
n>t
i
t,j
⎪
j=1
⎩ i=1
tactics of bears. First, we estimated cause-specific mortality rates for
collared caribou calves in a competing risks framework using a nonparametric cumulative incidence function (CIF; Heisey & Patterson,
2006), with mortality causes classified as black bear, coyote or other.
(6)
We used a right-censored design with time-at-risk (days) based on
time since capture (Fieberg & Delgiudice, 2009). We conservatively
assumed that survival timelines corresponded with age because we
We estimated the daily per-capita rate of caribou calf predation
by black bears (Rt) during the calving season as:
captured most calves when they were ≤2.5 days old, but we recognize
that this assumption likely biased our estimates of neonatal survival
Rt =
upward (Gilbert, Lindberg, Hundertmark, & Person, 2014). We censored all calves at 180 days or prior to that if their radio transmitter
×D
(7)
where Kt,b was the daily number of bear kills for ordinal day t
(Equation 4), Z was the area (km2) of a 100% MCP encompassing the
was lost or detached prematurely.
Next, we estimated the average number of caribou calves born in
locations of kill sites during the calving season, and D was the estimated average density (bears/100 km2) of bears in Middle Ridge
Middle Ridge North (C) from 2003 to 2013 as:
North from 2009 to 2012 (4.85 bears/100 km2; see Appendix S3).
∑2013
C=
Kt,b
Z
100
y=2003 Ny × M × Fy × Py
(3)
2013 − 2003
We estimated the per-capita rate of caribou calf predation by black
bears (V181,b) to 30 June as:
where y referred to the years 2003 through 2013, Ny was the population estimate of the herd, M was the estimated proportion of the
V181,b =
herd that calved in Middle Ridge North (0.95; Fifield et al., 2013), Fy
was the estimated proportion of adult females in the herd, and P y
Q181,b
Z
100
×D
(8)
where Q181,b was the cumulative number of calves killed by bears on
was the estimated productivity of the herd. We used the Petersen
the last day of the calving season (30 June; Equation 5), Z was the
method to estimate Ny from single session mark-resight aerial sur-
area (km2) of a 100% MCP encompassing the locations of kill sites
veys (see Mahoney, Virgl, Fong, Maccharles, & McGrath, 1998 for
during the calving season, and D was the estimated average density
details). We used aerial classification surveys flown in November-
(4.85 bears/100 km2) of bears in Middle Ridge North from 2009 to
April to estimate Fy and in June to estimate P y (Mahoney & Schaefer,
2012.
2002). We then calculated the daily number of cause-specific kills
(Kt,j) for ordinal day t and cause j as:
⎧∑
n
⎪ (C × Bi × (Ht−(i−1),j − Ht−i,j ))
⎪
Kt,j = ⎨ i=1
t
⎪ ∑ (C × B × (H
i
t−(i−1),j − Ht−i,j ))
⎪ i=1
⎩
2.9 | Foraging tactics of black bears
To evaluate the foraging tactics of black bears, we developed an
n≤t
(4)
n>t
RSF estimating the relative probability of occurrence of bears using
a GLMM with a binomial distribution, logit link, and individual-year
as the random intercept (Gillies et al., 2006). We compared the re-
where i referred to ordinal days 1 through n for the dates of calving,
source attributes of used locations of bears (coded as 1) with an
C was the average number of caribou calves born in Middle Ridge
equal number of randomly sampled available locations (coded as
North (Equation 3), Bi was the estimated proportion of calves born
0) drawn from within the 100% MCP of each bear in each calving
�880
|
Journal of Animal Ecology
RAYL et al.
TA B L E 1 Ranked list of candidate
models for resource selection functions
estimating the relative probability of
selection of black bears, Newfoundland,
Canada, 2008–2013. Numbers of
parameters (K), second-order Akaike
information criteria (AICc), differences in
AICc (∆AICc) and AICc weights (ω) are
presented
Model
K
AICc
11
15,185.7
0.0
1.0
Vegetation + Ants + Caribou calves
+ Kill sites
6
15,255.5
69.8
0.0
Vegetation + Ants + Caribou calves
5
15,380.1
194.4
0.0
Vegetation + Ants + Kill sites
5
15,390.0
204.3
0.0
Caribou calves + Kill sites
4
15,432.8
247.1
0.0
Vegetation + Ants
4
15,473.1
287.4
0.0
Ants
3
15,511.0
325.3
0.0
a
b
Vegetation + Ants + Caribou calves
+ Kill sitesd + Daily Killse
c
∆AICc
ω
Kill sites
3
15,517.1
331.4
0.0
Caribou calves
3
15,610.4
424.8
0.0
Null (intercept)
2
15,650.6
464.9
0.0
Vegetation
3
15,653.5
467.8
0.0
a
Occurrence of vegetation.
Occurrence of ant colonies.
c
Occurrence of caribou calves.
d
Occurrence of kills sites of caribou calves killed by black bears.
e
Daily number of caribou calves killed by black bears and two-way interactions between occurrence
variables (vegetation, ants, caribou calves, kill sites) and daily kills.
b
season. For each bear, we randomly assigned available locations
for all main-effect variables ≤4.10; Dormann et al., 2013). We de-
to a specific day drawn with replacement from the distribution of
rived maximum-likelihood estimates for GLMMs using adaptive
days of the corresponding calving season used locations of that
Gauss–Hermite approximation with five integration points (Bolker
bear. We then extracted predicted values of the relative abun-
et al., 2009). To evaluate the predictive ability of the top RSF models,
dance of vegetation, the relative abundance of ant colonies, the
we used 100 repetitions of 4-fold cross-validation (using Huberty’s
relative probability of caribou calf occurrence, and the relative
(1994) rule of thumb to determine the training-to-testing ratio) with
probability of kill site occurrence to used and available bear loca-
10 bins of equal size, calculating the average Spearman rank correla-
tions. We considered these occurrence variables (vegetation bio-
tion (rs) ± SE between the withheld data and the ranked bins (Boyce,
mass, ant colonies, caribou calves, kill sites; Figure 2) and two-way
Vernier, Nielsen, & Schmiegelow, 2002). We conducted all analyses
interactions between these occurrence variables and daily kills as
in program r version 3.0.2 (R Development Core Team, 2016), using
potential explanatory variables. These two-way interactions al-
lme4 to fit GLMMs.
lowed us to test our prediction that changes in the availability of
calves influenced the foraging tactics of bears; support for models
that included these interactions would indicate that bears adjusted
their habitat selection in response to changing calf availability. We
developed a list of candidate models (Table 1), and rescaled all variables between 0 and 1 (except ant colonies, which already ranged
3 | R E S U LT S
3.1 | Occurrence of vegetation, ant colonies and
caribou calves
from 0 to 1) to facilitate model convergence and interpretability
During the calving season, the occurrence of vegetation was great-
(Bastille-Rousseau et al., 2011). Our candidate RSFs took the form:
est in open landcover (barrens, heathland, wetlands) and lowest in
(
)
w(x) = exp β1 x1 + ⋯ + βu xuij + βkills x(kills)ij + ⋯ + βu xu × βkills x(kills)ij+γ0j (9)
forested landcover (conifer scrub, conifer forest; see Appendix S2:
Table S4). There was a strong positive correlation between NDVI and
vegetation growth (average conditional R2 = .68; see Appendix S1).
where w(x) represented the RSF scores, βu was the selection coef-
The relative abundance of ant colonies was greatest in conifer scrub,
ficient for explanatory variable xu for the ith observation and the jth
followed by barrens, heathland, wetlands and conifer forest (see
individual-year, βkills was the selection coefficient for daily kills x(kills),
Appendix S2: Table S4).
and γ0j was the random intercept for the jth individual-year.
The top-ranked RSF model (AICc weight >0.99) estimating the oc-
We used Akaike information criteria for small sample size (AICc;
currence of caribou calves included all landcover types, aspect and the
Burnham & Anderson, 2002) to identify the top-ranked caribou calf,
presence of a functional response (see Appendix S2: Table S1). Calves
kill site, and bear RSFs from our candidate models and assessed the
selected barrens and west and flat aspects, and avoided water (see
top models for multicollinearity using the variance inflation factor
Appendix S2: Table S5). Functional responses were revealed in variable
(VIF; Graham, 2003). We detected no multicollinearity issues (VIF
coefficients for two-way interactions between a specific landcover type
�RAYL et al.
and its proportion within a 5-km radius; selection for conifer forest and
conifer scrub increased as their proportion in the surrounding area increased, whereas selection for heathland decreased as its proportion in-
Journal of Animal Ecology
|
881
3.3 | Availability of caribou calves
(predation phenology)
creased in the surrounding area. The model had strong predictive ability
From 2003 to 2013, we estimated that an average of 5,524 calves
(rs = .89 ± .00).
were born in Middle Ridge North each year (range: 4,483–8,468; see
Appendix S2: Table S7). Predation by black bears was the dominant
3.2 | Occurrence of kill sites
source of mortality for these calves (see Appendix S2: Figure S1),
removing 32% of the calving cohort by 6 months of age (Figure 4).
As hypothesized (H1), we found that landscape attributes were related
Bears killed almost five times as many calves as coyotes to 30 June
to the probability of caribou calves being killed by black bears dur-
(1,301 vs. 275), and nearly two and a half times as many calves to
ing the calving season (see Appendix S2: Table S2). The top-ranked
6 months of age (1,763 vs. 725; Figure 4). Initially, the daily number
model (AICc weight = 0.98) included local-scale landcover types within
of calves killed by bears rapidly increased as calves accumulated on
a 100-m radius of the kill site, landscape-scale landcover types within
the landscape, peaking on 7 June when 66 calves were killed, before
a 5-km radius of the kill site, and elevation (Figure 3, see Appendix S2:
more gradually declining (Figure 4, see Appendix S2: Figure S2). We
Table S6). We found only mixed support for our prediction that bears
estimated that 49 bears lived in the MCP encompassing calving sea-
would be more successful at killing calves in areas with greater stalk-
son kill sites (1,012 km2), and that these bears killed an average of
ing cover, however. Calves were most vulnerable to bear predation in
0.63 calves/day to 30 June (27 calves/bear).
areas with greater proportions of local-scale conifer scrub or water
and at higher elevations, and least vulnerable in areas with greater
proportions of open landcover types and conifer forest (Figure 3). Risk
3.4 | Foraging tactics of black bears
for calves decreased in areas with greater proportions of landscape-
As hypothesized (H2), we found that black bears were actively
scale conifer scrub. The model was fairly robust to cross-validation
hunting caribou calves by selecting areas where calves were most
(rs = .65 ± .02).
vulnerable to bear predation (selection for kill sites; Figure 5, see
F I G U R E 3 Relative probability of
occurrence of kill sites of caribou calves
killed by black bears as a function of (a)
local-scale (proportion of a landcover type
within a 100-m radius) conifer scrub, (b)
local-scale water, (c) elevation (m) and (d)
landscape-scale (proportion of a specific
landcover type within a 5-km radius)
conifer scrub. Rug plots depict the x-axis
values of kill sites for all variables found to
influence predation risk (95% confidence
interval did not overlap 0). (e and f)
Local-and landscape-scale variables not
found to influence predation risk (95%
confidence interval overlapped 0). Each
predictor variable was plotted within
its observed range (except for local-
scale conifer scrub, water, barrens and
heathland variables, which were truncated
at 0.96, so that the total proportion of
local-scale landcover never exceeded 1
across the plotted range), while holding
elevation and landscape-scale landcover
variables constant at their mean values
and local-scale landcover variables
constant at 0.01
�882
|
Journal of Animal Ecology
RAYL et al.
dynamically adjusting its foraging patterns as the abundance and
vulnerability of its prey changed.
As we had hypothesized (H1), landscape features shaped patterns of predation risk for caribou calves. The strong influence of
landscape heterogeneity on mortality risk we observed is consistent with the prediction that ambush predators produce powerful
point-source cues that prey can reliably associate with increased risk
(Preisser, Orrock, & Schmitz, 2007; Schmitz, 2008). To our knowledge, our study is among the first to demonstrate that landscape
heterogeneity directly mediates vulnerability of ungulate neonates
to predation, although it has been shown previously in other systems with adult prey (Gervasi et al., 2013; Hebblewhite et al., 2005;
Hopcraft et al., 2005; Kauffman et al., 2007). Ungulate populations
frequently experience strong variation in juvenile survival with high
and stable adult female survival (Gaillard, Festa-Bianchet, Yoccoz,
Loison, & Toigo, 2000). Thus, juvenile survival often governs ungulate population trajectories. The pattern of early and intense predation we observed is common across systems with ursid mortality of
ungulate neonates (Zager & Beecham, 2006). Elsewhere, this predation phenology has been thought to play a key role in ungulate
population dynamics because bears kill most neonates before body
condition begins to mediate vulnerability (Adams et al., 1995; Barber-
Meyer, Mech, & White, 2008; Griffin et al., 2011; Zager & Beecham,
F I G U R E 4 (a) The estimated daily cumulative number of caribou
calves killed by black bears, coyotes and other causes, and the
estimated daily number of alive calves. (b) The estimated daily
number of caribou calves from Middle Ridge North that were killed
by black bears, coyotes and other causes, Newfoundland, Canada,
2003–2013
2006). Consequently, bear predation of calves frequently represents
an additive source of mortality. Although we were unable to assess
whether bear predation was additive or compensatory, calf predation is currently strongly influencing caribou population dynamics
in Newfoundland (Weir, Morrison, Luther, & Mahoney, 2014). Given
the influential role bears and other calf predators often play in determining ungulate population trajectories, we suggest more emphasis
Appendix S2: Table S8). We also found that bears selected for areas
should be placed on analyses such as ours, which examine the under-
with an increased probability of encountering calves (selection for
lying mechanisms of calf predation risk. In systems with substantial
caribou calves). Bears also selected for ant colonies, and avoided
calf predation, identifying features associated with increased risk of
vegetation. We also found support for our hypothesis (H3) that the
predation may offer solutions for management and conservation of
abundance and temporal vulnerability of caribou calves influenced
declining ungulate populations (cf. Whittington et al., 2011).
the foraging tactics of black bears; the top-ranked RSF (AICc weight
The design of our kill site RSF did not allow us to distinguish
>0.99) included two-way interactions between occurrence variables
whether landscape features influenced predation by affecting the
and daily kills. This model included all of the occurrence variables,
outcome of predator–prey encounters or by affecting the occur-
daily kills, and two-way interactions between occurrence variables
rence of predator–prey encounters. However, we speculate that
and daily kills. As daily kills increased, selection for the occurrence of
local-scale landscape features likely influenced calf vulnerability
kill sites increased, whereas selection for the occurrence of ant colo-
principally by altering the outcome of black bear-caribou calf en-
nies and caribou calves decreased. This model had strong predictive
counters. The relationship between local-scale landscape features
ability (rs = .96 ± .01).
and calf vulnerability was generally consistent with the expected
hunting tactics of an ambush predator. At the local-scale, open
4 | D I S CU S S I O N
landcover types were associated with low levels of predation risk
(Figure 3). The riskiest landscape feature, the proportion of local-
scale conifer scrub, provided dense cover that may have facilitated
We combined models of predation phenology, dynamic resource dis-
capture of calves when it occurred at higher proportions. Likewise,
tributions and RSFs to examine the spatiotemporal context of preda-
increased proportions of local-scale water elevated the risk of bear
tor–prey interactions from the perspective of a predator. Our results
predation, perhaps because vegetation in riparian areas and along
demonstrated that landscape heterogeneity influenced caribou calf
shorelines provided cover and water bodies limited escape routes.
vulnerability to black bear predation, suggested that bears were ac-
Additionally, encounters between calves and bears may have been
tively hunting calves, and provided evidence of a generalist predator
higher in areas with these landcover characteristics. Interestingly,
�RAYL et al.
Journal of Animal Ecology
|
883
F I G U R E 5 Relative probability of
selection of black bears as a function
of (a) the occurrence of vegetation, (b)
the occurrence of ant colonies, (c) the
occurrence of caribou calves and (d) the
occurrence of kill sites of caribou calves
killed by black bears. Rug plots depict the
x-axis values of telemetry locations of
black bears (the height of the tick marks
in (b) indicates the number of locations at
that value). (e–h) The relative probability
of selection of black bears as a function
of the occurrence variables and the daily
number of caribou calves killed by black
bears (daily kills). Each predictor variable
was plotted within its observed range
while holding all other variables constant
at their mean values, except daily kills
in panels (e)-(h), which was plotted at
the discrete values labelled in the panel
legends
although local-scale conifer forest could provide ambush cover for
A common assumption of many food web and predator–prey
black bears, predation risk did not increase with increasing propor-
models is that the per-capita rate at which predators kill prey varies
tions of this cover type. This may be because bears were primar-
exclusively as a function of prey density (the functional response;
ily resting when using this landcover type. Among landcover types,
Holling, 1959; Solomon, 1949; but see Fortin et al., 2015). Our re-
bears were least active in conifer forest (N.D. Rayl, unpubl. data),
sults, however, highlight the role that spatiotemporal heterogeneity
and it was the least productive landcover type for vegetation and
in prey abundance and prey vulnerability play in mediating the mor-
ants (see Appendix S2: Table S5). The considerable effect of eleva-
tality patterns of ungulate neonates succumbing to ursid predation.
tion and the lesser influence of the proportion of landscape-scale
Early in the calving season, when caribou calves were highly vulner-
conifer scrub on mortality risk indicated that large-scale features
able to predation, the daily number of calves killed by black bears
also affected the risk of mortality, likely by altering the frequency of
appeared to vary as a function of the abundance of calves (Figure 4,
bear-calf encounters.
see Appendix S2: Figure S2). Until 7 June, daily kill rates varied
�884
|
Journal of Animal Ecology
positively and linearly with calf abundance in a type-I functional
response (kill rate = 0.0003 × (alive calves) + 0.0553, F1,17 = 624.4,
RAYL et al.
current computing capacity, it would be challenging to quantify
uncertainty in our space-use predictions. Further, our conclusions
p < .001, R2 = .974; Holling, 1959). Less than 3 weeks after the onset
about the foraging tactics of black bears are based upon a sample
of calving, however, the daily kill rate declined sharply. Presumably,
size of only nine individuals. Although we estimate that these nine
this decoupling of kill rates and calf abundance occurred because
individuals comprise more than 18% of the black bear population
increasing mobility of calves lowered the risk of bear predation. As
living in our study area, it is a still a small sample size from which
a result, later in the calving season, predation rate was independent
to draw inferences. For these reasons, our results and conclusions
of calf abundance, and at a single calf abundance value bears killed
should be viewed with caution.
calves at two different rates (see Appendix S2: Figure S2).
The foraging tactics of carnivores are flexible, dynamic and
Interestingly, unlike prior studies examining black bear–caribou
complex. We examined a period of only 6 weeks, and in that short
calf interactions (Bastille-Rousseau et al., 2011; Latham et al., 2011),
time, observed large fluctuations in the availability of a major re-
we found compelling evidence that suggested bears with access to
source for black bears, caribou calves and what appeared to be
caribou were actively hunting calves (in support of H2). These differ-
reciprocal changes in the space-use tactics of bears. These find-
ing results may be due to caribou calves occurring at a much higher
ings clearly underscore the value of incorporating temporal het-
density in our system, variation in space-use strategies of parturient
erogeneity into analyses of resource selection. As ecologists, we
caribou among systems, or because previous studies were unable to
have become increasingly sophisticated at accounting for spatial
evaluate whether bears selected areas where calves were vulnera-
heterogeneity when examining the resource selection patterns of
ble to predation. In our system, black bears exhibited the strongest
animals, but we rarely include similarly complex temporal variation
selection for areas where they were more likely to kill caribou calves.
in our investigations. In this study, we offer one possible frame-
This is consistent with findings from other studies examining the
work for integrating temporal dynamics of prey availability into
space use of large ambush predators (Hopcraft et al., 2005). Bears
investigations of predator–prey interactions and carnivore space
also strongly selected areas where they were more likely to encoun-
use, and demonstrate the power of such an approach. In doing
ter adult female caribou. Together, these results suggest that bears
so, we reveal new insights into how variation in prey availability
were actively hunting caribou calves during the calving season.
may mediate predator–prey dynamics, and move towards a mech-
To date, the influence of seasonal variability in the abundance
and vulnerability of prey on predation has not received sufficient
anistic understanding of the foraging tactics of a large, generalist
predator.
attention in studies of carnivore ecology (Pereira, Owen-Smith, &
Moleón, 2014). Here, we begin to address this deficit, and, in support of our hypothesis (H3), provide novel evidence that changing
AC K N OW L E D G E M E N T S
prey availability influenced the foraging tactics of a generalist pred-
This study was funded, and N.D.R. was supported, by the Institute
ator (Figure 5). Classical theory predicts that as the availability of
for Biodiversity, Ecosystem Science & Sustainability and the
profitable prey declines, foragers should expand their diet to include
Sustainable Development and Strategic Science Division of the
less profitable prey items (Charnov, 1976; MacArthur & Pianka,
Government of Newfoundland and Labrador Department of
1966). Our findings suggest that bears may have compensated for
Environment and Conservation. Additional funding was provided
capturing fewer calves by consuming more ant protein (although
by the Safari Club International Foundation, and the University
overall selection for ant-rich areas was always low). Additionally,
of Massachusetts, Amherst. We thank S. Ellsworth, S. Gullage,
bears appeared to switch from a hunting tactic that maximized suc-
T. Hodder, J. McGinn, A. Mouland, F. Norman and T. Murphy for
cess rates (Figure 5h) to a hunting tactic that maximized encounter
logistical support. We thank N. Brooks, R. Curran, B. Efford, C.
rates (Figure 5g) based upon the efficiency with which they could
Gosse, T. Hodder, J. Maloney, M. McGrath, J. Neville, D. O’Leary,
catch calves. If so, this would be the first evidence of which we are
P. Saunders, B. Slade and P. Tremblett, for help with caribou calf
aware of a predator exhibiting facultative switching between hunt-
captures, calf telemetry and calf mortality retrievals. We thank
ing modes. These results reveal the dynamic nature of predation
T. Porter and P. Tremblett for assisting with bear captures. We
risk, and highlight the challenge caribou face in reliably discerning
thank J. Maloney and B. Slade for safe flying. We thank Editor J.M.
and responding to the spatial and temporal distribution of that risk.
Gaillard, Associate Editor A. Loison, S. Focardi, G. Hilderbrand and
Additional research is needed to clarify the connection between the
two anonymous reviewers for valuable comments that improved
foraging ecology and space-use strategies of these bears, and to
this manuscript. Any use of trade, firm or product names is for
examine whether female caribou respond to the shifting space-use
descriptive purposes only and does not imply endorsement by the
patterns of predatory bears.
U.S. Government.
We did not account for estimation uncertainty in our analyses.
Instead, we assumed that the availability of caribou calves and the
occurrence of vegetation, ant colonies, calves and kill sites were all
AU T H O R S ’ C O N T R I B U T I O N S
known without error. At present, we do not have data to estimate
N.D.R., G.B.-R., T.K.F. and J.F.O. conceived the study; N.D.R., G.B.-R.
uncertainty in the availability of caribou calves. Additionally, given
and M.A.M. performed all analyses; S.P.M., C.E.S., R.D.O., T.K.F.,
�Journal of Animal Ecology
RAYL et al.
J.F.O., D.L.M. and L.P.W. secured funding; N.D.R., G.B.-R., M.A.M.,
S.P.M., C.E.S. and K.P.L. collected and prepared the data; N.D.R.
wrote the manuscript, and all authors contributed to revisions and
gave final approval for publication.
DATA AC C E S S I B I L I T Y
All data were collected by the Newfoundland and Labrador
Department of Environment and Conservation. Data are available
from the Dryad Digital Repository: https://doi.org/10.5061/dryad.
t081d (Rayl et al., 2018).
ORCID
Nathaniel D. Rayl
http://orcid.org/0000-0003-3846-2764
Guillaume Bastille-Rousseau
Matthew A. Mumma
http://orcid.org/0000-0001-6799-639X
http://orcid.org/0000-0003-1954-6524
REFERENCES
Adams, L. G., Singer, F. J., & Dale, B. W. (1995). Caribou calf mortality
in Denali National Park, Alaska. Journal of Wildlife Management, 59,
584–594. https://doi.org/10.2307/3802467
Barber-Meyer, S. M., Mech, L. D., & White, P. J. (2008). Elk calf survival
and mortality following wolf restoration to Yellowstone National
Park. Wildlife Monographs, 169, 1–30. https://doi.org/10.2193/2008004
Basille, M., Fortin, D., Dussault, C., Bastille-Rousseau, G., Ouellet, J. P.,
& Courtois, R. (2015). Plastic response of fearful prey to the spatiotemporal dynamics of predator distribution. Ecology, 96, 2622–2631.
https://doi.org/10.1890/14-1706.1
Bastille-Rousseau, G., Fortin, D., Dussault, C., Courtois, R., & Ouellet,
J.-P. (2011). Foraging strategies by omnivores: Are black bears
actively searching for ungulate neonates or are they simply opportunistic predators? Ecography, 34, 588–596. https://doi.
org/10.1111/j.1600-0587.2010.06517.x
Bastille-Rousseau, G., Potts, J. R., Schaefer, J. A., Lewis, M. A., Ellington,
E. H., Rayl, N. D., … Murray, D. L. (2015). Unveiling trade-offs in
resource selection of migratory caribou using a mechanistic movement model of availability. Ecography, 38, 1049–1059. https://doi.
org/10.1111/ecog.01305
Bastille-Rousseau, G., Schaefer, J. A., Lewis, K. P., Mumma, M. A.,
Ellington, E. H., Rayl, N. D., … Murray, D. L. (2016). Phase-dependent
climate-predator interactions explain three decades of variation in
neonatal caribou survival. Journal of Animal Ecology, 85, 445–456.
https://doi.org/10.1111/1365-2656.12466
Bergerud, A. T. (1974). The role of the environment in the aggregation,
movement and disturbance behaviour of caribou. In V. Geist & F.
Walther (Eds.), The behaviour of ungulates and its relation to management (pp. 552–584). Morges, Switzerland: International Union for
Conservation of Nature and Natural Resources Publications New
Series 24.
Bergerud, A. T. (1975). The reproductive season of Newfoundland
caribou. Canadian Journal of Zoology, 53, 1213–1221. https://doi.
org/10.1139/z75-145
Boileau, F., Crête, M., & Huot, J. (1994). Food habits of the black bear,
Ursus americanus, and habitat use in Gaspésie Park, Eastern Quebec.
Canadian Field Naturalist, 108, 162–169.
|
885
Bolker, B. M., Brooks, M. E., Clark, C. J., Geange, S. W., Poulsen, J. R.,
Stevens, M. H. H., & White, J.-S. S. (2009). Generalized linear
mixed models: A practical guide for ecology and evolution. Trends
in Ecology and Evolution, 24, 127–135. https://doi.org/10.1016/j.
tree.2008.10.008
Boyce, M. S., Vernier, P. R., Nielsen, S. E., & Schmiegelow, F. K. A. (2002).
Evaluating resource selection functions. Ecological Modelling, 157,
281–300. https://doi.org/10.1016/S0304-3800(02)00200-4
Burnham, K. P., & Anderson, D. R. (2002). Model selection and multimodel
inference: A practical information-theoretic approach. New York, NY:
Springer.
Charnov, E. L. (1976). Optimal foraging, the marginal value theorem. Theoretical Population Biology, 9, 129–136. https://doi.
org/10.1016/0040-5809(76)90040-X
Clutton-Brock, T. H., & Harvey, P. H. (1978). Mammals, resources
and reproductive strategies. Nature, 273, 191–195. https://doi.
org/10.1038/273191a0
Coulson, T., Milner-Gulland, E. J., & Clutton-Brock, T. (2000). The relative
roles of density and climatic variation on population dynamics and
fecundity rates in three contrasting ungulate species. Proceedings of
the Royal Society B: Biological Sciences, 267, 1771–1779. https://doi.
org/10.1098/rspb.2000.1209
Dormann, C. F., Elith, J., Bacher, S., Buchmann, C., Carl, G., Carré, G., …
Lautenbach, S. (2013). Collinearity: A review of methods to deal with
it and a simulation study evaluating their performance. Ecography, 36,
027–046. https://doi.org/10.1111/j.1600-0587.2012.07348.x
Fieberg, J., & Delgiudice, G. D. (2009). What time is it? Choice of time
origin and scale in extended proportional hazards models. Ecology,
90, 1687–1697. https://doi.org/10.1890/08-0724.1
Fifield, D. A., Lewis, K. P., & Gullage, S. E. (2013). Application of distance
sampling to determine calving ground abundance and aggregation
of parturient females in the middle ridge herd, June 2012. Technical
Bulletin No. 005, Sustainable Development and Strategic Science,
Department of Environment and Cons. St. John’s, Newfoundland
and Labrador, Canada.
Fortin, D., Buono, P., Schmitz, O. J., Courbin, N., Losier, C., Drapeau, P.,
… Mainguy, J. (2015). A spatial theory for characterizing predator
– Multiprey interactions in heterogeneous landscapes. Proceedings
of the Royal Society B, 282, 20150973. https://doi.org/10.1098/
rspb.2015.0973
Fuller, A. K., Harrison, D. J., & Vashon, J. H. (2007). Winter habitat selection by Canada lynx in Maine: Prey abundance or accessibility? Journal of Wildlife Management, 71, 1980–1986. https://doi.
org/10.2193/2006-288
Gaillard, J.-M., Festa-Bianchet, M., Yoccoz, N. G., Loison, A., & Toigo, C.
(2000). Temporal variation in fitness components and population
dynamics of large herbivores. Annual Review of Ecology, Evolution,
and Systematics, 31, 367–393. https://doi.org/10.1146/annurev.
ecolsys.31.1.367
Gervasi, V., Sand, H., Zimmermann, B., Mattisson, J., Wabakken, P., &
Linnell, J. D. C. (2013). Decomposing risk: Landscape structure and
wolf behavior generate different predation patterns in two sympatric ungulates. Ecological Applications, 23, 1722–1734. https://doi.
org/10.1890/12-1615.1
Gilbert, S. L., Lindberg, M. S., Hundertmark, K. J., & Person, D. K.
(2014). Dead before detection: Addressing the effects of left truncation on survival estimation and ecological inference for neonates. Methods in Ecology and Evolution, 5, 992–1001. https://doi.
org/10.1111/2041-210X.12234
Gillies, C. S., Hebblewhite, M., Nielsen, S. E., Krawchuk, M. A.,
Aldridge, C. L., Frair, J. L., … Jerde, C. L. (2006). Application
of random effects to the study of resource selection by animals. Journal of Animal Ecology, 75, 887–898. https://doi.
org/10.1111/j.1365-2656.2006.01106.x
�886
|
Journal of Animal Ecology
Graham, M. H. (2003). Confronting multicollinearity in ecological multiple regression. Ecology, 84, 2809–2815. https://doi.
org/10.1890/02-3114
Griffin, K. A., Hebblewhite, M., Robinson, H. S., Zager, P., Barber-Meyer,
S. M., Christianson, D., … White, P. J. (2011). Neonatal mortality of
elk driven by climate, predator phenology and predator community
composition. Journal of Animal Ecology, 80, 1246–1257. https://doi.
org/10.1111/j.1365-2656.2011.01856.x
Hebblewhite, M., Merrill, E., & McDermid, G. (2008). A multi-scale test
of the forage maturation hypothesis in a partially migratory ungulate population. Ecological Monographs, 78, 141–166. https://doi.
org/10.1890/06-1708.1
Hebblewhite, M., Merrill, E. H., & McDonald, T. L. (2005). Spatial decomposition of predation risk using resource selection functions: An example in a wolf-elk-predator-prey system. Oikos, 1, 101–111. https://
doi.org/10.1111/j.0030-1299.2005.13858.x
Heisey, D. M., & Patterson, B. R. (2006). A review of methods to estimate
cause-specific mortality in presence of competing risks. Journal of
Wildlife Management, 70, 1544–1555. https://doi.org/10.2193/0022
-541X(2006)70[1544:AROMTE]2.0.CO;2
Holling, C. S. (1959). The components of predation as revealed by a study
of small mammal predation of the European pine sawfly. Canadian
Entomologist, 91, 293–320. https://doi.org/10.4039/Ent91293-5
Holling, C. S. (1965). The functional response of predators to prey density and its role in mimicry and population regulation. Memoirs of the
Entomological Society of Canada, 97, 1–60.
Holt, R. D. (1977). Predation, apparent competition, and the structure
of prey communities. Theoretical Population Biology, 12, 197–229.
https://doi.org/10.1016/0040-5809(77)90042-9
Hopcraft, J. G. C., Sinclair, A. R. E., & Packer, C. (2005). Planning
for success: Serengeti lions seek prey accessibility rather than
abundance. Journal of Animal Ecology, 74, 559–566. https://doi.
org/10.1111/j.1365-2656.2005.00955.x
Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic regression
(2nd ed.). New York, NY: John Wiley & Sons Inc. https://doi.
org/10.1002/0471722146
Huberty, C. J. (1994). Applied discriminant analysis. New York, NY: Wiley.
Kauffman, M. J., Varley, N., Smith, D. W., Stahler, D. R., MacNulty, D. R.,
& Boyce, M. S. (2007). Landscape heterogeneity shapes predation in
a newly restored predator-prey system. Ecology Letters, 10, 690–700.
https://doi.org/10.1111/j.1461-0248.2007.01059.x
Krebs, C. J., Boutin, S., Boonstra, R., Sinclair, A. R. E., Smith, J. N. M.,
Dale, M. R. T., … Turkington, R. (1995). Impact of food and predation
on the snowshoe hare cycle. Science, 269, 1112–1115. https://doi.
org/10.1126/science.269.5227.1112
Latham, A. D. M., Latham, M. C., & Boyce, M. S. (2011). Habitat selection
and spatial relationships of black bears (Ursus americanus) with woodland caribou (Rangifer tarandus caribou) in northeastern Alberta.
Canadian Journal of Zoology, 89, 267–277. https://doi.org/10.1139/
z10-115
Latombe, G., Fortin, D., & Parrott, L. (2014). Spatio-temporal dynamics in the response of woodland caribou and moose to the passage
of grey wolf. Journal of Animal Ecology, 83, 185–198. https://doi.
org/10.1111/1365-2656.12108
Lent, P. C. (1974). Mother-infant relationships in ungulates. In V. Geist
& F. Walther (Eds.), The behaviour of ungulates and its relationship to
management (pp. 14–55). Morges, Switzerland: International Union
for the Conservation of Nature and Natural Resources.
Lima, S. L. (2002). Putting predators back into behavioral predator-prey
interactions. Trends in Ecology and Evolution, 17, 70–75. https://doi.
org/10.1016/S0169-5347(01)02393-X
Lima, S. L., & Dill, L. M. (1990). Behavioral decisions made under the risk
of predation: A review and prospectus. Canadian Journal of Zoology,
68, 619–640. https://doi.org/10.1139/z90-092
RAYL et al.
MacArthur, R. H., & Pianka, E. R. (1966). On optimal use of a patchy
environment. The American Naturalist, 100, 603–609. https://doi.
org/10.1086/282454
Mahoney, S. P., Lewis, K. P., Weir, J. N., Morrison, S. F., Luther, J. G.,
Schaefer, J. A., … Latifovic, R. (2016). Woodland caribou calf mortality in Newfoundland: Insights into the role of climate, predation and
population density over three decades of study. Population Ecology,
58, 91–103. https://doi.org/10.1007/s10144-015-0525-y
Mahoney, S. P., & Schaefer, J. A. (2002). Long-term changes in demography and migration of Newfoundland Caribou. Journal of Mammalogy,
83, 957–963. https://doi.org/10.1644/1545-1542(2002)083<095
7:LTCIDA>2.0.CO;2
Mahoney, S. P., Virgl, J. A., Fong, D. W., Maccharles, A. M., & McGrath,
M. (1998). Evaluation of a mark-resighting technique for woodland
caribou in Newfoundland. Journal of Wildlife Management, 62, 1227–
1235. https://doi.org/10.2307/3801986
Manly, B. F. J., McDonald, L. L., Thomas, D. L., McDonald, T. L., &
Erickson, W. P. (2002). Resource selection by animals: Statistical design
and analysis for field studies (2nd ed.). Dordrecht, the Netherlands:
Kluwer Academic Publishers.
Moreau, G., Fortin, D., Couturier, S., & Duchesne, T. (2012). Multi-level
functional responses for wildlife conservation: The case of threatened caribou in managed boreal forests. Journal of Applied Ecology,
49, 611–620. https://doi.org/10.1111/j.1365-2664.2012.02134.x
Mumma, M. A., Soulliere, C. E., Mahoney, S. P., & Waits, L. P. (2014).
Enhanced understanding of predator-prey relationships using
molecular methods to identify predator species, individual
and sex. Molecular Ecology Resources, 14, 100–108. https://doi.
org/10.1111/1755-0998.12153
Murray, D. L., Boutin, S., & O’Donoghue, M. (1994). Winter habitat selection by lynx and coyotes in relation to snowshoe hare abundance.
Canadian Journal of Zoology, 72, 1444–1451. https://doi.org/10.1139/
z94-191
Mysterud, A., & Ims, R. A. (1998). Functional responses in habitat use:
Availability influences relative use in trade-off situations. Ecology, 79,
1435–1441. https://doi.org/10.1890/0012-9658(1998)079[1435:
FRIHUA]2.0.CO;2
Noyce, K. V., Kannowski, P. B., & Riggs, M. R. (1997). Black bears as ant-
eaters: Seasonal associations between bear myrmecophagy and ant
ecology in north-central Minnesota. Canadian Journal of Zoology, 75,
1671–1686. https://doi.org/10.1139/z97-794
Orians, G. H., Cochran, P. A., Duffield, J. W., Fuller, T. K., Gutierrez, R.
J., Hanemann, W. M., … Yaska, G. (1997). Wolves, bears, and their
prey in Alaska: Biological and social challenges in wildlife management.
Washington, DC: National Academy Press.
Pereira, L. M., Owen-Smith, N., & Moleón, M. (2014). Facultative predation and scavenging by mammalian carnivores: Seasonal, regional
and intra-guild comparisons. Mammal Review, 44, 44–55. https://doi.
org/10.1111/mam.12005
Preisser, E. L., Orrock, J. L., & Schmitz, O. J. (2007). Predator hunting mode and habitat domain alter nonconsumptive effects in
predator-prey interactions. Ecology, 88, 2744–2751. https://doi.
org/10.1890/07-0260.1
R Development Core Team. (2016). R: A language and environment for statistical computing. Vienna, Austria: R Development Core Team.
Rayl, N. D., Bastille-Rousseau, G., Organ, J. F., Mumma, M. A., Mahoney,
S. P., Soulliere, C. E., … Fuller, T. K. (2018). Data from: Spatiotemporal
heterogeneity in prey abundance and vulnerability shapes the foraging tactics of an omnivore. Dryad Digital Repository, https://doi.
org/10.5061/dryad.t081d
Rayl, N. D., Fuller, T. K., Organ, J. F., McDonald, J. E., Mahoney, S. P.,
Soulliere, C., … Murray, D. L. (2014). Mapping the distribution of a prey
resource: Neonate caribou in Newfoundland. Journal of Mammalogy,
95, 328–339. https://doi.org/10.1644/13-MAMM-A-133.1
�Journal of Animal Ecology
RAYL et al.
Rayl, N. D., Fuller, T. K., Organ, J. F., McDonald, J. E., Otto, R. D., BastilleRousseau, G., … Mahoney, S. P. (2015). Spatiotemporal variation in the
distribution of potential predators of a resource pulse: Black bears
and caribou calves in Newfoundland. Journal of Wildlife Management,
79, 1041–1050. https://doi.org/10.1002/jwmg.936
Schmitz, O. J. (1998). Direct and indirect effects of predation and predation
risk in old-field interaction webs. The American Naturalist, 151, 327–342.
Schmitz, O. J. (2008). Effects of predator hunting mode on grassland
ecosystem function. Science, 319, 952–954. https://doi.org/10.1126/
science.1152355
Sikes, R. S., & Gannon, W. L. & American Society of Mammalogists.
(2011). Guidelines of the American Society of Mammalogists for
the use of wild mammals in research. Journal of Mammalogy, 92,
235–253.
Solomon, M. E. (1949). The natural control of animal populations. Journal
of Animal Ecology, 18, 1–35. https://doi.org/10.2307/1578
Tilman, D. (1978). Cherries, ants and tent caterpillars: Timing of nectar
production in relation to susceptibility of caterpillars to ant predation. Ecology, 59, 686–692. https://doi.org/10.2307/1938771
Weir, J. N., Morrison, S. F., Luther, J. G., & Mahoney, S. P. (2014). Caribou
Data Synthesis-Progress Report #2. Status of the Newfoundland
Population of Woodland Caribou. Technical Bulletin No. 008, St.
John’s, Newfoundland and Labrador: Sustainable Development and
Strategic Science, Government of Newfoundland and Labrador.
Whittington, J., Hebblewhite, M., Decesare, N. J., Neufeld, L., Bradley,
M., Wilmshurst, J., & Musiani, M. (2011). Caribou encounters
with wolves increase near roads and trails: A time-to-event approach. Journal of Applied Ecology, 48, 1535–1542. https://doi.
org/10.1111/j.1365-2664.2011.02043.x
|
887
Wirsing, A. J., Steury, T. D., & Murray, D. L. (2002). Relationship between
body condition and vulnerability to predation in red squirrels and
snowshoe hares. Journal of Mammalogy, 83, 707–715. https://doi.org
/10.1644/1545-1542(2002)083<0707:RBBCAV>2.0.CO;2
Young, K. V., Brodie, E. D. J., & Brodie, E. D. I. (2004). How the horned
lizard got its horns. Science, 304, 65. https://doi.org/10.1126/
science.1094790
Zager, P., & Beecham, J. (2006). The role of American black bears and brown
bears as predators on ungulates in North America. Ursus, 17, 95–108.
https://doi.org/10.2192/1537-6176(2006)17[95:TROABB]2.0.CO;2
Zieminski, C. J. (2016). Trophic relationships among caribou calf predators
in Newfoundland. MS thesis, University of Massachusetts, Amherst,
MA.
S U P P O R T I N G I N FO R M AT I O N
Additional Supporting Information may be found online in the
supporting information tab for this article.
How to cite this article: Rayl ND, Bastille-Rousseau G, Organ
JF, et al. Spatiotemporal heterogeneity in prey abundance
and vulnerability shapes the foraging tactics of an omnivore.
J Anim Ecol. 2018;87:874–887. https://doi.org/10.1111/13652656.12810
�
https://cpw.cvlcollections.org/files/original/f62b0ac5001ba605f933541a09eec53b.pdf
e21dfe23e2b71058afcfc3c4b84327f6
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
Journal Articles
Description
An account of the resource
CPW peer-reviewed journal publications
Text
A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text.
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
Spatiotemporal heterogeneity in prey abundance and vulnerability shapes the foraging tactics of an omnivore
Description
An account of the resource
<ol>
<li>Prey abundance and prey vulnerability vary across space and time, but we know little about how they mediate predator–prey interactions and predator foraging tactics. To evaluate the interplay between prey abundance, prey vulnerability and predator space use, we examined patterns of black bear (<i>Ursus americanus</i>) predation of caribou (<i>Rangifer tarandus</i>) neonates in Newfoundland, Canada using data from 317 collared individuals (9 bears, 34 adult female caribou, 274 caribou calves).</li>
<li>During the caribou calving season, we predicted that landscape features would influence calf vulnerability to bear predation, and that bears would actively hunt calves by selecting areas associated with increased calf vulnerability. Further, we hypothesized that bears would dynamically adjust their foraging tactics in response to spatiotemporal changes in calf abundance and vulnerability (collectively, calf availability). Accordingly, we expected bears to actively hunt calves when they were most abundant and vulnerable, but switch to foraging on other resources as calf availability declined.</li>
<li>As predicted, landscape heterogeneity influenced risk of mortality, and bears displayed the strongest selection for areas where they were most likely to kill calves, which suggested they were actively hunting caribou. Initially, the per-capita rate at which bears killed calves followed a type-I functional response, but as the calving season progressed and calf vulnerability declined, kill rates dissociated from calf abundance. In support of our hypothesis, bears adjusted their foraging tactics when they were less efficient at catching calves, highlighting the influence that predation phenology may have on predator space use. Contrary to our expectations, however, bears appeared to continue to hunt caribou as calf availability declined, but switched from a tactic of selecting areas of increased calf vulnerability to a tactic that maximized encounter rates with calves.</li>
<li>Our results reveal that generalist predators can dynamically adjust their foraging tactics over short time-scales in response to changing prey abundance and vulnerability. Further, they demonstrate the utility of integrating temporal dynamics of prey availability into investigations of predator–prey interactions, and move towards a mechanistic understanding of the dynamic foraging tactics of a large omnivore.</li>
</ol>
Bibliographic Citation
A bibliographic reference for the resource. Recommended practice is to include sufficient bibliographic detail to identify the resource as unambiguously as possible.
Rayl, N.D., G. Bastille-Rousseau, J.F. Organ, M.A. Mumma, S.P. Mahoney, C.E. Soulliere, K.P. Lewis, R.D. Otto, D.L. Murray, L.P. Waits, and T.K. Fuller. 2018. Spatiotemporal heterogeneity in prey abundance and vulnerability shapes the foraging tactics of an omnivore. Journal of Animal Ecology 87:874-887. <a href="https://doi.org/10.1111/1365-2656.12810" target="_blank" rel="noreferrer noopener">https://doi.org/10.1111/1365-2656.12810</a>
Creator
An entity primarily responsible for making the resource
Rayl, Nathaniel D.
Bastille-Rousseau, Guillaume
Organ, John F.
Mumma, Matthew A.
Mahoney, Shane P.
Soulliere, Colleen E.
Lewis, Keith P.
Otto, Robert D.
Murray, Dennis L.
Waits, Lisette P.
Fuller, Todd K.
Subject
The topic of the resource
Black bear (<em>Ursus americanus</em>)
Caribou (<em>Rangifer tarandus</em>) calves
Cause-specific survival analysis
Foraging tactics
Kill rates
Predation risk
Trophic interaction
Ungulate
Extent
The size or duration of the resource.
14 pages
Date Created
Date of creation of the resource.
2018-02-16
Rights
Information about rights held in and over the resource
<a href="http://rightsstatements.org/vocab/InC-NC/1.0/">In Copyright - Non-Commercial Use Permitted</a>
Format
The file format, physical medium, or dimensions of the resource
application/pdf
Language
A language of the resource
English
Is Part Of
A related resource in which the described resource is physically or logically included.
Journal of Animal Ecology
Type
The nature or genre of the resource
Article