Mind the cat: Conservation management of a protected dominant scavenger indirectly affects an endangered apex predator

Full text

1
Citation: Krofel, M., Jerina, K. 2016. Mind the cat: Conservation management of protected dominant
scavenger indirectly affects an endangered apex predator. – Biological Conservation, 197: 40-46. doi:
10.1016/j.biocon.2016.02.019
Mind the cat: Conservation management of a protected dominant scavenger indirectly affects an
endangered apex predator
Miha Krofel* and Klemen Jerina
Department of Forestry and Renewable Forest Resources, Biotechnical Faculty, University of
Ljubljana, Večna pot 83, SI-1001 Ljubljana, Slovenia
* Corresponding author; e-mail: miha.krofel@gmail.com; tel.: +386 51 228 717
Abstract
Interspecific interactions are among the key factors influencing the structure of animal communities
and have high relevance for conservation. However, managers, conservationists and decision-makers
rarely consider the potential side-effects of single-species carnivore management for the
conservation of other carnivores. We studied how management of protected brown bears (Ursus
arctos) affected interspecific interactions with an endangered apex predator, the Eurasian lynx (Lynx
lynx) in Slovenia. Due to large body size and superb olfactory abilities, bears are one of the most
important dominant scavengers and regularly usurp kills from other large predators, a process known
as kleptoparasitism. At the same time, bears throughout the world are usually actively managed
through zone-specific culling regimes, supplemental feeding, and translocations. This can
considerably alter bear densities and activity patterns and in turn influence interactions among
carnivores. Overall, we observed that bear scavenging pressure resulted in substantial energetic
losses for Eurasian lynx. The probability of lynx losing kills to bears ranged from 8 to 74% and strongly
depended on local bear densities and monthly bear movement rates. Kleptoparasitic interaction
intensity differed almost 3-fold between different bear management zones. Furthermore, the
presence of a bear feeding site increased the odds of lynx losing kills by 5-fold compared to areas
>1000 m from these sites. We suggest that existing bear-feeding regimes should be reconsidered in
order to reduce unwanted side-effects of this controversial practice on endangered apex predators.
We also call attention to the importance of considering impacts of interspecific interactions in
wildlife management and conservation.
Keywords: wildlife management, interspecific interaction, kleptoparasitism, cascading effects, Lynx
lynx, Ursus arctos
Highlights
 We examined loss of roe deer carcasses killed by lynx to scavenging bears
 Bear management measures indirectly increased loss of kills for the lynx
 Kleptoparasitism increased cumulatively with bear densities and bear activity
 Lynx lost more kills near bear feeding sites

2
1 Introduction
Interspecific interactions have profound effects on ecosystem function and community structure
(Begon et al. 2006). Understanding the underlying mechanisms that influence interspecific
interactions is increasingly an important aspect of animal conservation (Creel et al. 2001; Moleón et
al. 2014). Despite the potential to alter entire communities, wildlife managers rarely consider
possible negative side-effects of management decisions on interspecific interactions (Linnell and
Strand 2000; Ordiz et al. 2013; Selva et al. 2014). More empirical knowledge is needed for better
conservation and management that accounts for interactions across multiple levels of ecosystems
(Lozano et al. 2013; Périquet et al. 2014). This is particularly true for strongly interacting species, such
as large mammalian carnivores due to their cascading effects on numerous species and terrestrial
ecosystems worldwide (Estes et al. 2011; Ripple et al. 2014).
Researchers are increasingly concerned about unwanted or unexpected impacts of specific
management actions involving large carnivores. For example, hunting increases infanticide in African
lions (Panthera leo; Loveridge et al. 2007; Whitman et al. 2004) and brown bears (Ursus arctos;
Gosselin et al. 2015; Swenson et al. 1997), decreases pack stability in wolves (Canis spp.) and
increases hybridization with domestic dogs (Moura et al. 2014; Rutledge et al. 2010). For cougars
(Puma concolor) and African lions, hunting changes their distribution and movement patterns
(Davidson et al. 2011; Maletzke et al. 2014). Hunting also changes brown bear activity and foraging
behaviour (Ordiz et al. 2012). Changes in abundance, sociality, foraging, spatial distribution and
movement patterns have also been reported as a consequence of carnivores exploiting readily
available human-provided foods (Newsome et al. 2015; Oro et al. 2013). On the other hand, much
less is known about the effects of these measures beyond the managed species (Périquet et al.
2014). Consequently, carnivore management programs rarely consider the indirect effects on other
apex predators via changes in interspecific interactions.
Interspecific interactions among carnivores frequently occur at kill sites (Atwood and Gese 2008). The
stealing of kills or kleptoparasitism is recognized as an important part of large carnivore ecology with
the potential to change entire ecological communities (Allen et al. 2014). High levels of
kleptoparasitism can directly threaten predators (Carbone et al. 1997; Gorman et al. 1998).
Kleptoparasitic interactions among bears and solitary felids provide an opportunity to study these
interactions. Solitary felids that kill large prey are characterized by a prolonged consumption process
of their kills (Jobin et al. 2000; Stander et al. 1997) and are regularly exposed to kleptoparasitism in
their ranges worldwide (Krofel et al. 2012a). As the largest terrestrial scavengers with superb
olfactory abilities, bears are one of the most important dominant scavengers and kleptoparasites in
the Holarctic region (Allen et al. 2014; Krofel et al. 2012a; Murphy et al. 1998). At the same time,
ursids are often actively managed either through hunting and management removals (Kaczensky et
al. 2013; Nielsen et al. 2004) or, in case of endangered populations, through translocations (Clark et
al. 2002). In addition, bear movements, local densities, diet and other life history traits can be greatly
altered through human-caused changes of habitat and food availability (Apps et al. 2004; Güthlin et
al. 2011; Kavčič et al. 2015; Penteriani et al. 2010). However, it is poorly understood how
management of dominant scavengers like bears affect their interactions with other predators.
Our research focuses on how management of protected brown bears in Slovenia influences
interspecific interactions with a sympatric apex predator, the Eurasian lynx (Lynx lynx). The highly
endangered Dinaric lynx population is impacted by kleptoparasitism from brown bears, through
substantial energetic losses and potential reduction in reproductive success. On average, bears
usurped one third of lynx kills and despite increasing their kill rate, lynx are not able to fully
compensate the losses. (Krofel et al. 2012a). These kleptoparasitic interactions were highest during
the bear mating season and lowest in the denning period (Krofel et al. 2012a). Brown bears in the

3
region are intensively managed through a zoning system of culling and supplemental feeding, which
was shown to considerably alter bear distribution, local densities, diet and activity patterns (Jerina
and Adamič 2008; Jerina et al. 2013; Kavčič et al. 2015; Steyaert et al. 2014). We speculated that
these management actions could influence interactions between bears and the lynx (Krofel et al.
2012a). Here we tested this hypothesis. We predicted that the proportion of lynx kills usurped by
bears would cumulatively increase with: 1) higher local bear densities, 2) higher bear movement
rates, and 3) proximity to bear feeding sites.
2 Material and methods
2.1 Study area and study species
The study was conducted in the Northern Dinaric Mountain Range in Slovenia (45˚25'–45˚47'N,
14˚15'–14˚50'E) in mixed temperate forests dominated by fir and beech (Omphalodo-Fagetum s.
lat.). The altitudes range from 200 m to the peak of Mount Snežnik at 1 796 m. The climate is a mix of
influences from the Alps, the Mediterranean sea and the Pannonia basin with annual temperature
averaging 5–8˚ C, ranging from average maximum of 32˚ C to a minimum of –20˚ C, and average
annual precipitation of 1 400–3 500 mm.
The study area encompasses the north-western part of the transboundary Alps-Dinaric–Pindos
brown bear population. Here bears are under strong influence of various human activities and
management measures, which created a large gradient in bear densities. Bears were nearly
extirpated in the late 19th century, but since the 1940s, their numbers and distribution increased due
to conservation measures, including establishment of the Core Bear Protective Area (CBPA) of 3 500
km2 within the Dinaric Range in 1966, where bear hunting was strictly regulated (Simonič 1994). In
contrast, bears outside this area (mostly dispersing individuals) experienced higher harvest rates and
consequently bear densities there have remained low (Jerina and Adamič 2008; Krofel et al. 2010).
Currently, bears are present in approximately half of the country, although the majority (95%) of
bears are concentrated in 19% of Slovenian territory. The average density of brown bears in most of
the lynx range in Slovenia is estimated at 12 bears/100 km2, with local densities exceeding 40
bears/100 km2 (Jerina et al. 2013).
Today the most important bear management practices are hunting and supplemental feeding. In
Slovenia, 75% of bear mortality is human-caused (Jerina and Krofel 2012) and 20% of the brown bear
population is removed annually through legal hunting (Krofel et al. 2012b). Supplemental feeding in
the central part of the CBPA is intensive, with high-energy supplemental food, especially corn,
available to bears year-round and in high quantities (on average, 12 500 kg/100 km2 annually) at
numerous feeding sites. Supplemental food represents 34% of dietary energy content ingested by
bears in this area (Kavčič et al. 2015). Locally intensive supplemental feeding likely increases carrying
capacity and may result for some of the highest recorded densities and reproduction rates of brown
bears worldwide (Jerina et al. 2013; Kavčič et al. 2015; Reding 2015). It has also been observed that
intensive supplemental feeding affects habitat use of bears in Slovenia (Jerina et al. 2012) and likely
shortens bear denning periods by as much as 20% compared to areas without supplemental feeding;
currently average denning period for bears in Slovenia lasts 75 days (Krofel et al. 2013a).
Eurasian lynx are the largest felid in Europe and along with the grey wolf (Canis lupus), the main
predator of wild ungulates on the continent (Jedrzejewski et al. 2011). In most of Europe, lynx
specialize in hunting European roe deer (Capreolus capreolus), which they typically consume in a
course of several days (Breitenmoser and Breitenmoser-Würsten 2008). Lynx in Slovenia are part of
the Dinaric lynx population, one of the most threatened populations in Europe (Kaczensky et al.
2013; Sindičić et al. 2013). The population is rapidly declining in Slovenia with estimated 15–25

4
residential animals (Kos et al. 2012). In the study area, lynx hunt mainly wild ungulates, which
together represent 88% of biomass consumed (Krofel et al. 2011). Roe deer is the main prey species
(79% of consumed biomass), with edible dormouse (Glis glis) and red deer (Cervus elaphus) as
important alternative prey, each representing approximately 7% of consumed biomass.
2.2 Locating kills and telemetry
We measured lynx predation, lynx prey consumption, and bear movements using telemetry. During
2005-2011, eight lynx (five females and three males) and 33 bears (14 females and 19 males) were
captured and equipped with telemetry collars (five lynx and all bears with GPS-VHF collars and three
lynx with VHF collars) using standard protocols (see Krofel et al. 2013b and Jerina et al. 2012 for
details on capture and immobilization of lynx and bear, respectively). GPS collars were scheduled to
attempt 7-8 GPS fixes per day for lynx and 12-24 fixes per day for bears.
We used snow-tracking and GPS location cluster analysis of lynx telemetry data to locate kill sites
with prey remains of ungulates killed by lynx (see Krofel et al. 2013b for details). At each kill site we
checked for signs of bear presence (footprints, hair, scat, or characteristic signs of consumption – e.g.
large broken bones or crushed skull) or monitored the carcass consumption with the use of
automatic infra-red video cameras with motion detectors (Fig. 1; Krofel et al. 2012a). Only carcasses
of roe deer, the main lynx prey, were included in this study. Kleptoparasitic interaction (i.e. kill being
found by bears) was noted only when bears usurped the kill during the time while lynx were still
feeding on them. Lynx in the study area fed on roe deer for 4.4 days on average if kills were not
usurped by bears (Krofel et al. 2012a). We typically visited the kill sites the day after lynx abandoned
the kill site, but on some occasions (n = 13) we arrived earlier to install the video system at the kill
site (median time of visit: 4.5 days after the kill was made). When a kill site was too old to reliably
asses it, these data was not included in the analysis.
Figure 1: Still photographs from a video showing a female Eurasian lynx feeding on a roe deer she
killed (A) and a brown bear usurping the kill (B).

5
2.3 Analysing effects of bear densities, movement rates and supplemental feeding sites
For each lynx kill site we determined the local bear density. We used raster map of local bear
population densities in Slovenia with 1 km2 resolution, which was produced using voting
classifications method based on GPS telemetry data, records of bear removals, systematic and
opportunistic direct observations and signs of bear presence, and non-invasive genetic samples
(Jerina et al. 2013). Data for estimating bear densities was obtained in the same period as lynx kill
site monitoring. Since precise data on local bear densities were available only for Slovenia, we
excluded kill sites located in neighbouring Croatia from the analysis.
Frequency of the lynx-bear kleptoparasitic interactions changes seasonally and is strongly correlated
(r=0.89) with the bear daily movement rate (Krofel et al. 2012a). We used bear telemetry data to
calculate average daily movements (i.e. sum of linear distances between consecutive GPS locations;
Jerina et al. 2012) for each month of the year. We attributed the corresponding bear movement rate
to each lynx kill site according to the month when the predation event occurred.
Based on local bear densities and month-specific movement rates we also created a new variable,
index of total path walked daily by all bears around given kill site in given month (total bear path
length), which represents an interaction (product) of both variables. This interaction (product) could
be understood as a proxy for probability of a kill being randomly found by bears and could be
biologically meaningfully interpreted already without the main effects of both variables. Thus we
used it in the models also without the main effects of variables.
To analyse effects of supplemental feeding on the kleptoparasitic interactions, we measured distance
from each lynx kill site to the nearest bear feeding site. Because effects of feeding sites on bear space
use are markedly non-linear (close to feeding sites the space use of bears steeply decreases with
distance to the feeding site, but at greater distances effects are not detected anymore; Jerina et al.
2012), we categorized this variable into three classes (<500 m, 500-1000 m, and >1000 m from the
feeding site) and thus include it in the analysis as a factor.
Bear finding a lynx kill was regarded as a binary event (i.e. bear either finds the remains or not) and
we used generalized linear mixed models (GLMM; binomial error and a logit link function) with bear
finding the lynx kill as a dependent variable, local bear density, monthly bear movement rate, and
total bear path length as independent covariates, and distance to the closest bear feeding site as a
factor. In addition, we included lynx ID as a random factor in all GLMMs. We calculated all possible
models and explored structure of all candidate models with ΔAICc scores ≤2 and used them for
model averaging to obtain robust parameter estimates (Burnham and Anderson 2002). For easier
interpretation of the results, we also produced correlation matrix for the relationships among the
predictor variables and dependent variable (Appendix B) and calculated odd ratios (change in
predicted probability of a lynx kill being found by bears) for changes in values of each independent
variable from the first to the last decile, while values of the other variables remained constant. To
demonstrate relative importance of the results we also calculated probabilities for kill being found by
bears for various combinations of independent variables’ values (for the first and the last deciles), as
well as for different bear management zones.
Supplemental feeding affects density and spatial distribution of bears on different scales. On a large
scale, supplemental feeding likely increases carrying capacity for bears since it represents one of the
main food sources (Kavčič et al. 2015). In addition, it affects bear densities on a local scale, where
preferential habitat use has been observed in the vicinity of feeding sites (Jerina et al. 2012).
However, this may in part be a consequence of local hunters placing feeding sites in more suitable
habitats for bears, where bear densities would be high regardless of supplemental feeding. To

6
account for this, we used a more conservative approach to analyse effects of feeding sites. We first
produced weighted averaged GLMM in a similar manner as described above, but without including
the variable “distance to the feeding site” (conservative GLMM; Appendix A). Thus all explained
variance connected with the bear densities, including variance potentially resulting from hunters
placing feeding sites in more suitable habitats for bears (which might otherwise be attributed to the
effect of supplemental feeding), was allocated to the variable “local bear density”. Next, we
calculated predicted probabilities of kleptoparasitic event for each lynx kill site from the conservative
GLMM and subtracted them from observed values (whether the kleptoparasitic event took place or
not). Thus we obtained residual values from the conservative GLMM, which range from -1 to 1 and
where negative values indicate that actual probability of kleptoparasitism was overestimated and
vice versa. If presence of a feeding site affected the probability of kleptoparasitism, the residual
values should decrease with the distance to the feeding site. Due to non-linear effects of feeding
sites on bear habitat use (see above), we used rank non-parametric correlation to test for
interactions between residual values and distance to the nearest feeding site. We also visually
inspected the residuals by dividing them in five classes (each containing the same sample size) in
respect to the distance to the closest feeding site and for each class calculated average residual
values and CI (for p = 95%).
3. Results
We found 117 lynx kill sites among which 81 were suitable for further analysis. The probability of a
lynx kill being usurped by bears was affected by local bear density, bear movement rates for a given
month, their interaction (total bear path length), and distance to the nearest bear feeding site (Table
1, Figs. 2 and 3). The best model explaining the probability of kleptoparasitism included distance to
the feeding site and total bear path length (Nagelkerke R2 = 0.27). Four additional candidate models
with combinations of local bear density, movement rate, total bear path length, and distance to the
feeding site had ΔAICc scores ≤2 (Table 1). Total bear path length and distance to the feeding site
were included in four out of five models and bear density and movement rate in two models.
Bivariate correlation analyses revealed significant correlations between dependant variable (event of
kleptoparasitism) and all independent variables (rmin=0.229, p < 0.05; Appendix B).
Local bear densities at kill sites ranged from 0.2 to 38.6 bears/100 km2 (mean 16.9 bears/100 km2).
Localities of lynx kills usurped by bears had on average 36% higher bear densities (mean: 21.0, CI:
18.1-23.9, n = 20) compared to lynx kill sites not found by bears (mean: 15.5, CI: 13.3-17.7, n = 61;
Mann-Whitney U = 307.5; p < 0.0001).
Across the combinations of months and bear densities (while keeping the variable supplemental
feeding at fixed value), the predicted probability of kleptoparasitism ranged from 8% (the lowest
decile of bear densities and month with the lowest bear movement rate) to 74% (the highest decile
of bear densities and month with the highest movement rate; Table 1, Average model). Inside the
CBPA (average density 14.0 bears/100 km2) the predicted probability of kleptoparasitism was 2.75-
fold higher compared to the bear distribution range outside this management zone (average density
0.6 bears/100 km2; Table 1, Model 3).
The odds of kleptoparasitism increased 4-times from areas with the lowest to the highest decile of
bear densities (i.e. 8 and 28 bears/100 km2, respectively; Table 1, Model 3), 8.3-times from the
lowest to the highest decile of bear movement rate (1.7 and 8 km/day, respectively; Table 1, Model
3), 10.5-times from the lowest to the highest decile of total bear path length values (Table 1, Average
model) and 5-times from far (>1000 m) to close (<500 m) distance to the nearest bear feeding site
(Table 1, Average model).

7
Very similar results were obtained with a more conservative approach, when distance to the nearest
bear feeding site was analysed separately, based on the residual values from the GLMM model
without distance to the feeding sites (conservative GLMM; Appendix A). Probability of
kleptoparasitism (residual values) decreased with distance from the feeding site (Spearman Rank
Order Correlation r = -0.321, n = 81, p = 0.004), but the effects were detected only until distances
were approximately 1 km from the nearest feeding site (Fig. 4). Effects of bear density, movement
rate and total bear path length remained similar in the conservative GLMM (see Appendix A for exact
values).
Figure 2: Proportion of lynx kills usurped by bears during the time when carcass was still being used
by lynx in relation to the local (1 km2) bear density and average monthly bear movement rate within
the range observed in the Dinaric Mountains in Slovenia.
Figure 3: Proportion of lynx kills usurped by bears during the time when carcass was still being used
by lynx in relation to the local (1 km2) bear density (A), average monthly bear movement rate (B), and
interaction (product) between bear density and movement rate (total bear path length; C). Vertical
bars indicate confidence intervals (p = 0.95), horizontal bars indicate limits of given class (each
containing equal sample size), and lines on top indicate sample distribution in the gradient of
independent variable.

8
Table 1: Parameter estimates and test statistics for the best generalised linear mixed models (ΔAICc
≤2) explaining probability of bear kleptoparasitism on lynx kills. Distance 0-500 m from the nearest
feeding site served as a contrast (estimate = 0) for the remaining levels of that variable. ωi = model
Akaike’s weights; a for change from the first to the last decile of the variable.
Model
Variable
Estimate
SE (β)
Odd ratioa
ΔAICc
ωi
Nagelkerke R2
1
Total bear path length
0.93
0.33
12.0
0
0.36
0.27
Distance to the feeding site
500-1000 m
-1.00
0.86
0.37
>1000 m
-1.57
0.69
0.21
2
Total bear path length
0.88
0.29
10.5
1.3
0.19
0.19
3
Bear movement rate
0.77
0.33
8.3
1.5
0.17
0.28
Bear density
0.56
0.31
4.1
Distance to the feeding site
500-1000 m
-1.05
0.89
0.35
>1000 m
-1.62
0.74
0.20
4
Total bear path length
0.76
0.44
7.6
1.7
0.15
0.28
Bear movement rate
0.25
0.45
2.0
Distance to the feeding site
500-1000 m
-1.07
0.88
0.34
>1000 m
-1.70
0.74
0.18
5
Total bear path length
0.96
0.44
13.0
2.0
0.13
0.27
Bear density
0.05
0.41
1.1
Distance to feeding place
500-1000 m
-1.02
0.87
0.36
>1000 m
-1.60
0.74
0.20
Average
model
Total bear path length
0.88
0.35
10.5
0.26
Bear density
0.56
0.31
4.1
Bear movement rate
0.52
0.47
4.2
Distance to the feeding site
500-1000 m
-1.03
0.87
0.36
>1000 m
-1.61
0.72
0.20

9
Figure 4: Residual values from the generalised linear mixed model explaining probability of bear
kleptoparasitism on lynx kills in relation to distance from the nearest bear feeding site. Vertical bars
indicate standard deviation and horizontal bars limits of each class.
4. Discussion
In a large part of the bear distribution range, bear densities, habitat use, and movement patterns are
under strong influence of management measures (Apps et al. 2004; Gosselin et al. 2015; Kavčič et al.
2015). Because bears regularly interact with other species in the ecosystem, bear management can
induce cascading effects. In Slovenia, management-induced perturbations of the brown bear
population affected the endangered Dinaric population of Eurasian lynx by modulating interactions
between these two keystone carnivores.
The probability of lynx losing its kill to a scavenging bear was related to the local bear density and
bear movement rates. The importance of the interaction between both parameters indicates that
they both act multiplicatively and thus create considerable spatial and seasonal variation in
interaction intensity. In our study area, the predicted probability of lynx kill being lost to bears
ranged from 8 to 74% for combinations of months and lynx distribution range. These results provide
strong support that by affecting bear densities, managers indirectly influence the amount of food
that lynx lose due to bear kleptoparasitism. In Slovenia, bear densities have been strongly regulated
by zone-specific hunting regimes for many decades and about 20% of the population is culled
annually (Krofel et al. 2012b). At the same time, the supplemental feeding in the CBPA zone provides
34% of the total dietary energy content ingested by bears, which is believed to be the reason for one
of the highest observed concentrations and reproductive rates for brown bears worldwide (Kavčič et
al. 2015). Zone-specific bear management thus created remarkably varied conditions for lynx
regarding their interactions with bears. For example, for a lynx living inside the CBPA the predicted
probability of losing kill to a bear are almost 3-fold higher compared to a lynx living in the bear
distribution range outside this management zone.

10
Furthermore, we observed that supplemental feeding of bears modulated bear-lynx interactions
even beyond the effects on local bear densities. When controlling for bear densities at 1 km2 scale,
the presence of bear feeding sites locally increased odds for kleptoparasitism 5-fold. This probably
reflects changes in the use of space by bears induced by supplemental feeding, which has already
been observed in a bear telemetry study (Jerina et al. 2012). The strongest effects of feeding site
presence were detected only up to a 675 m radius (Fig. 4). However, when the high density of these
sites is considered (on average one feeding site per every 2.7 km2), a substantial (45%) part of the
CBPA is thus affected. Therefore, by avoiding the vicinity of bear feeding sites, lynx could
substantially reduce its vulnerability to kleptoparasitism. Further research will be needed to test
whether lynx actually adjust their hunting efforts in respect to the distribution of the bear feeding
sites and local bear densities. Elsewhere, for example, it has been observed that cheetahs (Acinonyx
jubatus) avoid hunting in areas with higher densities of lions, which regularly usurp cheetah kills
(Cooper et al. 2007).
In addition to affecting local bear densities and space use, supplemental feeding could affect lynx-
bear interactions through its impact on bear movement rates, which had a similar importance as
bear density in our study. On one hand, the presence of abundant human-provided food can reduce
the amount of daily activity of bears (Beckmann and Berger 2003), which would decrease the
probability of kleptoparasitism. On the other hand, overall annual movement activity in bears is
strongly affected by the length of the denning period, which can last over 7 months for brown bears
(Manchi and Swenson 2005) and it has been shown that availability of human-provided food reduces
the time period bears spend in a den (Beckmann and Berger 2003). Compared to the neighbouring
region in Italy, where no supplemental feeding is practised, bears in Slovenia were observed to
shorten their denning period by 20% (Kavčič et al. 2015; Krofel et al. 2013a).
Pigeon et al. (2011) showed that climate change caused a shortening of the bear denning period in
Alberta. The strong connection between bear movement activity and interaction intensity observed
in our study thus indicates the possible effect of predicted future climate change on interspecific
interactions among large carnivores. Similarly, since the bear denning period generally increases
towards northern regions (Manchi and Swenson 2005), we expect that potential for kleptoparasitism
decreases with latitude. At the same time, bear densities are typically substantially lower in northern
regions (Jerina et al. 2013). A combination of lower densities and a longer denning period probably
best explains why the frequency of lynx-bear kleptoparasitic interactions in Sweden (Mattisson et al.
2011) is 94% lower compared to our study area.
4.1 Conservation and management implications
Human-caused perturbations of interspecific interactions between Eurasian lynx and brown bears
could have important implications for lynx conservation and management of its prey. Apex predators
are thought to often function close to physiological energetic limits (Gorman et al. 1998; but see
Scantlebury et al. 2014). Thus, additional energetic pressure due to increased prey losses, which can
be substantial in the case of Eurasian lynx, in combination with higher risk of injuries due to
increased hunting rate, could have demographic effects on lynx populations (Krofel et al. 2012a). This
may be especially important for threatened populations, which already suffer from other serious
threats, such as inbreeding and poaching in the case of the Dinaric population (Sindičić et al. 2013).
We suggest that including the effects of kleptoparasitism in conservation actions for Eurasian lynx
populations coexisting with bears where bear densities are high (e.g. Dinaric, Balkan, and Carpathian
lynx populations) could benefit lynx recovery programs. For example, when funds for conservation
are limited, more effort could be focused on areas with lower bear densities (given that there are no
differences in other threats), where there is a better chance of preserving at least part of the

11
predator population. A similar recommendation can be used when planning reintroduction of a
potentially vulnerable carnivore.
In response to kleptoparasitism, lynx in Slovenia compensate losses by increasing their kill rate by
23% (Krofel et al. 2012a). We suggest that wildlife managers should take into account scavenger-
driven cascading effects in predator-prey interactions and appropriately adjust management of prey
species when needed.
Since scavenging is an important natural process, we believe that it would be unwise to attempt to
prevent this interaction (e.g. by radical culling of dominant scavengers), as this would contradict the
general premise of nature conservation, which strives to preserve the ecological integrity of
ecosystems and their processes (Ray et al. 2013; Ripple et al. 2014). Moreover, dominant scavengers
like bears are often protected and threatened themselves. However, we do urge managers and
conservationists to pay attention not to artificially increase local scavenger densities without
considering indirect effects of management measures on apex predators and other species directly
or indirectly affected by dominant scavengers. Several conservation initiatives already led to
overpopulation of some large carnivores, especially when populations were confined to small
reserves (Hayward et al. 2007). Even more common are superabundant scavenger communities due
to human-provided foods, which can create local high concentrations of facultative scavengers
(Cortes-Avizanda et al. 2009; Selva et al. 2014). The observed impact of bear supplementary feeding
on endangered Eurasian lynx population in Slovenia provides another caution against uncritical
promoting of supplementary feeding practices. In the case of Slovenia we recommend that bear
feeding intensity should be reduced, which could be achieved by gradual reduction in the number of
feeding sites or the amount of food provided per site, especially in the season of increased
kleptoparasitic interactions.
Since bears throughout the world are actively managed through hunting, reintroductions, and
supplemental feeding or baiting (Clark et al. 2002; Kaczensky et al. 2013; Kavčič et al. 2013), effects
similar to those observed in our study could be expected also for other predators and scavengers
that co-exist with healthy bear populations, such as cougars in North America, tigers (Panthera tigris)
and leopards (Panthera pardus) in Asia, and wolves throughout the Holarctic. In addition to bears,
other dominant scavengers can also importantly affect apex predators (Cooper 1991; Gorman et al.
1998; Jedrzejewska and Jedrzejewski 1998), indicating a general need for wildlife managers to
broaden their focus from single-species management to community- or ecosystem-focused approach
and include evaluation of potential cascading effects of their management plans into decision-
making processes, especially when managing dominant scavengers, apex predators, and other
strongly interacting species.
Acknowledgements
We would like to thank M. Jonozovič, F. Kljun, A. Marinčič, H. Potočnik, N. Ražen, T. Skrbinšek, and A.
Žagar for their help with the fieldwork. We are also grateful to S.M. Wilson, T.A. Nagel and S.M.J.G.
Steyaert for their valuable input in reviewing the early draft and improving the English. This study
was partly financed by the Slovenian Environmental Agency (projects no. 2523-09-100075 and 2523-
08-100547), the European Union (INTERREG IIIA Neighbourhood Program Slovenia/Hungary/Croatia
2004–2006, project “DinaRis”), the Ministry of Agriculture, Forestry and Food (project V4-0497) and
the Slovenian Research Agency (projects P1-0184 and J4-7362). MK was supported by the research
grant from the Pahernik foundation.

12
References
Allen, M.L., Elbroch, L.M., Wilmers, C.C., Wittmer, H.U., 2014. Trophic facilitation or limitation?
Comparative effects of pumas and black bears on the scavenger community. Plos One 9, e102257.
Apps, C.D., McLellan, B.N., Woods, J.G., Proctor, M.F., 2004. Estimating grizzly bear distribution and
abundance relative to habitat and human influence. Journal of Wildlife Management 68, 138-152.
Atwood, T.C., Gese, E.M., 2008. Coyotes and recolonizing wolves: social rank mediates risk-
conditional behaviour at ungulate carcasses. Animal Behaviour 75, 753-762.
Beckmann, J.P., Berger, J., 2003. Rapid ecological and behavioural changes in carnivores: the
responses of black bears (Ursus americanus) to altered food. Journal of Zoology 261, 207-212.
Begon, M., Townsend, C.R., Harper, J.L., 2006. Ecology: from individuals to ecosystems, 4th edition
edn. Blackwell Publishing Ltd., Oxford, U.K.
Breitenmoser, U., Breitenmoser-Würsten, C., 2008. Der Luchs: Ein Grossraubtier in der
Kulturlandschaft. Salm Verlag, Wohlen and Bern.
Burnham, K.P., Anderson, D.R., 2002. Model selection and multimodel inference: a practical
information-theoretic approach, 2nd edn. Springer, New York.
Carbone, C., DuToit, J.T., Gordon, I.J., 1997. Feeding success in African wild dogs: Does
kleptoparasitism by spotted hyenas influence hunting group size? Journal of Animal Ecology 66,
318-326.
Clark, J.D., Huber, D., Servheen, C., 2002. Bear reintroductions: Lessons and challenges. Ursus 13,
335-345.
Cooper, A.B., Pettorelli, N., Durant, S.M., 2007. Large carnivore menus: factors affecting hunting
decisions by cheetahs in the Serengeti. Animal Behaviour 73, 651-659.
Cooper, S.M., 1991. Optimal hunting group-size: the need for lions to defend their kills against loss to
spotted hyaenas. African Journal of Ecology 29, 130-136.
Cortes-Avizanda, A., Carrete, M., Serrano, D., Donazar, J.A., 2009. Carcasses increase the probability
of predation of ground-nesting birds: a caveat regarding the conservation value of vulture
restaurants. Animal Conservation 12, 85-88.
Creel, S., Spong, G., Creel, N., 2001. Interspecific competition and the population biology of
extinction-prone carnivores, In Carnivore Conservation. eds J.I. Gittleman, S.M. Funk, D.W.
Macdonald, R.K. Wayne, pp. 35-60. Cambridge University Press, Cambridge.
Davidson, Z., Valeix, M., Loveridge, A.J., Madzikanda, H., Macdonald, D.W., 2011. Socio-spatial
behaviour of an African lion population following perturbation by sport hunting. Biological
Conservation 144, 114-121.
Estes, J.A., Terborgh, J., Brashares, J.S., Power, M.E., Berger, J., Bond, W.J., Carpenter, S.R., Essington,
T.E., Holt, R.D., Jackson, J.B.C., Marquis, R.J., Oksanen, L., Oksanen, T., Paine, R.T., Pikitch, E.K.,
Ripple, W.J., Sandin, S.A., Scheffer, M., Schoener, T.W., Shurin, J.B., Sinclair, A.R.E., Soule, M.E.,
Virtanen, R., Wardle, D.A., 2011. Trophic downgrading of planet Earth. Science 333, 301-306.
Gorman, M.L., Mills, M.G., Raath, J.P., Speakman, J.R., 1998. High hunting costs make African wild
dogs vulnerable to kleptoparasitism by hyaenas. Nature 391, 479-481.
Gosselin, J., Zedrosser, A., Swenson, J.E., Pelletier, F., 2015. The relative importance of direct and
indirect effects of hunting mortality on the population dynamics of brown bears. Proceedings of the
Royal Society B-Biological Sciences 282.
Güthlin, D., Knauer, F., Kneib, T., Kuchenhoff, H., Kaczensky, P., Rauer, G., Jonozovič, M., Mustoni, A.,
Jerina, K., 2011. Estimating habitat suitability and potential population size for brown bears in the
Eastern Alps. Biological Conservation 144, 1733-1741.
Hayward, M.W., O'Brien, J., Kerley, G.I.H., 2007. Carrying capacity of large African predators:
Predictions and tests. Biological Conservation 139, 219-229.
Jedrzejewska, B., Jedrzejewski, W., 1998. Predation in vertebrate communities: The Białowieża
Primeval Forest as a case study. Springer, Heidelberg.

13
Jedrzejewski, W., Apollonio, M., Jedrzejewska, B., Kojola, I., 2011. Ungulate-large carnivore
relationships in Europe, In Ungulate management in Europe: problems and practices. Eds R.J.
Putman, M. Apollonio, R. Andersen, pp. 284-318. Cambridge University Press, Cambridge.
Jerina, K., Adamič, M., 2008. Fifty years of brown bear population expansion: effects of sex-biased
dispersal on rate of expansion and population structure. Journal of Mammalogy 89, 1491-1501.
Jerina, K., Jonozovič, M., Krofel, M., Skrbinšek, T., 2013. Range and local population densities of
brown bear Ursus arctos in Slovenia. European Journal of Wildlife Research, 1-9.
Jerina, K., Krofel, M., 2012. Analysis of removed brown bears in Slovenia 2007–2010 on the basis of
age determined by teeth-grinding. Final report. Biotechnical Faculty, University of Ljubljana,
Ljubljana.
Jerina, K., Krofel, M., Stergar, M., Videmšek, U., 2012. Factors affecting brown bear habituation to
humans: a GPS telemetry study. Final report. University of Ljubljana, Biotechnical Faculty, Ljubljana.
Jobin, A., Molinari, P., Breitenmoser, U., 2000. Prey spectrum, prey preference and consumption
rates of Eurasian lynx in the Swiss Jura Mountains. Acta Theriologica 45, 243-252.
Kaczensky, P., Chapron, G., Von Arx, M., Huber, D., Andren, H., Linnel, J.D.C., 2013. Status,
management and distribution of large carnivores – bear, lynx, wolf & wolverine – in Europe. Part 1.
IUCN/SSC Large Carnivore Initiative for Europe.
Kavčič, I., Adamič, M., Kaczensky, P., Krofel, M., Jerina, K., 2013. Supplemental feeding with carrion is
not reducing brown bear depredations on sheep in Slovenia. Ursus, 111-119.
Kavčič, I., Adamič, M., Kaczensky, P., Krofel, M., Kobal, M., Jerina, K., 2015. Fast food bears: brown
bear diet in a human-dominated landscape with intensive supplemental feeding. Wildlife Biology
21, 1-8.
Kos, I., Koren, I., Potočnik, H., Krofel, M., 2012. Status and distribution of Eurasian lynx (Lynx Lynx) in
Slovenia from 2005 to 2009. Acta Biologica Slovenica 55, 49-63.
Krofel, M., Filacorda, S., Jerina, K., 2010. Mating-related movements of male brown bears on the
periphery of an expanding population. Ursus 21, 23-29.
Krofel, M., Groff, C., Špacapan, M., Jerina, K., 2013a. Winter sleep with room service: effects of
supplemental feeding on denning behavior of brown bear (Ursus arctos). In 22nd International
conference on bear research and management, p. 16, Provo, Utah.
Krofel, M., Huber, D., Kos, I., 2011. Diet of Eurasian lynx Lynx lynx in the northern Dinaric Mountains
(Slovenia and Croatia): Importance of edible dormouse Glis glis as alternative prey. Acta
Theriologica 56, 315-322.
Krofel, M., Jonozovič, M., Jerina, K., 2012b. Demography and mortality patterns of removed brown
bears in a heavily exploited population. Ursus 23, 91-103.
Krofel, M., Kos, I., Jerina, K., 2012a. The noble cats and the big bad scavengers: effects of dominant
scavengers on solitary predators. Behavioral Ecology and Sociobiology 66, 1297-1304.
Krofel, M., Skrbinšek, T., Kos, I., 2013b. Use of GPS location clusters analysis to study predation,
feeding, and maternal behavior of the Eurasian lynx. Ecological Research 28, 103-116.
Linnell, J.D.C., Strand, O., 2000. Interference interactions, co-existence and conservation of
mammalian carnivores. Diversity and Distributions 6, 169-176.
Loveridge, A.J., Searle, A.W., Murindagomo, F., Macdonald, D.W., 2007. The impact of sport-hunting
on the population dynamics of an African lion population in a protected area. Biological
Conservation 134, 548-558.
Lozano, J., Casanovas, J.G., Virgos, E., Zorrilla, J.M., 2013. The competitor release effect applied to
carnivore species: how red foxes can increase in numbers when persecuted. Animal Biodiversity
and Conservation 36, 37-46.
Maletzke, B.T., Wielgus, R., Koehler, G.M., Swanson, M., Cooley, H., Alldredge, J.R., 2014. Effects of
hunting on cougar spatial organization. Ecology and Evolution 4, 2178-2185.
Manchi, S., Swenson, J.E., 2005. Denning behaviour of Scandinavian brown bears Ursus arctos.
Wildlife Biology 11, 123-132.

14
Mattisson, J., Andrén, H., Persson, J., Segerström, P., 2011. Influence of intraguild interactions on
resource use by wolverines and Eurasian lynx. Journal of Mammalogy 92, 1321-1330.
Moleón, M., Sánchez-Zapata, J.A., Selva, N., Donázar, J.A., Owen-Smith, N., 2014. Inter-specific
interactions linking predation and scavenging in terrestrial vertebrate assemblages. Biological
Reviews 89, 1042-1054.
Moura, A.E., Tsingarska, E., Dąbrowski, M.J., Czarnomska, S.D., Jędrzejewska, B., Pilot, M., 2014.
Unregulated hunting and genetic recovery from a severe population decline: the cautionary case of
Bulgarian wolves. Conservation Genetics 15, 405-417.
Murphy, K.M., Felzien, G.S., Hornocker, M.G., Ruth, T.K., 1998. Encounter competition between bears
and cougars: some ecological implications. Ursus 10, 55-60.
Newsome, T.M., Dellinger, J.A., Pavey, C.R., Ripple, W.J., Shores, C.R., Wirsing, A.J., Dickman, C.R.,
2015. The ecological effects of providing resource subsidies to predators. Global Ecology and
Biogeography 24, 1-11.
Nielsen, S.E., Herrero, S., Boyce, M.S., Mace, R.D., Benn, B., Gibeau, M.L., Jevons, S., 2004. Modelling
the spatial distribution of human-ccaused grizzly bear mortalities in the Central Rockies ecosystem
of Canada. Biological Conservation 120, 101-113.
Ordiz, A., Bischof, R., Swenson, J.E., 2013. Saving large carnivores, but losing the apex predator?
Biological Conservation 168, 128-133.
Ordiz, A., Stoen, O.G., Saebo, S., Kindberg, J., Delibes, M., Swenson, J.E., 2012. Do bears know they
are being hunted? Biological Conservation 152, 21-28.
Oro, D., Genovart, M., Tavecchia, G., Fowler, M.S., Martínez-Abraín, A., 2013. Ecological and
evolutionary implications of food subsidies from humans. Ecology Letters 16, 1501-1514.
Penteriani, V., Delgado, M.D.M., Melletti, M., 2010. Don't feed the bears! Oryx 44, 169-170.
Périquet, S., Fritz, H., Revilla, E., 2014. The Lion King and the Hyaena Queen: large carnivore
interactions and coexistence. Biological Reviews, n/a-n/a.
Pigeon, K., 2011. Denning behaviour and climate change: linking environmental variables to denning
of grizzly bears in the Rocky Mountains and boreal forest of Alberta, Canada, In Foothills Research
Institute Grizzly bear program. 2010 annual report. Eds G. Stenhouse, K. Graham, pp. 11-25, Hinton,
Alberta.
Ray, J., Redford, K.H., Steneck, R., Berger, J., 2013. Large Carnivores and the Conservation of
Biodiversity. Island Press.
Reding, R., 2015. Effects of diversionary feeding on life history traits of brown bears. Master thesis.
University of Natural Resources and Life Sciences, Vienna
Ripple, W.J., Estes, J.A., Beschta, R.L., Wilmers, C.C., Ritchie, E.G., Hebblewhite, M., Berger, J.,
Elmhagen, B., Letnic, M., Nelson, M.P., Schmitz, O.J., Smith, D.W., Wallach, A.D., Wirsing, A.J., 2014.
Status and ecological effects of the world’s largest carnivores. Science 343.
Rutledge, L.Y., Patterson, B.R., Mills, K.J., Loveless, K.M., Murray, D.L., White, B.N., 2010. Protection
from harvesting restores the natural social structure of eastern wolf packs. Biological Conservation
143, 332-339.
Scantlebury, D.M., Mills, M.G.L., Wilson, R.P., Wilson, J.W., Mills, M.E.J., Durant, S.M., Bennett, N.C.,
Bradford, P., Marks, N.J., Speakman, J.R., 2014. Flexible energetics of cheetah hunting strategies
provide resistance against kleptoparasitism. Science 346, 79-81.
Selva, N., Berezowska-Cnota, T., Elguero-Claramunt, I., 2014. Unforeseen effects of supplementary
feeding: ungulate baiting sites as hotspots for ground-nest predation. Plos One 9, e90740.
Simonič, A., 1994. The legal protection of the brown bear in Slovene territory — Past and present,
and some suggestions for the future. In Rjavi medved v deželah Alpe-Adria: Zbornik posvetovanja.
Ed. M. Adamič, pp. 11-75. Ministrstvo za Kmetijstvo & Gozdarstvo RS and Gozdarski inštitut
Slovenije, Ljubljana.

15
Sindičić, M., Polanc, P., Gomerčić, T., Jelenčič, M., Huber, Đ., Trontelj, P., Skrbinšek, T., 2013. Genetic
data confirm critical status of the reintroduced Dinaric population of Eurasian lynx. Conservation
Genetics 14, 1009-1018.
Stander, P.E., Haden, P.J., Kaqece, Ghau, 1997. The ecology of asociality in Namibian leopards.
Journal of Zoology 242, 343-364.
Steyaert, S.M.J.G., Kindberg, J., Jerina, K., Krofel, M., Stergar, M., Swenson, J.E., Zedrosser, A., 2014.
Behavioral correlates of supplementary feeding of wildlife: Can general conclusions be drawn?
Basic and Applied Ecology 15, 669-676.
Swenson, J.E., Sandegren, F., Soderberg, A., Bjarvall, A., Franzen, R., Wabakken, P., 1997. Infanticide
caused by hunting of male bears. Nature 386, 450-451.
Whitman, K., Starfield, A.M., Quadling, H.S., Packer, C., 2004. Sustainable trophy hunting of African
lions. Nature 428, 175-178.

16
Appendices A and B. Supplementary data
Appendix A: Mind the cat: Conservation management of protected dominant scavenger
indirectly affects an endangered apex predator
Miha Krofel, Klemen Jerina
Table A.1: Parameter estimates and test statistics for the average generalised linear mixed model
explaining probability of bear kleptoparasitism on lynx prey with excluded effects of distance to the
closest bear feeding site (“conservative GLMM”). a for change from the first to the last decile of the
variable.
Model
Variable
Estimate
SE (β)
Odd ratioa
Average
model
Total bear path length
0.83
0.33
9.3
Bear density
0.52
0.41
3.8
Bear movement rate
0.58
0.30
5.1

17
Appendix B: Mind the cat: Conservation management of protected dominant scavenger
indirectly affects an endangered apex predator
Miha Krofel, Klemen Jerina
Values of the continuous variables (bear movement rate, bear density, and total bear path length)
were non-normally distributed, one variable was ordinal (distance to the nearest feeding site) and
one variable was binary (event of kleptoparasitism). To construct correlation matrix we used:
Spearman's rho (for pairs of continuous variables), point-biserial correlation (for pairs of binary and
continuous variables) and Kendall's tau b correlation (for pairs of binary and ordinal variables).
Table B.1: Correlation matrix for the relationships among the dependent variable (event of
kleptoparasitism) and predictor variables. * correlation is significant at the 0.05 level (2-tailed). **
correlation is significant at the 0.01 level (2-tailed).
Bear
movement rate
Bear
density
Total bear path
length
Distance to the
feeding place
Event of
kleptoparasitism
Bear movement rate
1.000
0.225*
0.793**
0.142
0.229*
Bear density
0.225*
1.000
0.707**
-0.130
0.314**
Total bear path length
0.793**
0.707**
1.000
0.025
0.340**
Distance to the feeding place
0.142
-0.130
0.025
1.000
-0.437**
Event of kleptoparasitism
0.229*
0.314**
0.340**
-0.437**
1.000