ChapterPDF Available

Tigers and Wolves in the Russian Far East: Competitive Exclusion, Functional Redundancy and Conservation Implications

CHAPTER 10
Tigers and Wolves in the Russian Far East:
Competitive Exclusion, Functional Redundancy,
and Conservation Implications
Dale G. Miquelle, Philip A. Stephens, Evgeny N. Smirnov,
John M. Goodrich, Olga J. Zaumyslova, and
Alexander E. Myslenkov
Abundant evidence indicates that predators can have profound indirect effects on
many aspects of the diversity of a region through their direct effects on prey
species. It follows that conserving predators is likely to be an essential component
of conserving intact floral and faunal complexes. However, if its role can be ful-
filled by an alternative species, the extinction of a predator need not lead to radi-
cal changes in the biota. In particular, in areas where multiple members of the
predatory guild are present, the loss of one predator may be compensated for by
other members of the guild. Similarly, where one predator has historically ex-
cluded a competitor, extinction of the former may permit the latter to colonize
the area, once again compensating for the loss. This idea of functional redundancy
(e.g., Walker 1992) among the predator guild relies heavily on close similarities in
the effects of predators on their prey. If different species of carnivores are not func-
tionally equivalent (i.e., if they have varying impacts on ungulate populations),
then competitive exclusion, or any replacement of one large carnivore by another,
could have important cascading effects on community structure.
Differences in body size, morphology, life history, and predation behavior
among large carnivores can all potentially lead to differences in their impacts on
prey populations and each other. For instance, comparisons of felid and canid life
history strategies, social structures, and hunting techniques suggest that selection
of prey and impact on prey populations can vary substantially (Kleiman and
179
Eisenberg 1973; Eisenberg 1984; Kunkel et al. 1999; Husseman et al. 2003). At the
same time, within-guild competition among predators, expressed directly through
avoidance, food usurpation, and outright killing, and indirectly by reducing prey
abundance (Palomares and Caro 1999; Creel et al. 2001) can lead to competitive
exclusion of one or more carnivore species by another. Unfortunately, although
there are many comparisons of within-guild food habits of predators, comparisons
of the relative impact of different species of predators on ungulate populations
within an ecosystem (J2drzejewska and J2drzejewski 1998; Sinclair et al. 2003) are
fewer. Similarly, although examples of competitive exclusion among large carni-
vores exist (Fuller and Keith 1981; Creel and Creel 1996; Durant 1998; Tannerfeldt
et al. 2002), explorations of the impact of this phenomenon on prey species are rare.
In this chapter, we present a case study of how two predators, the Amur tiger
(Panthera tigris altaica) and the gray wolf (Canis lupus), interact competitively and
exert top-down pressures in the Sikhote-Alin ecosystem, Russian Far East. Specif-
ically, we use a combination of historical records, more recent ecological data
(from Sikhote-Alin and Bia1owie3a Primeval Forest, a similar, well-studied area in
Poland), and modeling techniques to address two principal questions: (1) What is
the relationship between tigers and wolves in Sikhote-Alin, and is there evidence
of competitive exclusion? and (2) Do these two species have similar impacts on
prey, suggesting some degree of functional redundancy in this system? We pro-
vide evidence that tigers do exclude wolves from this system but that human per-
secution can lead to a reversal of that process. Our analyses suggest the direct
effects of these two carnivores on prey species are different and, consequently, their
roles in influencing ecosystem structure are unlikely to be equivalent.
Study Areas
The southernmost Russian Far East is dominated by the Sikhote-Alin Mountains,
a coastal range that parallels the coast of the Sea of Japan from Vladivostok 1000 km
north to the mouth of the Amur River (Fig. 10.1). The principal forest type of the
original ecosystem is a mixed-composition forest dominated by Korean pine (Pinus
koraiensis) and a variety of broadleaf species. In disturbed areas, Mongolian oak
(Quercus mongolica) is dominant. As with plant communities, the faunal complex
180 From Largely Intact to Human-Dominated Systems
is a mixture of Asian, Himalayan, and boreal species. The ungulate complex is rep-
resented by seven species: red deer (Cervus elaphus), wild boar (Sus scrofa), sika deer
(Cervus nippon), roe deer (Capreolus capreolus), Manchurian moose (Alces alces
cameloidus), musk deer (Moschus moschiferus), and ghoral (Nemorhaedus caudatus).
In addition to tigers and wolves, Far Eastern leopards (Panthera pardus orientalis),
lynx (Felis lynx), brown bears (Ursus arctos), and Himalayan black bears (Ursus
thibetanus) occur in the region. Of these, tigers and wolves are the only large
Tigers and Wolves in the Russian Far East 181
0 100 200 300 400 Kilometer s
N
EW
S
Tiger distrib ution based
on 1996 survey
SeaofJ
apan
A
m
u
r
R
i
v
e
r
U
s
s
u
r
i
R
i
v
e
r
Lake
Khanka
EastMan
c
h
u
ria
n
Mountai
n
s
Khabarovski
Krai
Primorski
Krai
People's
Republic
of China
S
i
k
h
o
t
e
-
A
l
i
n
M
o
u
n
t
a
i
n
s
Figure 10.1
Primary features of the Southern Russian Far East, and tiger distribution based on a
1996 survey (Matyushkin et al. 1999).
carnivores that are widely distributed and depend upon the larger ungulates. Leop-
ards are restricted to the East Manchurian Mountains, bears have not been shown
to be important predators of ungulates, and lynx focus on roe deer and occasion-
ally red deer calves (Okarma et al. 1997; Goodrich et al., unpubl. data).
Sikhote-Alin Zapovednik (SAZ), an IUCN Category I protected area, is situ-
ated in the central Sikhote-Alin Mountains, and extends from the coast of the Sea
of Japan across the divide to the western side of the range (see Fig. 10.1). When
established in 1935 it covered 11,570 km2, but its size has varied dramatically.
Reaching a nadir of 990 km2in 1951, it was restored to 3100 km2in 1960, and is
presently 4000 km2. Red deer are the most abundant ungulate in SAZ followed
by roe deer (Stephens et al., in prep.) Wild boar numbers fluctuate widely, but the
species is generally common.
Where necessary (see following), we compare information on prey selection
in SAZ primarily to Bia1owie3a Primeval Forest (BPF) in Poland, where prey com-
position is similar to SAZ, where wolves are the dominant large predator, and
where an extensive database of both predators and prey exists (J2drzejewska and
J2drzejewski 1998, this volume; J2drzejewski et al. 2002). BPF retains essentially
the same ungulate complex as SAZ, with red deer, roe deer, and wild boar the
dominant species.
Data Analysis and Modeling Methods
To assess the relationship between tigers and wolves, and their impact on prey, we
reviewed long-term monitoring data from SAZ and across the Sikhote-Alin Moun-
tains and then modeled predation by the two species using data from both SAZ
and BPF. We outline these approaches below.
Relationship between Tigers and Wolves in the Sikhote-Alin Ecosystem
We derived the relative abundance of tigers and wolves in the Sikhote-Alin ecosys-
tem from anecdotal and historical accounts of their distributions and abundances,
as well as from data derived from the archives of the Primorski Krai Department
of Hunting Management and “Chronicles of Nature” from SAZ. Track counts (in-
182 From Largely Intact to Human-Dominated Systems
cluding expert assessment of tracks to derive absolute estimates) on standardized
survey routes in winter in SAZ provide a basis for assessing changes in abundance
over time (Smirnov and Miquelle 1999). Biases may exist in such data but we as-
sume that these remain relatively constant from year to year, allowing an assess-
ment of changes in relative abundance over time.
Impacts of Wolves and Tigers on Prey
The degree of similarity between the impacts of these two predators on prey de-
pends on what they eat and the extent to which they limit prey populations. The
first of these can be determined by looking at empirical data on diet breadth and
prey selection. Unfortunately, there are no comparable estimates of the number
of prey taken by the two different predators in SAZ. Consequently, we used data
from the literature to parameterize models of prey removal by these species.
Diet Breadth and Prey Selection of Tigers and Wolves
We compared data on food habits from SAZ for two periods: 1962 to 1972 when
both tigers and wolves were present, using data from Gromov and Matyushkin
(1974); and 1992 to 2002 when wolves were absent and tiger numbers were con-
siderably higher (Smirnov and Miquelle 1999). To supplement these data, we also
used information on wolf diets from BPF for the period 1986 to 1996 (J2drzejew-
ska and J2drzejewski 1998).
In SAZ, data on prey selection from 1992 to 2002 were obtained by locating
kills made by both radiocollared (Goodrich et al. 2001) and uncollared tigers
(Miquelle et al. 1996). We combined data for both collared and uncollared tigers
after finding no significant variation in the ratio of prey species found (Miquelle
et al. 1996). Kills were identified to species and, where possible, categorized into
sex–age classes as adult males, adult females, yearlings, or young of the year.
We compared diet selection using Shannons diversity (H) and equitability (E)
indices. We also estimated Horn’s index of diet overlap between: (a) tigers and
wolves in SAZ (1962–1972); and (b) tigers in SAZ (1992–2002) and wolves in BPF
(1986–1996 (all formulae available in Krebs 1989). Indices for wolves and tigers in
SAZ from 1962 to 1972 were based on number of kills, but for tigers in 1992 to
2002 we converted data to biomass, using weights from Bromley and Kucherenko
Tigers and Wolves in the Russian Far East 183
(1983), Danilkin (1999), and J2drzejewska and J2drzejewski (1998) to derive com-
parable dietary diversity indices for tigers in SAZ and wolves in BPF (% biomass
in Table 4.8, autumn–winter diets in western part, J2drzejewska and J2drzejew-
ski 1998). For wolves in BPF we used data only for wild prey items and allocated
“undetermined deer” in J2drzejewska and J2drzejewski (1998) to proportions of
red deer and roe deer in wolf diets in western BPF (1984–1984) (as reported in Fig.
4.10 of J2drzejewska and J2drzejewski 1998).
Estimates of relative abundance of red deer and wild boar from both study
periods in SAZ were determined from winter counts of fresh tracks (less than 24
hours) adjusted by the relative daily travel distance of each ungulate species
(Stephens et al. in prep.) found along permanent routes within SAZ (Stephens et
al. in prep.). Population composition of red deer and wild boar (the two most com-
mon prey) was determined by trained observers (scientists and forest guards)
recording sex–age composition of all observed groups and individuals on a year-
round basis. We considered data on predation and population structure for the
winter period only (November–April). We used a multinomial test to compare se-
lection by tigers and wolves for the three most important prey species (red deer,
wild boar, and roe deer) with estimates of relative prey abundance as the expected
ratios, and used chi-square analyses to compare kill selection for each sex–age class
within a species (red deer and wild boar) to the proportion of that class found in
the population.
Following Karanth and Sunquist (1995) and Kunkel et al. (1999), we used
Chesson’s (1978) index of selectivity (also know as Manly’s alpha, Krebs 1989), to
compare dietary preferences of wolves and tigers. We compared dietary responses
of tigers to changes in red deer density by using relative abundance estimates of
red deer and tiger kill composition (Miquelle et al. 1996), both averaged over ap-
proximately five-year intervals from 1962 to 1999.
Tiger and Wolf Predation on Prey Populations
The impact of predators on prey is dependent on three factors: (1) the density and
productivity of the prey population, (2) the amount that each predator kills, and
(3) the density at which the predators occur. Neither the productivity of prey nor
the daily requirements of predators are known for wolves and tigers in SAZ; how-
ever, we were able to derive estimates of both from the literature on conspecifics
184 From Largely Intact to Human-Dominated Systems
in similar systems. Productivity of the prey population varies with population den-
sity (because this affects population growth) and also depends on body mass. Red
deer (and elk) population dynamics are well studied (Clutton-Brock et al. 1982;
Houston 1982) and these represent the most important prey species for tigers and
wolves in SAZ. Consequently, we limited our assessments of predator impacts
on prey to this prey species. Our conclusions, however, should be qualitatively sim-
ilar for a multiprey system. The dynamics of the red deer population were
assumed to be of a ramped density-dependent form (e.g., Fowler 1987; McCul-
lough 1992), with a simple, linear decline in population growth rate only above a
threshold at 0.6 K(where Kis the environmental carrying capacity in the absence
of predation). Below this threshold, mean population growth, r, was assumed to
be constant. Biomass production was estimated by assuming that the average adult
weighed approximately 180 kg (this assumes a male to female ratio of 0.66, with
mean adult masses of 149 kg for females and 224 kg for males, Bromley and
Kucherenko 1983). Where necessary, we estimated standing biomass of prey
assuming that an adult female would represent an individual of approximately
average mass.
The density at which predators occur (the third factor underlying the impact
of predators and prey) is complex, relying on prey density, predator population
growth rates, and, potentially, predator social structure. Prey density is itself a
product of the extent of predation, leading to potentially circular logic. Further-
more, flexibility in social structure (for example, the rate at which territory size
changes) is difficult to quantify and may also be a function of prey density. As a re-
sult of these complications, we used two modeling approaches to assess the poten-
tial impact of predators on prey: an energy balance model and a simulation model.
Energy Balance Model The energy balance model combined empirical rela-
tionships between predator and prey densities (hereafter, “numerical responses”),
with estimates of prey productivity (see earlier) and predator kill rates. From these
it is possible to estimate, for any given prey carrying capacity, the prey density at
which prey productivity and prey consumption by predators are in balance. This,
in turn, provides an estimate of the proportion by which the prey population will
be depressed below its potential carrying capacity.
Predator–Prey Simulation Model We generated simple matrix models to describe
the dynamics of predators and prey. Productivity of the prey (red deer) was based
Tigers and Wolves in the Russian Far East 185
on a stochastic version of the dynamics already described. Parameters used for
tigers and wolves are summarized in Appendix 10.1. Predator dynamics were
linked to prey availability through energetic constraints on survival and repro-
duction. Because estimation of functional responses is problematic (Marshal and
Boutin 1999), energetic constraints were modeled using a simple depletion ap-
proach (Sutherland 1996). Specifically, we assumed that during each time step,
predators could remove all required prey (red deer) from the environment down
to some critical threshold density (below which predation is no longer energeti-
cally viable). Predator consumption rates were the same as those used for the equi-
librium model. Predators that could not obtain their requirements during any time
step were assumed to die or disperse. Social group sizes, reproductive behavior,
dispersal behavior, and presence of transient animals were all modeled on the basis
of empirical data (e.g., Mech 1974; Hayes and Harestad 2000a; Sunquist and Sun-
quist 2002; Kerley et al. 2003; Goodrich, unpubl. data).
A key factor determining the ability of predators to respond to changes in
prey availability and, thus, to fully exploit prey populations, is their flexibility re-
garding territory size. Predator territory size is known to vary with prey densi-
ties across habitats for both wolves (Fuller et al. 2003) and tigers (Miquelle et al.
1999). However, flexibility (i.e., the rate of change) of territory sizes in response
to changes in prey availability within a single area is poorly understood. For the
wolf at least, Fuller (1989) provided evidence that territories may expand, contract,
disappear, or be established in response to changes in prey availability or distribu-
tion, but the rate at which these changes occur is unclear. In the absence of de-
nitive knowledge, we derived territory size from empirical predator–prey
relationships, and retained a stable territory size through simulations, with the as-
sumption that large carnivores will be conservative in adjusting territory size to
changes in prey density.
Research Findings
Analyses of the competitive relationship of tigers and wolves in the Sikhkote-Alin
ecosystem provide necessary background for considering their relative influence
on prey populations in temperate forest ecosystems.
186 From Largely Intact to Human-Dominated Systems
Relationship between Tigers and Wolves in the Sikhote-Alin Ecosystem
Considered common throughout the region in the late 1800s and early 1900s,
Amur tigers were driven to historical lows in the 1940s due to human persecu-
tion (Kaplanov 1948; Kucherenko 2001). With hunting of tigers outlawed in 1947,
recovery of the tiger population continued for approximately 40 years, reaching
an apparent peak in the late 1980s and early 1990s.
Distribution and abundance of wolves follows an inverse pattern to tigers.
Wolves were absent or exceedingly rare in the southern Russian Far East at the
end of the 19th and beginning of the 20th centuries. Abramov (1940) believed that
wolves appeared in the Sikhote-Alin only after the beginning of the 20th century,
coincident with the abrupt range reduction of tigers. Yudin (1992) suggested that
wolves arrived earlier by infiltrating human-dominated regions (the Ussuri Basin
and Lake Khanka regions) where tigers were largely eliminated during coloniza-
tion by Russians in the late 1800s and early 1900s. In the broken forests and mead-
ows of these regions, wolves survived on a combination of roe deer and domestic
livestock (Yudin 1992). When tiger numbers dropped in the Sikhote-Alin and East
Manchurian Mountains in the 1930s and 1940s, this peripheral population acted as
a source for expansion and colonization. Wolf populations across the region de-
creased coincident with recovery of the tiger population since the 1940s. Today
wolves are rare across the range of tigers, being found in scattered pockets, and
usually as solo individuals or small groups.
Numerous anecdotal accounts suggest an inverse correlation between wolf
and tiger numbers. For instance, in the absence of tigers, wolves apparently sur-
vived (at least intermittently) in relatively large numbers in the Pogranichniy Raion
(district) of the Lake Khanka region (Yudin 1992) but disappeared from this frag-
ment of habitat with the arrival of tigers (Matyushkin et al. 1999). Tigers have
since disappeared again from this fragment, with reports of wolves returning
(Pikunov, pers. comm.).{AQ: pls give date of pers. comm.}
Accounts of historical shifts in the abundance of tigers and wolves are espe-
cially well documented in the Zapovedniks, where long-term monitoring has been
conducted. For instance, in Lazovski Zapovednik, Bromley (1953) reported that,
although wolves were formerly absent, 105 wolves “had to be destroyed” in
the 1940s, coincident with the low density of tigers. Wolf numbers declined
Tigers and Wolves in the Russian Far East 187
consistently from the 1960s through the 1980s, at the same time as tiger popula-
tions were recovering in the reserve. Wolf tracks were rarely observed in the 1990s,
and no tracks were registered in 1992 and 1993, whereas tiger numbers were high
and stable (Khramtsov 1995).
The population dynamics of tigers and wolves are best documented in SAZ.
Bromley (1953) noted that elder native people on the eastern slopes of the Sikhote-
Alin Mountains (including Terney Raion, where the Zapovednik is located) had
no recollection of wolves occurring in that region prior to the 1930s (Abramov
1940), coincident with a depleted tiger population (Kaplanov 1948). Although
rangewide tiger numbers began increasing in the 1950s, tigers were still virtually
absent in SAZ in the early 1960s (Matyushkin et al. 1981; Smirnov and Miquelle
1999), whereas wolves remained common despite efforts to control their numbers
even within the Zapovednik (Gromov and Matyushkin 1974) (Fig. 10.2). Restora-
tion of reserve size and better protection led to recolonization by tigers in 1963,
and recovery of the tiger population through the mid-1990s (Fig. 10.2) (Matyush-
kin et al. 1981; Smirnov and Miquelle 1999). Based on fieldwork in SAZ through
the early 1970s, Gromov and Matyushkin (1974) argued against the perception,
common in Russia, that wolves are driven to low densities or extinction in the pres-
ence of tigers. In retrospect, it is clear that their observations were made during
a period of disequilibrium, when both species coexisted in SAZ in moderate num-
bers. As tiger numbers continued to increase, records of wolves in the Zapoved-
nik decreased and became rare (see Fig. 10.2).
Despite the clear inverse correlation between wolf and tiger numbers, the
mechanism driving population declines of wolves is unclear. Gromov and Matyush-
kin (1974) reported both usurpation of wolf kills by tigers and scavenging of tiger
kills by wolves. The former has probably not been a primary factor influencing wolf
distribution. Although usurpation has been documented in cougar (Puma con-
color)–wolf systems (Murphy 1998; Kunkel et al. 1999), it is not as common in these
forest habitats as in open savannas (Fanshawe and Fitzgibbon 1993; Caro 1994; Creel
2001). Gromov and Matyushkin (1974) believed that tigers did not prey directly on
wolves, and along with others (Yudin 1992) proposed that wolves actively avoid
areas used by tigers, resulting in spatial separation of the two species as tiger abun-
dance increases, with wolves remaining only in peripheral areas. However, simple
displacement and avoidance seem unlikely to explain the dramatic decrease in wolf
numbers across such large areas. Although there are only four records of a tiger
188 From Largely Intact to Human-Dominated Systems
killing a wolf (Miquelle et al. 1996; Makovkin 1999), Amur tigers are notorious for
killing dogs (Makarov and Tag irova 1989; Miquelle et al. in press); official records in-
dicate 104 dogs killed by tigers in or near SAZ (where dogs are illegal) between 1957
and 2002. Tiger predation on another canid, the dhole (Cuon alpinus), has also been
reported on several occasions (Venkataraman 1995; Karanth and Sunquist 2000). Al-
though rarely observed, direct killing of one predator by another is suspected to
play an important role in limiting many predator species (Palomares and Caro 1999;
Woodroffe and Ginsberg, this volume). Thus, despite lack of clear evidence, we pro-
pose that direct killing of wolves by tigers has likely been an important element in
reducing wolves to a functionally insignificant role in the Sikhote-Alin ecosystem.
Impacts of Wolves and Tigers on Prey
Selection of prey species by tigers and wolves appears surprisingly similar both within
and across study sites, but selection for sex–age classes varies. Our models also sug-
gest that the impact of tiger and wolf predation on prey populations will also vary.
Tigers and Wolves in the Russian Far East 189
Tiger abundance
Wolf track records
0
5
10
15
20
25
30
35
1957
1960
1963
1966
1969
1972
1975
1978
1981
1984
1987
1990
1993
1996
1999
2002
0
5
10
15
20
25
30
35
Year
Figure 10.2
Estimates of wolf abundance (------) (based on the total number of tracks
reported/year) and tiger abundance (—) (derived from an expert assessment based
on number and distribution of tiger tracks/year) in SAZ 1957–2002. Data taken
from the SAZ’s “Chronicles of Nature” database.
Diet Breadth and Prey Selection of Tigers and Wolves
We located 389 remains of 15 species of wild mammals and birds killed by tigers
in SAZ between 1992 and 2002. During a similar span (1986–1996) in the western
part of BPF, J2drzejewska and J2drzejewski (1998) reported 15 wild species (and
plant material) taken by wolves, based on scats and kills. During a shorter time
frame with smaller sample sizes, Gromov and Matyushkin (1974) reported that
co-occurring tigers and wolves used five and six species, respectively, in SAZ (Table
10.1). Diversity of tiger and wolf diets appeared similar based on comparisons of
data from Gromov and Matyushkin (1974) and, again, when based on more recent
data on tigers in SAZ and wolves in BPF (Table 10.2). Shannon’s diversity and eq-
uitability indices put greater emphasis on the diversity of more common species
and, hence, large numbers of species taken very infrequently do little to suggest
increased diversity. Indeed, the highest diversity score was calculated for wolves
during 1962 to 1972 (see Table 10.2), in spite of the narrower range of species
recorded (see Table 10.1). Available evidence thus suggests that diet breadth is sim-
ilar for these two predators.
Overlap in species preyed upon by wolves and tigers was high (see Tables 10.1,
10.2). In all the studies, red deer, wild boar, and roe deer constituted 79 to 97% {AQ:
79 to 90% meant?} of kills made, with both tigers and wolves relying on
red deer as their primary prey (57–65%) in both SAZ and BPF (see Table 10.1). Wild
boar represented the majority of the remaining prey taken by tigers, but wolves
relied more on roe deer (BPF) or musk deer (SAZ) than wild boar (see Table 10.1).
Despite a reliance on red deer as their major prey, important differences in diets
of tigers and wolves existed. Tigers showed prey selectivity in both study periods
(1962–1972: χ2= 51.52, df = 2, P < 0.001; 1992–2002: χ2= 563.4, df = 2, P < 0.001),
taking red deer in proportion to their relative abundance but showing a strong se-
lection for wild boar and against roe deer (Fig. 10.3a,b; see Table 10.2). This pattern
was consistent across both time periods, irrespective of whether wolves were pres-
ent. For the short time that data on both species were collected in SAZ, wolves ap-
peared to take a slightly greater percentage of red deer than were available, took a
smaller percentage of wild boar than were available, and clearly selected against roe
deer (see Fig. 10.3a, Table 10.2). Wolf predation in BPF (J2drzejewska and J2drze-
jewski 1998; J2drzejewski et al. 2002) showed similar patterns to the limited data
that exist in SAZ, but with a stronger preference for red deer, and greater avoidance
of wild boar (Fig. 10.3c). Wolves in SAZ used a surprisingly high percentage of
190 From Largely Intact to Human-Dominated Systems
Ed.:
Probability
variable (“P”)
shifts between
upper -and
lowercase,
throughout:
Standardize??
Table 10.1
Food habits of coexisting tigers and wolves in Sikhote-Alin Zapovednik, Russian Far
East, 1962–1972, for tigers in Sikhote-Alin 1992–2002, and for wolves in an ecosystem
with a similar ungulate complex in Bia1owie3a Primeval Forest, Poland
% Occurrence % Biomass
Sikhote-Alin, Sikhote-Alin, Bia1owie3a
1962–197211992–2002 1986–19962
Tigers Wolves Tigers Wolves
Prey Species (n3= 40) (n3= 77) (n3= 389) (n= 528.2 kg)
Red deer 57.5 64.9 63.9 65.4
Wild boar 27.5 7.8 24.0 12.8
Roe deer 5 6.4 0.9 19.2
Sika deer 4.9
Musk deer 14.4 0.04
Bison 1.8
Goral 0.4
Brown bear 7.541.342.4
Asiatic black bear 1.8
Harbor seal 0.6
Moose 2.5 5.2 0.6
Badger 0.4
Raccoon dog 0.03 0.2
Red fox 0.02
Hedgehog 0.2
Brown hare 0.1
Red squirrel 0.015
Bank vole 0.015
Vole (undet. species) 0.015
White-tailed sea eagle 0.01
Ural owl 0.001
Great spotted
woodpecker 0.015
Nuthatch 0.015
Tawny owl 0.015
Reptile (undet. species) 0.015
Amphibian (undet. species) 0.015
Plant material 0.10
Total 100 100 100 100
1From Gromov and Matyushkin 1974.
2From J2drzejewska and J2drzejewski 1998.
3n= number of kills or scats
4Brown and Asiatic bears combined
5Recorded as present by J2drzejewska and J2drzejewski 1998. Minimum value is presented
here to estimate biomass.
musk deer, which are most common in higher-altitude, coniferous forests, a habi-
tat that tigers rarely use, according to our more recent studies (Miquelle et al. 1999).
J2drzejewska and J2drzejewski (1998) suggested that the density of red deer
was the primary factor driving wolf selection of ungulate prey, with proportion
of red deer in the diet increasing with density (Fig. 10.4d). Our data suggest no
such pattern for tigers: the contribution of red deer did not change (Fig. 10.4a),
and percentage of wild boar and roe deer in the diet did not decrease with in-
creasing densities of red deer (Fig. 10.4b,c), as was apparently the case in BPF (Fig.
10.4e,f ) (J2drzejewska and J2drzejewski 1998: 203). Interpretation of data in SAZ
is confounded by increasing densities of a number of prey species over time
(Stephens et al. in prep.). However, the fact that red deer abundance does not ap-
pear to drive prey selection may also be due to the fact that tigers, in contrast to
wolves, show a strong preference for wild boar, a species that fluctuates greatly
in abundance and might consequently confound any potential relationships.
Tiger selection of red deer sex–age classes closely mirrored herd composition
of red deer, averaged over the study period from 1992 to 2002 (χ2= 1.92, df = 2,
192 From Largely Intact to Human-Dominated Systems
Table 10.2
Indices of dietary diversity (Shannon index H), equitability (E), diet overlap (Horn’s
index), and preference for prey (Chesson’s or Manly’s alpha) for tigers and wolves in
Sikhote-Alin Zapovednik, Russian Far East, and Bia1owie3a Primeval Forest, Poland
Wolves
Tigers Wolves Tigers Bia1owie3a
Sikhote-Alin Sikhote-Alin Sikhote-Alin Forest
1962–1972 1962–1972 1992–2002 1986–1996a
n(# kills or scats) (40) (77) (389) (344)
Dietary diversity, H1.60 1.65 1.58 1.40
Dietary equitability 0.60 0.64 0.40 0.35
Diet overlapb0.89 0.89 0.86 0.86
Preferences (alpha)
Red deer 0.19 0.46 0.15 0.62
Wild boar 0.79 0.48 0.83 0.22
Roe deer 0.02 0.05 0.02 0.16
aData from J2drzejewski and J2drzejewska 1998.
bOverlap comparisons between (1) wolves and tigers in SAZ, 1962–1972, and (2) tigers in
SAZ 1992–2002, and wolves in BPF, 1984–1994.
0
10
20
30
40
50
60
70
Red deer Wild boa
r
Roe dee
Musk dee
r
Moose
0
10
20
30
40
50
60
Red deer Wild boa
r
Roe dee
Musk dee
r
Moose
0
10
20
30
40
50
60
70
80
Red deer Wild boa
r
Roe deer Moose
Proportion/ %
(a)
(b)
(c)
Figure 10.3
Relative abundance of ungulate species ( ), and selection of those species by tigers
( ) and wolves ( ): (a) tigers and wolves in Sikhote-Alin Zapovednik (1962–1972);
(b) tigers in Sikhote-Alin Zapovednik (1992–2002); and (c) wolves in the western part
of Bia1owie3a Primeval Forest, 1984–1994.
p < 0.382) (Fig. 10.5a). We also found no evidence that tigers preyed selectively
on sex–age classes of wild boar across all years combined (χ2= 2.31, df = 3, p=
0.5) (Fig. 10.5b). Sex–age composition of the wild boar population fluctuated more
dramatically than that of red deer (see Fig. 10.5b), and though there may have been
more subtle within-year changes in selection of wild boar sex–age classes, sample
sizes of boar kills limit meaningful yearly comparisons.
194 From Largely Intact to Human-Dominated Systems
0
20
40
60
80
100
45 50 55 60 65
0
10
20
30
40
50
60
45 50 55 60 65
0
2
4
6
8
10
45 50 55 60 65
0
20
40
60
80
100
0123456
0
10
20
30
40
50
60
0123456
0
10
20
30
40
50
60
0123456
(a)
(b)
(d)
(e)
(c) (f)
Relative track density of red deer Red deer density/km
2
% roe deer in kills % wild boar in kills % red deer in kills
Figure 10.4
Diet selection responses to changes in densities of red deer. All trend lines were fitted
by least squares regression and are linear, except for (d), which was best described by
a logarithmic function. The R2values are given in parentheses following here, to give
an indication of the strength of the relationships. Selection by tigers in SAZ, in rela-
tion to red deer abundance, for (a) red deer (R2= 0.13), (b) wild boar (R2= 0.10), and
(c) roe deer (R2= 0.23); selection by wolves in Bia1owie3a Primeval Forest, in relation
to red deer density, for (d) red deer (R2= 0.99), (e) wild boar (R2= 0.89), and (f ) roe
deer (R2= 0.91) (d–f reproduced from J2drzejewska and J2drzejewski 1998).
In existing studies of wolf predation on red deer or elk, wolves preferentially
fed on calves (Okarma et al. 1995; Mech et al. 2001; Husseman et al. 2003), but
selection of adult males and females apparently varied over years in association
with vulnerability (Mech et al. 2001).
Tiger and Wolf Predation on Prey Populations
Parameters underlying the two modeling approaches were largely derived from
the literature. Demographic parameters for red deer (e.g., Clutton-Brock et al.
1982; Houston 1982) suggested an approximate mean population growth rate of
r= 0.3 in the absence of density constraints. Above 0.6 K(where density
Tigers and Wolves in the Russian Far East 195
0
10
20
30
40
50
Males Females Yearlings Young
0
10
20
30
40
50
60
70
Males Females Calves
(a)
(b)
Proportion
Age class
Figure 10.5
Sex–age composition ( ) and tiger kill composition ( ) of (a) red deer and (b) wild
boar populations, SAZ, 1992–2002. Error bars show 95% confidence intervals.
constraints begin to act) this was assumed to decline linearly (to r= 0 at K). Esti-
mates of kill rates were 5.1 kg wolf–1d–1 (based on live prey eaten, J2drzejewski et
al. 2002) and approximately 8 kg tiger–1d*1 (based on a food requirement of 5–6 kg
of meat per day, Sunquist et al. 1999). The wolf numerical response has been sub-
jected to considerable scrutiny but, most recently, Eberhardt et al. (2003) have
shown that, within the range of available data, it is well represented by a linear re-
gression through the origin. The gradient of this line (Eberhardt and Peterson
1999) is equivalent to approximately 0.1 wolves km–2 for each 4.9 red deer km–2
(adjusted for the size of red deer in SAZ). For tigers, we took data from 13 sites
to construct the numerical response. This was found to be a Type II response, best
represented by a Michaelis–Menton function of the form T= aP / (b+ P), where
Tis tiger density, Pis prey density, and aand bare constants (see Fig. 10.6). The fit
of this function was very highly significant (R2= 0.484, F12,11 = 10.32, p < 0.001).
These parameters were all used to develop the two modeling approaches.
Energy Balance Model Biomass production and requirement curves were con-
structed for a range of habitat carrying capacities, from 2 to 10 km–2 deer. Biomass
production increases linearly up to 0.6 K, because mean growth rate is constant
below this point. Above 0.6 K, production declines curvilinearly (Fig. 10.6a). Com-
bining the numerical responses with energetic requirements of individual animals
provides energy requirement curves for tiger and wolf populations (Fig. 10.6c).
These results suggest constant prey depletion by wolves, but the curvilinear nu-
merical response of tigers results in a relative reduction in energy offtake with in-
creasing prey carrying capacity. Taken together, the production and requirement
curves allow predictions of predator–prey equilibria (Fig. 10.6d). Due to the linear
nature of both the wolf numerical response and the initial slope of the prey bio-
mass production curve, prey depletion by wolves is predicted to be constant (with
prey populations reduced by slightly over 28% of carrying capacity), irrespective of
the initial carrying capacity of the prey population. By contrast, the Type II nu-
merical response indicated for tigers suggests that the role of prey in limiting tiger
densities declines with increasing prey density. As prey density increases, reduc-
tion of prey by tigers becomes gradually less significant, declining from approxi-
mately 23% of very low prey carrying capacities (2 km–2 deer) to 15% when prey
carrying capacity is 10 km–2.
Predator–Prey Simulation Model Using existing estimates of prey biomass and
196 From Largely Intact to Human-Dominated Systems
tiger density, we derived territory size of tigers based on the assumption that each
territory of an adult resident tigress contains 3.3 tigers (a female, a third of a male,
one to three cubs or a young daughter, and one transient) (Fig. 10.7a). For wolves,
we derived a relationship between prey availability and territory size (Fig. 10.7b),
based on data from Fuller et al. (2003). Using the predicted territory sizes, simu-
lation models of predation suggested that wolves could deplete prey to a greater
extent than tigers. This result was consistent in every scenario of prey availability
and environmental stochasticity (Fig. 10.8). The proportion by which prey popu-
lations were reduced below Kby tigers was typically in the range of 18 to 25% and
Tigers and Wolves in the Russian Far East 197
0
50
100
150
200
250
300
350
024681012
0.00
0.05
0.10
0.15
0.20
0.25
0.30
246810
0
50
100
150
200
250
300
350
400
450
024681012
0.00
0.04
0.08
0.12
0.16
0.20
0 10 20 30 40 50 60
Pro
p
ortional reduction of
p
re
y
Annual demand/k
g
/
km2Annual
p
roduction/k
g
/
km2
Predator densit
y
/km2
Prey carrying capacity/km2
Prey carrying capacity/km2
Prey carrying capacity/km2
Prey density/km2
(a)
(c)
(b)
(d)
Figure 10.6
(a) Production curve for the deer population when K= 10 km–2. (b) Numerical re-
sponse for the tiger ( ) (data from Karanth 1991; Karanth and Nichols 2000;
Miquelle et al. 1999; Schaller 1967; Støen and Wegge 1996; Tamang 1982; Thapar
1986); the solid line shows the fitted regression (see text); the dashed line shows the
predicted response for wolves, using data from Eberhardt and Peterson (1999). (c)
Demand curves for wolf (——) and tiger (-------) populations, predicted from their
numerical responses and food requirements. (d) Prey depletion by wolves (——) and
tigers (-------) as a function of prey carrying capacity, K. Depletion, as used here, indi-
cates the proportion by which the prey population is reduced below its carrying ca-
pacity in the absence of predation.
never exceeded 30%. Tiger impact on prey decreased with higher prey density due
to constraints on tiger density imposed by territoriality (see Fig. 10.7a), except
when stochasticity was high. In this case, tigers were better able to limit prey when
initial prey abundance was high, because this reduced the possibility of prey be-
198 From Largely Intact to Human-Dominated Systems
0
100
200
300
400
500
600
0 1000 2000 3000 4000 5000 6000 7000 8000
0
200
400
600
800
1000
1200
1400
1600
1800
0 2000 4000 6000 8000 10000 12000 14000
Ungulate biomass/kg/km2
Ungulate biomass index
Territory size/km2
Figure 10.7
Relationships between prey availability and predator territory size.
(a) Tiger territory size, S(assuming 3.3 individuals per territory), and prey biomass,
B; the solid line shows the least-squares regression, S= 22270 B–0.7764, F11 = 42.5, p<
0.001. Mass data sources as for Figure 6 and Sunquist (1981). (b) Wolf territory size,
S, and ungulate biomass index, I; the solid line shows the least squares regression,
S= 134138 I–0.7126, F25 = 17.7, p< 0.001, data from Table 6.3 in Fuller et al. (2003).
0.00
0.20
0.40
0.60
0.80
1.00
2345678910
0.00
0.20
0.40
0.60
0.80
1.00
23456789
10
0.00
0.20
0.40
0.60
0.80
1.00
23456789
10
Proportion by which prey population is reduced below carrying capacity
Initial carrying capacity of prey population/km2
(a)
(b)
(c)
Figure 10.8
Prey depletion (proportion by which prey are reduced below carrying capacity) by
wolves ( ) and tigers ( ) when territory sizes are set according to empirical relation-
ships between prey availability and predator territory size. The amount of environ-
mental stochasticity to which the prey population was subject varied from (a) none,
to (b) moderate (prey population coefficient of variation, CV = 0.15), and (c) high
(CV = 0.35). Rates of depletion are derived from simulations at the territory size
predicted for the given level of prey abundance.
coming so scarce that tigers could not hunt effectively. By contrast, wolves typi-
cally reduced prey density to approximately 50% below carrying capacity, and by
as much as 80% when prey were abundant and stochasticity was low. Wolves also
showed a tendency to deplete prey more when the prey were initially abundant,
because high rates of depletion in low prey availability scenarios led to frequent
crashes in prey numbers.
Competitive Exclusion and Functional Redundancy in
Tigers and Wolves
Mech (1974) suggested that few animals compete successfully with the wolf, but
the tiger appears to be an exception. Wolves do not occur across most of the range
of tigers in Southeast Asia. Whether this negative correlation is causal is unknown,
but evidence from the Russian Far East strongly suggests that tigers depress wolf
numbers either to the point of localized extinction or to such low numbers as to
make them a functionally insignificant component of the ecosystem. Wolves appear
capable of escaping competitive exclusion only when human persecution decreases
tiger numbers. Although there are now many studies documenting the impact of
intraguild competition among large carnivores (Creel et al. 2001), examples of one
large carnivore driving another to localized extinction are relatively rare.
Woodroffe and Ginsberg (this volume) argue that the evidence for intraguild
redundancy is poor but warn that data are extremely restricted. Our analyses sug-
gest that the top-down influences of tigers and wolves (and, therefore, their
broader impacts on biodiversity) are likely to differ substantially. Although diet
breadth was similar, and both tigers and wolves depended on red deer as the pri-
mary prey species, there were important differences in prey selection. J2drzejew-
ska and J2drzejewski (1998) argued that wolves strongly selected for red deer, and
that changes in red deer density determined percentages of other prey that were
taken by wolves. We found no such evidence to support this effect with Amur
tigers. In contrast, tigers demonstrated a very strong preference for wild boar. Vari-
ation in preference, which may be due to the relative vulnerability of boar to am-
bush and cursorial predation, could alter the relative impact of tigers and wolves
on ungulates in a multiprey system. In particular, when boar are common, tiger
200 From Largely Intact to Human-Dominated Systems
predation on red deer might be even lower than our models suggested, while wolf
predation on red deer would remain high.
All available evidence suggests that wolves select for vulnerable age classes of
Cervus elaphus (J2drzejewska and J2drzejewski 1998; Kunkel et al. 1999; Mech et al.
2001; Husseman et al. 2003). In contrast, data from SAZ suggest tigers exhibit vir-
tually no selectivity among sex–age classes of red deer. Our findings are consistent
with observations that solitary, ambush predators are not selective for vulnerable
sex–age classes (Okarma et al. 1997; Kunkel et al. 1999), although such results are
by no means universal (e.g., Karanth and Sunquist 1995).
Although these results reaffirm both theoretical and empirical indications that
cursorial and ambush predators will differentially select prey (Kleiman and Eisen-
berg 1973; Husseman et al. 2003), we propose that absolute levels of prey deple-
tion by predators may be even more important in determining differences between
the top-down influences of tigers and wolves within terrestrial ecosystems. Both
of our modeling approaches suggested that, for the majority of conditions,
limitation of prey populations by wolves is likely to be considerably higher than
by tigers. Our equilibrium approach relies partially on the accuracy of the nu-
merical responses underlying it. Due to the fact that data on predator–prey systems
may have been collected at a point in time when predators and prey were not in
equilibrium, caution should be used in interpreting these results (Eberhardt et al.,
2003). Nevertheless, there are three reasons why we have confidence in the gen-
eral differences in the numerical responses of tigers and wolves used for this
model. First, the numerical response used for wolves is the result of very thorough
analyses of the underlying data (Eberhardt and Peterson 1999). Second, several
of the data points underlying the tiger numerical response represent mean pred-
ator and prey densities averaged over a long period, decreasing the potential error
associated with estimates from a single point in time. Third, evidence for density
predictions of the numerical responses is borne out by the analysis of Carbone
and Gittleman (2002), in which wolves fall substantially below the predictions of
numbers per unit prey biomass. This is in agreement with the numerical response
derived by Eberhardt and Peterson (1999), which indicates that 10,000 kg of prey
supports only 48 kg of wolf (approximately half of the biomass predicted by Car-
bone and Gittleman 2002). Indeed, for the same prey biomass, the numerical re-
sponses underlying our equilibrium model predict higher biomasses of tigers than
Tigers and Wolves in the Russian Far East 201
of wolves over the range of prey density values that exist for wolf–prey systems.
This suggests that differences in predator density do not explain our predictions of
higher depletion by wolves. Rather, the results arise from differences in food
consumption.
Estimates of energy requirements made using standard allometric relation-
ships (Nagy et al. 1999) suggest that wolves eat more than would be expected for
their body mass, whereas tigers eat less than would be expected. Although it is pos-
sible that there are some differences in assimilation efficiencies between these two
predators, it is likely that most of these differences can be explained by two aspects
of their life history: hunting mode and sociality.
As ambush predators, tigers hunt by stalking followed by very short rushes
(Yudakov and Nikolaev 1992), with high success rates (Hornocker 1970). Chases
by tigers rarely extend beyond 150 m (Miquelle et al., unpubl. data). Yudakov and
Nikolaev (1992) reported 54% and 38% success of Amur tigers hunting wild boar
and red deer, respectively. Because tigers are solitary, intervals between kills are
high (six to nine days) (Sunquist and Sunquist 2002; Miquelle and Goodrich.,
unpubl. data). Collectively, high success rates, short chase distances, and long
intervals between kills result in low energy expenditures for tigers. By contrast,
wolves have low success rates, averaging 14% (based on individual prey) over 14
studies (Mech and Peterson 2003). As cursorial predators, they may chase prey
for kilometers (Husseman et al. 2003; Peterson and Ciucci 2003). Although living
in groups provides wolves the capacity to kill larger prey and obtain greater
biomass per kill (Gittleman 1989), group living also has costs. In particular, shar-
ing kills results in less energy acquired per individual per kill. Consequently, kill
rates must be considerably higher for wolves, leading to greater travel distances as
packs search for prey. An additional cost of group living is the time and energy ex-
pended on social interactions, a cost that solitary species like tigers largely avoid.
In summary, higher kill rates (a consequence of sociality) and the associated
greater travel distances, as well as greater energy expended in cursorial hunting
(greater chase distances) and social interactions, all result in greater energy de-
mands for wolves. These factors likely explain much of the difference in food con-
sumption between wolves and tigers. As additional support for this argument, it
is worth noting that the African wild dog (Lycaon pictus), another cursorial, social
predator for which good data exist, is also renowned for its greater than predicted
energy demands (e.g., Gorman et al. 1998).
202 From Largely Intact to Human-Dominated Systems
Due to limitations in our understanding of prey dynamics, predator selectiv-
ity, and the flexibility of large predators to adapt territory sizes to reflect changing
prey availability, our simulation modeling approach was necessarily coarse. Nev-
ertheless, our simulations also suggest that wolves can limit prey to a much greater
degree than can tigers. These results are supported by empirical data. Although
few direct estimates of harvest rates have been made for tigers, in the high-
ungulate-biomass systems of the Indian continent, offtake has been estimated at
less than 10% (Schaller 1967; Støen and Wegge 1996). Our model, which predicted
10% depletion when K= 20 km–2 deer (equivalent to a prey biomass density of
3000 kg km–2), is in agreement with these estimates. For wolves, offtake ranges
from minimal (less than 10–15% of the standing biomass of red deer removed each
year) (Glowacinski and Profus 1997; J2drzejewski et al. 2002) to much more sig-
nificant (35% reduction in moose populations, Messier 1994, 15–30% reduction in
elk populations, Eberhardt et al. 2003).
In conclusion, it appears that, despite the fact that wolves and tigers have his-
torically replaced one another as the top carnivore in the Sikhote-Alin ecosystem,
they are extremely unlikely to be functionally equivalent. Not only are there dif-
ferences in prey preferences but there are substantial differences in the extent to
which these two predators impact prey populations.
Conservation Implications
In Russia, where there is a strong hunting tradition that is based on maximum sus-
tainable yield, large carnivores are primarily viewed as competitors to human
hunters. The impact of wolves on ungulate populations in Russia has received at
least as much, if not more, attention than in North America (Filonov and Kalet-
skaya, 1985; Yudin, 1992). The general conclusion in Russia is that wolves can sig-
nificantly depress ungulate populations and should be controlled to maintain high
ungulate yields for hunters. A similar conclusion concerning wolf impact on un-
gulate populations has been reached by many in North America, but, whereas wolf
control is practiced across much of Russia, in North America it has been contro-
versial (Gasaway et al. 1992; Van Ballenberghe and Ballard 1994).
Our conclusions that tiger predation is unlikely to limit ungulate populations
to the same extent as wolf predation concur with the opinions of local biologists
Tigers and Wolves in the Russian Far East 203
(Kucherenko 1974; Dunishenko 1987). Nonetheless, the relationship between
tigers and Russian hunters is less than amicable, and “intraguild” killing of com-
petitors extends to the hunter–tiger relationship. Abundant evidence suggests that
competition killing is one of the primary motivations for tiger poaching in Rus-
sia (Miquelle et al. in press). Due to this inimical relationship, and because new
regulations provide nongovernmental hunting groups with wide-ranging respon-
sibilities to manage game species, hunters will be key stakeholders in determining
the future of tigers in the Russian Far East (Miquelle et al. in press). Finding com-
mon ground with hunters must therefore be a primary task for those wishing to
conserve the Amur tiger. Paraphrasing one argument for tiger conservation, local
conservationists and biologists have proposed to local hunters that, “while tigers
may not be desirable, they prevent wolves from becoming abundant . . . and we
all know that wolves are worse than tigers in depressing prey numbers, so it is to
your advantage to tolerate the tiger.” Our models support local perceptions of the
relative capacity of these two predators to impact prey populations and suggest
that this “backdoor” rationale for tiger conservation has a biological basis. This ar-
gument does little to foster a more balanced perspective on carnivores, and fur-
ther enforces a negative perspective on wolves. However, it appears that if Amur
tiger conservation is a priority in the Russian Far East, maintaining high numbers
of tigers in the Sikhote-Alin ecosystem will come at the cost of low wolf numbers
due to competitive exclusion. Although promoting tiger conservation as a mech-
anism to control wolves may not encourage an appreciation of large carnivores or
the intricacies of ecosystem processes, it provides a message understandable at the
local level, appears to have a real biological basis, and is likely to elicit a more ef-
fective and rapid response than other approaches.
The absence of functional redundancy has implications not only for tiger con-
servation strategies but for biodiversity conservation as well. Because large carni-
vores are not necessarily functionally similar, saving “a” large carnivore is not
equivalent to saving ecosystem integrity or ecosystem processes. The definition
of “integrity” and the types of processes saved will depend on the carnivore re-
tained in the system. Understanding the potential differences in the way large car-
nivore species structure communities is therefore a necessary prelude to defining
biodiversity conservation strategies.
Conservation of large carnivores is obviously not synonymous with biodi-
versity conservation. For example, it provides no guarantee that other rare species,
204 From Largely Intact to Human-Dominated Systems
hotspots, or centers of endemism will be retained. Nonetheless, we believe that
conservation of carnivores can help achieve these other conservation objectives.
This chapter (and many others in this book) demonstrates that large carnivore
ecology is largely driven by their relationship with prey species. Tigers and wolves,
and indeed most large carnivores, are habitat generalists, and as such, minimum
prey density is perhaps the key habitat parameter determining their presence.
Large area requirements are not an intrinsic characteristic of large carnivores but
a consequence of their need for adequate prey (Karanth and Stith 1999; Miquelle
et al. 1999). Carnivore prey requirements therefore help define the minimum suit-
able area needed for biodiversity conservation but not necessarily the specific lo-
cations. Carnivore habitat is not spatially fixed but can be created by managing the
prey base, a well-understood process that largely requires adequate protection
from human harvest. Hotspots, areas of high endemism and habitat for rare spe-
cialists, can also be carnivore habitat because the exact parcel of land is less im-
portant than the presence of suitable prey for large carnivores. The charisma and
large area requirements of large carnivores thereby provide a mechanism for
achieving other conservation objectives. In this context, carnivore conservation
is not synonymous with biodiversity conservation but as a mechanism to define
ecosystem processes, identify minimum area requirements, and generate public
interest, one of the necessary tools to achieve it.
Summary
Through their direct effects on prey species, predators can have profound indi-
rect effects on many aspects of biodiversity. Where functional redundancy exists
within the predator guild, however, conservation of a particular carnivore may not
be essential to maintain the wider biodiversity of an area. Unfortunately, few stud-
ies have evaluated functional overlap of large carnivores. Here, we consider the
impact of two predators, the Amur tiger and the gray wolf, on each other and on
prey populations in the Sikhote-Alin ecosystem of the Russian Far East. Using his-
torical data, we show that wolves do replace tigers as the top predator when
anthropogenic influences depress tiger numbers; however, recovery of tiger num-
bers leads to competitive exclusion of wolves by tigers. Proportions of prey species
in diets are similar, but, whereas wolves show a preference for red deer and select
Tigers and Wolves in the Russian Far East 205
more vulnerable age classes, tigers exhibit a clear preference for wild boar and no
apparent selection for any sex–age class. Two modeling efforts support the con-
tention that wolves are more likely to deplete prey populations to a much greater
extent than tigers. Local conservationists have used evidence for competitive ex-
clusion and the apparent differential impact on prey to convince local hunters of
the need to conserve tigers (as a means of reducing wolf numbers and wolf im-
pact on ungulates). Although large carnivore conservation is not synonymous with
biodiversity conservation, the charisma, large area requirements (related directly
to prey requirements), and plasticity in habitat requirements of most large carni-
vores provide a mechanism for achieving other conservation objectives.

Maurice Hornocker and Howard Quigley of the Hornocker Wildlife Institute con-
ceptualized and initiated the Siberian Tiger Project, and we are grateful for their
expertise, support, and dedication to research and conservation of tigers. The
Siberian Tiger Project is funded by The National Fish and Wildlife Foundation’s
“Save the Tiger Fund,” 21st Century Tiger, The National Geographic Society, The
National Wildlife Federation, Exxon Corporation, The Charles Engelhard Foun-
dation, Disney Wildlife Fund, Turner Foundation, Gary Fink, Richard King Mel-
lon, and the Wildlife Conservation Society. Director A. A. Astafiev and former
Assistant Director of Science M. N. Gromyko of Sikhote-Alin State Biosphere Za-
povednik provided the logistical, administrative, and political support necessary
to conduct fieldwork in Sikhote-Alin Zapovednik, and the Russian State Commit-
tee for Environmental Protection provided permits for capture of tigers. We thank
I. Nikolaev, B. Schleyer, N. Reebin, A. Reebin, A. Kostirya, I. Seryodkin, V.
Melnikov, A. Saphonov, V. Schukin, and E. Gishko for their assistance with data
collection. P.A. Stephens was supported by a grant by the U.S. Forest Service, In-
ternational Programs, to participate in this work. One anonymous reviewer pro-
vided valuable comments, as did all editors of this book, who, more importantly,
had the patience and tolerance to guide us through formulation of this chapter.
206 From Largely Intact to Human-Dominated Systems
Tigers and Wolves in the Russian Far East 207
Appendix 10.1
Life-history parameters used in the predator population models
Parameter Tiger Wolf Sources
Survival
Maximum age 25 15 (Mech 1974; Mazák 1981)
Background 0.95 (females > 1 yr) 0.90 (males > 1 yr) (Peterson and Page 1988;
survival ratea0.90 (adults > 2 yrs) 0.80 (yearlings) Hayes and Harestad 2000a;
0.75 (cubs < 1 yr) 0.75 (pups < 1 yr) Kerley et al. 2003)
Fecundity and birth
Age at first 4 yrs 2 yrs (Mech 1974; Mazák 1981;
reproduction Kerley et al. 2003)
Annual probability 0.55c1.00d(Mech 1974; Fritts and Mech
of female 1981; Mech 1981; Fuller
reproduction 1989;Kerley et al. 2003)
Mean (± SD) 2.38 (± 1.15) 6.00 (± 0.50) (Mech 1974; Mazák 1981;
litter sizebMech 1981; J2drzejewska et
al. 1996; Kerley et al. 2003)
Sex ratio at birth 0.41 0.50e(Kerley et al. 2003)
(males per offspring)
aBackground survival rates reflect mortality from causes other than food limitation. The figures used
were selected to reflect mortality in the absence of anthropogenic causes. Survival rates are expressed as
annual equivalents.
bLitter sizes in the model were drawn from normal distributions described by these parameters but were
reduced if food was limiting.
cThe territories of male and female tigers are known to overlap. However, it was assumed that one male
could mate with no more than three females in any one year.
dOnly the alpha female could breed in any wolf pack (e.g., Hayes and Harestad 2000a).
eAssumed, in the absence of detailed information.
... Unlike felids and the extinct nimravids which use(d) their powerful forelimbs to drag down prey and exact a powerful killing bite with shortened snouts and mandibles, the wolf, along with other canids and hyaenids, have lost the ability to supinate their paws (Andersson andWerdelin 2003, Andersson 2004), leaving their jaws and teeth the sole method of killing prey by inflicting numerous shallow bites from multiple individuals, causing prey to bleed out (Gittleman 1989). Standard allometric relationships reported by Nagy et al. (1999) indicate that wolves eat more than would be expected for their body mass, which Miquelle et al. (2005) attribute to the sociality and hunting mode of this canine: wolves are long-limbed cursorial predators that may 119 chase prey for kilometers (Husseman et al. 2003, Peterson andCiucci, 2003) and display low rates of hunting success, averaging a 14% success rate based on individual prey (Mech and Peterson 2003). Group living, while providing the capacity for wolves to kill larger prey (cooperative hunting hypothesis) and obtain more biomass per kill (Gittleman 1989), also cost individual pack members as sharing kills limits the total energy acquired for each pack member per kill (Schmidt and Mech 1997). ...
... Additional energy is also spent on social interactions, a cost that solitary predator species largely avoid. Greater expended energy from social living and cursorial hunting leads to more frequent hunting bouts and greater distances travelled search for prey; kill rates must consequently be high for wolves to compensate for this energy expenditure (Miquelle et al. 2005). ...
Thesis
Full-text available
Introduction: Vast amounts of knowledge about wolf (Canis lupus) ecology, life history, anatomy, and physiology have been gathered over the decades; however, the relationship the wolf shares with its prey, specifically in terms of how the morphology of its skull is related to its prey in a naturally selective context, has received far less attention. Methods: I use multivariate statistics to analyze a dataset composed of 23 linear measurements taken from the crania, teeth, and mandibles of 583 museum-curated North American wolves (333 males, 250 females). A geographically varying index of mean pooled prey weight (MPPW) was developed through literature review against which variables reflecting skull size and shape would be compared. Results: Wolves are sexually dimorphic, with males appearing significantly larger on average than females in all linear skull measurements. A 5-Factor PCA on raw skull data revealed that PC 1 accounted for ~76.5% of the observable variation in male and female data. The overall size of any skull, male or female, is best represented primarily by its condylobasal and jaw lengths (CBL, JL) and secondarily by its zygomatic width (Zyg B). A factor value (F1 Score) associated with PC 1 was ascribed to skulls, and this value equates with its greatest length (GSL; premaxilla to sagittal crest). The largest skulls were harvested from mainland Alaska and northwestern Canada. Moderately sized skulls were from arctic Canada, southeast Alaska, and northern Ontario. The smallest skulls were from south-central Ontario and New Mexico. Linear regression indicated significantly positive relationships for wolves of both sexes between MPPW and F1 Score, GSL, and Zyg B. These relationships, especially between MPPW and Zyg B, are stronger in males, suggesting that males are better adapted than females for hunting and killing large prey. Skull size is also significantly positively related to prey species richness, suggesting that greater availabilities of food may help to maximize growth potentials in wolves. I also calculated the average bite forces (BFs) of wolves at the mandibular canine and the carnassial using 3D digital modelling techniques and the "dry skull method" developed by Thomason (1991). BFs were significantly positively related to MPPW, and this relationship was stronger for females. This may be due to greater natural variance in males; it may also suggest that females are adapted to process flesh with greater efficiency before ingestion, possibly a consequence of nutritional requirements associated with reproduction. BFs are most strongly associated with the size of the jaw muscles, which are intricately linked to Zyg B. The relationship between BFs and Zyg B is stronger than the relationship between MPPW and BFs, indicating that prey size only indirectly influences bite force via skull morphology. Conclusions: This study determines the presence of significant links between prey size, prey availability, and skull size in wolves, and is the first to distinguish between sex and among geographic regions in estimating bite forces using dry skulls. This study would benefit from (1) additional skulls from specific regions and (2) data on regional prey densities to compare against skull sizes. Future research may involve genetic analyses to examine if regional differences in wolf skull morphology are also reflected by genotype, and isotope analyses to better characterize the regional diets of geographically diverse wolf populations.
... Interaction with a larger carnivore results into wolves being displaced by Amur tigers Panthera tigris (Miquelle et al., 2005), according to habitat suitability modeling for both species (Voloshina et al., 2014). Study period spans 1960-1989 and includes 566 wolf and 2,543 tiger locations within and around Lazo Nature Reserve, Russian Far East. ...
Article
Full-text available
This paper provides an overview of wolf research in Russia at the beginning of the 21st century. Wolf research covered various directions, including population density estimation, management methods and minimization of human-wildlife conflicts, general and behavioral ecology, behavior, wolf population genetics and morphology, paleontology, dog domestication, helminthology and the wolves’ role in the rabies transmission. Some studies are performed with state-of-art methodology using molecular genetics, mathematical modeling, camera traps, and GPS telemetry.
... The lowlands surrounding the Sudety Mountains (Bohemian, Moravian, and Silesian) were overwhelmed by packs of robust steppe wolves, whose remains have been recorded regularly from open-air sites (Herr 1924;Kowalski 1959;Wiśniewski et al. 2009;Marciszak et al. 2019Marciszak et al. , 2020. In addition, the presence of wolves might be a limiting factor, because as a solitary hunter, the leopard usually avoided aggressive competitions with large dogs (Heptner and Sludskij 1972;Eaton 1979;Spassov and Raychev 1997;Miquelle et al. 2005). Given the similar size of wolves and leopards, in a one-on-one confrontation, the leopard has an advantage, but the confrontation of a solitary cat with a pack of wolves is a different story (Figure 3). ...
Article
The Pleistocene history of the leopard ( Panthera pardus ) in Europe has been documented by the material obtained from 312 localities, with the last dated ∼1.1 Myr. The relatively small and gracile form of the leopard was very rare during the late Early and Middle Pleistocene. Only after the disappearance of the jaguar ( Panthera gombaszoegensis ) did P . pardus spread widely in Europe, increasing in size and ecologically substituting P. gombaszoegensis . The number of late Middle Pleistocene localities with leopard remains, younger than 300 kyr, increased considerably. The leopard reached the maximum extension of its geographical range in the Late Pleistocene. The Iberian Peninsula was the last European refuge for this cat. Six sites, the Naciekowa, Obok Wschodniej, Radochowska, and Wschodnia Caves from the Sudety Mountains and the Biśnik and Dziadowa Skała Caves from the Kraków-Częstochowa Upland, have documented the presence of the leopard in Poland between MIS 10/9 and MIS 3. These records are from rocky regions with rugged terrain and are located in the territory of Silesia (southern Poland). A newly obtained radiocarbon date (43–42 kyr) from the Radochowska Cave directly confirms the occurrence of P. pardus in the Sudety Mountains in the middle part of MIS 3.
... While considerable conservation investments are targeted towards understanding ecology and distribution patterns for both species (Walston et al., 2010), such efforts have rarely taken account of possible effect of competitive and intra-guild interactions on their abundance and distribution. Tigers are known to be socially dominant over other species, with competitive interactions affecting the time of activity, habitat use and abundance of other sympatric carnivores (Seidensticker, 1976;McDougal, 1988;Miquelle et al., 2005). Studies on tigers and leopards at local scales have investigated the mechanisms facilitating their coexistence (Johnsingh, 1992;Karanth & Sunquist, 1995Wang & Macdonald, 2009;Wegge et al., 2009) or competition between them Odden et al., 2010), suggesting that regional ecological factors could play a major role in shaping carnivore assemblages. ...
... The sub-adult leopardess 26,277 had the largest home range as she traversed vast areas between protected areas, tea plantations, crop fields and settlements from the western to the eastern part of the landscape. Home range sizes are an artifact of the spatial distribution and abundance of wild prey 41,42 . The smallest leopard home ranges are reported from the alluvial flood plains of Nepal 39 and eastern, southern Africa 43,44 where there is abundance of wild prey, whereas the largest reported home ranges prior to this study (103 km 2 ) were reported from Iran 40 . ...
Article
Full-text available
With a rise in human induced changes to natural habitats, large predators are forced to share space and resources with people to coexist within multiple-use landscapes. Within such shared landscapes, co-occurrence of humans and predators often leads to human-carnivore conflicts and pose a substantial challenge for biodiversity conservation. To better elucidate large carnivore space use within a hotspot of human-wildlife conflicts, we used GPS data for leopards (N = 6) to identify behavioral states and document spatial patterns of resource selection in response to season and human activity periods within a fragmented landscape of North Bengal, eastern India. We identified two major behavioral states (i.e. resting and travelling). From the resource selection models, we found leopards selected habitats with dense to moderate vegetation cover and proximity to water while resting and travelling within the landscape. During the dry season, when risk of human-leopard conflicts is highest, leopards selected tea plantations, forest patches but avoided protected areas. These results suggest a potential for increase in human-carnivore conflicts and a strategy to conserve large predators within multiple-use landscapes of South Asia.
... With all of these species experiencing significant range contractions and further being endangered by loss of prey base, they typify the problems faced by large carnivore guilds across the world (Wolf & Ripple, 2016. Studies that have investigated mechanisms facilitating sympatry amongst these species have documented that tigers are socially dominant with competitive interactions affecting the diet, activity patterns, habitat use and abundance of other sympatric carnivores (Seidensticker, 1976;Seidensticker et al., 1990;Karanth & Sunquist, 1995Miquelle et al., 2005;Harihar et al., 2011). In prey-rich habitats, studies suggest that differential prey selection in terms of species, body-size and age-sex classes facilitate sympatry (Karanth & Sunquist, 1995;Andheria et al., 2007). ...
Article
Full-text available
Inferring the mechanisms that facilitate sympatry amongst large mammalian carni-vores in anthropogenically altered habitats is critical if conservation initiatives aimed to recover populations are to be effective. In this study, we assessed the utilization of space, time and prey resources by tigers Panthera tigris, leopards P. pardus and dholes Cuon alpinus in Manas National Park (MNP), India-a site where prey densities are not limiting, yet anthropogenic influences alter the spatial and temporal behaviour of prey. We use photographic capture data and predator scats collected over three sampling years (2014-15, 2015-16 and 2016-17) to assess patterns of (a) spatial use using a multispecies species occupancy framework , (b) time-activity patterns and overlap between predators and prey using non-parametric circular kernel-density functions, (c) fine-scale spatio-temporal behaviour by comparing time-to-encounters of subsequent events and (d) predator diets by analysing prey remains in predator scats. Our results highlight that the predators segregate through fine-scale spatio-temporal avoidance rather than displaying population level changes in space-use, activity patterns or food habits. Overlap in space-use between tigers and leopards was high and time-activity patterns of the predators closely matched those of prey, suggesting that predators likely maximize resource acquisition in this prey-rich environment. Ungulate prey dominated the diet of predators, resulting in the high dietary overlap. From our results, we infer patterns of sympatry among large carnivores in the face of anthropogenic influences and highlight the need to understand interspecific interactions within a community before initiating conservation actions aimed at recovering these endangered species.
... In one clear case, the tracking of a reintroduced GPS collared female tiger in the Khingan Nature Reserve yielded positive evidence of accidental wolf kills, followed by regular tiger hunting and a subsequent change in the home range of the wolf pack . This proved the previously observed yet unverified correlation between the increase of tiger presence and the decrease or even complete relocation of wolves (Miquelle et al., 2005;Yudin, Yudina, 2009), thus corroborating the aforementioned hypothesis. A wolf pursued by a female leopard during 2016 was also observed in the Caucasus Nature Reserve and was confirmed by this observation as well. ...
Book
Full-text available
This monograph provides a comprehensive review and analysis of the available literature on the monitoring of big cats. Special attention is paid to the most up-to-date methods based on recent advances in technology, resulting in useful tools to remotely and noninvasively study animals in natural habitats, essential when working with rare species. Existing large- and small-scale approaches to monitoring big cats are described. Methods of monitoring the habitat conditions of the species and their dynamics, as well as the basics of modeling territories with suitable conditions for leopards, are suggested. The whole range of field sampling methods that enable data to be processed using contemporary techniques is described. Moreover, methods of processing collected data (obtained via GPS collars, photo and video recorders, and hormonal and molecular genetic analysis) as well as examples of results are considered.
... Law enforcement, frontline training and capacity building, national enabling policies and transnational actions are important for long-term survival of tiger population. However, in a large number of researches, often the focus has been on habitat restoration, corridor connectivity and maintenance of prey density (Miquelle et al 2005), vis-à-vis tiger resilience and low rate of extinction. It is observed that in the event of paucity of tiger population data, selective harvesting (poaching/targeted killing) of tiger population, conflict and mortality of dispersing tiger in a fragmented habitat, negative effect due to cultural dependency on forest and ethical unrest are not given enough weightage to plan conservation policies. ...
... Most leopard photo-captures were recorded during 0600-0659 hours (128 images), followed by 1700-1759 hours (126 images) (Figure 2), suggesting a crepuscular activity pattern. This activity pattern among leopards may be also explained probability by an avoidance from more dominant species such as tigers (Sunquist & Sunquist, 2002;Stephen, 2004;Miquelle et al., 2005;Mark Rayan & Linkie, 2016). When tigers hunt in certain areas, leopards tend to avoid hunting in the same place (Wang & Macdonald, 2009;Athreya et al., 2013). ...
Article
Full-text available
In Southeast Asia, the population range of the Leopard (Panthera pardus) is rapidly shrinking. In Peninsular Malaysia, temporal activity patterns of leopards are poorly known in its largest protected area-Taman Negara National Park. Our study obtained 1,263 photos of leopards from 235 camera trap stations over 31,333 trap nights covering the Kelantan, Pahang, and Terengganu portions of Taman Negara National Park. Our results show that leopard activity peaks around 0600-0659 hours (dawn) and 1700-1759 hours (dusk). This study represents a fresh attempt to document the crepuscular nature of leopards in Taman Negara National Park.
... Äðóãîé ïðèìåð. Áûëî èçâåñòíî, ÷òî òàì, ãäå âîçðàñòàåò ÷èñëåííîñòü òèãðà, ÷èñëåííîñòü âîëêà ëèáî ñíèaeàåòñÿ, ëèáî îí âîîáùå ïîêèäàåò äàííûé ðàéîí (Miquelle et al., 2005;Þäèí, Þäèíà, 2009), íî ïðÿìûõ äîêàçàòåëüñòâ íå áûëî. Ëèøü ìå÷åíèå GPS-îøåéíèêîì ðåèíòðîäóöèðóåìîé òèãðèöû è åå ïîñëåäóþùåå òðîïëåíèå íà òåððèòîðèè Õèíãàíñêîãî çàïîâåäíèêà äàëè äîñòîâåðíóþ èíôîðìàöèþ ñíà÷àëà î ñëó÷àéíûõ äîáû÷àõ òèãðèöåé âîëêîâ, à çàòåì î åå öåëåíàïðàâëåííîé îõîòå è ïîñëåäóþùåì ñìåùåíèè ó÷àñòêà îáèòàíèÿ ñòàè âîëêîâ (Êàñòðèêèí è äð., 2015). ...
Book
Full-text available
В монографии проанализированы и обобщены практически все доступные литературные источники, посвященные мониторингу крупных хищных. Особое внимание уделено самым современным методам, основанным на активном развитии технологий и позволяющим дистанционно и неинвазивно исследовать животных в естественных условиях, что принципиально важно при работе с редкими видами. Представлены существующие подходы к мониторингу крупных кошачьих, включающие как мониторинг крупного масштаба, так и более подробный мелкомасштабный. Предложены методы мониторинга условий обитания вида и их динамики и основы моделирования пригодных для обитания леопардов территорий. Приведен весь спектр способов сбора данных в полевых условиях, позволяющий обрабатывать их самыми современными методами. Рассмотрены варианты способов обработки собранных данных (ГИС на основе данных спутниковых ошейников и информации, полученной с автоматических фото- и видеорегистраторов, гормональных и молекулярно-генетических исследований) и примеры полученных результатов.
ResearchGate has not been able to resolve any references for this publication.