The Alexander Archipelago
Wolf: A Conservation
David K. Person, Matthew Kirchhoff,
Victor Van Ballenberghe, George C. Iverson,
and Edward Grossman
DAVID K. PERSON is a graduate fellow, Alaska Cooperative Fish and Wildlife
Research Unit at the University of Alaska Fairbanks, Fairbanks, AK 99775;
MATTHEW KIRCHHOFF is a research wildlife biologist, Alaska Department of Fish
and Game, P.O. Box 240020 Douglas, AK 99824; VICTOR VAN BALLENBERGHE is
a research wildlife biologist, U.S. Department of Agriculture, Forest Service, Pacific
Northwest Research Station, Forestry Sciences Laboratory, 3301 C Street, Suite 200,
Anchorage, AK 99503-3954; GEORGE C. IVERSON is the regional ecology program
leader, U.S. Department of Agriculture, Forest Service, Alaska Region, P.O. Box
21628, Juneau, AK 99801; and EDWARD GROSSMAN is a wildlife biologist with the
U.S. Department of the Interior, Ecological Services, 3000 Vintage Boulevard Suite
205, Juneau, AK 99801.
Conservation and Resource
Assessments for the Tongass Land
Management Plan Revision
Charles G. Shaw III
The Alexander Archipelago Wolf:
A Conservation Assessment
David K. Person
Victor Van Ballenberghe
George C. Iverson
U.S. Department of Agriculture
Pacific Northwest Research Station
General Technical Report PNW-GTR-384
In cooperation with:
U.S. Fish and Wildlife Service
Alaska Department of Fish and Game
Person, David K.; Kirchhoff, Matthew; Van Ballenberghe, Victor; Iverson,
George C.; Grossman, Edward. 1996. The Alexander Archipelago wolf: a
conservation assessment. Gen. Tech. Rep. PNW-GTR-384. Portland, OR: U.S.
Department of Agriculture, Forest Service, Pacific Northwest Research Station.
42 p. (Shaw, Charles G., III, tech. coord.; Conservation and resource assessments
for the Tongass land management plan revision).
We summarized the scientific information available for the Alexander Archipelago
wolf (Canis lupus ligoni) in the Tongass National Forest of southeast Alaska. Infor-
mation concerning the morphology, distribution, taxonomy, genetics, and ecology
of wolves are presented. Three issues for the conservation of wolves in southeast
Alaska are discussed: loss of long-term carrying capacity for deer due primarily to
extensive timber harvesting, increased mortality of wolves associated with improved
human access from roads, and continued high levels of harvest of wolves by humans.
Continued timber harvesting at current levels and by current methods will likely have
adverse consequences for some segments of the wolf population. Although some
short-term regulatory changes and the management of road access may need to be
considered to keep wolf harvest at a sustainable level, the most important considera-
tion is to maintain long-term carrying capacity for deer, the principal prey for most of
the wolf population. A series of old-growth forest reserves may provide an effective
strategy to increase the likelihood that wolves will persist where extensive timber
harvesting has occurred, or is planned.
Keywords: Alexander Archipelago wolf, Canis lupus ligoni, effects of logging on
wildlife, population dynamics of wolves, predator-prey dynamics, roads and wolf
mortality, Tongass National Forest, southeast Alaska.
The Alexander Archipelago wolf (Canis lupus ligoni) occupies most of southeast
Alaska from Yakutat Bay to Dixon Entrance except for Admiralty, Baranof, and
Chichagof Islands. Based on common cranial characteristics, the Alexander
Archipelago wolf was considered by early taxonomists to be a distinct subspecies.
Recent taxonomic work suggests that these wolves may have originated from a
larger subspecific group (C. l. nubilus) that at one time inhabited most of the
contiguous Western United States. Wolves probably entered southeast Alaska
sometime after the Wisconsin glaciation, following the northward expansion of
black-tailed deer (Odocoileus hemionus) along the coast. The hypothesis of a
southern origin is supported by recent genetic research showing that wolves in
southeast Alaska share a common allele not found in a sample of wolves from
interior Alaska or the Yukon. The population is relatively isolated from other wolf
populations by water and mountain barriers.
A study of the ecology of wolves in southeast Alaska was conducted on Prince
of Wales and Kosciusko Islands from 1992 to 1995. Average home-range size of
radio-telemetered wolves was 280 square kilometers (109 mi2), with 75 percent of
the radio locations for each pack within “core areas” averaging 124 square kilometers
(48 mi2). Pack sizes ranged from 2 to 12, with 7 to 9 typical in early autumn. Annual
rates of dispersal averaged 39 percent; 71 percent of dispersers were adults ≥2 years
old. Dispersal distances were relatively short (13 to 182 kilometers [5 to 71 mi]) pre-
sumably due to inter-island water barriers. Wolf movements were more restricted
during the denning and pup-rearing season (mid-April through August), when home
ranges were 50 percent smaller than in winter. Of the 14 dens located in this study,
all were in old-growth forest within 100 meters (328 feet) of fresh water. One den
was under a large log; all others were in cavities beneath the roots of large trees.
Sitka black-tailed deer (O. h. sitkensis) were the primary prey of wolves. Deer remains
occurred in 90 percent of wolf feces (scats) examined from Prince of Wales Island.
Deer occurred exclusively in 45 percent of the scats. The only other prey occurring
with >10 percent frequency was beaver (Castor canadensis). Other prey consumed
in small quantities included black bears (Ursus americanus), mustelids, other small
mammals, birds, and salmon (Oncorhynchus spp.). Using information on diet com-
position, consumption rates, and body size of prey, we estimated that wolves on
islands in southern southeast Alaska consumed an average of 26 deer per wolf per
year (SE = 4.1).
Most of the wolves in southeast Alaska occur on the large islands south of Frederick
Sound. These islands (game management units 2 and 3) support approximately 60
to 70 percent of the total population. By extrapolating from empirical population esti-
mates for Prince of Wales Island, we estimated the autumn 1994 population of wolves
in southeast Alaska at slightly over 900 animals (SE = 216). Hunting, trapping, and
illegal killing accounts for a high percentage of the mortality in wolves. Based on
analysis of trapping and hunting mortality by wildlife analysis area (WAA), we de-
termined that mortality was correlated with the linear kilometers of road within WAAs.
Indeed, reported wolf harvest increased twofold when the length of road below 370
meters (1200 ft) elevation exceeded 95 kilometers (59 mi), regardless of the size
of the WAA. This corresponded to an approximate road density of 0.4 kilometer per
square kilometer (≈0.7 mi/mi2), most of which were open to human access. Between
1993 and 1995, the average annual mortality in a total sample of 24 radio-collared
wolves on Prince of Wales Island was 50 percent (SE = 13 percent). If applied to the
overall wolf population on Prince of Wales Island, this rate of mortality would not be
Wolf populations are closely tied to population levels of their ungulate prey. For
southeast Alaska, we predicted the number of deer required per wolf to attain
equilibrium between deer and wolves by using a Monte Carlo simulation of a model
that calculated equilibrium ratios for wolves and their ungulate prey. We assumed a
high average finite rate of increase for deer (1.3), a mean predation rate of 26 deer
per wolf per year, and a human harvest of deer equal to 21 percent of the annual
increment. Our results suggest that 170 to 180 deer per wolf are needed for a
95-percent probability of equilibrium, provided that mortality of deer due to preda-
tion is primarily additive. We cannot suggest a minimum deer population because we
do not know what would constitute a minimum viable wolf population either demo-
graphically or genetically. Nevertheless, if we expect to sustain the current post-
denning population of 250 to 300 wolves on Prince of Wales Island (along with sub-
sistence and sport harvests of deer) with a high probability of attaining equilibria, then
sufficient habitat is needed to support 42,500 to 54,000 deer.
Our review raises a number of issues concerning the long-term sustainability of
wolves in southeast Alaska. Many more data are needed on wolf population structure,
genetic structure, and predator-prey relations to fully address these issues and the
overall question of viability. The Alexander Archipelago wolf exists in small numbers
in a rapidly changing insular environment. Projected growth in human population,
increasing road access, and the continuing loss and fragmentation of high-quality
deer habitat will increase the risk of not maintaining a viable, well-distributed popula-
tion of wolves in southeast Alaska. The area of most immediate concern is game
management unit 2, including Prince of Wales and Kosciusko Islands.
Management actions that address risks to wolf populations include modifying hunting
and trapping regulations as necessary, limiting construction of new roads and effec-
tively closing some existing ones, and modifying timber harvest strategies to minimize
fragmentation and loss of critical deer winter range. Habitat to support a minimum
density of 5 deer per square kilometer (13 deer/mi2), where deer are the primary prey
for wolves, would provide for current levels of deer harvest by hunters, trappers, and
wolves. In areas less productive for deer, maintaining current densities of deer is
particularly important. Setting aside contiguous blocks of habitat within each biogeo-
graphic province that are large enough to encompass at least one wolf pack core
home range (200 square kilometers [76 mi2]) would markedly increase the likely
persistence of wolves, especially if the reserves contain high-quality deer habitat
sufficient to support an average density of deer equal to 7 deer per square kilometer
(18 deer/mi2). The objective of maintaining undisturbed blocks of habitat within each
biogeographic province is to assure the persistence of several wolf packs that will
serve as source populations capable of replacing wolves that periodically disappear
from adjacent disturbed lands.
Purpose and Need
In 1990, wolves in southeast Alaska were identified by a USDA Forest Service-
sponsored interagency committee as a species for which there may be concerns
about viability or distribution as a result of extensive timber harvesting in the Tongass
National Forest.1In December 1993, the Biodiversity Legal Foundation (Boulder,
CO) and an independent biologist from Haines, AK, filed a petition with the U.S. Fish
and Wildlife Service (FWS) requesting that wolves in southeast Alaska be listed as
a threatened subspecies pursuant to the Endangered Species Act of 1973, as
amended. The FWS ruled that listing was not warranted at this time, but added:
“However, it is clear by our analysis that without significant changes to the existing
Tongass Land Management Plan, the long-term viability of the Alexander Archipelago
wolf is seriously imperiled.”2Following that notice, this assessment was prepared
under the auspices of the memorandum of understanding among the Forest Service,
Alaska Department of Fish and Game, and the Fish and Wildlife Service. They, along
with the Regional Forester, Alaska Region, requested that a conservation assess-
ment be prepared to summarize available scientific information on the status and
ecology of wolves in southeast Alaska, and to identify issues regarding their long-
term viability and distribution.
1Suring, L.; Crocker-Bedford, D.C.; Flynn, R.W. [and others].
1993. A proposed strategy for maintaining well-distributed,
viable populations of wildlife associated with old-growth
forests in southeast Alaska. Review draft. 278 p. On file
with: U.S. Department of Agriculture, Forest Service, Alaska
Region, P.O. Box 21628, Juneau, AK 99802-1628.
2U.S. Fish and Wildlife Service. 1994. Report on petition
to list the Alexander Archipelago wolf under provision of
the Endangered Species Act, 12 month finding; under
signature of Mollie Beattie, Director. 25 p. (p. 19). On
file with: Alaska Department of Fish and Game, Douglas,
This conservation assessment was chartered under the Tongass [National Forest]
Land Management Plan revision and the interagency memorandum of understanding
(MOU) among the Alaska Region of the Forest Service, the Alaska Region of the
U.S. Fish and Wildlife Service, and the Alaska Department of Fish and Game to
conserve species tending toward listing.3Our report summarizes the most current
information available on the taxonomy, genetics, distribution, ecology, and population
dynamics of wolves in southeast Alaska. Much of the information has not been pub-
lished. In particular, we have drawn heavily from ongoing studies on wolves in south-
east Alaska that is cosponsored by the USDA Forest Service, the Alaska Department
of Fish and Game, the U.S. Fish and Wildlife Service, and the University of Alaska,
Fairbanks. Results from that work are preliminary.
3The primary objective of the MOU was to foster inter-
agency cooperation for the conservation of candidate and
sensitive species to avoid the need for listing and protection
under the Endangered Species Act. The wolf, in addition to
the marbled murrelet (Branchyramphus mamoratus) and
northern goshawk (Accipter gentilis), was identified as
a priority species for conducting a conservation assess-
ment to achieve MOU objectives.
1 A Brief Description of Southeast Alaska
Description and Distribution of Alexander Archipelago Wolves
21 Prey Populations
This page has been left blank intentionally.
Document continues on next page.
A Brief Description of
Southeast Alaska comprises a narrow strip of mainland and a chain of islands known
as the Alexander Archipelago wedged between the Pacific Ocean and the Coast
Mountains of Canada (fig 1). The archipelago includes hundreds of islands ranging in
size from less than 1 hectare (2.4 acres) to over 6,000 square kilometers (2,340 mi2)
that extend over 500 kilometers (310 mi) in a line parallel to the mainland. Elevations
less than 800 meters (≈2,500 ft) are covered by temperate rain forests of Sitka spruce
(Picea sitchensis (Bong.) Carr.), western hemlock (Tsuga heterophylla (Raf.) Sarg.),
western redcedar (Thuja plicata Donn ex D. Don), and Alaska-cedar (Chamaecyparis
nootkatensis (D. Don) Spach.) interspersed with muskegs. Subalpine and alpine
zones of vegetation exist above 800 meters (≈2,600 ft), with the highest elevations
covered by rock and ice. Several river drainages connect the region to interior British
Columbia and the Yukon. Weather conditions are highly variable, with annual pre-
cipitation ranging between 130 and 400 centimeters (≈50 to 160 in). Accumulation of
snow is greatest on the mainland and northernmost islands and becomes intermittent
in the southern portion of the archipelago. About 80 percent of the land in southeast
Alaska is contained within the Tongass National Forest, the largest National Forest
in the United States.
Based on a sample of 37 wolves representing most of the Alexander Archipelago
and the adjacent mainland, taxonomist E.A. Goldman described the Alexander
Archipelago wolf (Canis lupus ligoni) as smaller, shorter haired, and darker in normal
coloration than wolves in northern and interior areas of Alaska (Goldman 1944).
Wood (1990) also described the Alexander Archipelago wolf as smaller with shorter
and coarser hair than interior Canadian or Alaskan wolves. On islands in the southern
portion of the archipelago, a black color phase constitutes about 20 percent of the
population, gray or brown about 80 percent, and white or nearly white wolves less
than 1 percent (Wood 1990). On the northern mainland, the black color phase is
more common, accounting for about 50 percent of the wolves in the harvest (Morgan
1990). Adult wolves in southeast Alaska weigh an average of 39.5 kilograms (87 lb),
with females weighing about 6.8 kilograms (15 lb) less than males (Alaska Depart-
ment of Fish and Game [ADFG] 1960, Wood 1990). Wolves from some of the major
island groups in the archipelago may be somewhat smaller than the average for
the region. For example, adults from a sample of 24 radio-collared wolves from
Kosciusko and Prince of Wales Islands averaged 34 kilograms (75 lb) (n =11,
SE = 4.5 kilograms [10 lb]), with females weighing about 1 kilogram (2.2 lb) less
than males (Person and Ingle 1995).
Wolves occur on the mainland and all larger islands in southeast Alaska except
Admiralty, Baranof, and Chichagof Islands (fig. 2). Klein (1965) hypothesized that
wolves colonized the region from the south following the postglacial northward
expansion of black-tailed deer (Odocoileus hemionus). Carbon-14 dates of fossil
bones from deer found on Prince of Wales Island suggest that deer arrived at least
8,000 years B.P. (Jull 1993). Although most of the islands in the southern half of the
archipelago are used by wolves, only the largest, including Prince of Wales, Kuiu,
Kupreanof, Mitkof, Etolin, Revillagigedo, Kosciusko, and Dall Islands, likely support
packs of wolves that persist longer than a few years. For example, Prince of Wales
and the islands immediately adjacent to it (including Kosciusko and Dall Islands)
Figure 1—Southeast Alaska and the Tongass National Forest (USDA Forest Service 1991).
constitute game management unit 2 (GMU 2, fig. 2), so-designated by the ADFG.
Within GMU 2, only the three largest islands—Prince of Wales, Kosciusko, and
Dall—are known to have been continuously occupied by wolves for more than
20 years (Person and Ingle 1995). Wolf packs may include several smaller islands
(e.g., Baker, Lulu, Noyes, Tuxekan, Marble, Thorne) in their home ranges or may
exclusively inhabit smaller islands for a few years, but they are unable to persist
permanently (Klein 1996, Person and Ingle 1995).
From a continental perspective, wolf populations in southeast Alaska are relatively
isolated. The Coast Mountains to the east are glaciated or snow-covered year-round,
and only six rivers or passes penetrate the 800 kilometers-long (500-mi) cordillera.
The most significant of these are the Stikine, Taku, Whiting, and Unuk Rivers.
Although no direct evidence exists regarding the dispersal of wolves through these
passes or river valleys, wolves do occur in the valleys, and some level of inter-
change between interior Canada and the coast may occur. Nevertheless, the sig-
nificant morphological differences between southeast Alaskan wolves and wolves
from interior Canada and Alaska (Friis 1985, Nowak 1983, Pedersen 1982) suggest
that the level of interchange is small or admixture has occurred only recently.
Figure 2—Distribution of Alexander Archipelago wolves in southeast Alaska. Wolves occupy areas shown
in black. Map also shows game management units (GMUs) in southeast Alaska.
Within the Alexander Archipelago, dispersal of wolves among individual islands (or
island groups) probably is restricted by water barriers. The absence of wolves (or
wolf fossils) on the northern islands of the archipelago attests to the effectiveness
of Frederick Sound, Stephen’s Passage, and Icy Strait as barriers to movement by
wolves. Dense populations of brown bears (Ursus arctos) on Admiralty and Chichagof
Islands also may contribute to the lack of successful colonization by wolves. Clarence
Strait, which runs north-south along the eastern shore of Prince of Wales Island,
restricts movement from the mainland and from Revillagigedo Island to GMU 2.
Although wolves may reach Prince of Wales Island from the mainland via a chain
of stepping-stone islands at the north end of Clarence Strait, at least eight directed
swims (some 1.6 to 3.2 kilometers [1 to 2 mi] long) would be needed to complete
such a crossing. Evidence exists suggesting that wolves are capable of swimming
up to 4 kilometers (2.5 mi; Person and Ingle 1995), but the shapes and distribution
of land masses, currents, and the high frequency of storms combine with distance
to impede dispersal among the major island groups in the archipelago. Exchange
between mainland and island populations is probably greatest in the vicinity of Mitkof
Island and the Stikine River Delta, where the swimming distance is short at low tide.
Water barriers between some major island groups and the mainland also could limit
interchange between these island wolves and wolves immigrating from interior Alaska
Wolves in southeast Alaska are legally hunted and trapped. A maximum of five
wolves per hunter may be shot from 1 August to 30 April. Trappers may harvest
wolves from 10 November to 30 April with no bag limit. The annual reported harvest
in southeast Alaska from 1990 to 1995 has averaged about 194 wolves (SE = 35,
GMUs 1 to 5). Hunting and trapping seasons are regulated and administered by the
ADFG through the Board of Game. Subsistence hunting and trapping on Federal
public lands are regulated and administered by the Federal Subsistence Board.
State and Federal regulations regarding wolf harvest currently are identical.
Early taxonomists recognized 24 subspecies of gray wolves (Canis lupus) in North
America (Goldman 1944, Hall 1981) based on morphometric analyses of skull char-
acteristics. Wolves from southeast Alaska were included in this classification as a
subspecies, C. l. ligoni, unique to the Alexander Archipelago; their range extended
from Dixon Entrance north to Yakutat Bay including the coastal mainland and all
islands south of Frederick Sound. Pedersen (1982) performed an extensive
morphometric analysis of wolves in Alaska by using a transect sampling strategy
to determine if the four Alaskan subspecies (C. l. pambasileus, C. l. tundrarum,
C. l. alces, and C. l. ligoni) described by Goldman constituted gradual clinal vari-
ations or distinct, geographically correlated populations. Pedersen (1982) reported
that the three subspecies from interior Alaska (C. l. pambasileus, C. l. tundrarum,
and C. l. alces) are indistinguishable from each other and probably represented clinal
variation within a single population, but concluded that the abrupt morphometric dif-
ferences between C. l. ligoni and interior Alaskan wolves suggest that wolves from
southeast Alaska are a distinct population, possibly warranting subspecific classifi-
cation. Nowak (1983) and Friis (1985) also concluded that C. l. ligoni is morphometri-
cally distinct from interior Alaskan and Canadian wolves; nevertheless, both authors
suggested that the Alexander Archipelago wolf is related to wolves that historically
occupied coastal British Columbia, Vancouver Island, and the contiguous Western
United States. Nowak (1996) recently proposed a revised taxonomy for extant pop-
ulations of wolves in which C. l. ligoni is considered an isolated population of C. l.
nubilus (fig. 3), a subspecies including wolves in central Canada and Minnesota.
This latest taxonomic revision is likely to be the most widely accepted system among
biologists and taxonomists. Regardless of the subspecific status of C. l. ligoni, mor-
phological data strongly suggest a southern affinity, supporting the hypothesis that
wolves colonized southeast Alaska from the south. As taxonomist Ronald Nowak
(1994) pointed out in his letter:
The original wolf populations of the entire western United States, and probably of coastal
British Columbia, have been extinct for many years. Wolves of another subspecies have
recently expanded into much of southwestern Canada and the northwestern United
States. Essentially, the wolf population of Southeast Alaska is surrounded on one side
by the sea and on the land side by wolves of another subspecies. It is probably isolated
from its nearest living relatives (Minnesota to central Canada) by about 2,000 miles.
The Southeast Alaska wolf population also can be viewed as a remnant U.S. pop-
ulation.…It is as distinctive a population segment as can be found among North
Studies incorporating allozyme electrophoresis of nuclear DNA (Kennedy et al. 1991)
and restriction fragment analysis of mitochondrial DNA (mtDNA; Wayne et al. 1992)
concluded that wolf populations in North America show little genetic differentiation
among geographic regions. Wayne et al. (1992) identified 13 mtDNA genotypes in
a sample of wolves representing most of their current range in North America and
estimated sequence divergence to be small, ranging from 0.10 percent to 0.83 per-
cent. Indeed, one genotype was virtually ubiquitous and occurred in wolves from
Manitoba, northern Alaska, Yukon, Alberta, Montana, and Vancouver Island.
These data from mtDNA contrast with the current taxonomic system that recognizes
24 distinct subspecies of wolves in North America. At the very least, they support the
revised taxonomic system proposed by Nowak (1996) that identifies only five distinct
groups, and may suggest that wolves in North America are a single large population
(Wayne et al. 1992). Nonetheless, genetic data are based primarily on results from
restriction fragment analysis of mtDNA, which is informative only above the level of
0.5 percent genetic divergence (Shields 1995) and may have insufficient resolution
to detect recent divergence in wolf populations. Further, none of the genetic studies
included C. l. ligoni in the samples.
Results from a preliminary genetic analysis of mtDNA from C. l. ligoni identified a
fixed allelic substitution in wolves from southeast Alaska distinct from those surveyed
in northern Alaska and the Yukon (Shields 1995). In addition, genetic variation at
eight other nucleotides within the mtDNA genomes of northern Alaska and Yukon
wolves was not observed in any of the samples from southeast Alaskan wolves.
This study used direct sequencing of a 310-base pair portion of the control region
of PCR-amplified mtDNA, which is capable of detecting any genetic variation existing
within the segment of the mtDNA genome surveyed. These data are significant
because only a small portion of the mtDNA genome (<2 percent) was analyzed,
thereby suggesting that additional genetic variation may exist within the entire
genome. Nonetheless, sample sizes were small. Only 38 samples (29 from southeast
Alaska, 9 from northern Alaska and the Yukon) from a total of 50 yielded amplified
DNA adequate for analysis. A larger sample of wolves from outside southeast Alaska
may reveal a closer affinity between C. l. ligoni and other Alaskan and Canadian
wolves; nonetheless, data currently available suggest that the Alexander Archipelago
wolf is distinct from wolves in northern Alaska and the Yukon (Shields 1995). The
genetic relations between C. l. ligoni and wolves from coastal and interior British
Columbia, Montana, and Minnesota have not been investigated. Shields (1995)
speculates, based on the revised wolf taxonomy of Nowak (1996), that the Alexander
Archipelago wolf may show genetic affinity with historic wolf populations from coastal
British Columbia and Vancouver Island.
Figure 3—Current distribution of gray wolf subspecies in North
America reflecting the revised taxonomy proposed by Nowak
(1996): (1) Canis lupus arctos, (2) C. l. baileyi, (3) C. l. lycaon,
(4) C. l. nubilus, and (5) C. l. occidentalis. Map is adapted from
Nowak (1996) and range boundaries are approximate.
Shields (1995) observed no evidence that genetically unique subpopulations of
wolves occur on individual islands within the Alexander Archipelago. One DNA
cytotype occurred on six different islands, thereby suggesting that some gene flow
exists across water barriers. Alternatively, 12 DNA cytotypes occurred among Alex-
ander Archipelago wolves, indicating that genetic diversity was not being reduced by
bottlenecking (Shields 1995). These data are limited because of small sample sizes,
and no analysis of allelic frequency was possible. In addition, analyses of nuclear
DNA, rather than mtDNA, may be necessary to resolve questions about gene flow
and rates of migration within the Alexander Archipelago wolf population. Data sug-
gest that migration events have occurred in the past, but are uninformative about
the frequency of migration because mtDNA genotypes are transmitted intact through
the female germ line and may proliferate in a population through a small number of
matriarchal lines (Lehman et al. 1991).
Ecology Home range and pack size—Wolves are social animals that travel in packs and
actively defend territories from encroachment by other individuals or packs (Mech
1970). In southeast Alaska, minimum convex polygon (MCP) home ranges for wolf
packs on Revillagigedo Island averaged 279 square kilometers (108 mi2; range 79 to
447 square kilometers [30 to 170 mi2], n = 7; Smith et al. 1987). On Prince of Wales
and Kosciusko Islands, pack home ranges (average of 95-percent adaptive kernel
[Worton 1989] and 95-percent MCP home ranges) based on a minimum of 40 inde-
pendent radio locations averaged 280 square kilometers (109 mi2; range 101 to 419
square kilometers [39 to 163 mi2], n = 7; Person, in prep.). Core areas where wolf
activity was concentrated, were estimated to be 124 square kilometers (48 mi2)
and were 55 to 60 percent smaller than total home ranges (based on the average
of 75 percent adaptive kernel and 75 percent MCP home ranges).
MCP home ranges of packs (based on a minimum of 40 independent radio locations)
during the denning and pup-rearing period (15 April-1 August) were about 50 percent
smaller than during other times of the year (Person and Ingle 1995). This result con-
trasts with other studies that note wolf home ranges during summer and winter to be
similar in size (Fritts and Mech 1981, Fuller 1989, Potvin 1988, Van Ballenberghe et
al. 1975). Summer home ranges reported for wolves on Prince of Wales Island were
not unusually small when compared with summer home ranges reported elsewhere
for well-established wolf packs that primarily preyed on deer; however, winter home
ranges were substantially larger than those reported in other studies. For example,
Fuller (1989) reports average winter and summer home ranges of 116 and 110
square kilometers (45 and 43 mi2), respectively, for wolves in a portion of Minnesota
where deer density was high. On Prince of Wales Island, summer home ranges for
five wolf packs averaged about 100 square kilometers (39 mi2) whereas winter home
ranges for the same packs averaged about 240 square kilometers (93 mi2; Person,
Pack sizes on Revillagigedo Island averaged 5.4 wolves (range 2 to 12, n = 8; Smith
et al. 1987). Pack sizes on Prince of Wales and Kosciusko Islands were larger,
averaging 7 to 9 wolves (range = 2 to 12, n = 8; Person and Ingle 1995) in early
autumn before the trapping season. Pack size and composition can change over
time, with individuals or small groups of wolves occasionally splitting from the main
pack, only to rejoin it days or weeks later (Person, in prep.). Indeed, all members of
a wolf pack rarely are observed together, except during winter. Pack sizes are difficult
to estimate unless repeated, direct observations are made.
Food habits—Wolves prey on large ungulates such as deer, mountain goats
(Oreamnos americanus), moose (Alces alces), caribou (Rangifer tarandus), and elk
(Cervus elaphus) (Mech 1970). In southeast Alaska, deer are the primary prey on the
islands and portions of the mainland (Garceau 1960, Kohira 1995, Smith et al. 1987).
In areas of the mainland where deer are scarce, important prey include beaver
(Castor canadensis), mountain goat, and moose (Smith et al. 1986a, Wood 1990).
Wolves in southeast Alaska have access to spawning salmon during late summer
and early autumn (Kohira 1995, Smith et al. 1986b, Wood 1990). Wolves also feed
opportunistically on harbor seals (Phoca vitulina), mustelids, small mammals, birds,
and marine invertebrates (Garceau 1960; Kohira 1995; Merriam 1966, 1968; Smith
et al. 1986b).
Data from 182 microscopically examined wolf scats collected between autumn 1992
and autumn 1994 on Prince of Wales Island showed deer remains in 90 percent of
the scats (Kohira 1995). Deer occurred exclusively in 45 percent of the scats, but
often were found in conjunction with beaver (30 percent of the scats). The remains
of beaver, the only other prey species that occurred with >10 percent frequency,
were found in 31 percent of the scats, with the highest incidence occurring during
late spring. Only one scat contained exclusively beaver remains. Other prey con-
sumed in small quantities included black bear (Ursus americanus), mink (Mustela
vison), marten (Martes americana), river otter (Lutra canadensis), small mammals,
birds, and fish (Kohira 1995).
In a sample of 363 scats from Revillagigedo Island, 74 percent and 20 percent of
the scats contained deer and beaver remains, respectively (Smith et al. 1987). About
30 percent of the scats collected in summer contained remains of fawns, accounting
for 41 percent of all scats containing deer.
We combined the data from Prince of Wales and Revillagigedo Islands (n = 545
scats) and estimated the percentage of prey species by volume in the diet by using
a regression model described by Floyd et al. (1978) and later modified by Weaver
(1993). Results of a Monte Carlo simulation (n = 1,000) of the regression function
indicated that 77 percent (SE = 10 percent) of the diet was composed of deer. Beaver
composed only 13.7 percent (SE = 9 percent) of the diet, and all other species
contributed <10 percent. These data strongly suggest that wolves occurring on the
islands of southeast Alaska depend on the availability of deer and raise questions
about the ability of alternative prey to sustain wolves in the absence of deer.
Reproduction and denning—Age of first breeding is about 22 to 34 months, and
litters of 3 to 7 young usually are produced (Mech 1970, Stephenson 1989). Most
packs include a pair of breeding adults plus adults that may or may not breed.
Mean litter sizes in Alaska, as indicated by counts of blastocysts, range from
4.6 to 7.2, depending on prey availability per wolf (Boertje and Stephenson 1992,
Gasaway et al. 1992). The reproductive potential of wolves is high, and natural or
human-caused mortality, rather than the failure of a pack to breed or produce pups,
often is the major factor limiting the growth of wolf populations (Fuller 1989, Rausch
Wolves in southeast Alaska use dens, or areas near dens, from mid-April through
early July. Activity peaks around den sites from early May to the third week in June.
Of 14 den sites examined between October 1992 and October 1995, all were in
old-growth forest within 100 meters (330 ft) of fresh water (Person and Ingle 1995).
One den was under a large log; all others were in cavities beneath the roots of large
trees (mean diameter at breast height >90 centimeters [35 in]). Eight of 11 dens
visited during the year they were used were adjacent to ponds or streams with
active beaver colonies.
Litter sizes estimated from direct observations at dens on Prince of Wales Island
ranged from one to six pups ( = 3.7, SE = 1.8, n = 6) during the 1995 denning
period (Person, in prep.). Direct observations made during early autumn revealed
that only one pup from the total number seen at the dens had disappeared, sug-
gesting high pup survivorship during summer. This outcome may be a consequence
of the availability of salmon at the time pups are weaned, enabling them to exploit an
additional food supply at a time when they are inexperienced hunters and occupy
a low position in the dominance hierarchy of the pack.
Mortality—Humans account for a high percentage of mortality in both protected and
heavily exploited wolf populations (Ballard et al. 1987, Fuller 1989, Peterson et al.
1984). In southeast Alaska, 1,163 wolves were reported killed by hunters and trappers
from 1990 to 1995. About 60 percent of the mortality was from trapping (including
snaring) and 40 percent from hunting (i.e., ground shooting).1The highest reported
mortality occurred in December and January during the peak of the trapping season.
In exploited populations, mortality from natural causes (e.g., starvation, accidents,
disease, and fighting) is small, typically averaging 5 to 10 percent per year (Fuller
1989). A more substantial cause of mortality results from unreported or illegal killing
of wolves by people. Of 17 radio-collared wolves on Prince of Wales Island that died
during a 3-year study, nine (53 percent) were legally killed by humans, five (29 per-
cent) were killed by humans, but not reported, and three (18 percent) died from
natural causes (Person, in prep.). Considering the additive effects of natural and
unreported mortality, total mortality could be 35 to 50 percent higher than that
reported. Some bias may exist against reporting legally killed wolves with radio
collars; therefore, the actual unreported mortality from hunting or trapping may be
less than that suffered by radio-collared wolves. Nevertheless, reported mortality
may substantially underestimate total mortality.
The highest annual harvest of wolves in southeast Alaska is in GMU 2. To estimate
mortality we pooled sex and age classes and used a staggered-entry Kaplan Meir
survival rate estimation method (Pollock et al. 1989) with 2-week intervals. Total
mortality (natural, legal, and illegal harvest) in a sample of 19 radio-collared wolves
from autumn 1993 to spring 1994 was 61 percent (SE = 11 percent); mortality in a
sample of 13 radio-collared wolves from autumn 1994 to spring 1995 was 38 percent
(SE = 13 percent) (Person and Ingle 1995; Person, in prep.). Lower mortality in the
second year reflected a shift in trapping effort to the southern half of GMU 2, away
from the study area. From these estimates, the most probable average annual
mortality in the study area on Prince of Wales and Kosciusko Islands for June 1993
to June 1995 was 50 percent (SE = 13 percent, total sample for both years was 24
1Alaska Department of Fish and Game. Unpublished data.
On file with: Division of Wildlife Conservation, Douglas, AK
Table 1—Number of wolves per 1000 square kilometers (390 mi2) reported
during fall and winter in various areas of North America where deer are the
LocationYears Wolves Source
Prince of Wales Island
Van Ballenberghe et al. 1975
Nelson and Mech 1986a
Pimlott et al. 1969
Pimlott et al. 1969
Hatter and Janz 1994
Hebert et al. 1982
Person, in prep.
Figure 4—Distribution of wolf numbers in southeast Alaska by game
management units (GMUs) as predicted by habitat capability for deer
(USDA Forest Service 1991).
The variance of this model can be approximated by the following equation:
assuming that s and b are not significantly correlated and are density independent
(see appendix 2 for the derivation of these models). The model also represents a
point estimate in time and assumes the wolf population is closed (a reasonable
assumption for Prince of Wales and Kosciusko Islands). Survival rate is the most
important factor influencing λ, with birth rates of secondary importance. Substituting
0.5 (SE = 0.13) for s and 0.52 (SE = 0.50) for b (see appendix 2 for how these
values were estimated), we estimated λ to be 0.69 (SE = 0.23). Comparing this
estimate with a simulated population with λ = 1 and SE = 0.23, the results are dif-
ferent (t = 16.3, df = 14, P < 0.001), thereby indicating that the population declined
between June 1993 and June 1995. We caution that because of the limitations of the
data, λ should be interpreted only with respect to the significance of the difference
from one and not as a precise estimate of the actual value. These results are con-
sistent with observations made in the field by biologists and trappers who believe that
wolves on Prince of Wales and Kosciusko Islands were at a population peak during
winter 1992-93 and have declined since, owing primarily to trapping and hunting (in
GMU 2, 86, 105, 103, 85, and 99 wolves were reported killed during the 1991-92,
1992-93, 1993-94, 1994-95, and 1995-96 trapping seasons, respectively4). We
cannot extrapolate from these results to assess population trends for wolves outside
of GMU 2. Trapping records indicate that harvest intensity is probably greatest in
GMU 2, but wolf numbers could be increasing or remaining stationary in other areas
while declining in GMU 2.
Wolf-deer interactions—Much emphasis has been placed on the role of wolf
predation in limiting or regulating ungulate populations (Gasaway et al. 1983,
1992; Mech 1970; Messier 1994; Peterson and Page 1988; Theberge 1990; Van
Ballenberghe and Ballard 1994). The focus has been primarily on the effects of
predation on moose and caribou. Most authors have concluded that wolf predation
can limit moose and caribou populations under some circumstances, particularly
when multiple predators, multiple prey species, or severe weather are involved
(Gasaway et al. 1992, Van Ballenberghe and Ballard 1994).
Much less attention has been directed toward the quantitative analysis of predator-
prey dynamics with respect to wolves and deer (Fuller 1989, Hatter and Janz 1994,
Lewis and Murray 1993), although much is known qualitatively about the effects of
wolf predation on deer (Fuller 1991; Hoskinson and Mech 1976; Nelson and Mech
1981, 1986b). In Minnesota, severe winter weather caused a decline in numbers of
white-tailed deer (Odocoileus virginianus) in some areas (Mech and Karns 1977).
Wolves delayed the recovery of these deer populations, but they did not cause the
original decline (Mech and Karns 1977). Van Ballenberghe and Hanley (1984) hypoth-
esized that three things tend to increase the probability of wolves suppressing deer
populations in southeast Alaska: (1) periodic winters of deep snow that result in
increased killing, (2) the patchiness of winter habitat for deer that reduces the time
4Alaska Department of Fish and Game. Unpublished data.
On file with: Division of Wildlife Conservation, Douglas, AK
Var VarVar( )
( )s ( )b,
wolves spend searching for prey, and (3) the availability of marine food resources
that buffer wolves from cyclic fluctuations in their terrestrial prey. Mech and Karns
(1977) report that in a declining deer herd, surviving deer inhabit overlapping edges
of wolf-pack territories and wolves tend not to hunt in these areas, or buffer zones, to
avoid potentially fatal encounters with their neighbors. Klein (1981) suggests that in
southeast Alaska, where the boundaries of many pack home ranges are shorelines,
the potential for wolves to reduce deer numbers is increased because there are
fewer buffer zones among territories.
In the early 1960s, four wolves were transplanted to Coronation Island in an exper-
iment to observe the effects of wolf predation on Sitka black-tailed deer (Klein 1996).
The wolves reached a peak population of 13 after 4 years and apparently reduced
deer numbers dramatically. As their food supply dwindled, the wolves declined in
number. The last wolf on Coronation Island was shot in the late 1960s, ending the
experiment (Merriam et al. 1994), and the deer population has since rebounded.
Coronation Island is too small (78 square kilometers [30 mi2]) and supports insuf-
ficient alternative prey to sustain wolves in the absence of deer. The experiment
suggests that islands in southeast Alaska comparable in size to Coronation Island
are unlikely to support wolves permanently unless they are near other land masses,
thus enabling wolves to periodically emigrate or immigrate. Further, deer density
probably was near carrying capacity when wolves were introduced (Klein 1996).
Consequently, the resiliency of the deer population to wolf predation (i.e., its ability
to offset mortality from predation by increasing reproduction) probably was limited
because of lower reproductive rates in deer resulting from the overbrowsed and
degraded condition of the habitat (Klein 1996). The results from Coronation Island
suggest that if wolves and deer are confined to small islands of habitat, then predator-
prey disequilibria may result.
The stability of deer and wolf populations depends on several factors, including the
rate of predation by wolves, the number of deer killed by hunters and other predators,
and the ratio of deer to wolves. The most important determinants, however, are the
rate of growth of the deer population and the frequency of stochastic events such
as severe winters. These latter factors are related to habitat quantity and quality.
For example, as deer or other ungulate populations approach carrying capacity,
intraspecific competition for food will reduce successful reproduction and increase
chronic mortality (Caughley and Sinclair 1994; McCullough 1979, 1987) and thus
reduce the fraction of the deer population that can be exploited by wolves and other
predators without causing the prey population to decline. A predator-prey equilibrium
(McCullough 1979) may result, in which predation reduces deer density sufficiently
to stimulate greater reproduction in deer, thereby making more deer available to the
predators. The stability of such an equilibrium is contingent on the availability of
suitable habitat for deer, such that the deer population can maintain a rate of growth
sufficient to offset losses to predation. Stochastic events, such as severe winters or
reductions in the abundance and quality of suitable habitat, may perturb the conditions
of equilibria and result in widely fluctuating wolf and deer populations. By limiting
immigration and emigration of both deer and wolves, the insular conditions commonly
encountered in southeast Alaska increase the probability that population fluctuations
may end in local crashes of deer populations. The fluctuations exhibited by the
population trajectories of wolves and moose on Isle Royale, MI, certainly illustrate the
consequences of insularity with respect to a mammalian predator-prey community
(Peterson and Page 1988).
The recovery of deer populations depends on the reproductive potential of the sur-
vivors, immigration from other areas, and the numerical response of wolves and other
predators (including humans) to declining deer density. If alternative species of prey
exist so that wolf numbers do not decline sufficiently to permit deer to increase (or if
mortality of deer from hunting or other predators remains high), or if the reproductive
potential of the deer population is reduced because of habitat loss, then recovery of
deer populations after a crash could take a long time (Van Ballenberghe and Hanley
1984). For example, deer populations in GMU 3 (including Kuiu, Kupreanof, and
Mitkof Islands) crashed during the early 1970s as a result of a series of hard winters
(Olson 1979). In 1975, deer hunting in GMU 3 was halted completely and remained
closed throughout most of the GMU for 18 years. Today, deer are still at levels far
below carrying capacity over most of the GMU (Kirchhoff 1994). The persistently low
deer numbers illustrate that interactions of deer and their predators (in this instance,
wolves and black bears) may, under some circumstances, retard recovery of the deer
population. Although deer also declined in GMU 2 following the same hard winters,
snowfall on the more southern islands was not as heavy, and more of the landscape
in GMU 2 was sheltered by productive old-growth forest. In marked contrast to the
result in GMU 3, deer in GMU 2 did not decline as far and rebounded much more
quickly after the hard winters.
The rate at which wolves kill large mammals ranges widely given the availability,
size, and vulnerability of prey. Direct estimates of predation rates on Sitka black-
tailed deer are not available. We therefore estimated predation rate by deriving a
series of equations incorporating average weights of deer, proportion consumed by
wolves, average weights of wolves, and their daily requirements for food intake. We
calculated the weight of edible meat from each deer carcass:
wcons = weight in kilograms of meat consumed per deer carcass,
= average weight in kilograms of adult deer,
= proportion of adult deer carcass consumed,
= proportion of adults in the deer population,
= average weight in kilograms of fawns,
= proportion of fawn carcass consumed, and
= proportion of fawns in the deer population.
Next we estimated the weight of deer meat consumed per wolf per year:
Cwolf = weight of deer meat in kilograms consumed per wolf per year,
wwolf = average weight in kilograms of wolves,
)( )() ()()( ),
wolf wolfwolf wolf
= ( )()( )(,
= daily per capita food requirement in kilograms of food per kilogram of wolf,
= proportion of deer in the diet.
Finally, an estimate of the number of deer killed per wolf per year (PR) was
and its variance:
(for the parameter estimates used, see appendix 3). From inputting of parameter
values appropriate for wolves and Sitka black-tailed deer in southeast Alaska, we
estimated predation rate to be 26 deer per wolf per year (SE = 4.1). This calculation
assumed that wolves exhibit a type I functional response (Holling 1966) resulting in
the rate of predation remaining constant despite changes in the density of prey.
To evaluate the potential for equilibria between wolves and deer given our estimate
of predation rate, we used a model proposed by Keith (1983) that combines pre-
dation rate, rate of growth of the prey population, and the fraction of the recruitment
to the prey population removed by hunting or other predators to calculate the number
of ungulate prey per wolf (equilibrium ratio) necessary for equilibrium to occur:
N = number of deer per wolf necessary for equilibrium prior to parturition,
PR = number of deer killed per wolf per year,
= finite rate of increase of deer in the absence of predation measured prior to
H = fraction of annual recruitment to the deer population removed by hunting or
Van Ballenberghe and Hanley (1984) used this model, or equilibrium function, to
explore the influence of changes in λ on equilibrium ratios brought about by habitat
changes. They concluded that reductions in λ dramatically increase the number of
deer per wolf needed for equilibrium and reduces the probability that stability between
deer and wolves will occur. Unfortunately, the equilibrium function (eq. 8) is deter-
ministic and does not reflect the uncertainty associated with the input parameters.
We altered the model to correct for this attribute by replacing the input variables
with random samples drawn from the probability density function (pdf) of each
parameter as defined by their sampling distributions or some appropriate hypo-
f, for and
= vector of equilibrium solutions for the number of deer per wolf necessary
for equilibrium prior to parturition,
= vector of random samples from the pdf of PR with µ = a and σ = b,
= vector of random samples from the pdf of λ with µ = a and σ = b, and
= vector of random samples from the pdf of H with µ = a and σ = b.
The model assumes that predation by wolves and other predators is a source of
additive mortality, that λ encompasses all reproduction, immigration, emigration,
and compensatory mortality, and that each input parameter is an independent ran-
dom variable. The variance of may be approximated by the following equation:
A Monte Carlo simulation was conducted to estimate the equilibrium ratio for deer
and wolves in GMU 2 using the following parameter estimates:
—Connolly’s (1981) review suggests finite rates of increase as high as 1.4 for
established populations of mule and black-tailed deer. For these model simulations,
we assumed a relatively high λ (µ = 1.3, SE = 0.05) because deer seem to be
well below carrying capacity in GMU 2, based on browse and pellet group surveys
(Kirchhoff 1993). About 95 percent of the values fell between 1.2 and 1.4.
—The analysis in this report suggests a predation rate of 26 deer per wolf per
year in areas where deer are the primary prey species (>75 percent of the diet).
Actual rates of killing attained will be higher or lower in any given year depending on
availability of alternate prey, fawn-to-adult ratios in the kill, and snow conditions that
make deer more or less vulnerable to wolves. Our estimate is higher than predation
rates reported for wolf predation on white-tailed deer (Fuller 1989) because Sitka
black-tailed deer are about 25 to 30 percent smaller. For these model simulations,
we estimated the standard error for predation rate to be 4.1. Thus, 95 percent of the
values fell between 18 and 34 deer per wolf per year.
R a b
PR a b ,
λa b ,
Ha b ,
Var Var Var
—We assumed that the only other significant predation on deer in GMU 2 was
from hunting. Although black bears do prey on fawns, the extent of this predation is
unknown and no estimate is possible. Consequently, equilibrium ratios predicted by
the simulation may underestimate the true values if bear predation is significant. In
our simulation, the value H refers to the proportion of the annual increment (popula-
tion increase) of deer removed each year by hunting. To obtain this value, we esti-
mated the current deer population on Prince of Wales Island from surveys of deer
pellet groups on winter range (productive old-growth forest below 450 meters [≈1,500
ft]) (Kirchhoff 1993, Kirchhoff and Pitcher 1988). From a standard conversion factor
(Kirchhoff 1990), mean pellet-group density was converted to an equivalent deer
density of 14.6 deer per square kilometer (38/mi2; n = 11 watersheds). Assuming
deer winter range is approximately equal to productive forest land, the current popu-
lation of deer on Prince of Wales Island may be close to 42,000 deer (14.6 deer per
square kilometer 2895 square kilometer of winter range). Because this population
estimate is lower than the habitat capability of 55,000 deer based on the USDA
Forest Service (1996) model, we assumed a high finite rate of increase in the popu-
lation (1.3). With the population growing at this rate, the annual increment is 12,600
deer (0.342,000). The average deer harvest from 1993 to 1995 on Prince of
Wales Island was about 2,700 deer, which represented 21 percent of the annual
increment.5In the model, we assumed the proportion of the annual increment
removed by hunting each year was 0.21 (SE = 0.05). The large standard error
(25 percent of the mean) reflects the uncertainty associated with this parameter
estimate. Population estimates based on pellet groups generally have low precision,
and the error associated with the habitat suitability model for deer is compounded
within its structural and functional components. In addition, hunting effort and suc-
cess may differ widely owing to weather conditions.
The results of 10,000 simulations predicted a mean equilibrium ratio of 113 deer per
wolf (SE = 29) (fig. 5a). A cumulative frequency plot of the results (fig. 5b) shows that
there would be a 95-percent probability of equilibrium occurring if the ratio was 170
to 180 deer per wolf. Thus, if the objective is to maintain adequate prey to sustain
250 to 300 wolves on Prince of Wales and Kosciusko Islands (the current estimated
population) with a 95-percent chance of equilibrium and still support subsistence and
sport deer harvesting and current levels of wolf harvest, then sufficient habitat for
42,500 to 54,000 deer is needed. For comparative purposes, the habitat suitability
model for deer (USDA Forest Service 1996) predicts the current habitat capability of
Prince of Wales and Kosciusko Islands (on Federal lands) to be about 55,000 deer
(margin of error unknown).
We caution that equilibrium does not necessarily imply stability, which is a measure
of the tendency for a system to return to a condition of equilibrium after being per-
turbed (May 1974). Nonetheless, a large equilibrium ratio, particularly for wolves and
deer confined to small geographic or habitat islands, would suggest a low probability
of stability between predator and prey. The potential for equilibrium, and consequently
stability, is diminished if habitat loss causes residual deer populations to exist at a
5Alaska Department of Fish and Game. Unpublished data.
On file with: Division of Wildlife Conservation, Douglas, AK
lower density that is also closer to carrying capacity. Increased intraspecific compe-
tition for food within the remaining deer habitat will likely lower reproduction. For
example, if λ were reduced to 1.15 and all other parameters remained the same,
then the mean equilibrium ratio predicted by equation (9) is 219 deer per wolf
(SE = 82). Over 380 deer per wolf would be needed for a 95-percent probability
of equilibrium. The implication of this example is that equilibrium, and therefore
stability, would be difficult to achieve and tenuous at best if carrying-capacity for
Our review raises a number of questions about the long-term viability and distribution
of wolves in southeast Alaska. The Alexander Archipelago wolf exists in limited
numbers in an insular environment undergoing rapid change. Issues focus on three
general areas: (1) a decline in carrying capacity for deer; (2) the effect of road use by
humans on mortality and displacement of wolves; and (3) continued exploitation of
wolves. These changes are related to past and planned timber harvesting on Federal
and private lands.
Figure 5—(A) Frequency plot of the number of deer per wolf
(equilibrium ratio) predicted by Monte Carlo simulation of the
equilibrium function. (B) Cumulative probability plot showing the
number of deer per wolf needed for 80- and 95-percent probability
Prey Populations Deer carrying capacity—Maintaining viable, well-distributed wolf populations ulti-
mately will depend on maintaining habitat to support a relatively abundant, well-
distributed, and stable population of deer. Short-rotation clearcut logging of old-
growth forests in southeast Alaska will reduce habitat capability for Sitka black-
tailed deer (USDA Forest Service 1991, 1996). This conclusion is supported by an
extensive body of research spanning 30 years on forest succession following logging
(Alaback 1982, Alaback and Herman 1988, Alaback and Juday 1989), silvicultural
practices (Alaback and Tappeiner 1984, DellaSala et al. 1994, Doerr and Sandburg
1986, Nyberg et al. 1989), deer habitat relations (Bloom 1978, Kirchhoff and Schoen
1987, Klein 1964, Klein and Olson 1960, Schoen and Kirchhoff 1985, Schoen et al.
1988, Van Ballenberghe and Hanley 1984, Wallmo and Schoen 1980, Yeo and Peek
1992), and nutritional ecology of deer (Hanley 1984, 1987; Hanley and McKendrick
1985; Hanley et al. 1987, 1989; McArthur et al. 1993; Parker et al. 1993) in southeast
Alaska. Short-rotation clearcut logging replaces productive old-growth forest, which is
important deer winter habitat, with even-aged, second-growth stands of much lower
habitat value for deer (Wallmo and Schoen 1980). Although young clearcuts may
temporarily produce abundant forage, typically it is of poorer nutritional quality than
forage available in old-growth forest stands (Hanley et al. 1989), and some under-
story plant species may never reappear on sites under short-rotation, even-aged
management (Alaback 1982). Moreover, what forage is present is not available to
deer during periods of deep snow.
Within 30 years of clearcutting, regenerating conifers shade out most understory
vegetation (Alaback 1982), creating poor habitat conditions for deer (Wallmo and
Schoen 1980). These stands represent a serious problem for deer because the
habitat is poor in all seasons, and these conditions persist for a long time (150 to
200 years) (Alaback 1982, Wallmo and Schoen 1980). Although precommercial
thinning may delay crown closure in second-growth stands, the improvements in
their value as winter habitat for deer are modest and short-lived (<20 years) (Alaback
and Tappeiner 1984).
Old-growth forest stands composed of relatively tall, large-canopied trees are par-
ticularly important to deer during severe winters. High-volume old growth intercepts
snow more effectively than low-volume old growth (Kirchhoff and Schoen 1987) and
receives disproportionately high use by deer during periods of deep snow (Schoen
and Kirchhoff 1990). Under the current Forest plan (USDA Forest Service 1996, alt.
9), about 61 percent of the original productive old growth on Federal lands in GMU 2
will be cut within the next 100 years. About 15 to 20 percent of the land in GMU 2
is owned by Native corporations under the provisions of the Alaska Native Claims
Settlement Act of 1971 (ANSCA). Between 65 and 81 percent of the original produc-
tive old growth on Native lands in GMU 2 will likely be cut within the next 20 years
(USDA Forest Service 1991). Logging on both Federal and Native lands in GMU 2
will likely result in the conversion to young-growth forest of 75 to 80 percent of the
original productive old-growth forest within 100 years. Timber harvesting on Federal
and Native-owned lands in GMU 3 will remove about 48 to 50 percent of the original
productive old-growth forest and, in combination with logging in GMU 2, will potentially
affect over 65 percent of the wolf population in southeast Alaska (fig. 4).
This loss of habitat will likely reduce deer populations, and consequently affect deer-
wolf dynamics. As our simulations have shown, habitat changes that lower repro-
ductive rates in deer will reduce the probability of equilibria between wolves and deer
and increase the sensitivity of equilibria to perturbation. In planning timber harvests,
efforts to minimize the loss of long-term deer carrying-capacity at geographic scales
consistent with the size of wolf pack home ranges, would likely reduce the probability
of disequilibria between wolves and deer.
Forest fragmentation—In southeast Alaska, forests are a mosaic of high-, mod-
erate-, and low-volume stands interspersed with open muskegs and alpine terrain.
This natural fragmentation is intensified when clearcut logging converts individual
forest stands into a patchwork of smaller old-growth stands isolated by clearcuts.
During periods of deep snow, deer confined to isolated stands of old-growth forest
consume the available food resources and suffer higher rates of mortality from mal-
nutrition than deer in unfragmented old-growth forest (Kirchhoff 1994). If wolves also
are present, then forest fragmentation may focus wolf predation on specific sites
where deer are concentrated and relatively vulnerable (Hebert et al. 1982, Janz
1989, Nelson and Mech 1986b). For example, logging roads in southeast Alaska are
typically constructed along the coastline or valley bottoms, with clearcuts spaced out
along the roads on the lower hillsides. Residual patches of old-growth forest among
clearcuts are connected by these roads, which are occasionally traveled by wolf
packs. Deer concentrate in these patches during winter when snow depths prevent
them from using open areas, and may suffer higher mortality from wolf predation.
From Vancouver Island, British Columbia, McNay and Voller (1995) report the
efficiency of predation by wolves to be high in fragmented, heavily logged land-
scapes. They calculated the annual survival rate for adult resident deer at low
elevations to be 73 percent and concluded this level, if continuous, is inadequate
to sustain the deer population. In contrast, migratory deer, which made greater use
of high-elevation habitats and completely avoided low-elevation second-growth
forests in winter, were able to avoid predation and consequently had high rates
of survival (95 percent). McNay and Voller (1995: 145) cautioned that:
...Retention of older, intact forests is basic to rebuilding deer populations. Forest har-
vesting, hence, road building and spatial isolation of winter habitats, may intensify
predation on 1 segment of the deer population (resident deer) and indirectly impede
recruitment to the other segment (migratory deer). The result likely contributes to
declines in deer populations and an overall loss of population resiliency.
Avoiding further habitat fragmentation, especially at lower elevations and in the beach
fringes where wolves commonly prey on deer in winter, should enhance the likelihood
of maintaining deer populations in southeast Alaska. Alternative methods of timber
harvest, such as single-tree selection could minimize the effects of logging, especially
in areas of critical habitat for deer.
Studies in Wisconsin, Michigan, Ontario, and Minnesota indicate a strong relation
between road density and the presence or absence of wolves (Fuller 1989, Jensen
et al. 1986, Mech et al. 1988, Thiel 1985). These studies show that wolves generally
failed to survive in areas with road densities >0.6 kilometer per square kilometer
(0.9 mi/mi2) whereas they persist in similar areas with lower densities of roads. Mech
(1989) speculated that excessive mortality experienced by wolves in roaded areas is
compensated for by individuals that disperse from adjacent roadless areas. Citing the
expansion of wolf populations in Minnesota, Wisconsin, and Michigan, Mech (1995)
suggested that road density is not really a deterrent, provided human populations
are tolerant of the presence of the wolves. Nevertheless, in an analysis of habitat
selection by colonizing wolves, Mladenoff et al. (1995) report that roadless areas
are preferred, and occupation of roaded landscapes occurs after roadless areas are
saturated. In all the circumstances described in which wolves successfully inhabited
roaded or developed areas, adjacent roadless areas were present.
There are several plausible explanations for the absence of wolves in densely roaded
areas. In some instances, wolves may avoid roaded areas depending on the type of
use the roads receive (Thurber et al. 1994). In most instances, their absence may be
a direct result of mortality associated with roads (Berg and Kuehn 1982, Mech 1977,
Van Ballenberghe et al. 1975). Substantial human-caused mortality can occur even
when wolves are completely protected from legal hunting and trapping (Fritts and
Mech 1981, Fuller 1989). Fuller (1989) reported that 80 percent of the identified
mortality in a protected study area in Minnesota was human caused. Similarly, Mech
(1989) noted that 60 percent of the mortality in a roaded study area was human
caused, whereas no mortality was human caused in an adjacent roadless area.
Over the past 40 years, timber harvest has been most extensive in GMUs 2 and 3.
Road densities in 50 percent of the Wildlife Analysis Areas (WAA) in GMU 2 and
17 percent of the Wildlife Analysis Areas in GMU 3 currently exceed 0.6-kilometer-
per-square kilometer (USDA Forest Service 1996, Tongass land management plan
geographic database), a density suggested to be inimical to wolves in other areas.
The current Forest plan for the Tongass National Forest calls for road densities 3 to
4 times higher than exist today in GMUs 1A, 2, and 3 (fig. 6). Although wolves on
Prince of Wales Island currently use areas with road densities greater than 0.6
kilometer per square kilometer (0.9 mi/mi2), their “core areas” are generally located in
the least densely roaded portion of the home range, and the wolf activity that does
occur in densely roaded areas occurs primarily at night (Person, in prep.).
Figure 6—Kilometers of existing road and projections of
future kilometers of road to 2090 for GMUs 1A, 1B, 2, and
3 (USDA Forest Service 1991). The GMUs encompass
most of the range of wolves in southeast Alaska. The data
do not include roads on state, private, or Native lands.
Based on harvest statistics from 1990 to 1995 in GMUs 2 and 3 and tabulated by
WAA, wolves experienced significantly higher mortality from hunting and trapping
in WAAs with higher road densities (r = 0.59, P < 0.001, n = 56). We calculated road
density by using the area within a WAA below 370 meters (≈1,200 feet) elevation
as the denominator. Wolves spend most of their time at low elevations (Person,
in prep.), and calculations of road density should reflect this relation. Analysis of
variance indicated a significant effect of road density on the 6-year average wolf
harvest (P < 0.002, df = 49) reported for WAAs in GMUs 2 and 3 after controlling for
the size of WAAs (a square root transformation of wolf harvest was used to stabilize
the variance). Harvest rates increased sharply in WAAs exceeding 0.3 kilometer of
road per square kilometer (0.49 mi/mi2) (fig. 7a and 7b). Indeed, in two areas on
Prince of Wales Island (Kasaan Peninsula and Big Salt Lake), trappers and hunters
eliminated all known resident wolves (Person, in prep.). In both areas, road density
exceeded 0.6 kilometer per square kilometer. Moreover, two female wolves with radio
collars, each paired with a male, dispersed into the vacant areas and attempted to
establish packs. Both wolves were killed (one legally, one illegally) by hunters within
We regressed the 6-year (1990-95) average wolf harvest from WAAs in GMUs 2
and 3 against the linear kilometers of road within a WAA. To stabilize the variance,
we used a square-root transformation of the harvest data for wolves. We used linear
distance of road rather than road density, because most roads are along valley
bottoms and through forest stands that are the most productive for timber and deer.
Overall road density in a WAA may be low, but the roads that do exist are likely to
go through or be located adjacent to areas used heavily by wolves.
Figure 7—(A) Mean numbers of
wolves killed per Wildlife Analysis
Area (WAA) during the regular
trapping and hunting seasons by
the density of roads contained
within a WAA after controlling for
the size of the WAA. Harvest
numbers represent the 6-year
average harvest from 1990 to
1995. The filled circles represent
the mean and the bars represent
the 95-percent confidence intervals.
(B) Plot showing the mean number
of wolves killed per Wildlife Analysis
Area (WAA) during the regular trap-
ping and hunting seasons by the
density of roads contained within a
WAA after controlling for the size
of the WAA. Harvest numbers
represent the square root of the
6-year average harvest from 1990
to 1995 to stabilize the variance.
The filled circles represent the
mean and the bars represent the
95-percent confidence intervals.
Regression analysis indicated a significant positive relation between kilometers of
road and wolf harvest (r2= 0.33 , df = 55, P < 0. 001) (fig. 8). The model predicted
that wolf mortality due to hunting and trapping increased twofold when the linear
distance of road exceeded 95 kilometers (59 mi), threefold when it exceeded 171
kilometers (106 miles), and fourfold when more than 234 kilometers (145 mi) of road
existed within a WAA. The average area below 370 meters (≈1200 ft) in elevation for
WAAs in GMUs 2 and 3 is 233 square kilometers; (89 mi2); therefore, the distance
values correspond to densities of road equivalent to 0.41, 0.73, and 1.0 kilometer
per square kilometer, respectively (0.66, 1.19, and 1.63 mi/mi2). We caution that
the model explains only 33 percent of the variation in the data, and the hyperbolic
prediction intervals around the regression lines are large; therefore, considerable
deviation from predicted values is likely to be observed.
Means by which wolf mortality associated with roads can be better controlled would
benefit from further investigation and consideration. This approach would include
evaluating management of road use. For example, infrared traffic sensors and direct
observations revealed that several roads posted as closed to motorized traffic on
Prince of Wales Island were frequently driven over, and in two cases, gates were
Figure 8—Plots showing regres-
sions of the square root of the
6-year average wolf kill from 1990
to 1995 for GMUs 2, 3, and 2 and
3 combined versus the linear
kilometers of road within a Wildlife
Analysis Area (WAA).
repeatedly vandalized in 1994 and 1995 to enable motorized access.6Further, the
construction of gates or the removal of culverts may not be sufficient to eliminate
all-terrain vehicle and snowmobile traffic.
Exploitation refers to the deliberate killing of wolves by people, either by trapping
(and snaring) or shooting. For discussion purposes we divide the wolf kill into two
categories, legal and illegal harvest. The legal harvest by hunters and trappers
accounts for most of the annual take, and is typically managed by adjusting season
lengths, harvest limits, or restrictions on methods of take. Illegal harvest includes
wolves that are killed but are either not salvaged or otherwise not presented to
ADFG for sealing of the pelts (as required by regulation). Illegal harvest is difficult
to estimate, monitor, and control.
Legal and illegal exploitation of wolves will almost certainly increase as human popu-
lations increase, and as expanded access to the interior of islands brings people and
wolves into more frequent contact. Furthermore, as human populations increase and
the demand grows for a deer population that, under current habitat management, is
simultaneously declining, wolves will inevitably be viewed by some as undesirable
competitors. These trends in human population, access, and rates of exploitation
are particularly evident on Prince of Wales Island. As recently as 1960, the human
population on Prince of Wales Island numbered slightly more than 1,000 people.
Air and ferry service were limited, and only a few kilometers of road existed outside
a handful of small communities. Currently the population on the island and in the
immediate area has grown to over 7,000 people (U.S. Department of Commerce
1996). The population has increased 13 percent since the 1990 census (U.S. Depart-
ment of Commerce 1996). Since 1954, about 4,800 kilometers (3,000 miles) of road
have been built on private and Federal lands on Prince of Wales Island.7 8These
roads, originally built for access to timber, now provide a transportation network
among communities and provide people with ready access to many otherwise in-
accessible areas of the island. This rapid rate of development is expected to con-
tinue. For the Ketchikan Area (GMUs 1A and 2), the kilometers of existing road will
double in 30 years; the Stikine Area (GMUs 1B and 3) likely will have twice as many
roads in 20 years (USDA Forest Service 1991).
Even though most of the wolves harvested in southeast Alaska are taken along the
shore from boats (55 percent), a large and growing percentage are now taken from
the road system (44 percent).9Roads on Prince of Wales Island access 19 of 22
Wildlife Analysis Areas, representing 78 percent of the land area in GMU 2. Except
for the Thorne River area, all WAAs on the Island are accessible by skiff. Due to the
relative ease of access and the continued efforts of trappers, wolves on Prince of
Wales Island are heavily exploited. Total annual mortality on the island may exceed
45 percent (Person and Ingle 1995) in some years.
6Person, D.K. Unpublished data. On file with the author.
7USDA Forest Service. GIS database. Tongass National
Forest, Regional Office, Juneau, AK 99801.
8Alaska Department of Fish and Game. Internal report.
9 Feb. 1995. On file with: Habitat Division, Ketchikan,
9Alaska Department of Fish and Game. Unpublished data.
On file with: Division of Widlife Conservation, Douglas, AK 99824.
Although trapping effort will undoubtedly diminish if wolf populations decline and
animals become harder to catch, the relative ease of access and liberal hunting and
trapping seasons may still result in harvest that is unsustainable. Deer populations
currently may be sufficiently high to enable wolf numbers to rebound after several
years of intensive trapping and hunting pressure. Nonetheless, if deer carrying
capacity continues to decline in the future, then the resiliency of wolf populations
likely will diminish, compounding the effects of overharvesting wolves by delaying
or reducing the potential for their population recovery. In areas experiencing heavy
mortality, regulatory changes may be needed to bring total wolf mortality within
sustainable limits. Based on our analysis of birth rates and population size for wolves
on Prince of Wales and Kosciusko Islands, we estimate the per capita birth rate for
wolves to be approximately 0.33 (SE = 0.15). The buffering effects of immigration
and emigration are probably limited for most of the wolves in southeast Alaska;
consequently, total annual mortality should not exceed reproduction to maintain
current population levels. Thus, to maintain current population levels, a level of
mortality (from all causes; including natural, legal, and illegal harvest) for wolves
in southeast Alaska is likely to be less than or equal to 30 to 35 percent.
Wolves, deer, and humans have coexisted in southeast Alaska for centuries and
perhaps thousands of years. The natural environment of the past succeeded in
supporting species that are of concern in the managed environment of the present.
Given the overwhelming complexity of ecological systems and the state of scientific
knowledge about them, we use the past as our default position, against which future
management plans are compared. We suggest that the following questions be con-
sidered in making management decisions that may effect wolves and their habitat.
Can long-term deer carrying capacity be maintained at levels similar to current
conditions after logging, particularly if clearcut logging is continued on a large
This is the most important consideration for the long term. Results from research
on postharvesting treatments of clearcut stands, such as precommercial thinning,
suggest that benefits to deer from thinning are short-lived. The current state of
knowledge offers no likely solutions to the problem of maintaining long-term deer
carrying capacity while using extensive short-rotation clearcut logging. Other timber
harvesting strategies, such as limited selection harvests, may ameliorate some of
the impacts of logging on deer carrying capacity, but the effect of these alternative
harvesting methods on deer is unknown at present.
If carrying capacity is reduced, then what is the minimum carrying capacity for
deer that will support a viable and well-distributed population of wolves?
We cannot suggest a minimum deer population because we do not know what would
constitute a minimum viable wolf population either demographically or genetically.
Many more data are needed concerning the population structure, genetic structure,
and predator-prey relations of wolves in southeast Alaska before this question can
be answered. This question also must be addressed separately for the major island
groups and possibly for individual sections of the mainland that are divided by major
glaciers or river drainages. Our analyses suggest that sustaining the estimated cur-
rent population of 250 to 300 wolves on Prince of Wales Island would require at least
42,500 to 54,000 deer (6 to 8 deer per square kilometer [16 to 21 deer/mi2]) for a
95-percent probability of equilibrium. As deer carrying capacity declines in the years
after extensive timber harvest, we would expect both deer numbers and reproductive
rates to decline as well. In this situation, a disequilibria between wolves and deer
likely would occur. Whether new equilibria would result is not known, but even if
they do, they probably would be more difficult to achieve and less stable. At lower
equilibria, there may be insufficient deer numbers to support current subsistence
and sport deer harvest levels.
If roads are built, can access be managed effectively to reduce wolf mortality?
Effective management of roads to control human access should help to reduce wolf
mortality; however, closing roads has not been fully successful because of practical
difficulties and opposition voiced by residents within the communities affected by road
closures. The demands of a growing human population for access to the Tongass
National Forest for commercial and recreational purposes likely will make manage-
ment of access even more difficult, while doing so will perhaps be even more impor-
tant. A systematic access management approach that considers the biological require-
ments of wolves would be beneficial. For example, not building roads, or closing
existing ones, in core activity areas or near denning sites would be beneficial. Of
course, this type of action would require reliable information concerning the move-
ments and activities of wolves in areas where access management is proposed.
Considerable time and expense would be necessary to collect such information.
Can wolf mortality from trapping and hunting be controlled by regulatory
Regulatory changes could have a marked effect on wolf mortality, nevertheless, the
effect may be limited by the logistical difficulties of enforcing regulations in an area
the size of the Tongass National Forest. Moreover, in areas where a large portion
of the human population is employed in industrial or agricultural occupations, the
number of wildlife law violations may be high (Berger and Daneke 1988). Educational
programs concerning wolves, such as those available in Minnesota (International
Wolf Center, Ely, MN), where a generation of people have grown up while wolves
have been protected, might help the public understand the need to comply with
regulatory changes in southeast Alaska and thus help to assure the effectiveness
of regulations in managing human-caused mortality of wolves.
Although this report presents a substantial body of scientific information concerning
wolves in southeast Alaska, the data are temporally and geographically limited in
scope and caution must be used when drawing long-term, forest-wide conclusions
from them. Nevertheless, we believe that continued extensive timber harvesting,
especially short-rotation clearcut logging, will have adverse consequences for wolves
and place some population segments at risk (particularly those wolves in GMU 2 and
perhaps those in GMU 3).
In our opinion, the long-term consequences of continued short-rotation, clearcut tim-
ber harvesting and the short-term consequences of road access and legal or illegal
wolf harvest result in increased risk to wolf viability. The Alexander Archipelago wolf
exists as a population, perhaps less than 1,000 animals, that is potentially divided
into smaller subgroups associated with different portions of the mainland and island
clusters. Our analysis suggests that the largest portion of the population inhabits
areas where extensive timber harvest has occurred or is planned. Maintaining a
minimum average density of deer equal to 5 deer per square kilometer (13 deer/mi2)
in wildlife analysis areas that currently support deer numbers greater than or equal
to 5 deer per square kilometer and where deer are the principal species of prey for
wolves would reduce long-term risk to wolf viability. This density of deer is likely to
support wolves and sustain the current subsistence harvest of deer by humans. In
other wildlife analysis areas where deer are the primary prey, but current densities
of deer are lower than 5 deer per square kilometer, maintaining deer numbers would
reduce long-term risk to wolf viability.
In addition, in biogeographic provinces where extensive timber harvesting has oc-
curred or where extensive harvesting is planned (see USDA Forest Service 1996 for
description of biogeographic provinces), maintaining large, unfragmented, and un-
roaded blocks of habitat within each biogeographic province where wolves occur
would reduce long-term risk to wolf viability. Making each “reserve” large enough to
encompass the core activity areas of at least one wolf pack markedly increases the
likelihood of their effectiveness. Core areas are often disjunct and the utilization
distributions are frequently multimodal (Person, in prep.). Analysis of radio-telemetry
data from wolves on Prince of Wales and Kosciusko Islands suggests that the
75-percent confidence ellipse (Jennrich and Turner 1969) or the 85-percent adaptive
kernel home range (Worton 1989) would be adequate to delimit an area that com-
pletely overlaps core areas regardless of the shape of the utilization distribution
(fig. 9). Both home range models indicate that risk to wolves increases if reserves
are less than 200 square kilometers (78 mi2) in size. An average density of 7 deer
per square kilometer (18 deer/mi2) has a high probability of sustaining 7 to 11 wolves
(depending on the density of deer in the matrix that is overlapped by the wolf pack
home range) of which 2 to 4 would likely disperse from the reserve annually. Given
the high mortality rate for dispersers, we expect that at least one dispersing wolf
would survive for more than one year and be available to replace wolves eliminated
from the matrix.
Establishing at least one reserve per 800 square kilometers (≈300 mi2) of area
within each biogeographic province where wolves occur would meet the above
need. The spacing of reserves is unlikely to be critical because wolves probably
are highly mobile within provinces. The equilibrium model would predict that the
proposed density of reserves would have a high probability of supporting about
30 wolves per 1000 square kilometers if the matrix of land between reserves can
support a minimum of 5 deer per square kilometer.10Although the matrix of land
10This density was computed as follows: no. of wolves
supported by reserves + no. of wolves supported by matrix
= ([200 × 7]/180)+[600 × 5]/80) = 24 wolves/800 square
kilometers = 30 wolves/1000 square kilometers.
between reserves should be capable of sustaining wolves, the expectation is that
the density of wolves in the matrix would be lower than in the reserves and indi-
vidual packs may periodically disappear from these lands due to hunting and trap-
ping (legal or illegal). The matrix of developed lands would presumably function
as a wolf population sink siphoning off dispersers from the reserves, and thus
increase the probability of stability between wolf and deer populations within the
reserves. The matrix also would enable wolves to travel among reserves.
The objective of the reserves is to retain enough high-quality habitat with limited
human access to assure that some packs persist as source populations for those
wolves affected by extensive development. If the strategy described above was
applied to Prince of Wales and Kosciusko Islands, then nine reserves would be
Figure 9—Radio locations for three wolf packs on central Prince of Wales Island, March
1993 to September 1995. The radio locations clearly indicate areas where activity is
concentrated and illustrate the disjunct nature of the usage distributions. The area repre-
sented by the average of the 75-percent bivariate-normal confidence ellipse home ranges
(shown) and the 85-percent adaptive kernel home ranges for all wolf packs monitored on
Prince of Wales and Kosciusko Islands is suggested as the minimum size of reserves
that will encompass the core activity areas of at least one wolf pack.
created totaling about 180 000 hectares (437,000 acres). If timber harvest was
limited and road construction prohibited in the reserves, consistent with reserve
objectives, then risk to long-term wolf viability would be reduced. We predict that a
total of 100 to 200 wolves would be sustained on Prince of Wales and Kosciusko
Islands depending on the capability of the matrix to support deer over the long
term and the intensity of human-caused mortality of wolves in the matrix.
We emphasize that this strategy is appropriate only for biogeographic provinces that
currently support (or have the potential to support) dense populations of deer, such
as those provinces within GMUs 2 and 3. Data are lacking concerning home ranges,
movements, activity, and food habits of wolves inhabiting areas that can support only
low densities of prey. Attempting to maintain the density of current prey populations
and minimizing road construction would benefit wolf populations in these areas. More
detailed and precise information should result from further research.
We thank the University of Alaska Fairbanks, USDA Forest Service, U.S. Fish and
Wildlife Service, National Biological Service, and Alaska Department of Fish and
Game for support and assistance in completing this paper; and the Northwest Scien-
tific Association for assistance with manuscript review. M. Kirchhoff was supported
by Federal Aid in Wildlife Restoration Study 2.11 while working on this assessment.
We also thank L. Beier, R.T. Bowyer, M. Brown, C. Crocker-Bedford, E. Debevec,
B. Dinneford, M. Ingle, M. Hicks, D. Kellyhouse, R. Lande, D. Larsen, M. McNay,
B. Noon, R. Nowak, J. Schoen, C. Schwartz, W. Smith, and K. Taylor for their in-
sight and editorial comments. We are grateful to the anonymous reviewers who
offered critical comments and helped to strengthen this report.
Alaback, P.B. 1982. Dynamics of understory biomass in Sitka spruce-western
hemlock forests of southeast Alaska. Ecology. 63(6): 1932-1948.
Alaback, P.B.; Herman F.R. 1988. Long-term response of understory vegetation
to stand density in Picea-Tsuga forests. Canadian Journal of Forest Research.
Alaback, P.B.; Juday, G.P. 1989. Structure and composition of low elevation
old-growth forests in research natural areas of southeast Alaska. Natural
Areas Journal. 9: 27-39.
Alaback, P.B.; Tappeiner, J.C. 1984. Response of understory vegetation to thin-
ning in the Sitka spruce-western hemlock forests of southeastern Alaska.
Establishment report. On file with: USDA Forest Service, Pacific Northwest
Research Station, 2770 Sherwood Lane, Suite 2A, Juneau, AK 99801-8545.
Alaska Department of Fish and Game. 1960. Reproduction, growth, and mortality
of wolves, southeast Alaska. Douglas, AK: Alaska Department of Fish and Game;
annual report of progress; investigations project; W6-R-1; job 2.
Ballard, W.B.; Whitman, J.S.; Gardner, C.L. 1987. Ecology of an exploited wolf
population in south-central Alaska. Wildlife Monographs. 98. 54 p.
Berg, W.O.; Kuehn, D.W. 1982. Ecology of wolves in north-central Minnesota. In:
Harrington, F.H.; Paquet, P.C., eds. Wolves of the world. Park Ridge, NJ: Noyes
Berger, J.; Daneke, D. 1988. Effects of agricultural, industrial, and recreational
expansion on frequency of wildlife law violations in the central Rocky Mountains,
USA. Conservation Biology. 2: 283-289.
Bloom, A.R. 1978. Sitka black-tailed deer winter range in the Kadashan Bay area,
southeast Alaska. Journal of Wildlife Management. 42: 108-112.
Boertje, R.D.; Stephenson, R.O. 1992. Effects of ungulate availability on wolf
reproductive potential in Alaska. Canadian Journal of Zoology. 70: 2441-2443.
Caughley, G.; Sinclair, A.R.E. 1994. Wildlife ecology and management. Cambridge,
MA: Blackwell Scientific Publishing. 334 p.
Connolly, G. 1981. Limiting factors and population regulation. In: Wallmo, O.C., ed.
Mule and black-tailed deer of North America. Lincoln, NE: University of Nebraska
DellaSala, D.A.; Engel, K.A.; Volsen, D.P. [and others]. 1994. Effectiveness of
silvicultural modifications of young-growth forest for enhancing wildlife habitat on
the Tongass National Forest, southeast Alaska. [Place of publication unknown]:
[Publisher unknown]; final report to USDA Forest Service, Alaska Region; [contract
Doerr, J.G.; Sandburg, N.H. 1986. Effects of precommercial thinning on understory
vegetation and deer habitat utilization on Big Level Island in southeast Alaska.
Forest Science. 32: 1092-1095.
Floyd, T.J.; Mech, L.D.; Jordan, P.A. 1978. Relating wolf scat content to prey
consumed. Journal of Wildlife Management. 42: 528-532.
Friis, L.K. 1985. An investigation of subspecific relationships of the grey wolf, Canis
lupus, in British Columbia. Victoria, BC: University of Victoria. 164 p. M.S. thesis.
Fritts, S.H.; Mech, L.D. 1981. Dynamics, movements, and feeding ecology of a
newly protected wolf population in northwestern Minnesota. Wildlife Monographs.
80. 79 p.
Fuller, T. 1989. Population dynamics of wolves in north-central Minnesota. Wildlife
Monographs. 105. 41 p.
Fuller, T. 1991. Effect of snow depth on wolf activity and prey selection in north-
central Minnesota. Canadian Journal of Zoology. 69: 283-287.
Garceau, P. 1960. Food habits and hunting behavior of wolves in southeast Alaska.
Douglas, AK: Alaska Department of Fish and Game; annual report of progress;
investigations project; 1959-60 segment; project W-6-R-1; job 3; 486-490.
Gasaway, W.C.; Boertje, R.D.; Grangaard, D.V. [and others]. 1992. The role of
predation in limiting moose at low densities in Alaska and Yukon and implications
for conservation. Wildlife Monographs. 120. 59 p.
Gasaway, W.C.; Stephenson, R.O.; Davis, J.L. [and others]. 1983. Interrelation-
ships of wolves, prey, and man in interior Alaska. Wildlife Monographs. 84. 50 p.
Gese, E.M.; Mech, L.D. 1991. Dispersal of wolves (Canis lupus) in northeastern
Minnesota, 1969-1989. Canadian Journal of Zoology. 69: 2946-2955.
Goldman, E.A. 1944. Classification of wolves. In: Young, S.P.; Goldman, E.A.,
eds. The wolves of North America. New York: Dover Publications. Part 2.
Hall, E.R. 1981. The mammals of North America. New York: John Wiley and Sons.
1181 p. Vol. 2.
Hanley, T.A. 1984. Relationships between Sitka black-tailed deer and their habitat.
Gen Tech. Rep. PNW-168. Portland, OR: U.S. Department of Agriculture, Forest
Service, Pacific Northwest Forest and Range Experiment Station. 21 p.
Hanley, T.A. 1987. Physical and chemical response of understory vegetation to deer
use in southeastern Alaska. Canadian Journal of Forest Research. 17: 195-199.
Hanley, T.A.; Cates, R.G.; Van Horne, B.; McKendrick, J.D. 1987. Forest stand-
age-related differences in apparent nutritional quality of forage for deer in south-
eastern Alaska. In: Provenza, F.D.; Flinders, J.T.; McArthur, E.D., comps. Pro-
ceedings: Symposium on plant-herbivore interactions; 7-9 August 1985; Snowbird,
UT. Gen. Tech. Rep. INT-222. Ogden, UT: U.S. Department of Agriculture, Forest
Service, Intermountain Research Station: 9-17.
Hanley, T.A.; McKendrick, J.D. 1985. Potential nutritional limitations for black-tailed
deer in a spruce-hemlock forest, southeastern Alaska. Journal of Wildlife Man-
agement. 49: 103-114.
Hanley, T.A.; Robbins, C.T.; Spalinger, D.E. 1989. Forest habitats and the nutri-
tional ecology of Sitka black-tailed deer: a research synthesis with implications for
forest management. Gen. Tech. Rep. PNW-230. Portland, OR: U.S. Department of
Agriculture, Forest Service, Pacific Northwest Research Station. 52 p.
Hatter, I.W.; Janz, D.W. 1994. Apparent demographic changes in black-tailed deer
associated with wolf control on northern Vancouver Island. Canadian Journal of
Zoology. 72: 878-884.
Hebert, D.; Youds, J.; Davies, R. [and others]. 1982. Preliminary investigations of
the Vancouver Island wolf (C.l. crassodon). In: Harrington, F.H.; Paquet, P.C., eds.
Wolves of the world: perspectives of behavior, ecology, and conservation. Park
Ridge, NJ: Noyes Publishing: 54-70.
Holling, C.S. 1966. The functional response of predators to prey density and its role
in mimicry and population regulation. Memoirs Entomology Society of Canada.
Hoskinson, R.L.; Mech, L.D. 1976. White-tailed deer migration and its role in wolf
predation. Journal of Wildlife Management. 40: 429-441.
Janz, D.W. 1989. Wolf-deer interactions on Vancouver Island—a review. In: Wolf-prey
dynamics and management. Rep. WR-40. Victoria, BC: B.C. Ministry of Environ-
Jennrich, R.I.; Turner, F.B. 1969. Measurement of non-circular home range. Journal
of Theoretical Biology. 22: 227-237.
Jensen, W.F.; Fuller, T.K.; Robinson, W.L. 1986. Wolf (Canis lupus) distribution on
the Ontario to Michigan border near Sault Ste. Marie. Canadian Field Naturalist.
Johnson, L.J. 1987. Reproductive potential of Sitka black-tailed deer in southeast
Alaska. Douglas, AK: Alaska Department of Fish and Game; federal aid in wildlife
restoration project; project W-22-4 and W-22-5; job 2.8R. 29 p.
Jull, A.J.T. 1993. Letter dated May 7 providing C14analysis results to T.H. Heaton.
On file with: T.H. Heaton, University of South Dakota, Vermillion, SD 57069.
Keith, L.B. 1983. Population dynamics of wolves. In: Carbyn, L., ed. Wolves in
Canada and Alaska: their status, biology, and management. Canadian Wildlife
Service Report. Series 45: 66-77.
Kennedy, P.K.; Kennedy, M.L.; Clarkson, P.L.; Liepins, I.S. 1991. Genetic
variability in natural populations of the gray wolf, Canis lupus. Canadian Journal
of Zoology. 69: 1183-1188.
Kirchhoff, M.D. 1990. Evaluation of methods for assessing deer population trends in
southeast Alaska. Douglas, AK: Alaska Department of Fish and Game; federal aid
in wildlife restoration project; project W-22-6, W-23-1, W-23-2, and W-23-3; study
IIB-2.9; final report. 35 p.
Kirchhoff, M.D. 1994. Effects of habitat fragmentation on deer in southeast Alaska.
Juneau, AK: Alaska Department of Fish and Game; federal aid in wildlife restor-
ation project; project W-23-3, W-23-4, W-23-5, W-24-1, W-24-2; study 2.10. 60 p.
Kirchhoff, M.D.; Pitcher, K.W. 1988. Deer pellet-group surveys in southeast Alaska.
Douglas, AK: Alaska Department of Fish and Game; federal aid in wildlife restor-
ation project; project W-22-6; job 2.9. 113 p.
Kirchhoff, M.D.; Schoen, J.W. 1987. Forest cover and snow: implications for deer
habitat in southeast Alaska. Journal of Wildlife Management. 51: 28-33.
Kirchhoff, M.J. 1993. Report—deer pellet-group surveys in southeast Alaska.
Douglas, AK: Alaska Department of Fish and Game. [Unpaged].
Klein, D.R. 1964. Range to related differences in growth of deer reflected in skeletal
ratios. Journal of Mammalogy. 45: 226-235.
Klein, D.R. 1965. Postglacial distribution patterns of mammals in the southern
coastal regions of Alaska. Arctic. (10)1: 7-20.
Klein, D.R. 1981. The problems of over-population of deer in North America. In:
Jewell, P.A., ed. Problems in management of locally abundant wild animals.
New York: Academic Press: 119-127.
Klein, D.R. 1996. The introduction, increase, and demise of wolves on Coronation
Island. In: Carbyn, L.N.; Fritts, S.H.; Seip, D.R., eds. Ecology and conservation
of wolves in a changing world. Occas. Publ. 35. [Place of publication unknown]:
Canadian Circumpolar Institute: [pages unknown].
Klein, D.R.; Olson, S.T. 1960. Natural mortality patterns of deer in southeast Alaska.
Journal of Wildlife Management. 24: 80-88.
Kohira, M. 1995. Diets and summer habitat use by wolves on Prince of Wales
Island, southeast Alaska. Fairbanks, AK: University of Alaska Fairbanks. 97 p.
Lehman, N.; Eisenhawer, A.; Hansen, K. [and others]. 1991. Introgression of
coyote mitochondrial DNA into sympatric North American gray wolf populations.
Evolution. 45: 104-119.
Lewis, M.A.; Murray, J.D. 1993. Modeling territorial and wolf to deer interactions.
Nature. 366: 738-740.
May, R.M. 1974. Stability and complexity in model ecosystems. Princeton, NJ:
Princeton University Press. 265 p.
McArthur, C.; Robbins, C.T.; Hagerman, A.E.; Hanley, T.A. 1993. Diet selection by
a generalist browser in relation to plant chemistry. Canadian Journal of Zoology.
McCullough, D.R. 1979. The George Reserve deer herd: population ecology of a
K-selected species. Ann Arbor, MI: The University of Michigan Press. 271 p.
McCullough, D.R. 1987. The theory and management of Odocoileus populations. In:
Wemmer, C.M., ed. Biology and management of the Cervidae. Washington, DC:
Smithsonian Institution Press: 535-549.
McNay, R.S.; Voller, J.M. 1995. Mortality causes and survival estimates for adult
female Columbian black-tailed deer. Journal of Wildlife Management. 59(1): 138-46.
Mech, L.D. 1970. The wolf: the ecology and behavior of an endangered species.
Minneapolis, MN: University of Minnesota Press. 384 p.
Mech, L.D. 1977. Productivity, mortality, and population trends of wolves in north-
eastern Minnesota. Journal of Mammalogy. 58: 559-574.
Mech, L.D. 1989. Wolf population survival in an area of high road density. American
Midland Naturalist. 121: 387-389.
Mech, L.D. 1995. The challenge and opportunity of recovering wolf populations.
Conservation Biology. 9: 270-278.
Mech, L.D.; Fritts, S.H.; Radde, G.L.; Paul, W.J. 1988. Wolf distribution and road
density in Minnesota. Wildlife Socitey Bulletin. 16: 85-87.
Mech, L.D.; Karns, P.D. 1977. Role of the wolf in a deer decline in the Superior
National Forest. NC-148. St. Paul, MN: U.S. Department of Agriculture, Forest
Service, North Central Forest Experiment Station. 23 p.
Merriam, H. 1966. Deer report. Juneau, AK: Alaska Department of Fish and Game;
federal aid in wildlife restoration; project W-6-R-6 and W-15-R-1.
Merriam, H. 1968. Deer report. Juneau, AK: Alaska Department of Fish and Game;
federal aid in wildlife restoration; project W-15-R-2 and W-15-R-3.
Merriam, H.; Schoen, J.; Hardy, D. 1994. Sitka black-tailed deer. Wildlife Notebook
Series. Juneau, AK: Alaska Department of Fish and Game. 2 p.
Messier, F. 1994. Ungulate population models with predation: a case study with the
North American moose. Ecology. 75: 478-488.
Mladenoff, D.J.; Sickley, T.A.; Haight, R.G.; Wydeven, A.P. 1995. A regional
landscape analysis and prediction of favorable gray wolf habitat in the northern
Great Lakes region. Conservation Biology. 9: 279-294.
Morgan, S.O. 1990. Wolf. Juneau, AK: Alaska Department of Fish and Game;
annual report of survey-inventory activities, federal aid in wildlife restoration
project; W-23-2; study 14.0. 158 p.
Nelson, M.E.; Mech, L.D. 1981. Deer social organization and wolf predation in
northeastern Minnesota. Wildlife Monographs. 77. 53 p.
Nelson M.E.: Mech, L.D. 1986a. Deer population in central Superior National Forest,
1967-1985. Res. Pap. NC-271. [St. Paul, MN]: U.S. Department of Agriculture,
Forest Service [North Central Forest Experiment Station]. 8 p.
Nelson, M.E.; Mech, L.D. 1986b. Relationship between snow depth and gray wolf
predation on white-tailed deer. Journal of Wildlife Management. 50: 471-474.
Noon, B.R.; Biles, C.M. 1990. Mathematical demography of spotted owls in north-
west California. Journal of Wildlife Management. 54: 1-10.
Nowak, R. 1983. A perspective on taxonomy of wolves in North America. In: Carbyn,
L.N., ed. Wolves in Canada and Alaska. Rep. Ser. 45. [Place of publication
unknown]: Canadian Wildlife Service: 10-19.
Nowak, R. 1994. Letter to U.S. Fish and Wildlife Service, Ecological Services,
Juneau, AK. 15 July. 2 p.
Nowak, R. 1996. Another look at wolf taxonomy. In: Carbyn, L.N.; Fritts, S.H.; Seip,
D.R., eds. Ecology and conservation of wolves in a changing world. Occas. Publ.
35. [Place of publication unknown]: Canadian Circumpolar Institute: [pages
Nyberg, J.B.; McNay, R.S.; Kirchhoff, M.D. [and others]. 1989. Integrated man-
agement of timber and deer: coastal forests of British Columbia and Alaska. Gen.
Tech. Rep. PNW-GTR-226. Portland, OR: U.S. Department of Agriculture, Forest
Service, Pacific Northwest Research Station.
Olson, S.T. 1979. The life and times of the black-tailed deer in southeast Alaska.
In: Wallmo, O.C.; Schoen, J.W., eds. Sitka black-tailed deer: proceedings of a
conference; [dates of meeting unknown]; Juneau, AK. Ser. R10-48. [Place of
publication unknown]: U.S. Department of Agriculture, Forest Service, Alaska
Parker, K.L.; DelGiudice, G.D.; Gillingham, M.P. 1993. Do urinary urea nitrogen
and cortisol ratios of creatinine reflect body-fat reserves in black-tailed deer?
Canadian Journal of Zoology. 71: 1841-1848.
Pedersen, S. 1982. Geographical variation in Alaskan wolves. In: Harrington, F.H.;
Paquet, P.C., eds. Wolves of the world: perpsectives of behavior, ecology, and
conservation. Park Ridge, NJ: Noyes Publishing: 345-361.
Person, D.K. [In prep.]. The Alexander Archipelago wolf: the ecology of a frag-
mented predator population. Fairbanks, AK: University of Alaska Fairbanks. Ph.D.
Person, D.K.; Ingle, M.A. 1995. Ecology of the Alexander Archipelago wolf and
responses to habitat change. Unpublished progress report 3. 39 p. On file with:
Alaska Department of Fish and Game, P.O. Box 240020, Douglas, AK 99824.
Peterson, R.O.; Page, R.E. 1988. The rise and fall of Isle Royale wolves, 1975-
1986. Journal of Mammalogy. 69: 89-99.
Peterson, R.O.; Woolington, J.D.; Bailey, T.N. 1984. Wolves of the Kenai
Peninsula, Alaska. Wildlife Monographs. 88. 52 p.
Pimlott, D.H.; Shannon, J.A.; Kolenosky, G.B. 1969. The ecology of the timber
wolf in Algonquin Provincial Park. Res. Rep. Wildl. 87. [Place of publication
unknown]: Ontario Department of Lands and Forests. 92 p.
Pollock, H.C.; Connof, M.J.; Winterstein, S.R. 1989. Estimation and analysis
of survival distributions for radio-tagged animals. Biometrics. 45: 99-109.
Potvin, F. 1988. Wolf movements and population dynamics in Papineau-LaBelle
reserve, Quebec. Canadian Journal of Zoology. 66: 1266-1273.
Rausch, R.A. 1967. Some aspects of the population ecology of wolves, Alaska.
American Zoology. 7: 253-265.
Schoen, J.W.; Kirchhoff, M.D. 1985. Seasonal distribution and home-range
patterns of Sitka black-tailed deer on Admiralty Island, southeast Alaska.
Journal of Wildlife Management. 49: 96-103.
Schoen, J.W.; Kirchhoff, M.D. 1990. Seasonal habitat use by Sitka black-tailed
deer on Admiralty island, Alaska. Journal of Wildlife Management. 54: 371-378.
Schoen, J.W.; Kirchhoff, M.D.; Hughes, J.H. 1988. Wildlife and old-growth forests
in southeastern Alaska. Natural Areas Journal. 8: 138-145.
Seber, G.A.F. 1983. The estimation of animal abundance. New York: Macmillan.
Shields, G.F. 1995. Genetic variation among the wolves of the Alexander
archipelago. Fairbanks, AK: University of Alaska Fairbanks; final report;
Alaska Department of Fish and Game. 7 p. Available from: Alaska Depart-
ment of Fish and Game, P.O. Box 240020, Douglas, AK 99824.
Smith, C.A.; Wood, R.E.; Beier, L.; Bovee, K.P. 1986a. Wolf-deer-habitat
relationships in southeast Alaska. Juneau, AK: Alaska Department of Fish
and Game; federal aid in wildlife restoration progress report; project W-22-4;
job 14.13. 20 p.
Smith, C.A.; Young, E.L.; Land, C.W.; Bovee, K.P. 1986b. Wolf-deer-habitat
relationships in southeast Alaska. Juneau, AK: Alaska Department of Fish and
Game; federal aid in wildlife restoration progress report; project W-22-3; job
14.14. 20 p.
Smith, C.A.; Young, E.L.; Land, C.W.; Bovee, K.P. 1987. Predator induced
limitations on deer population growth in southeast Alaska. Juneau, AK: Alaska
Department of Fish and Game; federal aid in wildlife restoration progress report;
project W-22-4, W-22-5. W-22-6; job 14.14R. 20 p.
Stephenson, R.O. 1989. Wolf Canis lupus. Wildl. Notebook Ser. Juneau, AK: Alaska
Department of Fish and Game. 2 p.
Suring, L.; DeGayner, G. 1988. Habitat capability model for gray wolves in south-
east Alaska, version 2.1. Juneau, AK: U.S. Department of Agriculture, Forest
Service, Alaska Region, Fish and Wildife. 13 p.
Theberge, J.B. 1990. Potentials for misinterpreting impacts of wolf predation through
prey:predator ratios. Wildlife Society Bulletin. 18: 188-192.
Thiel, R.P. 1985. The relationship between road densities and wolf habitat suitability
in Wisconsin. American Midland Naturalist. 113: 404-407.
Thurber, J.M.; Peterson, R.O.; Drummer, T.D.; Thomasma, S.A. 1994. Gray wolf
response to refuge boundaries and roads in Alaska. Wildlife Society Bulletin.
U.S. Department of Agriculture, Forest Service. 1991. Supplement to the draft
environmental impact statement, Tongass National Forest land management
plan revision. R10-MB-149. Juneau, AK. 994 p.
U.S. Department of Agriculture, Forest Service. 1996. Revised supplement to the
draft environmental impact statement, Tongass National Forest land management
plan revision, proposed revised Forest plan. R10-MB-314c. Juneau, AK. 575 p.
U.S. Department of Commerce, Bureau of the Census. 1996. Current population
reports. Series P25-1111. Washington, DC.
Van Ballenberghe, V. 1983. Extraterritorial movements and dispersal of wolves in
southcentral Alaska. Journal of Mammalogy. 64: 168-171.
Van Ballenberghe, V.; Ballard, W.B. 1994. Limitation and regulation of moose
populations: the role of predation. Canadian Journal of Zoology. 72: 2071-2077.
Van Ballenberghe, V.; Erickson, W.; Byman, D. 1975. Ecology of the timber wolf
in northeastern Minnesota. Wildlife Monographs. 43. 43 p.
Van Ballenberghe, V.; Hanley, T.A. 1984. Predation on deer in relation to old-
growth forest management in southeastern Alaska. In: Meehan, W.R.; Merrell,
T.R.; Hanley, T.A., eds. Fish and wildlife relationships in old-growth forests.
Morehead City, NC: American Institute Fishery Research Biologist: 291-296.
Wallmo, O.C.; Schoen, J.W. 1980. Response of deer to secondary forest suc-
cession in southeast Alaska. Forest Science. 26: 448-462.
Wayne, R.K.; Lehman, N.; Allard, M.W.; Honeycutt, R.L. 1992. Mitochondrial
DNA variability of the gray wolf: genetic consequences of population decline and
habitat fragmentation. Conservation Biology. 6: 559-569.
Weaver, J.L. 1993. Refining the equation for interpreting prey occurrence in gray
wolf scats. Journal of Wildlife Management. 57: 535-538.
Wood, R.E. 1990. Game management in unit 1A. In: Morgan, S.O., ed. Wolf.
Juneau, AK: Alaska Department of Fish and Game; federal aid in wildlife
restoration; annual report of survey-inventory activities; project W-23-2; study
Worton, B.J. 1989. Kernel methods for estimating the utilization distribution in
home-range studies. Ecology. 70: 164-168.
Yeo, J.J.; Peek, J.M. 1992. Habitat selection by female Sitka black-tailed deer
in logged forests of southeastern Alaska. Journal of Wildlife Management.
The model used to estimate the wolf population on Prince of Wales and Kosciusko
Islands was as follows:
Wpop = total wolf population,
A = area in square kilometers,
= average home range in square kilometers,
= proportion of home range overlapping neighboring wolf home ranges,
Wpack = average number of wolves per pack, and
= proportion of dispersers in wolf population.
Parameter estimates used in the Monte Carlo simulation of the model were as follows:
1. Pack territories were calculated from empirical data (mean = 280 square
kilometers [109 mi2], SE = 78).a
2. Percentage of overlap among territories was calculated to approximate exclusive
use areas by a single pack (mean = 0.04, SE = 0.01).
3. Pack sizes were estimated from aerial observations (mean = 9.0, SE = 2.6 in fall
1994; mean = 5.6, SD = 3.0 in spring 1995).b
4. Percentage of dispersers was determined from empirical data (mean = 0.29,
SE = 0.09).
5. Land area of Prince of Wales and Kosciusko Islands is 6,808 square kilometers
The variance of the estimate is approximated by the following equation derived from
the “delta” method (Seber 1983):
The equation assumes that all variables are independent random samples with
mean = µ and variance = . Ten thousand simulations were conducted, each
incorporating a random permutation of the input parameters.
aThe home range estimate is an average of MCP and
95-percent adaptive kernel home ranges.
bPack size estimates include data from packs not
containing radio-collared pack members.
The Lotka-Leslie model used in this paper was derived in the following way. The
formal model is expressed as:
lx = survivorship to age class x,
bx = number of female offspring per female of age x, and
= finite rate of increase assuming a stable-age distribution.
Recalling that lx = (sx-1)(sx-2)(sx-3)...,
sx = annual survival rate for age class x,
the model can be rewritten as a function of survival rates:
We assume that b and s for age classes >2 are constant. For wolves, b0 = b1 = 0,
and b2 represents the birth rate at age of first breeding, which is likely to be lower
than the birth rate of older adults (b). Therefore the equation can be simplified as:
We then may solve for:
If we assume that s0 = s1 = s and b = b2 then the equation simplifies to:
The estimate of variance is obtained by the “delta” method described by Seber
which yields the following variance estimator:
01 0 1
s s b
s s s b
s s s
s s b
s s s b
0 1 2
s s b
s s s b
λ b g
Survival rates were estimated from the survivorship of radio-collared wolves. The
birth rate was estimated by the following formula:
fp = average number of female pups per pack (mean = 1.8, SE = 0.9),
= proportion of female dispersers in the female population (mean = 0.33,
SE = (0.12), and
fad = average number of females per pack in spring (mean = 2.3, SE = 1.9).
The variance of the estimate was approximated by:
Var VarVar Varb
The wolf predation rate on Sitka black-tailed deer in southeast Alaska was calculated
in the following way:
In southeast Alaska, adult bucks average 54.4 kilograms (120 lb; Merriam et al.
1994) and adult does average 42.2 kilograms (93 lb; Johnson 1987). Assuming a
60:40 female-to-male sex ratio in the adult population (due to buck-only hunting
and higher natural mortality in males), the average adult deer weighs about 47 kilo-
grams (104 lb; 42.2 kilograms × 0.6 plus 54.4 kilograms × 0.4). Winter fawns (<1 yr)
average 19.5 kilograms (43 pounds; Johnson 1987). Predation rates on fawns ver-
sus adults are highly variable in the literature, and the factors affecting that ratio are
poorly understood (Fuller 1989). We conservatively assumed that an average of one
in three deer killed is a fawn (33 percent). About 74 percent of adult deer carcasses
are consumed (Ballard et al. 1987, Fuller 1989) and 88 percent of fawn carcasses
(Floyd et al. 1978). Thus, the average edible weight of each deer carcass is 29.1 kil-
ograms or 64 pounds ([104 × 0.66 × 0.74]+[43 × 0.33 × 0.88]).
In other parts of North America where deer are the primary prey, the mean con-
sumption rate of wolves is 0.04 kilogram of food per kilogram of wolf body weight
(0.09 lb of food per pound of wolf body weight) per day (range 0.06 to 0.10; Fuller
1989). An average wolf on Prince of Wales Island weighs 30.4 kilograms (67 lb;
n=26; Person, in prep.); thus, each wolf consumes 2.72 kilograms (6.0 lb) per wolf per
day (0.09 × 67). Given that requirement, and assuming a diet composed 77 percent
of deer, each wolf kills about 26 deer per year [(365 × 2.72 × 0.77)/29.1)].