See discussions, stats, and author profiles for this publication at: [Link]

net/publication/241277361

Wolves

Chapter · January 2011

CITATIONS READS
0 121

1 author:

Jane M. Packard
Texas A&M University
86 PUBLICATIONS   1,312 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

well-being of social ungulates View project

Wolf behavior View project

All content following this page was uploaded by Jane M. Packard on 29 May 2014.

The user has requested enhancement of the downloaded file.

 This article was originally published in the Encyclopedia of Animal Behavior
published by Elsevier, and the attached copy is provided by Elsevier for the
author's benefit and for the benefit of the author's institution, for non-
commercial research and educational use including without limitation use in
instruction at your institution, sending it to specific colleagues who you know,
and providing a copy to your institution’s administrator.

All other uses, reproduction and distribution, including without limitation

commercial reprints, selling or licensing copies or access, or posting on open
internet sites, your personal or institution’s website or repository, are
prohibited. For exceptions, permission may be sought for such use through
Elsevier's permissions site at:

[Link]

Packard J.M. (2010) Wolves. In: Breed M.D. and Moore J., (eds.) Encyclopedia
of Animal Behavior, volume 3, pp. 611-620 Oxford: Academic Press.

© 2010 Elsevier Ltd. All rights reserved.

 Author's personal copy

Wolves
J. M. Packard, Texas A&M University, College Station, TX, USA
ã 2010 Elsevier Ltd. All rights reserved.

Introduction In contrast, the hunting behavior of adult wolves is

highly flexible, varying with the food web in the neigh-
Among the large-bodied social carnivores, the foraging borhood where each individual grows up (Figure 1).
behavior of gray wolves is similar to that of lions, African Individuals fine-tune their innate hunting abilities directly
wild dogs, dholes, hyenas, and killer whales, in that indi- by trial and error, as well as indirectly by joining group
viduals hunt in groups as well as in singles and pairs. hunts. In the arctic, some wolf families follow migratory
However, communicative and reproductive behaviors of caribou that feed on sparse lichens, traveling across vast
gray wolves differ in distinctive ways from cat-like expanses of tundra. In wooded regions where more vari-
(Feliformia) and other dog-like (Caniformia) species in ety of food is available, caribou do not migrate and
the taxonomic order Carnivora. Teasing apart which fixed neither do wolves. Further south, in the boreal forest of
aspects of wolf behavior are relatively more instinctive Isle Royale National Park, some lone wolves and pairs
(associated with heritable genotypes), and which flexible supplement their moose diet with beaver and snowshoe
aspects are relatively more variable due to individual hares; they learn to take the prey species that is most
experiential learning and behavioral plasticity has fasci- readily available. When food resources change, switching
nated behavioral researchers for over half a century. tactics appear to be an intelligent adaptive behavioral
Recent advances in technology have led to some unex- trait on the flexible end of the instinct–learning contin-
pected answers and generated new questions about wolf uum (see below).
behavior, at both the genomic and ecological levels of In regions where there are no longer large native prey
behavioral systems. species, and the food web offers less variety, wolves even
Both the foraging and communication behaviors of scavenge from garbage dumps. Indeed, one hypothesis
wolves illustrate how some actions may be categorized as about the divergence of the ancestors of domestic dogs,
more ‘instinctive,’ and others more ‘learned’ on the contin- which most taxonomists now recognize as a subspecies of
uum that ethologists define as ranging from fixed reflex to the gray wolf, is related to appearance of a novel resource
flexible intelligence. For example, when a pup does a ‘leap- in the food web: refuse discarded by human hunters and
pounce’ the first time she hears a vole rustling in grass or early agriculturalists. An associated hypothesis states that
sees a grasshopper moving, the action appears instinctive. the social cognition abilities of wolves were a preadapta-
Pups do not have to learn the reflex-like action, it just tion for domestication by humans worldwide.
occurs; however, they do refine their capture skills with This behavioral flexibility in hunting, which varies with
trial and error learning. The leap-pounce is common in the complexity of the food web, is not unique to wolves but
coyotes and foxes, as well as in all subspecies of wolves. rather, is one of many examples of social carnivores fine-
Applying the comparative method, an ethologist would tuning their actions through mechanisms of learning from
infer that the genetic basis of ‘leap-pounce’ is highly herita- individual experience, alone as well as in social contexts.
ble in canids. Exactly how much learning in wolves occurs because of
Also on the fixed end of the continuum of instinct/ social transmission remains a question for future inquiry.
learning, caching behavior appears in untutored wolf pups In the sections that follow, I elaborate more on the research
before they are weaned. A predictable sequence of caching that has answered many questions about behavioral flexi-
motions (placing a food object on the ground, scraping bility and intelligence in individual wolves: subtle com-
debris over the object with the bridge of the nose, and munication, problem-solving, and learning in the social
tamping down with the muzzle) appear in several species context of family groups. From the perspective of wolf
of the dog-like (Canini) and fox-like (Vulpini) tribes in the populations, we will explore how behavioral flexibility is
subfamily Caninae. Likewise, the whines that newborn also linked with adaptive responses to environmental fluc-
wolf pups emit when separated from the warmth of their tuations: territoriality, deferred reproduction and dispersal
mother and siblings appear to be a more instinctive reflex mechanisms.
than a learned one. Therefore, the fixed actions (e.g., Before synthesizing recent studies and intriguing ques-
whine, cache and leap-pounce) that are similar across tions for future research about wolf behavior, we first need
closely related canine species are hypothesized to have a historical perspective on how certain hypotheses have
been retained in the ancestral part of the genome (phylo- not stood the test of time. Unfortunately, many of these
genetic inference). discarded hypotheses about wolves and myths in the

611
Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620
 Author's personal copy
612 Wolves

so that all wolves in a pack know their roles in a domi-

nance hierarchy. For example, Douglas Pimlott quoted
Niko Tinbergen’s interpretation of a strict hierarchy in
the sled dogs he observed in Greenland. At the time, it
seemed reasonable to infer that those dogs that look more
like wolves would behave like wolves. No published
evidence about wolf behavior was available prior to
the 1940s, so the hypothesis remained untested for dec-
ades. The popular notion of born losers and winners was
reinforced by the insightful anecdotes about personal
experiences with dogs and wolves published by Konrad
Lorenz. However, Lorenz also noted both the persistent
personality traits that varied across breeds of dogs, and the
extreme changes in one individual deprived of his pri-
mary social companion.
Two seminal publications introduced a different inter-
pretation of social interactions in wolf packs, both empha-
sizing the family structure. In Alaska, Adolph Murie
observed a wolf family caring for pups near a den,
Figure 1 Arctic wolf on Ellesmere Island.
now in Denali National Park. In a Swiss zoo, Rudolph
Schenkel described in more detail how the food begging
behaviors of pups developed into solicitous appeasement
signals in juveniles interacting with both parents. He
popular literature persist, as remnants of earlier times. interpreted these interactions as the social glue that
The subsequent review subsection is aimed at encourag- holds the wolf family together. Both studies emphasized
ing critical thinking about what evidence is needed to test the influence of age on the dynamics of dominance inter-
hypotheses about wolf behavior and how difficult it has actions, as both parents and older siblings cared for pups.
been to obtain. Is wolf pack structure more like a pecking order or like
a caring family? Understanding the ancestral roots of dog
behavior was a compelling justification for the multiple
Review: History of Canid Behavioral postwar studies that emerged in the 1950s and continued
Research into the 1980s. An American team led by John P. Scott
investigated the development and heritability of behavior
Visual signals used by dogs in communication had been in dogs, in the context of comparative studies of wolves. In
noted in the behavioral literature long before wolves were Europe, Erik Zimen examined ontogeny of behavior in
studied either in captivity or the field. Charles Darwin wolf/dog hybrids, inquiring in what ways arrested devel-
illustrated his hypothesis about the principle of antithesis opment in dogs might illustrate the principle of neoteny,
by contrasting images of the upright posture of an alarmed the persistence of juvenile characteristics into adulthood.
dog in response to a person approaching in the distance, Benson Ginsburg’s research group examined questions
compared to the crouching posture that the dog switched associated with the hypothesis that evolution of social
to as soon as it recognized the person as its ‘master.’ In cognition in wolves would have been accelerated if they
terms used in the nineteenth century, this expression of had been isolated in groups that benefited from helpers
emotion signaled an unambiguous change in motivation, caring for young. Among others, Mike Fox and Mark
from dominance to submission. In modern terms, the Bekoff studied behavioral development in litters raised
change in visual signal conveyed the information that without parents, comparing species considered to repre-
the dog was unlikely to escalate attack in response to a sent a continuum of solitary foxes, semisocial coyotes, and
familiar care-giving companion. Unfortunately, the mis- social wolves.
perception that some individuals are always dominant and As evidence from more packs emerged, the variation in
others are always subordinate has persisted despite the group structure became clear (Figure 2). Behavior in some
original context of the drawings that Darwin used; his groups fit the model of a pecking order and others were more
drawings illustrate how one individual can rapidly change like a caring family, as we will examine in more detail in a
signals as it gathers more information about a stimulus later section. Separate research teams examined ontogeny of
(e.g., cues about familiarity). behavior in long-term studies of hand-reared wolves assem-
In the popular literature, the anthropomorphic myth bled to form reproductive groups. These lines of inquiry
persists that a dominant male is needed to enforce order were led by John Fentress, John Rabb, Erik Klinghammer,

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
Wolves 613

hemisphere. In addition, the genetic diversity of wolves is

illustrated by multiple subspecies isolated in fragmented
habitats; one subspecies even lives in ambivalent symbio-
sis with humans, that is, domestic dogs.

Modern Synthesis: Nested Hierarchical

Systems

The social structure of wolves has been analyzed in terms

of three levels of selection, each nested within the other:
family groups, subpopulations, and ecosystems. Viewed
from a systems perspective, at the first level, individual
Figure 2 The direct stare of the father is enough to interrupt decisions affect survival and reproduction of other wolves
courtship by his adult son in a captive family group. within the same family group (wolf pack). At the second
level of analysis, the social environment, each family
group is influenced by the actions of other family groups
Erik Zimen, R. Derix, Dave Mech, and Ulysses Seal. Evi- within the neighborhood. The groups of individuals that
dence was clear that not all adult-sized wolves reproduce in interact most frequently within a neighborhood are tech-
packs. However, integrated studies of behavior and physiol- nically called a deme, or subpopulation, within a fragment
ogy led to rejection of the hypothesis that nonbreeders were of habitat relatively separated from other fragments.
always physiologically stressed because of the behavior of At the third level of analysis, the physical ecosystem,
dominants. This left open the question why packs contain wolf populations are influenced by biotic (e.g., prey, com-
adult-sized wolves that do not breed. peting species, diseases) and abiotic factors (e.g., winter
Do wolves need to cooperate to kill large prey? On a severity, fire cycles, drought cycles). From a theoretical
parallel track, the landmark study of wolves on Isle Royale perspective, each of these systems is viewed as nested,
by Dave Mech opened several lines of research about because individuals fit within groups, subpopulations, and
wolves as predators, including interaction with the ecosystems.
dynamics of foodwebs. In the early 1960s, the ecosystem This theoretical framework of nested biological systems
on Isle Royale was relatively stable, and the evidence becomes important when scientists apply sociobiological
supported the hypothesis that individual wolves needed models to test hypotheses about wolf social behavior. For
to put aside their own interests in reproduction for the example, several general hypotheses about evolution of
sake of group hunting and to avoid wiping out their food cooperative breeding have been proposed: (1) eusocial: obli-
supply. The notion that wolves had to cooperate to be gate reproductive suppression under extremely harsh con-
successful at hunting large dangerous prey like moose ditions in ecosystems where individuals do not survive
fit nicely with this model, but only for a few years. Over outside a breeding colony; (2) conditional suppression: repro-
the subsequent decades, Rolf Peterson has monitored ductive behavior is reversibly turned on and off in adults,
dynamic peaks and lows in wolf, moose, beaver, and depending on the social environment (groups and neigh-
other carnivore populations on Isle Royale. There have borhoods); and (3) deferred reproduction: the average onset of
been years when the moose population crashed because of first reproduction is delayed by the interaction of social and
overbrowsing and years when wolves supplemented their ecosystem factors (e.g., body size, nutritional condition,
diet with beaver and snowshoe hares, then crashed when olfactory signals, competition for mates). In the following
availability of all prey species was very low. Winter sever- sections, specific evidence from wolf behavior will be
ity and fire ecology have added to the complexity of these synthesized in a manner needed to test each of these
ecosystem dynamics. three general models.
As we will examine, evidence from additional field stud-
ies led to rejection of the hypothesis that wolves are obli-
gated to cooperate in hunting. Under some foodweb Individuals Within Family Groups
conditions, wolves coordinate hunting and pup-rearing
activities, but in other conditions they do not. Hunting Food provisioning within a family group is key to under-
large prey permits large group size, but does not require it. standing the social environment of canids. Although indi-
Such diverse and dynamic conditions fit evolutionary vidual wolves may leave their natal group and spend
models that predict that behavioral plasticity would be at varying amounts of time alone during the transition to
a genetic premium for this large-bodied social carnivore another group, all wolves are born into and develop within
that is widely distributed across all biomes of the northern a family group. Wolf litter size may vary from 1 to 10,

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
614 Wolves

usually 5–6 pups. Group size rarely exceeds 15 wolves, initiate nursing without having previously given birth to a
depending on whether a breeding pair is disrupted, how litter, despite speculation in the popular literature.
many females breed, and how long offspring of varying The social context of learning expands during weeks
ages remain in the family group. Although the highest 5–8, as pups encounter family members that deliver food
reported count was 42, the usual group size is 6–8 wolves, by regurgitation and carrying pieces of carcass. For exam-
consisting of a breeding pair with 4–6 offspring (Figure 3). ple, in a pack on Ellesmere Island, regurgitations were
However, this varies with the history of each family group directed to the pups (81%), the nursing female (14%),
as well as environmental changes in the wolf population and other auxiliaries (6%). All adult wolves regurgitated
and ecosystem, for example, prey type. food, including the breeding pair, yearlings, and a post-
Unlike most polygynous mammals, pup care by more reproductive female. The breeding female and pups
than the mother matters in monogamous wolves. Pups are received most regurgitations from the breeding male.
unlikely to survive to puberty without the care of parents Regurgitations by the breeding female were directed
and/or older siblings. In a study of 148 pooled cases of exclusively to pups. Wolves respond by regurgitation to
territorial breeding wolves, at least one pup survived the muzzle licking by another familiar wolf, a multimodal
loss of a parent in 84% of the cases, presumably because signal that changes meaning with age, social context, and
pups were cared for by other members of the family the presence/absence of food.
group. Pups were more likely to survive the loss of a During the transition stage from dependency on milk
parent in large groups with auxiliary nonbreeders than to solid food (6–10 weeks), not only do pups learn to
in small groups. Older pups were more likely to survive recognize familiar kin, but they are also rewarded by
the loss of a parent than younger pups. food when they approach or follow adults. Detailed
Born in an earthen den or shallow scrape on the sequence analysis of interactions, in both pups and adults,
surface, altricial wolf pups are cared for exclusively by have illustrated that individuals learn the physical and
the mother for 2–3 weeks. Their eyes are closed until social consequences of their actions. At this transition
12–14 days, and their first reflexive topo-taxic responses stage, bouts of chase and wrestling play are typically
are to touch, warmth, smell, and taste. Hearing matures 1–3 h between naps and feedings.
more slowly. Urination by pups in response to the mother At 7–8 weeks, bite strength is sufficient for pups to feed
licking the urogenital region is an example of a parent– from opened small carcasses, such as arctic hares in the
offspring signal. As pups develop coordination to stand Ellesmere Pack. In the weaning process for one litter,
and walk, early learning begins to expand from the social frequency of suckling bouts that occurred outside the
context of littermates and soliciting care from the mother den decreased gradually, as the nurser initiated bouts at
to include interactions with physical objects. longer intervals interrupted more bouts and pups per-
Between 3 and 5 weeks, pups explore the entrance of sisted less when interrupted. Parent–offspring conflict
the den, retreating from unconditioned stimuli that elicit was not obligate, although it may be conditional on food
an alarm bark and approaching the soft squeaking vocaliza- delivery and food storage by caching. On the average, by
tions and multimodal stimuli they have learned to associate 11 weeks, pups no longer suckled and began to follow
with the nursing female. In rare circumstances, more than adults on foraging trips. Activity centers focused on dens
one nursing female may share a den. However, currently and rendezvous sites may change several times, as a litter
there is no evidence that pseudopregnant female wolves is carried to a new location by the breeding female or
moves in response to disturbance.
Sound analysis indicates the vocal repertoire increases
from four to nine call types as pups mature. Barks and
howls are examples of vocal signals that have been studied
in wolves. Pups bark in response to alarming stimuli and
howl when separated from the group or in response to
other howls. One hypothesis of the adaptive function of
these signals is safety in numbers. Pups are vulnerable to
predation by bears and unfamiliar wolves from neigh-
boring packs.
The first agonistic signals used by wolf pups occur in
the context of food. When conflict escalates over a large
food item, such as a rabbit carcass, pups learn the con-
sequences of uninhibited bites from a sibling. They learn
the subtlety of signs (e.g., hard stare, snarl, ear posture,
Figure 3 Sibling wolf pups in a family group on Ellesmere partial lunge) that predict likely escalation to uninhibited
Island. biting. Subtle signs of de-escalation include: look-away,

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
Wolves 615

lie-down, ears-back, lip-licking, crouch, roll-over, pawing, without the help of others. Transitions among behavioral
tuck-tail, and tail-wagging. To the extent that individuals tactics will be discussed in more detail in the next sec-
vary in temperament at birth, each learns coping styles tions, because they are influenced by factors at both the
influenced by the contingencies of their interactions with population and foodweb levels of ecosystems.
siblings and the context of resources, although this com- Comparing canid species, large body size is correlated
plex process is not completely understood because it is with later age of first reproduction. For example, on
so variable. average, large-bodied wolves reach puberty in their sec-
Current behavioral evidence does not support the ond winter, 1 year later than smaller bodied coyotes.
hypothesis that the dominance hierarchy within a litter Puberty in wolves may be accelerated or delayed by a
determines which individuals breed later in life. Behav- couple of years because of interactions of nutritional and
ioral profiles of individuals vary on several dimensions in social factors.
addition to a shyness/boldness continuum. Multivariate There is no evidence to support the hypothesis that
quantitative studies determined that affiliate and play social stress in wolves turns off the physiological readiness
behaviors explain more variation in wolf behavior than to breed after puberty. For example, fecal cortisol was
agonistic actions within intact family groups. Disrupted higher for breeders than nonbreeders in samples from
families that have lost one or more parents are more free-ranging wolves in Iberia. In one captive study, non-
variable and conflict is more likely to escalate as described breeding adults cycled normally; females ovulated and
in the next section. males produced sperm. In another captive pack, the posi-
Born in late spring, juvenile wolves are not quite adult- tive correlation between stress hormones and aggression
sized by their first winter. The synchronized birth season was skewed by one individual with abnormal adrenal
fits the functional hypothesis that the young are born at a hypertrophy. An early-winter peak in testosterone has
time when food is readily available. Those that were born been correlated with rates of scent marking and escalated
later would have been unlikely to survive the rigors of conflict among males.
their first winter. Neonates born in winter would have Deferred reproduction best explains the variation in
risked exposure and malnutrition in times of scarce food. reproductive tactics of wolves in family groups. In nuclear
This genetic basis for seasonal reproduction has been families, food provisioning shapes asymmetric relations
modified in domestic dogs, which breed year-round. On between parents and offspring. During breeding season,
average, birth dates occur weeks earlier in wolf popula- parents are more attracted to mating signals from each other
tions at lower compared to higher latitudes, although the than from offspring (Figure 4). Adolescent wolves are less
mechanism is still not entirely understood. attracted to mating signals from siblings than parents, an
Group size expands seasonally, as pups are born, and attraction likely not reciprocated. Older wolves, both parents
declines as family members disperse or die. For example, and siblings, are likely to interrupt sexual activity by
on average in Denali, only half the pups of the year younger wolves of the same sex. However, the subtle signs
remained with the family through the first winter. Of of asymmetric mate choice and same-sex rivalry are only a
those that remained, only half were still with the family matter of probability and may shift within weeks when
through the second winter. Only a few wolves remained one or both breeders are removed.
with their natal group past the third winter.
Despite the popular notion that young wolves are
driven out of the pack by conflict with parents and sib-
Pair-bonded parents
lings, the data suggest that dispersal mechanisms are a
complex interaction of individual maturation, relation-
P 3 P
ships within the group, food availability, and scent marks
in the neighborhood. During their lifetimes, individuals
may switch among the following categories of tactics: (1) O O
‘biding’ auxiliaries are nonreproductive members of a
territorial group; (2) ‘dispersing’ floaters leave the group
and wander alone or in transient groups that pass through
or between group territories; and (3) ‘breeding’ parents O O
defend the territory where they forage and reproduce,
Courtship
attacking outsiders of the same sex. Both sexes switch (acts/15 min)
among these tactics. O
0.1 to 0.3
This evidence of developmental plasticity has led to 0.4 to 0.9
rejection of the hypothesis that the wolf social system fits Copulations: 3 NP:YR1 Over 1.0
the model of eusociality. Dispersing wolves do not coop- Figure 4 Relations among parents and their offspring during
erate in parental care and do successfully catch prey the breeding season in a captive pack.

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
616 Wolves

Wolf mating signals are multimodal, including phero- stimulates smooth muscle contraction increasing internal
mones and tactile, auditory, and visual cues. Sexual fertilization rate due to sperm and egg movement into
dimorphism is minor, on average males are 20% heavier; the uterine horns and (2) reduced probability of extra-pair
however, no visual signals distinguish the sexes. Howling copulation during the male postejaculation refractory
is an auditory signal with the function of advertising period.
presence of wolves; loners are more likely to howl when The seasonal canid reproductive cycle is unusual
traveling in sparsely than densely populated regions. among mammals because (1) there is only one estrus
Pheromones deposited in urine and feces are the primary whether or not a female becomes pregnant, (2) the post-
mode for advertising sexual identity in dispersers. Dis- ovulatory growth of the corpus luteum is roughly the
persers join up at all times of the year; males are more same duration for both pregnant and nonpregnant
likely to find single females during proestrus vaginal females, and (3) because of elevated prolactin in both
bleeding, usually 2–8 weeks before ovulation. Plasma males and females during the spring pup-rearing season.
estrogen is elevated in females during proestrus and The cascade of hormonal changes following ovulation
estrus, which occurs during January and February. Syn- may stimulate growth of nipples, hair loss from the
chronized estrus usually peaks mid-February, although belly, abdominal swelling, denning behavior, and milk
the exact timing varies with body condition, the social production, although these symptoms are highly variable
environment, and latitude. among individuals and change with age. Seasonal peaks
Proximate mechanisms of reproduction have been well in prolactin are associated with den-digging and food-
studied in both wolves and dogs. During 1–7 days of provisioning by both sexes (breeders and nonbreeders).
estrous, each breeding pair of wolves remains in close Thus, it is difficult to diagnose pregnancy on the basis
proximity (Figure 5), exchanging mutually stimulating of external indicators; more accurate internal indicators
olfactory, visual, and tactile signs of copulatory readiness, may be obtained by sonography and measuring the hor-
for example, sniffing, sequential-urinating, chinning, darting, mone relaxin.
ears-together, head-flick, and paws-to-shoulders. Within Although social monogamy is typical of smaller nuclear
days of spontaneous ovulation, females signal readiness families of wolves, extended and disrupted families may
to stand by averting the tail to the side of the vulva, a include polygynous and polyandrous relationships. Since
reflex that is a fixed action. Ovulation is associated with a postpubertal wolves retain the physiological readiness
peak in plasma luteinizing hormone (LH), which coin- to breed, loss of one or more breeding parents may de-
cides with a drop in elevated estrogen (Figure 5). stabilize dominance relationships. For example following
With experience, male wolves quickly learn to orient deaths of the fathers in two Yellowstone packs, immigration
mounting to the rear. The ejaculatory reflex follows pen- of an unrelated breeding male was followed by multiple
etration and penile thrusting. Subsequently, tissues swell litters. A low frequency of plural breeding has been re-
in the penile bulb, a reflex keeping the pair locked in a corded in several field studies. Congenial relations among
postcopulatory tie that usually lasts 20 min; the range multiple breeding females are usually unstable and persist
in duration (3–30 min) is conditional on interruptions no more than a few years.
by familiar rivals. Hypotheses about the function of Agonistic interactions vary with both the immediate
the postcopulatory tie include (1) oxytocin release that presence of resources and the social environment within
each group of wolves. Resources include food, mates, and
pups. Factors likely associated with escalated conflict are
complex, including the quality of the resource, proximity
to the resource, motivation (e.g., satiation, reproductive
cycle, adrenal activity), personality (e.g., inherited tem-
perament and learned coping styles), and relationships
(e.g., learned contingencies of interactions among specific
individuals). All these factors influence the complexity of
dominance hierarchies (e.g., linear, triadic, age-graded,
branched sex-specific, multinodal), which may change
within a group over time as well as varying among groups
depending on age/sex composition.
The variation in age/sex composition of wolf groups
depends not only on internal factors, but also on the
interactions among groups within populations and dy-
namic patterns of food availability within ecosystems.
Figure 5 A male stands near his resting mate, guarding her Interactions of internal and external factors are elaborated
from rivals while she is in estrous. in the following section.

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
Wolves 617

Family Groups Within Fragmented of their natal group, (5) dispersers return to a familiar
Subpopulations group (siblings or offspring) after turnover in the breeding
pair, and (6) dispersers immigrate into groups that have lost
The social structure of wolf populations includes territo- a same-sex breeder. Overall, genetic variation is likely to
rial reproductive groups and floaters that move between be lower between groups than between breeders within
and through resident territories. In colonizing populations each group.
with low territory density, floaters are more likely to join Auditory and olfactory communication influence dis-
up and start new breeding units in the gaps between tance between groups; in contrast to the momentary and
territories. In established populations, the number of ambiguous information conveyed in howls, scent marks
breeding groups is relatively constant despite turnover as may last for days. In response to playbacks of strange-
groups break up and new groups are formed. Wolves that group howls, groups that reply are more likely to remain
are slow to disperse from their natal family are ‘biders’ in place compared to groups that do not reply and retreat.
(nonbreeding auxiliaries) waiting for a chance to breed Response rate is positively correlated with group size,
when the opportunity arises. breeding condition, and presence of a resource. Single
Within wolf populations ranging from those that are wolves are more likely to approach silently when the
recolonizing an area (Yellowstone) to well-established playback is a solo howl. The prevalence of scent marks,
(Denali, northern Minnesota), genetic relatedness is lower both urine and feces, on trails near junctions and at the
between breeders within a group than it is on the average edge of territories has been described as an ‘olfactory bowl.’
between breeding groups. Wolf groups are semiclosed, usu- Breeders urine-mark on conspicuous objects at a higher
ally accepting immigrants only upon loss of a breeder. Mate rate than nonbreeders, and the urination rate is highest in
turnover ranges from 1 to 6 years, varying among popula- newly formed pairs. Pairs deposit urine marks sequentially
tions. As elaborated in the following section, hypothesized in the same location, a double-marking behavior that may
mechanisms to explain this pattern of nonrandom genetic function in intimidating rivals and stimulating mates.
dispersal include the following ones: (1) individuals choose Group howls occur when resting wolves arise and
mates that are distantly related over those that are close gather together prior to traveling, as well as when they
relatives, (2) breeders defend their mates from same-sex come together after separation. During a group howl,
rivals in neighboring groups, and (3) dispersal between individuals rub bodies, touch noses, and circle with wag-
groups is influenced by mate choice and same-sex rivalry. ging tails. Individuals that hold the tail high are more
Given a choice of mates, wolves of both sexes are predicted likely to respond with an over-the-muzzle bite to nose-
to be more attracted to unrelated than to related individuals. licking by wolves with a lower tail posture. Similar to pups
Although it is unlikely that wolves have a mechanism for soliciting food provisioning, adults that receive an over-
directly detecting genetic relatedness, familiarity is highly the-muzzle bite do not retreat from the group. Whichever
correlated. Given a choice, unfamiliar individuals are more individual departs with a confident gait is likely to be
attractive mates than family members. Contrary to predic- followed by those that are more solicitous. However, if a
tion, inbreeding has occurred when a parent had no other key food provider does not join the departing group,
choice than offspring and when siblings copulated in the group cohesion may deteriorate.
absence of parents. Group decisions on movements vary between wolf
Same-sex combat may explain the intense, uninhibited packs, as well as seasonally within each group. Alone,
conflict between wolf groups, resulting in documented adult wolves can easily travel 40 km in half a day. In
death of breeders, biders, and floaters. Fights between general, movements revolve around pups in the spring/
groups have escalated during extra-territorial intrusions. summer and the breeding female in the winter. In other
Although breeders may be more likely to escalate conflict seasons, the individual leading a traveling line of wolves is
with same-sex rivals, in the excitement of a fight, all group likely to be a breeder. Evidence supporting the hypothesis
members may mob a victim that displays defensive sig- that the male is more likely to be a leader is ambiguous
nals, for example, tucked-tail, ears-flat-back, and arched- and depends on the definition of leading behavior. An
back. While the sample size is not definitive, small groups alternative hypothesis is that variation in leading behavior
are less likely to escalate than large groups. Social monog- may relate to which individual is most consistently asso-
amy is reinforced and extra-pair copulations are reduced ciated with food acquisition, which likely changes with the
by same-sex combat between groups. age, experience, and personality of group members.
Several categories of dispersal between groups have Within a given latitude, group home range size is posi-
been documented: (1) biders immigrate from an unrelated tively correlated with group size in colonizing wolf popula-
neighboring group to one that has lost a same-sex breeder, tions but only marginally so under saturated conditions.
(2) one group divides into neighboring groups, (3) disper- For example, in northwestern Minnesota and Yellowstone,
sers may travel distances as long as 1000 km, (4) dispersers recolonizing groups initially were spaced far enough apart
meet up and establish a new breeding group within 100 km that there were no shared boundaries. As the open areas

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
618 Wolves

filled in, groups defended adjacent boundaries and poten- deer, fallow deer, and mouflon (Appenine mountains
tial for elastic expansion of home range was limited, pre- of Italy).
sumably by encounters with sign left by neighboring packs. Where wild ungulate populations have died out in
Genetic variability between wolf populations is likely parts of Israel and Italy, wolves scavenge at garbage
related to the connectedness of habitat patches. In some dumps in addition to hunting whatever small animals
homogeneous stands of forest, wolves may travel hun- and livestock are vulnerable. Domestic animals (e.g.,
dreds of kilometers without encountering areas of low goats, sheep, pigs, cattle, and dogs) are primary prey in
wolf density. In other landscapes, they travel hundreds northwestern Spain, and the eastern Caucasus of Russia,
of kilometers through areas without wolves before en- wherever wild ungulates are scarce and livestock graze in
countering the sign of wolf presence. From the same litter, or near forests. In northern Finland, wolves hunt semido-
some individuals have dispersed short distances and mestic reindeer.
others long distances. Overall, both sexes are equally In seasonal environments, wolves may opportunisti-
likely to disperse, although dispersal in some populations cally feed lower on the food chain when fruits become
has been biased toward males and others were biased available in the summer. Seeds of raspberries and blue-
toward females. berries have been found in scats (defecations), as have
In summary, an overall model includes neighborhoods cultivated fruits (e.g., grapes, cherries, apples, pears, figs,
of relatively low genetic variation which are nested within plums, and melon). The frequency of grass in wolf scats
habitat fragments that vary in degree of connectivity. ranges from 14% to 43%, based on studies from both
Sexual competition limits the openness of family groups continents.
to immigration of unrelated individuals; however, disrup- The influences of foraging on wolf populations are
tion of monogamous relationships facilitates movement of evident in the variation of wolf territory size. The corre-
individuals between groups. Mate choice tends to favor lation between latitude and mean territory size is highly
outbreeding, although inbreeding occurs when choices significant. The mean estimated territory size ranges from
are limited. Gene flow occurs via dispersal between frag- 137 km2 in Wisconsin to over 2600 km2 on Ellesmere
mented habitats. Island. This variation is also correlated with (1) lower
prey biomass at higher latitudes, (2) smaller ungulate
body size at higher latitudes, and (3) lower productivity
Populations Within Fluctuating of the plants upon which herbivores feed at higher
Ecosystems latitudes.
In a meta-analysis of 38 studies, about one-third of
Sociobiological theory predicts that species adapted to the variation in wolf territory size is positively correlated
fluctuations of their social and ecological environments with prey biomass. Other factors contributing to the vari-
will evolve behavioral traits with a high degree of plastic- ation included (1) wolf density, (2) interaction between
ity. Variation in the distribution of resources is most likely wolf density and rate of wolf population increase, and
to influence the distribution of reproductive females, (3) interaction between the mean territory size and the
including group size. In turn, the distribution of females rate of wolf population increase. For example, mean terri-
likely influences male tactics for defending females and tory size in regions where wolves hunt deer (199 km2) is
offspring from the risks of encounters with rival males. one-quarter the size in moose regions (817 km2), possibly
Secondarily, predation separately influences evolution of because moose are harder to catch and wolves travel
behavioral traits in males and females. further between kills.
Ecosystems inhabited by wolves range from Arctic Do individual wolves benefit from cooperative hunting
tundra (80  N latitude) to desert mountains (less than tactics? In contrast to lions, food acquisition per wolf
40  N latitude). Foodwebs within these diverse ecosystems decreases with hunting group size. Single adult wolves
vary from simple to complex. Examples of simple food- can kill a moose, bison, or musk-ox; however, calves and
webs include (1) blackbuck (India), (2) arctic hares and sick adults are more vulnerable to single wolves. Most
musk oxen (Ellesmere Island), (3) migratory caribou sup- hunting sequences described for wolves have been simple
plemented by small mammals during denning (Alaskan and straightforward. Field biologists differ in opinions
Brooks Range), (4) white-tailed deer and snowshoe hare about whether hunting tactics of wolves show evidence
(Minnesota), (5) moose and white-tailed deer (eastern of cooperation, defined in terms of ambushing prey and
Canada), and (6) red deer and wild boar (Spain and Poland). relay running.
More complex food webs include (1) moose, caribou, and When group size increases, it is most likely due to
Dall sheep (Denali National Park); (2) moose, snowshoe recruitment of young inexperienced wolves, adding little
hare, and beaver (Isle Royale); (3) elk, mule deer, bison, advantage to capture success by the group. One hypothe-
mountain sheep, caribou, mountain goat, and small mam- sis is that young wolves may benefit from group hunt-
mals (western Canada); and (4) red deer, wild boar, roe ing in that they learn the consequences of their own

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
Wolves 619

interactions with prey as well as observing the conse- the plants. When moose are unhealthy, they are more
quences of actions by more experienced group members. vulnerable to wolves, and wolf predation has more of an
However, in the Mexican wolf reintroduction, even inex- impact than when moose are relatively invulnerable. Vul-
perienced captive-reared wolves learned to kill elk within nerability of prey to wolves is also increased by snow
3 weeks after release. Evidence to test this ‘trade school’ conditions and harsh winters.
hypothesis about the function of group size would be very Since Isle Royale is a closed system on an island, the
difficult to obtain because of welfare issues and limited fluctuations in plants, herbivores, and carnivores are more
visibility of wolf hunts in forested ecosystems and rugged accentuated. However, similar dynamics exist in other
terrain. fragments of forested habitat that are more open systems.
Alternatively, large groups wolves may have a compet- The linkages among components of each system are
itive advantage in interactions with other predators (e.g., harder to measure in regions where wolves and their
black bears) and scavengers at carcasses (e.g., ravens, prey disperse over larger distances. Large expanses of
eagles, foxes, coyotes, wolverines, and bobcats). In one forest are not homogeneous; local conditions function as
study, the percentage of carcasses consumed by other sources and sinks in terms of the dynamics of wolf popu-
scavengers was inversely correlated with wolf group size. lations on a broader scale of analysis.
Groups of wolves were more likely than singles to attack Theoretical questions about the stability of predator
denning black bear or chase them off a carcass. However, and prey populations due to wolf foraging ecology are still
grizzly bears usually displace wolves at carcasses indepen- actively debated. However, researchers agree on three
dent of the number of wolves present. Single coyotes and generalizations: (1) wolf impact is highest on the juvenile
foxes are vulnerable to being killed by wolves. Wolves age class of prey; (2) where wolf populations are increas-
rarely consume carcasses of competitors. Few interactions ing, the impact of predation is higher; and (3) the com-
have been recorded between wolves and felids (e.g., bob- bined impact of predation by wolves and bears is more
cat, lynx, mountain lion, and Siberian tiger). likely to tip prey populations into a declining trend.
Do breeding wolves benefit from helpers at the den? Disease outbreaks also contribute to the instability of
One hypothesis is that auxiliaries may contribute more to wolf populations. Over a 30-year study of canine parvovi-
provisioning under conditions of food abundance than rus in northeastern Minnesota, pup mortality increased
when food is scarce. However, in good times, breeders 70% in one region and varied from 40 to 60% over a
are likely to be more successful at prey delivery, and larger scale. The rate of growth for the infected wolf
auxiliaries are more likely to disperse. More auxiliaries population was 4% as against 16–58% in other wolf
in a group do not always increase the probability that pups populations. Changes in dispersal potentially related to
will be attended around the clock. Some evidence points spread of disease included (1) fewer dispersing juveniles,
to auxiliaries returning to intercept provisioning at times (2) mortality of entire groups, and (3) a higher probability
when breeders are likely to return to pups. Older off- of adults dispersing following disruptions due to death
spring may compete with younger siblings, under scarce of breeders.
food conditions. Further studies are needed to fully In summary, variation in the canid genome has been
answer this question. shaped over geologic time scales by glacial cycles that
Do nonbreeding auxiliaries benefit by inheriting a ter- repeatedly displaced northern populations and blocked
ritory when breeders are displaced? The current working or opened dispersal routes between continents. The
hypothesis suggests that the answer depends again on the behavioral plasticity that permitted wolves to invade eco-
interaction of wolf density and prey availability. Under systems as diverse as deserts, forests, mountains, and tun-
conditions of low wolf density and high food availability, dra also permits individuals to adapt within lifetimes to
dispersing floaters are more likely to start a new breeding ecological changes in prey availability resulting from
unit than biders are likely to inherit a territory. However, shorter cycles (e.g., fire, precipitation, plant succession).
when wolf density is high and prey density is low, mortality Interactions of factors within dynamic ecosystems make it
is higher in dispersers than biders. Extra-territorial forays very difficult to test behavioral models on the basis of
and encounters with neighboring groups result in deaths of costs and benefits in terms of ultimate fitness. Given the
breeders under these conditions. Under conditions where behavioral plasticity in social carnivores, it is all too
breeders die, biders are more likely to inherit a territory. tempting to infer the adaptive significance of cooperative
This conditional model of switching tactics and variable foraging despite the paucity of definitive evidence.
pay-offs for wolves foraging in groups has emerged from
studies of ecosystem fluctuations. On Isle Royale, the body
condition of moose is correlated with browse forage quality. Some Current Questions
The forage for moose has changed over decades because
of plant succession in patches disturbed by forest fire, as Recent expansion of research in social cognition and the
well the direct impacts of moose and other herbivores on canid genome have opened promising perspectives for

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

 Author's personal copy
620 Wolves

reframing questions about behavioral mechanisms in The challenges of teasing apart epigenetic influences
wolves. For example, the behavioral traits currently inter- during social learning raise questions about research
preted as cooperative may also be viewed as by-products design and sample size. Given the genetic and behavioral
of directional selection favoring large body size. In the variation within and between breeds of dogs, which indi-
following section, alternative models that are emerging in viduals should be chosen to compare with wolves? Given
diverse lines of research that view dogs as a subspecies of equivalent variation in wolves, which should be chosen to
wolf are briefly outlined. compare with dogs? To compare social cognition in dogs
The genetics of domestic and wild canids are under and wolves, should they both be raised in human families
investigation by the research team coordinated by Elaine or in canine families? What are the implications of separ-
Ostrander and Robert Wayne. Questions arise about how ating both dogs and wolves from social companions dur-
contrasts in body size as dramatic as a great Dane and ing development?
Chihuahua illustrate the big variation that can result from Answers to questions about how behavioral variation is
small changes in the timing of action of ‘controller genes.’ related to the interaction of genetic variation and envi-
Not only does the dog subspecies of wolves contain the ronmental variation are still elusive. However, the prog-
genes of their ancestors, but substantial numbers of muta- ress in understanding some of these linkages within the
tions have also accumulated during artificial selection canid genome has accelerated in the past decade and
over thousands of years of dog breeding. holds promise for reframing future questions about the
New lines of research are opening in the study of evolution of behaviors on both ends of the continuum
which dog breeds conserve more of the genotypes typical between fixed actions and behavioral plasticity.
of wild wolves from separate continents. This raises addi-
tional questions about how breed differences, associated See also: Conservation and Behavior: Introduction;
with diversity of human culture, reflect variation in per- Domestic Dogs; Social Cognition and Theory of Mind;
ceptual systems (e.g., sight hounds, scent hounds), cogni- Spotted Hyenas.
tive abilities (e.g., shepherd breeds, guard/rescue breeds),
emotional systems (e.g., retriever breeds, fighting breeds),
and energetic systems (e.g., sled dogs, lap dogs). Further Reading
Dogs are not just neotenous wolves. Breed differences
may be productively viewed as experiments in differential Asa CS and Valdespino C (1998) Canid reproductive biology: An
reproduction, which has changed the timing by which integration of proximate mechanisms and ultimate causes. American
Zoologist 38: 251–259.
behavioral and morphological systems develop. Behavioral Bekoff M, Daniels TJ, and Gittleman JL (1984) Life history patterns and
studies of directional selection for tameness in silver foxes the comparative social ecology of carnivores. Annual Review of
raised similar questions about how changes in the develop- Ecology and Systematics 15: 191–232.
Brainerd SM, Andren H, Bangs EE, et al. (2008) The effects of breeder
ment of neurotransmitter systems, for example, serotonin, loss on wolves. Journal of Wildlife Management 72: 89–98.
dopamine, epinephrine, may be linked in unexpected ways Darwin C (1872) The Expression of the Emotions in Man and Animals.
to timing of reproductive cycles and morphological traits Chicago, IL: University of Chicago Press.
Ferguson G and Smith DW (2006) Decade of the Wolf: Returning the
such as a curled tail and white star on the chest. Wild to Yellowstone, p. 256. Guilford, CT: Globe Pequot Press.
Wolves are not just the ancestors of dogs. The genotype Lorenz K (1954) Man Meets Dog. London: Methuen Press.
associated with black coat color in wolves appears to have Mech LD (1970) The Wolf: The Ecology and Behavior of an Endangered
Species. Garden City, NY: Natural History Press.
originated in domestic canids and to have spread through Mech LD, Adams G, Meier TJ, Burch JW, and Dale BW (1998) The
wild populations of wolves. Questions arise about where Wolves of Denali, p. 227. Minneapolis, MN: University of Minnesota
mutations appeared and how they have persisted in canid Press.
Mech LD and Boitani L (eds.) (2003) Wolves: Behavior, Ecology
lineages with histories of genetic bottlenecks, likely shaped and Conservation, p. 448. Chicago, IL: University of Chicago
by alternating phases of inbreeding and outbreeding. What Press.
are the implications for evolutionary models of behavioral Packard JM and Mech LD (1980) Population regulation in wolves. In:
Cohen MN, Malpass RS, and Klein HG (eds.) Biosocial Mechanisms
processes in wolves and other wild canid species? of Population Regulation, pp. 135–150. New Haven, CT: Yale
On parallel lines of investigation, more detailed ques- University Press.
tions about social cognition are progressing for canids. For Packard JM, Mech LD, and Ream RR (1992) Weaning in an arctic wolf
pack: Behavioral mechanisms. Canadian Journal of Zoology 70:
example, questions about empathy and reconciliation after 1269–1275.
conflict have recently been examined for wolves. Intriguing Peterson RO (2007) The Wolves of Isle Royale: A Broken Balance, p. 192.
differences between dogs and wolves have been documen- Ann Arbor, MI: University of Michigan Press.
Smith D, Meier T, Geffen E, et al. (1997) Is incest common in gray wolf
ted in observational learning and the recruitment of social packs? Behavioral Ecology 8: 384–391.
companions in problem-solving tasks. Questions arise about Vonholdt BM, Stahler DR, Smith DW, Earl DA, Pollinger JP, and Wayne
what are appropriate problem-solving tasks: contraptions RK (2008) The genealogy and genetic viability of reintroduced
Yellowstone grey wolves. Molecular Ecology 17: 252–274.
humans devise to control extraneous variables or tasks that Zimen E (1981) The Wolf, a Species in Danger, p. 373. New York, NY:
arise in foraging? Delacorte Press.

Encyclopedia of Animal Behavior (2010), vol. 3, pp. 611-620

View publication stats