Ecology and reproductive strategy of an Afroalpine specialist: Ethiopian wolves in the Bale Mountains

Abstract

Ethiopian wolves Canis simensis are confined to seven ranges of Afroalpine habitats in Ethiopia, where they combine conspicuous sociability with specialised, solitary foraging for a narrow range of Afroalpine rodent species. A detailed field study in the Bale Mountains in 1988-1992 yielded information on the behavioural ecology of this rare carnivore, and was followed up by other field studies on population biology, ecological requirements and genetics. Here we present a review of the current state of knowledge of Ethiopian wolves’ biogeography, diet, foraging behaviour, spatial organization, territoriality, social structure, mating behaviour, reproductive biology dispersal, biogeography, and genetics. We conclude with remarks on the cost the wolves’ specialisation to the Afroalpine ecosystem poses to their long-term conservation.

Full text

Walia-Special Edition on the Bale Mountains 61
Ecology and Reproductive Strategy of an Afroalpine Specialist: Ethiopian Wolves in
the Bale Mountains
Claudio Sillero-Zubiri1*, Dada Gottelli2, Jorgelina Marino1, Deborah Randall1,3, Lucy Tallents1 and
David W. Macdonald1
1 Wildlife Conservation Research Unit, University of Oxford, Tubney House, Abingdon Road,
Tubney OX13 5QL, UK
2 Institute of Zoology, Zoological Society of London, Regent’s Park, London NW1 4RY, UK
3 Frankfurt Zoological Society, PO Box 100003, Addis Ababa, Ethiopia
*Email: claudio.sillero@zoo.ox.ac.uk
Abstract
Ethiopian wolves Canis simensis are conned to seven ranges of Afroalpine habitats in Ethiopia,
where they combine conspicuous sociability with specialised, solitary foraging for a narrow range
of Afroalpine rodent species. A detailed eld study in the Bale Mountains in 1988-1992 yielded
information on the behavioural ecology of this rare carnivore, and was followed up by other eld
studies on population biology, ecological requirements and genetics. Here we present a review of
the current state of knowledge of Ethiopian wolves’ biogeography, diet, foraging behaviour, spatial
organization, territoriality, social structure, mating behaviour, reproductive biology dispersal,
biogeography, and genetics. We conclude with remarks on the cost the wolves’ specialisation to the
Afroalpine ecosystem poses to their long-term conservation.
Introduction
At ca. 20 kg, the Ethiopian wolf Canis simensis differs from such typical, medium-sized canids as
the coyote C. latrans in its unusually long legs and a long muzzle (Sillero-Zubiri and Gottelli 1994).
Restricted to rodent-rich Afroalpine habitat within the Ethiopian highlands, its diurnal habits and
distinctive coat render this species conspicuous. A bright tawny rufous fur, with a characteristic
pattern of white marks, a thick black and white bushy tail and broad, pointed ears result in a rather
‘foxy’ appearance (Sillero-Zubiri and Marino 2004). This, and its reliance upon small prey, misled
early European naturalists to name this species the Simien fox. Uncertainty over its taxonomy led
to an array of alternative vernacular names, including the Simien jackal, Abyssinian wolf, ky kebero
(Amharic for red jackal) and jedala faarda (Orominia for horse’s jackal).
Unlike other medium- to large-sized canids, which typically are generalist predators and
widely distributed (Macdonald 1992). Ethiopian wolves combine conspicuous sociability with
specialised, solitary foraging for a narrow range of Afroalpine rodents. Today, these wolves are
conned to Afroalpine pockets in a handful of Ethiopian mountains, and total less than 600 individuals,

Walia-Special Edition on the Bale Mountains 62
distributed in seven small and fragmented populations (Marino 2003a). One of at least nineteen
species of mammals restricted to the Afroalpine grasslands and heathlands of Ethiopia (Yalden
and Largen 1992), Ethiopian wolves evolved in the isolation of this huge mountain massif, which
comprises 80% of Africa’s land above 3000m a.s.l. (Yalden 1983; Malcolm and Ashena 1997).
The dominant herbivores in these high altitudes are rodents, particularly molerats (Rhyzomidae)
and grass rats (Muridae), adapted to the extreme diurnal temperature uctuations, and these are
the main prey of Ethiopian wolves (Sillero-Zubiri and Gottelli 1995a; Sillero-Zubiri et al. 1995a,
1995b). As top predators of the Afroalpine ecosystem, Ethiopian wolves attain densities as high
as 1.2 adult per km² in prime habitats (Sillero-Zubiri and Gottelli 1995b) and adult wolves have
no known predators except man. Ironically, the specialisation on Afroalpine rodents that was once
the basis of the species’ success is now the force that constrains Ethiopian wolves to a fragmented
habitat (Yalden and Largen 1992; Marino 2003), and heightens the risk of local extinctions in the
face of stochastic and anthropogenic factors (Sillero-Zubiri and Macdonald 1997; Haydon et al.
2002, 2007).
Study Area
Field studies of Ethiopian wolves in the Bale Mountains began full-time in 1988, and still continue,
having expanded to other populations in Ethiopia since 1997. Most work has taken place in four study
areas: Web Valley (3450 m a.s.l.), Sanetti Plateau (4000 m a.s.l.), Morebawa (3600-3800 m a.s.l.),
and Tullu Deemtu (3800-4300 m a.s.l.) in the central massif of the Bale Mountains National Park
of southern Ethiopia (7°S, 42°E). This is the largest realm of Afroalpine habitat in Africa, spanning
over 1,000 km² and harbouring over half of the global Ethiopian wolf population (Marino 2003a;
Randall et al. this edition). The rst three study areas represent typical open-grassland Afroalpine
habitat and sustain the highest wolf densities (ca. 1.2 wolf/km² in Web and Sanetti, ca. 0.9 wolf/km2
in Morebawa) (Gottelli and Sillero-Zubiri 1992; Sillero-Zubiri 1994; Tallents 2007). Tullu Deemtu
is characterised by Helichrysum dwarf-scrub, also a common habitat type, which sustains a much
lower wolf density (ca. 0.25 wolf/km²).
The Solitary Wolf as the Top Predator of the Afroalpine Rodent Community
Diet
The diet of Ethiopian wolves was studied by scat analysis (689 droppings) and 946 hours of watching
focal animals that yielded 811 attempts to kill prey, of which 361 corresponded to successful kills/
feeds (Sillero-Zubiri and Gottelli 1995a). Rodents accounted for 96% of all prey occurrences in
droppings and 97% by volume of undigested faecal material. Wolf prey included six rodent species,
Stark’s hare Lepus starkii, cattle, birds, insects and undigested sedge leaves, Carex monostachya.
Giant molerat Tachyoryctes macrocephalus – mean weight 618 g - was the main component in the

Walia-Special Edition on the Bale Mountains 63
overall diet (36% of total prey occurrences) and was present in 69% of all faecal samples, whereas
diurnal rats Arvicanthis blicki, Lophuromys melanonyx and Otomys typus (respective mean weight
126 g, 94 g and 100 g) together accounted for 59% of occurrences and appeared in 78% of the
samples. These four species together accounted for 86% of prey occurrences and no signicant
differences were found for these main four prey items between months or between dry and wet
seasons (Sillero-Zubiri and Gottelli 1995a).
Direct observations indicated a higher incidence of large prey (hare, rock hyrax Procavia
capensis capillosa, birds, lambs, and antelopes) than suggested by scat analysis. Of all feeding
instances observed, 69% were grass rats while giant molerat kills accounted for 22% of all successful
attempts. Giant molerats formed the bulk of the prey by weight (40%), while diurnal rats were second
(23%), although taken more often. Carrion, hares, hyraxes, and birds contributed the remaining
36.5% of the total prey weight, of which 12% was scavenged from livestock carcasses.
The diet was broadly similar at the three sites (Web Valley, Sanetti and Tullu Deemtu), with
giant molerat as the single most important food item. In areas where this species is absent or rare it is
often replaced by the common molerat Tachyoryctes splendens. For instance in the Bale Mountain’s
Gaysay Valley, common molerats constituted 32 % of all animals eaten (Malcolm 1997), and in
Menz, central Ethiopia, 31% of occurrences - 17% by volume - in the wolf diet (Ashena 2001).
Analysis of faeces from wolf populations across the Ethiopian highlands conrmed this dietary
specialization, amply dominated by diurnal rodents even where molerats were absent or rare and
livestock abundant (Marino et al. 2010).
Foraging behaviour
During 946 hours of focal observation away from dens, wolves spent 43% of their time foraging
(Sillero-Zubiri and Gottelli 1995a). They foraged solitarily throughout the day, travelling widely at a
walk or trot, covering large areas of their home range. Peaks of foraging activity were synchronised
with the activity of diurnal rodents above the ground. The wolves used various hunting strategies:
molerats were commonly stalked, while zigzag and hole-checks were aimed at grass rats. Although
foraging wolves were mostly observed alone, their daily hunting ranges overlapped considerably.
Of 35 occasions in which more than one wolf was present during kills involving rats, only 23%
were within 10 m. In the remaining observations, wolves shared the same foraging area, but did not
appear to interfere with each others’ foraging attempts or prey captures. Occasionally small packs
hunted hares, antelope calves, and sheep. In 12 of 20 attempts to catch hares, two to four wolves
hunted simultaneously. In the northern grasslands wolves have been observed in packs of three to
four animals hunting reedbuck Redunca redunca (n = 3) and a mountain nyala calf Tragelaphus
buxtoni (Sillero-Zubiri pers. obs.).
Rodents and Ethiopian wolf distribution
The role of the Afroalpine rodent community in limiting the distribution of Ethiopian wolves was
studied by looking at the relationship between wolf abundance and the species composition, relative

Walia-Special Edition on the Bale Mountains 64
abundance and activity pattern of the rodent community in various habitats. Combined biomass of
all diurnal rodents and hares in the Afroalpine grassland habitats was estimated at 24 kg/ha in Sanetti
and 26 kg/ha in Web, with giant molerats contributing about one third of this biomass (we assumed
that an average molerat weighed 618 g (n = 11), and occurred at a biomass of 10-25 kg/ha, with
patches of up to 55 kg/ha) (Sillero-Zubiri et al. 1995a, 1995b). Stark’s hares averaged 2,250 g (n =
4), giving a projected biomass of 0.4-0.7 kg/ha. A more recent study by Tallents (2007) estimated
lower prey biomass, with diurnal rodents represented by 9-12 kg/ha in the short pastures of the Web
Valley (with a peak of 17.1 kg/ha), 4-7 kg/ha in the meadows of Sanetti and Morebawa, and the
uniform Helichrysum heaths of Tullu Deemtu harbouring ca. 1.5-2.5 kg/ha.
Indices of giant molerat biomass for Helichrysum dwarf-scrub and Ericaceous belt were only
1/5 and 1/150 respectively of those in Afroalpine grasslands (Sillero-Zubiri et al. 1995b). Positive
correlations between wolf density and molerat abundance in four areas (Tullu Deemtu, Sanetti,
Web and the Ericaceous belt) suggested that molerats were a vital determinant of wolf presence
(Sillero-Zubiri et al. 1995b). Because they are roughly six times the weight of any other rodent,
hunting T. macrocephalus is likely to be considerably more efcient than hunting a smaller species.
Nonetheless, the positive correlation between wolf abundance and an index of biomass of smaller
rodents showed that the giant molerat was not the only determinant of wolf distribution. The biomass
index for grass rats (in kilograms per 100 transect snap-trap nights) was highest on Sanetti Plateau,
followed, in order, by Web Valley, montane grasslands, the Ericaceous belt and Tullu Deemtu (Table
1). A. blicki and L. melanonyx were the most numerous species in Afroalpine grasslands. Ethiopian
wolf density, measured both from observation and road counts, correlated positively with the total
biomass index and the biomass index for diurnal species, but not for nocturnal species. Also a
positive correlation was detected between rodent burrows and wolf sign (droppings or diggings)
along habitat assessment transects. A similar correlation was found between wolf signs and the
average index of fresh giant molerat signs (Sillero-Zubiri et al. 1995a, 1995b).
Large mammal densities in the Afroalpine grasslands are low and, in any case, they might
be largely unavailable to the wolves. Rodents were the most abundant, conveniently-sized prey,
and easiest to catch. Their availability was more predictable, insofar as their abundance was
closely associated to different habitat types ((Sillero-Zubiri et al. 1995a, 1995b, Marino 2003b,
Tallents 2007). The predictability of the rodent prey may be one selective pressure favouring pack
territoriality (Sillero-Zubiri 1994).

Walia-Special Edition on the Bale Mountains 65
Table 1. Ethiopian wolf density (individuals/km²) and biomass index, weighted for sub-habitat
area, for diurnal and nocturnal snap-trapped rodent prey. The biomass index represents the biomass
(kg) contributed per 100 trap nights using data from all months. Mean weights used as follows:
Arvicanthis blicki: 126g; Lophuromys melanonyx: 94g; L. avopunctatus: 49g; Stenocephalemys
griseicauda: 101.5g; S. albocaudata: 129.5g; Otomys typus: 100g. Home ranges (km2 ± SD) were
estimated as average 100% minimum convex polygons of wolf packs in Bale between 1988-1991.
Group size is the average number of adult and subadults (mean ± SE) in a pack.
Web Sanetti Montane Tullu Ericaceous
Valley Plateau Grassland Deemtu Belt
Biomass index:
Diurnal rats 2.7 2.9 1.6 0.4 0.4
Nocturnal rats 1.8 2.1 1.2 1.4 1.7
TOTAL 4.4 5.0 2.8 1.8 2.1
Ethiopian wolf density:
Road counts 1.0 1.2 0.1 0.2 0.1
Observation 1.2 1.2 0.3 0.2 0.1
Pack home ranges: 6.5 ± 2.1 5.5 ± 1.3 7.4 13.4 ± 2.0 -
Group sizes: 6.7 ± 0.7 4.9 ± 0.3 4.5 ± 0.3 2.6 ± 0.4 -
Wolf Packs Carve Out the Precious Suitable Habitat Available
Pack home ranges and rodent biomass
While the abundant rodent fauna characteristic of the Afroalpine habitat constitutes a very rich
and predictable source of food, this habitat is very restricted in geographic distribution. Between
1988 and 1992 all areas supporting a substantial rodent biomass in Bale were occupied by resident
wolf packs, organised into discrete groups that were spatially and temporally stable (Sillero-Zubiri
and Gottelli 1995b). Groups were composed of 2 to 13 adults and subadults (> 1 year old) and the
average group size for all 14 known packs in Web and Sanetti was 5.9 ± 0.5 (mean ± SE), with Web
packs signicantly larger than Sanetti’s (Table 1). Tullu Deemtu packs were notably smaller and
averaged 2.6 ± 0.4. Subsequent studies indicated similar group size values not differing between
Web and Sanetti (Marino 2003b; Tallents 2007), although across a longer time period group size in
either areas was also affected directly or indirectly by disease (Marino 2003b).
Home ranges of resident wolves overlapped almost completely with other pack members
and entirely contained the home ranges of pups and juveniles (81 to 87 % intragroup annual home
range overlap between adult-adult and adult-subadult dyads, n = 4 packs) (Sillero-Zubiri and Gottelli
1995b). Home ranges of individual residents ranged between 2.0 - 15.0 km² (n = 92) and most of this
variability was attributed to habitat type (Table 1). For instance, combined home ranges (i.e. pack
home ranges estimated as minimum convex polygons) in Afroalpine grasslands averaged 6.5 km² ±

Walia-Special Edition on the Bale Mountains 66
2.1 and 5.5 km² ± 1.3 in Web and Sanetti, respectively (n = 7 packs). In Helichrysum dwarf-scrub
home ranges were twice as large, and explained by the different density of prey species (Table 1).
On the other hand, the home ranges of three non-resident females in Web overlapped widely with
other packs and ranged from 8.5 - 18.7 km², their mean range being signicantly larger than those of
resident dominant females. These oater females disperse and occupy narrow ranges between pack
ranges, awaiting a breeding opportunity (Sillero-Zubiri et al. 1996a).
A follow up study added a population recovery period after the 1992 rabies epizootic and
showed similar home range sizes overall, but the packs that survived attained signicantly larger
territories than before at high density, or than any new packs during the recovery (Marino 2003b).
Tallents (2007) recorded similar results to those of Sillero-Zubiri and Gottelli (1995b) after a rabies
outbreak in 2003, but with somewhat larger home ranges and greater variation. Small ranges,
particularly those recorded in the grasslands and herbaceous communities of Web and Sanetti, are
a reection of the great density of the food resources available in some Afroalpine habitats. The
ranges observed are among the smallest, and density among the highest, reported for all eight Canis
species (Sillero-Zubiri 1994). Established relationships between metabolic rate, body weight and
size of home range in mammals would predict home ranges of 42 km², (Carbone et al. 1999), nearly
eight times the mean values observed in Web and Sanetti.
Different factors drive the numbers of the different age-sex classes in a pack. Home range
size is correlated with group size (Sillero-Zubiri and Gottelli 1995b; Marino 2003b) and the number
of adult males (Tallents 2007) and territories were enlarged whenever a reduction in group size in
a neighbouring pack allowed it, which is indicative of an ‘expansionist strategy’ (sensu Kruuk and
Macdonald 1985). A second adult female was more likely to be found in territories with a greater
proportion of the best molerat habitat, and more subadults were found in higher quality territories
(i.e. with a greater prey density). The advantage of the expansionist strategy is apparent in the
greater areas of high-quality foraging habitat, and greater per capita prey density, in larger territories
(Marino 2003b; Tallents 2007). Under intense competition for rodent-rich grasslands, pack group
size may determine the outcome of territorial boundary clashes and the maintenance of a high quality
range may be the greatest advantage of group-living (Sillero-Zubiri and Macdonald 1998).
Marking and territoriality
Studies of scent-marking behaviour and inter-pack aggression in Ethiopian wolf packs provided
additional evidence of territoriality (Sillero-Zubiri and Macdonald 1998). Movements and activity
at the periphery of ranges was characterised by ‘border patrols’ during which groups of pack
members of both sexes trot and walk along the territory boundary. In 167 km of border patrols
totalling 68 hours, 1,208 scent marks were deposited at an overall rate of 7.2/km. Raised-leg
urinations were the most frequently deposited scent mark (4.7/km), followed by ground scratching
(2.3/km). Defecations and squat urinations during border patrols were rare (0.23/km and 0.04/km
respectively). Scent-marking rates were highest along or near territory boundaries (mean number of
scent-marks deposited per kilometre signicantly greater (F(1,313) = 6.40, P = 0.012) during patrols
than at other times) where wolves vigorously over-marked neighbours’ scent-marks. Most direct

Walia-Special Edition on the Bale Mountains 67
encounters between neighbouring wolves at territory borders were aggressive and involved repeated
chases (102 out of 119 encounters) and the larger group was most likely to win (the larger group won
in 77% of cases, whereas victorious and defeated groups were the same size in 15% of encounters).
Ethiopian wolf packs in Bale occur at saturation density, in a system of highly stable
tessellated territories (Sillero-Zubiri and Gottelli 1995b). Frequent scent-marking, inter-pack
encounters and aversion to strangers’ marks probably constrain each pack to its territory, while
positive feedback keeps each territory boundary adequately marked. A further function of scent-
marking may be to indicate sexual and social status. Wolves in Bale are seasonal breeders and in any
given year mating was synchronised to a period of one to three weeks in the latter part of the rainy
season (August-October), suggesting that a social mechanism triggered mating (Sillero-Zubiri et al.
1998). Scent-marking might allow females to monitor their reproductive condition reciprocally and
synchronise their oestrus. On the other hand, neighbouring packs’ males may gather information on
the receptivity of females. While border encounters occurred throughout the year, peak intrusion
pressure coincided with the mating season. Fifty out of 169 observed encounters between wolves
of neighbouring packs consisted of territorial intrusions by small groups of neighbouring males
attracted by a receptive resident female. Highly seasonal mating may be connected to the occurrence
of a philandering mating system in Ethiopian wolves (Sillero-Zubiri 1994; Sillero-Zubiri et al.
1996a).
Philopatry and the Risk of Inbreeding
Dispersal and philopatry
Lack of suitable habitat places a tight constraint on dispersal in Ethiopian wolves. In Bale immigration
was rare, births and deaths predominated over transfers between packs, and all pack members were
close kin. Relatedness values from microsatellite data were similar for adults (females R = 0.39,
males R = 0.33) and subadults (R = 0.30) (Fig. 1), and approximate relatedness values expected
among half-siblings in this population (Randall et al. 2007). With kin of opposite sex residing in
the same group, natal philopatry provided the potential for inbreeding, in a situation of severely
limited dispersal opportunities (Sillero-Zubiri 1994). Although dispersal was rare, both behavioural
observations and genetic evidence suggest that which occurs is sex biased (Sillero Zubiri et al.
1996a; Randall et al. 2007). In Web and Sanetti, behavioural observations between 1988 and 1992
suggested that 63% of females dispersed at, or shortly before, sexual maturity at two years, some
becoming ‘oaters’ (Sillero-Zubiri et al. 1996a). Under stable demographic conditions Ethiopian
wolf males are philopatric (Randall et al. 2007), however short and long distance male dispersal has
been observed following demographic disturbance due to disease outbreak (Marino 2003b). Pack
ssion and males seeking extra-pack sexually receptive females during the breeding season can also
act as male dispersal into neighbouring territories (see below). The population sex-ratio of adults
was biased toward males at 1.9:1 ± 0.07 (SE), with the mean pack sex-ratio of adults at 2.6:1 ± 0.2
(Sillero-Zubiri and Gottelli 1995b).

Walia-Special Edition on the Bale Mountains 68
Female breeding slots are the most coveted
Adaptive explanations of sex-biased dispersal include avoidance of reproductive competition - for
breeding status or resources - (Dobson 1982) or of inbreeding (Harvey and Ralls 1986). In Ethiopian
wolves observations of same-sex aggression prior to female dispersal support the competition-
for-breeding-status hypothesis (Sillero-Zubiri 1994; Sillero-Zubiri and Macdonald 1998), while
inbreeding avoidance also plays a role (Randall et al. 2007). In Bale, only the dominant female in
each pack typically breeds (Sillero-Zubiri et al. 1996a; Marino 2003b; Randall et al. 2007), indicating
a high level of reproductive competition. Between 1988 and 1992 each breeding female was clearly
dominant over her daughters, and, in all study packs with more than one subordinate female the
lowest ranking female left the group at 18-28 months-old (Sillero-Zubiri 1994). During this period 14
subordinate females emigrated or disappeared from focal packs, whereas only four entered a different
pack and two returned to their natal group, suggesting that approximately 57% of dispersing females
either died or failed to nd residence in the study population. Of the 14 females known prior to
dispersal, 10 settled as oaters next to their natal territory (Sillero-Zubiri et al. 1996a).
Figure 1. Mean pairwise relatedness (± SE) for (a) individuals with known kinship, and (b)
individuals with unknown kinship. Numbers above bars indicate (a) number of dyads or (b) number
of packs from which mean values were calculated (from Randall et al. 2007).

Walia-Special Edition on the Bale Mountains 69
No new packs were formed between 1998 and 1992 (Sillero-Zubiri and Gottelli 1995b)
but an apparent attempt by a subordinate female to split a pack suggested that ssion could be a
mechanism for pack formation (this ended when the subordinate’s litter succumbed, probably killed
by the dominant female) (Sillero-Zubiri et al. 1996a). During this period female ascendancy to
breeding status, either by immigration or inheritance, only occurred after the death of a dominant
and so the chances of a female ever securing a breeding position were low. Five out of 10 dominant
females retained their breeding position throughout four years of observation, whereas the remainder
maintained that role until they died. Breeding openings occurred at an average of 0.12 ± 0.09
opportunities for a subordinate female per year per pack. During contests for a breeding position,
resident females appeared to have an advantage over oaters (three breeding females were replaced
by their daughters after their deaths).
In late 1991 a rabies epidemic decimated the population in Bale and resulted in the
disintegration of three out of ve packs in Web, causing the sudden opening of potential breeding
opportunities (Sillero-Zubiri et al. 1996b; similar epidemics took place in 2003 and 2008, see
Randall et al. 2004 and Johnson et al. 2009). Rather than forming smaller, new breeding units,
the surviving packs maintained their social cohesion and readjusted their territorial boundaries to
occupy the habitat available (Marino 2003b). After a new epidemic in 2003, seven out of nine packs
survived, and again packs remained cohesive and territory boundaries shifted (Tallents 2007). It
took ve years after the rst die-off for new packs to start forming, in one case by the ssion of a
large pack, and twice by the grouping of dispersing individuals, mostly from neighbouring packs.
This suggested that the establishment of a new group depends not only on the availability of high-
quality habitat but also on the presence of sufcient number of helpers to defence the new territory
from the ‘expansionist’ packs that survived (Marino 2003b). The population reduction also opened
breeding vacancies to subordinate females, but social factors delaying formation of new breeding
units maintained population growth low at low densities (Marino 2003b; Marino et al. 2006).
Among Ethiopian wolves, inbreeding avoidance may be an additional adaptive advantage
of female-biased dispersal and may underlie the observed mating behaviour. Male philopatry and
the long tenure of breeding females increase the potential for incest within groups, which may be
countered in part by female dispersal (Sillero-Zubiri et al. 2004). A detailed genetic study of Ethiopian
wolves also found that although mean pairwise relatedness within packs (R = 0.39) was signicantly
greater than that estimated from random assignment of individuals to packs, breeding pairs were
most often unrelated (Randall et al. 2007). This strongly suggests that female-biased dispersal also
reduces the number of incestuous matings in the population. However inbreeding is not entirely
absent in the population. Kinship analysis showed that 11 of 15 (73%) of successful matings were
between unrelated individuals (R = -0.26 to 0.22, P < 0.05). Four apparently incestuous matings
resulted in 9 of 47 (19%) pups being potentially inbred (Randall et al. 2007). Given the potential
negative consequences of inbreeding for the long-term genetic and demographic viability of the
population, inbreeding should be monitored in this population, particularly in the face of further
habitat loss and/or disease outbreaks.

Walia-Special Edition on the Bale Mountains 70
Extra pack copulations and multiple paternity
During the short mating season, the dominant female exercises choice in accepting when and with
which male she mates. Of 30 observed instances of mating that involved copulation between 1988
and 1992, only nine (30%) took place with males from the female’s pack, whereas the other 21
(70%) involved males from other packs (Sillero-Zubiri et al. 1996a, Fig. 2). Within packs, females
copulated only with the dominant male and rejected all mating attempts by lower ranking males
(Sillero-Zubiri 1994; Sillero-Zubiri et al. 1996a). In contrast, mate choice with regard to male status
was not apparent when a female courted and mated with outside males. Microsatellite DNA analysis
has conrmed the occurrence of multiple-paternity in litters (Gottelli et al. 1994; Randall et al. 2007)
and extra-pack paternity (EPP) accounted for 28% of all resolved paternities, occurring in 45% of
litters (Randall et al. 2007). Multiple and extra-pack paternity suggest that extra-pack copulations
(EPCs) are an important reproductive (or tness) tactic for both male and possibly female wolves.
Extra-pack and multiple-paternity may rival male philopatry and female-biased dispersal in
importance as an outbreeding mechanism in a situation where habitat constraints impede dispersal.
An alternative, but non-exclusive, explanation may be the prevention of infanticide from neighbours
(Wolff and Macdonald 2004) who, in this competitive milieu, could benet by killing the offspring
of neighbouring packs.
alpha beta other
alpha beta other
0
20
40
60
80
100
Percentage of observations
copulations (n=30)
rejects (n=46)
Intra-pack males
Extra-pack males
Figure 2. Frequency with which Ethiopian wolf females were observed in sexual encounters with
males from their packs or neighbouring packs.

Walia-Special Edition on the Bale Mountains 71
Cooperative Breeding
Role of helpers at the den
Studies of wolf social behaviour conducted during 1988-1992 included detailed observations
of behaviour at 20 dens where pups emerged. In those, all wolves helped to rear the litter of the
dominant female, guarding the den, chasing potential predators, and regurgitating or carrying rodent
prey to feed the pups (Sillero-Zubiri 1994). Given the high degree of relatedness among group
members (Randall et al. 2007), subordinate wolves may increase the indirect component of their
inclusive tness by acting as helpers or, if competition within the group is intense, subordinates
may be induced to help as a ‘payment’ for remaining in the territory. Ethiopian wolf males tend to
help throughout their lives and never disperse; the dominant male at least shared paternity, whereas
subordinate males appear generally to have no probability of fathering the pups, nevertheless they
still help. Subordinate females help more intensely than do males for one or two years before
dispersing or inheriting the breeding position. The balance of costs and benets to all participants
in a cooperative breeding system have been widely debated (e.g., Solomon and French 1997). One
aspect of the debate is whether groups containing many helpers deliver more food and care to the
young than do smaller groups.
The development of the young wolves is divisible into three broad stages (Sillero-Zubiri
1994). First, early denning (birth to four weeks) when the pups are conned to the den and are
entirely dependent on milk. Second, mixed nutritional dependency (week ve to week 11) when
milk is supplemented by solid foods such as rodents provisioned by all pack members until pups are
completely weaned. And third, post-weaning dependency (week 12 to six months) when the pups
subsist almost entirely on solid foods supplied by breeders and non-breeding helpers. Juveniles are
considered independent after six months, when they cease receiving appreciable quantities of food
from adults. Juvenile become subadults and are ‘recruited’ at one year of age.
Although the mother and putative father spend more time at the den on average than do
other wolves, some non-breeders spend more time at the den than do the breeders themselves. The
proportion of time pups were left unattended declined signicantly as the number of helpers in the
pack increased. Pack size may thus inuence anti-predator behaviour, because baby-sitters were
active in deterring and chasing potential predators. Unattended young might be taken by spotted
hyaenas Crocuta crocuta, domestic dogs, honey badgers Melivora capensis and eagles Aquilla
verreauxi, A.rapax. On the other hand, no evidence has been found that increases in pack size result
in measurable increases in numbers of pups at any age.
In the original 1982-1992 study observations were made of nine wolf packs during the
breeding season to quantify the amount of solid food provisioned to pups. Non-maternal food
provisioning for 17 litters constituted 67% of feedings observed other than nursing. Independent of
the number of donors, there were signicant differences in the rate of food contributions per hour
by individuals of different breeding status, sex or age (F(5,119) = 9.08, P < 0.0001; Table 2). Breeders
contributed signicantly more food than did non-breeders, and females more than males. Breeding

Walia-Special Edition on the Bale Mountains 72
females contributed more than any other wolves, dominant males came second, and non-breeding
males contributed on average the least food. When the net contribution rate was considered (i.e.
food items contributed minus items eaten by the individual helper), breeding females were still the
most generous individuals, followed by subadult females, which contributed more than any other
non-breeder.
Table 2. Individual contributions of Ethiopian wolves to cooperative pup-care in relation to
reproductive status, sex, and age during 2,115h of den observations. Values are mean ± SD. Sample
size was 17 breeding males, 18 breeding females, 49 non-breeding adult males, 11 non-breeding
adult females, 16 subadult males and 12 subadult females. Baby-sitting measured as the percentage
of observation time in which individuals of a given category were present within 200m of the den.
Feeding measured in number of solid food items (i.e. whole rodents or regurgitations) contributed
per hour.
Breeders Non-breeders
Behaviour/Age Male Female Males Females
mean ± SD mean ± SD mean ± SD mean ± SD
Babysitting
Visits per hour:
Adults 1.9 ± 0.6 2.6 ± 0.7 1.1 ± 0.7 1.3 ± 1.2
Subadults 1.3 0.7 1.6 ± 1.1
Percentage of time:
Adults 17.3 ± 8.6 23.6 ± 11.4 11.4 ± 9.5 11.4 ± 9.5
Subadults 8.6 ± 8.4 11.1 ± 8.8
Grooming rate:
Adults 0.04 ± 0.05 0.06 ± 0.06 0.02 ± 0.03 0.04 ± 0.08
Subadults 0.03 ± 0.04 0.03 ± 0.05
Food Provisioning
Hourly total food contribution:
Adults 0.06 ± 0.05 0.12 ± 0.09 0.03 ± 0.03 0.04 ± 0.06
Subadults 0.03 ± 0.03 0.06 ± 0.07
Hourly net food contribution
Adults 0.05 ± 0.04 0.10 ± 0.08 0.02 ± 0.03 0.04 ±0.04
Subadults 0.02 ± 0.04 0.06 ± 0.04
The prediction of a positive correlation between the total amount of food delivered to the
pups, and the number of non-breeding helpers present was not supported in this study, since the
presence of helpers did not increase feeding frequency at the den (rs = 0.18, n = 7, P > 0.05). However,
while the total food-provisioning rate did not increase signicantly with the number of contributors
to the den, the share contributed by non-breeding helpers did do so. Food contributions by non-
breeders were accompanied by reduced parental input in pup rearing - reducing food contributions
by the dominant male (rs = -0.63, n = 7, P > 0.05) and female (rs = -0.85, n = 7, P < 0.05) - and hence
a reduction in energy expenditure by the breeding pair (Fig. 3).

Walia-Special Edition on the Bale Mountains 73
0
0.1
0.2
0.3
0.4
0 2 4 6 8 10
Number of helpers
Mean feeding rate (contributions/hour)
dominant pair
non-breeders
Figure 3. Rate of feeding pups (contribution of solid foods per hour) by the dominant pair in
relation to the number of non-breeder helpers.
Pup survival
Between one and six pups emerged from each breeding den, and the observed average litter size for
20 litters born between 1988 and 1991 was 4.1 ± SE 0.36. While there was no signicant difference
in litter size between years (F(3,16) = 1.10, P = 0.379), the presence of an additional nursing female
at the den was associated with smaller litter size (see Allo-suckling below). The number of pups
emerging from the den was not signicantly correlated with the number of adults and subadults at the
den (rs = -0.26). The hypothesis that the number of non-breeding helpers enhances the reproductive
output of the group was not supported either. Pups survival at emergence was not correlated with
the number of non-breeding helpers (rs = -0.26, n = 20, P > 0.05), nor was survival at whelping (rs =
-0.28, n = 20, P > 0.05). Similarly, there was no signicant correlation with survival at six months,
one or two years and the number of non-breeding helpers.
Marino (2003b) looked at relationships between wolf group size and reproduction or survival
in a time series between 1988 and 2000 and conrmed the absence of a simple, direct relationship
between them. Instead, wolf density within territories appears to be negatively correlated to the
number of pups emerging from the den and pup survival and the survival of adults at the population
level correlates with density. Intriguingly, Tallents (2007) studied a subset of packs and found
litter size from emergence to 11 months was correlated with the number of adult males in the
pack, potentially through their role in securing high quality territories. There may be other effects
mediated by prey availability, with the impact of helpers possibly only being felt when times are
hard. Between weaning and ten months, juvenile survival was strongly correlated with the extent
of high quality foraging habitat, a pattern probably mediated through access to high prey densities
for the provisioning adults and subadults, and also through the juveniles’ own foraging success as

Walia-Special Edition on the Bale Mountains 74
they moved towards independence. Additionally, juveniles aged 10-12 months had lower survival
in territories with a greater degree of overlap with neighbouring packs, indicating they may be
vulnerable to foraging competition and intraspecic aggression, particularly during the mating
season when territorial incursions increase (Tallents 2007).
Allo-suckling
The most extreme manifestation of cooperative care by Ethiopian wolves involves nursing the
offspring of the dominant female, or allo-suckling (Sillero-Zubiri 1994, Sillero-Zubiri et al. 2004).
Of the 20 successful breeding attempts observed eight dens had a subordinate female acting as allo-
suckler. These females, two years-old females or older, often were closely related to the breeder and
at least two showed signs of pregnancy. Allo-suckling obviously has the potential to confer benets
to infants, and reduce the mother’s energetic costs. For example pups with access to two lactating
females were suckled signicantly more often (0.43 ± 0.05 SE bouts per hour versus`0.26 ± 0.03;
t = 2.78, df = 60, P = 0.007). Pups whose mother was assisted by an allo-suckler received a higher
energetic input per capita until weaning (weeks 4 to 18) and enjoyed better survival than did those
nursed by their mother alone. Dominant females apparently benet from allo-suckling by sharing
the costs of lactation, and thus lowering their per capita suckling frequency, without affecting a
reduction in the pups’ overall milk intake. For a more detailed treatment of energetic contributions
of allo-suckling see Sillero-Zubiri et al. (2004).
Although one might expect up to double the average litter size of pups in those with two dens
that was not the case; mean litter size on emergence from 12 dens with only one nursing female was
5.1 ± SE 0.35, signicantly larger than that from eight dens with an allo-suckler, 2.6 ± SE 0.39 (t =
4.88; df = 16; P = 0.0002). Tallents’ data (2007) supports these results, with a lower proportion of
the litter surviving from three to eight months in packs with more adult and subadult females. The
presence of an allo-suckler was associated with distinct social unease in the pack and evident tension
between the dominant and subordinate females, suggesting that female aggression inside the den
may have an inuence on pup mortality prior to emergence (Sillero-Zubiri 1994).
The foregoing results raise several interesting puzzles, which we hope the continuing
research of the Ethiopian Wolf Conservation Programme will resolve. First, to the extent that the
allo-sucklers do indeed make a long-term contribution to pup survival, this is initially disguised
by the counter-intuitive earlier effect of litter reduction. Although the helpers in general, and allo-
sucklers in particular, appear to work assiduously for the well-being of the pups, and notwithstanding
the rather large size of our data set, demonstrating any survival benet is difcult at best. Perhaps
such benets are conditional upon circumstances. One intriguing speculation is that males nursed
by two females do well: one such male grew up to acquire the dominant male position in his pack,
another became dominant male in a pack with six adult males, and three survived a rabies epizootic
in which nearly all other pack members perished (Sillero-Zubiri et al. 1996b). Additionally, the
benet of the presence of an allo-suckler in a pack of Ethiopian wolves might be contingent on
the availability of prey. As suggested above, in a good year, unassisted females may not need help,
but allo-suckler assistance might be important in harsh years. In a scenario where the chances of

Walia-Special Edition on the Bale Mountains 75
successful dispersal are very low, concentrating resources in fewer, tter, individuals might raise
their prospects of securing a dominant position, and eventually breeding status.
Biogeography and Genetics: The Big Picture
Phylogenetic analyses of mitochondrial DNA indicate that the closest living relatives of Ethiopian
wolves probably are the gray wolves Canis lupus and coyotes (Gottelli et al. 1994; Wayne and
Gottelli 1997; Vilà et al. 1999) and the estimated coalescence for the species is of ≈ 102,602 years
ago (Gottelli et al. 2004) - from a mean sequence divergence of ~ 1.0% and assuming a divergence
rate of 10%/Myr for the canid control region I as in Vilà et al. 1999). The history of the species thus
appears to be relatively short. One evolutionary interpretation is that the Ethiopian wolf is a relict
form of a Pleistocene invasion into East Africa of a gray wolf-like progenitor (Wayne et al. 2004),
pre-adapted to the cold temperatures of the Ethiopian highlands. Fossils of gray wolf-like canids are
known from Europe from the late Pleistocene, and land bridges into Northeast Africa existed when
the climate was generally cooler and drier than present (Kingdon 1990). At the onset of the last
glaciation, approximately 100,000 years ago, the Afroalpine habitats in Ethiopia were vast allowing
this pre-adapted canid to ourish; while competition with well-established tropical carnivores may
have prevented it from advancing further south into East Africa (Gottelli et al. 2004).
The end of the Pleistocene roughly 12,000 years ago brought the most recent change in the
climate, and the extensive Ethiopian Afroalpine steppes shrunk to their present state, reducing the
habitat available to Ethiopian wolves by an order of magnitude (Gottelli and Sillero-Zubiri 1992;
Gottelli et al. 2004). Microsatellite and mitochondrial DNA suggest that small population sizes and
multiple population bottlenecks may have characterised the recent evolution of Ethiopian wolves
due to climatic effects on habitat availability (Gottelli et al. 2004) and disease outbreaks. Indeed,
the Ethiopian wolf appears to have the most limited genetic variability at the population level of any
extant canid (Gottelli et al. 1994, Wayne et al. 1997).
Across the extant populations and within populations, mitochondrial DNA and nuclear
microsatellites revealed a strong genetic structuring, indicating isolation and lack of gene ow
(Gottelli et al. 2004, Randall et al. 2010, Gottelli et al. unpublished data). The most parsimonious
genetic subdivisions across populations corresponded to three mountain areas: Wollo/NE Shoa,
Simien/Mt. Guna/Mt. Choke, and Arsi/Bale, a pattern closely matching the geographical history of
Afro-alpine fragmentation. The highest genetic variability was found in the populations north of the
Rift valley (Gottelli et al. 2004, unpublished data). The Arsi Mountains population, represented in
the study by current and historical genetic samples, also showed that haplotypes and alleles unique to
this population have been lost over the last 100 years. There is also signicant genetic structuring at
the population level, both within packs and within subpopulations in the Bale Mountains, emerging
as result of intense sociality, limited dispersal, and localized disease outbreaks causing population
bottlenecks (Randall et al. 2010). Given the current levels of genetic isolation and small population
sizes the conservation and genetic management of these populations is critical to preserve them as
the main reservoir of genetic variability in the species.

Walia-Special Edition on the Bale Mountains 76
The Cost of Specialisation: A Conservation Challenge
The apparently sterile Afroalpine steppes of Ethiopia support a rodent biomass which is spatially
and temporally predictable, and comparable to other rodent-rich habitats elsewhere in Africa, which
may explain why the Ethiopian wolf is the only canid to specialize so completely on rodents (Sillero-
Zubiri et al. 1995a). The rodents’ distribution and diurnal activity also concur with Ethiopian wolves’
diurnal and solitary foraging habits, and their connement to Afroalpine habitats over 3,000 m a.s.l..
Global warming during the last 10,000 years progressively conned the Afroalpine ecosystem to
the highest mountains, and today 60% of all Ethiopian land above 3,000 m has been converted to
farmland (Marino 2003a).
Ethiopian wolves face threats that arise from their isolation, small size, and the increasing
contact with humans and disease transmission from domestic dogs (Sillero-Zubiri and Marino
2004). Transmission of rabies is the main threat for wolves in Bale population and can have serious
consequences for small populations (Haydon et al. 2002, 2006; Randall et al. 2004, 2006). Elsewhere
habitat loss is ever increasing the risk of population extinctions. Two small populations became
extinct when suitable wolf range shrunk below 20 km² in recent years (Marino 2003a), but seven
still survive in Afroalpine ranges across the country, with Bale the largest (Marino 2002a; Randall
et al. this edition). Aspects of the demography of wolves in Bale, particularly their high adult
survivorship, also stress the resilience and stability of wolf populations (Marino 2003b; Marino
et al 2006; Haydon et al. 2002). Although their small, fragmented populations are a poor omen for
Ethiopian wolves, their concentration in a few clearly dened sites, their charisma and, we hope, a
fair understanding of their biology, lend hope that with unwavering commitment from all concerned
and adherence to a well-founded management plan (Sillero-Zubiri and Macdonald 1997; Sillero-
Zubiri et al. 2004) they will survive. Protective measures require the consolidation of the General
Management Plan for the Bale Mountains National Park (OARDB 2007) and other protected areas,
and active efforts to monitor and protect all remaining populations.
Acknowledgements
We thank the Ethiopian Wildlife Conservation Authority and the Bale Mountains National Park
for permission to undertake this research, and all the staff of the Ethiopian Wolf Conservation
Programme (EWCP) who assisted collecting data in the last 20 years. EWCP is a partnership
between the Oxford University’s WildCRU and the Ethiopian authorities under the auspices of the
IUCN/SSC Canid Specialist Group. We thank Born Free Foundation, Frankfurt Zoological Society,
Wildlife Conservation Network, Wildlife Conservation Society, Peoples’ Trust for Endangered
Species and many others for their nancial support.

Walia-Special Edition on the Bale Mountains 77
References
Ashena, Z.T. 2001. Common property resource management of an Afroalpine habitat: supporting
a population of the critically endangered Ethiopian wolf (Canis simensis). DPhil Thesis.
Durrel Institute of Conservation and Ecology. University of Kent, UK.
Ashena, Z.T., T. Coulson, Sillero-Zubiri, C. andLeader-Williams, N. 2005. Behaviour and ecology
of the Ethiopian wolf in a human-dominated landscape outside protected areas. Animal
Conservation, 8: 113-121.
Carbone, C., Mace, G.M., Roberts S.C., and Macdonald, D.W. 1999. Energetic constraints on the
diet of terrestrial carnivores. Nature, 402: 286-288
Dobson, F.S. 1982. Competition for males and predominant juvenile male dispersal in mammals.
Animal Behaviour, 30: 1183-1192.
Gottelli, D., and Sillero-Zubiri, C. 1992. The Ethiopian wolf - an endangered endemic canid. Oryx,
26: 205-214.
Gottelli, D., Sillero-Zubiri, C., Applebaum, G.D., Roy, M.S., Girman, D.J., Garcia-Moreno, J.,
Ostranders, E.A. and Wayne, R.K. 1994. Molecular genetics of the most endangered canid:
The Ethiopian wolf Canis simensis. Molecular Ecology, 3:301-312.
Gottelli, D., Marino, J., Sillero-Zubiri, C. and Funk, S.M. 2004. The effect of the last glacial age
on speciation and population genetic structure of the endangered Ethiopian wolf (Canis
simensis). Molecular Ecology, 13: 2275-2286.
Harvey, P.H., and Ralls, K. 1986. Do animals avoid incest? Nature, 320: 575-576.
Haydon, D.T., Laurenson, M.K. and Sillero-Zubiri, C. 2002. Integrating epidemiology into
population viability analysis: managing the risk posed by rabies and canine distemper to the
Ethiopian wolf. Conservation Biology, 16: 1372-1385.
Haydon, D.T., Randall, D.A., Matthews, L., Knobel, D.L., Tallents, L.A., Gravenor, M.B., Williams,
S.D., Pollinger, J.P., Cleveland, S., Woodhouse, M.E.L., Sillero-Zubiri, C., Marino, J.,
Macdonald, D.W. and Laurenson, M.K. 2006. Low-coverage vaccination strategies for the
conservation of endangered species. Nature, 443: 692-695.
Johnson, N., Manseld, K.L. Marston, D.A., Wilson, C., Goddard, T., Selden, D., Hemson, G.,
Edea, L., van Kesteren, F., Shiferaw, F., Stewart, A.E., Sillero-Zubiri, C., Fooks, A.R. 2010.
A new outbreak of rabies in rare Ethiopian wolves (Canis simensis). Archives of Virology
155: 1175–1177
Kingdon, J. 1990. Island Africa. Collins, London.
Kruuk, H., and Macdonald, D.W. 1985. Group territories of carnivores: empires and enclaves.
Pp. 521-536 In Behavioral Ecology: Ecological Consequences of Adaptive Behavior (R.M.
Sibly and R.H. Smith, eds.). Blackwell Scientic Publications, Oxford, UK.
Malcolm, J. 1997. The diet of the Ethiopian wolf (Canis simensis Rüeppell) from a grassland area of
the Bale Mountains, Ethiopia. African Journal of Ecology, 35: 162-164.
Malcom, J.R. and Ashena, Z.T. 1997. Conservation of Afroalpine habitats. Pages 61-63 In The
Ethiopian Wolf Status Survey and Conservation Action Plan. (Sillero Zubiri, C. and &
Macdonald, D.W. (Eds)). IUCN, Gland, Switzerland & Cambridge, UK.
Marino, J. 2003a. Threatened Ethiopian wolves persist in small isolated Afroalpine enclaves. Oryx,
37:62-71.
Marino, J. 2003b. Spatial ecology of Ethiopian wolf evolution, population dynamics and territoriality.
D.Phil. dissertation. Oxford University, Oxford, UK.
Marino, J., Sillero-Zubiri, C. and Macdonald, D.W. 2006 Trends, dynamics and resilience of an
Ethiopian wolf population. Animal Conservation, 9: 49-58.
Marino, J., R. Mitchell, and P.J. Johnson. 2010. Dietary specialization and climatic-linked variations
in extant populations of Ethiopian wolves. Journal of African Ecology, 48: 517-525

Walia-Special Edition on the Bale Mountains 78
OARDB. 2007. Bale Mountains National Park General Management Plan. Oromia Agriculture and
Rural Development Bureau, Oromia National Regional State, Ethiopia.
Randall D.A. 2006. Determinants of genetic variation in the Ethiopian wolf, Canis simensis. D.Phil.
dissertation, University of Oxford, UK.
Randall, D.A., Williams, S.D., Kuzmin, I.V., Rupprecht, C.E., Tallents, L.A., Tefera, Z., Argaw,
K., Shiferaw, F., Knobel, D.L., Sillero-Zubiri, C. and Laurenson, M.K. 2004. Rabies in
endangered Ethiopian wolves. Emerging Infectious Diseases, 10 (12): 2214-2217.
Randall, D.A., Marino, J., Haydon, D.T., Sillero-Zubiri, C., Knobel, D.L., Tallents, L.A., Macdonald,
D.W., Laurenson, M.K. 2006. An integrated disease management strategy for the control of
rabies in Ethiopian wolves. Biological Conservation, 131: 151-162.
Randall, D.A., Pollinger, J.P., Wayne, R.K., Tallents, L.A., Johnson, P.J., and Macdonald, D.W.
2007. Inbreeding is reduced by female-biased dispersal and mating behavior in Ethiopian
wolves. Behavioral Ecology, 18: 579-589.
Randall, D.A., Pollinger, J.P., Argaw, K, Macdonald, D.W., and Wayne, R.K. 2010. Fine scale genetic
structure in Ethiopian wolves imposed by sociality and population bottlenecks. Conservation
Genetics, 11: 89-102
Sillero-Zubiri, C., and Gottelli. D.1994. Canis simensis. Mammalian Species, 1-6.
Sillero-Zubiri, C., and Gottelli, D. 1995a. Diet and feeding behavior of Ethiopian wolves (Canis
simensis). Journal of Mammalogy, 76:531-541.
Sillero-Zubiri, C., and Gottelli, D. 1995b. Spatial organization in the Ethiopian wolf Canis simensis:
Large packs and small stable home ranges. Journal of Zoology London, 237:65-81.
Sillero-Zubiri, C., Gottelli, D. and Macdonald, D.W. 1996a. Male philopatry, extra-pack copulations
and inbreeding avoidance in Ethiopian wolves (Canis simensis). Behavioral Ecology and
Sociobiology, 38: 331-340.
Sillero-Zubiri, C., Hoffmann, M. and. Macdonald, D.W (Eds.). 2004. Canids: foxes, wolves, jackals
and dogs: status survey and conservation action plan, second edition. IUCN Canid Specialist
Group, Gland, Switzerland and Cambridge, UK.
Sillero-Zubiri, C., Johnson, P.J. and Macdonald, D.W. 1998. A hypothesis for breeding synchrony
in Ethiopian wolves (Canis simensis). Journal of Mammalogy, 79: 853-858.
Sillero-Zubiri, C., King, A.A. and Macdonald, D.W. 1996b. Rabies and mortality in Ethiopian
wolves (Canis simensis). Journal of Wildlife Diseases, 32: 80-86.
Sillero-Zubiri, C., and Macdonald, D.W. 1997. The Ethiopian wolf: status survey and conservation
action plan. IUCN, Gland, Switzerland and Cambridge, UK.
Sillero-Zubiri, C., and Macdonald. D.W. 1998. Scent-marking and territorial behaviour of Ethiopian
wolves Canis simensis. Journal of Zoology London, 245: 351-361.
Sillero-Zubiri, C. and Marino, J. 2004. Ethiopian wolf (Canis simensis). Pp 167-174. In Canids:
Foxes, Wolves, Jackals and Dogs: Status Survey and Conservation Action Plan (Sillero-
Zubiri, C., Hoffmann, M. and Macdonald, D.W. (Eds.)). Second edition. IUCN Canid
Specialist Group, Gland, Switzerland and Cambridge, UK.
Sillero-Zubiri, C., Marino, J., Gottelli, D. and Macdonald, D.W. 2004. Afroalpine ecology, solitary
foraging and intense sociality amongst Ethiopian wolves. In The Biology and Conservation of
Canids (Macdonald, D.W., and Sillero-Zubiri, C. (Eds.)). Oxford University Press, Oxford.
pp. 311-323
Sillero-Zubiri, C., Tattersall, F.H. and Macdonald, D.W. 1995a. Bale mountains rodent communities
and their relevance to the Ethiopian wolf (Canis simensis). African Journal of Ecology, 33:
301-320.
Sillero-Zubiri, C., Tattersall, F.H. and Macdonald, D.W. 1995b. Habitat selection and daily activity
of giant molerats Tachyoryctes macrocephalus: Signicance to the Ethiopian wolf Canis
simensis in the Afroalpine ecosystem. Biological Conservation, 72: 77-84.

Walia-Special Edition on the Bale Mountains 79
Solomon, N. G., and French, J.A. Eds. 1997. Cooperative Breeding in Mammals. Cambridge
University Press, Cambridge, UK.
Tallents, L. A. 2007. Determinants of reproductive success in Ethiopian wolves. Oxford University,
Oxford UK.
Vila C, Amorim, I.A., and Leonard, J.A. 1999 Mitochondrial DNA phylogeography and population
history of the grey wolf, Canis lupus. Molecular Ecology, 8: 2089-2103.
Wayne, R., Geffen, E and Vila, C. 2004. Population and conservation genetics of canids. In The
biology and conservation of canids. (Macdonald, D.W. and Sillero-Zubiri, C (Eds.)). Oxford
University Press, Oxford, UK. Pp 55-84
Wayne, R, and Gottelli, D. 1997. Systematics, population genetics and genetic management of
the Ethiopian wolf. In The Ethiopian Wolf - Status Survey and Conservation Action Plan.
(Sillero-Zubiri, C and Macdonald, D.W. (Eds.)) IUCN/SSC Canid Specialist Group. Gland,
Switzerland and Cambridge, UK. Pp. 43-50.
Wolff J.O and Macdonald, D.W. 2004. Promiscuous females protect their offspring. Trends in
Ecology and Evolution,19: 127-134.
Yalden, D.W. 1983. The extent of high ground in Ethiopia compared to the rest of Africa. Sinet:
Ethiopian Journal of Science, 6: 35-38.
Yalden, D.W., and Largen, M.J. 1992. The endemic mammals of Ethiopia. Mammal Review, 22:
115-150.