ORIGINAL PAPER
Reproductive sharing and proximate factors mediating
cooperative breeding in the African wild dog (Lycaon pictus)
Penny A. Spiering & Michael J. Somers &
Jes?s E. Maldonado & David E. Wildt &
Micaela Szykman Gunther
Received: 18 May 2009 /Revised: 20 October 2009 /Accepted: 21 October 2009 /Published online: 10 November 2009
# Springer-Verlag 2009
Abstract Although dominant African wild dogs (Lycaon
pictus) are generally believed to be the sole breeders within
a pack, earlier behavioral and endocrine data suggest that
reproduction could be shared with subordinates. We
performed an extensive behavioral, demographic, and
genetic evaluation of a wild dog population in South Africa
to examine the level of such sharing and the proximate
mechanisms influencing reproductive contributions of each
sex. While a majority of pups were born to dominants
because of a lack of subordinate potential breeders, we
discovered a substantial portion of reproductive sharing
between dominants and subordinates. Compared with alpha
females that mated annually, subordinate beta females bred
in 54.5% of years whereas thetas never bred. The three top-
ranking males all sired pups (56.0%, 32.0%, and 12.0%,
respectively) when three or more adult males were present.
With only two pack males, alpha and beta individuals
shared reproduction nearly equally (55.2% and 44.8%,
respectively), and litters of mixed paternity were discovered
on eight of 15 (53.3%) occasions. A skewed adult sex-ratio
and frequent alpha mortalities for females and behavioral
aggression in males allowed most individuals to attain
dominant status in their lifetime, creating a constantly
shifting social hierarchy. Genetic parentage results
corresponded to reported hormone profiles, suggesting
physiological suppression in some lower-ranked individuals
of both sexes. Thus, a combination of demographic,
behavioral, and hormonal proximate factors mediates
reproductive partitioning in wild dogs. We conclude that
reproductive sharing can be significant in this species,
especially for males that have less robust suppressive
mechanisms than females.
Keywords Dominance . Lycaon pictus . Multiple
parentage . Proximate mechanisms . Reproductive sharing .
Subordinate breeding
Introduction
Cooperative breeders live in social groups where reproduc-
tion is partitioned among members to varying degrees, and
individuals other than parents help care for offspring
(Faulkes and Bennett 2001). In some such species, such
as the naked mole rat (Heterocephalus glaber, Clarke and
Faulkes 1997), dwarf mongoose (Helogale parvula, Rood
1990), and jackal (Canis mesomelas and Canis aureus,
Moehlman 1979), social dominants nearly completely
monopolize copulations, and reproduction is rare or
nonexistent for subordinate adults, even of prime age.
Communicated by A. Schulte-Hostedde
P. A. Spiering (*) : D. E. Wildt : M. S. Gunther
Center for Species Survival, Conservation and Research Center,
National Zoological Park, Smithsonian Institution,
Front Royal, VA 22630, USA
e-mail: SpieringP@si.edu
P. A. Spiering : J. E. Maldonado
Center for Conservation and Evolutionary Genetics,
National Zoological Park, Smithsonian Institution,
Washington, DC 20008, USA
P. A. Spiering : M. J. Somers
Centre for Wildlife Management, University of Pretoria,
Pretoria 0002, South Africa
M. J. Somers
Centre for Invasion Biology, University of Pretoria,
Pretoria 0002, South Africa
M. S. Gunther
Department of Wildlife, Humboldt State University,
Arcata, CA 95521, USA
Behav Ecol Sociobiol (2010) 64:583?592
DOI 10.1007/s00265-009-0875-6
Other species express intermediate levels of reproductive
partitioning, with some subordinates being reproductively
successful, as in the Seychelles warbler (Acrocephalus
sechellensis, Richardson et al. 2001), redfronted lemur
(Eulemur fulvus rufus, Kappeler and Port 2008), and
Jamaican fruit-eating bat (Artibeus jamaicensis, Ortega et
al. 2008). Still, others express almost equal division in
reproductive capacity among group members, as in the
coati (Nasua narica, Russell 1983) and banded mongoose
(Mungos mungo, Creel and Waser 1991). This substantive
diversity among species suggests that there are complex
mechanisms regulating reproductive sharing. Such infor-
mation is crucial for understanding the processes of sexual
selection, reproductive skew, and inclusive fitness. Yet,
studies to precisely identify these driving factors are
challenging due to the lack of large-scale genetic and
behavioral analyses of wild populations that include lineage
relationships.
Comprehensive studies reveal that proximate mecha-
nisms of subordinate reproductive suppression may encom-
pass behavioral, physiological, and/or demographic factors
(French 1997). Behavioral mechanisms, such as aggression
(Reyer et al. 1986), can profoundly influence distribution of
reproduction with dominants physically blocking a sub-
ordinate's access to mates, as observed in the gray wolf
(Canis lupus, Packard et al. 1985). Mate guarding is a
similar tactic (Birkhead and Moller 1992) used by the
mustached tamarin (Saguinus mystax, Huck et al. 2004).
Physiological suppression by means of stifling a prominent
reproductive hormone (i.e., testosterone) has been effective,
as seen in the naked mole-rat (Faulkes and Abbott 1991).
An indirect strategy may also occur, as when subordinates
produce elevated glucocorticoids (stress hormones) that, in
turn, compromise normal reproduction, as reported for the
song sparrow (Melospiza melodia, Wingfield and Silverin
1986). A combination of behavioral and hormonal cues is
also well known to limit subordinate reproduction in the
common marmoset (Callithrix jacchus, Barrett et al. 1993)
and cotton-top tamarin (Saguinus oedipus, Savage et al.
1988). In the context of demographics, group size and
composition are significant. For example, chances for
subordinate male savannah baboons (Papio cynocephalus,
Alberts et al. 2003) to breed are vastly improved in larger
troops due to increased energetic costs of mate guarding for
dominants. Thus, the mechanisms associated with repro-
ductive contributions appear as diverse as the phenotypes of
the cooperatively breeding species studied to-date.
To improve our understanding of the phenomenon of
reproductive sharing, we chose the endangered African
wild dog (Lycaon pictus) as a study species, largely because
of its well-recognized, highly cooperative, and complex
social system (Estes and Goddard 1967). Although natu-
rally found at low densities, packs as large as 27 individuals
are known to range an average of 12 km daily to hunt prey
(Creel and Creel 2002). The standard model of a wild dog
pack consists of a dominant breeding pair, several subor-
dinate non-breeding adults (usually siblings of the same-sex
dominant individual), and subordinate offspring of the
alpha pair (Girman et al. 1997). After a birth (usually once
per year in a given pack), all group members cooperate in
provisioning the lactating female in the den and feeding/
protecting pups after emerging from underground (Malcolm
and Marten 1982; Creel and Creel 2002). After 1 year of
age, offspring disperse as far as 250 km from the natal pack
territory in single-sex, sibling groups (Fuller et al. 1992) in
search of opportunities to join other dispersers or an already
established group (Frame et al. 1979). Once a new breeding
pack forms, a clear social dominance hierarchy develops
within each gender, with reproductively ?prime?, middle-
aged males and the oldest female usually holding dominant
positions (Creel and Creel 2002).
The prevailing view has been that dominant female and
male wild dogs are the sole breeders within a pack, with
most pups resulting from this pairing (Malcolm and Marten
1982; Girman et al. 1997). One study estimated that alpha
females and males achieve ~96% of total annual reproduc-
tive success through mating and offspring production
compared with only ~10% for occasionally breeding
subordinates (Girman et al. 1997). However, exceptions to
alpha reproductive dominance have been reported, with
behavioral observations suggesting that subordinate females
have produced 19% of litters in Kruger National Park
(South Africa; Reich 1981), 24% in the Selous (Tanzania;
Creel and Creel 2002), and 25% in the Serengeti ecosystem
(Tanzania; Malcolm and Marten 1982). Only three small-
scale genetic analyses have compared reproductive con-
tributions between wild dog social classes. In a study of
nine packs in Kruger National Park, Girman et al. (1997)
found that subordinate females and males produced only
8% and 10% of pups, respectively. Creel and Creel (2002)
examined two litters in the Selous and found one pup in
each litter not sired by the alpha male. Moueix (2006)
reported that at least one pup in each of five sampled
litters from Madikwe Game Reserve and Pilanesburg
National Park (South Africa) was not offspring of the
alpha male.
We conducted the first large-scale molecular genetics,
demographic, and behavioral assessment of a wild dog
population in KwaZulu-Natal (KZN) Province, South
Africa, in order to answer three important questions about
the relationship between social rank and reproductive
opportunities in this species. First, how much reproduction
is shared between dominants and subordinates in a breeding
pack? Second, what behavioral, demographic, or physio-
logical factors are influencing distribution of reproductive
opportunities? And third, does the degree of reproductive
584 Behav Ecol Sociobiol (2010) 64:583?592
sharing or the mechanisms that control reproductive
contributions vary between sexes?
We hypothesized that the incidence of shared reproduc-
tion among dominant and subordinate African wild dogs
was higher than previously reported for this species. We
also expected that a combination of demographic, behav-
ioral, and hormonal factors determined the distribution of
breeding opportunities and that mechanisms mediating
reproductive sharing were different for males and females.
To test these assumptions, we evaluated extensive behav-
ioral and genetic data collected through radio-telemetry
monitoring, frequent behavioral observations, individual
fecal sampling, and occasional field immobilizations.
Specifically, we compared social rank and genetic parent-
age results, examined behavioral processes maintaining
dominance hierarchies, compared our extensive genetic
results to hormonal data for dominants and subordinates
collected earlier by others (Creel et al. 1997), and evaluated
demographic processes at work in the population.
Materials and methods
Study population
Demographic and behavioral information was collected on
African wild dogs in KZN from January 2001 through August
2008. This population began successfully breeding and
expanding in 2001 after the release of artificially assembled
packs into a single protected area. By August 2008, further
reintroductions, natural dispersals, and pack formations
boosted the population to 88 dogs in eight different groups
living in three protected areas (Spiering et al. 2009). During
the study period, the population included 257 individuals
that comprised ten packs and 36 total pack years, with
successful breeding occurring in 32 of these years.
Data on pack composition (number of dogs, sex, age
classes, and litter size at first emergence from the den),
location, and reproductive status (i.e., breeding, non-
breeding, pregnant, lactating) were collected once monthly
minimally and as often as ten times per month for more
accessible packs. Packs and dispersing groups were located
by radio-telemetry, observations made from a vehicle or on
foot, and individual wild dogs identified by unique coat
patterns and photographic records.
Determining dominance
The alpha male and female in a given pack were recognized
on the basis of: (1) reciprocal male and female scent-
marking behavior (Frame et al. 1979); (2) obvious
co-incidental male and female movement; and (3) mutual
offensive and defensive maneuvers in agonistic encounters
with other adult pack members (Girman et al. 1997). The
dominance hierarchy was also inferred from gestures of
subordination, including laying the ears flat against the
head and/or rotating the head away from a higher ranking
individual (van Lawick 1970) as well as passive submission
that included a subordinate rolling onto its back in the
presence of a more dominant dog (Schenkel 1967).
In this study, we considered three ways for an individual
to gain the dominant position within a pack. First, a wild
dog could become dominant by default as the only adult of
their sex in the pack. Alternatively, there could be inter-
animal competition without physical aggression with a
same-sex adult at the initial pack-bonding phase or when
the hierarchy was disrupted due to death of the alpha
individual. Lastly, the current dominant could be over-
thrown by a competitor through fighting. As dominant
females never disperse (Creel et al. 2004), we assumed that
any such individuals missing from a pack had died.
Subordinate siblings of the alpha pair were considered
?potential breeders?, whereas subordinate offspring of the
dominant pair were not because alpha individuals appar-
ently share breeding with siblings, but rarely with offspring
(Girman et al. 1997; McNutt and Silk 2008; Spiering et al.,
unpublished observations).
Genetic sampling and genotyping
Biomaterials for molecular genetic evaluations were col-
lected from January 2003 through January 2008. Wild dog
tissue and blood samples were obtained opportunistically
during immobilization operations for translocation and
collaring and when a wild dog carcass was located
(Spiering et al. 2009). Non-invasive collection of feces
allowed securing representative samples from a significant-
sized population (n=113 wild dog individuals and ten
packs). Fecal samples were collected fresh from known
individuals within 5 to 30 min of deposition and then kept
in a cool bag for up to 4 h before storing in labeled, plastic
freezer bags at ?20?C until genetic analysis.
Sample collection and detailed DNA extraction protocols
are described in Spiering et al. (2009). In brief, DNA was
extracted from scat using a QIAamp DNA Stool Mini Kit
and from tissue and blood using a QIAamp Tissue and
Blood Kit (QIAGEN). Genetic analyses were completed
using 19 microsatellites selected from the 2006 Interna-
tional Society for Animal Genetics domestic dog (Canis
familiaris) panel that were consistent with other wild dog
genetic studies in southern Africa (Moueix 2006). All
individuals were genotyped at 17 dinucleotide microsatel-
lite loci (AHT130, AHT137, AHTh171, AHTh260,
AHTk211, AHTk253, CXX279, FH2848, INRA21,
INU030, INU055, LEI004, REN54P11, REN105L03,
REN162C04, REN169D01, and REN247M23) and two
Behav Ecol Sociobiol (2010) 64:583?592 585
tetranucleotide loci (FH2054 and FH2328). These loci are
commonly used for determining parentage in domestic dogs
and, therefore, were selected because they are widely
distributed throughout the genome and highly polymorphic.
The polymerase chain reaction protocols are discussed in
Spiering et al. (2009). A combination of the multiple tubes
approach (Taberlet et al. 1996) and the maximum likelihood
method (Miller et al. 2002) were used to overcome the
potential for fecal DNA genotyping errors (Spiering et al.
2009). To detect and eliminate sampling error, we com-
pared matched tissue or blood with feces, analyzed
duplicate samples for individuals, and used a significant
number of microsatellite markers to verify unique individ-
uals in the dataset (Waits et al. 2001).
Parentage analysis
During the study period, 220 pups emerged from 30 litters,
and 86 of these offspring were sampled for parentage
analyses using the likelihood-based approach in the Cervus
3.0.3 software package (Marshall et al. 1998). The simulation
program in Cervus was used to establish the critical
difference in natural logarithm of the odds ratio (LOD score)
between the first and second most likely candidate parents
(at >95% confidence). Only adults from within the pack with
a given set of offspring were considered candidate parents
because no extra-group copulations have been reported for
this species (and analyses later confirmed that all parentage
was assigned to pack members). As most wild dog packs are
comprised of a group of brothers and an unrelated group of
sisters (Girman et al. 1997), we completed all simulations
with and without the advanced simulation option that
includes relatives among candidate parents. Assignments
using the advanced option with relatives did not differ from
assignments not using the function.
A lower than expected frequency of heterozygotes
indicating a high frequency of null alleles was detected at
locus INU030, which, consequently, was excluded from
further analysis. No other locus deviated from Hardy?
Weinberg equilibrium. For the 18 loci used, the overall
probability of exclusion was 0.991 for the first parent and
0.999 for the second. Critical LOD values were calculated
to assign: (1) maternity, with paternity unknown (in cases
where multiple females appeared to be pregnant); then (2)
paternity, with known maternity (in cases where multiple
adult males were present); and (3) the parent pair, with
sexes known (to verify assignments). Each breeding pack
was simulated and assigned parentage separately, which
allowed entering pack-specific data, including the proportions
of potential parents sampled and the relatedness between
candidate parents. All statistical analyses were performed
using JMP software version 3.2.2 (SAS Institute Incorporated)
with results presented as means ? standard error of the mean.
Results
Reproductive sharing
Maternity
The age of first-litter production for a female varied from
1.3 to 5.0 years (mean, 3.2?0.3 years). The number of
litters produced per breeding female ranged from one to
seven (mean, 2.1?0.4 litters). Alpha females produced 32
litters in 36 breeding years, resulting in an 88.9?5.3%
annual probability of breeding for dominants. In the
remaining 4 years, packs formed late in the year or a
dominant died, thereby causing reproductive failure. Alpha
females gave birth to 93.0% of the 86 pups that were
genetically sampled (22 litters), whereas the remaining
7.0% were produced by beta counterparts. However, 68 of
the 86 pups (79.1%) whelped by alpha females were born
to packs comprised of only one adult female at that time.
When multiple adult females were present in a pack,
physical signs of pregnancy (i.e., greatly increased weight,
enlarged teats) were never observed in more than two
females at one time. Although a few matings were observed
in 2 years involving theta females, none subsequently
appeared pregnant on the basis of physical appearance or
offspring production. Alpha individuals mated every year
compared to beta counterparts that bred in 6 of 11 breeding
years (54.5?15.7%; Table 1; Wilcoxon test, T41=5.04, P<
0.001). Female subordinates of all ranks gave birth to six
litters in 16 individual years, resulting in a 37.5?12.5%
probability of a subordinate breeding. In breeding years,
with multiple adult females present, six alpha females
whelped 12 sampled pups (66.7%), and three subordinates
produced six sampled pups (33.3%; Table 1). There was no
significant difference in average number of pups per litter for
years with single (7.4?0.6 pups, 26 litters) versus multiple
(8.8?1.4, 3 litters; Wilcoxon test, T27=0.80, P=0.43) births.
Although sample size was small, on the three occasions
when two litters were living within the same den (i.e.,
creched), the proportion of emerging pups whelped by the
alpha female (54.1?21.0%) was similar to that for sub-
ordinates (45.9?20.5%; t test, t4=0.37, P=0.80). There was
also no significant difference in percentage survival to 1 year
for pups born to alpha versus beta mothers (83.4?16.7% and
100.0?0.0%, respectively; t test, t2=?1.00, P=0.42).
Paternity
Males first bred from 1.1 to 5.0 years of age (mean, 2.9?
0.2 years), which was not significantly different from
females (3.2?0.3 years; Wilcoxon test, T31=0.78,
P=0.44). During the study interval, individual breeding
males produced from one to five litters each (mean, 1.7?
586 Behav Ecol Sociobiol (2010) 64:583?592
0.3 litters). Based on genetic analyses, alpha males were
confirmed to have sired 72.1% of pups compared with
22.1% for beta and 5.8% for theta counterparts (Table 2).
Alpha males sired offspring in 20 of 24 sampled litters,
resulting in an 83.3?7.7% chance of breeding annually.
However, for 32 pups in six litters (37.2% of sampled
pups), only one adult male was present and was confirmed
to have sired all pups.
Without exception, if present in a pack, at least one
subordinate male was observed mating with the alpha
female, a subordinate female, or both. In the presence of
multiple potentially breeding males, the dominant indi-
vidual sired 55.5% compared with the beta at 38.9% and
theta at 5.6% of all offspring (based on 54 pups in 17
litters; Table 2). When only two adult males were present,
the incidence of reproductive sharing was high, and the
percentage of pups sired did not differ significantly
between the alpha (55.2%) and beta (44.8%; t test,
t14=0.75, P=0.47) males, with similar annual probabilities
of siring offspring (alpha, 80.0?16.4% versus beta, 60.0?
15.7%; t28=1.18, P=0.25). In the presence of three or
more males in a pack, the dominant individual produced
24% more pups than the beta that, in turn, sired 20% more
offspring than the theta (Table 2). The overall annual
probability of breeding for the alpha males (86.4?7.5%)
was not significantly different than for the beta counter-
parts (73.3?11.8%; Wilcoxon test, T35=0.98, P=0.33),
but both were higher than for theta individuals (28.6?
18.4%; T27=3.45, P=0.002 and T20=2.09, P=0.048,
respectively).
Table 1 Number of years with multiple African wild dog females pregnant and number of sampled pups (18 pups over 11 pack years) parented
by dominant versus subordinate females
Pack Year No. of adult
females
Multiple
pregnant
females?
No. alpha
female offspring
No. beta female
offspring
Observations
Imfolozi 2001 2 No 1 0
Crocodile 2003 3 Yes NGD NGD Multiple males mated with all femaels.
No pups emerged from den.
2004 2 No 3 0 Males mated with both females, but no
pregnancy in the beta.
Mkhuze 2005 3 No 1 0 Offspring only from the alpha female.
2006 3 Yes 0 1 Alpha female died before whelping. Beta
female whelped ~2 weeks later.
Thanda 2006 2 Yes 3 1
2007 2 No NGD NGD Pups and alpha females killed at den.
2008 3 Yes NGD NGD Alpha and beta female litters killed by
lions at den.
Ume 2007 3 Yes 2 4 Beta female whelped ~2 weeks after
alpha females.
2008 2 Yes NGD NGD Large litter of 14 pups emerged. No
genetic samples.
Veggie 2007 2 No 2 0
Mean 2.5 54.5% 66.7% 33.3%
NGD indicates no genetic data available
Table 2 Numbers and mean percentages of sampled pups emerging from dens sired by alpha, beta, and theta African wild dog males
No.
pups
No.
litters
Alpha male pups
(%)
Beta male pups
(%)
Theta male pups
(%)
Subordinate males total
(%)
All sampled pups 86 24 72.1 22.1 5.8 27.9
Alpha male only packs 32 6 100.0 NP NP NP
Multiple adult male packs 54 17 55.5 38.9 5.6 44.5
Two adult males 29 9 55.2 44.8 NP 44.8
Adult males ?3 25 8 56.0 32.0 12.0 44.0
NP indicates individual of that rank not present
Behav Ecol Sociobiol (2010) 64:583?592 587
In 15 breeding years when subordinate males were
present, multiple sires fathered pups in eight single litters,
with the alpha male siring the majority of three mixed-sire
litters. Two males were present in eight breeding years
where the dominant individual sired 58.8?16.5% of pups,
which was not significantly different than the 41.3?16.5%
for beta individuals (t test, t14=0.75, P=0.47). In the seven
litters with three potential male breeders, the alpha male
sired 55.4?12.3% of young compared with 30.3?5.5% and
14.3?10.2% for the beta and theta individuals, respectively
(Table 3). For seven litters and 35 pups, there was no
statistically significant difference between the proportion of
pups surviving to 1 year that had been sired by dominant
(78.8?12.9%) versus subordinate males (93.8?5.0%; t test,
t13=?1.14, P=0.30).
Factors mediating reproductive opportunities
Population demographics
Although the average lifespan of African wild dogs in our
population was 2.0?0.2 years, dispersing individuals that
formed breeding packs lived 4.1?0.3 years, which was
similar to the mean of 4.7?0.5 years for animals gaining
dominance status (Wilcoxon test, T46=?0.91, P=0.37).
During the breeding season in May, pack size ranged from
two to 23 adults and yearlings (mean, 8.1?0.8 individuals/
pack). At emergence from the den, litter size varied from
two to 14 pups (mean, 7.6?0.6 pups/litter) with a sex ratio
near parity (0.51?0.04). The number of potentially breed-
ing adults (intra-pack adults that were siblings, half-
siblings, fathers or uncles, but not offspring of the dominant
pair) varied from two to seven (mean, 3.6?0.5 adults/pack).
The remainder of each pack (i.e., 54.7?6.5% of individu-
als) was comprised of offspring from the alpha pair,
generally considered to be non-breeders and confirmed as
such in this study. As observed in other wild dog
populations (McNutt and Silk 2008), adult and yearling
males outnumbered females slightly but not significantly
within packs (males, 54.9?3.9% versus females, 45.1?
2.6%; t test, t62=?1.13, P=0.21). However, this pattern was
significant when considering only potentially breeding
adults (males, 59.7?3.2% versus females, 40.3?5.8%;
t62=?3.19, P=0.003). Of the 32 successful breeding years
evaluated, only 11 (34.4%) were years when more than one
adult female potential breeder was in the pack (i.e.,
generally there was a lack of subordinate females available
for breeding). In contrast, multiple adult males were present
in 22 (68.9%) of these same breeding years.
Influence of demographics and behavior on rank
Only 25.3% of the wild dogs in our study population survived
to disperse from the natal group and form a breeding pack. An
Table 3 Percentages of litters sired by alpha, beta, and theta African wild dog males in 15 breeding years with subordinate male potential
breeders present. Numbers represent percentages of pups sired within single litters
Pack Year (no. of
litters)
No. subordinate
males
No. sampled
pups
Alpha male offspring
(%)
Beta male offspring
(%)
Theta male offspring
(%)
Imfolozi 2001 (1) 1 1 100.0 0 NP
2002 (1) 1 5 100.0 0 NP
2003 (1) 1 10 20.0 80.0 NP
2004 (1) 1 3 100.0 0 NP
Crocodile 2004 (1) 2 3 33.3 33.3 33.3
2005 (1) 2 3 0 33.3 66.7
2006 (1) 2 7 71.4 28.6 0
2007 (1) 1 3 100.0 0 NP
Juma 2005 (1) 1 2 0 100.0 NP
2006 (1) 1 1 0 100.0 NP
2007 (1) 4 3 66.7 33.3 0
Mkhuze 2005 (1) 3 1 100.0 0 0
Thanda 2006 (2) 1 4 50.0 50.0 NP
Ume 2007 (2) 2 6 66.7 33.3 0
Veggie 2007 (1) 2 2 50.0 50.0 0
Total pups 54 30 21 3
Mean % of litters with two adult males in pack 58.8 41.3 NP
Mean % of litters with ?3 adult males in pack 55.4 30.3 14.3
NP indicates individual of that rank not present
588 Behav Ecol Sociobiol (2010) 64:583?592
even smaller percentage (13.3%) became dominant in the
pack during their lifetime, which was similar to the proportion
of dogs that eventually reproduced (15.2%). Of all dogs
surviving dispersal and living in a breeding pack for at least
1 year, 76.5% eventually became dominant.
On average, dominance tenures lasted 2.4?0.3 years, with
alpha females tending to have non-significantly longer such
periods (2.7?0.5 years) than males (2.1?0.4 years; t test,
t38=0.17, P=0.86). Of the packs having multiple, potentially
breeding adult males (n=8 packs, 19 pack years), there was
major variation in the frequency of alpha male replacement.
One pack remained stable for more than 2 years, one
changed dominants randomly three times in 4 years, and
others experienced dominance switches annually or biannu-
ally. Within packs with multiple females (n=7 packs, 14
pack years), alpha females were sustained in multiple years
for three packs, but, in all other cases, dominants changed
annually or biannually, usually as a result of alpha female
mortality. In years with multiple, potentially breeding adults,
male dominance changes occurred in 65.6?5.3% of pack
years compared with 41.7?4.1% for females (Wilcoxon test,
T31=1.15, P=0.31). Although rates of hierarchical change
were similar between sexes, the causes leading to reestab-
lishment of social dominance roles were different. Changes
in dominance status were more likely to occur as a result of
aggression in males than in females (males, 45% versus
females, 5%; Fig. 1; t test, t38=?3.21, P=0.003). In contrast,
females most often earned dominance via non-aggressive
competition at pack formation, after the death of an alpha
individual, or by default as being the sole surviving adult
female in the pack (Fig. 1). A higher probability of achieving
social dominance was observed with increasing age, as all
males and females in our population surviving to 6 years of
age or older were dominant. In each case, these dogs were
the sole individual of their sex left in the pack and, therefore,
attained dominance (and became the only reproducers) by
simply outlasting competitors.
Discussion
Results from our integrated demographic, behavioral, and
genetic analysis of social dominance and parentage in a
multi-pack population of African wild dogs supported the
hypothesis that reproduction was shared among adults
substantially more than previously reported. While an
earlier behavior-genetics study reported that lower-ranking
individuals rarely bred (subordinate females and males
producing only 8% and 10% of offspring, respectively;
Girman et al. 1997), our detailed genetic testing revealed
significant reproductive sharing with, and pup production
by, subordinate adults, especially males. In the presence of
multiple adult potential breeders, we found that subordinate
females whelped 33.3% and subordinate males sired 44.5%
of pups. Subordinate females became pregnant in 37.5% of
individual years with beta individuals reproductively
successful in 54.5% of years. It is unclear if the litters
sampled by Girman et al. (1997) were born into packs with
several potential breeders or if packs consisted only of a
single alpha pair and their offspring. This would greatly
alter results of subordinate breeding, as sharing reproduc-
tion has been reported with siblings of the alpha pair but
never with adult offspring. When both single alpha pair
packs and multiple potential breeder packs were consid-
ered, our results revealed female percentages (7.0%) similar
to Girman et al. (1997), but subordinate male reproduction
was still substantially higher (27.9%).
Contrary to a previous investigation suggesting that
male and female subordinates reproduce at similarly low
levels (Girman et al. 1997), our study confirmed that the
proportion of subordinates able to breed and produce
offspring was higher for males than females. This observa-
tion corroborated earlier speculation by Creel et al. (1997)
who, in interpreting gonadal hormone profiles, predicted that
shared paternity should be more common than shared
maternity in the African wild dog. We determined that
female reproduction was only shared with the beta female
regardless of number of available female potential breeders.
In contrast, when there were three or more adult males in a
pack, each almost always sired some pups in a given litter.
Reproductive partitioning extended as far as the third ranking
male, a mechanism absent in females where the third
position apparently offers no direct reproductive benefit.
Our discovery of shared reproduction among pack
members, even including multiple sires in a given litter,
suggests that this strategy contributes to ensuring the fitness
of the pack and gene diversity in progeny (Gottelli et al.
2007). This is advantageous in an unpredictable environ-
ment or when the risk of disease-related mortality is high
(Sherman et al. 1998). Others have maintained that
increased genetic variation could simply be the result of
multiple matings rather than its selective force (Wolff and
Macdonald 2004). The wild dog may have also evolved a
multiple mating strategy to promote post-copulatory sperm
competition (Madsen et al. 1992) or to encourage all males
to contribute equally to caring for young (Nakamura 1998).
50
60
*
0
10
20
30
40
Fr
eq
ue
nc
y 
(in
 %
)
*
Non-aggressive
Competition
Aggression Default
Fig. 1 Frequency distribution of the mechanisms used to achieve
dominance for 20 female (open bars) and 20 male (solid bars) African
wild dogs. An asterisk indicates a significant difference (P=0.003)
Behav Ecol Sociobiol (2010) 64:583?592 589
Our results revealed that a combination of factors
appeared to mediate reproductive partitioning within the
African wild dog breeding pack. Within the ecological
conditions of KwaZulu-Natal, both demographic and
behavioral mechanisms created a constantly shifting dom-
inance hierarchy and social system. Although genetic
analyses confirmed that dominants indeed parented the
majority of offspring, we found that alpha male and female
dominance tenures were relatively short, with frequent
within-pack hierarchical changes that permitted multiple
adults to gain top-tier status and reproduce. Although only
~25% of African wild dogs born into this population
survived long enough to join a breeding pack, 75% of those
succeeding at this endeavor eventually became alpha. The
slightly longer (non-significant) dominance tenures for
females compared with males was similar to observations
of Creel et al. (1997), but there also was substantial inter-
pack variation in dominance stability for both sexes.
Regardless, we were most impressed with the overall short
lifespan of the average wild dog in breeding packs
(ca. 4 years) that, in turn, stimulated a constantly changing
society where frequent alpha deaths provoked dominance
changes, ranging from inheritance of dominance to dis-
persals to search for new mates. The incidence of mortal-
ities measured in our study was similar to that reported by
Woodroffe et al. (2007) for other wild dog populations in
Africa. Therefore, it is reasonable to suspect that a
continually changing hierarchical social system is a
common trait of this species regardless of regional location.
The basis for abbreviated dominance tenures was
markedly different between males and females, with the
former changing via aggression and the latter mostly by
nonviolent competition at pack-bonding. Whereas there
was little physical antagonism among females (regardless
of social status), dominant males often reaffirmed domi-
nance by aggression with subordinates frequently challeng-
ing alpha male status. Both genders displayed the capacity
to advance their status opportunistically upon the death of
an alpha individual. Exactly how a wild dog individual
secures the advantage to win dominance in these situations
is unknown and is fascinating given that most competitive
interactions often were between closely related and
physically-similar individuals (i.e., siblings). More study
is warranted, especially exploring the potential of mate
choice by either sex in influencing who attains the alpha
(and breeding) positions.
It was clear from our investigation that a female often
achieved dominance by default when she alone joined a
male dispersal group or when the death of a dominant
sibling resulted in her being the sole surviving adult female
in the pack. Our results supported the findings of Creel and
Creel (2002) and Somers et al. (2008) that the probability of
being socially dominant increases with female age, as all
females in our population surviving to 6 years or older
became dominant. In fact, there was an overall skewing of
sex ratio for mature wild dogs to favor males (60%) despite
there being near-gender parity in the pup age class.
Frequent adult mortalities led to a prevalence of packs
with only a single female, with multiple adult females
occurring during only about one third of all pack breeding
years. This was not surprising given the findings of others
(Frame et al. 1979; Reich 1981; Creel and Creel 2002) who
have noted progressively increasing male bias in older age
classes. In the KZN population, generally, males had more
contemporaries than females in a breeding pack and, thus,
more competitors that decreased the chance of securing
dominance after the death of an alpha male. This explained
why only 20% of males achieved dominance by default and
rather relied on constantly striving to win dominance by
aggressively and/or competitively overthrowing siblings. In
contrast, females experienced less competition within the
pack and perhaps developed a patient ?waiting? strategy,
seeking out active reproductive opportunity only after the
alpha female's death. Interestingly, subordinate males
reproduced more than subordinate females, indicating that
more competitors were not necessarily a significant
obstacle to males achieving reproductive success.
When breeding packs did contain multiple adult females,
reproductive success of subordinates was noticeably higher
than reported by Creel and Creel (2002), probably because
our population often was comprised of newly formed packs
with less stability in dominance hierarchies. In three of 11
breeding years, packs with multiple female potential
breeders had formed shortly before the breeding season,
and, on many occasions, copulations were observed
between several males and females before pack dominance
hierarchy was established. This is not unusual as others
have observed multiple females and males copulating in
earlier African wild dog studies (Malcolm and Marten
1982; Girman et al. 1997). Occasionally, this has resulted in
births to subordinate females with the litters then sometimes
killed by the alpha female (Reich 1981; Malcolm and
Marten 1982; Fuller et al. 1992). Although our sample size
for these genetic comparisons was modest, we observed no
infanticide on the three occasions where two litters were
produced in a pack simultaneously, and similar pup
numbers were whelped by the alpha and beta females.
Our study is the first to show a parallel between the
extent of reproductive sharing and behavioral and endo-
crine correlates of rank reported for the African wild dog in
the Selous (Creel et al. 1997). For instance, Creel et al.
(1997) found that female dominant African wild dogs
excreted higher estrogen and progestin concentrations in
feces and were more aggressive during mating periods than
all subordinates. Although it was predicted that subordinate
females in the Selous did not ovulate due to high baseline
590 Behav Ecol Sociobiol (2010) 64:583?592
estrogens and estrogen-to-progestin ratios (Creel et al.
1997), we suspect that most beta females in our population
completed a normal ovarian cycle, in part because at least
five of these individuals actually produced pups. However,
this ovarian activity likely was temporally delayed based on
betas always appearing visibly pregnant later and whelping
pups 2 to 4 weeks after the alpha. Since no females that
ranked third or lower in the dominance hierarchy became
pregnant in our population, it was possible that these
individuals were physiologically suppressed and failed to
ovulate. This seemed especially likely as little aggression
was observed between females to suggest behavioral
mechanisms of limiting lower ranking individuals' access
to males. Creel et al. (1997) also reported differences in
androgen excretion among males in the hierarchy, the
highest values being in the top-tier dog, which was believed
to present advantages in reproductive capability and
aggression. Our results, however, demonstrated clearly that
beta and theta males were physiologically capable of siring
young, but differences in gonadal steroid production may
have affected fertility in males ranking third and lower.
Nonetheless, most reproductive benefits probably are
conferred behaviorally by dominants being more likely to
block access to females in estrus through enhanced
antagonistic, female guarding activities.
In conclusion, our study has demonstrated that reproductive
success of the African wild dog is influenced by (1) a
continuously dynamic social system where marked hierarchi-
cal shifts ensure that a diversity of individuals eventually
become alpha reproducers and (2) a greater than previously
reported incidence of reproductive sharing, especially involv-
ing subordinate males. Both beta males and females played a
significant role in producing viable young that appeared as
robust as those from the dominant pair. Although the alpha
position was considered most attractive to all pack members,
this top-tier status was not necessary to contribute offspring.
Nevertheless, short life spans and an aspiration to be dominant
appeared to drive a constantly shifting social order, with males
mainly relying on behavioral aggression and females remain-
ing vigilant for nonviolent opportunities to escape possible
hormonal suppression. We suspect that this strategy assists in
sustaining genetic diversity and more effectively maintains
heterogeneity than in other cooperative breeders (i.e., naked
mole rat, Clarke and Faulkes 1997; dwarf mongoose, Rood
1990), where reproduction is almost exclusively monopolized
by alpha individuals. This is not the case in the African wild
dog where more individuals are contributing genes to
offspring, even when only a single litter is produced annually.
Acknowledgments We thank Ezemvelo KwaZulu-Natal Wildlife,
especially the management teams at Hluhluwe-Imfolozi Park and the
uMkhuze section of iSimangaliso Wetland Park. We also appreciate
the assistance of Thanda Private Game Reserve and their wildlife
management team. We are grateful to Rob Fleischer, Emily Latch,
Sarah Haas, Kalon Armstrong, and Nancy Rotzel of the Smithso-
nian's Center for Conservation and Evolutionary Genetics for
laboratory support and assistance with protocols and procedures.
Sarah Arnoff, Jan Graf, Gabriella Flacke, Mariana Venter, Carla
Naude-Graaff, Sboniso (Zama) Zwane, and Chris Kelly provided
invaluable assistance in monitoring and sample collection in the
field. This research was supported by funding from the Smithsonian
Institution Undersecretary for Science Endowment, University of
Pretoria, Rotterdam Zoo Thandiza Fund, Humboldt State University,
Conservation Endowment Fund of the Association of Zoos and
Aquariums, Disney Wildlife Conservation Fund, Knowsley Safari
Park, DST-NRF Centre of Excellence for Invasion Biology, Khaki
Fever Work Wear, Pittsburgh Zoo Conservation Fund, and the
Morris Animal Foundation. International travel was generously
provided by British Airways.
References
Alberts SC, Watts HE, Altmann J (2003) Queuing and queue-jumping:
long-term patterns of reproductive skew in male savannah
baboons, Papio cynocephalus. Anim Behav 65:821?840
Barrett J, Abbott DH, George LM (1993) Sensory cues and the
suppression of reproduction in subordinate female marmoset
monkeys, Callithrix jacchus. J Reprod Fertil 97:301?310
Birkhead TR, Moller AP (1992) Sperm competition in birds:
evolutionary causes and consequences. Academic, London
Clarke FM, Faulkes CG (1997) Dominance and queen succession in
captive colonies of the eusocial naked mole-rat, Heterocephalus
glaber. Proc R Soc Lond B 264:993?1000
Creel S, Creel NM (2002) The African wild dog: behavior, ecology,
and conservation. Princeton University Press, Princeton
Creel S, Waser P (1991) Failures of reproductive suppression in dwarf
mongooses (Helogale parvula): accident or adaptation? Behav
Ecol 2:7?15
Creel S, Creel NM, Mills MGL, Monfort SL (1997) Rank and
reproduction in cooperatively breeding African wild dogs:
behavioral and endocrine correlates. Behav Ecol 8:298?306
Creel S, Mills MGL, McNutt JW (2004) Demography and population
dynamics of African wild dogs in three critical populations. In:
Macdonald DW, Sillero-Zubiri C (eds) Biology and conservation
of wild canids. Oxford University Press, Oxford
Estes RD, Goddard J (1967) Prey selection and hunting behaviour of
the African wild dog. J Wildl Manage 31:52?69
Faulkes CG, Abbott DH (1991) Social control of reproduction in
breeding and non-breeding naked mole-rats (Heterocephalus
glaber). J Reprod Fertil 93:427?435
Faulkes CG, Bennett NC (2001) Family values: group dynamics and
social control of reproduction in African mole-rats. Trends Ecol
Evol 16:184?190
Frame L, Malcolm J, Frame G, Lawick HV (1979) Social organization
of African wild dogs (Lycaon pictus) on the Serengeti plains,
Tanzania, 1967?1978. Z Tierpsychol 50:225?249
French JA (1997) Proximate regulation of singular breeding in
Callitrichid primates. In: SolomonNG, French JA (eds) Cooperative
breeding in mammals. Cambridge University Press, Cambridge
Fuller TK, Kat PW, Bulger JB, Maddock AH, Ginsberg JR, Burrows
R, McNutt JW, Mills MGL (1992) Population dynamics of
African wild dogs. In: Barret DR, McCullogh RH (eds) Wildlife
2001: populations. Elsevier Science, London
Girman DJ, Mills MGL, Geffen E, Wayne RK (1997) A molecular
genetic analysis of social structure, dispersal, and interpack
relationships of the African wild dog (Lycaon pictus). Behav
Ecol Sociobiol 40:187?198
Behav Ecol Sociobiol (2010) 64:583?592 591
Gottelli D, Wang J, Bashir S, Durant SM (2007) Genetic analysis
reveals promiscuity among female cheetahs. Proc R Soc Lond B
274:1993?2001
Huck M, Lottker P, Heymann EW (2004) Proximate mechanisms of
reproductive monopolization in male moustached tamarins
(Saguinus mystax). Am J Primatol 64:39?56
Kappeler PM, Port M (2008) Mutual tolerance or reproductive
competition? Patterns of reproductive skew among male redfronted
lemurs (Eulemur rufus). Behav Ecol Sociobiol 62:1477?1488
Madsen T, Shine R, Loman J, Hakansson T (1992) Why do female
adders copulate so frequently? Nature 355:440?441
Malcolm JR, Marten K (1982) Natural selection and the communal
rearing of pups in African wild dogs (Lycaon pictus). Behav Ecol
Sociobiol 10:1?13
Marshall TC, Slate J, Kruuk LEB, Pemberton JM (1998) Statistical
confidence for likelihood-based paternity inference in natural
populations. Mol Ecol 7:639?655
McNutt JW, Silk JB (2008) Pup production, sex ratios, and
survivorship in African wild dogs, Lycaon pictus. Behav Ecol
Sociobiol 62:1061?1067
Miller CR, Joyce P, Waits LP (2002) Assessing allelic dropout and
genotype reliability using maximum likelihood. Genetics
160:357?366
Moehlman PD (1979) Jackal helpers and pup survival. Nature
277:382?383
Moueix C (2006) Genetic verification of multiple paternity in two
free-ranging isolated populations of African wild dogs (Lycaon
pictus). MSc thesis, University of Pretoria, Pretoria
Nakamura M (1998) Multiple mating and cooperative breeding in
polygynandrous alpine accentors. Anim Behav 55:259?275
Ortega J, Guerrero JA, Maldonado JE (2008) Aggression and
tolerance by dominant males of Artibeus jamaicensis: strategies
to maximize fitness in harem groups. J Mammal 89:1372?1378
Packard JM, Seal US, Mech DL, Plotka ED (1985) Causes of
reproductive failure in two family groups of wolves (Canis
lupus). Z Tierpsychol 69:24?40
Reich A (1981) The behavior and ecology of the African wild dog
(Lycaon pictus) in the Kruger National Park. PhD thesis, Yale
University, New Haven
Reyer H, Dittami JP, Hall MR (1986) Avian helpers at the nest: are
they psychologically castrated? Ethology 71:216?228
Richardson DS, Jury FL, Blaakmeer K, Komdeur J, Burke T (2001)
Parentage assignment and extra-group paternity in a cooperative
breeder: the Seychelles warbler (Acrocephalus sechellensis). Mol
Ecol 10:2263?2273
Rood JP (1990) Mating relationships and breeding suppression in the
dwarf mongoose. Anim Behav 28:143?150
Russell JK (1983) Altruism in coati bands: nepotism or reciprocity?
In: Wasser SK (ed) Social behavior of female vertebrates.
Academic Press, New York, pp 263?290
Savage A, Ziegler TE, Snowdon CT (1988) Sociosexual development,
parent bond formation and mechanisms of fertility suppression in
the female cotton-top tamarins (Sanguinus oedipus oedipus). Am
J Primatol 14:345?359
Schenkel R (1967) Submission: its features and function of the wolf
and dog. Amer Zool 7:319?329
Sherman PW, Seeley TD, Reeve HK (1998) Parasites, pathogens, and
polyandry in honey bees. Am Nat 151:392?396
Somers MJ, Graf JA, Szykman M, Slotow R, Gusset M (2008)
Dynamics of a small re-introduced population of wild dogs over
25 years: Allee effects and the implications of sociality for
endangered species recovery. Oecologia 158:239?247
Spiering PA, Gunther MS, Wildt DE, Somers MJ, Maldonado JE
(2009) Sampling error in non-invasive genetic analyses of an
endangered social carnivore. Conserv Genet. doi:10.1007/
s10592-009-9880-6
Taberlet P, Griffin S, Goossens B, Questiau S, Manceau V, Escaravage
N, Waits LP, Bouvet J (1996) Reliable genotyping of samples
with very low DNA quantities using PCR. Nucleic Acids Res
2:3189?3194
van Lawick H (1970) Wild dogs. In: van Lawick H, van Lawick-
Goodall J (eds) The innocent killers. Houghton Mifflin Co.,
Boston, pp 47?101
Waits LP, Luikart G, Taberlet P (2001) Estimating the probability of
identity among genotypes in natural populations: cautions and
guidelines. Mol Ecol 10:249?256
Wingfield JC, Silverin B (1986) Effects of corticosterone on territorial
behavior of free-living Song Sparrows, Mdospha mtiodia. Horm
Behav 20:405?417
Wolff JO, Macdonald DW (2004) Promiscuous females protect their
offspring. Trends Ecol Evol 19:127?134
Woodroffe R, Davies-Mostert H, Ginsberg J, Graf J, Leigh K,
McCreery K, Robbins R, Mills G, Pole A, Rasmussen G, Somers
M, Szykman M (2007) Rates and causes of mortality in
endangered African wild dogs Lycaon pictus: lessons for
management and monitoring. Oryx 41:215?223
592 Behav Ecol Sociobiol (2010) 64:583?592