Behav Ecol Sociobiol (2005) 58: 414?422
DOI 10.1007/s00265-005-0928-4
ORIGINAL ARTICLE
Bradley C. Fedy ? Bridget J. M. Stutchbury
Territory defence in tropical birds: are females as aggressive
as males?
Received: 30 January 2004 / Revised: 28 January 2005 / Accepted: 8 February 2005 / Published online: 23 April 2005
C? Springer-Verlag 2005
Abstract Much of our knowledge concerning the func-
tions of territorial behaviour and how territories are de-
fended by individuals comes from research on birds. The
vast majority of this work has focused on temperate zone
breeding territoriality in which territories are defended
most obviously by males. Our understanding of the female
role in territory defence is limited because they are less
conspicuous and much harder to observe. We studied sex
roles in territory maintenance and defence in a duetting,
resident neotropical passerine, the white-bellied antbird
(Myrmeciza longipes). This species maintains territories
and pair bonds year round and both sexes sing and actively
participate in territory defence. We performed a series of
playback experiments throughout the dry (non-breeding)
and wet (breeding) seasons. We exposed territorial pairs to
three types of stimuli including: (1) single sex, male only
songs, (2) single sex, female only songs, and (3) both sex
songs/duets. Contrary to findings for most other tropical
species, individuals defended their territories with equal
levels of aggression regardless of stimuli. Furthermore, sex
roles were very different, with males responding more ag-
gressively than females to all stimuli throughout both sea-
sons. Both males and females consistently responded more
aggressively to territorial intrusions during the dry sea-
son than during the wet season, likely because food abun-
dance is low in the dry season and territory value is high.
Our analysis of duetting behaviour suggests that duets do
not serve a significant role in mate guarding, or territory
defence.
Communicated by A. Cockburn
B. C. Fedy (


)
Department of Biology, York University,
4700 Keele St.,
Toronto, Ontario, M3J 1P3, Canada
e-mail: bfedy@interchange.ubc.ca
Present address:
B. J. M. Stutchbury
Centre for Applied Conservation Research, University of
British Columbia,
3rd Floor Forest Sciences Centre,
Vancouver, British Columbia, V6T 1Z4, Canada
Key words Sex roles . Territory defence . Tropics .
White-bellied antbirds
Introduction
Territory defence in birds has been extensively studied,
but primarily in temperate zone species where breeding
territories are defended for a short period in spring and
summer, and male defence via song and border disputes is
conspicuous (Catchpole and Slater 1995). Female territory
defence against other females is relatively under-studied
in birds because it is more subtle and therefore difficult
to observe (Amundsen 2000). Tropical passerines offer a
unique opportunity to ask new questions about territorial
behaviour because in many species both sexes sing and
defend the territory and territory defence takes place all
year round (Stutchbury and Morton 2001). Are males and
females equally aggressive in territory defence? Do they
direct their aggression primarily at same-sexed intruders,
or does the pair co-operate and repel all intruders regardless
of sex?
The evolution of sex roles is primarily driven by the op-
erational sex ratio (OSR) (Emlen and Oring 1977), which
in turn affects the intensity of sexual selection within each
sex. Factors that influence OSR include: (1) the relative
abundance of each sex (Emlen and Oring 1977), (2) the po-
tential for mating (Clutton-Brock and Vincent 1991), and
(3) the stringency of mate choice (Owens and Thompson
1994). These factors appear to be equal for male and fe-
male tropical passerines that have a monogamous mating
system and yearlong territoriality (Stutchbury and Morton
2001). Furthermore, both males and females require territo-
ries and are capable of defending them alone (Levin 1996b;
Morton and Derrickson 1996; Fedy and Stutchbury, unpub-
lished data). Greenberg and Gradwohl (1983) concluded
that both sexes of dot-winged antwrens (Microrhopia quix-
ensis) participate equally in territory defence behaviour.
However, this is not the case for all species. In other species
of antbirds, males are more aggressive than females (Willis
1973; Morton and Derrickson 1996; Bard et al. 2002).
415
In several tropical bird species, individuals of a pair de-
fend their territories primarily against intruders of the same
sex (Freed 1987; Levin 1996b; Morton and Derrickson
1996; Appleby et al. 1999; Morton et al. 2000). In
playback experiments, both Levin (1996b) and Morton and
Derrickson (1996) found that females responded more
aggressively to same sex songs than to duets or male
songs. Same-sex territory defence by females may deter
trespassing, and eviction, by other females (e.g. McElroy
and Ritchison 1996). This type of same-sex aggression
in females has also evolved in several other taxa with
monogamous mating patterns, including several species of
monogamous antelope (the dikdik Madoqua kirkii, Broth-
erton and Rhodes 1996; and the klipspringer Oreotragus
oreotragus,, Roberts and Dunbar 2000). Alternatively,
year round territory defence by males and females may
lead to co-operative territory defence within a pair. Hall
(2000) documented apparent co-operative behaviour in a
series of playback experiments investigating the function
of duetting in magpie-larks (Grallina cyanoleuca). In that
study, females maintained both territory and pair bond by
attempting to exclude intruders of either sex in apparent
co-operation with their mates.
We studied sex roles in territory maintenance and defence
in the white-bellied antbird (Myrmeciza longipes), a resi-
dent neotropical passerine which maintains territories and
pair bonds year round. Both males and females sing and
exhibit aggressive territory defence and pairs routinely duet
throughout the year. Like other species of antbirds (Formi-
cariidae), both sexes contribute equally to nest building,
incubation and care of the young. We performed a series
of playback experiments throughout the dry and wet sea-
sons. We exposed territorial pairs to three types of stimuli
including: (1) single sex, male only songs, (2) single sex,
female only songs, and (3) both sex songs/duets. Because
the threat of eviction from a territory applies only to a same
sex competitor we predicted that individuals would defend
their territories more aggressively from same sex intruders
than opposite sex (Levin 1996b; Morton and Derrickson
1996; Appleby et al. 1999). Territorial intrusions are ex-
pected to pose a greater threat to resident birds in the wet
season than in the dry season because the majority of breed-
ing occurs during the wet season, mates are likely fertile
and thus intruders present an immediate risk of cuckoldry.
Furthermore, if an intruder evicts a resident during this
period, the evicted bird likely will have difficulty finding
a new territory in time to breed that season. Thus, given
the importance of territory ownership for breeding, we also
predicted that individuals would respond more strongly to
playbacks during the wet season than the dry season.
Duets generally have the two key characteristics of
a constant lag time between the contributions of the
two individuals and a repeated and predictable structure
(Langmore 2002). Several non-mutually exclusive hy-
potheses exist for the function of avian duetting (Hall 2004).
One key hypotheses is that duetting may represent a form
of co-operative territorial defence between pair members
(Wickler 1976, Wickler and Seibt 1980; Hall 2000; Grafe
and Bitz 2004; but see Farabaugh 1982). If this hypoth-
Fig. 1 Tracing of a white-bellied antbird (Myrmeciza longipes) duet.
Black male song, white female song. Females join the male song
which results in song overlap
esis is correct, then duets should be perceived as more
threatening territorial signals than solo songs. If duetting
is a more threatening signal, then the proportion of duets
sung by the defenders should increase during playbacks
compared with natural background rates. Furthermore, duet
playbacks should elicit a more aggressive response from the
territory owners than single sex playbacks. White-bellied
antbirds are unusual because their duetting appears to be al-
most entirely under female control. A duet is formed when
the female begins singing during a male song (Fig. 1). In
most other species, either sex can join the song of their
mate to create a duet (Farabaugh 1982), and evidence
seems to suggest that it may be more common among
species for the male to join the female to create the duet
(see Langmore 1998). Therefore, we can also predict that
duets, as a form of female defence, will be more common
in response to female song playbacks than to male song
playbacks.
Methods
Study site and study species
White-bellied antbirds inhabit second growth and edge
habitat in lowland tropical areas from northern Brazil to
central Panama (de Schauensee 1966; Wetmore 1972).
They are a medium-sized (X bar + ?SE = 29 g ? 0.3,
range = 25?32.5 g), sexually dimorphic, passerine of
the Formicariidae family and males are slightly heav-
ier than females (males: X bar + ?SE = 28.7 ? 0.3 g,
n=42; females: X bar + ?SE = 27.5 ? 0.3 g, n=29; t-test
P=0.005). They are insectivorous birds that associate in
pairs throughout the year, forage almost exclusively in the
ground leaf litter, and do not follow ant swarms.
Research was conducted from 15 January to 30 July 2000
in Parque Nacional Soberania, Panama (9?7?N, 79?40?W)
at the junction of the Panama Canal and the Chagres River.
This area is a tropical lowland forest receiving approxi-
mately 2.6 m of rain per year (Karr 1971), almost all of
which falls during the wet season. Central Panama has a
pronounced dry season from mid-December to late April
(Windsor 1990). In general, the non-breeding season for
ground foraging, insectivorous birds in this area corre-
sponds with the dry season and lasts approximately from
January to April. Breeding can begin in late April, when
the wet season begins and litter arthropod populations in-
crease (Levings and Windsor 1982). However, breeding is
416
highly asynchronous and evidence of breeding can still be
found into December for species with similar ecology to
white-bellied antbirds (Robinson et al. 2000).
Birds were caught in mist nets by broadcasting a con-
specific song on their territory and every individual was
banded with a unique colour combination (n=48 males
and 34 females). Some birds were flushed into the nets
when necessary. Territory boundaries were determined by
mapping bird locations during song bouts, border disputes
and by assessing responses to playback of recorded songs
(Falls 1981).
Natural song rates
Data on natural song rates were collected between 0620
and 1230 hours throughout the non-breeding and breeding
seasons. Sessions lasted 1 h with a minimum of 24 h be-
tween sampling sessions on any one territory (total 140 h,
X bar + ?SE = 8.8 h ? 0.52 per territory, n=16 territo-
ries). Researchers sat quietly on the edge of the territory
and recorded all vocalisations of the focal pair. Distance
between pairs was recorded every time a pair duetted. Total
song rate refers to the number of songs/h for a focal pair and
includes solo and duet songs. Natural song rate data were
not normally distributed (Shapiro-Wilk P=0.01), therefore
non-parametric tests were used in all analyses of these
data.
Playback experiments
Playback experiments were conducted in the dry season
(10 March to 12 April) and the wet season (21 May to
18 June) between 0600 and 1230 hours. Each pair of ter-
ritorial birds (n=15 pairs) was subject to three playback
experiments in both seasons. Each experiment involved
broadcasting a song stimulus to the focal birds. The three
different stimuli were: (1) male song only, (2) female song
only, and (3) duets. The order of stimuli presentation was
randomised for each pair within each season. Experiments
were spaced between 20 and 48 h apart and neighbouring
territories were not tested on the same day.
Playback tapes were made from high-quality recordings
of territorial individuals located a minimum of 3 km away
from the focal pair?s territory. This distance is well out-
side the dispersal range of white-bellied antbirds (Fedy
and Stutchbury 2004) and is sufficient to ensure that all
individuals were exposed to unfamiliar songs. All record-
ings were made within 5 m of the subject using a Sony
TC-D5 PRO II recorder with a Sennheiser MKH 816 di-
rectional microphone. Playback tapes mimicked the typical
singing pattern and amplitude of aggressive white-bellied
antbirds as determined by observing natural boundary dis-
putes and response to simulated intrusions. Throughout
the study we observed 32 boundary disputes (13 territo-
ries). During these disputes average male song rate was
2.0 songs/min (SD 0.5), average female song rate was
1.1 songs/min (SD 0.4) and the average number of songs
sung as a duet was 0.7/min (SD 0.3). Each playback tape
was made by looping a 1-min segment of a natural recording
for 10 min (male song only: 2 songs/min; female song only:
1 song/min; duets: 2 male only songs and one duet/min).
Four tapes of different individuals were made for each of
the three stimuli (male, female, duets n=12 tapes) to avoid
the potentially confounding affects of psuedoreplication
(Kroodsma 1989). One of the four tapes was chosen ran-
domly for each trial within an experiment.
The playback speaker was concealed within a pair?s ter-
ritory a minimum of 15 m from the nearest boundary and
observers retired to cover a minimum of 10 m away. Birds
showed no response to our presence during the playback
experiments. To determine the focal birds? distance to the
playback speaker, 5-m intervals around the speaker were
marked off with flagging tape. Between each trial the play-
back speaker was moved 5?10 m within the territory to
avoid habituation to any particular location. Focal birds
were silent for a minimum of 10 min before each exper-
iment began. This species has low natural song rates and
therefore the situation of 10 min of no singing was not un-
common. Songs were broadcast for the first 10 min of the
experiments and were followed by a 10-min post-playback
silent period. We recorded the behaviour of the resident
male and female for the entire duration (20 min). Six vari-
ables were collected for all birds during playback experi-
ments: total time responding?time from the first song or
movement towards the speaker until the bird moved >15 m
away and stopped singing (TTR; s), the latency to first vo-
calization of any kind (LFV; s), number of songs (NS),
closest approach to speaker (CA; m), time spent between
20 and 5 m to the speaker (T20; s), and time spent <5 m to
the speaker (T5; s). A strong response by territory owners
would show high TTR, NS, T20, T5 and low values for
LFV and CA. To quantify the intensity of response to the
playback experiments a principal components analysis was
performed on the variables by the use of the correlation ma-
trix (James and McCullough 1990; McGregor 1992). Prin-
cipal components analysis does not accommodate missing
response values, so it was necessary to insert values for
LFV and CA for birds that did not respond. In these cases
the individual was assigned an LFV of 1200 s (20 min, the
entire duration of the experiment) and a CA of 30 m (the
largest recorded in responding birds). All response vari-
ables measuring the intensity of an individual?s response
were entered without transformation, except for CA which
was log transformed to achieve a closer approximation of a
normal distribution. Means are presented ?SE throughout
and all data were analysed using SPSS v. 10.
Results
Natural song rates
White-bellied antbirds have low natural song rate and al-
though average male song rate was only 12?17 songs/h,
in two watches we recorded up to and over 200 songs/h,
whereas for 43% of watches no songs were recorded from
417
Fig. 2 Natural song rates
throughout the day. Each circle
represents a sampling session.
Closed circles dry sessions,
open circles wet season
sessions. Unlike many other
tropical passerines,
white-bellied antbirds do not
sing a dawn chorus in either the
dry or wet season
either the male or female of a pair. The average number of
songs/hour for males and females combined did not vary by
time of day (0630?1230 hours) throughout the sampling pe-
riod (Fig. 2, ANOVA: F1,146=2.4, P=0.12), and there was
no dawn chorus. Average total song rate/h in the dry sea-
son (12.0?2.9) was not significantly different than in the
wet season (17.6?4.1 Wilcoxon signed ranks test: n=16,
z=?0.83, P=0.41). When males and females were con-
sidered separately the song rate/h of either sex (including
solo songs and duets) did not differ between the dry and
wet seasons (males: dry 8.3?1.9, wet 13.5?3.1, n=16,
z=?1.32, P=0.19 females: dry 3.7?1.1, wet 4.2?1.2,
z=?0.06, n=16, P=0.96). Average male song rate per hour
(10.9?1.6) was significantly higher than female song rate
per hour (3.9?0.8, z=?3.16, n=16, P<0.001) across both
seasons (Fig. 3). The number of duets given per hour did not
differ between the dry (2.7?1.0) and wet seasons (2.6?0.9,
z=?0.17, n=16, P=0.87).
Playback experiments
Song rate for both males and females increased signifi-
cantly from natural levels during playback experiments.
During playback trials individuals averaged 18.6?1.9
songs/20 min. Natural song rates of the same birds were sig-
Fig. 3 Box plots showing natural song rates between seasons
(n=16). Grey boxes dry season, white boxes wet season
nificantly lower 2.6?0.4 songs/20 min (z=?4.78, n=30,
P<0.001). In a few cases, territorial individuals did not re-
spond to the playback experiments. In the dry season, males
(n=15) responded to 100% of the experiments (45/45 tri-
als) and females (n=15) responded to 91% of experiments
(41/45). Both sexes decreased the number of experiments
they responded to in the wet season. Males responded to
88% of the experiments (40/45) and females to 71% of
experiments (32/45). Over both seasons combined males
responded to 94% of experiments (85/90) and females to
81% (73/90).
The principal components analysis (PCA) showed that
PC1 varied positively with TTR, NS, T20 and T5. The
range of loadings varied little from T5 (0.78) to T20 (0.95).
PC1 varied negatively with the other two variables (LFV,
?0.87; CA, ?0.74). PC1 accounted for 74% of the total
variation in the original data set. Data were then analysed
using a repeated measures ANOVA (RMA).
There was a significant difference in the intensity of
response to playbacks between the dry and wet seasons
(MANOVA: F1,15=6.4, P=0.02) and in the intensity of re-
sponses from the different sexes (MANOVA: F1,15=19.7,
P<0.001). Thus, on average, both sexes respond more ag-
gressively to playbacks in the dry season than in the wet
season and males consistently respond more aggressively
than females throughout both seasons. The different stimuli
(male song, female song, duet) did not elicit significantly
different responses (MANOVA: F2,15=0.13, P=0.88). Fur-
thermore, analysis of interaction terms revealed no sig-
nificant interactions between season, stimuli or sex (sea-
son*sex, MANOVA: F2,15=1.4, P=0.27; stimuli ? sex,
MANOVA: F2,15=1.3, P=0.31, season ? stimuli ? sex,
MANOVA: F2,15=0.8, P=0.44). An RMA was also con-
ducted to see if the presentation order of the tapes had any
effect on the intensity of the birds? responses and no signifi-
cant difference was found within the tape order or the inter-
action terms (tape order, MANOVA: F2,15=0.0, P=0.99;
season ? tape order, MANOVA: F2,15=1.2, P=0.32).
Figure 4 shows the actual values for three of the key vari-
ables (TTR, NS, LFV) which contributed to the calculation
of PC1 and reflects the findings of the principal components
analysis.
418
0
5
10
15
20
25
30
35
40
0
100
200
300
400
500
600
700
800
900
1000
duets male
only
female
only
duets male
only
female
only
duets male
only
female
only
duets male
only
female
only
duets male
only
female
only
duets male
only
female
only
To
ta
l t
im
e 
re
sp
on
di
ng
(se
co
nd
s ?
 
se
, 
n
=
15
) 
0
100
200
300
400
500
600
700
800
900
1000
0
5
10
15
20
25
30
35
40
0
100
200
300
400
500
600
700
800
900
0
100
200
300
400
500
600
700
800
900
N
um
be
r o
f s
on
gs
 (?
 se
, n
=1
5)
La
te
nc
y 
to
 fi
rs
t v
oc
al
iza
tio
n
(se
co
nd
s ?
 
se
, 
n
=
15
)
NON-BREEDING BREEDING
To
ta
l t
im
e 
re
sp
on
di
ng
(se
co
nd
s ?
 
se
, 
n
=
15
) 
N
um
be
r o
f s
on
gs
 (?
 se
, n
=1
5)
La
te
nc
y 
to
 fi
rs
t v
oc
al
iza
tio
n
(se
co
nd
s ?
 
se
, 
n
=
15
)
Fig. 4 Male and female responses to the three different playback stimuli. Black bars male response, white bars female response. Variables
reported are: total time responding (s), number of songs, and latency to first vocalisation (s)
Since duetting results from females singing after the male
initiates his song, there are two ways to gauge the frequency
with which females duet and hence use duetting in terri-
torial aggression. First, what proportion of female songs
are given as duets with the male, and second, what propor-
tion of male songs are joined by female song to create a
duet? We averaged the proportion of female songs in duets
across both seasons in response to male song only and fe-
male song only playbacks. Duets occurred in 54%?8%
of female songs in response to female song playback and
56%?8% in response to male song playbacks. This was
not significantly different (z=?0.47, n=15, P=0.64). We
conducted a paired analysis of natural duetting behaviour
and duets sung during playbacks to determine if a sim-
ulated intrusion increases the frequency of duetting by
females. An increase in duetting during intrusions may
indicate that duets are a more aggressive form of territorial
signal than solo songs. We averaged the proportion of songs
sung in duets throughout all playback experiments (n=6
experiments per individual) and compared this against the
natural duet rate for the same individuals. During the play-
back experiments females joined 24% of male songs to
419
create a duet, compared with 27% for natural song obser-
vations (z=?0.97, n=14, P=0.33). Similarly, duets com-
prised 60% of all female songs given during playbacks,
and 56% of all female songs during natural observations
(z=?1.16, n=14, P=0.25). Summarising, both predictions
regarding the possible role of female duetting in territory
defence were rejected.
Discussion
The territorial behaviour of white-bellied antbirds is not
typical compared with other similar species (Levin 1996b;
Morton and Derrickson 1996), demonstrating that our un-
derstanding of the prevalent tropical territorial system of
year round defence is still quite limited. Sex roles in terri-
tory defence were very different, and birds responded with
equal aggression to same sex and opposite sex playbacks.
Song rates are typically low and there is no conspicuous
dawn chorus. This is likely due to low intruder pressure,
compared with temperate zone species, and stable terri-
tory boundaries with known neighbours (Greenberg and
Gradwohl 1986).
Seasons
White-bellied antbirds responded more strongly to play-
backs during the dry season. They feed on leaf-litter arthro-
pods which increase dramatically in abundance during the
wet season (Levings and Windsor 1982). The stronger re-
sponses during the dry season may be the result of a higher
territory value, due to a greater competition for rare food
resources. This suggests that territory defence may be re-
lated to food abundance. Territory defence may be more
important when food is limited during the dry season to
ensure access to food and avoid competition with intrud-
ers over food resources. These findings agree with other
theoretical and empirical studies on the role of food re-
sources in territorial behaviour in that an increase in food
resources led to a decrease in territorial aggression (Brown
1964; Gill and Wolf 1975; Dubois and Giraldeau 2004). Of
course, maintaining territories in the dry season would also
secure a breeding position in the wet season, when nesting
actually occurs.
Constraints during the breeding (wet) season may ac-
count for some of the disparity observed in aggression
levels between the two seasons. Attendance at the nest,
or young, could constrain participation in territory defence
during the breeding season due to time or energy restric-
tions. If birds were attending to eggs or young it may be
too costly in terms of time lost foraging and feeding young
to engage in active territory defence. Territory defence may
also expose young to predators as territory defence is very
conspicuous and young seem to stay close to their parents.
We were unable to determine the exact stage of breeding
during playbacks in the breeding season, so are unable to
directly compare the defence behaviour at different nest
stages.
Both sexes also responded to a greater proportion of the
playback experiments during the dry season. It is unlikely
that focal birds did not hear the playbacks as white-bellied
antbird songs are loud and can often be heard across two
territories (B. Fedy, personal observation). Territory bound-
aries were well established and stable and we knew the
extent of every marked bird?s territory. Furthermore, radio-
telemetry has shown it is very rare for a white-bellied
antbird to engage in forays that are further than two ter-
ritories away (Fedy and Stutchbury 2004). Therefore, we
can conclude that birds chose to ignore the intrusion when
they did not respond to the playback.
Natural song rates were low (males: 11/h, females: 4/h)
and similar between both seasons. Although birds are more
aggressive during the dry season in response to intrusions,
they do not show a higher natural song rate in the absence
of intruders. It is unlikely that low song rates were the
result of limited food abundance because natural song rates
did not increase with higher food abundance in the wet
season.
Sex roles
Males are more aggressive in territory defence than fe-
males. An inequality in sex roles has been documented in
this type of territorial system before, but not to the same
degree (Morton and Derrickson 1996). Male white-bellied
antbirds consistently responded to more of the playback ex-
periments and were more aggressive than females regard-
less of stimuli or season. This sex difference in territorial
aggression may be connected to the benefits of retaining
versus switching mates and territories. Levin (1996a) and
Morton et al. (2000) conducted removal experiments on
their duetting study species. In both species, it was found
that territorial vacancies (both experimentally created and
natural) were filled very rapidly (less than 1 day). This rapid
replacement rate was observed in another duetting species,
the buff-breasted wren (S. Gill and B. J. M. Stutchbury
unpubl. data). Furthermore, Levin (1996a) found that bay
wrens (Thryothorus nigricapillus) would quickly increase
their song rate after the loss of a mate and this was inter-
preted as a means of attracting a new mate. In an even more
obvious effort at mate attraction, widowed dusky antbirds
(Cercomacra tyrannina) begin singing a unique song
within minutes of their mate?s disappearance (Morton and
Derrickson 1996). Such long distance mate attraction ef-
forts do not exist in white-bellied antbirds (Fedy and
Stutchbury 2004). Temporary removals of white-bellied
antbirds showed no evidence of long-distance song ad-
vertisement in widowed males or females (Fedy and
Stutchbury 2004). Removed birds were not replaced
rapidly, with replacements occurring after 3 days in sev-
eral cases. Therefore, white-bellied antbirds do not face
the same intense/short term threat of replacement which
exists in other species. Furthermore, this suggests that in
white-bellied antbirds retaining a familiar mate is ben-
eficial as they do not advertise or switch readily. For
white-bellied antbirds, sex roles in territory defence may
420
therefore be tightly linked to the traditional notion of pair
bonds. When both sexes readily switch mates and break
long-held bonds, territory defence is focussed primarily on
same sex individuals (Levin 1996b; Morton and Derrickson
1996). When mate switching is less common, as in white-
bellied antbirds, territory defence is equal against all intrud-
ers regardless of sex. This implies the pair is co-operating in
territory defence, in the sense that each individual benefits
by preventing intruders from displacing either themselves
or their mate.
White-bellied antbirds are ground foraging birds which
likely leads to high predation risk for adults. Therefore,
mates may be important for predator vigilance as well as
for breeding partners. In a species where life is dangerous,
familiar mates may become an important resource for sur-
vival. Morton and Stutchbury (2000) found that adult dusky
antbirds, which forage in dense vine tangles, had a survival
rate of 82%. In our study of white-bellied antbird only
55% of individuals banded in one year survived to the next
(Fedy and Stutchbury, unpublished data). Furthermore, us-
ing radio telemetry we have found that pair members spend
over 60% of their time within 5 m of each other (Fedy and
Stutchbury, unpublished data). This reliance on mates for
vigilance could partially explain why white-bellied antbirds
do not readily switch or advertise for new mates. Each in-
dividual within a pair receives benefits from their mates
in the form of the non-depreciable ?resource? of predator
vigilance which would moderate the conflict between the
sexes and allow them to act in a more co-ordinated man-
ner in territory defence. Therefore, predation risk could be
playing a significant role in the expression of sex roles in
white-bellied antbirds.
The higher levels of aggression shown by males may be
the result of a greater threat to their territory position than
to females. If males are more likely to be ousted from a
territory, males should respond more strongly to male in-
truders than to female intruders; however, males responded
equally to all stimuli. Males would face a greater competi-
tion for their territorial position if there were a male-biased
sex ratio. However, our data from removal experiments
suggests this is not the case. We confirmed the presence
of similar numbers of male and female floaters, and male
and female vacancies were equally likely to get filled in
response to natural and artificially created vacancies (Fedy
and Stutchbury 2004).
Breeding activities may also influence sex roles and ag-
gression. Similar male and female mating effort could pro-
duce similar levels of aggression by the sexes in both
seasons. We observed that males were more aggressive
than females during the breeding season. Although males
and females likely contribute equally to nest building and
care of the young, females still need to invest energy
(Walsberg 1983) and/or dietary protein (Ramsay and Hous-
ton 1998) into egg production. This may partially influence
the unequal sex roles observed during the breeding period.
However, this does not explain the differences in aggression
levels between the sexes observed during the non-breeding
season and is not likely to be a significant factor influencing
the disparity in aggression between the sexes.
The ultimate mechanism influencing the difference in
aggression levels between males and females is not imme-
diately obvious. However, sexual selection in white-bellied
antbirds is likely stronger on males than females, which
may explain the sex role differences in territory defence.
Males are bigger than females and the species is sexually di-
morphic. Furthermore, males sing more than females and
are consistently more aggressive to territorial intrusions.
Despite the similar threat of replacement between the sexes,
competition among males may be high due to control of
high quality territories and/or mates. If males are primar-
ily responsible for territory defence then female choice of
mates may in turn depend on the male?s aggression levels
and ability to maintain a high quality territory throughout
the entire year.
Duets
Duets represent a more aggressive form of territory defence
than solo songs in some species (Hall 2000). Thus, we pre-
dicted that males and females would respond more strongly
to duet playbacks than to single sex playbacks. This was
not the case as individuals responded with equal levels of
aggression to single sex and duet playbacks. We also pre-
dicted an increase in the proportion of duets in response
to the playbacks than observed in natural vocalizations.
However, the intense threat posed by the playback experi-
ments did not result in a greater proportion of duets. Female
white-bellied antbirds create duets by beginning to sing af-
ter the male song begins, but before it ends. Consequently,
if birds were more aggressive to intruders of the same sex
(Levin 1996b; Morton and Derrickson 1996), more duets
should ensue in response to female only playbacks than
to male only playbacks. Because white-bellied antbirds do
not differentiate their responses between same and oppo-
site sex playbacks, it is not surprising, in retrospect, that
they did not differ in the proportion of duets sung in re-
sponse to female only and male only playbacks. Duets may
also function as a form of mate guarding (Hall 2004). If
this were the case in white-bellied antbirds we would have
observed an increase in duets in response to female only
playbacks. No such increase was observed and thus we can
conclude that duetting in white-bellied antbirds likely does
not function as a form of mate guarding.
Farabaugh (1982) explored the idea that individuals may
join in duets to avoid the possibility of mis-directed ag-
gression from their mates. This could be the case in white-
bellied antbirds as males respond with equal levels of ag-
gression to male and female songs. It would therefore be
beneficial for the female to let her mate know where she
is to avoid mis-directed aggression from her mate. This
would also allow the male to better focus his efforts on
the intruding birds. The use of interactive playback ex-
periments (e.g. Mennill et al. 2002) could be an excellent
tool for investigating this possible function of duetting.
Temporary removal of a female would allow investigators
to playback her song to her mate with two treatments:
(1) a trial in which the playback does not duet with the
421
resident male, and (2) an alternate experiment in which the
playback results in duetting with the male. The avoidance
of mis-directed aggression hypothesis would predict lower
aggression in the male when the female playback overlaps
with his songs to form duets.
Acknowledgements Earlier versions of the manuscript were im-
proved by valuable comments from Eugene Morton, Laurene Rat-
cliffe and two anonymous reviewers. We thank the Smithsonian Trop-
ical Research Institute for logistical support in Panama. Rachel Houli-
han provided field assistance and Michael Friendly assisted with
statistical analysis. We also thank Stan and Pat Rand for providing
transportation throughout much of the project. B.C.F. was supported
by a Natural Sciences and Engineering Research Council (NSERC)
Scholarship and York University Graduate Studies Scholarships. Re-
search was supported by York University Graduate Studies and Sigma
Xi grants to B.C.F., NSERC grant to B.J.M.S., and a Smithsonian
Scholarly Studies Grant to E.S. Morton. Research was conducted
under permits issued by the Autoridad Nacional del Ambiente
References
Amundsen T (2000) Why are female birds ornamented? Trends Ecol
Evol 15:149?155
Appleby BM, Yamaguchi N, Johnson PJ, MacDonald DW (1999)
Sex-specific territorial responses in tawny owls Strix aluco. Ibis
141:91?99
Balthazart J (1983) Hormonal correlates of behavior. In: Farner DS,
King JR, Parkes KC (eds) Avian Biology, vol 7. Academic
Press, New York, pp 221?365
Bard SC, Hau M, Wikelski M, Wingfield JC (2002) Vocal distinc-
tiveness and response to conspecific playback in the spotted
antbird, a neotropical suboscine. Condor 104:387?394
Brotherton PNM, Rhodes A (1996) Monogamy without biparental
care in a dwarf antelope. Proc R Soc Lond B 263:23?29
Brown JL (1964) The evolution of diversity in avian territorial
systems. Wilson Bull 76:160?169
Catchpole CK, Slater PJB (1995) Bird song. Cambridge University
Press, UK
Clutton-Brock TH, Vincent ACJ (1991) Sexual selection and the
potential reproductive rates of males and females. Nature
351:58?60
de Schauensee RM (1966) The species of birds of South America and
their distribution. Academy of Natural Sciences of Philadelphia,
Philadelphia
Dubois F, Giraldeau L-A (2004) Reduced resource defence in an
uncertain world: an experimental test using captive nutmeg
manikins. Anim Behav 68:21?25
Emlen ST, Oring LW (1977) Ecology, sexual selection, and the
evolution of mating systems. Science 197:215?223
Falls JB (1981) Mapping territories with playback: an accurate
census method for songbirds. Stud Avian Biol 6:86?91
Farabaugh SM (1982) The ecological and social significance of
duetting. In: Kroodsma DE, Miller EH (eds) Acoustic commu-
nication in birds, vol 2. Academic Press, New York, pp 85?124
Fedy BC, Stutchbury BJM (2004) Territory switching and floating in
white-bellied antbird (Myrmeciza longipes) a resident tropical
passerine in Panama. Auk 121:486?496
Freed LA (1987) The long-term pair bond of tropical house wrens:
advantage or constraint? Am Nat 130:507?525
Gill FB (1990) Ornithology. Freeman, New York.
Gill FB, Wolf LL (1975) Economics of feeding territoriality in the
golden-winged sunbird. Ecology 56:333?345
Grafe TU, Bitz JH (2004) Functions of duetting in the tropical
boubou, Laniarius aethiopicus: territorial defence and mutual
mate guarding. Anim Behav 68:193?201
Greenberg R, Gradwohl J (1983) Sexual roles in the dot-winged
antwren (Microrphopias quixensis), a tropical forest passerine.
Auk 100:919?925
Hall ML (2000) The function of duetting in magpie-larks: conflict,
cooperation or commitment? Anim Behav 60:667?677
Hall ML (2004) A review of hypotheses for the functions of avian
duetting. Behav Ecol Sociobiol 55:415?430
Hau M, Wikelski M, Soma KK, Wingfield JC (2000) Testosterone
and year-round territorial aggression in a tropical bird. Gen
Comp Endocr 117:20?33
James FC, McCullough CE (1990) Multivariate analysis in ecology
and systematics: panacea or Pandora?s box? Annu Rev Ecol
Syst 21:126?166
Karr JR (1971) Structure of avian communities in selected Panama
and Illinois habitats. Ecol Monogr 41:207?233
Ketterson ED, Nolan V Jr, Wolf L, Ziegenfus C (1992) Testos-
terone and avian life histories: effects of experimentally
elevated testosterone on behavior and correlates of fitness
in the dark-eyed junco (Junco hyemalis). Am Nat 140:980?
999
Kroodsma DE (1989) Suggested experimental designs for song
playbacks. Anim Behav 37:600?609
Langmore NE (1998) Functions of duet and solo songs of female
birds. Trends Ecol Evol 3:136?140
Langmore NE (2002) Vocal duetting: definitions, discoveries and
directions. Trends Ecol Evol 17:451?452
Levin RN (1996a) Song behavior and reproductive strategies
in a duetting wren, Thryothorus nigricapillus: I. Removal
experiments. Anim Behav 52:1093?1106
Levin RN (1996b) Song behavior and reproductive strategies
in a duetting wren, Thryothorus nigricapillus: II. Playback
experiments. Anim Behav 52:1107?1117
Levings SC, Windsor DM (1982) Seasonal and annual variation in
litter arthropod populations. In: Leigh EG, Rand AS, Windsor
DM (eds) The ecology of a tropical forest: seasonal rhythms and
long-term changes. Smithsonian Institution Press, Washington
DC, pp 355?387
McElroy DB, Ritchison G (1996) Effect of mate removal on singing
behavior and movement patterns of female northern cardinals.
Wilson Bull 108:550?555
McGregor PK (1992) Quantifying responses to playback: one,
many, or composite multivariate measures? In: McGregor PK
(ed) Playback and studies of animal communication. Plenum,
London, pp. 79?96
Mennill DJ, Ratcliffe LM, Boag PT (2002) Female eavesdropping
on male song contests in songbirds. Science 296:873
Morton ES, Derrickson KC (1996) Song ranging by the dusky
antbird, Cercomacra tyrannina: ranging without song learning.
Behav Ecol Sociobiol 39:195?201
Morton ES, Stutchbury BJM (2000) Demography and reproductive
success in the dusky antbird, a sedentary tropical passerine. J
Field Ornith 71:493?500
Morton ES, Derrickson KC, Stutchbury BJM (2000) Territory
switching behavior in a sedentary tropical passerine, the
dusky antbird Cercomacra tyrannina. Behav Ecol 11:648?
653
Owens IPF, Thompson DBA (1994) Sex differences, sex ratios and
sex roles. Proc R Soc Lond B 258:93?99
Ramsay SL, Houston DC (1998) The effect of dietary amino acid
composition on egg production in blue tits. Proc R Soc Lond B
265:1401?1405
Roberts SC, Dunbar RIM (2000) Female territoriality and the func-
tion of scent-marking in a monogamous antelope (Oreotragus
oreotragus). Behav Ecol Sociobiol 47:417?423
Robinson TR, Robinson WD, Edwards EC (2000) Breeding ecology
and nest-site selection of song wrens in central Panama. Auk
117:345?354
Stutchbury BJM, Morton ES (2001) Behavioral ecology of tropical
birds. Academic Press, San Diego
Walsberg GE (1983) Avian ecological energetics. In: Rarner DS,
King JR, Parkes KC (eds) Avian biology vol. 7. Academic
Press, London, pp 161?220
Wetmore A (1972) The birds of the Republic of Panama part 3 ?
Passeriformes: Dendrocolaptidae (Woodcreepers) to Oxyrunci-
dae (Sharpbills). Smithsonian Institution Press, Washington DC
422
Wickler W (1976) Duetting songs in birds: biological significance
of stationary and non-stationary processes. J Theor Biol
61:493?497
Wickler W, Seibt U (1980) Vocal duetting and the pair bond II.
Unisono duetting in the African forest weaver, Symplectes
bicolour. Z Tierpsychol 59, 217?226.
Wikelski M, Hau M, Robinson WD, Wingfield JC (1999) Seasonal
endocrinology of tropical passerines ? a comparative approach.
In: Adams NJ, Slotow RH (ed) Proceedings of the 22nd
International Ornithology Congress, Durban, Johannesberg:
Birdlife South Africa, pp 1224?1241
Willis EO (1973) The behavior of oscellated antbirds. Smithson
Contrib Zool 144:1?57
Windsor DM (1990) Climate and moisture variability in a tropical
forest: Long-term records from Barro Colorado Island, Panama.
Smithsonian Contributions to Earth Sciences No. 29
Wingfield JC (1994) Regulation of territorial behavior in the
sedentary song sparrow, Melospiza melodia morphna. Horm
Behav 28:1?15
Wingfield JC, Hegner RE, Dufty AM Jr, Ball GF (1990) The
?challenge hypothesis?: theoretical implications for patterns of
testosterone secretion, mating systems, and breeding strategies.
Am Nat 136:829?846