ANIMAL BEHAVIOUR, 2003, 65, 285–295
doi:10.1006/anbe.2003.2049, available online at http://www.sciencedirect.com
Reed warblers guard against cuckoos and cuckoldry
N. B. DAVIES*, S. H. M. BUTCHART*, T. A. BURKE†, N. CHALINE† & I. R. K. STEWART†
*Department of Zoology, University of Cambridge
†Department of Animal and Plant Sciences, University of Sheffield
(Received 25 January 2002; initial acceptance 5 April 2002;
final acceptance 8 July 2002; MS. number: 7216)
Previous studies have shown that reed warblers, Acrocephalus scirpaceus, are more likely to reject a cuckoo,
Cuculus canorus, egg if they have seen a cuckoo at their nest. This suggests that they would benefit from
watching out for cuckoos. We tested whether presentations of a cuckoo mount near the nest (to simulate
nest inspection) led to increased nest attendance by the warblers. Cuckoo presentations at completed
nests before laying, when males guarded their females closely, led to desertion at 40% of nests before any
eggs were laid (there were no desertions after presentations of a jay, Garrulus glandarius, a nest predator).
In the remaining cases, there was no effect of the cuckoo on nest attendance before laying began, but a
marked increase in male nest attendance (compared with jay and no-presentation controls) on the days
the first and second eggs were laid. Cuckoo presentations at the one-egg stage led to the same increase in
male nest attendance as did the prelaying presentations. Increased male nest attendance at the
one–two-egg stage was not at the expense of mate guarding, because this declined anyway when laying
began, and it did not lead to increased paternity loss compared with controls. Overall, 15% of broods had
one or two extrapair young (6% of all young extrapair). We conclude that male reed warblers do increase
nest guarding in response to cuckoos, but only after their females have begun egg laying, when there are
less likely to be costs in lost paternity. Females did not increase nest guarding, perhaps because they need
to spend more time foraging during the egg-laying period. Our results suggest that cuckoos should be
secretive not only when they lay but also when they monitor host nests beforehand.

2003 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.
Reed warblers, Acrocephalus scirpaceus, are among the
favourite hosts of the common cuckoo, Cuculus canorus,
in Europe. Considering large geographical areas, the
average parasitism rate is low; for example over the whole
of Britain it is just 5% (Brooke & Davies 1987). However,
parasitism of local populations may vary from 0 to 60%
(Schulze-Hagen 1992) and there may be substantial variation between years because local cuckoo populations are
often small and therefore susceptible to chance fluctuations and extinction (Lindholm 1999). Consequently,
adult reed warblers that return to breed at a particular site
may encounter variable parasitism rates, even during
their brief lives, and their offspring (which often disperse
to breed at other sites) may experience very different
parasitism rates from those of their parents (Lindholm
1999).
This variability creates a problem for the hosts.
Parasitism by common cuckoos is costly because the
newly hatched cuckoo chick ejects all the host eggs, and
any host young, from the nest. However, the main line of
Correspondence: N. B. Davies, Department of Zoology, Downing
Street, Cambridge CB2 3EJ, U.K. (email: [email protected]).
T. A. Burke, N. Chaline and I. R. K. Stewart are at the Department of
Animal and Plant Sciences, The University, Sheffield S10 2TN, U.K.
0003–3472/02/$30.00/0

host defence, namely egg rejection, is costly too because
hosts may damage their own eggs while attempting to
eject a cuckoo egg, and they may also make recognition
errors and eject one of their own eggs instead of the
cuckoo egg, which is mimetic (Davies & Brooke 1988;
Marchetti 1992; Welbergen et al. 2001). Thus egg rejection pays only above a certain level of parasitism (Davies
& Brooke 1989; Lotem et al. 1995; Davies et al. 1996).
Reed warblers have adapted to their variable world, and
to this trade-off between the costs of parasitism and the
costs of defences, by increasing egg rejection at times and
in places where parasitism rates are higher (Brooke et al.
1998; Lindholm & Thomas 2000; see also Alvarez 1996).
Individuals probably vary their rejection in relation to
their assessment of local cuckoo abundance. For example,
reed warblers are more likely to reject eggs if they have
seen a cuckoo at their nest, and so are more certain that
they have been parasitized. It has been shown experimentally that the sight of a cuckoo mount on their nest
stimulates reed warblers to increase their rejection of
model cuckoo eggs (Davies & Brooke 1988; see Moksnes
& Røskaft 1989 and Moksnes et al. 1993 for the
same effect with meadow pipits, Anthus pratensis).
Videorecordings of natural parasitism reveal that reed
285
2003 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.
286
ANIMAL BEHAVIOUR, 65, 2
warblers are also more likely to reject a real cuckoo egg if
they have seen the cuckoo lay than if they were absent
during laying (Moksnes et al. 2000). This is probably why
the cuckoo has evolved rapid laying, to decrease the
chance that it alerts the hosts (Davies & Brooke 1988).
These observations suggest that hosts would benefit by
spending time near their nests to watch out for cuckoos,
and thereby determine whether they are likely to be
parasitized. Hosts may have several opportunities to see a
cuckoo at their nest, because the female cuckoo may also
visit a nest prior to laying to check on its exact location,
and to monitor its progress to ensure she times her
parasitism to coincide with the host’s laying period
(Chance 1940; Moksnes et al. 2000). Both male and
female reed warblers reject cuckoo eggs (Davies & Brooke
1988), so both could gain valuable information by nest
guarding. However, nest guarding is a costly defence too,
because it reduces feeding time (Komdeur & Kats 1999).
This cost will be particularly acute for reed warblers,
because their nesting territories in the reeds are small and
they often leave to feed in bushes up to 150 m away
(Catchpole 1972; Davies & Green 1976). The period that
cuckoos inspect nests and lay coincides with the time that
the female reed warbler needs to feed intensively to form
her eggs, and with the time that the male is likely to have
to guard her to protect his paternity (Birkhead & Møller
1992).
Our aim in this study was to test whether the presence
of cuckoos near their nest, in the period prior to laying
and during laying, stimulated reed warblers to guard their
nests more intensively and, if so, whether this was costly
for males in terms of reduced mate guarding, and hence
paternity loss.
METHODS
Study Area and Study Species
Most of the study took place from May to July in
2000 and 2001, on Wicken Fen, Cambridgeshire, U.K.,
the site of our previous work. Reed warblers are socially
monogamous, with occasional polygyny (Leisler &
Catchpole 1992). They arrive from their African winter
quarters from late April to the end of May. Males defend
territories in the reeds, Phragmites australis, and sing for
much of the day until they attract a female (Catchpole
1973). Nests are built by the female alone and suspended
from the reed stems, usually over water. Most clutches are
of four eggs (range two to five). Males and females share
incubation about equally, although only the female
develops a vascularized brood patch, and both feed the
nestlings and fledglings (Duckworth 1992). Some pairs
rear two broods in a season but most have time for just
one.
We studied reed warblers nesting along two waterways,
an 820-m stretch of Wicken Lode (a channel constructed
in Roman times) and the 400-m stretch of Cross Dyke.
There were reed beds fringing both sides of the waterways, 2–4 m in depth, and each male defended a linear
territory along one bank. We put markers in the reeds
every 20 m so we could map territories and nests. Adults
were caught in mist nets and colour ringed for individual
identification. In 2000, the study stretches contained 35
pairs, and one male with two females, and in 2001, when
some of the reeds had been cut, there were 20 pairs. Of
these 20 pairs in 2001, 10 males and eight females had
been present in 2000, but most paired up with new mates
in 2001 (only three pairs remained together over the
2 years). No bird experienced the same experimental
treatment more than once over the 2 years.
Cuckoos arrived, also from their African winter quarters, in late April to May. Up to three males and two
females were heard or seen daily along the stretches
where we studied the reed warblers, until around
5–10 July, when the cuckoos left, their departure coinciding with the completion of the last reed warbler clutches
of the summer. Parasitism of reed warblers on Wicken Fen
has declined during the last 15 years, because of a decline
in cuckoos, from 26 and 16% of nests in 1985 and 1986,
respectively, to 2–6% in 1995–1997 (Brooke et al. 1998).
Among our colour-ringed reed warbler population, only
one of 54 nests was parasitized in 2000, and none of 35
nests in 2001, although including other nests monitored
on Wicken Fen in 2001 the parasitism rate was 6%
(N=135 nests).
The study was done under Home Office and English
Nature licences.
Behavioural Observations
Females were observed from the start of nest building
through to clutch completion, to measure the intensity of
male mate guarding, which was scored as the percentage
of time that the male was within 3 m of his mate (which
usually meant she was in his view). We excluded periods
when either bird was sitting on the nest.
To measure nest guarding, we observed nests for
30-min periods, from before laying through to clutch
completion, and scored the proportion of time that the
male, female, or both were either sitting on the nest or
were within 3 m. At some randomly chosen experimental
nests we presented a cuckoo mount, placed level with the
nest rim and 10 cm away, to simulate a cuckoo nest visit.
Previous experiments showed that the reed warblers
treated a mount as if it were a real cuckoo; they mobbed
it with loud ‘skurr’ calls, often snapped their bills and
occasionally struck it (Duckworth 1991): like the sight of
a real cuckoo at the nest, it stimulated increased egg
rejection (Davies & Brooke 1988). We left the cuckoo in
place for 5 min from the time that the first nest owner
returned to within 3 m of the nest. The cuckoo was
presented once at each experimental nest, but at various
stages of laying, to examine its effects on nest attendance.
Prelaying presentations (no-egg stage) were done at completed nests, 1–3 days before the first egg was laid. We
also did presentations during laying (the days the first,
second, third and fourth eggs were laid). On the day of
presentation, nests were watched both before and 1–2 h
after the cuckoo was removed, and we then watched on
subsequent days to measure any long-term effects.
As a control for this experiment, we did the same
presentation using a jay mount, Garrulus glandarius, at
DAVIES ET AL.: GUARDING VERSUS CUCKOOS AND CUCKOLDRY
Table 1. Characterization of the six microsatellite loci used in the paternity analysis of the reed warbler (using
samples from 2001, N=92 individuals)
Locus
Character
Fluorescent label
Annealing temperature, Ta (°C)
MgCl2 concentration (mM)
Size range (bp)
Number of alleles
Observed heterozygosity, Ho
Expected heterozygosity, He
Excl. 1
Excl. 2
Ase 18
Ase 25
Ase 37
Ase 48
Ase 58
Ppi 2
6-FAM
58
1.5
173–183
6
0.835
0.768
0.367
0.546
HEX
60
2.0
196–341
32
0.946
0.962
0.840
0.913
6-FAM
58
1.5
235–251
6
0.685
0.691
0.268
0.437
6-FAM
58
2.0
267–364
18
0.913
0.888
0.622
0.767
HEX
60
1.5
195–249
12
0.804
0.874
0.587
0.741
HEX
55
2.0
246–291
20
0.946
0.917
0.700
0.823
Excl. 1: Exclusion probability of the locus for the first parent. Excl. 2: Exclusion probability of the locus for the
second parent (with the first parent assigned). Ho, He and exclusion probabilities calculated with Cervus 2.0
(Marshall et al. 1998).
other randomly chosen nests. Jays are common nest
predators of reed warblers and other small birds on
Wicken Fen. The reed warblers also mobbed the stuffed
jay, but from further away, and they gave fewer calls,
never bill-snapped and never attacked it (see also
Duckworth 1991). Their greater caution makes good
sense because jays can also kill adult birds. Only one
cuckoo and one jay mount were used. However, the
distinctive responses of the reed warblers to each did not
differ significantly from those recorded in previous experiments using four other cuckoo mounts and two other jay
mounts (unpublished data).
In 2000–2001, the cuckoo presentations were made
both before and during laying, while the jay was presented only at the one-egg stage. To test whether the
increased nest desertion caused by prelaying cuckoo
presentations was a specific response to the cuckoo, we
repeated the prelaying presentations in May to early June
2002, this time presenting cuckoo and jay alternately at
successive nests. These experiments were done along the
Reach Lode, a waterway adjoining Wicken Fen and
3.5 km from our 2000–2001 study sites, to minimize the
chance that we retested the same birds.
Cuckoos usually lay in the afternoon (Chance 1940),
but nonlaying nest visits may occur in the morning too
(Moksnes et al. 2000). We standardized presentations of
the jay and cuckoo by doing them all in the afternoons
between noon and 1700 hours B.S.T.
Microsatellite Analysis
Blood (ca. 20 l) was collected from the brachial vein of
adults and 6–7-day-old nestlings and stored in ethanol.
DNA was extracted from 1 l of whole blood using a
resin-based technique (Walsh et al. 1991) for the samples
collected in 2000 and a salt-based technique (Bruford
et al. 1998) for the samples collected in 2001. Samples
from adults that were present in both years (N=18) were
extracted and genotyped again in 2001 as a control.
For the paternity analysis of the 2000 samples (36
broods, 129 nestlings) we used four fluorescently labelled
primers isolated from a congeneric species, the Seychelles
warbler, Acrocephalus sechellensis (Ase25, Ase37, Ase48 and
Ase58, Richardson et al. 2000). Two nonlabelled loci,
Ase18 (Richardson et al. 2000) and the magpie, Pica pica,
locus Ppi2 (Martinez et al. 1999), were used to investigate
uncertain parental mismatches and confirm likely genetic
sires. For the 2001 samples (16 broods, 57 nestlings),
all six primers were fluorescently labelled and used to
genotype all birds. Details of the primers tested to identify
the six that were suitable for this study can be found on
the Sheffield Molecular Genetics Facility Passerine
primer cross-utility database, accessed via http://
www.shef.ac.uk/misc/groups/molecol/birdmarkers.html.
DNA was amplified by the polymerase chain reaction
(PCR, Ellegren 1992). PCRs were performed in a 10-l
volume containing 10–50 ng DNA, 1.0 M of each
primer, 0.2 mM of each dNTP, 1.5–2.0 mM MgCl2
(Table 1) and 0.05 units of Taq DNA polymerase
(Thermoprime + , ABGENE, Epsom, Surrey, U.K.), in the
manufacturer’s buffer at a final concentration of 20 mM
(NH4)2SO4, 75 mM Tris-HCl pH 9.0 and 0.01% (w/v)
Tween. The reaction profile for each locus was 94C for
120 s, followed by 35 cycles of 94C for 30 s, Ta (see
Table 1) for 30 s, and 72C for 30 s. The forward primer
of each marker was 5 end-labelled with one of two
fluorescent phosphoramidites (Table 1), so that PCR
products of similar sizes from HEX-labelled and FAMlabelled loci could be combined and run in a single
multiplex load. The PCR products were electrophoresed,
together with an internal size standard (T500, Genesize),
through an Applied Biosystems (ABI) 377 DNA sequencer.
Gels were analysed using ABI Genescan software (version
3.1) and Genotyper DNA fragment analysis software
(version 2.5).
Exclusion probabilities were calculated with Cervus 2.0
(Marshall et al. 1998; Table 1). The combined probability
of exclusion using the four fluorescent markers was 0.994
in 2000. For the 2001 birds the combined probability
with six markers was 0.997. The frequency of null alleles
was low for all loci in both years.
Paternity analysis was performed with Cervus 2.0
(Marshall et al. 1998). First, we surveyed all of the
females in the study population to assign a genetic
287
ANIMAL BEHAVIOUR, 65, 2
mother to each nestling. In each case, the social mother
was identified as the genetic mother, except for a
maternal mismatch with a single nestling sampled in
2001 at a single locus, which was considered a mutation.
With the mother known, we then surveyed the whole
male population to assign the genetic father and
compared this against the observed social father.
In the samples collected in 2000, six nestlings (4.6%)
were mismatched with their social father at at least two of
the four fluorescent loci and were thus assumed to have
resulted from extrapair fertilizations (EPFs). A further
three chicks were mismatched at just one locus (two
mismatches at Ase25 and one at Ase58). Since single-locus
mismatches may be merely due to mutation, these three
nestlings were compared against their social parents at
two further loci polymorphic in reed warblers, Ase18 and
Ppi2. These primers were not fluorescently labelled, so the
PCR products were electrophoresed through 6% polyacrylamide gels and visualized using silver staining
(Bassam et al. 1991). The three contentious nestlings did
not match their social father at either of these loci, and
since a sample of nestlings that matched both parents at
the four fluorescent loci also matched both parents at the
two additional loci, the three mismatched chicks were
considered to have also resulted from EPFs.
To identify the genetic fathers of the nine extrapair
nestlings, we examined the profiles of all of the males in
the study population to find any that shared the nestlings’ nonmaternal allele at each of the four fluorescent
loci. Using this method, we identified the extrapair sires
for six of the nine nestlings. For further confirmation,
these sires were compared against the extrapair nestlings,
together with the social mother, at the two nonlabelled
loci. For all six nestlings, the profiles of the suggested
extrapair sires matched the nestlings’ nonmaternal allele
at both loci, and hence we concluded that they were the
genetic fathers.
In the samples collected in 2001, three nestlings (5.3%)
mismatched the social male at at least three of the six
loci. No single-locus mismatches were detected. None of
the other males present in 2001 could be assigned
as fathers of the extrapair nestlings. The fathers were
therefore assumed to come from outside the study area.
All statistical tests are two tailed.
RESULTS
Mate Guarding
Males defended, on average, a 22-m stretch of reeds
along one bank of the waterway (range 11–35 m, N=19).
They paired up after 1–20 days (mean 7 days, N=17) and
the female began to build a nest within 2 days of settling
on the territory, usually close to the site where the male
had been singing most intensively. Nests were completed
in 4–7 days, and the first egg was laid 5–9 days after the
nest was begun.
Males spent most of their time close to their females
both during nest building and up to the laying of the first
egg (Fig. 1). They chased away other males that
approached, and followed the female as she foraged in
the reeds and whenever she left the territory to feed in
100
Time mate guarding (%)
288
75
2
6
7
10
13
10
5
1
50
11
25
0
–7
–6
–5
–4
–3
–2
–1
1
2
egg eggs
Nest stage
Figure 1. Mean±SE percentage of time that male reed warblers
guarded their mates (<3 m) during the prelaying stage (scored as
days before first egg) and early-laying stage (one and two eggs).
Data are from a total of 27 pairs, with number of pairs watched
(>10 min) on each day indicated above error bars.
bushes nearby. Sometimes a male failed to see his female
fly off, and he then sang intensively in the reeds or
searched through the territory until he found her. Intruders were secretive and the guarding male usually
spotted them before we did, and chased them off before
we could identify them. Hence we could only measure
the rate of chases, not intrusion rate itself. For 15 pairs
with a total of at least 1 h of observation, the mean
rateSE was 1.60.5 chases/h. There was no significant
variation with laying stage (Wilcoxon signed-ranks tests,
comparing days 7–3 with 2–1 for 11 pairs, and
days 2–1 with the day of the first egg for seven pairs:
T + =31 and 8, respectively, P>0.36). Twelve of the 13
different individuals identified as intruders were males.
They were usually trespassing on a neighbouring
territory, but were sometimes as far as four territories
away from home.
Mate guarding was significantly more intense in the
2 days prior to the laying of the first egg (days 2 and
1) than earlier (days 7–3: Wilcoxon signed-ranks
test for 11 pairs: T + =63.5, P=0.002; Fig. 1). It then
dropped dramatically on the day the first egg was laid
(Wilcoxon test for seven pairs comparing prelaying
period plus completed nest with day of first egg: T + =28,
P=0.008; Fig. 1), and remained low on the day of the
second egg too. After the first egg was laid, the male, in
particular, began to spend more time on the nest (see
below). However, males that did less nest guarding on the
day of the first egg did not spend significantly more time
mate guarding (Spearman rank correlation between
% time spent nest guarding and % time mate guarding
that day: rS = 0.137, N=11, P=0.67). Therefore, the
sudden drop in mate guarding was not a consequence of
increased nest duty; males simply followed the females
much less once the first egg appeared.
DAVIES ET AL.: GUARDING VERSUS CUCKOOS AND CUCKOLDRY
50
Male
Female
Total
75
40
Time to see cuckoo (min)
Time attending nest (%)
100
50
25
30
20
10
0
1
2
3
4
No. of eggs
Figure 2. Mean±SE percentage of time spent attending the nest
during the laying period, by males, females, and in total. Data are
from 30-min watches of 12 control pairs observed each day during
the laying of clutches of four eggs.
Nest Attendance At Control Nests
Figure 2 shows attendance at 12 nonexperimental nests
where females laid a clutch of four eggs and where we did
nest watches every day during the laying period. Almost
all nest attendance involved one adult sitting on the nest
(mean of 97% nest attendance time for the 12 nests).
Change-overs were rapid, with the sitting bird slipping
away as its partner announced its imminent arrival with a
soft ‘churr’. Total nest attendance increased significantly
during the laying period (Friedman ANOVA: 23 =26.84,
P<0.0001; Fig. 2). However, the contribution of the two
sexes changed markedly. Most of the attendance before
the last egg was laid was by the male, who was significantly more attendant than the female on the days the
first, second and third eggs were laid (Wilcoxon signedranks tests: T + >66, N=12, P<0.017 on all 3 days). Male
attendance did not vary across the laying period
(Friedman ANOVA: 23 =4.73, P=0.19) but female attendance gradually increased (23 =25.63, P<0.001), so that by
the day the fourth and final egg was laid there was no
significant difference between the sexes (Wilcoxon:
T + =15.5, P=0.22).
Male attendance varied considerably between nests. For
example, on the day of the first egg it varied from 0 to
80% of the time and on the day of the second egg it
varied from 17 to 73%. Individual males were consistent
in their attendance between these 2 days (Spearman rank
correlation: rS =0.60, N=12, P=0.049). We examined
whether this variation was related to female nest attendance, laying date, the number of cuckoos we saw or heard
during the nest watch, or the number of female neighbours who were in their fertile period (7 days before
laying to the day of the penultimate egg), as a measure of
the opportunities for males to pursue extrapair copulations. However, there was no significant correlation with
any of these variables (r10 =0.080–0.286, P=0.10–0.52)
0
0
25
50
75
Nest attendance (% time)
100
Figure 3. The time taken for the first reed warbler to see the cuckoo
mount in relation to the % total time spent attending the nest prior
to the presentation. Spearman rank correlation: rS = −0.451, N=28
nests, P=0.019.
and a multiple regression was not significant (F4,30 =1.52,
P=0.22).
Cuckoo and Jay Presentations
Time to spot the cuckoo
The greater the nest attendance prior to cuckoo
presentation, the quicker the reed warblers spotted the
cuckoo (Fig. 3). It could be argued that this is confounded
by the greater nest attendance during the later stages of
egg laying. Perhaps increased motivation to defend the
nest when there are more eggs might cause the increased
speed of reaction to the cuckoo, rather than attendance
time per se. To test this, we regressed nest attendance on
laying stage and calculated the residuals, which were
normally distributed. We regressed the time it took to
spot the cuckoo on these residual nest attendance times
and still found a significant effect (F1,40 =6.7, P=0.013), so
nest attendance was important even after controlling for
laying stage.
We could watch the reed warblers’ arrival in detail at 37
nests: in 23 cases the pair arrived together; in nine cases
the male arrived first, followed by the female within
3 min (probably attracted by his mobbing calls); in four
cases, only the male arrived before the 5-min trial terminated; and in one case, only the female did so. Thus in 32
of the 37 experiments (86%) both adults saw the cuckoo.
The pair were more likely to arrive together in trials
prior to laying (14 of 16 cases), when the male was
guarding his female closely, than at the one–three-egg
stage (nine of 21 cases; 21 =5.91, P<0.02), when the male
was more likely to be alone. During the one–three-egg
stage, the male was more likely to spot the cuckoo first
(11 cases) than was the female (one case), as expected
from his greater nest attendance.
289
ANIMAL BEHAVIOUR, 65, 2
Desertions before laying
Effects on nest attendance
Figure 4 compares nest attendance before and after the
cuckoo on the day of the presentation. There was no
effect for presentations at the no-egg stage, but a marked
increase in attendance after the cuckoo when presented at
the one-egg stage, owing to a significant increase in male
attendance. Presentations at the two-egg stage also led to
increased total attendance (although of marginal significance with the small sample size). At later stages there was
no effect, but nest attendance was by then already at a
high level at control nests.
This increase was not a general response to any large
bird next to the nest, because the jay caused a small, but
significant, decrease in total and male nest attendance
(Fig. 4).
For nests that were not deserted, we continued watches
on subsequent days to see if there were any long-term
effects (Fig. 5). Prelaying presentations of the cuckoo did
not cause a significant increase in nest attendance
on subsequent days before the laying of the first egg
(Wilcoxon test: T + =8, N=8, P>0.5). However, compared
with control nests (Mann–Whitney U tests), prelaying
presentations of the cuckoo led to increased nest attendance on the day the first egg was laid (U=29.5, N1 =12,
N2 =9, P=0.08) and on the day the second egg was laid
(U=17.5, N1 =12, N2 =8, P=0.018; Fig. 5). These increases
elevated nest attendance to the same levels as at nests
that had experienced the cuckoo presentation during
laying itself (comparing one-egg and two-egg attendance
after prelaying versus one-egg cuckoo presentations,
Mann-Whitney U tests: U>40, N1 =9, N2 =11, P>0.45 for
both comparisons; Fig. 5).
Cuckoo presentations at the one-egg stage also caused
elevated nest attendance at the two-egg stage (compared
Total nest attendance (% time)
100
Male nest attendance (% time)
After seven of the 17 (41%) cuckoo presentations at the
no-egg stage in 2000–2001, the reed warblers deserted,
apparently before any eggs were laid. The female dismantled the nest and used the material to build a new
one at another site 3–51 m (mean 22 m) away. This
desertion rate was much higher than at control nests
monitored daily from nest completion, where only three
of 99 (3%) were apparently deserted before an egg was
laid (21 =22.18, P<0.001). Prelaying desertions were a
specific reaction to the cuckoo, because when we repeated
the presentations in 2002, at a site 3.5 km away, this time
with the jay as a control, there was a similar response to
the cuckoo (five of 13 nests, 38%, deserted before eggs
laid) but none to the jay (none of 14 nests deserted;
Fisher’s exact test: P=0.016).
Once a clutch had been started, cuckoo presentations
did not significantly increase the desertion rate; for
clutches that survived depredation for at least 6 days after
clutch completion, three of 28 (11%) were deserted after
cuckoo presentations at the one–four-egg stage, compared with three of 64 (5%) clutches with no cuckoo
(21 =0.38, P>0.5). Jay presentations at the one-egg stage
did not cause desertion (none of six surviving clutches
deserted).
Female nest attendance (% time)
290
(a)
**
†
75
50
*
25
0
100
(b)
*
75
50
*
25
0
100
(c)
Before presentation
After presentation
75
50
25
0
0
1
2
3–4
Cuckoo
presentation
1
Jay
presentation
Nest stage (no. of eggs)
Figure 4. Effect of cuckoo and jay mounts on percentage of time
spent attending the nest comparing before and after presentation
on the day of the experiment. Data are mean±SE % time attending
the nest (<3 m) for: (a) either male or female, (b) male and (c)
female reed warblers. Significance levels from Wilcoxon signed-ranks
tests (†P=0.07; *P<0.05; **P<0.01). Number of pairs tested with
cuckoo mount was 14 at no-egg, 11 at one-egg, six at two-egg and
seven at three–four-egg stages; and with jay mount, 10 at one-egg
stage.
to control nests: U=45, N1 =10, N2 =21, P=0.011) but not
at the three–four-egg stages, when attendance was high
anyway; Fig. 5). As with the response on the day of
presentation, this long-term effect was not a general
response to any large bird next to the nest because jay
presentations at the one-egg stage had no effect on nest
attendance the following day (compared to two-egg stage
controls: U=78, N1 =8, N2 =21, P=0.81).
DAVIES ET AL.: GUARDING VERSUS CUCKOOS AND CUCKOLDRY
(compared with controls) on the days of the first egg and
second egg. Cuckoo presentations at the one-egg stage led
to the same increased level of nest attendance that day
and the next day (two-egg stage) as prelaying presentations. On subsequent days (third and fourth eggs) the
presentations had no significant effects, but by then
attendance at control nests was at a high level anyway.
Although both members of the pair usually saw, and
mobbed, the cuckoo, these increases in nest attendance
were entirely due to responses by the male reed warbler.
Nest attendance (% time)
100
75
50
Presentation stage
No eggs
One egg
Control
25
0
0
1
2
No. of eggs
3
4
Figure 5. Mean±SE % nest attendance, by either male or female
reed warblers, at various stages of laying (0=before any eggs laid).
Data are from 12 control nests (no cuckoo presentation), 14 nests
after the cuckoo mount had been presented at the no-egg stage,
and 11 nests after the cuckoo had been presented at the one-egg
stage.
Considering responses of male and female reed
warblers separately, there were no significant effects of
cuckoo presentations on female nest attendance on subsequent days, compared with controls (P=0.39–0.88). The
only effects were increased attendance, compared with
controls, by males (increased one-egg nest attendance
after prelaying cuckoo presentations: U=27, N1 =9,
N2 =12, P=0.054; increased two-egg nest attendance after
one-egg cuckoo presentations: U=35.5, N1 =10, N2 =21,
P=0.0032).
In summary, for nests that were not deserted, the sight
of a cuckoo in the prelaying stage had no effects on nest
attendance before laying began, but led to an increase
Paternity at Control and Experimental Nests
We determined paternity for 52 broods. In 12 of these
there was one (N=8) or two (N=4) unhatched eggs, and in
five broods one (N=4) or two (N=1) nestlings died before
sampling. So for 17 broods (33%) our paternity measures
were incomplete. However, there was no difference across
treatments in the proportion of these incompletely sampled broods (10 out of 29 nests where we had presented a
cuckoo, compared with six of 16 nests with no cuckoo;
21 =0.015, P>0.9; one of seven nests with a jay presentation). Therefore our analysis should not be biased by
these missing samples.
The top part of Table 2 summarizes the paternity data
from 29 broods where we had presented the cuckoo at
various stages during laying. Reed warblers lay one egg
per day, soon after dawn. Because eggs are fertilized
ca. 24 h before they are laid (Birkhead & Møller 1992),
cuckoo presentations in the afternoon of the day of the
penultimate egg will have been made after all the eggs
had already been fertilized, as will have all the presentations on the day of the final egg. Any increase in male
nest attendance caused by these late-stage cuckoo presentations should, therefore, not have affected paternity. So,
for analysis, we included these cases (starred in Table 2,
summarized in the seventh row) with controls, where no
cuckoo or jay presentations had been done (eighth row,
Table 2). We compared them with the cases where we had
Table 2. Frequency of extrapair paternity for experimental broods, where a cuckoo mount had been presented either before or during laying,
and control broods where no presentations were made, or a jay mount had been presented at the one-egg stage
No. of extrapair young in
each brood
No. of broods with
extrapair young
Total no. of
extrapair young
0/4×6, 0/2×2, 1/1
0/4×4, 0/3×4, 0/2*
0/5, 2/4×2, 0/3×3*
0/5, 0/4×2*
0/4*, 2/3*
1/9
0/9
2/6
0/3
1/2
1/29
0/30
4/22
0/13
2/7
Cuckoo summary†
Eggs still to be fertilized
After all eggs fertilized*
0/5×2, 0/4×10, 2/4×2, 0/3×4, 0/2×2, 1/1
0/4×3, 0/3×3, 2/3, 0/2
3/21
1/8
5/75
2/26
No cuckoo or jay
Jay: 1 egg
0/5, 1/5, 0/4×4, 1/4, 0/3×8, 1/3
2/5, 0/5, 0/4×4, 0/2
3/16
1/7
3/57
2/28
Treatment
Cuckoo
Stage: 0
1
2
3
4
Total
egg
egg
eggs
eggs
eggs
8/52
(15.4%)
*Clutches in which all the eggs had been fertilized by the time the cuckoo was presented (see text).
†Collates data from the no-egg to the four-egg stages into two categories.
12/186
(6.5%)
291
292
ANIMAL BEHAVIOUR, 65, 2
presented the cuckoo at an earlier stage during laying,
when there were still eggs to be fertilized, and therefore
when increased male nest attendance could potentially
be at the expense of lost paternity (sixth row, Table 2).
However, there was no significant difference, either in the
proportion of broods with one or more extrapair young
(21 =0.037, P>0.8), or in the proportion of all young that
were sired by extrapair males (21 =0.026, P>0.8). Of all
broods sampled, eight (15.4%) had one or more extrapair
young, and 6.5% of all young (N=186) were sired by
extrapair males (Table 2). There were no cases of extrapair
maternity.
We identified the extrapair fathers for four of the eight
broods (see Methods for details), and determined their
breeding stage in their own territories at the time the
focal female was probably most susceptible to extrapair
fertilizations (the 2 days prior to laying, see Colegrave
et al. 1995). One male was from the next territory but one
(nests 35 m apart); he fathered one of five chicks and his
own female had just completed her clutch. Another male
was also from the next territory but one (nests 59 m
apart); he fathered two of three chicks and his own female
was in mid-laying. A third male was again from the next
but one territory; he fathered two of four chicks and was
unpaired at the time, but paired up 3 weeks later (nest
built 81 m from the focal nest). Finally, one male did not
have a territory on our study stretch (but may have had
one nearby); he was caught for ringing 195 m from the
focal nest and fathered one of four chicks.
DISCUSSION
Responses to Cuckoos
Female cuckoos are a double threat to reed warblers;
they may parasitize the nest and they may depredate
complete clutches or broods to force the hosts to lay a
replacement clutch, which they can then parasitize
(Gärtner 1981). Previous studies have shown that reed
warblers have specific responses to adult cuckoos near
their nest that reflect these threats. First, reed warblers
attack a cuckoo, by close mobbing and by striking it. This
contrasts with their more wary approach to a jay or a
sparrowhawk, Accipiter nisus, which, unlike cuckoos, are
also a danger to the adults themselves. Reed warblers
cease reacting to cuckoos once their young fledge, and
the cuckoo is no longer a danger, whereas alarms to a jay
or sparrowhawk continue, as both are capable of killing
fledglings (Duckworth 1991). Second, the sight of a
cuckoo near the nest stimulates increased egg rejection
(Davies & Brooke 1988; Moksnes et al. 2000). Our results
here show two other responses to adult cuckoos: nest
desertion and increased nest attendance.
Nest desertion
Presentation of a cuckoo mount prior to laying led to
desertion before any eggs were laid at 40% of nests,
compared to no desertions after jay presentations and 3%
desertion at control nests with no presentations. If the
sight of a cuckoo near the nest signals a high probability
of parasitism once laying begins, then it may pay the
female reed warbler to abandon the current nest and start
again, because she can build a new nest within 4–7 days.
The strong response to the cuckoo, but not to the jay,
would be adaptive if a cuckoo was more likely to return to
an empty nest. Furthermore, if a nest is parasitized
successfully this is more costly because the hosts then
spend 6–7 weeks rearing a cuckoo chick, whereas
after depredation they are free to start a new clutch
straightaway.
Once laying began, the cuckoo presentations did not
cause increased desertion. At this later stage desertion
would be more costly because the female would already
be committed to laying the clutch and she would not be
able to build a new nest in time to save the remaining
eggs. By contrast, female moorhens, Gallinula chloropus,
will sometimes desert a clutch at the one–two-egg stage in
response to conspecific brood parasitism. In this species
the male helps the female to construct a new nest, which
can therefore be made ready within 12 h, so that the
female can continue laying at the new site without a
break in the laying sequence (McRae 1995).
Increased nest attendance
Cuckoo presentations caused a marked increase in nest
attendance during the early laying period (days of first
and second eggs), but not before laying began, and the
increase was due entirely to a response by the male, even
though both members of the pair had usually seen and
mobbed the cuckoo. Increased male nest attendance
was a specific response to the cuckoo because jay
presentations led to a small decrease in male nest attendance. Are these responses what we would expect if reed
warblers trade-off maintaining a vigil for cuckoos with
other essential activities during the egg-laying period?
The lack of an effect of cuckoo presentations until the
warblers laid their first egg makes good sense, for two
reasons. First, reed warblers have the perfect defence
against cuckoo eggs laid before they begin their own
clutch, because they reject any eggs that appear before the
female warbler herself begins to lay (Davies & Brooke
1988). After laying has begun, however, most mimetic
cuckoo eggs are accepted unless the warblers have seen a
cuckoo at their nest (Davies & Brooke 1988; Moksnes et
al. 2000). Therefore increased nest attendance after the
laying of the first egg will increase the chance that
the warblers gain information about the likelihood of
parasitism. Female cuckoos parasitize nests only after the
hosts have begun their clutch, but they may check nests
prior to laying (Chance 1940; Moksnes et al. 2000). Hence
the sight of a cuckoo near the nest before laying has
begun should alert the hosts to an increased chance of
parasitism later on. This may explain why presentations
of the cuckoo made during the prelaying stage led to the
same increased levels of nest attendance after laying
began, as the presentations made during early laying
itself.
Second, the 2 days immediately before egg laying is the
period when extrapair matings are most likely to lead to
fertilizations (Westneat 1994; Colegrave et al. 1995). This
probably explains why male mate guarding was most
DAVIES ET AL.: GUARDING VERSUS CUCKOOS AND CUCKOLDRY
intense then (Fig. 1), and may also contribute to why
males did not respond to prelaying cuckoo presentations
until laying began.
In some cuckoo hosts, only the females incubate and
only females reject cuckoo eggs (Lotem et al. 1995;
Palomino et al. 1998). However, in reed warblers both
sexes incubate and both reject (Davies & Brooke 1988), so
potentially both would gain from increased nest attendance. Why, then, did only males respond to the cuckoo
presentations? Our results showed that increased male
attendance after laying had begun did not lead to
increased paternity loss, so the main costs are likely to be
energetic (Komdeur & Kats 1999). Perhaps these costs
would be greater for the laying female because she needs
to spend more time foraging to form reserves for her eggs.
Why No Paternity Costs?
Although extrapair matings are probably most potent
prior to the onset of laying, they can nevertheless lead to
fertilizations after laying has started (Birkhead et al. 1988;
Davies et al. 1992; Westneat 1994; Colegrave et al. 1995).
Reed warblers lay one egg per day, at dawn, and since eggs
are fertilized about 24 h before they are laid, the second
egg is fertilized soon after the first egg is laid. But extrapair matings on the day of the first egg could still
potentially fertilize the third and fourth eggs. Why, then,
did we not find paternity costs from increased male nest
attendance on the first day of laying, when half the
clutch was still at risk? Certainly, the levels of extrapair
paternity (15% of broods with one or two extrapair
young, 6% of all young extrapair) suggests that reed
warblers face a significant threat from other males.
One possible explanation is that females are less likely
to solicit matings from their social mate once they have
laid their first egg (Arvidsson 1992; Birkhead & Møller
1993; Sheldon & Burke 1994). If this also applied to
solicitation of extrapair matings, then reduced mate
guarding once laying begins would not be very costly,
because the pair male’s sperm would be numerically
dominant in the female’s storage tubules as a result of
prelaying inseminations (Birkhead & Møller 1993). Our
data on intrusion rates were not detailed, but studies of
other species have shown reduced intrusions by neighbouring males once the first egg is laid, which suggests
that the main threat of extrapair matings is before laying
(Westneat 1987; Hasselquist et al. 1995). This may
explain why male reed warblers reduced their mate guarding after the first egg, even at control nests, and were
prepared to increase nest attendance then in response to
the cuckoo presentations.
Why Guard by Sitting?
Our evidence for the idea that increased nest attendance by reed warblers reflects guarding against cuckoos is
three-fold: first, the sight of a cuckoo at the nest caused
increased nest attendance, whereas the sight of a jay
did not; second, increased nest attendance caused the
warblers to spot a cuckoo more quickly; third, as shown
previously (Davies & Brooke 1988; Moksnes et al. 2000),
spotting a cuckoo alerted them to parasitism and led to
increased egg rejection. These results suggest that cuckoos
should be secretive not only when they lay, but also when
they monitor host nests prior to laying, because their
activities could lead to host desertion or increased host
vigilance.
If the function of increased nest attendance during
early laying is to watch for cuckoos, why do reed warblers
sit in the nest rather than merely perch nearby? One
possibility is that sitting might physically block a
parasite’s access. In North America, yellow warblers,
Dendroica petechia, sometimes respond to the presence of
a female brown-headed cowbird, Molothrus ater, by rushing to sit on the nest, which may prevent this parasite
from laying (Hobson & Sealy 1989; but see Tewksbury
et al. 2002). However, although the larger hosts of the
common cuckoo can occasionally repel a cuckoo (Molnar
1944), reed warblers are too small to do this, and will
leave the nest if a cuckoo approaches (Moksnes et al.
2000). Second, sitting might dissuade the parasite from
laying, because it can more easily see that the hosts are in
attendance and would therefore be alerted to reject the
parasite egg. However, reed warbler nests are usually in
dense reeds, so the cuckoo probably cannot see the
nest before she glides down from her tree perch to lay.
Furthermore, there is no evidence that the cuckoo is less
likely to lay if reed warblers are present at their nest
(Moksnes et al. 2000). Sitting may simply be a more
comfortable, and perhaps energetically cheaper, way to
guard the nest, and it may also reduce the conspicuousness of both the guarding adult and the clutch to passing
predators.
Alternative Hypotheses
We now consider three other possible benefits of
increased nest attendance. First, it could represent guarding against predators. For example, male Seychelles
warblers, Acrocephalus sechellensis, do not incubate, but
while the female is away foraging they guard the clutch
against depredation by an endemic weaverbird (Komdeur
& Kats 1999). However, the main nest predators of reed
warblers are members of the crow family (Corvidae),
mink, Mustela vison, and weasels, Mustela nivalis, all of
which are large predators that the warblers would be
unable to chase away. Our presentations confirmed that
the warblers regarded jays (a corvid) as dangerous,
because they led to decreased nest guarding.
Second, nest guarding may enable hosts to defend
against conspecific brood parasitism (McRae 1996). However, this is unlikely to apply to reed warblers because we
found no evidence for parasitism by conspecifics, nor has
any been reported for the related great reed warbler, A.
arundinaceus (Hasselquist et al. 1995), Australian reed
warbler, A. australis (Welbergen et al. 2001), or sedge
warbler, A. schoenobaenus (Buchanan & Catchpole 2000).
Finally, because increased nest attendance involved the
male sitting on the clutch, it may reflect a decision to
initiate incubation. Although male reed warblers do not
develop a vascularized brood patch, they can warm the
293
294
ANIMAL BEHAVIOUR, 65, 2
eggs to some extent (Duckworth 1992). However, it is not
obvious why the sight of a cuckoo near the nest would
favour the earlier onset of incubation. Even if earlier
incubation led to earlier hatching of the first-laid eggs,
this would not defeat the cuckoo because newly hatched
cuckoo chicks eject reed warbler nestlings as readily as
eggs. Nevertheless, while we believe that guarding against
cuckoos provides the most likely explanation for
increased nest attendance by male reed warblers, the role
of males in incubation, and possible sexual conflicts over
hatching asynchrony (Slagsvold et al. 1994) would be
worth exploring further.
Acknowledgments
We thank the Natural Environment Research Council for
funding this study, the National Trust for permission to
work on Wicken Fen, Chris Thorne and the Wicken Fen
Group for ringing facilities and the Home Office and
English Nature for licences. The molecular work was done
at the NERC-funded Sheffield Molecular Genetics Facility
and we thank Deborah Dawson and Andy Krupa for
assistance.
References
Alvarez, F. 1996. Model cuckoo Cuculus canorus eggs accepted by
rufus bush chats Cercotrichas galactotes during the parasite’s
absence from the breeding area. Ibis, 138, 340–342.
Arvidsson, B. L. 1992. Copulation and mate guarding in the willow
warbler. Animal Behaviour, 43, 501–509.
Bassam, B. J., Caetano-Anolles, G. & Gresshoff, P. M. 1991. Fast
and sensitive silver staining of polyacrylamide gels. Analytical
Biochemistry, 196, 80.
Birkhead, T. R. & Møller, A. P. 1992. Sperm Competition in
Birds: Evolutionary Causes and Consequences. London: Academic
Press.
Birkhead, T. R. & Møller, A. P. 1993. Why do male birds stop
copulating while their partners are still fertile? Animal Behaviour,
45, 105–118.
Birkhead, T. R., Pellatt, J. E. & Hunter, F. M. 1988. Extra-pair
copulation and sperm competition in the zebra finch. Nature, 334,
60–62.
Brooke, M. de L. & Davies, N. B. 1987. Recent changes in host
usage by cuckoos Cuculus canorus in Britain. Journal of Animal
Ecology, 56, 873–883.
Brooke, M. de L., Davies, N. B. & Noble, D. G. 1998. Rapid decline
of host defences in response to reduced cuckoo parasitism:
behavioural flexibility of reed warblers in a changing world.
Proceedings of the Royal Society of London, Series B, 265,
1277–1282.
Bruford, M. W., Hanotte, O., Brookfield, J. F. Y. & Burke, T. 1998.
Multilocus and single-locus DNA fingerprinting. In: Molecular
Genetic Analysis of Populations: a Practical Approach. 2nd edn (Ed.
by A. R. Hoezel), pp. 287–336. Oxford: IRL Press.
Buchanan, K. L. & Catchpole, C. K. 2000. Extra-pair paternity in the
socially monogamous sedge warbler Acrocephalus schoenobaenus
as revealed by multilocus DNA fingerprinting. Ibis, 142, 12–20.
Catchpole, C. K. 1972. A comparative study of territory in the
reed warbler Acrocephalus scirpaceus and sedge warbler A.
schoenobaenus. Journal of Zoology, 166, 213–231.
Catchpole, C. K. 1973. The functions of advertising song in the
sedge warbler Acrocephalus schoenobaenus and the reed warbler
A. scirpaceus. Behaviour, 46, 300–320.
Chance, E. P. 1940. The Truth About the Cuckoo. London: Country
Life.
Colegrave, N., Birkhead, T. R. & Lessells, C. M. 1995. Sperm
precedence in zebra finches does not require special mechanisms
of sperm competition. Proceedings of the Royal Society of London,
Series B, 259, 223–228.
Davies, N. B. & Brooke, M. de L. 1988. Cuckoos versus reed
warblers: adaptations and counter-adaptations. Animal Behaviour,
36, 262–284.
Davies, N. B. & Brooke, M. de L. 1989. An experimental study of
co-evolution between the cuckoo Cuculus canorus and its hosts. II.
Host egg markings, chick discrimination and general discussion.
Journal of Animal Ecology, 58, 225–236.
Davies, N. B. & Green, R. E. 1976. The development and ecological significance of feeding techniques in the reed warbler
Acrocephalus scirpaceus. Animal Behaviour, 24, 213–229.
Davies, N. B., Hatchwell, B. J., Robson, T. & Burke, T. 1992.
Paternity and parental effort in dunnocks Prunella modularis:
how good are male chick-feeding rules? Animal Behaviour, 43,
729–745.
Davies, N. B., Brooke, M. de L. & Kacelnik, A. 1996. Recognition
errors and probability of parasitism determine whether reed
warblers should accept or reject mimetic cuckoo eggs. Proceedings
of the Royal Society of London, Series B, 263, 925–931.
Duckworth, J. W. 1991. Responses of breeding reed warblers
Acrocephalus scirpaceus to mounts of sparrowhawk Accipiter nisus,
cuckoo Cuculus canorus, and jay Garrulus glandarius. Ibis, 133,
68–74.
Duckworth, J. W. 1992. Effects of mate removal on the behaviour
and reproductive success of reed warblers Acrocephalus scirpaceus.
Ibis, 134, 164–170.
Ellegren, H. 1992. Polymerase chain reaction (PCR) analysis of
microsatellites: a new approach to studies of genetic relationships
in birds. Auk, 109, 886–895.
Gärtner, K. 1981. Das Wegnehmen von Wirtsvogeleiern durch
den Kuckuck Cuculus canorus. Ornithologische Mitteilungen, 33,
115–131.
Hasselquist, D., Bensch, S. & von Schantz, T. 1995. Low frequency
of extra-pair paternity in the polygynous great reed warbler
Acrocephalus arundinaceus. Behavioral Ecology, 6, 27–38.
Hobson, K. A. & Sealy, S. G. 1989. Responses of yellow warblers
to the threat of cowbird parasitism. Animal Behaviour, 38, 510–
519.
Komdeur, J. & Kats, R. K. H. 1999. Predation risk affects trade-off
between nest guarding and foraging in Seychelles warblers.
Behavioral Ecology, 6, 648–658.
Leisler, B. & Catchpole, C. K. 1992. The evolution of polygamy in
European reed warblers of the genus Acrocephalus: a comparative
approach. Ethology, Ecology and Evolution, 4, 225–243.
Lindholm, A. K. 1999. Brood parasitism by the cuckoo on patchy
reed warbler populations in Britain. Journal of Animal Ecology, 68,
293–309.
Lindholm, A. K. & Thomas, R. J. 2000. Differences between
populations of reed warblers in defences against brood parasitism.
Behaviour, 137, 25–42.
Lotem, A., Nakamura, H. & Zahavi, A. 1995. Constraints on egg
discrimination and cuckoo–host co-evolution. Animal Behaviour,
49, 1185–1209.
McRae, S. B. 1995. Temporal variation in responses to intraspecific brood parasitism in the moorhen. Animal Behaviour, 49,
1073–1088.
McRae, S. B. 1996. Brood parasitism in the moorhen: brief
encounters between parasites and hosts and the significance of an
evening laying hour. Journal of Avian Biology, 27, 311–320.
Marchetti, K. 1992. Costs to host defence and the persistence of
parasitic cuckoos. Proceedings of the Royal Society of London, Series
B, 248, 41–45.
DAVIES ET AL.: GUARDING VERSUS CUCKOOS AND CUCKOLDRY
Marshall, T. C., Slate, J., Kruuk, L. E. B. & Pemberton, J. M. 1998.
Statistical confidence for likelihood based paternity inference in
natural populations. Molecular Ecology, 7, 639–655.
Martinez, J. G., Soler, J. J., Soler, M., Møller, A. P. & Burke, T.
1999. Comparative population structure and gene flow of a
brood parasite, the great spotted cuckoo (Clamator glandarius),
and its primary host, the magpie (Pica pica). Evolution, 53, 269–
278.
Moksnes, A. & Røskaft, E. 1989. Adaptations of meadow pipits
to parasitism by the common cuckoo. Behavioral Ecology and
Sociobiology, 24, 25–30.
Moksnes, A., Røskaft, E. & Korsnes, L. 1993. Rejection of
cuckoo Cuculus canorus eggs by meadow pipits Anthus pratensis.
Behavioral Ecology, 4, 120–127.
Moksnes, A., Røskaft, E., Hagen, L. G., Honza, M., Mørk, C. &
Olsen, P. H. 2000. Common cuckoo Cuculus canorus and host
behaviour at reed warbler Acrocephalus scirpaceus nests. Ibis, 142,
247–258.
Molnar, B. 1944. The cuckoo in the Hungarian plain. Aquila, 51,
100–112.
Palomino, J. J., Martin-Vivaldi, M., Soler, M. & Soler, J. J. 1998.
Females are responsible for ejection of cuckoo eggs in the rufous
bush robin. Animal Behaviour, 56, 131–136.
Richardson, D. S., Jury, F. L., Dawson, D. A., Salgueiro, P.,
Komdeur, J. & Burke, T. 2000. Fifty Seychelles warbler
(Acrocephalus sechellensis) microsatellite loci polymorphic in
Sylviidae species and their cross-species amplification in other
passerine birds. Molecular Ecology, 9, 226–231.
Schulze-Hagen, K. 1992. Parasitierung und Brutverluste durch den
Kuckuck (Cuculus canorus) bei Teich-und Sumpfrohrsänger
(Acrocephalus scirpaceus, A. palustris) in Mittel-und Westeurope.
Journal für Ornithologie, 133, 237–249.
Sheldon, B. C. & Burke, T. 1994. Copulation behaviour and
paternity in the chaffinch. Behavioral Ecology and Sociobiology, 34,
149–156.
Slagsvold, T., Amundsen, T. & Dale, S. 1994. Selection by sexual
conflict for evenly spaced offspring in blue tits. Nature, 370,
136–138.
Tewksbury, J. J., Martin, T. E., Heil, S. J., Kuehn, M. J. & Jenkins,
J. W. 2002. Parental care of a cowbird host: caught between the
costs of egg removal and nest predation. Proceedings of the Royal
Society of London, Series B, 269, 423–429.
Walsh, P. S., Metzger, D. A. & Higuchi, R. 1991. Chelex 100 as a
medium for simple extraction of DNA for PCR-based typing from
forensic material. Biotechniques, 10, 506.
Welbergen, J., Komdeur, J., Kats, R. & Berg, M. 2001. Egg
discrimination in the Australian reed warbler (Acrocephalus
australis): rejection response toward model and conspecific eggs
depending on timing and mode of artificial parasitism. Behavioral
Ecology, 12, 8–15.
Westneat, D. F. 1987. Extra-pair copulations in a predominantly
monogamous bird: observations of behaviour. Animal Behaviour,
35, 865–876.
Westneat, D. F. 1994. To guard mates or go forage: conflicting
demands affect the paternity of male red-winged blackbirds.
American Naturalist, 144, 343–354.
295