IBI_085.

fm Page 614 Monday, September 9, 2002 11:15 AM

Ibis (2002), 144, 614 – 622

European Cuckoo Cuculus canorus parasitism and

Blackwell Science, Ltd

host’s rejection behaviour in a heavily parasitized Great

Reed Warbler Acrocephalus arundinaceus population
CSABA MOSKÁT 1* & MARCEL HONZA 2
1
Animal Ecology Research Group of the Hungarian Academy of Sciences, c/o Hungarian Natural History Museum,
H-1083 Budapest, Ludovika ter 2., Hungary
2
Institute of Vertebrate Ecology, Academy of Sciences of the Czech Republic, Kvetná 8, 603 65, Brno,
The Czech Republic

An unusually high frequency (64%) of European Cuckoo Cuculus canorus parasitism was
found in Great Reed Warbler Acrocephalus arundinaceus clutches in central Hungary. Sixty-
four per cent of the parasitized clutches contained one Cuckoo egg, 23% contained two,
10% had three and 3% had four. This means that 58% of the Cuckoo eggs were found in
multiply parasitized clutches. In multiple parasitism the laying second Cuckoo removed an
egg from the clutch randomly, so preferred neither the host eggs, nor the concurrent Cuckoo
egg. Host response towards the parasitic eggs showed 66% acceptance, 12% ejection, 20%
desertion and 2% egg burial. We found great variation in both the host and the parasitic egg
colour and pattern. This reduces the chance that the parasitic egg’s appearance matched that
of the hosts’ but, in spite of this, almost perfect mimesis was found in 28% of the Cuckoo
eggs. Poorly mimetic Cuckoo eggs were more frequently rejected by Great Reed Warblers
than parasite eggs that were very similar to the host eggs. This high level of mimicry some-
times makes it difficult for the observer to identify the parasitic egg, especially when it is
similar in size to the host eggs. It is also difficult for the host, as shown by the relatively high
recognition error and ejection cost.

The European Cuckoo Cuculus canorus (hereafter through many stages of evolutionarily stable and
referred to as the Cuckoo) is an obligate brood unstable systems (Lotem & Nakamura 1998). At the
parasite that lays its eggs in the nests of several host end of the process, well-developed egg recognition
species (Wyllie 1981, Moksnes & Røskaft 1995). Dif- ability of the host and a reduced rate of acceptance
ferent geographical areas typically have one main of the parasitic egg may force the Cuckoo to switch
host species, and the eggs of the particular Cuckoo to a new host. There is evidence that in intermediate
population show at least some degree of mimicry of stages of the arms race an evolutionary equilibrium
the host’s eggs (Wyllie 1981, Moksnes & Røskaft exists between accepter and rejecter individuals of
1995, Davies & Brooke 1998). The arms race the host population (see e.g. Lotem et al. 1992,
hypothesis (Dawkins & Krebs 1979) explains the Davies 1999, Takasu 1999). When a host species
coevolution of egg morphology in time between the is newly parasitized it may reach a high frequency
Cuckoo and the hosts (Davies & Brooke 1988, of parasitism at first (Yamagishi & Fujioka 1986,
1989a, 1989b, Øien et al. 1995, Soler & Møller Nakamura 1990), for example, c. 80% during 10 years
1996, Davies 1999, 2000) as a multistep evolution- in the case of the Azure-winged Magpie Cyanopica
ary process, with evolving mimicry of the Cuckoo cyana parasitized by the Cuckoo (Takasu et al. 1993),
eggs on the one hand, and the evolving egg recogni- but later this high level of parasitism might decline
tion ability and antiparasite behaviour of the host (Nakamura et al. 1998), and in many Cuckoo–host
population on the other hand. This process may go systems it usually is much less than 10% (Davies &
Brooke 1989b, Schulze-Hagen 1992, Johnsgard 1997).
*Corresponding author. The intermittent arms race hypothesis (Soler et al.
Email: moskat@ludovika.nhmus.hu 1998) suggests that a brood parasite–host interaction

© 2002 British Ornithologists’ Union

IBI_085.fm Page 615 Monday, September 9, 2002 11:15 AM

Cuckoo parasitism on the Great Reed Warbler 615

in a short-term study may seem to be an arms race, usually arrive in early May, the latest among the
but in a long-term study appears to be described as Acrocephalus warblers (Moskát & Báldi 1999) and
an evolutionary equilibrium due to cycles in the start breeding from mid-May. Between 15 May and
abundance of cuckoos and parasites. Although a par- 15 July 1998, and 16 May and 15 July 1999, we sys-
ticular study on a parasitized host population may tematically checked the reedbeds at least twice a week,
reveal the presence of the arms race or the evolutionary trying to find as many warbler nests as possible in the
equilibrium, on a wider geographical and time scale nest-building stage. As c. 96% of the nests were found
these local processes look like an intermittent arms race. in the nest-building or early egg-laying stages, we were
In Hungary the Great Reed Warbler Acrocephalus able to monitor the hosts’ rejection of Cuckoo eggs.
arundinaceus is the main host of the Cuckoo in wet- Each nest was controlled daily, non-parasitized nests
land habitats (Molnár 1944). The Robin Erithacus were visited at least 2 days after incubation started,
rubecula is the main host in woodlands (Varga 1994) or at least on six additional days after parasitism
and accepts Cuckoo eggs that usually are a poor mimic happened, to observe potential reactions of the hosts
of the eggs of the Hungarian Robin population. towards the parasitic eggs. All eggs were marked with
Experiments have revealed that the Red-backed waterproof pens to monitor any exchange of eggs
Shrike Lanius collurio in northern Hungary is a strong carried out by different competing Cuckoo females.
rejecter of eggs that are unlike its own, so this species We measured the length and breadth of host and
is probably one of the major hosts that has been parasitic eggs. Almost all clutches were photographed
abandoned by the Cuckoo (Moskát & Fuisz 1999). on Kodak Grey Card using Kodak 100 ASA film and
The paper by Molnár (1944) on Cuckoo parasitism a ring flash. The mimesis of the Cuckoo egg in relation
in Hungarian Great Reed Warblers is often cited as the to host eggs was estimated from the slides, using the
description of one of the best cases of mimicry of spotted, five-point scale developed for Cuckoo eggs (Moksnes
non-plain coloured Cuckoo eggs ever found (e.g. & Røskaft 1995): (1) Perfect mimesis: the Cuckoo egg
Southern 1954, Davies & Brooke 1991, Moksnes was hard to pick out from the clutch, because both
& Røskaft 1995), and the approximately 50% para- the colour and patterning showed a high resemblance.
sitism frequency he found is one of the highest fre- (2) Good mimesis: only small deviations were found
quencies of Cuckoo parasitism on any host species between the parasitic and the host eggs, but after a
(e.g. Johnsgard 1997). The goal of the present paper close inspection of the clutch the small deviations in
is to confirm whether this high level of parasitism colour and pattern made the parasitic egg easy to
and the ‘almost perfect mimicry’ (Molnár 1944) of the recognize. (3) Moderate mimesis: it was easy to identify
Cuckoo egg still exists in the Great Reed Warbler and the Cuckoo egg, but either the colour or the pattern
Cuckoo relationships in Hungary. We also intend to of the Cuckoo egg showed similarity to the host’s
find out how the host is able to counter the effects eggs. (4) Poor mimesis: both the colour and the
of parasitism, and what type of rejection behaviour pattern of the parasitic egg differed from those of
toward the parasitic eggs occurs in this population. the host’s eggs. (5) Bad mimesis (non-mimetic): the
Cuckoo egg showed no similarity to the host’s eggs in
any respect. This type of quantification of the Cuckoo
S T U D Y A RE A A N D M E T H O D S
egg mimicry has already proved to be useful in studies
The study was conducted in central Hungary, about on brood parasitism (e.g. Moksnes et al. 1993a).
30–40 km south of Budapest, around Bugyi, Apaj, Most of the statistical tests were carried out using
and Kiskunlacháza (47°09′N, 19°05′E). We searched SPSS (Norusis 1990). All means are given ±1 stand-
for Great Reed Warbler nests in 3- to 5-m-wide reed- ard error. Fitting of the Poisson series was carried out
beds along both sides of canals in 1998 and 1999. using STATISTICA (StatSoft 1995).
The Great Reed Warbler is a typical edge species in
Hungary (Báldi & Kisbenedek 1999), so this was its
RESULTS
favoured habitat; furthermore, no nest of any other
passerine species was found in the reedbeds in our
Frequency of parasitism and host
study area. Great Reed Warblers preferred reeds that
response types
were well developed, sheltered from the wind, and
avoided dry reed patches from the previous year. In total, 193 Great Reed Warbler clutches were
(More details of habitat are available in Moskát & found: 103 in 1998 and 90 in 1999. Only nests with
Honza 2000.) In Hungary, Great Reed Warblers at least one host or Cuckoo egg were considered.

© 2002 British Ornithologists’ Union, Ibis, 144, 614 – 622

IBI_085.fm Page 616 Monday, September 9, 2002 11:15 AM

616 C. Moskát & M. Honza

Table 1. Number of Cuckoo eggs in Great Reed Warbler Table 2. Response types of Hungarian Great Reed Warblers
clutches in central Hungary (combined data from 1998 and towards parasitic Cuckoo eggs.
1999).
Host’s response type
Number of Type of
Cuckoo eggs no. of clutches no. of Cuckoo eggs parasitism Acceptance Ejection Desertion Egg burial Total
in parasitized
clutches n % n % Single 41 13 14 3 71
Multiple 67 6 18 1 92
0 70 36.3 – – all 108 19 32 4 163
1 79 40.9 79 42.2 % 66.3 11.7 19.6 2.4 100.0
2 28 14.5 56 29.9
3 12 6.2 36 19.3
4 4 2.1 16 8.6
Total 193 100.0 187 100.0
frequent response types (acceptance, ejection and
desertion; χ2 = 6.140, df = 2, P = 0.046). Desertion
was the most common type of rejection (n = 32,
From the total number of Great Reed Warbler 19.6%) that occurred mainly when Cuckoos para-
clutches, 65 (63.1%) and 58 (64.4%) were parasit- sitized recently completed, but empty host nests. The
ized in 1998 and 1999, respectively. As parasitism frequency of ejection was 11.7% (n = 19), about
rate did not differ significantly (χ2 = 0.037, df = 1, one-third of all types of rejections. The disappear-
P = 0.847), we combined the data-sets from the two ance of host eggs, termed ‘cost of ejection’ when the
years. host successfully ejected the parasitic egg or ‘recog-
Besides the unusually high frequency of para- nition error’ when the parasitic egg was accepted
sitism, we also found an extraordinarily high rate of (i.e. only host eggs were ejected, Davies & Brooke
multiple parasitism (Table 1). Almost one-quarter (1988)), was observed in six and 11 cases, respect-
(22.8%) of the Great Reed Warbler nests were mul- ively (Table 3). The total number of host eggs lost
tiply parasitized, and more than half of the Cuckoo in these ways was 32. In one case the host ejected
eggs (57.8%) were found in multiply parasitized gradually, all of the five eggs in a clutch within
clutches. These multiply parasitized clutches con- 4 days, but eventually the Cuckoo egg also. In 13
tained 2, 3 or even 4 Cuckoo eggs in the same clutch, clutches the Cuckoo egg was ejected without any
generally at the same time. The number of Cuckoo rejection cost. We observed no case of recognition
eggs found in host clutches showed no significant error in which Great Reed Warblers ejected a host
difference from the Poisson series when all types of egg from unparasitized clutches. The rarest rejec-
clutches, i.e. containing 0, 1, 2, 3 and 4 Cuckoo eggs tion type was egg burial (2.4%). This behaviour was
were used (χ2 = 2.309, df = 2, P = 0.315). observed in the pre-egg-laying stage, when Great
One hundred and eight (66.3%) parasitic eggs Reed Warblers built over the nest above the parasitic
were accepted by the Great Reed Warblers, and 55 egg only when there was no egg of their own in the
(33.7%) were rejected by ejection, desertion or egg nest.
burial (Table 2). Host reactions differed somewhat Most of the parasitic eggs were laid during the host
between single and multiple parasitism, but these egg-laying stage, but early layings (e.g. 2 days before
differences were only weakly significant for the most the laying of the first host egg) and parasitism on

Table 3. Successful and unsuccessful ejection attempts of Cuckoo eggs and disappearance of host eggs from parasitized Great Reed
Warbler clutches.

Ejections:
Without With ejection cost With recognition error
ejection cost Cuckoo egg Only host egg
Cuckoo egg
only +1 host egg +3 host eggs +5 host eggs 1 host egg 2 host eggs 3 host eggs

13 4 1 1 6 1 4

© 2002 British Ornithologists’ Union, Ibis, 144, 614–622

IBI_085.fm Page 617 Monday, September 9, 2002 11:15 AM

Cuckoo parasitism on the Great Reed Warbler 617

Table 4. Stage of parasitism for each Cuckoo egg in relation to response types of the host. (Stage is expressed in number of days after
clutch was started, e.g. day of laying the first host egg is 1, or the value −1 means that the Cuckoo egg was found in the nest 2 days
before the host laid her first egg.) Under the title ‘others’ we collected the cases of predation, damage of nests caused by storm, and
unusual laying behaviour of the Cuckoo, like laying mistakenly into abandoned nest, or old empty nest.

Host’s response type Total

Time of Removed
parasitism Acceptance Ejection Desertion Egg burial by Cuckoo Others n %

–1 2 0 0 0 0 1 3 1.6
0 –2 33 7 25 4* 8 1 78 41.7
3–6 37 10 5 0 3 3 58 31.0
7– 9 4 1 0 0 0 0 5 2.7
Unknown 32 1 2 0 1 3 39 20.9
Deserted nest 0 0 0 0 0 3 3 1.6
Old empty nest 0 0 0 0 0 1 1 0.5
Total (n) 108 19 32 4 12 12 187 100.0
Total (%) 57.8 10.2 17.1 2.1 6.4 6.4 100.0

*All cases of egg burial happened at time = 0.

incubated clutches or abandoned nests were also set of eight (Table 5). These results indicate that if
rarely observed (Table 4). Table 2 shows the hosts’ only one egg type (i.e. host or Cuckoo) was present,
responses to Cuckoo eggs. Table 4 contains more Cuckoos chose one egg from the available pool,
cases, and also shows the fate of all Cuckoo eggs, regardless of the type of the egg.
including the categories ‘removed by Cuckoo’ and The most interesting cases occurred when the laying
‘others’. For this reason the calculated acceptance Cuckoo had the chance to select either a host or a
rate and rates for rejections decrease somewhat in parasitic egg from the clutch during parasitism. From
Table 4. The category ‘removed by Cuckoo’ included the 23 clutches which had both types of eggs, altogether
cases when the parasitic egg was exchanged by 55 host eggs and 28 parasitic eggs, in total, Cuckoos
another laying Cuckoo. We were able to monitor this removed 18 host eggs and 8 parasitic eggs (Table 5).
phenomenon by numbering the eggs and analysing The frequency of the selected host and Cuckoo
the photographs. The category ‘others’ covered a eggs did not differ significantly from availability
high variety of cases (predation, laying in an old (χ2 = 0.079, df = 1, P = 0.779), i.e. the probability of
empty nest, etc.). selecting one egg at random from the clutches.

Egg removal by laying female Cuckoos Mimicry of the Cuckoo eggs

In four cases Cuckoos did not remove an egg when In the case of the Great Reed Warbler, the host and
they laid their own. Besides these, the simplest situ- parasite eggs are usually about the same size. The
ations were when female Cuckoos did not have the mean length of host eggs (23.05 ± 0.058 mm; clutch
chance to select an egg from the parasitized clutch: means: 23.00 ± 0.101) was greater than those
(i) Thirteen Cuckoo eggs were laid into empty nests, of the parasitic eggs (22.26 ± 0.086 mm), while
where Cuckoos were unable to remove an egg. Cuckoo eggs (16.57 ± 0.059 mm) were larger in
(ii) Twenty-eight clutches contained only host or only mean breadth than the Great Reed Warbler eggs
parasitic eggs: (a) From 21 clutches containing only host (16.25 ± 0.033 mm; clutch means: 16.23 ± 0.053).
eggs, in 18 cases at least one host egg was removed These differences between the clutch means (n = 82)
by the Cuckoo, which means the removal of 22 host of the measured host eggs (n = 294) and the parasitic
eggs altogether from the total of 48 Great Reed War- eggs (n = 91) proved to be significant for both the
bler eggs. (b) In each of the six clutches out of seven length and the breadth of the eggs (length: t = 5.59,
containing only parasitic egg(s) Cuckoos removed df = 171, P < 0.01; breadth: t = 4.29, df = 171, P < 0.01
at least one egg, which means seven eggs out of the at corrected significance level). An elongation index

© 2002 British Ornithologists’ Union, Ibis, 144, 614 – 622

IBI_085.fm Page 618 Monday, September 9, 2002 11:15 AM

618 C. Moskát & M. Honza

Table 5. Eggs removed by Cuckoo from Great Reed Warblers’ clutches where only one type of egg was present in the clutch, and eggs
removed where at least two types of eggs were present in the clutch. (Comment: additionally 13 empty nests were naturally parasitized
by the Cuckoo, and Cuckoos did not remove any egg from four additional nests where eggs were present.)

Cuckoo could select from host and

Cuckoo could choose from only one egg type Cuckoo eggs

No. of nests No. of eggs No. of nests No.of eggs

(egg removal / (egg removed / (egg removal/ (egg removed/
Removed egg type nests parasitized) eggs available) nests parasitized) eggs available)

Host 18/ 21 22/ 48 16 / 23 18/55

Cuckoo 3/ 4 4/ 5 7/ 19*† 7/23
Artificial Cuckoo 3/ 3 3/ 3 1/ 5† 1/5
All parasitic eggs 6/ 7 7/ 8 8 / 23 8/28
All eggs 24/ 28 29 / 56 23 / 23 26/83

*Two clutches contained two, one clutch contained three Cuckoo eggs.
†One clutch contained one artificial plus one real Cuckoo egg.

(length/breadth) also revealed a significant differ- Table 6. Mimicry of Cuckoo eggs in relation to Great Reed
ence between the clutch means of the host and the Warbler host eggs measured on the 5-state scale proposed by
Moksnes et al. (1993a), and host response to parasite’s egg.
parasite population (index: 1.41 ± 0.006, 1.34 ± 0.007
for the Great Reed Warbler and the Cuckoo, respec- Mimicry
tively; t = 7.96, df = 171, P < 0.01 at corrected Host (or Cuckoo)
significance level). For this reason the Great Reed reaction Perfect Good Moderate Poor Bad
Warbler eggs looked more elongated than the more
Acceptance 29 22 22 13 3
rounded form of the Cuckoo eggs.
Ejection 0 3 4 3 2
Interestingly, we found perfect, or more frequently Desertion 1 1 3 1 2
only almost perfect, mimicry of the Cuckoo eggs in Removed by Cuckoo 3 1 3 1 0
a high proportion (28.2%), although good and mod-
n 33 27 32 18 7
erate mimesis were also found in similar frequencies % of Cuckoo eggs 28.2 23.1 27.3 15.4 6.0
(23.1% and 27.3%, respectively; Table 6). Perfect
mimicry is illustrated by the fact that at the begin-
ning of our study we did not identify the first Cuckoo
egg that was a perfect match. We checked a nest almost all main types of host eggs, but many Cuckoo
known to have one poorly mimetic Cuckoo egg, but eggs were unlike any egg-type of the Great Reed
on the day of hatching we realized that two Cuckoo Warbler, and possibly belonged to another gens.
chicks had hatched. Mimicry of the Cuckoo egg influenced the reac-
Although some Cuckoo eggs were extremely good tion of the host towards the parasitic egg (Table 6).
matches for Great Reed Warbler eggs, the matches The best-matching cuckoo eggs were almost never
were not perfect, because of high interclutch vari- rejected, but rejection rate increased as the degree of
ation. Although a detailed analysis of matching of matching decreased. The difference between accept-
the host and parasitic eggs is beyond the scope of this ance and rejection across the mimicry categories was
paper, we should consider the high interclutch vari- significant (χ2 = 12.65, df = 4, P = 0.013).
ability in the host, and the high egg variability among
different female Cuckoos (Fig. 1). The most com-
DISCUSSION
mon ground colour of host eggs was ivory, white,
light blue or yellowish-green, but sometimes the
Frequency of parasitism
ground colour was greenish. Three main pattern
types are known: separated small dark dots, sepa- We found 64% of nests parasitized, a high value
rated or overlapping light brown, dark brown and when compared with other studies. Generally, the
bluish-grey spots (the commonest type), and adjacent frequency of Cuckoo parasitism varies from 2 to
brownish patches. Perfect mimesis was found to 20(−50)% in different geographical areas. There is

© 2002 British Ornithologists’ Union, Ibis, 144, 614–622

IBI_085.fm Page 619 Monday, September 9, 2002 11:15 AM

Cuckoo parasitism on the Great Reed Warbler 619

A similar phenomenon was reported from Spain,

when the Great Spotted Cuckoo Clamator glandarius
started to parasitize new local populations of the
Magpie Pica pica (Soler 1990). Varga (1994) reported
a yearly parasitism rate of 28.2–42.9%, mean 36.7%,
for the Robin in northern Hungary between 1965
and 1977, but the parasitism frequency decreased
between 1978 and 1991 to a mean of 17.8% (range:
2.4–30.5%). Over a longer period, only lower values
of parasitism rates persisted. Wyllie (1981), Perrin de
Brichambaut (1997) and Johnsgard (1997) showed
overviews of parasitism rate found in different
Cuckoo gentes in the western Palearctic. Generally it
was below 10%. There are many data from Britain,
where the major hosts suffer from 0.3 to 5.5% para-
sitism frequency (Davies & Brooke 1998). In Norway,
the main Cuckoo host, the Meadow Pipit Anthus
pratensis, was also parasitized at a low frequency
(0–6%) (Moksnes & Røskaft 1987). The Great Reed
Warbler is an important Cuckoo host in Japan,
where its parasitism rate is 20% (Lotem et al. 1992).
In the Czech Republic, where the Reed Warbler
Acrocephalus scirpaceus is the primary Cuckoo host
(with a parasitism rate of 18%), the Great Reed
Warbler is a much rarer, secondary host with a rate
of 5% (Moksnes et al. 1993a). From Hungary, where
the Great Reed Warbler is the primary host of the
Cuckoo, Molnár (1944) reported a parasitism rate of
50%. The present study found an even higher level
(64%). Schulze-Hagen (1992) collected data on
parasitism rate from different central and western
European countries for the Reed Warbler and the
Marsh Warbler Acrocephalus palustris. Based on 52
Figure 1. Great Reed Warbler and European Cuckoo egg types studies, Cuckoo parasitism was generally higher for
from parasitized clutches in Hungary. (a) The most common host
egg type (on the left) with light bluish ground colour and
the Reed Warbler than for the Marsh Warbler (aver-
moderate intensity of isolated dark brown, light brown and grey age parasitism rates: 8.3 ± 7.9%, and 6.3 ± 6.6% for
spots and blotches, parasitized with two Cuckoo eggs. (b) Host the Reed and the Marsh Warblers, respectively), and
egg (on the left) with ivory ground colour and dense, adjacent in one case an exceptionally high value, 63%, was
brown patches, found together with two Cuckoo eggs. (c) Perfect found for the Reed Warbler.
mimicry of the Cuckoo egg (on the left) in respect to host eggs.
(d) Good resemblance of the Cuckoo egg (on the right) to the
Molnár’s (1944) study area was situated c. 100–
host eggs. 120 km east of our study area, but the rivers, lakes,
fishponds and the dense network of the irrigation
channel system maintain habitat corridors between
evidence that the rate of Cuckoo parasitism of the the two study areas. For this reason it is surprising
Azure-winged Magpie peaked at about 50% (and that after nearly 60 years such a high or somewhat
locally and temporarily even 80%) c. 30 years after higher parasitism rate is still found in central Hun-
its breeding area overlapped that of the Cuckoo in gary. There are some other reports and observations
central Japan, and the Cuckoo started to parasitize of Cuckoo parasitism on the Great Reed Warbler
this species (Takasu et al. 1993, Nakamura et al. in Hungary from the period between Molnár’s and
1998). Shortly after this, a steady decline was our studies (Makatsch 1964, Csörgõ pers. comm.).
observed in the parasitism rate (Nakamura pers. Additionally, all of the bird checklists and monographs
comm.). published in Hungary have stated that the Great

© 2002 British Ornithologists’ Union, Ibis, 144, 614 – 622

IBI_085.fm Page 620 Monday, September 9, 2002 11:15 AM

620 C. Moskát & M. Honza

Reed Warbler was the most important Cuckoo stating that in Japan, where the Cuckoo is more
host in the lowlands of Hungary during this period common than in England, he also found no significant
(Horváth 1958; Keve 1960, 1984, Csörgõ 1984, Csörgõ evidence. Our results on egg removal proved that
1998, Magyar et al. 1998). laying female Cuckoos usually selected one egg from
parasitized clutches regardless of the egg type. If only
host or Cuckoo eggs were present in the clutch, the
Multiple parasitism
laying Cuckoo had no chance to select, but in these
In our study, 36% of the parasitized Great Reed cases Cuckoos also usually chose one egg from the
Warbler nests received more than one Cuckoo egg, and available set of eggs.
thus 58% of the Cuckoo eggs were found in multiply
parasitized nests. A consequence of multiple parasit-
Mimicry of the Cuckoo eggs in Great Reed
ism is a high level of competition among the Cuckoo
Warbler clutches in Hungary
chicks. Multiple parasitism is disadvantageous for
the Cuckoo, because only one chick can fledge. As Our study revealed that different levels of mimicry
Cuckoos usually lay their eggs in the laying stage of may coexist in the same area, habitat and host spe-
the host, we often observed that two Cuckoo chicks cies. The high variability of Cuckoo egg morphs in a
were hatched on the same day in multiple parasit- highly specialized Cuckoo gens indicates that the
ized nests. After a struggle between the Cuckoo Cuckoo is a ‘semi-generalist’ (as stated by Moksnes
chicks only one survived, because the other chick & Røskaft (1995) after checking 12 000 parasitized
was ejected from the nest like the eggs or young clutches of European passerines). They found poorly
nestlings of the host. Molnár (1944) also reported a matching Cuckoo eggs even in host species where
high rate of multiple parasitism in Hungarian Great many Cuckoo eggs were perfect mimics of the host
Reed Warblers, where 44% of the nests received 2–5 eggs. In our study area Great Reed Warbler clutches
Cuckoo eggs. Wyllie (1981) believed that a high rate show high interclutch variability. For this reason a
of multiple parasitism is a simple consequence of the Cuckoo egg, which perfectly mimics one egg type of
very high local density of the Cuckoo, relative to that the host species, sometimes does not match the host
of the host. eggs in the clutch in which it was laid. The degree of
mimicry becomes more a matter of chance, as when
a female Cuckoo with mimicry of the primary host
Do laying Cuckoos prefer to remove
lays her eggs into the nests of another host species.
competing Cuckoo eggs?
Egg mimicry by the Cuckoo has resulted from an
When the female Cuckoo lays her parasitic egg into evolutionary process driven by the host’s rejection of
any of the hosts’ clutches, she usually removes one parasitic eggs that differ from their own. This selec-
(rarely more) egg from the clutch. From 8 to 36 h tion process leads to the development and main-
after the Cuckoo chick hatches, it instinctively tenance of host-specific Cuckoo races, or gentes in
ejects any eggs or young from the nest (Wyllie 1981). geographical areas (Chance 1940, Brooke & Davies
So the first egg to hatch in a multiply parasitized 1988, Moksnes & Røskaft 1995, Øien et al. 1995).
clutch has an advantage over any other Cuckoo Egg mimicry may also be important for female–
egg(s). The Cuckoo egg laid earlier is therefore a female competition of the Cuckoos, because in the
great threat for any Cuckoo eggs laid later. Davies case of multiple parasitism mimicry of the parasitic
and Brooke (1988) stated that egg replacement egg reduces the chance that second Cuckoos will
behaviour may be regarded as a type of competition select the parasitic egg (Davies & Brooke 1988,
between laying Cuckoo females. Brooker and Brooker Brooker & Brooker 1989, 1990). In Japan, where
(1990) supposed that egg replacement behaviour four Cuckoo species breed sympatrically, inter-
is the selective force for mimicry of the parasite’s specific competition is also an important factor for
egg in cuckoo–host relationships. Unfortunately, the evolution of host use and egg colour in parasitic
there is no significant evidence for selective removal. cuckoos (Higuchi 1989, 1998).
Davies and Brooke (1988) found a weak tendency Besides the low frequency of ejection (12%),
for selective removal if artificial eggs placed in a rejection cost to the hosts was relatively high. We
clutch experimentally were unlike the host’s eggs explain the low frequency by the high level of
in appearance. Rothstein and Robinson (1998) refer mimicry of the Cuckoo eggs, which makes it difficult
to a personal communication with H. Nakamura, to recognize the parasitic egg, so the hosts can easily

© 2002 British Ornithologists’ Union, Ibis, 144, 614–622

IBI_085.fm Page 621 Monday, September 9, 2002 11:15 AM

Cuckoo parasitism on the Great Reed Warbler 621

make mistakes. As the signal-detection model by R E F E RE N C E S

Rodríguez-Gironés and Lotem (1999) suggests, in
the case of good mimicry and high interclutch vari- Báldi, A. & Kisbenedek, T. 1999. Species-specific distribution
ability, host females must use an extended learning of reed-nesting passerine birds across reed-bed edges:
effects of spatial scale and edge type. Acta Zool. Hung. 45:
phase, but the cost of mis-imprinting is high in 97–114.
multiple parasitism. Another possible reason why Brooke, M., de L. & Davies, N.B. 1988. Egg mimicry by
hosts sometimes broke their own eggs is that during Cuckoos Cuculus canorus in relation to discrimination by
puncture ejection the host’s bill may ricochet off hosts. Nature 335: 630 – 632.
the parasitic egg and damage their own eggs (Sealy Brooker, M.G. & Brooker, L.C. 1989. The comparative breed-
ing behaviour of two sympatric species of cuckoos, Hors-
1996). Partial clutch reduction may lead to nest field’s Bronze-cuckoo Chrysococcyx basalis and the Shining
desertion (Hill & Sealy 1994). Bronze-cuckoo C. lucidus in Western Australia: a new model
for the evolution of egg morphology and host specificity in
avian brood parasites. Ibis 131: 528–547.
Rejection behaviour of the host towards Brooker, L.C. & Brooker, M.G. 1990. Why are cuckoos host
the parasitic egg specific? Oikos 57: 301– 309.
Chance, E. 1940. The Truth About the Cuckoo. London: Country
Changes in antiparasite defence of the host in time Life.
are well described by the evolutionary arms race Csörgõ, T. 1984. The Great Reed Warbler. In Haraszthy, L. (ed.)
hypothesis (Dawkins & Krebs 1979, Davies & The Birds of Hungary: 181–182. Budapest: Mezögazda Kiadó.
Brooke 1988, Davies 1999). The development of (In Hungarian.)
Csörgõ, T. 1998. The Great Reed Warbler. In Haraszthy, L. (ed.)
antiparasite defence of the host is usually delayed The Birds of Hungary : 306–307. Budapest: Mezögazda Kiadó.
(Lotem & Rothstein 1995, Rothstein & Robinson (In Hungarian.)
1998). In Japan, the Great Reed Warblers accepted Davies, N.B. 1999. Cuckoos and cowbirds versus hosts: Co-
c. 38% of the real Cuckoo eggs (Lotem et al. 1995). evolutionary lag and equilibrium. Ostrich 70: 71–79.
Our value of acceptance is higher, but there is evid- Davies, N.B. 2000. Cuckoos, Cowbirds and Other Cheats.
London: T. & A.D. Poyser.
ence that even hosts with long histories of exposure Davies, N.B. & Brooke, M. de L. 1988. Cuckoos versus Reed
to brood parasitism accept parasitic eggs (Rothstein Warblers: adaptations and counteradaptations. Anim. Behav.
1975, Davies & Brooke 1989a). It may be important 36: 262 – 284.
in our results that the ejection rate was not very high Davies, N.B. & Brooke, M. de L. 1989a. An experimental
(12%), lower than desertion (20%). Egg burial was study of co-evolution between the Cuckoo Cuculus canorus
and its hosts. I. Host egg discrimination. J. Anim. Ecol. 58:
also infrequent (2%). In total, rejections were about 207– 224.
34%. This value is somewhat lower than was obser- Davies, N.B. & Brooke, M. de L. 1989b. An experimental study
ved for the Reed Warblers in the Czech Republic, of co-evolution between the Cuckoo Cuculus canorus and its
i.e. 42% (Moksnes et al. 1993a). hosts. II. Host egg markings, chick discrimination and general
Our results show that rejection behaviour towards discussion. J. Anim. Ecol. 58: 225 – 236.
Davies, N.B. & Brooke, M. de L. 1991. Co-evolution of the
the parasitic egg was more frequent when Cuckoo Cuckoo and its hosts. Sci. Am. 264: 92–98.
eggs matched host eggs poorly (see also Moksnes Davies, N.B. & Brooke, M. de L. 1998. Cuckoo versus hosts.
1992, Moksnes et al. 1993b). This concords with the Experimental evidence for coevolution. In Rothstein, S.I. &
model developed by Davies et al. (1996), which pro- Robinson, S.K. (eds) Parasitic Birds and Their Hosts: Studies
poses that the host has to optimize egg rejection in Coevolution: 59 –79. New York: Oxford University Press.
Davies, N.B., Brooke, M. de L. & Kacelnik, A. 1996. Recogni-
behaviour relative to the degree of mimicry and level tion errors and probability of parasitism determine whether
of parasitism. Reed Warblers should accept or reject mimetic Cuckoo eggs.
Proc. R. Soc. Lond. B 263: 925 – 931.
Dawkins, R. & Krebs, J.R. 1979. Arms races between and
We are indebted to Professors E. Røskaft, S.G. Sealy,
within species. Proc. R. Soc. Lond. B 205: 489–511.
N.B. Davies, R.B. Payne and an anonymous reviewer for their
Higuchi, H. 1989. Responses of the Bush Warbler Cettia
helpful comments on an earlier draft of the manuscript. diphone to artificial eggs of Cuculus cuckoos in Japan. Ibis
Permission for the fieldwork was provided by the Duna- 131: 94 – 98.
Ipoly National Park. The study was supported by grant Higuchi, H. 1998. Host use and egg color of Japanese Cuckoos.
no. T29570 of the Hungarian Scientific Research Fund In Rothstein, S.I. & Robinson, S.K. (eds) Parasitic Birds
(OTKA) to C.M., and also by grant no. 206/00/P046 of and Their Hosts: Studies in Coevolution: 80–93. New York:
the Grant Agency of the Czech Republic to M.H. The Oxford University Press.
study was also supported by the travel grant based on the Hill, D.P. & Sealy, S.G. 1994. Desertion of nests parasitized
bilateral agreement between the Hungarian and Czech by cowbirds: have Clay-coloured Sparrows evolved an anti-
Academies of Sciences to M.H. parasite defence? Anim. Behav. 48: 1063–1070.

© 2002 British Ornithologists’ Union, Ibis, 144, 614 – 622

IBI_085.fm Page 622 Monday, September 9, 2002 11:15 AM

622 C. Moskát & M. Honza

Horváth, L. 1958. Cuculiformes. In Székessy, V. (ed.) Fauna Nakamura, H., Kubota, S. & Suzuki, R. 1998. Coevolution
Hungariae, Vol. 21./7: 9 –10. Budapest: Akadémiai Kiadó. (In between the Common Cuckoo and its major hosts in Japan.
Hungarian.) In Rothstein, S.I. & Robinson, S.K. (eds) Parasitic Birds and
Johnsgard, P.A. 1997. The Avian Brood Parasites: Deception at Their Hosts: Studies in Coevolution: 94–112. New York:
the Nest. New York: Oxford University Press. Oxford University Press.
Keve, A. 1960. Nomenclator Avium Hungariae. (An Annotated Norusis, M.J. 1990. SPSS/PC+. Advanced Statistics 4.0 for the
Checklist of the Birds of Hungary.) Budapest: Hungarian IBM PC/ XT/ AT. Chicago: SPSS Inc.
Ornithological Institute. (In Hungarian and German.) Øien, I.J., Moksnes, A. & Røskaft, E. 1995. Evolution of vari-
Keve, A. 1984. Nomenclator Avium Hungariae (An Annotated ation in egg color and marking pattern in European passer-
Checklist of the Birds of Hungary). Budapest: Akadémiai ines: adaptations in a coevolutionary arms race with the
Kiadó. (In Hungarian and German.) Cuckoo, Cuculus canorus. Behav. Ecol. 6: 166–174.
Lotem, A. & Nakamura, H. 1998. Evolutionary equilibria in Perrin de Brichambaut, J. 1997. Parasitic behaviour of the
avian brood parasitism. An alternative to the ‘arms race- European Cuckoo Cuculus canorus: Regional comparisons
evolutionary lag’ concept. In Rothstein, S.I. & Robinson, S.K. and evolution in time. Alauda 65: 167–186. (In French.)
(eds) Parasitic Birds and Their Hosts: Studies in Coevolution: Rodríguez-Gironés, M.A. & Lotem, A. 1999. How to detect a
223–235. New York: Oxford University Press. Cuckoo egg: a signal-detection theory model for recognition
Lotem, A., Nakamura, H. & Zahavi, A. 1992. Rejection of and learning. Am. Natur. 153: 633 –648.
Cuckoo eggs in relation to host age: a possible evolutionary Rothstein, S. 1975. An experimental and teleonomic investiga-
equilibrium. Behav. Ecol. 3: 128 –132. tion of avian brood parasitism. Condor 77: 250–271.
Lotem, A., Nakamura, H. & Zahavi, A. 1995. Constraints on egg Rothstein, S. & Robinson, S.K. 1998. The evolution and ecology
discrimination and cuckoo-host co-evolution. Anim. Behav. of avian brood parasitism. In Rothstein, S.I. & Robinson, S.K.
49: 1185 –1209. (eds) Parasitic Birds and Their Hosts: Studies in Coevolution:
Lotem, A. & Rothstein, S.I. 1995. Cuckoo-host coevolution: 3 – 56. New York: Oxford University Press.
from snapshots of an arms race to the documentation of Schulze-Hagen, K. 1992. Parasitism and egg losses due to
microevolution. Trends Ecol. Evol. 10: 436 – 437. the Cuckoo (Cuculus canorus) in Reed and Marsh Warblers
Magyar, G., Hadarics, T., Waliczky, Z., Schmidt, A., Nagy, T. (Acrocephalus scirpaceus, A. palustris) in central and west-
& Bankovics, A. 1998. An Annotated List of the Birds of ern Europe. J. Ornithol. 133: 237– 249.
Hungary. Budapest: Hungarian Ornithological Institute and Sealy, S.G. 1996. Evolution of host defenses against brood
Hungarian Ornithological Society. (In Hungarian and English.) parasitism: implications of puncture-ejection by a small pas-
Makatsch, W. 1964. Ornithologische Beobachtungen in Ungarn. serine. Auk 113: 346 – 355.
Zool. Abh. Mus. Tierk. Dresden. 27: 129 –172. Soler, M. 1990. Relationships between the Great Spotted
Moksnes, A. 1992. Egg recognition in Chaffinches and Bram- Cuckoo Clamator glandarius and its corvid hosts in a recently
blings. Anim. Behav. 44: 993 – 995. colonized area. Ornis Scand. 21: 212–223.
Moksnes, A. & Røskaft, E. 1987. Cuckoo host interactions in Soler, J.J. & Møller, A.P. 1996. A comparative analysis of the
Norwegian mountain areas. Ornis Scand. 18: 168 –172. evolution of variation in appearance of eggs of European
Moksnes, A. & Røskaft, E. 1995. Egg-morph and host prefer- passerines in relation to brood parasitism. Behav. Ecol. 7:
ence in the Common Cuckoo (Cuculus canorus): an analysis 89–94.
of cuckoo and host eggs from European museum collections. Soler, M., Soler, J.J., Martinez, J.G., Pérez-Contreras, T. &
J. Zool. Lond. 236: 625 – 648. Møller, A.P. 1998. Micro-evolutionary change and population
Moksnes, A., Røskaft, E., Bicik, V., Honza, M. & Øien, I.J. dynamics of a brood parasite and its primary host: the inter-
1993a. Cuckoo Cuculus canorus parasitism on Acrocepha- mittent arms race hypothesis. Oecologia 117: 381–390.
lus warblers in Southern Moravia in the Czech Republic. Southern, H.N. 1954. Mimicry in cuckoos’ eggs. In Huxley, J.,
J. Ornithol. 134: 425 – 434. Hardy, A.C. & Ford, E.B. (eds) Evolution as a Process: 219–
Moksnes, A., Røskaft, E. & Korsnes, L. 1993b. Rejection of 232. London: George, Allen & Unwin.
Cuckoo (Cuculus canorus) eggs by Meadow Pipits (Anthus StatSoft, Inc. 1995. Statistica for Windows. Tulsa: StatSoft Inc.
pratensis). Behav. Ecol. 4: 120 –127. Takasu, F. 1999. The role of theoretical models explaining the
Molnár, B. 1944. The Cuckoo in the Hungarian Plain. Aquila 51: relationships between brood parasites and their hosts. Proc.
100 –112. Int. Ornithol. Congr. 22: 3146 –3164.
Moskát, C. & Báldi, A. 1999. The importance of edge effect in Takasu, F., Kawasaki, K., Nakamura, H., Cohen, J. & Shigesada, N.
line transect censuses applied in marshland habitats. Ornis 1993. Modeling the population dynamics of a cuckoo–host
Fenn. 76: 33 – 40. association and the evolution of host defenses. Am. Nat. 142:
Moskát, C. & Fuisz, T. 1999. Reactions of Red-backed Shrikes 819 – 839.
Lanius collurio to artificial Cuckoo Cuculus canorus eggs. Varga, F. 1994. Cuckoo Observations Around the Source of the
J. Avian Biol. 30: 175 –181. River Zagyva. Salgótarján, Hungary: Published by the author.
Moskát, C. & Honza, M. 2000. Effect of nest and nest site char- Wyllie, I. 1981. The Cuckoo. London: Batsford.
acteristics on the risk of Cuckoo Cuculus canorus parasitism Yamagishi, S. & Fujioka, M. 1986. Heavy brood parasitism by
in the Great Reed Warbler Acrocephalus arundinaceus. the Common Cuckoo Cuculus canorus on the Azure-winged
Ecography 23: 335 – 341. Magpie Cyanopica cyana. Tori 34: 91–96.
Nakamura, H. 1990. Brood parasitism by the Cuckoo Cuculus
canorus in Japan and the start of new parasitism on the Azure-
winged Magpie Cyanopica cyana. Jap. J. Ornithol. 39: 1–18. Received 4 October 2000; revision accepted 17 September 2001

© 2002 British Ornithologists’ Union, Ibis, 144, 614–622