17. Biologia 56(5):

Transkript

1 17. Grim T. & Honza M. 2001: Differences in behaviour of closely related thrushes (Turdus philomelos and T. merula) to experimental 17. parasitism by the common cuckoo Cuculus canorus. Grim T. & Honza M. 2001: Differences in behaviour of closely related thrushes (Turdus philomelos and T. merula) to experimental Biologia 56(5): parasitism by the common cuckoo Cuculus canorus. Biologia 56(5):

2 Biologia, Bratislava, 56/5: , 2001 Differences in behaviour of closely related thrushes (Turdus philomelos and T. merula) to experimental parasitism by the common cuckoo Cuculus canorus Tomáš Grim 1 & Marcel Honza 2 1 Laboratory of Ornithology, Faculty of Sciences, Palacký University, Tř. Svobody 26,CZ Olomouc, Czech Republic; grim@prfnw.upol.cz 2 Institute of Vertebrate Biology, AS CR, Květná 8, CZ Brno, Czech Republic; honza@brno.cas.cz Grim, T. & Honza, M., Differences in behaviour of closely related thrushes (Turdus philomelos and T. merula) towards experimental parasitism by the common cuckoo Cuculus canorus. Biologia, Bratislava, 56: , 2001; ISSN The common cuckoo Cuculus canorus parasitizes many passerines, but some common species sympatric with the brood parasite are rarely used as hosts. Potential host species may escape brood parasitism using methods such as high rejection of cuckoo eggs or high aggressiveness towards female parasite. We tested the responses of two common species, the song thrush Turdus philomelos and blackbird T. merula, not regularly parasitised by the cuckoo, to artificial cuckoo eggs and dummies. Both species rejected model parasitic eggs (song thrush 58.3%, blackbird 66.7%). Song thrushes showed very low levels of aggression toward a stuffed dummy, while blackbirds were very aggressive. Neither species discriminated between the cuckoo and control pigeon dummies. We observed one case of intraspecific nest parasitism in the song thrush. This is probably the first documentation of intraspecific nest parasitism in this species. Both our and previously published data indicate that the rejection behaviour of Turdus species evolved as a defence against intraspecific nest parasitism. This behaviour contributes to cuckoos avoiding these potential host species. However, other nonexclusive factors (e.g. diet composition) could explain more fully why thrushes are not victimized by the cuckoo. Key words: brood parasitism, nest defence, mimicry, aggression, co-evolution, egg rejection. Introduction The common cuckoo Cuculus canorus (Linnaeus, 1758) is a brood parasite whose eggs have been found in nests of more than one hundred species of small passerines (Moksnes & Rskaft, 1995). However, only five to ten host species are parasitised regularly (Rothstein & Robinson, 1998). Other common species of open-nesting passerines sympatric with the cuckoo (e.g. Turdus, Emberiza, Carduelis) are parasitised rarely or not at all (Moksnes & Rskaft, 1995). Although brood parasitism by the cuckoo and its North-American counterpart, the brown- 549

3 headed cowbird Molothrus ater (Boddaert, 1783) is currently subject to intensive research (for reviews see Ortega, 1998; Rothstein & Robinson, 1998), few authors have tried to explain low levels of parasitism in particular host species. Davies & Brooke (1989) and Moksnes & Rskaft (1992) found that several rare cuckoo hosts (e.g. the reed bunting Emberiza schoeniclus Linnaeus, 1758 and willow warbler Phylloscopus trochilus Linnaeus, 1758) show fine egg discrimination. Peer & Bollinger (1997) reported that low synchronization between breeding cycles of the host (common grackles Quiscalus quiscula Linnaeus, 1758) and parasite (brown-headed cowbirds) is responsible for low parasitism rates in common grackles. Brown-headed cowbirds avoid parasitising eastern kingbirds Tyrannus tyrannus (Linnaeus, 1758) because their eggs would be wasted kingbirds reject almost 100% of parasitic eggs (Sealy & Bazin, 1995). Mermoz & Fernandéz (1999) explained the low frequency of parasitism in scarlet-headed blackbirds Amblyramphus holosericeus (Scopoli, 1786) by shiny cowbirds Molothrus bonariensis (Gmelin, 1789) with the fact that the scarlet-headed blackbird shows a high level of nest attentiveness and therefore the parasitic female cannot lay without being noticed and attacked by the nest owners. Several hypotheses were proposed to explain why a particular bird species is not victimized by a brood parasite including host breeding success, egg acceptance/rejection status, intensity of nest defence, host care and diet, nest type and habitat (Ortega, 1998). In this paper we test two of these hypotheses (egg rejection and nest defence) in two common passerines: the song thrush Turdus philomelos (C.L. Brehm, 1831), and blackbird T. merula (Linnaeus, 1758). Observations of parasitism in the two species are extremely rare. Lack (1963) reported only 3 parasitized nests among 22,656 blackbird nests in England. Moksnes & Rskaft (1995) found only 11 and 21 cuckoo eggs laid in the nests of the two respective species (n = 11, 870 clutches of European passerines held in museum egg-collections). In addition, none of these cuckoo eggs matched the eggs of the song thrush and blackbird in appearance. These data indicate that neither species are regularly used as fosterers by the common cuckoo. Kleven et al. (1999) found that the size of the host had a significant effect on the growth of cuckoo nestlings cuckoo chicks cared for by larger host species, such as the great reed warbler Acrocephalus arundinaceus (Linnaeus, 1758), were significantly heavier at fledging than nestlings raised by a smaller host, the reed warbler Acrocephalus scirpaceus (Hermann, 1804). In the light of this finding it is interesting that no Turdus species is regularly used as a host by the cuckoo despite the body size of these potential hosts. Therefore we made an attempt to explain why the song thrush and blackbird are not parasitised by the common cuckoo. Responses of the two species to parasitic eggs were tested only in Great Britain (Davies & Brooke, 1989) and Norway (Moksnes et al., 1990). We focused on nest defence behaviour and parasite recognition abilities of the song thrush and blackbird, which have not been studied, to date. Methods The study was conducted in a deciduous forest nearby Dolní Bojanovice village (48 52 N, E), about 60 km SE of Brno, South Moravia, Czech Republic. Data were collected from 25 April to 30 June 2000 and For the egg experiments we used natural Chinese quail Coturnix chinensis (Linnaeus, 1766) eggs painted blue to mimic eggs of the cuckoo gent parasitising the redstart Phoenicurus phoenicurus (Linnaeus, 1758) (Moksnes & Rskaft, 1995). We introduced experimental eggs to host nests at the egg-laying or early incubation stages (from day 2 till day 6; day 0 is the day when the first egg was laid; the clutch size in both species tested is 3 5; Cramp, 1988). To minimize disturbance to the hosts, we did not remove any host eggs (as female cuckoos do) because the experimental removal of one host egg has no effect on the rejection rates of model eggs (Davies &Brooke, 1989). We experimentally parasitised 17 song thrush and 8 blackbird nests. The egg was considered accepted if it remained in the nest for six days. This criterion is used because almost all observed rejections in different host species appeared before the sixth day after the egg was introduced, see e.g. Moksnes et al. (1990); Davies &Brooke (1989) reported that most rejections took place within 3 days after clutch completion. Nest defence by hosts was tested with stuffed dummies of the common cuckoo and the feral pigeon Columba livia f. domestica (Gmelin, 1789) as a control. We chose the pigeon as a control species because it is about the same size and shape as a cuckoo but provides no threat to the either species tested. The experimental design followed the standard procedure suggested by Sealy et al. (1998). First, one of dummies (in a life-like position) was attached to a branch about 0.5 m from a focal nest. The head of the dummy was directed to the nest. After the first parent appeared near the nest and became aware of the dummy, reactions of nest owners were observed for 5 min from a hide set up 20 m from the focal nest. Presentation of the second mount at the same place was separated by a 30 min interval to avoid habituation or carry-over aggression (Sealy et al., 1998). The order in which models 550

4 Table 1. The outcome of the observed nesting attempts and results of model egg experiments at song thrush and blackbird nests. There were no significant interspecific differences in all measured outcomes (chi-square and Fisher s exact probabilities tests, all n.s.). Outcome Song thrush (n = 55) Blackbird (n = 25) Control Experiment Control Experiment Predated 27/38 5/17 12/17 2/8 Experimental egg accepted 5/12 2/6 Experimental egg ejected 4/12 3/6 Nest deserted 0/38 3/12 0/17 1/6 Note: some of successful experimental nests were predated after the egg experiments finished, therefore the breeding success is actually lower than is shown in the table as noted in the results. were presented was randomized. All experiments with stuffed dummies were performed by one author (T.G.) to avoid possible observer bias. The number of nests tested was 15 for the song thrush and 6 for the blackbird. In the song thrush 6 nests were tested at the egg stage and 9 at the nestling stage, while all blackbird nests were tested at the nestling stage. The intensity of nest defence varied from quiet watching of the nest from a distance to vigorous mobbing of the dummy. We adjusted our categorization of host responses to natural variation in responses observed during our experiments. Blackbirds usually quickly attacked the mount directly (the mount was immediately removed to avoid its destruction), so we did not quantify host reactions as a number of particular behaviours per 5 min of observation. Instead, we quantified the behaviour of both blackbirds and song thrushes on relative subjective scales (see e.g. Mermoz & Fernández, 1999). We recorded the behavioural variables suggested by Sealy et al. (1998) to enable interspecific comparisons (see e.g. Gill et al., 1997). We measured the delay in arrival of nest owners from the moment the dummy was attached near the focal nest (latency of response) in minutes, which should reflect the general level of nest attentiveness. Latency in response can be taken as a rough measure of time the parents spent at the nest which is crucial with respect to interaction with the brood parasitic cuckoo which lays extremely quickly to avoid host attacks. In addition we quantified several parameters of host reactions: vocalizations (0 to 3, i.e. from no vocalizations to very strong and permanent vocalizations), contact attacks (0 = none, 1 = one or more), overall level of response (0 = no response, i.e. silent watching of a dummy, 1 = few vocalizations, bird(s) usually more than 10 m from the dummy, 2 = more vocalizations, bird usually less than 10 m from the dummy, 3 = strong vocalizations and close passes, 4 = strong vocalizations and contact attacks). These parameters should reflect the risks taken by birds when defending their nests. Each nest was tested only once to avoid pseudoreplication. Results Egg discrimination Altogether we observed 66 song thrush and 31 blackbird nesting attempts (11 song thrush and 6 blackbird nests were not followed to fledging and their final fate is unknown). The proportions of successful breeding attempts (nests with a known fate) of song thrushes (41.8%, n = 55) and blackbirds (44.0%, n = 25) were not statistically significantly different (χ 2 = , d.f. = 1, P = 0.85). The main reason for nesting failure was the predation of clutch (Tab. 1). We experimentally parasitised 17 song thrush and 8 blackbird nests and only 12 and 6 artificially parasitised nests survived the 6 days acceptance/rejection criterion, for each species respectively. Song thrushes rejected 58.3% and blackbirds 66.7% of parasitic eggs (Tab. 1). Rejection methods used by both species were ejection of the parasite egg and desertion of the parasitised nest (Tab. 1). Blackbirds rejected parasitic eggs more quickly than song thrushes (medians: 1 and 4 days; Mann-Whitney test, Z = 2.038, P = 0.040, n = 7). At three predated song thrush nests the parasitic eggs were accepted for 1, 2 and 4 days before predation of the clutch (the remaining 2 nests were predated before the first check). At predated blackbird nests the parasitic eggs remained in the nests for 1 and 3 days before being predated. One song thrush and one blackbird deserted their experimentally parasitised nests. Both desertions occurred less than three days after the parasitic egg was added. This behaviour is probably a response to parasitism as no control nests (which were regularly checked but not used in experiments) were deserted. We observed one case of intraspecific nest parasitism in the song thrush. The parasitic egg 551

5 Table 2. Intraspecific comparison of intensity of song thrush and blackbird responses to cuckoo and pigeon dummies. Values are medians. There were no differences in responses to the two types of dummies in both host species in any of the measured behavioural parameters (Wilcoxon matched pairs tests, all n.s.). Latency of response was measured in minutes, other parameters measured on ordinal scale (see Methods). Parameter Song thrush (n = 15) Blackbird (n =6) Cuckoo Pigeon Cuckoo Pigeon Latency of response Vocalizations Contacts Overall level of nest defence Table 3. Interspecific comparison of intensity of song thrush and blackbird responses to stuffed dummies. Values are medians. The results of Mann-Whitney tests are shown. Latency of response was measured in minutes, other parameters measured on ordinal scale (see Methods). Differences which remained significant after sequential Bonferroni test (Rice, 1989) are indicated with an asterisk (P < 0.05). Parameter Song thrush Blackbird Z P (n = 15) (n =6) Latency of response n.s. Vocalizations * Contacts * Overall level of nest defence * was laid seven days after the host clutch completion (the appearance of eggs both before and after the host s laying period is considered as an indication of brood parasitism, see e.g. Ringsby et al., 1993). Five days later the parasitic egg was destroyed due to partial predation of the clutch. As far as we know, this is the first observation of intraspecific brood parasitism in the song thrush. Nest defence We found no differences in song thrush responses to mounts between the egg (n = 6) and nestling (n = 9) stage in latency of response, vocalizations, contacts and overall level of nest defence (Mann- Whitney tests, P > 0.05 in all cases). Therefore the data were pooled. Neither song thrushes or blackbirds distinguished between the cuckoo and the control species there were no differences between the responses of either species to either type of dummy in any of the measured behavioural parameters (Wilcoxon matched pairs tests, all n.s.; Tab. 2). There was no relationship between the latency of response and overall level of nest defence (song thrush: r s = 0.355, P = 0.194, n = 15; blackbird: r s = 0.655, P = 0.158, n = 6). The number of attacking individuals had no effect on the overall level of aggression in the song thrush (Mann-Whitney test, Z = 1.687, P = 0.092, n = 15). Interestingly, during five experiments on blackbirds the overall level of aggression was highest (level 4) and both parents attacked a dummy together in all five cases. In one experiment only the female was present showing no aggression at all. The male was never observed at this nest. This female also accepted the experimental blue egg, which remained in the nest for 15 days and was not ejected even after nestlings hatched. There were no significant differences in the latency of response between species (Tab. 3). However, the number of individuals performing nest defence behaviour differed between species in the song thrush usually only one parent responded, but in the blackbird both male and female attacked dummies (Fisher s exact probabilities test, P = 0.046). Moreover, blackbirds vocalized more strongly, attacked the mount directly with higher a probability and showed higher overall level of nest defence than song thrushes (Tab. 3). Song thrushes responded very weakly to dummies. They usually stayed more than 10 meters from the nest uttering very few vocalizations about 20 times per 5 min observation period. Comparable data for blackbirds are lacking because blackbirds uttered alarm calls very quickly and almost continu- 552

6 ously, moreover, the nest owners immediately attacked the mount directly and therefore the experiments were stopped earlier thus making sensible interspecific comparison of call rates impossible (see Methods). Only two studied song thrushes responded extremely aggressively and one even showed redirected aggression towards its own nest (this unusual observation is described in detail by Grim, 2000). On the other hand, blackbirds behaved very aggressively they continually and intensively vocalized and mobbed dummies vigorously. In conclusion, the low level of song thrush nest defence would present almost no risk to a female cuckoo. On the other hand, the extremely vigorous nest defence behaviour shown by blackbirds would threaten a female cuckoo s life. Three song thrush and three blackbird nests were tested with both parasitic eggs and dummies. Dummy experiments were performed after the egg experiments finished. Parents in four nests accepted parasitic eggs. It might be expected that acceptors are naive breeders (Lotem et al., 1992) that would also defend their nests poorly. However, intensity of response to mounts was at levels 1 and 2 for the two song thrush nests and 0 and 4 for the two blackbird nests. Parasitic eggs were rejected at two other nests. Intensity of response to dummies was at level 1 for one song thrush nest and 4 for the blackbird nest. Our data sets are too small to be analysed statistically, however, they show some inconsistency in the host responses to eggs versus dummies. Discussion Egg discrimination The breeding success of the study species found during our study seasons (song thrush: 41.8%, blackbird: 44.0%) is similar to that found in previous studies (e.g. Osborne & Osborne, 1980; Hatchwell et al., 1996). It is within the range of breeding success of bird species commonly used as hosts by the cuckoo. Therefore the level of breeding success in the song thrush and blackbird probably has no effect on the avoidance of these species by the cuckoo. The results of the egg experiments are similar to those reported by Davies & Brooke (1989). They found that song thrushes rejected 27.3% and blackbirds 59.1% of non-mimetic redstart type eggs (i.e. the same type as we used in our study). Overall levels of rejection of several types of alien eggs by song thrushes and blackbirds were 58.5% and 61.8% respectively (Davies & Brooke, 1989). These values are higher than the rejection rates of commonly parasitised hosts (Davies & Brooke, 1989). In a study by Moksnes et al. (1990) 20% of song thrush pairs rejected a model cuckoo egg and two blackbirds also rejected them. Four experimentally parasitised nests in our study were deserted. We consider this as a method of rejection because Davies & Brooke (1989) showed that desertion can serve as a method of rejection hosts tested by them deserted nests parasitised with non-mimetic eggs more frequently then nests parasitised with mimetic eggs. The cuckoo probably does not use some hosts (e.g. the reed bunting and spotted flycatcher Muscicapa striata Pallas 1764) because they show high rejection rates of parasitic eggs (Davies & Brooke, 1989). However, great reed warblers Acrocephalus arundinaceus in Hungary reject 39% of cuckoo eggs and are still heavily parasitised (Moskát & Honza, 2000). Other commonly used hosts are also rejecters (Davies & Brooke, 1989). Our data and the results of Davies & Brooke (1989) and Moksnes et al. (1990) show that both the song thrush and blackbird can recognize and reject parasitic eggs. However, these studies also indicate that the egg rejection behaviour of the song thrush and blackbird cannot alone explain the low level of parasitism in these species. We documented one case of intraspecific nest parasitism in the song thrush. We found no other report on intraspecific nest parasitism by the song thrush in the literature. However, Grendstad et al. (1999) reported that one of the studied redwing Turdus iliacus (Linnaeus, 1766) nests was parasitized by the closely related fieldfare Turdus pilaris (Linnaeus, 1758). Interestingly, Ringsby et al. (1993) detected a relatively high intraspecific parasitism rate in fieldfares (11.5%). Redwings discriminate against intraspecific eggs introduced to their nests and show stronger aggression toward conspecific compared to the female cuckoo. This strongly suggests that redwings evolved this behaviour as a defence against intraspecific and not interspecific nest parasitism (Grendstad et al., 1999). Nest defence Blackbirds displayed aggressive nest defence behaviour and readily attacked dummies with contact. The blackbird is generally very aggressive and intraspecific fights can lead to death (Cramp, 1988). Its parental anti-predatory behaviour is also reported to be very intense (Cramp, 1988). The song thrush is similarly described as aggressive by 553

7 Cramp (1988), however, in our study only two of 15 tested pairs attacked the dummy directly and the overall level of aggressiveness was very low compared to the blackbird. We found that both studied species did not discriminate between the cuckoo and control pigeon mounts. This result indicates that the behaviour of song thrushes and blackbirds was a result of generalized nest defence and not a response to a specific threat provided by the parasite. However, Duckworth (1991) found that reed warblers Acrocephalus scirpaceus (commonly parasitised by the cuckoo throughout their range) can differentiate cuckoos from similar sparrowhawks Accipiter nisus (Linnaeus, 1758). The reed warbler s response was more aggressive to a cuckoo than to a sparrowhawk during incubation, but the reaction to the cuckoo disappeared after fledging although the parents still responded strongly to a sparrowhawk at that stage. Therefore our results indicate that there probably was no coevolutionary arms-race between the cuckoo and Turdus species. This hypothesis is supported by the fact that none of the cuckoo eggs found in song thrush and blackbird nests were mimicking host species eggs (Moksnes & Rskaft, 1995). Fieldfares show aggression towards a hooded crow Corvus corone (Linnaeus, 1758) dummy (Meilvang et al., 1997) but they do not consider the cuckoo to be a potential threat (Moksnes & Rskaft, 1988). Grendstad et al. (1999) also concluded that redwings evolved alien egg rejection as a defence against intraspecific brood parasitism. We observed one case of intraspecific nest parasitism in the song thrush which also gives support to the hypothesis that Turdus species have not been in the co-evolutionary arms race with the cuckoo. The overall level of nest defence by song thrushes was very low. Thus, the song thrush would not prevent a female cuckoo trying to lay an egg in a song thrush nest. However, the very high intensity of blackbird aggression suggests that the cuckoo would risk serious injury if attacked at a blackbird nest during the parasitism act. Molnár (1944) reported several observations of dead female cuckoos under great reed warbler nests. Females were evidently killed by the nest owners. The blackbird is a much larger bird than the great reed warbler (100 g vs 30 g), thus, a female cuckoo would probably risk much more at a blackbird than warbler nest. The poor level of nest defence shown by the song thrush and strong aggression shown by blackbirds in our study are consistent with the results of dummy experiments performed by V. Bièík in Czech Republic (unpublished data; in verb). Mermoz & Fernández (1999) described a low rate of parasitism in scarlet-headed blackbirds Amblyramphus holosericeus with a non-specific life-history trait, i.e. high levels of nest attentiveness combined with aggressive host behaviour against all intruders. Scarlet-headed blackbirds do not recognize the shiny cowbird Molothrus bonariensis as a specific threat, but 98% of time the nest is guarded by at least one parent. The high level of nest attentiveness cannot play an important role in the cuckoo s avoidance of blackbirds and song thrushes as these species arrive at their nest with long delays (Tabs 1, 2), i.e. nest attentiveness is very poor and the probability of encountering a brood parasite at the host nest is consequently minimal. We tested whether the cuckoo avoids parasitising song thrush and blackbird nests because these species have low breeding success, strong egg rejection behaviour or intensive nest defence. We found that both species discriminated parasitic non-mimetic eggs. Moreover, blackbirds, although they do not recognize the female cuckoo as a specific threat (they attacked the pigeon with the same vigour), have a high level of generalized nest defence that could threaten the female cuckoo. However, both these behaviours are shared with commonly parasitised hosts (e.g. Phylloscopus warblers or the great reed warbler are also strongly aggressive to the cuckoo). Thus, these factors cannot alone explain why cuckoos avoid parasitising them. Ortega (1998) presents 12 hypotheses trying to explain the absence of parasitism in some passerine species. It is unlikely that most of these could explain the avoidance of Turdus species specifically (e.g. short host incubation period, insufficient amount of parental care, unsuitable habitat, well concealed hosts nests). However, other hypotheses are more likely to explain the absence of parasitism in thrushes. Moksnes & Rskaft (1988) hypothesized that Turdus nests could be too deep for a cuckoo nestling to evict host eggs or nestlings. However, Moksnes et al. (1990) reported an observation of a four-day old cuckoo chick ejecting a fieldfare nestling weighing 10.0 g. Turdus nestlings grow very quickly (Cramp, 1988), thus cuckoo nestling could have big problems competing with host chicks. In addition, cuckoos and cowbirds can live only on an insectivorous diet. Their breeding success in nests of granivorous birds is very low diet composition sufficiently explains the absence of parasitism in 554

8 e.g. greenfinch Carduelis chloris (Linnaeus, 1758), linnetc. cannabina (Linnaeus, 1758) and bullfinch Pyrrhula pyrrhula (Linnaeus, 1758) (Davies & Brooke, 1989). The song thrush and blackbird feed their nestlings mainly on molluscs and earthworms respectively (Cramp, 1988). These food items are probably indigestible for the cuckoo nestling that must be brought up on insects. However, this hypothesis could be rigorously tested only by cross-fostering experiments, i.e. transferring nestling cuckoos to Turdus nests. In a preliminary experiment it was found that a cuckoo nestling did not survive in a blackbird nest. In conclusion, our results suggest that the intensity of antiparasitic behaviour shown by song thrushes probably has a low effect on brood parasite choice. However, strong generalized nest defence by blackbirds probably constrains the use of this species by the cuckoo, although the low level of nest attentiveness in this species reduces the benefits of strong nest defence. Nevertheless, other hypotheses (food, nest type) need testing before decisive conclusions can be drawn. Our observations, coupled with previous findings, support the hypothesis that Turdus species have evolved rejection behaviour against intraspecific nest parasitism. Acknowledgements We are grateful to M. Èapek, V. Mrlík and P. Procházka for their help during the field work. Suggestions by two anonymous referees substantially improved the MS. The research was supported by Palacký University internal grant (no ) operated to T. Grim and grant 206/00/P046 operated to M. Honza. References Cramp, S. (ed.) The Birds of the Western Palearctic. Vol. V. Oxford University Press, Oxford & New York, 1063 pp. Davies, N. B. & Brooke, M. L An experimental study of co-evolution between the cuckoo, Cuculus canorus, anditshosts.i.hostseggdiscrimination. J. Anim. Ecol. 58: Duckworth, J. W Responses of breeding reed warblers Acrocephalus scirpaceus to mounts of sparrowhawk Accipiter nisus, cuckoo Cuculus canorus and jay Garrulus glandarius. Ibis 133: Gill, S. A., Grieef, P. M., Staib, L. M. & Sealy, S. G Does nest defence deter or facilitate cowbird parasitism? A test of the nesting-cue hypothesis. Ethology 103: Grendstad, L. C., Moksnes, A. & Rskaft, E Do strategies against conspecific brood parasitism occur in redwings Turdus iliacus? Ardea 87: Grim, T An interesting observation of redirected activity in the song thrush (Turdus philomelos). Sylvia 36(2): Hatchwell, B. J., Chamberlain, D. E. & Perrins, C. M The reproductive success of blackbirds Turdus merula in relation to habitat structure and choice of nest site. Ibis 138: Kleven, O., Moksnes, A., Rskaft, E. & Honza, M Host species affects the growth rate of cuckoo (Cuculus canorus) chicks. Behav. Ecol. Sociobiol. 47: Lack, D Cuckoo hosts in England. Bird Study 10: Lotem, A., Nakamura, H. & Zahavi, A Rejection of cuckoo eggs in relation to host age: a possible evolutionary equilibrium. Behav. Ecol. 3: Meilvang, D., Moksnes, A. & Rskaft, E Nest predation, nesting characteristics and nest defence behaviour of fieldfares and redwings. J. Avian Biol. 28: Mermoz, M. E. & Fernández, G. J Low frequency of shiny cowbird parasitism on scarletheaded blackbirds: anti-parasite adaptations or nonspecific host life-history traits? J. Avian Biol. 30: Moksnes, A. & Rskaft, E Response of fieldfares Turdus pilaris and bramblings Fringilla montifringilla to experimental parasitism by the cuckoo Cuculus canorus. Ibis130: Moksnes, A. & Rskaft, E Responses of some rare cuckoo hosts to mimetic cuckoo eggs and to foreign conspecific eggs. Ornis. Scand. 23: Moksnes, A. & Rskaft, E., Egg-morphs and host preference in the common cuckoo (Cuculus canorus): an analysis of cuckoo and host eggs from European museum collections. J. Zool. 236: Moksnes, A., Rskaft, E., Braa, A. T., Korsnes, L., Lampe, H. M. & Pedersen, H. C Behavioural responses of potential hosts towards artificial cuckoo eggs and dummies. Behaviour 116: Molnár, B Cuckoo in the Hungarian Plain. Aquila 51: Moskát, C.& Honza, M Effect of nest and nest site characteristics on the risk of cuckoo Cuculus canorus parasitism in the great reed warbler Acrocephalus arundinaceus. Ecography 23: Ortega, C Cowbirds and other brood parasites. The University of Arizona Press, Tucson, 374 pp. Osborne, P. & Osborne, L The contribution of nest site characteristics to breeding-success 555

9 among blackbirds Turdus merula. Ibis 122: Peer, B. D. & Bollinger, E. K Explanations for the infrequent cowbird parasitism on common grackles. Condor 99: Rice, W. R Analyzing tables of statistical tests. Evolution 43: Ringsby, T. H., Moksnes, A., Rskaft, E. & Lerkelund, H. E Do conspecific brood parasitism and antiparasite strategies occur in fieldfares Turdus pilaris? Fauna norv., Ser. C, Cinclus 16: Rothstein, S. I. & Robinson, S. K. (eds) Parasitic birds and their hosts. Studies in coevolution. Oxford University Press, New York & Oxford, 444 pp. Sealy, S. G. & Bazin, R. C Low-frequency of observed cowbird parasitism on eastern kingbirds host rejection, effective nest defense, or parasite avoidance? Behav. Ecol. 6: Sealy, S. G., Neudorf, D. L., Hobson, K. A. & Gill, S. A Nest defence by potentional hosts of the brown-headed cowbird: methodological approaches, benefits of nest defense, and coevolution. pp In: Rothstein, S. I. & Robinson, S. K. (eds) Parasitic birds and their hosts, Studies in coevolution, Oxford University Press, New York & Oxford. Received November 14, 2000 Accepted June 6,

10 18. Honza M., Grim T., Čapek M., Moksnes A. & Røskaft E. 2004: Nest defence, enemy recognition and nest inspection behaviour 18. of experimentally parasitised reed warblers Acrocephalus scirpaceus. Honza M., Grim T., Čapek M., Moksnes A. & Røskaft E. 2004: Nest defence, enemy recognition and nest inspection Bird behaviour Study 51(3): of experimentally parasitised reed warblers Acrocephalus scirpaceus. Bird Study 51(3):

11 Bird Study (2004) 51, Nest defence, enemy recognition and nest inspection behaviour of experimentally parasitized Reed Warblers Acrocephalus scirpaceus MARCEL HONZA 1 *, TOMÁS GRIM 2, MIROSLAV CAPEK JR 1, ARNE MOKSNES 3 and EIVIN RØSKAFT 3 1 Department of Avian Ecology, Institute of Vertebrate Biology, Academy of Sciences of the Czech Republic, Kvetná 8, Brno, Czech Republic, 2 Laboratory of Ornithology, Palacky University, Tr. Svobody 26, Olomouc, Czech Republic and 3 Department of Biology, Norwegian University of Science and Technology, NTNU, N-7491 Trondheim, Norway Capsule Reed Warblers in a regularly parasitized population do not recognize Cuckoo Cuculus canorus as a special enemy and do not change their behaviour at nest immediately after being parasitized. Aims To assess if an intruder near the nest influences the behaviour of the Cuckoo host. Methods Host responses to Cuckoo, control Pigeon dummies and human intruder were observed. Host behaviour at 71 nests was video-recorded for 30 minutes at four experimental groups of nests: Cuckoo dummy, Cuckoo dummy + Cuckoo egg, Pigeon dummy, human intruder. Results Reed Warblers did not respond differently to the Cuckoo and the control species. The experimental procedure had no significant effect on the behaviour of hosts during the study period. We were unable to find any differences in the time spent at the nest, clutch inspection behaviour and nest defence behaviour between morning and afternoon experimental groups. Our results do not support the hypothesis that afternoon laying by the Cuckoo is maintained by a selection pressure from the host. We observed no ejection or egg-pecking during the 30-min period after the experimental parasitism. Conclusions Low aggression and non-specificity of host responses in our study area are in line with the fact that the Reed Warbler is an intermediate rejecter of Cuckoo eggs as expected from the spatial habitat structure hypothesis. One important factor affecting reproductive success and consequently nest defence behaviour in many passerine birds is brood parasitism (Rothstein 1990), because the successful act of parasitism often reduces host fitness dramatically. Clearly, the best protection for a host against parasitism is to avoid being parasitized. Hosts can avoid parasitism by breeding in safe sites (Alvarez 1993, Øien et al. 1996, Honza et al. 1998, Moskát & Honza 2000, Clarke et al. 2001) or by a vigorous nest defence (Moksnes et al. 1990, Sealy et al. 1998, Grim & Honza 2001, Røskaft et al. 2002b). There is a huge interspecific variation in defensive behaviour of potential host species against territory intruders. Some host species discriminate between the parasite and control species that pose no threat to them (Burgham & Picman 1989, Moksnes et al. 1993a) and some hosts also respond differently to brood *Correspondence author. honza@brno.cas.cz parasites and predators (Duckworth 1991, Gill & Sealy 1996). Cuckoo Cuculus canorus hosts that reject at very high frequencies show little variation in their defence behaviour (Øien et al. 1999). Variation in the aggressive response of a host to the parasite is influenced by various factors, e.g. by an occurrence of Cuckoos in the particular locality (hosts breeding in sympatry with the parasite are more aggressive than those breeding in allopatry; Røskaft et al. 2002b). Habitat selection by the host also plays an important role host species breeding near trees only (Cuckoo vantage points) are more aggressive against a Cuckoo dummy and reject parasitic eggs at higher rates than host species breeding both near and further away from trees (Røskaft et al. 2002b, 2002c) probably because of gene flow between parasitized (near trees) and unparasitized (farther away from trees) populations. In the case where a host is not successful in deterring 2004 British Trust for Ornithology

12 Behaviour at nest of parasitized Reed Warbler 257 a brood parasite, it is adaptive to recognize and reject the parasitic egg (Rothstein 1990). There are great differences in reactions of various hosts towards parasitic eggs. Cuckoo hosts normally exhibit some delay in their response towards the parasitic egg (Davies & Brooke 1988, Moksnes et al. 1990, Amundsen et al. 2002), but Moksnes et al. (1994) and Soler et al. (2003) documented relatively short times to ejection in Chaffinches Fringilla coelebs, Blackcaps Sylvia atricapilla, Sub-Alpine Warblers S. cantillans and Blackbirds Turdus merula. Only a few studies have paid any attention to host behaviour immediately after the act of experimental Cuckoo parasitism: Moksnes et al. (1994) observed rejection behaviour of Chaffinches and Blackcaps and Martín-Vivaldi et al. (2002) video-recorded behaviour of three potential Cuckoo hosts to determine the effort needed to puncture experimentally added parasitic eggs (see also Soler et al. 2003). To our knowledge only one study has focused on host clutch inspection behaviour after the presentation of a stuffed Cuckoo dummy (Moksnes et al. 1993a). On the other hand several studies have been conducted on hosts of Brown-headed Cowbird Molothrus ater immediately after the act of parasitism (Rothstein 1977, Briskie & Sealy 1987, Sealy & Neudorf 1995, Sealy 1996, Sealy & Lorenzana 1998). Most of these studies did not, however, examine host behaviour in detail. More importantly, in all these studies (except Moksnes et al. 1993a) only experimentally parasitized nests were observed and there were no observations of unparasitized control nests. This makes a general interpretation of the results difficult because of problems of separating specific responses to parasitism and general nest defence behaviour. The Reed Warbler Acrocephalus scirpaceus is one of the most frequently used Cuckoo hosts in Europe (Moksnes & Røskaft 1995). However, the Reed Warbler is not very aggressive towards the Cuckoo (Duckworth 1991, Lindholm & Thomas 2000, Røskaft et al. 2002a, 2002b). We investigated whether host behaviour at the nest is influenced by the type of intruder (parasitic Cuckoo versus non-threatening intruder) during the presence of the intruder and also after it leaves the vicinity of the nest. As a more detailed knowledge of changes in all aspects of host behaviour during the course of a day is essential to understand parasitic adaptations of the Cuckoo (see also Moksnes et al. 2000) we performed experiments both in the morning and in the afternoon. We report behaviour of Reed Warblers in detail with respect to the time of day and the kind of intruder, immediately after nests have been tested experimentally. We made the following predictions. (1) If hosts can distinguish between their enemies and non-enemies, they should respond with higher intensity of nest defence to a brood parasite than to control dummies of a non-threatening species. Nest defence could be performed either as aggression (see Moksnes et al. 1990) or sitting in the nest to prevent parasites access to the nest (see Hobson & Sealy 1989). We predicted that Reed Warblers would show a higher level of nest defence towards a Cuckoo dummy than to a nonthreatening species. (2) The presence of a stuffed Cuckoo dummy in close proximity to the nest is reported to facilitate host egg discrimination behaviour (Davies & Brooke 1988, Moksnes & Røskaft 1989, Moksnes et al. 2000), i.e. a Cuckoo dummy increases the likelihood that a host will reject a Cuckoo egg. We therefore predicted that hosts should show a higher intensity of clutch inspection behaviour after being faced with a Cuckoo dummy compared with a nonthreatening control species. (3) The Cuckoo, unlike its hosts, lays its eggs in the afternoon (Wyllie 1981; but see Honza et al. 2002). Davies & Brooke (1988) suggested that this behaviour has evolved because the Cuckoo has a lower probability of encountering hosts during the parasitism act in the afternoon. Such behaviour should be adaptive because a physical presence of the host at its nest can serve as an effective defence against successful parasitism (Hobson & Sealy 1989; however, Moksnes et al found no support for this hypothesis). Alternative explanations for the unique laying pattern of the Cuckoo are that hosts spend more time inspecting their clutches in the morning than in the afternoon (Davies & Brooke 1988) or they are more aggressive in the morning. MATERIALS AND METHODS Study area and fieldwork The study sites were two pond systems in the southeastern part of the Czech Republic near Lednice (48 48 N, E) and Luzice (48 51 N, E). The two systems are 25 km apart. We searched systematically for Reed Warbler nests in the littoral vegetation during the breeding periods between 15 May and 30 June in 1997 and We located nests in vegetation consisting mostly of Common Reed Phragmites australis and to a lesser extent of Reedmace Typha angustifolia surrounding the ponds. The fish 2004 British Trust for Ornithology, Bird Study, 51,

13 258 M. Honza et al. ponds are situated in flat agricultural lowland landscape and are mostly surrounded by deciduous woods (see Hudec 1975 and Moksnes et al. 1993b, for more detailed descriptions of the study area). The frequency of Cuckoo parasitism in Reed Warbler nests in our study areas is 15.0% (Øien et al. 1998) and Reed Warblers reject 37.5% of parasitic eggs (29.9% by desertion, 7.6% by ejection; Øien et al. 1998). Experimental procedure and video recordings We conducted experiments on nests during the egglaying period (after the host had laid three or four eggs) to mimic a natural Cuckoo laying. A total of 71 nests was tested out of which 35 were tested during the morning (05:00 08:00 hours Central European Time, CET) and 36 during the afternoon (15:00 20:00 hours CET). To standardize our procedure, experiments were not conducted during rainy or windy days. We established four experimental groups. (1) Cuckoo dummy group (n = 30). The Cuckoo dummy was presented 1 m from the nest. (2) Cuckoo dummy + Cuckoo egg group (n = 11). After the presentation of the Cuckoo dummy, the nest was experimentally parasitized with a real Cuckoo egg from the study area resembling the eggs of the Great Reed Warbler Acrocephalus arundinaceus (the egg was removed after the video recording was finished, see later). (3) Pigeon Columba livia f. domestica group (n = 16). The nest owner s reactions were tested with a stuffed Pigeon as a control non-threatening species. (4) Human intruder group (n = 14). One researcher (M.H.) as a control only visited the nest. Each nest was tested only once (i.e. one type of experiment) to avoid pseudoreplication. To reduce the number of confounding variables (size, plumage colour, shape of a bird), we used Pigeon as the control species because its size and overall colour (grey) is similar to Cuckoo and it poses no threat to Reed Warblers (see also Sealy et al. 1998). Some researchers have used control species that are familiar to tested hosts (Moksnes & Røskaft 1989, Grim & Honza 2001), while others performed experiments with model species not occurring in the study area (Bazin & Sealy 1993). The Pigeon lives in sympatry with the Cuckoo in the nearest vicinity of reedbeds in our study area, therefore Reed Warblers probably have had a chance to gain prior experience with this species. Nevertheless, Sealy et al. (1998) suggested that prior experience (or its absence) with a control species should have no effect. The human intruder experiment was used to ascertain if responses to dummies are specific reactions to bird intruders or a general response to any form of nest disturbance. A Sony video camera (CCD-TR 660E Hi 8) was mounted on a stand and placed at a distance of 3 4 m and levelled 0.4 m above the nests. The camera was powered with a 12-volt car battery and provided with a shelter painted dark green to be inconspicuous. Since all the nests were situated above water, the battery was placed on the shore and connected to the camera with a cable. Reed Warblers are known to be tolerant to human presence (Lindholm & Thomas 2000). They start to feed their broods almost immediately after cameras or hides are placed near their nests (latency to start of feeding for cameras = 5.00 ± 0.99 min (mean ± sd), n = 5; latency to start of feeding for hides = 3.08 ± 2.10 min, n = 13) and the age of brood does not influence this latency (r = 0.04, P = 0.90, n = 13; T. Grim, M. Honza, B. Matysioková & K. Voslajerová unpubl. data). Furthermore, feeding frequencies observed at the nest immediately after placing cameras or hides are the same as those reported by Kilner et al. (1999) who left cameras near the nest for several hours for birds to habituate. Thus, the effect of cameras on Reed Warbler behaviour is undetectable and it is highly unlikely that it could confound our results (moreover, we were interested in among-group differences and all groups were treated identically with cameras). Also the behaviour of the closely related Great Reed Warbler is not significantly affected by a video camera near the nest (M. Honza & C. Moskát, unpubl. data). Finally, responses of Reed Warblers towards a stuffed Cuckoo in the current study were almost the same as those in a previous study (Røskaft et al. 2002a) where host behaviour was directly observed without cameras. Nevertheless, we left the place for 1 hour allowing the birds to habituate to the set-up. After 1 hour, we presented the Cuckoo or stuffed Pigeon dummy mounted on a wooden pole 1 m from the nest. The dummies were in perching position and at the same height as the focal nest. The behaviour of the nest owners was observed during a 5-min period from a distance of 10 m. We recorded latency time to arrival, latency time to alarm calling from the first arrival and time spent within 1 m of the Cuckoo/Pigeon dummy. The overall level of nest defence was rated on an ordinal scale: 0 = no response (no bird arrived during the 5-min period); 1 = silent watching; 2 = mobbing (i.e. flights around the dummy and alarm calls); 3 = contact attack(s); see Moksnes et al. (1990) for the description of these behaviours. All observations were made by one researcher (M.H.) to avoid possible observer bias British Trust for Ornithology, Bird Study, 51,

14 Behaviour at nest of parasitized Reed Warbler 259 After presentation of the dummy and clutch manipulation (see above), the dummy was removed and the camera was switched on. Nests were videotaped for 30 minutes. In the case of the human intruder group, we followed a similar pattern. The camera was set up 1 hour before an observer approached the nest. The observer stood near the nest for 2 min (almost all the pairs responded up to this time). After this period the camera was switched on and the observer left. The videotapes were analysed in the laboratory. From the tapes we recorded the arrival time (time to first sight of the nest owner/s), brooding time (time from when the camera was switched on until the host started brooding), look-1 (time spent with nest inspection behaviour, from the moment when the bird first looked into the nest until the moment when it first sat on the clutch) and look-2 (total time spent inspecting the nest after the bird sat on it. Look-1 is not included in look-2. Nest inspection behaviour is defined as the bird looking at the clutch. The number of arrivals at the nest (this variable should reflect the amount of host activity at the nest) and nest time (total amount of time spent at the nest) were also recorded. RESULTS Nest defence and enemy recognition (prediction 1) There were no statistically significant differences in any of the behavioural variables when Reed Warblers were facing Cuckoo or Pigeon dummies (Table 1). Absence of significant discrimination was not caused by absence of nest defence some Reed Warblers responded with alarm calling; however, only two individuals attacked the Cuckoo dummy. Pigeons were never attacked. The difference in the number of experiments with attacks on Cuckoo and Pigeon dummies was, however, not significant (Fisher s exact probabilities test, ns). Table 1. Responses of Reed Warblers to a Cuckoo dummy and a stuffed Pigeon. Values are medians (results of Mann Whitney tests are shown). Nest defence: 0 = no response, 1 = silent watching, 2 = mobbing, 3 = contact attack (see Methods for further explanation). Cuckoo Pigeon Variable (n = 41) (n = 16) U P Latency to first response (min) Latency to alarm calling after first response (min) Time spent within 1m from the nest (s) Median level of nest defence Clutch inspection behaviour (prediction 2) The experimental procedure had no significant effect on Reed Warbler s behaviour during the 30-min period immediately after the experiment (Kruskal Wallis ANOVA tests, P > 0.10 in all cases; Table 2). After seeing the Cuckoo dummy at the nest, Reed Warblers did not tend to arrive at their nests significantly earlier, they did not spend significantly more time with nest inspection behaviour and their activity was not significantly different from those who saw the Pigeon dummy near their nests. The addition of a Cuckoo egg to the host nest had no effect on Reed Warbler behaviour. There was no significant difference in the way a human intruder affected host behaviour compared to dummies (Table 2). Only six Reed Warbler pairs did not visit their nests during the 30-min period (two in Cuckoo morning, two in Cuckoo afternoon, one in Cuckoo + egg replacement morning, one in Pigeon afternoon). We observed no ejection or egg pecking of the Cuckoo egg during the 30-min observation period. Effect of time of day on host behaviour (prediction 3) Reed Warblers generally tended to spend slightly more Table 2. Behaviour of Reed Warblers during the first 30-min period following the dummy experiments. Results are means ± sd. See Methods for explanations of the measured variables. Cuckoo Cuckoo + egg Pigeon Human Kruskal Wallis P Variable (n = 30) (n = 11) (n = 16) (n = 14) test (H) Arrival time (s) 341 ± ± ± ± Brooding time (s) 353 ± ± ± ± Look-1 (s) 4.7 ± ± ± ± Look-2 (s) 9.9 ± ± ± ± No. of arrivals 2.5 ± ± ± ± Nest time (min) 18.5 ± ± ± ± British Trust for Ornithology, Bird Study, 51,

15 260 M. Honza et al. time at the nest in the morning (70.4% of the observation time) than in the afternoon (56.9%), although the differences were not statistically significant (Table 3). Furthermore, we found no significant differences in clutch inspection behaviour by Reed Warblers between the morning and afternoon experimental groups (Table 3). Finally, the intensity of nest defence was almost identical between the morning and afternoon (Table 3). Interestingly, host responses were highly variable among individuals. Individuals that quickly responded to the dummies also quickly arrived at their nests during the following 30-min period (r s = 0.404, n = 41, P = 0.009). Furthermore, Reed Warblers who arrived at their nests quickly reacted more immediately with alarm calling, while those individuals arriving later in response to the dummy tended to delay their alarm calling response (r s = 0.955, n = 41, P < 0.001). DISCUSSION Nest defence and enemy recognition (prediction 1) Our prediction 1 of a more intense nest defence behaviour against the Cuckoo dummy than against the Pigeon dummy was not supported. In general, the level of nest defence by Reed Warblers in our study area was relatively low, only 4.9% of pairs attacked the Cuckoo dummy which is in accordance with previous findings of Røskaft et al. (2002a, 2002b). A low level of aggression provides potential for errors in enemy recognition which could explain why responses to the Cuckoo and Pigeon were not significantly different. Absence of specific recognition of the Cuckoo could be explained by the fact that the Reed Warbler is an intermediate rejecter (rejection rate of natural Cuckoo eggs is 37.5% in our study areas; Øien et al. 1998) egg rejection and aggression against adult parasites evolve in concert, i.e. acceptors are less aggressive than rejecters (Røskaft et al. 2002b). Thus, intermediate rejecters could be expected to show low levels of aggression and also poorer abilities to recognize adult parasites. Our results are in line with this general relationship between parasitic egg and adult parasite related adaptations. Low level of aggression and poor enemy recognition in the Reed Warbler are also expected from the spatial habitat structure hypothesis (Røskaft et al. 2002c) species with some populations breeding in the vicinity of trees (Cuckoo observation points) and some breeding away from trees generally have lower egg rejection rates (Røskaft et al. 2002c) and less aggressive response to the parasite (Røskaft et al. 2002c) than species always breeding close to Cuckoo perches in trees. In the former (including Reed Warbler) there is lower selection pressure for antiparasitic adaptations (including specific enemy recognition) than in the latter. Thus, our results (low aggression, poor enemy recognition) accord with the spatial habitat structure hypothesis. Despite being regularly parasitized (Moksnes et al. 1993b, Øien et al. 1998) Reed Warblers in our study area behaved similarly to individuals from unparasitized populations in Britain (Lindholm & Thomas 2000; see also Røskaft et al. 2002). Duckworth (1991) found that Reed Warblers in his study area in England recognized Table 3. Effect of time of day on Reed Warbler behaviour at their nests during the first 30-min period following the dummy experiment. The data presented are medians (results of Mann Whitney tests are shown). Nest defence was rated on an ordinal scale (see Table 1). Sample sizes for morning/afternoon experiments are: Cuckoo 14/16, Cuckoo + egg 7/4, Pigeon 9/7, human 5/9. Variable Experimental procedure Morning Afternoon U P Time spent at the nest Cuckoo (% of total observation) Cuckoo + egg Pigeon Human Mean Nest inspection = Look-1 (s) Cuckoo Cuckoo + egg Pigeon Human Mean Nest defence (0 3) Cuckoo Cuckoo + egg Pigeon Human Total British Trust for Ornithology, Bird Study, 51,

16 Behaviour at nest of parasitized Reed Warbler 261 the Cuckoo as a special enemy while our results did not support the hypothesis. However, caution is needed in this comparison because the authors of the two studies did not follow the same experimental procedure we presented dummies 1 m from the focal nest while Duckworth (1991) placed dummies at two distances, either directly at the nest or 3 m from the nest. He found that Reed Warblers responded much more aggressively to the dummy on the nest compared to one 3 m from the nest (see also Røskaft et al. 2002). More importantly, Duckworth (1991) observed significant differences in Reed Warbler responses towards different mounts only when dummies were placed directly at the nest there were no significant differences when dummies were placed 3 m from focal nests. Nest inspection behaviour (prediction 2) We predicted that the host would show higher nest inspection activity after being exposed to a Cuckoo dummy than to a Pigeon dummy. We obtained no support for this prediction because Reed Warblers did not modify their behaviour in respect to the kind of intruder (Cuckoo, Pigeon or human). Including a Cuckoo egg in the nest in addition to the Cuckoo dummy did not influence the host behaviour. This could indicate that the sight of the Cuckoo at the nest was not a strong enough cue to indicate parasitism and thus release defence behaviour immediately after the act of parasitism. Moksnes et al. (1993a) have reported similar results for Meadow Pipit Anthus pratensis where the presence or absence of Cuckoo female and/or egg dummies had no significant effect on incubation and nest checking behaviours. An explanation may be that the Reed Warbler needs time to discover parasitism. Like many other species it usually rejects parasitic eggs several days after being parasitized (Davies & Brooke 1988, Moksnes et al. 1990, Grim & Honza 2001, Amundsen et al. 2002). On the other hand the Meadow Pipit, another frequently used Cuckoo host which also is an intermediate rejecter, deserts its nest very quickly (sometimes within a few minutes after the act of parasitism) and always within 24 h after experimental manipulation (Moksnes et al. 1993a). However, even in this quickly rejecting species there were no differences in the behaviour of individuals who could see a Cuckoo dummy at their nests and those who could not (Moksnes et al. 1993a). Davies & Brooke (1988) showed that a Cuckoo mount increases the probability that the Reed Warbler will reject the parasitic egg. Our observations indicate that this effect is not detectable immediately after the parasitism act. Why the Cuckoo lays in the afternoon (prediction 3) Reed Warblers did not spend more time at their nests in the morning compared with the afternoon, which gives no support for prediction 3. Davies & Brooke (1988) measured the temperature of Reed Warbler eggs in the morning and in the evening and found morning broods to be significantly warmer and concluded that Reed Warblers spent more time at their nests in the morning. However, our observations are in accordance with those obtained by Moksnes et al. (2000) who found no difference between the length of time that Reed Warblers spent at their nests in the forenoon and afternoon. Therefore, the absence of changes in nest attentiveness during the day does not support the view that Cuckoo afternoon laying has evolved because the host is away from its nest more during this part of the day. Cuckoo afternoon laying could also result from a higher intensity of clutch inspection behaviour and nest defence by hosts in the morning. Our study does not support either of these behaviours because Reed Warblers showed no differences in clutch inspection and nest defence behaviours during the day. Absence of immediate ejections We did not observe any pecking or ejections during the 30-min period after we added a parasitic egg to a focal nest. We used natural Cuckoo eggs which in our study area do not correspond very well to Reed Warbler eggs as judged by the human eye (Edvardsen et al. 2001). However, Reed Warblers in our study area accept 62.5% of these natural parasitic eggs (Øien et al. 1998) and even highly non-mimetic eggs are frequently accepted (43.7%; Stokke et al. 1999). Low aggression and absence of specific responses to the Cuckoo both during dummy experiments and in the 30-min period thereafter could result from the fact that the arms-race between the Cuckoo and Acrocephalus warblers in our study area is at a relatively early stage as indicated by host acceptance of badly matching Cuckoo eggs and a low match between parasitic and host eggs (Edvardsen et al. 2001). Honza et al. (2001) hypothesized that the fact that 86.3% of Cuckoo eggs laid in Reed Warbler nests belong to the Sylvia egg morph could be indicative of host switching due to construction of fish ponds in Moravia in the 2004 British Trust for Ornithology, Bird Study, 51,

17 262 M. Honza et al. 16th century. The low level of host defences in our study area could represent an example of evolutionary lag. Alternatively, this finding could be explained by gene flow between parasitized and unparasitized populations, see Røskaft et al. (2002). To explain poor Reed Warbler antiparasitic defences in our study area it is helpful to consider other sympatric hosts. The Great Reed Warbler is currently parasitized more frequently than the Reed Warbler (Kleven et al. 1999). Some 30 years ago Reed Warbler and Great Reed Warbler were equally common (Hudec 1975). At present the Great Reed Warbler is almost absent from the study area (but is still preferred as a host). If in the past Cuckoos also laid their eggs preferentially in the nests of the Great Reed Warbler then it is possible that the selection for specific host adaptations in the Reed Warbler was weak before the recent decline of the Great Reed Warbler in the study area. In conclusion, low level and specificity of host responses after presentation of stuffed dummies in our study area could be expected from the fact that the Reed Warbler is an intermediate rejecter of Cuckoo eggs (Røskaft et al. 2002b). This could result from a relatively short co-evolution between host and parasite (see also Edvardsen et al. 2001, Honza et al. 2001) and is in line with the spatial habitat structure hypothesis (Røskaft et al. 2002c). ACKNOWLEDGEMENTS We are grateful to E. Tkadlec for helpful comments on the manuscript. We thank O. Kleven, O. Mikulica, I. Øien and G. Rudolfsen for their help in the field. Financial support was received from the Grant Agency of the Czech Republic (grant 206/00/P046 and A ) and Ministry of Education grants VS and MSM This study has been carried out under permission given to M.H. and in accordance with the laws and ethical guidelines of the Czech Republic. REFERENCES Alvarez, F Proximity of trees facilitates parasitism by Cuckoos Cuculus canorus on Rufous Warblers Cercotrichas galactotes. Ibis 135: 331. Amundsen, T., Brobakken, P.T., Moksnes, A. & Røskaft, E Rejection of Cuckoo eggs in relation to female age in the bluethroat. J. Avian Biol. 33: Bazin, R.C. & Sealy, S.G Experiments on the responses of a rejecter species to threats of predation and Cowbird parasitism. Ethology 94: Briskie, J.V. & Sealy, S.G Responses of Least Flycatchers to experimental inter- and intraspecific brood parasitism. Condor 89: Burgham, M.C.J. & Picman, J Effect of Brown-headed Cowbirds on the evolution of Yellow Warbler anti-parasite strategies. Anim. Behav. 38: Clarke, A.L., Øien, I.J., Honza, M., Moksnes, A. & Røskaft, E Factors affecting Reed Warbler risk of brood parasitism by the Common Cuckoo. Auk 118: Davies, N.B. & Brooke, M.L Cuckoos versus Reed Warblers: adaptations and counter-adaptations. Anim. Behav. 36: Duckworth, J.W Responses of breeding Reed Warblers Acrocephalus scirpaceus to mounts of Sparrowhawk Accipiter nisus, Cuckoo Cuculus canorus and Jay Garrulus glandarius. Ibis 133: Edvardsen, E., Moksnes, A., Røskaft, E, Øien, I.J. & Honza, M Egg mimicry in Cuckoos parasitizing four sympatric species of Acrocephalus warblers. Condor 103: Gill, S.A. & Sealy, S.G Yellow Warbler nest defence: recognition of a brood parasite and an avian nest predator. Behaviour 133: Grim, T. & Honza, M Differences in behaviour of closely related thrushes (Turdus philomelos and T. merula) to experimental parasitism by the Common Cuckoo Cuculus canorus. Biologia 56: Hobson, K.A. & Sealy, S.G Responses of yellow warblers to the threat of Cowbird parasitism. Anim. Behav. 38: Honza, M., Øien, I.J., Moksnes, A. & Røskaft, E Survival of Reed Warbler Acrocephalus scirpaceus clutches in relation to nest position. Bird Study 45: Honza, M., Moksnes A., Røskaft E. & Stokke, B.G How are different Common Cuckoo Cuculus canorus egg morphs maintained? An evaluation of different hypotheses. Ardea 89: Honza, M., Taborsky, B., Taborsky, M., Teuschl, Y., Vogl, W., Moksnes, E. & Røskaft, E Behaviour of female common Cuckoos, Cuculus canorus, in the vicinity of host nests before and during egg laying: a radiotelemetry study. Anim. Behav. 64: Hudec, K Density and breeding of birds in the Reed swamps of Southern Moravian ponds. Acta Sci. Nat. Brno 9(6): Kilner, R.M., Noble, D.G. & Davies, N.B Signals of need in parent-offspring communication and their exploitation by the Common Cuckoo. Nature 397: Kleven, O., Moksnes, A., Røskaft, E. & Honza, M Host species affects the growth rate of Cuckoo (Cuculus canorus) chicks. Behav. Ecol. Sociobiol. 47: Lindholm, A.K. & Thomas, R.J Differences between populations of Reed Warblers in defences against brood parasitism. Behaviour 137: Martín-Vivaldi, M., Soler, M. & Møller, A.P Unrealistically high costs of rejecting artificial model eggs in Cuckoo Cuculus canorus hosts. J. Avian Biol. 33: Moksnes, A. & Røskaft, E Adaptations of Meadow Pipits to parasitism by the Common Cuckoo. Behav. Ecol. Sociobiol. 24: Moksnes, A. & Røskaft, E Egg-morphs and host preference in the Common Cuckoo (Cuculus canorus): an analysis of Cuckoo and host eggs from European museum collections. J. Zool. Lond. 236: Moksnes, A., Røskaft, E., Braa, A.T., Korsnes, L., Lampe, H.T. & Pedersen, H.C Behavioural responses of potential hosts towards artificial Cuckoo eggs and dummies. Behaviour 116: British Trust for Ornithology, Bird Study, 51,

18 Behaviour at nest of parasitized Reed Warbler 263 Moksnes, A., Røskaft, E. & Korsnes, L. 1993a. Rejection of Cuckoo (Cuculus canorus) eggs by Meadow Pipits (Anthus pratensis). Behav. Ecol. 4: Moksnes, A., Røskaft, E., Bicík, V., Honza, M. & Øien, I.J. 1993b. Cuckoo Cuculus canorus parasitism on Acrocephalus warblers in Southern Moravia in the Czech Republic. J. Ornithol. 134: Moksnes, A., Røskaft, E. & Solli, M.M Documenting puncture ejection of parasitic eggs by Chaffinches Fringilla coelebs and Blackcaps Sylvia atricapilla. Fauna Norv. Ser. C Cinclus 17: Moksnes, A., Røskaft, E., Hagen, G.L., Honza, M., Mørk, C. & Olsen, P.H Common Cuckoo Cuculus canorus and host behaviour at Reed Warbler Acrocephalus scirpaceus nests. Ibis 142: Moskát, C. & Honza, M Effect of nest and nest site characteristics on the risk of Cuckoo Cuculus canorus parasitism in the Great Reed Warbler Acrocephalus arundinaceus. Ecography 23: Øien, I.J., Honza, M., Moksnes, A. & Røskaft, E The risk of parasitism in relation to distance from Reed Warbler nests to Cuckoo perches. J. Anim. Ecol. 65: Øien, I.J., Moksnes, A., Røskaft, E. & Honza, M Costs of Cuckoo Cuculus canorus parasitism to Reed Warblers Acrocephalus scirpaceus. J. Avian Biol. 29: Øien, I.J., Moksnes, A., Røskaft, E., Edvardsen, E., Honza, M., Kleven, O. & Rudolfsen, G Conditional host responses to Cuckoo Cuculus canorus parasitism. In Adams, N.J. & Slotow, R.H. (eds) Proc. 22nd Int. Ornithol. Congr. University of Natal, Durban, South Africa: Røskaft, E., Moksnes, A., Meilvang, D., Bicík, V., Jemelíková, J. & Honza, M. 2002a. No evidence for recognition errors in Acrocephalus warblers. J. Avian Biol. 33: Røskaft, E., Moksnes, A., Stokke, B.G., Bicík, V. & Moskát, C. 2002b. Aggression to dummy Cuckoos by potential European Cuckoo hosts. Behaviour 139: Røskaft, E., Moksnes, A., Stokke, B.G., Moskát, C. & Honza, M. 2002c. The spatial habitat structure hypothesis of host populations explains the pattern of rejection behavior in hosts and parasitic adaptations in Cuckoos. Behav. Ecol. 13: Rothstein, S.I Cowbird parasitism and egg recognition in the northern oriole. Wilson Bull. 89: Rothstein, S.I A model system for coevolution: avian brood parasitism. Annu. Rev. Ecol. Syst. 21: Sealy, S.G Evolution of host defences against brood parasitism: implications of puncture-ejection by a small passerine. Auk 113: Sealy, S.G. & Lorenzana, J.C Yellow Warblers (Dendroica petechia) do not recognize their own eggs. Bird Behav. 12: Sealy, S.G. & Neudorf, D.L Male Northern Orioles eject Cowbird eggs: implications for the evolution of rejection behavior. Condor 97: Sealy, S.G., Neudorf, D.L., Hobson, K.A. & Gill, S.A Nest defense by potential hosts of the Brown-headed Cowbird: methodological approaches, benefits of defense, and coevolution. In Rothstein, S.I. & Robinson, S. Parasitic Birds and Their Hosts. Studies in Coevolution: Oxford University Press, New York & Oxford. Soler, M., Martin-Vivaldi, M. & Perez-Contreras, T Identification of the sex responsible for recognition and the method of ejection of parasitic eggs in some potential common Cuckoo hosts. Ethology 108: Stokke, B.G., Moksnes, A., Røskaft, E., Rudolfsen, G. & Honza, M Rejection of artificial cuckoo (Cuculus canorus) eggs in relation to variation in egg appearance among reed warblers (Acrocephalus scirpaceus). Proc. R. Soc. Lond. B 266: Wyllie, I The Cuckoo. London, Batsford. (MS received 27 February 2003; revised MS accepted 20 November 2003) 2004 British Trust for Ornithology, Bird Study, 51,

19 19. Grim T.: Does conspicuous nest defence by the blackcap attract nest defence helpers? 19. (subm.). Grim T.: Does conspicuous nest defence by the blackcap attract nest defence helpers? (subm.).

20 Does conspicuous nest defence attract nest defence helpers? Tomáš Grim 1 1 Department of Zoology Palacký University tr. Svobody 26 CZ Olomouc Czech Republic grim@prfnw.upol.cz Abstract One hypothesized function of conspicuous mobbing of predators by bird nest owners is to attract third-party predators (predator attraction hypothesis) or neighbouring birds (calling for help hypothesis). These may help the nest owners by distracting and/or attacking the mobbed intruder near the nest. I experimentally tested these hypotheses in the blackcap Sylvia atricapilla, a small passerine with a highly aggressive and conspicuous nest defence behaviour. I elicited blackcaps aggressive responses by presenting stuffed dummies of the brood parasitic common cuckoo Cuculus canorus and controls near their nests (n = 75 nests). At 32% nests blackcap s responses to dummies attracted birds from 21 passerine species (at 68% nests no neighbours appeared during trials). Up to 15 individuals were attracted per experiment. No potential predators of the cuckoo were attracted despite living in the study area, thus rejecting the predator attraction hypothesis. Most of the attracted birds were heterospecifics and rarely participated in mobbing, thus the calling for help hypothesis was not supported as well. The number of attracted birds was a positive function of the owner s intensity of nest defence, the most likely cue for attraction being vocal signals (rates of alarm calling) but not visual cues (rates of attacks). Suitable and unsuitable cuckoo hosts did not differ in their behaviour in the vicinity of defended nests. Based on discussion of possible costs and benefits for both blackcaps and attracted neighbours I conclude that the observed pattern of the positive correlation between the intensity of nest defence and the number of attracted birds is most likely a proximate by-product of the conspicuous nest defence by blackcaps and is selectively neutral for the study species. 1