Studies of microevolutionary dynamics and adap-tations have led to a better general understanding ofevolution. One particular evolutionary process thathas received much attention is the interaction be-tween different species leading to mutual responsesand reciprocal adaptations. Coevolution is defined asspecialised relationships between species that leads totandem evolutionary changes (Janzen 1980; Thomp-son 1994). Such interactions have intrigued re-searchers since the publication of Charles Darwin’s(1859) “The origin of species”. One of the most suit-able model systems for studying coevolution is theinteraction between avian brood parasites and theirhosts (Rothstein & Robinson 1998). Among the best-studied interspecific brood parasites are the CommonCuckoo Cuculus canorus (L.) and the Great SpottedCuckoo Clamator glandarius (L.) in Europe, and theBrown-headed Cowbird Molothrus ater (Boddaert) inNorth America.

The interaction between a brood parasite and it’shosts can be pairwise when one parasite species spe-

cialise on only one host species as is the case for theGreat Spotted Cuckoo and its Magpie Pica pica (L.)host in Spain (Soler & Møller 1990; Soler & Soler2000). In other host-parasite interactions, brood para-sites are utilising several host species and do not spe-cialise on any of them. This is the case in NorthAmerica where the generalist Brown-headed Cow-bird parasitizes more than 200 different host species(Friedmann 1963, 1971; Mayfield 1965, 1977; Fried-mann et al. 1977, Friedmann & Kiff 1985). In Eura-sia, interactions between the Common Cuckoo and itshosts represent a third possibility. The Cuckoo as aspecies is a generalist and is known to have para-sitized more than 100 different species in Europealone (Moksnes & Røskaft 1995). On the other hand,individual females are specialists, exploiting only oneor a few host species in which they (often) lay amimetic egg (Chance 1940; Baker 1942; Moksnes &Røskaft 1995; Davies 2000).

Hosts of brood parasites frequently suffer costs re-lated to egg loss, misdirected parental care and over-crowding (Petrie & Møller 1991; Øien et al. 1998;Power 1998), and such costs are in many cases se-vere. Thus, in Common Cuckoos the parasite chick

The enigma of imperfect adaptations in hosts of avian broodparasites

Bård G. STOKKE, Arne MOKSNES and Eivin RØSKAFT#

Department of Biology, Norwegian University of Science and Technology, NTNU, N-7491 Trondheim, Norway

Abstract Coevolutionary dynamics allow revealing ongoing microevolutionaryprocesses and adaptations. Interactions between obligate avian brood parasites andtheir hosts are suitable systems for the study of coevolution. This fascinating subjecthas interested both scientists and the common man for decades, but there are stillmany unanswered questions. One of the main puzzles is the apparent imperfect rejec-tion behaviour against parasite eggs among hosts of brood parasites leading to a situa-tion where many hosts experience severe costs of parasitism, which dramatically re-duce their fitness. In order to lower these costs, hosts are expected to evolve counter-adaptations against parasitism. However, many host species show no rejection or onlyintermediate rejection rates against non-mimetic parasite eggs. In this review we con-sider several hypotheses set out to explain this phenomenon, and then we link thesehypotheses with a consideration of various sources that can influence decision-mak-ing when hosts are faced with the possibility that a parasite egg is present in theirnest.

Key words Brood parasitism, Coevolution, Decision making, Host defence, Rejec-tion behaviour

SPECIAL FEATURE Interactions between avian brood parasites and their hosts

17

ORNITHOLOGICALSCIENCE

© The Ornithological Society of Japan 2005

(Received 11 October 2004; Accepted 7 February 2005)# Corresponding author, E-mail: roskaft@bio.ntnu.no

Ornithol Sci 4: 17–29 (2005)

ejects all the eggs and young of the host so that it be-comes the sole occupant of the nest (Wyllie 1981).Brown-headed Cowbird chicks, on the other hand,generally grow up together with the young of thehosts, but often outcompete these because of earlierhatching, which results in a size advantage, especiallyin small hosts (Payne 1997; Rothstein & Robinson1998). In addition to the obvious costs in current re-production, hosts may suffer costs in terms of low-ered future survival and reproduction (trade-offs inlife history theory; Stearns 1992; Roff 2002), but noefforts have so far been made to measure such costs(Rothstein & Robinson 1998; Winfree 1999). How-ever, it has recently been shown that increased eggproduction in one breeding season (e.g. due to aban-donment and relaying after being parasitized) iscostly, and may severely reduce future fitness in birds(Nager et al. 2001; Visser & Lessells 2001).

Due to the costs inflicted upon hosts by the broodparasite, there will be a strong selection favouringevolution of host defences. In the case of a host-para-site system in which female parasites specialise on asingle host species, the parasite will reply by evolv-ing more sophisticated trickery to counter recognitionmechanisms evolved by the host. This coevolutionbetween brood parasites and their hosts, in particularin the Common Cuckoo-host system, has led to a va-riety of adaptations on both sides, which has tradi-tionally been described as an evolutionary arms racecontaining several stages (Dawkins & Krebs 1979;Davies & Brooke 1989b; Rothstein 1990). Whenhosts start to reject non-mimetic parasite eggs, para-sites laying mimetic eggs will have a selective advan-tage (Brooke & Davies 1988). The parasite egg mim-icry may evolve to be a good match of the host eggs.Thus, parasite gentes or tribes may arise, which spe-cialise on one or a few host species (Jourdain 1925;Chance 1940; Baker 1942; Lack 1968; Gibbs et al.2000). The Common Cuckoo in Europe can be classi-fied into at least 16 such gentes (Wyllie 1981; Al-varez 1994; Moksnes & Røskaft 1995). However, ahost counteradaptation at this stage would be toevolve eggs with a more intricate surface pattern (i.e.signatures; Davies 2000), so that it is harder for theparasite to mimic its eggs. Additionally, the hosts canevolve a lower intraclutch variation and/or a higherinterclutch variation in egg appearance (Swynnerton1918; Victoria 1972; Rothstein 1974; Davies &Brooke 1989b; Øien et al. 1995; Soler & Møller1996; Lahti & Lahti 2002; Stokke et al. 2002b). Thelow intraclutch variation makes it easier for the hosts

to detect and reject foreign eggs (Stokke et al. 1999),whereas a high interclutch variation makes it moredifficult for parasites to mimic the eggs of a specifichost (Davies & Brooke 1989b). In Europe, therefore,hosts generally have a lower intraclutch variation andhigher interclutch variation in egg appearance thanpotential hosts in North America, indicating that thehost-parasite system in Europe has reached a moreadvanced level than the system in North America(Stokke et al. 2002b). In North America, hosts of theBrown-headed Cowbird often show an all or none re-sponse to parasite eggs, i.e. different hosts can beclassified as either rejecters or acceptors of the gener-ally non-mimetic cowbird eggs (Rothstein 1975a,1990; Rothstein & Robinson 1998). Due to that theBrown-headed Cowbird is a generalist and there ismany acceptor-hosts available, there is yet no signifi-cant selection pressure on the Cowbird for evolvingmimetic eggs. The arms race therefore seems to re-main at an early stage until the majority of hostspecies has evolved egg rejection forcing the cowbirdto evolve more advanced adaptations like egg mim-icry. Johnson and Lanyon (1999) have found that thegeneralists are younger as species than specialistsamong cowbirds, and Rothstein et al. (2004) havepointed out that it is likely that generalists will be-come more specialized in the future (but see Lanyon1992). Until now, therefore, the Brown-headed Cow-bird-host system may hardly be called a coevolution-ary arms race. Theoretically, the arms race may con-sist of several additional stages than those originallydescribed by Davies and Brooke (1989a, b). In Eu-rope, where some hosts have evolved a very high re-jection rate against parasite eggs and at the same timea high interclutch variation in egg appearance, thecuckoo may potentially evolve egg mimicry inde-pendently within specific host egg morphs so thateach cuckoo gens in reality split up into severalgentes. Each such gens may specialise on one hostegg morph, however, no such adaptations has untilnow been detected. The next step in the arms racewould be in the nestling stage, where hosts mayevolve chick rejection (see e.g. Grim et al. 2003). Inorder to counter this, parasites may as a responseevolve mimetic chicks in morphology or vocalisationas is the case among some African whydahs and in-digobirds (Payne 1973a, b, 1976).

B. G. STOKKE, A. MOKSNES and E. RØSKAFT

18

APPARENT IMPERFECT HOST BEHAVIOUR: LAG, EQUILIBRIUM

OR OTHER EXPLANATIONS?

Despite the fact that we would expect a high rejec-tion rate against parasite eggs, many hosts of broodparasites show very low or intermediate levels of re-jection of foreign non-mimetic eggs (Rothstein1975b; Davies & Brooke 1989a; Moksnes et al. 1990;Rothstein 1990; Brooke et al. 1998; Alvarez 1999;Grim 2002). The profound variation in rejection ratesof parasite eggs among hosts of brood parasites is in-triguing, and several hypotheses have been proposedto explain this phenomenon, all of which are consis-tent with the above mentioned arms race model. Wewould like to underline here that these hypotheses areconsistent with the arms race model even in the caseswhere parasites have not started to counteract the re-jection behaviour or when hosts still accept the para-site egg, because such an early phase is consistentwith stage one in the arms race model (Davies &Brooke 1989a, b).

1) Evolutionary lagAccording to the “evolutionary lag hypothesis”, a

lack of proper defence against parasitism is due to alag in the origin or spread of anti-parasite adaptations(Rothstein 1975a; Dawkins & Krebs 1979; Davies &Brooke 1989a; Rothstein 1990; Lotem & Rothstein1995). Thus, lag is more likely to explain lack of de-fences in recently established host-parasite systemsthan in systems where parasites and hosts have coex-isted for a long time (Rothstein & Robinson 1998;Winfree 1999). Lag in the evolution of anti-parasiteadaptations is basically the first step in the arms raceand proper genetic background can be regarded a pre-requisite for an arms race to take place rather than astage in the arms race itself. Many hosts of theBrown-headed Cowbird are acceptors of non-mimeticeggs (Rothstein 1990). This parasite has expanded itsrange dramatically in the last few centuries due to an-thropogenically induced habitat changes, and manynew hosts have become available (Rothstein &Robinson 1998). Comparing hosts that originallywere sympatric with the Brown-headed Cowbird withthose that have been explored more recently has putthe lag hypothesis to the test. The original Cowbirdhosts should show a higher rejection rate than newand naive hosts, and support for such a scenario hasbeen found (Hosoi & Rothstein 2000; Peer & Sealy2004). Some hosts of the Common Cuckoo also show

no or only moderate rejection of non-mimetic eggs,even though they are regularly parasitized. The Dun-nock Prunella modularis (L.) is an example of such aspecies, and the lack of defences against parasitism inthis species could be explained by the lag hypothesis(Brooke & Davies 1988).

2) Evolutionary equilibriumA contrasting hypothesis is the “evolutionary equi-

librium hypothesis”, stating that acceptance of para-site eggs is a result of a balance between various op-posing selection pressures (Zahavi 1979; Rohwer &Spaw 1988; Petit 1991; Lotem et al. 1992, 1995;Lotem & Nakamura 1998). Rejection costs andrecognition errors are especially important in this re-spect (Lotem & Nakamura 1998). Rejection costs areaccidental losses of their own eggs when rejecting orattempting to reject parasite eggs, whereas recogni-tion errors are erroneous rejections of their own eggswhen not parasitized (Lotem et al. 1995; Rothstein &Robinson 1998; Røskaft et al. 2002a; Stokke et al.2002a). There has been some confusion in the litera-ture over the concept “recognition error”. As statedabove, recognition error should be defined as hostsrejecting their own eggs in the absence of a parasiteegg in their nest (Rodríguez-Gironés & Lotem 1999).Rejection of own eggs in the presence of a parasiteegg (whether it is mimetic or not) could be inter-preted as a rejection cost as well as a recognitionerror, resulting in a situation where the two terms areinseparable. Røskaft et al. (2002a) used the term“possible rejection cost” in such cases.

Costs associated with the rejection of parasite eggsare likely to be highest when the parasite laysmimetic eggs (Davies & Brooke 1988; Davies et al.1996; Lotem & Nakamura 1998). Furthermore, theimpact of such costs is likely to be dependent on thehost-parasite system in question (Røskaft & Moksnes1998). As Common Cuckoo hosts normally lose alltheir eggs and young when successfully parasitized,only recognition errors are assumed to be of impor-tance as an opposing selection pressure againstproper host defence. Common Cuckoo hosts couldtherefore tolerate high rejection costs and it wouldstill be adaptive to reject foreign eggs (Røskaft &Moksnes 1998; Stokke et al. 2002a). In these hosts,intermediate rejection rates could thus be the result ofa balance between the costs of acceptance and thecosts of making recognition errors. Many hosts of theBrown-headed Cowbird, however, grow up togetherwith the parasite chick, and thus rejection costs can

19

The enigma of imperfect adaptations in hosts of avian brood parasites

also be important in this respect. The occurrence ofrejection costs has been thoroughly documented inprevious studies (Molnàr 1944; Rothstein 1976;Rothstein 1977; Davies & Brooke 1988; Davies &Brooke 1989a; Davies & Brooke 1989b; Rohwer etal. 1989; Røskaft et al. 1990; Røskaft et al. 1993;Lorenzana & Sealy 2001; Welbergen et al. 2001), butonly a few studies have revealed recognition errors innon-parasitized host clutches (Marchetti 1992). How-ever, loss of own eggs in non-parasitized nests canalso occur in species that are acceptors of foreigneggs, and can be due to jostling, partial predation orother reasons (Rothstein 1982b; Kemal & Rothstein1988; Lerkelund et al. 1993). To make the pictureeven more complicated there could be differentamounts of errors made among various populationsof the same species. Thus, Davies and Brooke (1988)found that recognition errors occurred within a ReedWarbler Acrocephalus scirpaceus (Hermann) popula-tion in England, whereas Røskaft et al. (2002a) foundno evidence for recognition errors even after expo-sure to a dummy Cuckoo among Reed Warblers andGreat Reed Warblers Acrocephalus arundinaceus (L.)in the Czech Republic.

The evolutionary equilibrium hypothesis is obvi-ously a part of the arms race model, because it de-scribes a stage where the hosts are expected to haveevolved traits in order to reject the parasite egg andtherefore the hypothesis evaluates the fitness of ac-ceptors and rejecters of parasite eggs. However, incases where parasites have evolved no counteradapta-tions, as in the case of the Brown-headed Cowbird, itmust be regarded as at a very early stage of the armsrace, where no coevolution has yet occurred. The testof the evolutionary equilibrium hypothesis rely on thepossibility to show that recognition errors or othercosts of rejection occur (Petit 1991; Soler et al. 1995;Røskaft et al. 2002a). Mathematical modelling byTakasu and co-workers (Takasu et al. 1993; Takasu1998a; Takasu 1998b) has provided some support forthis hypothesis.

3) The effect of geographically structured popula-tions

It has long been appreciated that the benefits of re-jection behaviour may depend on parasite density(Davies et al. 1996; Takasu 1998b). Recently, re-searchers have extended this idea, and acknowledgedthe importance of looking upon interactions betweenparasites and their hosts not only as isolatedentities/populations, but also as metapopulations

(Levins 1969; Lindholm 1999; Lindholm & Thomas2000; Soler & Soler 2000). The investigation of host-parasite dynamics in a metapopulation context hasmainly been applied to plants and their pathogens(see review in Kaltz & Shykoff 1998). However, themetapopulation approach has also been applied to thecoevolutionary interactions between Great SpottedCuckoos and their Magpie hosts (Soler et al. 1998;Martinez et al. 1999; Soler et al. 1999; Soler & Soler2000), and Lindholm (1999) has studied CommonCuckoo-Reed Warbler interactions in Britain using asimilar approach. Taking metapopulation dynamicsinto account, a new dimension is introduced to co-evolutionary studies. Coevolutionary interactionsamong spatially separated populations may lead to lo-cally distinct adaptations. However, these local adap-tations can be precluded by phenomena like extinc-tions, recolonisations, gene flow, variable selectionpressures and environmental stochasticity (Kaltz &Shykoff 1998; Martinez et al. 1999). Variation in co-evolutionary dynamics, local adaptations and effectsof stochasticity among various populations, can leadto spatial differences in adaptive traits or in otherwords lead to an increasing diversity in both geno-and phenotypes. Thus, the metapopulation approachcan help to explain why some host populations have amixture of acceptors and rejecters of parasite eggs,not necessarily because of costs associated with re-jection behaviour within specific populations butrather because of gene flow of “acceptor alleles”from non-parasitized host populations preventing theevolution of proper host defence in parasitized popu-lations. Martinez et al. (1999) found that the amountof gene flow between different Magpie populations inEurope is high, and Soler et al. (1999) obtained fur-ther support for the “rejecter-gene flow hypothesis”regarding this parasite-host system. This hypothesiscan explain the persistence of a high rejection rate incurrently non-parasitized host populations, and fur-thermore, the coexistence of both acceptors and re-jecters within a population, as well as the rapid in-crease in rejection rates in recently parasitized popu-lations by gene flow of “rejecter alleles” from areasof sympatry to allopatry with the parasite (see alsoSoler & Møller 1990; Briskie et al. 1992; Soler &Soler 2000).

The “spatial habitat structure hypothesis”, alsobased upon metapopulation dynamics, can explainthe variation in rejection behaviour among hosts bythe influence of gene flow and local adaptations in theevolution of host defences (Røskaft et al. 2002b, c).

B. G. STOKKE, A. MOKSNES and E. RØSKAFT

20

This hypothesis is based on the fact that most broodparasites, as the Common Cuckoo, must have accessto perches in trees where they can be on the lookoutfor host nests that they can parasitise (Alvarez 1993;Øien et al. 1996; Clotfelter 1998). Passerine birdsthat breed in the open are therefore believed to escapeparasitism and should not have evolved defencesagainst parasitism. At the other extreme, we findspecies that breed only near trees (i.e. in woodlandsand woodland borders). In such hosts, most popula-tions are utilised by brood parasites and should thusrapidly evolve rejection behaviour. Initially there willbe a high level of parasitism enforcing a strong selec-tion pressure upon the hosts for the evolution ofadaptations against parasitism. The brood parasitemight have problems in matching the speed of theevolution of host defences, and therefore only spo-radically evolves adaptations like egg mimicry. Thelevel of parasitism would thus decline as hosts evolvea proper defence. One prediction from the spatialhabitat structure hypothesis concerns host speciesthat can breed both near trees and far away fromtrees. Because only the populations that breed neartrees are exposed to parasitism and thus experience aselective pressure to evolve host defence, gene flowfrom non-parasitized populations can prevent theevolution of a proper host defence, leading to inter-mediate rejection rates in parasitized populations.Such net gene flow from non-parasitized to para-sitized populations is based upon the assumption thatthe mean fitness in parasitized populations (sinks) isless than in non-parasitized ones (sources), and thusmore individuals are produced in the latter popula-tions. This prediction also applies to single popula-tions, in which some individuals breed near treeswhile others breed further away from trees. In popu-lations fulfilling these criteria, host rejection behav-iour is moderate, cuckoo egg mimicry is good, andthe level of parasitism is high (Røskaft et al. 2002c).Recently, Røskaft et al. (in prep.) have modelled thedynamics of the “spatial habitat structure hypothe-sis”, and made a wider application of its use. Thus,the hypothesis can be used to predict host- and para-site adaptations in general as long as the host is spa-tially distributed in local populations, and that somepopulations remain non-parasitized. The speed of se-lection is in such cases dependent on migration ratesbetween populations. If migration rates are very high,as it is for many passerine species (Weatherhead &Forbes 1994), local adaptations might be impover-ished. In such cases the selection regime in source

populations will unduly influence evolution in thesink populations (Holt 1996).

ACCEPT OR REJECT? THE PROCESS OF DECISION-MAKING

The lag, equilibrium and metapopulation hypothe-ses can all explain the existence of varying occur-rences of rejection behaviour among hosts of broodparasites. They are all consistent with the hypothesisthat the interaction between parasites and their hostscan be described as coevolutionary arms races, al-though, particularly the lag hypothesis obviously canbe regarded as a starting point of the arms race whereno coevolution has yet occurred. The various hy-potheses may therefore explain different stages of thearms race (Davies 1999; Winfree 1999; Davies2000).

A topic that has only received little attention is theprocess of decision-making in hosts when they areevaluating if they are parasitized or not (Davies et al.1996; Rodríguez-Gironés & Lotem 1999). We nowwill demonstrate how this process (Fig. 1) is relatedto the considerable variability in defensive behaviourthat is found among host species. This link betweenprocess and response enables us to evaluate the previ-ously described evolutionary hypotheses in more de-tail.

1) Decision-making and clutch characteristicsHost decision-making is depending on cognitive

stimuli from the clutch, or in other words from char-acteristics of the clutch. Important in the host’s deci-sion making process, therefore, is signal detection,which really is the relationship between “signal” and“noise” in the decision-making (Rodríguez-Gironés& Lotem 1999). Important factors in this respect arehost intraclutch variation in egg appearance, parasiteegg mimicry and parasite induced clutch reduction(box 1, Fig. 1). When a host in general has a low intr-aclutch variation, there is no need for experience inorder to recognise the appearance of their own eggs(Rothstein 1974, 1978; Lotem et al. 1995). The lowintraclutch variation makes it very likely that all sub-sequent eggs laid by the female will look very similarto the first laid egg, and therefore a prolonged learn-ing period is not required (Lotem et al. 1995;Rodríguez-Gironés & Lotem 1999). ChaffinchesFringilla coelebs (L.) and Blackcaps Sylvia atri-capilla (L.) generally are very good rejecters of non-mimetic eggs, and also have a low intraclutch varia-

21

The enigma of imperfect adaptations in hosts of avian brood parasites

tion (Braa et al. 1992; Moksnes 1992; Moksnes &Røskaft 1992; Øien et al. 1995). In these species, re-jection of foreign eggs is strongly dependent on mim-icry between parasite and host eggs (Braa et al. 1992;Moksnes 1992; Honza et al. 2004; Stokke et al.2004). The better the mimicry, the poorer the rejec-tion ability. These results suggest that a threshold ex-ists in the cognitive abilities (McLean & Maloney1998), and that highly mimetic eggs fall below thisthreshold and thus are not perceived as foreign. How-ever, in species with higher intraclutch variation, therisk of making recognition errors may be higher.Therefore, in such species a prolonged learning pe-riod may be required in which the rejection behaviourand cognitive abilities of the host is adjusted withage, especially when they are utilised by parasiteslaying mimetic eggs (Lotem et al. 1995; Lotem &Nakamura 1998; Rodríguez-Gironés & Lotem 1999)(box 2, Fig. 1). A scenario, first proposed by Davies

and Brooke (1988), where an equilibrium betweenacceptors and rejecters may exist is when young in-experienced breeders have not yet learned the appear-ance of their own eggs and thus accept most of theeggs laid (both their own and parasite eggs) in theirfirst breeding attempt (Lotem et al. 1992, 1995). Ithas been proposed that such young and naive breed-ers also usually occupy nest-sites of lower quality(where they are more likely to be parasitized) thanolder more experienced females (Øien et al. 1996;Grim 2002) but see Payne & Payne (1998)). This sce-nario could facilitate parasitism and lower selectionfor parasite egg mimicry. This non-random distribu-tion of parasitism could explain imperfect egg mim-icry in parasites utilising hosts that need to learn theappearance of their eggs (Grim 2002). However, theeffect of age obviously differs among species. Lotemet al. (1992, 1995) found that both rejection behav-iour and intraclutch variation were correlated with

B. G. STOKKE, A. MOKSNES and E. RØSKAFT

22

Fig. 1. The process of host decision-making when assessing the possibility of being parasitized. The decision-making in hosts of brood parasites is depending on cognitive inputs in form of clutch characteristics (box 1). Theabilities of the host to recognise odd eggs are likely to be adjusted by experience and learning due to costs of mak-ing recognition errors (box 2). In addition, the occurrence of such costs may incur that some hosts need additionalstimuli when deciding whether or not to reject parasite eggs (box 3). In addition, the decision-making process willbe influenced by the genotype of the host (box 4). All these inputs lead to a host response, which can be seen as ei-ther acceptance or rejection of foreign eggs.

age in a Great Reed Warbler population in Japan.Several other studies, on the other hand, have not re-vealed any effects of age on rejection behaviour(Stokke et al. 1999; Marchetti 2000; Soler et al.2000a; Amundsen et al. 2002; Stokke et al. 2004), oregg signatures (Gosler et al. 2000).

Many brood parasites often remove one or severalhost eggs when depositing their own egg (Sealy1992; Davies & Brooke 1998; Moksnes et al. 2000;box 1, Fig. 1). This clutch reduction can affect thehost’s assessment of the nest content and thus also thehost rejection rate (Øien et al. 1999). If the parasiteremoves host eggs (or other parasite eggs) that aremost dissimilar to the parasite egg, i.e. eggs that fre-quently are the most dissimilar ones in the nest, thedegree of mimicry will become better as a result andthe host may be more likely to accept the parasiteegg. However, there is as far as we know no evidencefor such a scenario.

2) Decision-making and external cuesDue to recognition errors and rejection costs, hosts

possessing moderate or high intraclutch variation inegg appearance may need additional stimuli to theparasite egg when deciding whether to reject sucheggs or not (box 3, Fig. 1) (Rothstein 1982a; Davies& Brooke 1988; Moksnes & Røskaft 1989; Moksneset al. 1993; Sealy 1995; Alvarez 1996; Brooke et al.1998; Lindholm & Thomas 2000; Moksnes et al.2000; Soler et al. 2000a), especially when they areconfronted with mimetic eggs (Davies et al. 1996).Øien et al. (1999) predicted that species with low(�0%) or high (�100%) rejection rates should showfew costs and no conditional responses in relation torejection behaviour (Hill & Sealy 1994; Soler et al.2000b); their response towards parasite eggs shouldbe more or less fixed. Hosts that show high rejectionrates against foreign eggs and also have a low intra-clutch variation are thus likely to have overcome thecosts associated with recognition and rejection of par-asite eggs, and can therefore be described as winnersof the struggle against brood parasites. Evidence forthis scenario has recently been obtained regardingtwo hosts of the Common Cuckoo in Europe (Stokkeet al. 2002a, 2004; Honza et al. 2004). On the otherhand, hosts with intermediate rejection rates againstforeign eggs should experience larger costs and alsoshow more conditional responses in rejection behav-iour. The Reed Warbler is an intermediate rejecter ofnon-mimetic eggs (Brooke & Davies 1988; Davies &Brooke 1988, 1989b; Stokke et al. 1999; Lindholm &

Thomas 2000), and shows conditional responses inits rejection decisions (Davies & Brooke 1988; Mok-snes et al. 2000), suggesting a genetic basis for rejec-tion behaviour combined with some flexibility in thisbehaviour triggered by clues in the environment. Itseems highly likely that there exists some sort of ge-netic linkage between the evolution of clutch charac-teristics and the ability to recognise odd eggs in ReedWarblers (Stokke et al. 1999).

Zahavi (1979) proposed that acceptance of parasiteeggs can be based upon the existence of costs associ-ated with parasite retaliation or “mafia” behaviour,and Soler et al. (1995) found support for this in theGreat Spotted Cuckoo-Magpie system. In addition,nest-site limitations (Petit 1991) and the stage in thebreeding cycle or length of the breeding season(Moksnes et al. 1990, 1993; Brooker & Brooker1996, 1998) could also explain the acceptance of par-asite eggs in some situations.

3) The genetic foundationThe genetic constitution of the hosts, or in other

words the frequency of different genotypes in a hostpopulation (box 4, Fig. 1), will influence decision-making. Mutations, recombination and natural selec-tion will lead to local adaptations in hosts, but the ef-fect of drift and gene flow will disrupt this selection.The obvious lack of anti-parasite adaptations amongmany hosts today (especially among hosts of theBrown-headed Cowbird) can probably, as discussedabove, be explained by a lag in the evolution of suchtraits (Rothstein 1975a; Hosoi & Rothstein 2000;Peer & Sealy 2004). With a strong selection (highcosts of parasitism and/or high levels of parasitism)and a proper genetic background, hosts are assumedto eventually become good rejecters (i.e. “rejecter al-leles” approach fixation). However it has also beenstressed that rapid changes in host defences withinpopulations as well as differences in defences be-tween populations may reflect phenotypic plasticityrather than or in addition to evolutionary changes(Soler et al. 1994; Brooke et al. 1998; Robert & Sorci1999; Lindholm 2000; Lindholm & Thomas 2000).We can think of cases where it is not necessarilymore adaptive to be a rejecter of parasite eggs than tobe an acceptor (Lotem & Nakamura 1998), perhapswhen host intraclutch variation in egg appearance ishigh and the parasite egg is a good mimic of the hosteggs (Lotem et al. 1995; Davies et al. 1996; Takasu1998a; Rodríguez-Gironés & Lotem 1999). In suchcases, the costs of recognition and rejection can be

23

The enigma of imperfect adaptations in hosts of avian brood parasites

equal to or higher than the costs of acceptance andthus there could be a monomorphic host populationcontaining only acceptors, or a dimorphic populationwith an equilibrium between equally adapted accep-tors and rejecters (Lotem & Nakamura 1998). Thisequilibrium will be very dependent on the selectionpressure enforced upon the hosts for evolving de-fence mechanisms, and in particular the level of para-sitism is a critical factor (Lotem et al. 1992; Takasu etal. 1993; Davies et al. 1996). Thus, the equilibriumstate could be unstable, and the proportion of individ-uals that reject or accept foreign eggs could fluctuate(Soler & Soler 2000). As we have explained in previ-ous sections the situation can be even more compli-cated; when rejection behaviour is age dependent, orwhen hosts show conditional responses in their rejec-tion behaviour. In such cases, rejection is a betterstrategy on average, but the rejecter mutant will sup-press rejection due to rejection costs when the stimu-lus expression is below a critical threshold deter-mined by current perceptual abilities. There could al-ternatively exist a genotypic variability in adjustmentof the conditional response, if different genotypeswith different reaction norms are favoured in variousenvironmental conditions (Lotem & Nakamura1998). A hypothesis that could explain the existenceof both acceptors and rejecters within and amonghost populations based upon the existence of costsconnected to rejection behaviour is the “intermittentarms race hypothesis” (Soler et al. 1998), also termedthe “coevolutionary cycles model” (Rothstein 2001).According to this hypothesis that is based uponmetapopulation dynamics, hosts evolve better de-fences as the level of parasitism increases, and as aconsequence of this the fitness gains for parasites de-crease. Parasites that disperse will therefore have aselective advantage if they encounter host popula-tions that have no previous history of interactionswith the brood parasite. When the level of parasitismdeclines, hosts will lose their defences against para-sitism due to the costs associated with maintainingsuch traits. However, many hosts retain rejection be-haviour for a long time even though they are not cur-rently parasitized, indicating no or negligible costs inretention of such behaviour (Braa et al. 1992; Mok-snes & Røskaft 1992; Rothstein 2001; Stokke et al.2002a, but see also Cruz & Wiley 1989; Robert &Sorci 1999). The loss of rejection behaviour in suchhosts will thus only be due to random genetic drift.Rejection behaviour can even be retained throughspeciation events (Bolen et al. 2000; Rothstein 2001).

This scenario of long time retention of rejection be-haviour is termed the “single trajectory-model”(Rothstein 2001).

CONCLUSIONS: THE VALIDITY OF THE COMPETING HYPOTHESES

The occurrence of acceptor-hosts with a high intra-clutch and low interclutch variation and with no ad-justment of clutch characteristics or rejection behav-iour with age or conditional cues can be explained bythe “evolutionary lag hypothesis”. However, hoststhat have a large intraclutch variation but adjust thistrait as well as the rejection behaviour by learning,experience or conditional cues (due to risks of mak-ing recognition errors, rejection costs, and retaliationor nest-site limitations) may be explained by the“evolutionary equilibrium hypothesis”. However, it isimportant to underline that such an equilibrium statecould be unstable and eventually lead to the evolutionof more fine-tuned defensive mechanisms. This is, ofcourse, exactly what is expected from the lag hypoth-esis.

It is very difficult to separate the lag and the equi-librium hypotheses, and it could be argued that theyreally are not mutually exclusive but instead explaindifferent stages in the arms race (Davies 1999, 2000;Winfree 1999). In situations where the costs of recog-nition and rejection outweigh the costs of acceptance(equilibrium), it can be argued that selection has notyet reduced the costs of rejection (lag) (Winfree1999). If the hosts evolve a lower intraclutch varia-tion in egg appearance or improve their perceptualabilities, the risk of making recognition errors maydecrease, and the equilibrium state is distorted. Asdiscussed above, the “spatial habitat structure hypoth-esis” predicts that the level of parasitism and rejec-tion behaviour in various host population or speciesdepends on characteristics of the hosts’ breedinghabitat (Røskaft et al. 2002b, c). Intermediate rejec-tion rates are not necessarily due to the costs ofrecognition and rejection as stated by the equilibriumhypothesis. Such intermediate rejection rates canrather be due to gene flow of “acceptor alleles” fromnon-parasitized populations into parasitized popula-tions, thus efficiently preventing the evolution ofproper defence in such populations even when rejec-tion is adaptive. In order to separate the “equilib-rium” and “spatial habitat structure” hypotheses, fur-ther studies should focus on obtaining more data onrates of gene flow between populations of parasites

B. G. STOKKE, A. MOKSNES and E. RØSKAFT

24

and hosts as well as data on recognition errors fromnon-parasitized host nests and other costs of main-taining rejection behaviour. Only then will it be pos-sible to solve the enigma of apparently imperfect hostdefences.

ACKNOWLEDGMENTS

We want to thank Tomás Grim for comments on the manu-script. The Research Council of Norway, grant no.151641/432, funded BGS.

REFERENCES

Alvarez F (1993) Proximity of trees facilitates para-sitism by Cuckoos Cuculus canorus on Rufous War-blers Cercotrichas galactotes. Ibis 135: 331.

Alvarez F (1994) A gens of Cuckoo Cuculus canorusparasitizing Rufous Bush Chat Cercotrichas galac-totes. J Avian Biol 25: 239–243.

Alvarez F (1996) Model Cuckoo Cuculus canorus eggsaccepted by Rufous Bush Chats Cercotrichas galac-totes during the parasite’s absence from the breedingarea. Ibis 138: 340–342.

Alvarez F (1999) Attractive non-mimetic stimuli inCuckoo Cuculus canorus eggs. Ibis 141: 142–144.

Amundsen T, Brobakken PT, Moksnes A & Røskaft E(2002) Rejection of Common Cuckoo Cuculuscanorus eggs in relation to female age in theBluethroat Luscinia svecica. J Avian Biol 33: 366–370.

Baker ECS (1942) Cuckoo Problems. HF & G With-erby, London.

Bolen GM, Rothstein SI & Trost CH (2000) Egg recog-nition in Yellow-billed and Black-billed Magpies inthe absence of interspecific parasitism: implicationsfor parasite-host coevolution. Condor 102: 432–438.

Braa AT, Moksnes A & Røskaft E (1992) Adaptationsof Bramblings and Chaffinches towards parasitism bythe Common Cuckoo. Anim Behav 43: 67–78.

Briskie JV, Sealy SG & Hobson KA (1992) Behavioraldefenses against avian brood parasitism in sympatricand allopatric host populations. Evolution 46: 334–340.

Brooke MdL & Davies NB (1988) Egg mimicry byCuckoos Cuculus canorus in relation to discrimina-tion by hosts. Nature 335: 630–632.

Brooke MdL, Davies NB & Noble DG (1998) Rapid de-cline of host defences in response to reduced Cuckooparasitism: behavioural flexibility of Reed Warblersin a changing world. Proc R Soc Lond B 265: 1277–1282.

Brooker L & Brooker M (1998) Why do Splendid

Fairy-wren always accept cuckoo eggs? Behav Ecol9: 420–424.

Brooker M & Brooker L (1996) Acceptance by theSplendid Fairy-wren of parasitism by Horsfield’sBronze-cuckoo: further evidence for evolutionaryequilibrium in brood parasitism. Behav Ecol 7: 395–407.

Chance E (1940) The truth about the Cuckoo. CountryLife, London.

Clotfelter ED (1998) What cues do Brown-headed Cow-birds use to locate Red-winged Blackbird host nests?Anim Behav 55: 1181–1189.

Cruz A & Wiley JW (1989) The decline of an adapta-tion in the absence of a presumed selection pressure.Evolution 43: 55–62.

Darwin C (1859) The origin of species. Murray, Lon-don.

Davies NB (1999) Cuckoos and cowbirds versus hosts:Co-evolutionary lag and equilibrium. Ostrich 70: 71–79.

Davies NB (2000) Cuckoos, cowbirds and other cheats.T & AD Poyser, London.

Davies NB & Brooke MdL (1988) Cuckoos versus ReedWarblers: adaptations and counteradaptations. AnimBehav 36: 262–284.

Davies NB & Brooke MdL (1989a) An experimentalstudy of co-evolution between the Cuckoo, Cuculuscanorus, and its hosts. I Host egg discrimination. JAnim Ecol 58: 207–224.

Davies NB & Brooke MdL (1989b) An experimental-study of co-evolution between the Cuckoo, Cuculuscanorus, and its hosts. II Host egg markings, chickdiscrimination and general discussion. J Anim Ecol58: 225–236.

Davies NB & Brooke MdL (1998) Cuckoos versushosts. In: Rothstein SI & Robinson SK (eds) Para-sitic birds and their hosts: Studies in coevolution, pp59–79. Oxford University Press, New York.

Davies NB, Brooke MdL & Kacelnik A (1996) Recog-nition errors and probability of parasitism determinewhether Reed Warblers should accept or rejectmimetic Cuckoo eggs. Proc R Soc Lond B 263: 925–931.

Dawkins R & Krebs JR (1979) Arms races between andwithin species. Proc R Soc Lond B 205: 489–511.

Friedmann H (1963) Host relations of the parasitic cow-birds, Bull 233. Smithsonian Institution, UnitedStates National Museum, Washington DC.

Friedmann H (1971) Further information on the host re-lations of the parasitic cowbirds. Auk 88: 239–255.

Friedmann H & Kiff LF (1985) The parasitic cowbirdsand their hosts. Proc West Found Vertebr Zool 2:225–302.

25

The enigma of imperfect adaptations in hosts of avian brood parasites

Friedmann H, Kiff LF & Rothstein SI (1977) A furthercontribution to knowledge of the host relations of theparasitic cowbirds. Smithsonian Contrib Zool 235:1–75.

Gibbs HL, Sorenson MD, Marchetti K, Brooke MdL,Davies NB & Nakamura H (2000) Genetic evidencefor female host-specific races of the CommonCuckoo. Nature 407: 183–186.

Gosler AG, Barnett PR & Reynolds SJ (2000) Inheri-tance and variation in eggshell patterning in the GreatTit Parus major. Proc R Soc Lond B 267:2469–2473.

Grim T (2002) Why is mimicry in Cuckoo eggs some-times so poor? J Avian Biol 33: 302–305.

Grim T, Kleven O & Mikulica O (2003) Nestling dis-crimination without recognition: a possible defencemechanism for hosts towards Cuckoo parasitism?Proc R Soc Lond B 270: S73–S75.

Hill DP & Sealy SG (1994) Desertion of nests para-sitized by cowbirds; have Clay-coloured Sparrowsevolved an anti-parasite defence? Anim Behav 48:1063–1070.

Holt RD (1996) Adaptive evolution in source-sink envi-ronments: direct and indirect effects of density-de-pendence on niche evolution. Oikos 75: 182–192.

Honza M, Prochazka P, Stokke BG, Moksnes A,Røskaft E, Capek MJr & Mrlík V (2004) Are Black-caps current winners in the evolutionary struggleagainst the Common Cuckoo? J Ethol 22: 175–180.

Hosoi SA & Rothstein SI (2000) Nest desertion andcowbird parasitism: evidence for evolved responsesand evolutionary lag. Anim Behav 59: 823–840.

Janzen DH (1980) When is it coevolution? Evolution34: 611–612.

Johnson KP & Lanyon SM (1999) Molecular systemat-ics of the grackles and allies, and the effect of addi-tional sequence (cyt B and ND2). Auk 116: 759–768.

Jourdain FCR (1925) A study on parasitism in the cuck-oos. Proc Zool Soc Lond 1925: 639–667.

Kaltz O & Shykoff JA (1998) Local adaptation in host-parasite systems. Heredity 81: 361–370.

Kemal RE & Rothstein SI (1988) Mechanisms of avianegg recognition: adaptive responses to eggs with bro-ken shells. Anim Behav 36: 175–183.

Lack D (1968) Ecological adaptations for breeding inbirds. Methuen, London.

Lahti DC & Lahti AR (2002) How precise is egg dis-crimination in weaverbirds? Anim Behav 63: 1135–1142.

Lanyon SM (1992) Interspecific brood parasitism inblackbirds (Icterinae): a phylogenetic perspective.Science 255: 77–79.

Lerkelund HE, Moksnes A, Røskaft E & Ringsby TH

(1993) An experimental test of optimal clutch size ofthe Fieldfare; with a discussion on why brood para-sites remove eggs when they parasitize a host species.Ornis Scand 24: 95–102.

Levins R (1969) Some demographic and genetic conse-quences of environmental heterogeneity for biologi-cal control. Bull Entomol Assoc Am 15: 237–240.

Lindholm AK (1999) Phenotypic plasticity in ReedWarbler defence against brood parasitism. In: AdamsNJ & Slotow RH (eds) Proc 22 Int Ornithol Congr.pp 3107–3124. BirdLife South Africa, Johannesburg.

Lindholm AK (2000) Tests of phenotypic plasticity inReed Warbler defences against Cuckoo parasitism.Behaviour 137: 43–60.

Lindholm AK & Thomas RJ (2000) Differences be-tween populations of Reed Warblers in defencesagainst brood parasitism. Behaviour 137: 25–42.

Lorenzana JC & Sealy SG (2001) Fitness costs and ben-efits of Cowbird egg ejection by Gray Catbirds.Behav Ecol 12: 325–329.

Lotem A & Nakamura H (1998) Evolutionary equilibriain avian brood parasitism. In: Rothstein SI & Robin-son SK (eds) Parasitic birds and their hosts: Studiesin coevolution. pp 223–235. Oxford Univ Press, NewYork.

Lotem A, Nakamura H & Zahavi A (1992) Rejection ofCuckoo eggs in relation to host age: a possible evolu-tionary equilibrium. Behav Ecol 3: 128–132.

Lotem A, Nakamura H & Zahavi A (1995) Constraintson egg discrimination and Cuckoo-host co-evolution.Anim Behav 49: 1185–1209.

Lotem A & Rothstein SI (1995) Cuckoo-host coevolu-tion: from snapshots of an arms race to the documen-tation of microevolution. Trends Ecol Evol 10: 436–437.

Marchetti K (1992) Costs to host defence and the per-sistence of parasitic cuckoos. Proc R Soc Lond B248: 41–45.

Marchetti K (2000) Egg rejection in a passerine bird:size does matter. Anim Behav 59: 877–883.

Martinez JG, Soler JJ, Soler M, Møller AP & Burke T(1999) Comparative population structure and geneflow of a brood parasite, the Great Spotted Cuckoo(Clamator glandarius), and its primary host, theMagpie (Pica pica). Evolution 53: 269–278.

Mayfield H (1965) The Brown-headed Cowbird, withold and new hosts. Living Birds 4: 13–28.

Mayfield HF (1977) Brown-headed Cowbird: agent ofextermination? Am Birds 31: 107–113.

McLean IG & Maloney RF (1998) Brood parasitism,recognition, and response. In: Rothstein SI & Robin-son SK (eds) Parasitic birds and their hosts: Studiesin coevolution. pp 255–269. Oxford Univ Press, New

B. G. STOKKE, A. MOKSNES and E. RØSKAFT

26

York.Moksnes A (1992) Egg recognition in Chaffinches and

Bramblings. Anim Behav 44: 993–995.Moksnes A & Røskaft E (1989) Adaptations of Meadow

Pipits to parasitism by the Common Cuckoo. BehavEcol Sociobiol 24: 25–30.

Moksnes A & Røskaft E (1992) Responses of some rareCuckoo hosts to mimetic model Cuckoo eggs and toforeign conspecific eggs. Ornis Scand 23: 17–23.

Moksnes A & Røskaft E (1995) Egg-morphs and hostpreference in the Common Cuckoo (Cuculuscanorus): an analysis of cuckoo and host eggs fromEuropean museum collections. J Zool, Lond 236:625–648.

Moksnes A, Røskaft E, Braa AT, Korsnes L, Lampe HM& Pedersen HC (1990) Behavioural responses of po-tential hosts towards artificial Cuckoo eggs and dum-mies. Behaviour 116: 64–89.

Moksnes A, Røskaft E, Hagen LG, Honza M, Mørk C &Olsen PH (2000) Common Cuckoo Cuculus canorusand host behaviour at Reed Warbler Acrocephalusscirpaceus nests. Ibis 142: 247–258.

Moksnes A, Røskaft E & Korsnes L (1993) Rejection ofCuckoo (Cuculus canorus) eggs by Meadow Pipits(Anthus pratensis). Behav Ecol 4: 120–127.

Molnàr B (1944) The Cuckoo in the Hungarian plain.Aquila 51: 100–112.

Nager RG, Monaghan P & Houston DC (2001) The costof egg production: increased egg production reducesfuture fitness in gulls. J Avian Biol 32: 159–166.

Øien IJ, Honza M, Moksnes A & Røskaft E (1996) Therisk of parasitism in relation to the distance fromReed Warbler nests to Cuckoo perches. J Anim Ecol65: 147–153.

Øien IJ, Moksnes A & Røskaft E (1995) Evolution ofvariation in egg color and marking pattern in Euro-pean passerines: adaptations in a coevolutionary arms race with the Cuckoo, Cuculus canorus. BehavEcol 6: 166–174.

Øien IJ, Moksnes A, Røskaft E, Edvardsen E, Honza M,Kleven O & Rudolfsen G (1999) Conditional host re-sponses to Cuckoo Cuculus canorus parasitism. In:Adams NJ, Slotow RH (eds) Proc 22 Int OrnitholCongr. pp 3125–3145. BirdLife South Africa, Johan-nesburg.

Øien IJ, Moksnes A, Røskaft E & Honza M (1998)Costs of Cuckoo Cuculus canorus parasitism to ReedWarblers Acrocephalus scirpaceus. J Avian Biol 29:209–215.

Payne R (1997) Avian brood parasitism. In: Clayton DH& Moore J (eds) Host-parasite evolution: generalprinciples and avian models. pp 338–369. OxfordUniv Press, Oxford.

Payne RB (1973a) Behavior, mimetic songs and songdialects, and relationships of the parasitic indigobirds(Vidua) of Africa. In: Ornithological Monographs 11.p 333. American Ornithological Union, WashingtonDC.

Payne RB (1973b) Vocal mimicry of the Paradise Why-dahs (Vidua) and response of female whydahs to thesongs of their hosts (Pytilia) and their mimics. AnimBehav 21: 762–771.

Payne RB (1976) Song mimicry and species relation-ships among the West African Pale-winged Indigob-irds. Auk 93: 25–38.

Payne RB & Payne LL (1998) Brood parasitism byCowbirds: risks and effects on reproductive successand survival in Indigo Buntings. Behav Ecol 9:64–73.

Peer BD & Sealy SG (2004) Correlates of egg rejectionin hosts of the Brown-headed Cowbird. Condor 106:580–599.

Petit LJ (1991) Adaptive tolerance of Cowbird para-sitism by Prothonotary Warblers: a consequence ofnest-site limitation? Anim Behav 41: 425–432.

Petrie M & Møller AP (1991) Laying eggs in others’nests; intraspecific brood parasitism in birds. TrendsEcol Evol 6: 315–320.

Power HW (1998) Quality control and the importantquestions in avian conspecific brood parasitism. In:Rothstein SI & Robinson SK (eds) Parasitic birdsand their hosts: Studies in coevolution. pp 376–395.Oxford Univ Press, New York.

Robert M & Sorci G (1999) Rapid increase of host de-fence against brood parasites in a recently parasitizedarea: the case of Village Weavers in Hispaniola. ProcR Soc Lond B 266: 941–946.

Rodríguez-Gironés MA & Lotem A (1999) How to de-tect a cuckoo egg: a signal-detection theory model forrecognition and learning. Am Nat 153: 633–648.

Roff DA (2002) Life history evolution. Sinauer, Sunder-land MA.

Rohwer S & Spaw CD (1988) Evolutionary lag versusbill-size constraints: a comparative study of the ac-ceptance of cowbird eggs by old hosts. Evol Ecol 2:27–36.

Rohwer S, Spaw CD & Røskaft E (1989) Costs toNorthern Orioles of puncture-ejecting parasitic Cow-bird eggs from their nests. Auk 106: 734–738.

Røskaft E & Moksnes A (1998) Coevolution betweenbrood parasites and their hosts. In: Rothstein SI &Robinson SK (eds) Parasitic birds and their hosts:Studies in coevolution. pp 236–254. Oxford UnivPress, New York.

Røskaft E, Moksnes A, Meilvang D, Bicìk V,Jemelìková J & Honza M (2002a) No evidence for

27

The enigma of imperfect adaptations in hosts of avian brood parasites

recognition errors in Acrocephalus warblers. J AvianBiol 33: 31–38.

Røskaft E, Moksnes A, Stokke BG, Bicìk V & MoskátC (2002b) Aggression to dummy Cuckoos by poten-tial European Cuckoo hosts. Behaviour 139: 613–628.

Røskaft E, Moksnes A, Stokke BG, Moskát C & HonzaM (2002c) The spatial habitat structure of host popu-lations explains the pattern of rejection behavior inhosts and parasitic adaptations in cuckoos. BehavEcol 13: 163–168.

Røskaft E, Orians GH & Beletsky LD (1990) Why doRed-winged Blackbirds accept eggs of Brown-headedCowbirds? Evol Ecol 4: 35–42.

Røskaft E, Rohwer S & Spaw CD (1993) Cost of punc-ture ejection compared with costs of rearing cowbirdchicks for Northern Orioles. Ornis Scand 24: 28–32.

Rothstein SI (1974) Mechanisms of avian egg recogni-tion: possible learned and innate factors. Auk 91:796–807.

Rothstein SI (1975a) Evolutionary rates and host de-fenses against avian brood parasitism. Am Nat 109:161–176.

Rothstein SI (1975b) An experimental and teleonomicinvestigation of avian brood parasitism. Condor 77:250–271.

Rothstein SI (1976) Experiments on defenses CedarWaxwings use against Cowbird parasitism. Auk 93:675–691.

Rothstein SI (1977) Cowbird parasitism and egg recog-nition of the Northern Oriole. Wilson Bull 89: 21–32.

Rothstein SI (1978) Mechanisms of avian egg-recogni-tion: additional evidence for learned components.Anim Behav 26: 671–677.

Rothstein SI (1982a) Mechanisms of avian egg recogni-tion: which egg parameters elicit responses by re-jecter species? Behav Ecol Sociobiol 11: 229–239.

Rothstein SI (1982b) Successes and failures in avianegg and nestling recognition with comments on theutility of optimality reasoning. Am Zool 22: 547–560.

Rothstein SI (1990) A model system for coevolution:avian brood parasitism. Annu Rev Ecol Syst 21:481–508.

Rothstein SI (2001) Relic behaviours, coevolution andthe retention versus loss of host defences afterepisodes of avian brood parasitism. Anim Behav 61:95–107.

Rothstein SI, Patten MA & Fleischer RC (2004) Phy-logeny, specialization, and brood parasite-host coevo-lution: some possible pitfalls of parsimony. BehavEcol 13: 1–10.

Rothstein SI & Robinson SK (1998) Parasitic birds andtheir hosts: studies in coevolution. Oxford Univ

Press, New York.Sealy SG (1992) Removal of Yellow Warbler eggs in as-

sociation with Cowbird parasitism. Condor 94:40–54.

Sealy SG (1995) Burial of Cowbird eggs by parasitizedYellow Warblers: an empirical and experimentalstudy. Anim Behav 49: 877–889.

Soler JJ, Martinez JG, Soler M & Møller AP (1999) Ge-netic and geographic variation in rejection behaviorof Cuckoo eggs by European Magpie populations: anexperimental test of rejecter-gene flow. Evolution 53:947–956.

Soler M & Møller AP (1990) Duration of sympatry andcoevolution between the Great Spotted Cuckoo andits Magpie host. Nature 343: 748–750.

Soler JJ & Møller AP (1996) A comparative analysis ofthe evolution of variation in appearance of eggs ofEuropean passerines in relation to brood parasitism.Behav Ecol 7: 89–94.

Soler M, Palomino JJ, Martìn-Vivaldi M & Soler JJ(2000a) Lack of consistency in the response of Ru-fous-tailed Scrub Robins Cercotrichas galactotes to-wards parasitic Common Cuckoo eggs. Ibis 142:151–154.

Soler JJ & Soler M (2000) Brood-parasite interactionsbetween Great Spotted Cuckoos and Magpies: amodel system for studying coevolutionary relation-ships. Oecologia 125: 309–320.

Soler M, Soler JJ, Martinez JG & Møller AP (1994)Micro-evolutionary change in host response to abrood parasite. Behav Ecol Sociobiol 35: 295–301.

Soler M, Soler JJ, Martinez JG & Møller AP (1995)Magpie host manipulation by Great Spotted Cuckoos:evidence for an avian mafia? Evolution 49: 770–775.

Soler M, Soler JJ, Martinez JG, Peréz-Contreras T &Møller AP (1998) Micro-evolutionary change andpopulation dynamics of a brood parasite and its pri-mary host: the intermittent arms race hypothesis. Oe-cologia 117: 381–390.

Soler M, Soler JJ & Møller AP (2000b) Effect of GreatSpotted Cuckoo presence on Magpie rejection behav-iour. Behaviour 137: 213–220.

Stearns SC (1992) The evolution of life histories. Ox-ford Univ Press, Oxford.

Stokke BG, Honza M, Moksnes A, Røskaft E & Rudolf-sen G (2002a) Costs associated with recognition andrejection of parasitic eggs in two European passer-ines. Behaviour 139: 629–644.

Stokke BG, Moksnes A & Røskaft E (2002b) Obligatebrood parasites as selective agents for evolution ofegg appearance in passerine birds. Evolution 56:199–205.

Stokke BG, Moksnes A, Røskaft E, Rudolfsen G &

B. G. STOKKE, A. MOKSNES and E. RØSKAFT

28

Honza M (1999) Rejection of artificial Cuckoo (Cu-culus canorus) eggs in relation to variation in egg ap-pearance among Reed Warblers (Acrocephalus scir-paceus). Proc R Soc Lond B 266: 1483–1488.

Stokke BG, Rudolfsen G, Moksnes A & Røskaft E(2004) Rejection of conspecific eggs in Chaffinches:the effect of age and clutch characteristics. Ethology110: 459–470.

Swynnerton CFM (1918) Rejections by birds of eggsunlike their own: with remarks on some of theCuckoo problems. Ibis Tenth Series 6: 127–154.

Takasu F (1998a) Modelling the arms race in avianbrood parasitism. Evol Ecol 12: 969–987.

Takasu F (1998b) Why do all host species not show de-fense against avian brood parasitism: evolutionary lagor equilibrium? Am Nat 151: 193–205.

Takasu F, Kawasaki K, Nakamura H, Cohen JE &Shigesada N (1993) Modeling the population dynam-ics of a cuckoo-host association and the evolution of

host defenses. Am Nat 142: 819–839.Thompson JN (1994) The coevolutionary process. The

Univ of Chicago Press, Chicago.Victoria JK (1972) Clutch characteristics and egg dis-

criminative ability of the African Village WeaverbirdPloceus cucullatus. Ibis 114: 367–376.

Visser ME & Lessells CM (2001) The costs of egg pro-duction and incubation in Great Tits (Parus major).Proc R Soc Lond B 268: 1271–1277.

Weatherhead PJ & Forbes MRL (1994) Natal philopatryin passerine birds: genetic or ecological influences?Behav Ecol 5: 426–433.

Welbergen J, Komdeur J, Kats R & Berg M (2001) Eggdiscrimination in the Australian Reed Warbler (Acro-cephalus australis): rejection response toward modeland conspecific eggs depending on timing and modeof artificial parasitism. Behav Ecol 12: 8–15.

Winfree R (1999) Cuckoos, cowbirds and the persist-ence of brood parasitism. Trends Ecol Evol 14:338–343.

Wyllie I (1981) The Cuckoo. Universe Books, NewYork.

Zahavi A (1979) Parasitism and nest predation in para-sitic cuckoos. Am Nat 113: 157–159.

29

The enigma of imperfect adaptations in hosts of avian brood parasites