Behavioral Ecology The oicial journal of the ISBE International Society for Behavioral Ecology Behavioral Ecology (2016), 27(5), 1405–1412. doi:10.1093/beheco/arw061 Original Article Geographic variation in egg ejection rate by great tits across 2 continents Wei Liang,a Anders Pape Møller,b Bård Gunnar Stokke,c Canchao Yang,a Petr Kovařík,d Haitao Wang,e Cheng-Te Yao,f Ping Ding,g Xin Lu,h Arne Moksnes,c Eivin Røskaft,c and Tomáš Grimi aMinistry of Education Key Laboratory for Tropical Plant and Animal Ecology, College of Life Sciences, Hainan Normal University, 99 South Longkun Road, Haikou 571158, China, bEcologie Systématique Evolution, CNRS, Université Paris-Sud, AgroParisTech, Université Paris-Saclay, F-91400 Orsay, France, cDepartment of Biology, Norwegian University of Science and Technology, Realfagbygget (EU2-140), NO-7491 Trondheim, Norway, dNature Conservation Agency of the Czech Republic, Administration of Litovelské Pomoraví Protected Landscape Area, Husova 5, CZ-784 01 Litovel, Czech Republic, eSchool of Life Sciences, Northeast Normal University, 5268 Renmin Street, Changchun 130024, China, fMedium Altitude Experimental Station, Endemic Species Research Institute, Chichi, 15 Nantou 552, Taiwan, gCollege of Life Sciences, Zhejiang University, 866 Yuhangtang Road, Hangzhou 310058, China, hCollege of Life Sciences, Wuhan University, 299 Bayi Road, Wuhan 430072, China, and iDepartment of Zoology and Laboratory of Ornithology, Palacký University, 17. listopadu 50, CZ-771 46 Olomouc, Czech Republic Received 19 January 2015; revised 11 March 2016; accepted 21 March 2016; Advance Access publication 21 April 2016. Hosts of brood parasites may vary geographically in their ability to resist parasitism. In contrast, geographic variation in defenses, such as egg rejection, is not expected to be present or vary geographically in unsuitable hosts. We examined spatial patterns of resistance in the great tit Parus major, a passerine that is a textbook example of an unsuitable host for brood parasites because of its hole-nesting habits. We experimentally tested for spatial variation in foreign egg rejection in 395 nests across latitudinal gradients in China (5 populations) and Europe (7 populations). In China, egg rejection rates were very high but showed a latitudinal gradient from 100% in the south to 52% in the north. In Europe, rejection rates were very low (on average only 4%) and did not vary latitudinally. The egg ejection rate patterns matched geographic patterns of parasitism risk with rejection probabilities decreasing with latitude (a surrogate measure of the diversity of brood parasites). The present study for the first time challenges the idea that hole-nesting birds did not evolve resistance mechanisms against brood parasites and highlights the importance of large-scale geographic comparisons in ecological research. Key words: allopatry, conspecific parasitism, interspecific parasitism, life history traits, sympatry, trait loss. INTRODUCTION Parasites exploit variable numbers of hosts, ranging from strict specialists to generalists that may parasitize more than 100 diferent host species. Such patterns arise from coevolutionary interactions between hosts and parasites (Combes 2001). However, parasites may not exploit potential hosts if such hosts occupy niches that exclude parasites. For example, small-sized cavities used by birds as nest sites may prevent parasitism by larger parasitic cuckoos that Address correspondence to W. Liang. E-mail: liangwei@hainnu.edu.cn. © The Author 2016. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com cannot enter the nest hole, and small hole-nesters prefer smaller cavities as nest sites over larger ones (van Balen et al. 1982; Carlson et al. 1998). A parallel case of avoidance of parasitism is the close association between potential bird hosts and human habitation that can reduce the risk of brood parasitism because most parasitic cuckoos avoid close proximity to humans, especially in urbanized areas (Grim et al. 2011; Liang et al. 2013; Møller et al. 2016). Host resistance to parasitism is often consistent among populations (e.g., Grim et al. 2011). Thus, Moksnes et al. (1991), Moksnes and Røskaft (1995), and Davies (2000) classiied most hole-nesting bird species as “unsuitable” for cuckoo parasitism, with the sole 1406 exception of the hole-nesting redstart Phoenicurus phoenicurus that is a common cuckoo host, apparently because cavities used by this species generally have large entrances that allow the cuckoo female to enter the nest and the cuckoo chick to ledge (von Haartman 1981; Grim et al. 2009). In contrast, a recent comparative analysis by Medina and Langmore (2015) showed that nest type was not predictive of egg rejection rates in hosts of brood parasites, and even tits and other hole-nesters often use natural cavities with large entrances (van Balen et al. 1982) that should readily allow a cuckoo to lay its egg in such nests. Tits sometimes even reuse open nests of other species (Monrsós et al. 1999) although such instances are rare. Diferent populations of the same species were often found to be consistently acceptors or rejecters of model cuckoo eggs (Davies 2000). However, Møller and Soler (2012) recently reviewed the literature on consistency in host responses to brood parasites and found large intraspeciic variation in resistance in several host species (Cruz and Wiley 1989; Nakamura 1990; Briskie et al. 1992; Lindholm 1999; Soler, Martínez, et al. 1999; Moskát et al. 2002; Martín-Gálvez et al. 2006, 2007; Moskát et al. 2008; Stokke et al. 2008; Avilés et al. 2011; Soler et al. 2011; Langmore et al. 2012). These studies indicate that species-speciic consistency in resistance is far from the rule. This inding may also cast doubts on traditional categorization of host species as either suitable or unsuitable. A prime example of a group of apparently unsuitable hosts is species belonging to the families of tits (Paridae), lycatchers (Muscicapidae), treecreepers (Certhiidae), and nuthatches (Sittidae): these birds have an insect diet that should be optimal for raising a cuckoo chick (but see Yang et al. 2013), but they breed in holes which may not be accessible to cuckoos (Davies 2000). In a recent compilation of cases of brood parasitism by the common cuckoo Cuculus canorus in Europe (Møller et al. 2011), only 112 cases of parasitism in these hole-nesters were detected from a sample of 57 957 cases of parasitism (see also Grim et al. 2014). Within the above-outlined subset of hosts, the most common host species is the pied lycatcher Ficedula hypoleuca with 59 cases of parasitism, almost exclusively from Finland from the irst half of the 20th century (Stokke BG, et al., unpublished data). Although cuckoo eggs found in pied lycatcher nests are blue when perceived with human sight, such as those of the sympatric host race of the redstart, cuckoo eggs are in fact spectrally, perceptually, and chemically more similar to redstart than lycatcher eggs (Igic et al. 2012). This and the absence of rejection of even highly nonmimetic eggs by the pied lycatcher (von Haartman 1976; Davies and Brooke 1989) suggest that the recent reduction in the frequency of cuckoo parasitism in the pied lycatcher is caused by a large increase in the number of nest boxes for lycatchers that do not allow the cuckoo to lay its egg (Grim et al. 2014). This suggests that similar patterns may apply to other hole-nesters, like tits. Davies and Brooke (1989) documented a low rate of egg rejection by great tits Parus major (17%, N = 12 nests). However, all eggs rejected by great tits in this experiment were deserted. Because Davies and Brooke (1989) did not use nonmanipulated control nests, their data cannot be used to support the hypothesis that tits recognized and rejected foreign eggs. Indeed, Kempenaers et al. (1995) showed that desertion was similar in experimental and control tit nests (12% and 19% nests, respectively). Therefore, desertion cannot be considered an antiparasitic response in tits in this case, and these indings suggest that it is very likely that tits, at least in Europe, are pure acceptors of cuckoo eggs. However, the great tit’s range includes Asia and nothing is known about Asian tits responses to brood parasitism. Recent experimental evidence (Grim et al. 2014) shows that tits have better capacity Behavioral Ecology to raise the cuckoo chick (measured as growth rate) than any other currently suitable cuckoo host (Grim 2006). Also the number of cuckoo chicks that successfully ledged from naturally parasitized tit holes with large entrances is not negligible (Grim et al. 2014). These patterns, taken together, suggest that there is a potential for tits to be involved in interactions with cuckoos, especially in areas where smaller body-sized cuckoos live and where natural holes with large opening do not prevent cuckoos from parasitizing tits. This suggests that studying non-European tit populations may provide important insights into tit–cuckoo interactions speciically and host selection by parasitic birds generally. Therefore, the objectives of the present study were to reexamine the egg rejection capacity of the great tit, which constitutes a textbook example of an unsuitable host due to its use of holes as nest sites. We 1) tested for spatial heterogeneity in rejection behavior across 2 large-scale latitudinal gradients, which difered in number of sympatric brood parasites (China with multiple cuckoo species of various body sizes and Europe with a single cuckoo species of large body size); 2) assessed the extent to which diferent species of brood parasites afected the rates of egg rejection across multiple tit populations; and 3) tested whether ejection rate increased with the diversity of cuckoo species. We tested for rejection behavior in 395 nests of great tits across 12 populations from the Czech Republic, Norway, and Denmark in the west to China in the east. We recorded the presence or the absence of the 3 most common species of cuckoos at our study sites in China and the only cuckoo species in Europe to test whether rejection behavior could be predicted by the presence or the absence of any speciic cuckoo species. In addition, we tested if ejection behavior could be predicted by the local diversity of cuckoo species estimated as the total number of cuckoo species in each study site. We also included latitude as an additional predictor for 5 reasons. First, a recent study showed that latitude is a good predictor of egg rejection behavior in general (Medina and Langmore 2015). Second, cuckoos are typically very secretive and hard to detect (Erritzøe et al. 2012), and thus our estimates of local cuckoo presence may be imprecise (underestimated). Third, hosts typically show some dispersal and thus their responses in any particular study site (which we sampled) would relect larger-scale geographical patterns of cuckoo diversity (which we did not sample) (Soler, Martínez, et al. 1999). Fourth, brood parasite–host interactions are typically dynamic, with repeated local extinction and recolonization of both parasites and hosts. Therefore, point estimates of parasite presence often may not relect relevant long-term evolutionary pressures: currently parasitized populations may in fact be allopatric, whereas currently parasite-free population may in fact be sympatric at coevolutionary time scales (Thorogood and Davies 2013). Fifth, many factors other than brood parasitism may afect geographic patterns of host response to brood parasitism, and hence we included latitude as an additional variable to control for such potentially confounding efects. Therefore, we believe that latitude better represents potential parasite pressures on hosts than empirically but unreliably detected parasite presence/ absence. As any particular population may be deviant and not typical of general patterns, we focused on meta-replication, i.e., replication across multiple populations within species (Johnson 2002; Stokke et al. 2008; Grim et al. 2011). We predicted that egg rejection should be detected in China but not in Europe, and that population-level rejection rates should positively covary with cuckoo species diversity. Liang et al. • Geographic variation in egg ejection and diversity of brood parasites MATERIALS AND METHODS Study sites 1407 In 2008–2010, we collected extensive data in 12 populations in Norway, Denmark, the Czech Republic, and China (Figure 1; Supplementary Table 1). All study sites were forests or open woodland, sometimes bordering on urban environments. For detailed descriptions of all the study sites, see Kleven et al. (2007), Krist (2009), Yang et al. (2012, 2013), and Matrková and Remeš (2012). commixtus (Qiandaohu), and P. major hainanus (Diaoluoshan) in China and P. major major in Europe (Figure 1; Kvist et al. 2003; Päckert et al. 2005; Pavlova et al. 2006; Zhao et al. 2012; Johansson et al. 2013). The actual body sizes of the tits and their subspecies (Li et al. 1982; Cramp and Perrins 1993) are provided in Supplementary Table 3, as a recent study has shown that egg rejection is more likely to evolve when the parasite is relatively large compared with its host (Medina and Langmore 2015). We did not resolve whether P. major minor should be classiied as a separate species from P. major. Potential host species Potential brood parasites We experimentally tested the egg discrimination abilities of great tits (Figure 2; Supplementary Table 2). We sampled populations of P. major minor (Zuojia, Xiaolongmen, and Dongzhai), P. major In Europe, only the common cuckoo was sympatric in some populations. In China, the 3 most common sympatric brood parasites were the common cuckoo, Himalayan cuckoo Cuculus saturatus, and Vikhammer Zuoujia Stjørdal Xiaolongmen Røros Brovst Dongzhai Aalborg Qiandaohu Figure 1 Locations of study sites. Circles and squares refer to study sites in Europe and Asia, respectively. Sumperk Diaoluoshan Velky Kosir Behavioral Ecology 1408 Figure 2 Example of an experimental nest of the great tit with host eggs and a blue nonmimetic model egg. Photo by P. Kovařík. little cuckoo C. poliocephalus. We recorded their presence/absence in the study sites in China. The actual sizes of the cuckoos and their subspecies (Cheng et al. 1991) are provided in Supplementary Table 3. Other cuckoo species are present in some of the study sites (Yang et al. 2012, 2013); therefore, we in addition estimated diversity of brood parasites as the number of cuckoo species in each of our study sites. Egg experiments We used several treatments to robustly test for host egg discrimination abilities and to disentangle their origins (for speciic rationales see below). We employed blue (Figure 2) or white model eggs made of clay of a size and mass similar to that of Asian emerald cuckoo Chrysococcyx maculatus eggs (21.1–24.0 × 15.7–17.4 mm; 3.1–3.8 g, N = 65 model eggs) (Figure 2). All models were manufactured by a single person and consistently painted with the same artiicial nontoxic colors. Therefore, interpopulation diferences we recorded cannot be an artifact of variation in the cues presented to birds— models were the same across all study populations. In European populations, it turned out that the clay models were almost always accepted (see Results for details). Acceptance of hard nonpuncturable models may represent a methodological artifact in hosts that are unable to both puncture the clay or grasp the models due to having small bills (Martín-Vivaldi et al. 2002). Therefore, we performed additional experiments where we painted a randomly chosen host egg dark blue with a nontoxic marker (following Hauber et al. 2014; hereafter blue conspeciic treatment). Such experimental eggs were even more dissimilar to host eggs than blue and white artiicial models but were puncturable and thus provided a strong test of host egg discrimination abilities. Further, we employed another treatment where we added a single conspeciic egg to a focal host nest. Finally, some nests were only visited and eggs handled in the same way as in experimental nests (all treatments above) but no experimental eggs were added (hereafter: control nests). Throughout we followed standard protocols to ensure that our results are quantitatively directly comparable with previous studies of the common cuckoo. Speciically, experimental eggs were added to nests during the laying or early incubation period (most cuckoo hosts do not respond to foreign eggs diferently between laying vs. incubation stages: Davies and Brooke 1989; Moksnes et al. 1991; Grim et al. 2011). We removed a single host egg (like cuckoos do) in nests where an artiicial model or a conspeciic egg was added; we note that such removal has no efect on host responses (Davies and Brooke 1989; this study). Although we visited some nests daily (depending on logistic constraints), we managed to revisit some other nests only after 6 days to check host responses (6 days is the standard criterion to assess egg acceptance: Davies and Brooke 1989; Moksnes et al. 1991; Grim et al. 2011). This prevented us from assessing exact latency to rejection but allowed us to score the host responses as follows: 1) ejection when the model egg was ejected from the nest and hence was missing, whereas the hosts’ own eggs were incubated; 2) burial when the model egg was buried in the nest material; and 3) desertion when the clutch was left with cold eggs, and there were no signs of host presence. We employed multiple experimental treatments because use of some particular egg types may fail to reveal a realistic picture of host egg discrimination abilities (Hauber et al. 2015). First, for puncture ejecter hosts hard plastic model eggs may be impossible to eject (Martín-Vivaldi et al. 2002). Second, eggs that are too similar to host eggs may be accepted by hosts despite the ability of hosts to reject more dissimilar eggs (e.g., Hauber et al. 2014). Therefore, the use of too “mimetic” experimental eggs can lead to the erroneous conclusion that the particular host did not evolve an ability to reject foreign eggs (Grim 2005). Third, a host species may reject relatively nonmimetic eggs, but this may just be a by-product of adaptations that evolved in the context of conspeciic parasitism (López-de-Hierro and Moreno-Rueda 2010; Samas et al. 2014), although this latter problem is unlikely to be important in the present study species (Kempenaers et al. 1995; review in Griith et al. 2002). Fourth, experimental nests may be deserted not because of the introduction of experimental eggs but due to any unrelated disturbance; therefore, it is necessary to use control (unmanipulated) nests to determine whether desertion represents a speciic response to parasitism (Samas et al. 2014). For these reasons, it is necessary to use conspeciic, mimetic and nonmimetic experimental eggs and Liang et al. • Geographic variation in egg ejection and diversity of brood parasites control nests to reveal both host egg discrimination ability and the evolutionary origin of this ability (Grim 2005). Statistical analyses We performed generalized linear models (GLM) with a binomial error distribution and a logit link function. The binomial response variables were rejection, egg burial, and desertion with latitude and latitude squared (to account for nonlinear efects) as covariates and country and presence or absence of each of the focal cuckoo species as ixed efects (see Table 1). Because many diferent variables may vary geographically and therefore correlate with host responses to cuckoo model eggs, we attempted to account for the potential efects of such confounding variables by inclusion of latitude, but also presence or absence of individual cuckoo species and the total number of local cuckoo species at our study sites (see Table 2). Finally, we included color of model eggs, breeding stage (egg laying or incubation periods), clutch size, and year, as additional predictors to test for potential confounding efects of these variables. Nonsigniicant potential confounding efects were sequentially removed from full models. The estimation method for the GLM was the Firth adjusted maximum likelihood method. All analyses were performed in JMP (SAS 2012). RESULTS We detected 27 literature-reported cases of parasitism by the common cuckoo of the great tit in Europe (Møller et al. 2011, including cases of cuckoo chicks reported in Grim et al. 2014), but only a single case of common cuckoo parasitism of great tits in China. We did not detect any cuckoo parasitism in any of the study nests in any study site during the present study. We found only 1 case of 1409 conspeciic brood parasitism in China (n = 294 nests, 2008–2013) and a single case (2 eggs appeared in the nest in a single day) in the Czech Republic (n = 174 nests, 2009–2010). Responses to conspeciic eggs were in all cases acceptance (n = 53, Supplementary Table 2). These replicates were subsequently included in the global analysis because a dummy variable with conspeciic eggs coded as 0 or 1 did not explain a signiicant amount of variation. In control nests, we did not detect any cases of ejection (Supplementary Table 2). These replicates were subsequently included in the global analysis because a dummy variable with control nest coded as 0 or 1 did not explain a signiicant amount of variation. The statistical model of ejection of eggs included 5 predictors (Table 1). Ejection rate decreased signiicantly in a nonlinear fashion with increasing latitude (Figure 3; Table 1). The rejection rate was signiicantly higher in study areas with Himalayan cuckoos (Table 1). In contrast, there were no signiicant additional efects of color of model eggs, breeding stage (egg laying or incubation periods), clutch size, and year (results not shown). A statistical model for egg burial as response variable including latitude, latitude squared, and presence or absence of the 3 cuckoo species as predictor vari2 ables was not signiicant ( χ5 = 5.04, P = 0.07), nor was a model for nest desertion as response variable including latitude, latitude squared, and presence or absence of the 3 cuckoo species as predic2 tor variables ( χ5 = 5.21, P = 0.39). Rejection rate increased with the number of cuckoo species (Figure 4). This efect of number of cuckoos species had a χ12 = 72.12 (Table 2), whereas the combined efect for the 3 abun2 dant cuckoo species only had a χ3 = 17.42 (Table 1). DISCUSSION Table 1 GLM model of ejection of model eggs in relation to linear and quadratic terms of latitude and presence or absence of the common cuckoo, Himalayan cuckoo, and Asian lesser cuckoo Term χ2 P Estimate SE Intercept Latitude Latitude squared Common cuckoo Himalayan cuckoo Asian lesser cuckoo 38.695 19.774 12.173 3.045 13.698 0.675 <0.0001 <0.0001 0.0005 0.081 0.0002 0.412 −13.103 0.162 0.006 0.660 7.375 −1.296 2.700 0.043 0.002 0.392 2.279 2.072 We used large continental-scale data to examine patterns of antiparasite responses in a potential host of avian brood parasites that have traditionally been claimed to be unsuitable. Despite the presumed unsuitability as host due to its hole-nesting habit, some populations of great tits in China, but not in Europe, showed very high egg rejection rates (up to 100%). Data from 395 great tit nests revealed large variation in the rate of ejection of model eggs among populations decreasing with latitude and increasing with the presence of Himalayan cuckoos. In fact, ejection rate increased strongly with the diversity of all parasitic cuckoo species. All degrees of freedom (df) = 1. The model had the likelihood ratio χ25 = 198.41, P < 0.0001, and the goodness of it was χ2389 = 411.77, P = 0.20. 1.0 Term χ2 P Estimate SE Intercept Number of cuckoo species Latitude Latitude squared 29.952 72.117 13.210 11.728 <0.0001 <0.0001 0.0003 0.0006 −11.208 1.534 0.122 0.005 2.317 0.226 0.035 0.002 Ejection rate 0.8 Table 2 GLM model of ejection of model eggs in relation to number of cuckoo species, latitude, and latitude squared 0.6 0.4 0.2 All degrees of freedom (df) = 1. The model had the likelihood ratio χ23 = 170.70, P < 0.0001, and the goodness of it was χ2391 = 453.86, P = 1.00. 0.0 20 30 40 50 Latitude 60 Figure 3 Predicted rejection rates (with 95% conidence intervals) of model cuckoo eggs in relation to latitude. Behavioral Ecology 1410 Ejection rate 1.0 0.8 0.6 0.4 0.2 0.0 0 1 2 3 4 5 6 7 No. cuckoo species Figure 4 Predicted rejection rates (with 95% conidence intervals) in diferent host species in relation to the number of cuckoo species. Although great tits are common breeding birds in natural and managed forests in China, there is only a single record of cuckoo parasitism of the great tit, although this is likely to be due to the scarcity of data on cuckoo parasitism rather than an absence of parasitism (Grim et al. 2014). Studies of avian brood parasitism in China are rare, although the incidence of brood parasitism in China seems to be just as high as in Europe (Erritzøe et al. 2012). Consistent with this claim, Soler (2014) reported a similar mean (SE) rejection rate of model eggs from Asian cuckoo hosts (47% (12), N = 14 species), as for European hosts (59% (5), N = 52). Great tits in Europe are some of the most commonly studied birds in the world and still cases of brood parasitism by cuckoos are extremely rare (Grim et al. 2014; this study). This may represent a methodological artifact. Speciically, in Europe, great tits are typically studied only when they breed in nest boxes with tiny holes that efectively prevent brood parasitism by the single large-body-sized interspeciic parasite, and this may also prevent successful ledging by the cuckoo chick (reviewed in Grim et al. 2014). The rejection rate in 1 European population of great tits was relatively high, and we hypothesize that this may be the result of interspeciic competition among diferent hole-nesters such as great tits and pied lycatchers over scare nest cavities. However, great tits in Europe and China commonly use natural cavities with large entrances (mean [SE] width of natural cavities according to van Balen et al. (1982) for Europe: 3.9 cm [0.68], range 2.0–6.5 cm, N = 33 occupied holes; Liang et al. (2013) for China: 4.0 cm [1.23], range 1.8–6.5 cm, N = 61 occupied holes]. Empirical data from redstarts conirm that cuckoo females are regularly able to enter boxes and cuckoo chicks easily ledge from boxes with entrances at the upper parts of this size range (Grim et al. 2009). Therefore, natural tit cavities should not prevent cuckoo parasitism in many cases and allow for coevolution between tits and cuckoos. We did not ind any cases of cuckoo parasitism in this study. This is most likely because we used standard tit nest boxes with small entrances (entrance diameter = 3.5 cm). Although cuckoo females are able to squirt the egg even into small entrance cavities (Davies 2000), such layings cannot establish a viable cuckoo gens (host race)—cuckoo chicks would not ledge from a small entrance cavity and the strain would go extinct in the very irst generation (see also Grim et al. 2011; Samas et al. 2014). Therefore, a cuckoo gens specialized on parasitizing tits (or any other bird species) breeding in small cavities (natural or artiicial) cannot exist in principle. However, natural nests of tits (which we did not study) are most likely open to cuckoo parasitism and such parasitism would select for anti-cuckoo adaptations in the populations that also sometimes make use of nest boxes that we provided (see also Grim et al. 2014). We hypothesized that ejection behavior in great tits would depend on the local diversity of brood parasites in our study sites. We used the number of cuckoo species, as a measure of diversity of brood parasites. In addition, we included latitude (and latitude squared to account for nonlinear efects) as an additional explanatory variable because it is well known that numerous factors other than the diversity of parasites show latitudinal variation (Rohde 1998). Hence, we controlled statistically for such a potentially confounding efect. There is a steep latitudinal gradient of increasing diversity of cuckoos from Northern to Southern China, with no similar cline in Europe, where the common cuckoo is the sole brood parasite on small-sized hosts. We found a similar pattern when relating ejection rate in response to the number of sympatric cuckoo species. Indeed, we found evidence of signiicant heterogeneity among study areas with a particularly steep latitudinal cline once the efects of latitude and parasite species had been considered (Table 1). The stronger efect of number of cuckoo species compared with the presence or the absence of the 3 most abundant cuckoo species suggests that interaction efects in addition to the main efects contribute to the evolution of egg ejection. These patterns were present for ejection rate, but not for egg burial or desertion, demonstrating that there was no concomitant selection for a diversity of resistance behavior against cuckoo parasitism, but for speciic resistance based on egg ejection. Finally, we documented no rejection of conspeciic eggs in the present study (note that desertion was not a speciic response to parasitism), which is in line with previous indings (Kempenaers et al. 1995). Likewise, there is an absence of genetic evidence for conspeciic parasitism in tits according to parentage analyses of more than 12 500 nestlings (review in Griith et al. 2002). In line with this, we found only 2 cases of conspeciic brood parasitism across all study sites. This means that conspeciic brood parasitism cannot explain our results and provides evidence for ejection of our experimental eggs being a speciic antiparasite response that coevolved with locally abundant cuckoos, especially in Southern China. Two mechanisms may account for spatial intraspeciic variation in ejection rate among sites: gene low and local adaptation. Soler, Martínez, et al. (1999, 2001) showed for magpie Pica pica hosts that allopatric populations retained signiicant levels of resistance to cuckoos with this level depending on distance from areas of sympatry. The data that we have analyzed here do not allow for discrimination between the 2 hypothetical mechanisms (see also Thorogood and Davies 2013). However, the gradual decay in ejection rate with distance that we have documented (Figure 3) is consistent with an efect of dispersal. In other words, gene low between populations may maintain the egg rejection capacity to some extent in northern Chinese populations (with lower local cuckoo diversity and, by implication, parasite pressure), while local adaptation may lead to the high rejection in southern Chinese populations (with high diversity of cuckoos). Liang et al. (2013) have previously reported a strikingly similar case of strong ejection behavior against model cuckoo eggs in barn swallows Hirundo rustica from China, but not in Europe. The replication of these research indings in great tits suggests that sympatry of several species of cuckoos and the occurrence of 11 species of cuckoos in China alone may have resulted in the evolution of strong egg ejection. With the large number of cuckoo species in tropical South America, Africa, Asia, and Australia (Erritzøe et al. 2012), it is likely that similar cases will be revealed by future research. Liang et al. • Geographic variation in egg ejection and diversity of brood parasites In conclusion, high rates of ejection of model cuckoo eggs in great tits have evolved in sympatry with species of parasitic cuckoos in China with a gradually decreasing rate of rejection with increasing latitude as the species diversity of cuckoos decreases. Importantly, tits accepted all conspeciic eggs excluding an alternative hypothesis that conspeciic brood parasitism was the selective pressure behind the evolution of high rates of egg rejection in Asian tits. The present study highlights the importance of large geographic scale in ecological research: without studying multiple populations across 2 continents, it would not be feasible to challenge a traditional view, based on European ield sites, that any hole-nesting birds are unsuitable cuckoo hosts and do not evolve any anti-cuckoo defenses. Great tit populations living in Europe either lost or even did not evolve speciic anticuckoo adaptations in the ecological context where only a single large-sized cuckoo species does not represent a threat to great tits usually breeding in small-sized holes. This pattern provides an example of how general host ecology (general life history traits, Grim et al. 2011) may contribute to host escape from the burden of brood parasitism. SUPPLEMENTARY MATERIAL Supplementary material can be found at http://www.beheco. oxfordjournals.org/ FUNDING This study was supported by the National Natural Science Foundation of China (Nos. 31260514 to C.Y., 31071945 to H.W., 31071938, 31272328, and 31472013 to W.L.), Program for New Century Excellent Talents in University (NCET-13-0761 to C.Y.), and T.G. acknowledges the support from Human Frontier Science Program (awards RGY69/2007 and RGY83/2012) and the Czech Science Foundation (grant no. P506/12/2404). B.G.S. was funded by the Research Council of Norway (218144). The idea for this project was conceived during our stay at the “Centre for Advanced Studies (CAS)” in Oslo during August 2009–June 2010. For their help with various aspects of this study, we would like to thank C. Cheng, X. Cheng, E. Flensted-Jensen, J. Li, L. Liu, S. Li, M. Krist, J. Matrková, P. Samaš, S. Skořepa, D. Wang, J. Wang, J. Wu, Y. Zhang, and Z. Zhang. P. Samaš and N. Langmore kindly provided constructive comments. We declare that all the authors have no conlict of interest. Ethical Standards: experiments in this study comply with the current laws of local countries in which they were performed. Handling editor: Naomi Langmore REFERENCES Avilés JM, Vikan JR, Fossøy F, Antonov A, Moksnes A, Røskaft E, Shykof JA, Møller AP, Jensen H, Procházka P, et al. 2011. The common cuckoo Cuculus canorus is not locally adapted to its reed warbler Acrocephalus scirpaceus host. J Evol Biol. 24:314–325. Briskie JV, Sealy SG, Hobson KA. 1992. Behavioral defenses against avian brood parasitism in sympatric and allopatric host populations. Evolution. 46:334–340. Carlson A, Sandström U, Olsson K. 1998. Availability and use of natural tree holes by cavity nesting birds in a Swedish deciduous forest. Ardea. 86:109–119. Cheng T, Xian Y, Guan G. 1991. Fauna Sinica. Aves, Vol. 6. Columbiformes, Psittaciformes, Cuculiformes and Strigiformes. Beijing (China): Science Press. Combes C. 2001. Parasitism: the ecology and evolution of intimate interactions. Chicago (IL): University of Chicago Press. 1411 Cramp S, Perrins CM, editors. 1993. Handbook of the birds of Europe, the Middle East and North Africa. Vol. 7. Oxford: Oxford University Press. Cruz A, Wiley JW. 1989. The decline of an adaptation in the absence of a presumed selection pressure. Evolution. 43:55–62. Davies NB. 2000. Cuckoos, cowbirds and other cheats. London: T. & A. D. Poyser. Davies NB, Brooke MDL. 1989. An experimental study of co-evolution between the cuckoo, Cuculus canorus, and its hosts. I. Host egg discrimination. J Anim Ecol. 58:225–236. Erritzøe J, Mann CF, Brammer F, Fuller RA. 2012. Cuckoos of the world. London: Christopher Helm. Griith SC, Owens IPF, Thuman KA. 2002. Extra pair paternity in birds: a review of interspeciic variation and adaptive function. Mol Ecol. 11:2195–2212. Grim T. 2005. Mimicry vs. similarity: which resemblances between brood parasites and their hosts are mimetic and which are not? Biol J Linn Soc. 84:69–78. Grim T. 2006. Cuckoo growth performance in parasitized and unused hosts: not only host size matters. Behav Ecol Sociobiol. 60:716–723. Grim T, Rutila J, Cassey P, Hauber ME. 2009. The cost of virulence: an experimental study of egg eviction by brood parasitic chicks. Behav Ecol. 20:1138–1146. Grim T, Samaš P, Moskát C, Kleven O, Honza M, Moksnes A, Røskaft E, Stokke BG. 2011. Constraints on host choice: why do parasitic birds rarely exploit some common potential hosts? J Anim Ecol. 80:508–518. Grim T, Samaš P, Procházka P, Rutila J. 2014. Are tits really unsuitable hosts for the common cuckoo? Ornis Fennica. 91:166–177. Hauber ME, Samaš P, Anderson MG, Rutila J, Low J, Cassey T, Grim T. 2014. Life-history theory predicts host behavioural responses to experimental brood parasitism. Ethol Ecol Evol. 26:349–364. Hauber ME, Tong L, Bán M, Croston R, Grim T, Waterhouse GIN, Shawkey MD, Barron AB, Moskát C. 2015. The value of artiicial stimuli in behavioral research: making the case for egg rejection studies in avian brood parasitism. Ethology. 121:521–528. Igic B, Cassey P, Grim T, Greenwood DR, Moskát C, Rutila J, Hauber ME. 2012. A shared chemical basis of avian host-parasite egg colour mimicry. Proc Biol Sci. 279:1068–1076. Johansson US, Ekman J, Bowie RC, Halvarsson P, Ohlson JI, Price TD, Ericson PG. 2013. A complete multilocus species phylogeny of the tits and chickadees (Aves: Paridae). Mol Phylogenet Evol. 69:852–860. Johnson DH. 2002. The importance of replication in wildlife research. J Wildl Manage. 66:919–932. Kempenaers B, Pinxten R, Eens M. 1995. Intraspeciic brood parasitism in two tit Parus species: occurrence and responses to experimental parasitism. J Avian Biol. 26:114–120. Kleven O, Oigarden T, Foyn BE, Moksnes A, Røskaft E, Rudolfsen G, Stokke BG, Lifjeld JT. 2007. Low frequency of extrapair paternity in the common redstart (Phoenicurus phoenicurus). J Ornithol. 148:373–378. Krist M. 2009. Short- and long-term efects of egg size and feeding frequency on ofspring quality in the collared lycatcher (Ficedula albicollis). J Anim Ecol. 78:907–918. Kvist L, Martens J, Higuchi H, Nazarenko AA, Valchuk OP, Orell M. 2003. Evolution and genetic structure of the great tit (Parus major) complex. Proc Biol Sci. 270:1447–1454. Langmore NE, Feeney WE, Crowe-Riddell J, Luan H, Louwrens KM, Cockburn A. 2012. Learned recognition of brood parasitic cuckoos in the superb fairy-wren Malurus cyaneus. Behav Ecol. 23:798–805. Li G, Zheng B, Liu G. 1982. Fauna Sinica. Aves, Vol. 13: Passeriformes (Paridae – Zosteropidae). Beijing (China): Science Press. Liang W, Yang C, Wang L, Møller AP. 2013. Avoiding parasitism by breeding indoors: cuckoo parasitism of hirundines and rejection of eggs. Behav Ecol Sociobiol. 67:913–918. Lindholm AK. 1999. Brood parasitism by the cuckoo on patchy reed warbler populations in Britain. J Anim Ecol. 68:293–309. López-de-Hierro MDG, Moreno-Rueda G. 2010. Egg-spot pattern rather than egg colour afects conspeciic egg rejection in the house sparrow (Passer domesticus). Behav Ecol Sociobiol. 64:317–324. Martín-Gálvez D, Soler JJ, Martínez JG, Krupa AP, Richard M, Soler M, Møller AP, Burke T. 2006. A quantitative trait locus for recognition of foreign eggs in the host of a brood parasite. J Evol Biol. 19:543–550. Martín-Gálvez D, Soler JJ, Martínez JG, Krupa AP, Soler M, Burke T. 2007. Cuckoo parasitism and productivity in diferent magpie subpopulations predict frequencies of the 457bp allele: a mosaic of coevolution at a small geographic scale. Evolution. 61:2340–2348. 1412 Martín-Vivaldi M, Soler M, Møller AP. 2002. Unrealistically high costs of rejecting artiicial model eggs in cuckoo Cuculus canorus hosts. J Avian Biol. 33:295–301. Matrková J, Remeš V. 2012. Environmental and genetic efects on pigmentbased vs. structural component of yellow feather colouration. PLoS One. 7:e36640. Medina I, Langmore NE. 2015. The costs of avian brood parasitism explain variation in egg rejection behavior in hosts. Biol Lett. 11:20150296. Moksnes A, Røskaft E. 1995. Egg-morphs and host preference in the common cuckoo Cuculus canorus: an analysis of cuckoo and host eggs from European museum collections. J Zool. 236:625–648. Moksnes A, Røskaft E, Braa AT, Korsnes L, Lampe HM, Pedersen HC. 1991. Behavioural responses of potential hosts towards artiicial cuckoo eggs and dummies. Behaviour. 116:64–89. Møller AP, Díaz M, Liang W. 2016. Brood parasitism and proximity to human habitation. Behav Ecol. 27:1306–1311. Møller AP, Saino N, Adamík P, Ambrosini R, Antonov A, Campobello D, Stokke BG, Fossøy F, Lehikoinen E, Martin-Vivaldi M, et al. 2011. Rapid change in host use of the common cuckoo Cuculus canorus linked to climate change. Proc Biol Sci. 278:733–738. Møller AP, Soler JJ. 2012. A coevolutionary framework based on temporal and spatial ecology of host-parasite interactions: a missing link in studies of brood parasitism. Chin Birds. 3:259–273. Monrsós JS, Gómez J, Encabo SI, Bradt S, Barba E, Gil-Delgado JA. 1999. Open nesting in great tits Parus major. Ardeola. 46:89–91. Moskát C, Hansson B, Barabás L, Bártol I, Karcza Z. 2008. Common cuckoo Cuculus canorus parasitism, antiparasite defence and gene low in closely located populations of great reed warblers Acrocephalus arundinaceus. J Avian Biol. 39:663–671. Moskát C, Szentpeteri J, Barta Z. 2002. Adaptations by great reed warblers to brood parasitism: a comparison of populations in sympatry and allopatry with the common cuckoo. Behaviour. 139:1313–1329. Nakamura H. 1990. Brood parasitism by the cuckoo Cuculus canorus in Japan and the start of new parasitism on the azure-winged magpie Cyanopica cyana. Jap J Ornithol. 39:1–18. Päckert M, Martens J, Eck S, Nazarenko AA, Valchuk OP, Petri B, Veith M. 2005. The great tit (Parus major) – a misclassiied ring species. Biol J Linn Soc. 86:153–174. Pavlova A, Rohwer S, Drovetski SV, Zink RM. 2006. Diferent post-Pleistocene histories of Eurasian parids. J Hered. 97:389–402. Behavioral Ecology Rohde K. 1998. Latitudinal gradients in species diversity. Area matters, but how much? Oikos. 82:184–190. Samas P, Hauber ME, Cassey P, Grim T. 2014. Host responses to interspeciic brood parasitism: a by-product of adaptations to conspeciic parasitism? Front Zool. 11:34. SAS. 2012. JMP version 10.0. Cary (NC): SAS Inc. Soler JJ, Martínez JG, Soler M, Møller AP. 1999. Genetic and geographic variation in rejection behavior of cuckoo eggs by European magpie populations: an experimental test of rejecter-gene low. Evolution. 53:947–956. Soler JJ, Martínez JG, Soler M, Møller AP. 2001. Coevolutionary interactions in a host-parasite meta-population. Ecol Lett. 4:470–476. Soler JJ, Martín-Gálvez D, Martínez JG, Soler M, Canestrari D, AbadGómez JM, Møller AP. 2011. Evolution of tolerance by magpies to brood parasitism by great spotted cuckoos. Proc Biol Sci. 278:2047–2052. Soler M. 2014. Long-term coevolution between avian brood parasites and their hosts. Biol Rev Camb Philos Soc. 89:688–704. Stokke BG, Hafstad I, Rudolfsen G, Moksnes A, Møller AP, Røskaft E, Soler M. 2008. Predictors of resistance to brood parasitism within and among reed warbler populations. Behav Ecol. 19:612–620. Thorogood R, Davies NB. 2013. Reed warbler hosts ine-tune their defenses to track three decades of cuckoo decline. Evolution. 67:3545–3555. van Balen JH, Booy CJH, van Franeker JA, Osieck ER. 1982. Studies on hole-nesting birds in natural nest sites. 1. Availability and occupation of natural nest sites. Ardea. 70:1–24. von Haartman L. 1976. The reaction of a regular cuckoo host to foreign eggs. Ornis Fennica. 53:96–98. von Haartman L. 1981. Co-evolution of the cuckoo Cuculus canorus and a regular cuckoo host. Ornis Fennica. 58:1–10. Yang C, Liang W, Antonov A, Cai Y, Fossøy F, Stokke BG, Moksnes A, Røskaft E. 2012. Diversity of parasitic cuckoos and their hosts in China. Chin Birds. 3:9–32. Yang C, Stokke BG, Antonov A, Cai Y, Shi S, Moksnes A, Røskaft E, Møller AP, Liang W, Grim T. 2013. Host selection in parasitic birds: are open-cup nesting insectivorous passerines always suitable cuckoo hosts? J Avian Biol. 44:216–220. Zhao N, Dai C, Wang W, Zhang R, Qu Y, Song G, Chen K, Yang X, Zou F, Lei F. 2012. Pleistocene climate changes shaped the divergence and demography of Asian populations of the great tit Parus major: evidence from phylogeographic analysis and ecological niche models. J Avian Biol. 43:297–310.