Great spotted cuckoo fledglings are disadvantaged by magpie host parents when reared together with magpie nestlings
- First Online:
- Cite this article as:
- Soler, M., de Neve, L., Roncalli, G. et al. Behav Ecol Sociobiol (2014) 68: 333. doi:10.1007/s00265-013-1648-9
- 200 Views
The post-fledging period is a critical phase for juvenile survival, and parental care provided during this period is a key component of avian reproductive performance. Very little is known about the relationships between foster parents and fledglings of brood parasites. Here, we present the results of a 5-year study about the relationships between fledglings of the non-evictor brood parasitic great spotted cuckoo (Clamator glandarius) and its magpie (Pica pica) foster parents. Sometimes, great spotted cuckoo and magpie nestlings from the same nest can fledge successfully, but most often parasitic nestlings outcompete host nestlings and only cuckoos leave the nest. We have studied several aspects of cuckoo post-fledging performance (i.e. feeding behaviour, parental defence and fledgling survival) in experimental nests in which only cuckoos or both magpie and cuckoo nestlings survived until leaving the nest. The results indicate that great spotted cuckoo fledglings reared in mixed broods together with magpie nestlings were disadvantaged by magpie adults with respect to feeding patterns. Fledgling cuckoos reared in mixed broods were fed less frequently than those reared in only cuckoo broods, and magpie adults approached less frequently to feed cuckoos from mixed broods than cuckoos from only cuckoo broods. These results imply that the presence of host's own nestlings for comparison may be a crucial clue favouring the evolution of fledgling discrimination; and furthermore, that the risk of discrimination at the fledgling stage probably is an important selection pressure driving the evolution of the arms race between brood parasites and their hosts.
KeywordsBrood parasitismClamator glandariusFledgling discriminationFledgling periodNon-evictor brood parasitic chicksPica pica
In altricial birds, three main phases of parental care can be distinguished: the incubation period (from laying to hatching), the nestling period (from hatching until fledging) and the post-fledging period (from fledging until independence). Much is known about the incubation period and about the interactions between parents and nestlings during the nestling period in many species. In fact, theoretical background of bird ecology and evolution is based almost exclusively on the enormous amount of information published on the incubation and the nestling phase. However, in most species, very little is known about avian reproductive strategies and life history evolution during the post-fledging phase. Probably due to the difficulty of monitoring mobile young after they have left the nest (Kershner et al. 2004; Thompson and Ridley 2013), this is one of the least studied subjects of avian reproductive biology (Baker 1993; Martin 1996; Field and Brace 2004; Russell et al. 2004; Grüebler and Naef-Daenzer 2010; Matthysen et al. 2010). Nevertheless, parental care during the post-fledging period is considered a key component of avian reproductive performance because this phase is critical for juvenile survival (Royama 1966; Sullivan 1989; Grüebler and Naef-Daenzer 2010; Tarwater and Brawn 2010) as many fledglings with limited mobility (even flightless during the first days in many species) suffer heavily from predation (Sullivan 1989; Zann and Runciman 1994; Naef-Daenzer et al. 2001; Kershner et al. 2004; Yackel Adams et al. 2006; Grüebler and Naef-Daenzer 2010). Actually, in many species, care for fledglings tends to last at least as long as that for nestlings and in some cases much longer (Skutch 1976; McGowan and Woolfenden 1990; Russell et al. 2004).
Interspecific brood parasites, whose offspring are incubated and raised by members of other species (Rothstein 1990; Davies 2000; Roldán and Soler 2011), are not an exception, and very little is known about the relationships between foster parents and fledgling parasites (Davies 2000; Grim 2006a; Soler 2009). For example, there is only some anecdotal information for the best studied brood parasite species, the common cuckoo (Cuculus canorus; Wyllie 1981; Davies 2000), while we know something more about the post-fledging period of parasitic cowbirds (Molothrus spp.). For instance, it is known that parasitic fledglings of the brown-headed cowbirds (Molothrus ater) are fed more frequently than host fledglings (Woodward 1983) resulting in reduced survival of the latter (Payne and Payne 1998; Hauber 2003; Hoover 2003; Rasmussen and Sealy 2006). This is probably because brood parasitic fledglings display more intense and persistent begging than do host fledglings (Eastzer et al. 1980; Woodward 1983; Payne and Payne 1998; Hauber and Ramsey 2003), as is also the case during the nestling stage (Soler et al. 1999b; Kilner et al. 1999; Tanaka and Ueda 2005). Furthermore, there is also clear evidence that some host defences can appear at this stage. The baywing (Agelaioides badius), a host of the specialist screaming cowbird (Molothrus rufoaxillaris) and of the generalist shiny cowbird (Molothrus bonariensis), is willing to feed fledglings of the specialist parasitic cowbird, which visually and vocally mimic host fledglings, but is reluctant to feed non-mimetic fledglings from the generalist brood parasite (Fraga 1998; De Mársico et al. 2012).
Relationships between parasitic fledglings and their hosts have also been studied in the great spotted cuckoo (Clamator glandarius), a specialist non-evictor brood parasitic cuckoo that uses the magpie (Pica pica) as its main host. Brood parasitism by the great spotted cuckoo strongly affects the breeding success of its magpie host by outcompeting magpie nestlings (Soler 1990). However, sometimes great spotted cuckoo and magpie nestlings share the nest and then nestlings of both species can survive and fledge successfully. On average, 0.6 magpie chicks fledge per parasitized nest, whereas 3.5 magpie chicks fledge from unparasitized nests (Soler et al. 1996). In contrast to what happens in cowbirds, and in all known species in which parents gradually increase their reluctance to feed fledglings forcing them to feed themselves and hence promote their independence (Davies 1976, 1978; Woodward 1983; Moreno 1984; Buitron 1988; Husby and Slagsvold 1992), great spotted cuckoo fledglings reach independence without a decrease in feeding rate by their foster parents and were never seen feeding by themselves (Soler et al. 1994, 1995a). In addition, great spotted cuckoo fledglings reared in different nests usually join with conspecific fledglings and behave as if they were nest-mates (Soler et al. 1995a). These groups of fledglings are attended by more magpies than those involved in nestling care and the feeding rate per cuckoo increases with the number of cuckoos per group (Soler et al. 1995a). This grouping behaviour seems to be triggered by social learning of the recognition of conspecifics during development (i.e. recognition is not innate), given that experimentally cross-fostered great spotted cuckoo nestlings outside the parasite breeding range did not learn to recognize their own species and they did not form the typical groups of fledglings (Soler and Soler 1999).
Great spotted cuckoo parasitism can create two types of nest situations, those with cuckoo nestlings alone (a situation provoked mainly in multiparasitized nests by the early hatching of cuckoo eggs, which provokes the starvation of all magpie nestlings), and those in which cuckoo nestlings share the nest with magpie nestlings until fledging. The latter situation mainly occurs in nests parasitized with only one cuckoo egg (1.1 magpie nestling fledge per nest, N = 81, Soler et al. 1998). However, the probability that magpie nestlings survive decreases as the number of cuckoo eggs per nest increases (only 0.026 magpie fledglings per nest when four or more cuckoo eggs are laid per nest, N = 38; Soler et al. 1998). Cuckoo nestlings may suffer high mortality rate in mixed broods when they hatch later than magpie nestlings (Soler et al. 1998), but nothing is known about the post-fledgling period regarding mixed broods. There could be different costs for cuckoo fledglings raised together with or without magpie nestlings depending on the ability of adult magpies to discern between their own young. Thus, in order to understand the co-evolutionary relationships between magpies and great spotted cuckoos, it is critical to know what happens with fledgling cuckoos reared in mixed broods compared to fledgling cuckoos reared with other conspecifics.
The evolution of chick discrimination is mainly related to the absence of other efficient host defences at earlier stages, mainly absence of egg rejection behaviour (Grim 2006a; Britton et al. 2007). Several studies already suggested that the presence of host's own nestlings for comparison may be a crucial clue favouring the evolution of nestling discrimination (Davies and Brooke 1988; Lotem 1993). Previous results on the great spotted cuckoo–magpie system also provided support to this idea. Soler et al. (1995b) showed that parasitized magpies accepted and fed cuckoo nestlings experimentally introduced into their nests during the last part of the nestling period, while magpies from non-parasitized nests frequently were reluctant to feed the experimental cuckoo nestling, and even, sometimes, rejected it. This magpie response to experimentally introduced great spotted cuckoo nestlings into their nests should not be considered a host response which specifically evolved due to interaction with parasites, as it is well known that bird parents learn the appearance and begging calls of their chicks shortly before the end of the nestling period (Beecher et al. 1981; Lessells et al. 1991; Medvin et al. 1993; Soler et al. 1995b; Levréro et al. 2009).
It could be suggested that comparison between own and brood parasitic fledglings could be even a more important cue during the post-fledging stage than during the nestling stage because the appearance of fledglings could be compared better and more efficiently out of the nest without constrains imposed by the nest characteristics and when plumage development is much more advanced. Indeed, fledgling discrimination has until now only been reported for hosts of non-evicting parasites in which they have the opportunity of comparison (Fraga 1998; De Mársico et al. 2012). However, all currently available data come from the same host–brood parasite system, the one formed by the baywing host and its two cowbird brood parasite species (see above). This, probably, does not imply that cases of fledgling discrimination by hosts are very scarce; likely, it is the consequence of a lack of knowledge about the relationships between foster parents and fledgling parasites.
In addition, previous studies on the post-fledging period in brood parasites exclusively focused on feeding strategies (e.g. Woodward 1983; Soler et al. 1994; De Mársico et al. 2012). To our knowledge, there is no information about predator defence during the fledgling period in brood parasites, which is another crucial aspect of parental care (Royle et al. 2012). In fact, as far as we know, information on cuckoo nestlings is also extremely scarce and only two studies have explored this subject (Soler et al. 1999a; Honza et al. 2010). In general, predation is very important at the fledgling stage (see above), but the defence against predators could be even more critical for brood parasites because their more intensified begging behaviour is related to the attraction of nest predators (e.g. Hannon et al. 2009; Ibáñez-Álamo et al. 2012), which could be extended to the fledgling period given that begging differences between hosts and parasites are maintained at this stage (see references above). Thus, a differential defensive behaviour by adult magpies for cuckoo fledglings raised together with or without magpies could be an important, but so far unconsidered, cost for this parasite, even if they received a similar feeding rate as host fledglings.
The aim of the present study is to determine whether the nest situation of great spotted cuckoos (i.e. share the nest with host nestlings or only with conspecifics) could affect several aspects of great spotted cuckoo performance during the post-fledgling period. To investigate this, we experimentally created mixed nests with magpie and cuckoo nestlings and nests with only cuckoo nestlings, and repeated the experiment over five breeding seasons. If the presence of host's own fledglings for comparison is crucial for the evolution of parasite fledgling discrimination (see above), adult magpies should be more capable of discriminating great spotted cuckoo fledglings in mixed broods, and we predict differences between the two types of nests in feeding behaviour (i.e. feeding frequency and propensity/reluctance to feed the cuckoo fledglings) and/or defensive behaviour. Cuckoo fledglings from mixed broods should receive less feedings than those from only cuckoo broods and magpie adults should show a higher propensity to feed cuckoos in only cuckoo broods than in mixed broods. Furthermore, adult magpies would defend less the cuckoo fledglings from mixed broods than those from only cuckoo broods. As a result, we predict that cuckoos from mixed broods will survive less well than cuckoos raised together with other conspecifics.
Material and methods
Study area and general field methods
Field work was carried out during five consecutive breeding seasons (2008–2012) in a magpie population breeding in the Hoya de Guadix (37°10′ N, 3°11′ W). This area is a high-altitude plateau (approx. 1,000 m a.s.l.) with cereal crops (especially barley), groves of almond trees (Prunus dulcis) and some areas with dispersed holm-oak trees (Quercus rotundifolia).
All nests were located during nest-building or egg-laying, so clutch size and parasitism by great spotted cuckoos could be recorded and hatching date could be calculated accurately. Occurrence of brood parasitism was frequent in the study area. Between 2008 and 2012, a total of 339 out of 596 magpie nests were parasitized (parasitism rate = 56.8 %). At the predicted hatching date, nests were visited daily in order to create experimental broods by means of cross-fostering equally aged nestlings (see below). Once the experimental brood was created, nestlings were individually marked with a non-toxic marker on their tarsus and experimental nests were regularly visited along the nestling period.
Experimental set-up: cross-fostering of cuckoo and magpie nestlings
In naturally parasitized nests, great spotted cuckoo nestlings usually outcompete magpie nestlings (Soler et al. 1995c) mainly because they hatch several days earlier than magpie nestlings (Soler 1990; Soler et al. 1996). Thus, in order to ensure survival of both parasite and host nestlings until fledging, it was necessary to manipulate the nest content obtaining cuckoo and magpie nestlings of similar age and size within each nest. Cross-fostering of 1- or 2-day-old nestlings was done by carefully transporting them in an artificial cotton nest lined with tissues, maintaining the temperature in the car between 25 and 30 °C. Cross-fostered magpies and cuckoos did not differ in age and size (see Soler and De Neve 2012 for test statistics). Cross-fostering per se does not affect nestlings or host parents' behaviour (Soler and Soler 1999; Soler and de Neve 2012, 2013). In some cases in which we took one great spotted cuckoo chick that was alone in the nest, we left another cuckoo chick of similar age from a multiparasitized nest to avoid nest abandonment.
The cross-fostering manipulation used in this study was not only carried out to study the relationships between fledgling great spotted cuckoos and magpie foster parents but also for the purpose of testing other hypotheses related to food delivery by parents and begging behaviour by nestlings according to brood composition (see Soler and de Neve 2012, 2013 for further information on the performed cross-fostering experiments). For the purpose of this study, we considered two types of experimental broods: only great spotted cuckoos (one or three nestlings) and mixed broods with one cuckoo and one or two magpies. Given that transmitters are expensive, they constrained the number of fledglings we could provide with a tag. In the first years of the study, we mainly focused on providing cuckoos from mixed broods with a transmitter, given that their situation is rarer in natural conditions. Only few nests with only cuckoo chicks were provided with a transmitter in those years. However, in 2011, we expanded the study and provided a similar proportion of mixed broods and only cuckoo broods with a transmitter. For this reason, we only used nests from 2011 to test if cuckoos from mixed and those from only cuckoo broods received different feeding frequencies (see “Statistical analyses”).
The brood sizes used in the experimental nests (between one and three nestlings) are smaller than the typical brood size in magpie non-parasitized nests, but they are frequently found in natural conditions in naturally parasitized nests, where brood size is often only one or two nestlings (usually cuckoos).
Radio-tracking and observations of fledglings
Between 2008 and 2012, a total of 94 great spotted cuckoo nestlings from 77 different nests were equipped with radio transmitters. Thirty-five of them were from mixed broods and 42 from “only cuckoo” broods. Transmitters were provided to cuckoo nestlings a few days before fledging (when 16 ± 1 day old). In multiply parasitized nests (i.e. three nestling cuckoos), only the largest cuckoo nestling(s) per brood (which have more probabilities of survival) were fitted with a transmitter. Transmitters (Holohil PD-2) weighed approximately 4.5 g each, back-pack harness included (<3 % of fledgling body weight), and had a range of 1,000 m and a battery life of 14–26 weeks. They were attached using the back-pack method (2008 and 2010) (adapted from Hill et al. 1999) or the leg-loop harness method (2009, 2011 and 2012) (Rapole and Tipton 1991) with the antenna always oriented down the tail. Refined versions of both methods have been effective in small and medium size birds, without causing skin or plumage damage, or interfering with behaviour (Hill et al. 1999; Naef-Daenzer et al. 2001). Attaching the transmitters several days before the fledglings leave the nest also allowed them to become accustomed to it.
Observations on feeding behavior were performed during the seasons 2008–2011. At the expected fledging date of the cuckoo, nests were visited every 2–3 days in order to check if the cuckoo had fledged and to locate the chick. Observations started when the cuckoo was outside the nest. We made as many observations of each fledgling as possible. The age of the chick was included in all analyses (see “Statistical analyses”). Observations of interactions between fledglings and adult magpies were performed from the end of May until the middle of July during each breeding season. All the observations were made during the most active periods of both adult magpies and cuckoo fledglings, that is, from sunrise to 11 a.m. and from 6 p.m. until sunset. A total of 174 h of effective observations of radiotracked fledglings were carried out, in which 542 feedings to fledglings were observed.
Upon arrival at the territory, given that fledglings often remained well hidden in the tree canopy, first, we located the nestling/s by using the radio-tracking method (three element hand-held antennas O-5/8, receptor RX-98H (Televilt), which receive frequencies between 138 and 175 MHz until a flight distance of 1 km), and then we retreated and started observing on each observation day. Observations were made using binoculars (×10) and telescopes (×20–60) from the car or from a hide (depending on the characteristics of the territories). Each fledgling was located at least once every 3 days until it could not be tracked anymore because of mortality or abandonment of the study area. When a fledgling was detected, we noted the identity of the individual fledgling, the exact hour, whether the fledgling was alone or in a group with other fledglings, the behaviour of the fledgling and of the adult magpies (i.e. whether the magpie went towards the cuckoo to feed it, or whether the fledgling went towards the adult magpie to get the feeding). Once the situation of the fledgling cuckoo (i.e. alone or together with nest mates cuckoo fledglings) changed or when we lost sight of the fledgling, we considered it the end of the observation by noting the exact hour. The next observation started immediately if we still got sight of the fledgling in another situation or from the moment when we again got a visual of the same or another fledgling of the nest. A fledgling was observed for a mean of 85.1 ± 4.4 min on each observation day (95 % CL, 76.2–93.8 min, N = 123). The feeding rate to each fledgling was calculated as the number of feedings per hour. The number of nests available for analysing feeding rates (see “Statistical analyses”) was reduced due to predation events soon after fledging or because some territories were difficult to observe and did not provide reliable information on feeding rates.
When we were not able to locate a radio-tracked fledgling, the following day we tried again and, if the signal came from the same location as the previous day, we followed the signal with the hand-held antenna until finding the transmitter. A fledgling was considered depredated when it was found dead or when the transmitter was found with some indication of predation (like feathers or blood).
Fledgling defence trials
During the fledging period of 2011 and 2012, we performed 34 fledgling defence trials (11 only cuckoo broods, 23 mixed broods) in which one of us approached a recently fledged chick that was hidden in the tree canopy. During the first days after leaving the nest, great spotted cuckoo fledglings remained silent and immobile in the tree canopy close to the nest (Soler et al. 1994), which allowed us to test the defence behaviour of adult magpies when their cuckoo fledglings were threatened by one of us. Basically, the experimental procedure consisted in locating a recent fledgling by observing the tree of the nest and the near trees from a considerable distance (between 100 and 400 m) or by using a hand-held antenna, and, once detected in the tree canopy, one of us approached by walking slowly. We stopped at a distance of 1–5 m from the trunk of the tree in which the fledgling was hidden and waited for 10 min or until magpies approached. If no magpie was detected after the 10 min (N = 6), we considered the trial unsuccessful and it was not used in the analyses because we could not be sure that adult magpies were aware of our presence. When at least one adult magpie approached within a time frame of 10 min, we observed the defence behaviour for another 10 min. So, all observations of defence behaviour had duration of 10 min. This method has proven to be efficient in testing magpie nest defence (Redondo and Carranza 1989). During that 10 min interval, we measured (1) whether one or two magpies approached (i.e. if they came into our area of vision, usually less than 200 m), (2) the closest distance to the human [very close (0–10 m), close (11–20 m), intermediate (21–50 m), far (51–100 m) and very far (>100 m)], and (3) scolding intensity (1 = no scolding calls to the human, 2 = scolding for less than half the duration of the experiment and 3 = scolding for more than half the duration of the experiment). In order to have a complete perception of adults' defensive behaviour, we have calculated a defence index. Scolding intensity was first weighed (multiplied) by the distance index (ranging from 1 (very close) to 0.2 (very far)). The weighed scolding intensity of each adult was then summed to obtain the defence index towards each fledgling. This means that if two adults approached the intruder, the defence index (scolding intensity weighed for distance) of each adult was calculated and then summed to obtain the total defence intensity at that nest. If only one adult approached, then the defence index of that individual was used as total defence intensity of that nest. We are confident that adult magpies showed their natural anti-predator behaviour given that it has been shown that corvids defend their nests against humans and natural predators with the same frequency (Röell and Bossema 1982).
Feeding rate (feedings/hour) fitted a normal distribution after log transformation (Kolmogorov–Smirnov P > 0.2). To explore potential differences in feeding rate (dependent variable) between fledgling cuckoos and fledgling magpies (species fixed effect), we used the subset of mixed broods in which magpie fledglings were located while observing the cuckoo fledgling, and in which, at least one successful observation was made of feeding rates to both magpies and cuckoos from the same nest (N = 13 nests, 92 observations). A General Linear Mixed Model (GLMM) was fitted with year and fledgling identity nested within nest identity included as random factors to account for the non-independence of individuals from the same year and of different observations of the same individual and of individuals from the same nest.
Focusing only on fledgling cuckoos, we explored if cuckoos received different feeding rates depending on their nest situation (fixed effect: mixed broods versus only cuckoo broods). For this analysis, we only used data from 2011 because for this year most observations of only cuckoo broods were made (N = 10 nests, 24 observations). A GLMM was used in which nest was included as a random factor to account for the non-independence of different observations from the same nest. The brood size and age of fledglings, as well as their interactions with the fixed factor, were included in all models in order to account for a possible correlation with feeding rate. We provided only cuckoos proceeding from nests fledged in the middle of the breeding season (during a 3-week period) with tags, avoiding very early and very late nests.
To test the question if cuckoos have to approach more often to adult magpies to get food when they are raised in mixed broods compared to only cuckoo broods, we used again a GLMM with a repeated measures (RM) design. For each observation session of a cuckoo fledgling, we calculated the number of times per hour that the cuckoo approached the adult magpie to get fed and the number of times per hour that the adult magpie approached the cuckoo to feed it (hereafter referred to as “approach strategy”, repeated measure, within subject effect; i.e. testing for each observation the difference in frequency of magpies approaching cuckoos and cuckoos approaching magpies). The “nest situation” (mixed versus only cuckoo broods) was included as a fixed factor (between subjects) in the model. To test if the approach strategy was different for cuckoos raised in mixed broods or in only cuckoo broods, the interaction between “approach strategy” and “nest situation” was included in the model. Nest identity, nested within year, was included as a random factor to account for non-independence of observations made in the same nest and year.
Values of log-transformed feeding rates are back-transformed in the results to allow interpretation of real values. Due to this back-transformation, SE flags are slightly asymmetrical in the figures.
The defence index differed significantly between the three observers (F2,54 = 4.68, P = 0.013) and was therefore standardized for each observer (i.e. the mean value for each observer was extracted from the observed value). The index adjusted well to a normal distribution after standardization (Kolmogorov–Smirnov P > 0.2). To explore if there were differences in adults' defensive behaviour depending on the nest situation (only cuckoos vs. mixed broods), we used a GLMM with defence index as the dependent variable and nest situation as a fixed factor. Brood size and the fledgling's age were included as covariates, and year as a random factor.
We used a Generalized Mixed Model (GLZM, logit link function, binomial error) to analyse if fledgling cuckoos were differentially depredated depending on their nest situation (raised in mixed or only cuckoo broods). Brood size was included as a covariate in the model to control for a possible correlation with the probability of predation. Nest identity, nested in year, was included as a random factor to account for the non-independence of individuals from the same nest and year. Non-significant interactions and main effects were dropped from final models.
All analyses were performed with SAS 9.3 (SAS Institute Inc., Cary, NC, USA) and figures were made with STATISTICA 7.0 (StatSoft Inc. 2001–2004).
In mixed broods, feeding rates significantly differed between nests (GLMM Z = 2.24, P < 0.05), but did not differ between magpie and cuckoo fledglings (GLMM F1,18.5 = 0.11, P = 0.74). No other variables or interactions, included in the initial model, were retained (all P > 0.18).
From the 76 cuckoo fledglings (from 67 different nests) for which we had certainty about predation events during the post-fledging period, exactly 50 % of them were depredated. The probability of being predated did however not differ between cuckoo fledglings raised in mixed or only cuckoo broods (GLZM F1,9 = 0.11, P = 0.74, N = 76, before removing the factor from the model) and also brood size did not influence predation probability (GLZM F1,9 = 0.41, P = 0.54, N = 76). Only nest identity explained significant variation in predation events (GLZM Z = 3.49, P = 0.002).
Here, as far as we know, we present the first experimental results showing the importance of the nest situation for the performance of brood parasite fledglings. Furthermore, this is also the first time that, in addition to feeding behaviour, adult defence behaviour towards parasite fledglings is studied.
Feeding rate did not differ between magpie and great spotted cuckoo fledglings in mixed broods. This result is not in agreement with the one reported for cowbird fledglings, which usually elicit more parental care than host fledglings (Woodward 1983; Skutch 1996 (cited in Rasmussen and Sealy 2006); Payne and Payne 1998). Strikingly, adult hosts that were feeding cowbird fledglings almost never fed their own host fledglings (Woodward 1983; Rasmussen and Sealy 2006), which was interpreted as the cause of death of most hosts young soon after fledging (Woodward 1983; Rasmussen and Sealy 2006).
Furthermore, fledgling cuckoos from mixed broods tended to receive less feedings than those from only cuckoo broods and a clear significant result showed that during the first 3 weeks after leaving the nest, fledgling cuckoos reared in mixed broods were fed less frequently than those reared in only cuckoo broods (Fig. 1). These results might support the idea that magpies are capable of discriminating great spotted cuckoo fledglings when they are allowed to compare with their own fledglings. The fact that the difference disappeared 3 weeks after leaving the nest is probably due to the fact that, as fledgling cuckoos become older, they join in cuckoo groups that are communally fed by more magpies than those involved in rearing the cuckoos in the nest (Soler et al. 1995a). The most frequent situation is that great spotted cuckoos from mixed broods that become undernourished abandon their natal territories to join a group of fledglings in nearby areas (MS et al. unpublished), in which they get fed at a higher rate (Soler et al. 1995a). This ability to abandon less-efficient caregivers to look for more efficient ones has, as far as we know, never been found in fledglings of any brood parasite species; however, it has frequently been reported in some cooperative species in which juveniles switch between caregivers in order to maximize their provisioning rate (McGowan and Woolfenden 1990; Hodge et al. 2007; Thompson and Ridley 2013).
Another important result showing that magpies have some capacity to discriminate cuckoos from their own fledglings when these are present for comparison is that magpie adults tended to approach less frequently to feed cuckoos from mixed broods than cuckoos from only cuckoo broods (Fig. 2).
The fledgling defence experiment has shown that magpies defended cuckoo fledglings from mixed broods significantly more than cuckoo fledglings from only cuckoo broods (Fig. 3). This result is most intriguing because it is not in agreement with our expectations (supported by our results on feeding behaviour) that magpies have some ability of fledgling discrimination just after leaving the nest. This implies that cuckoo fledglings from mixed broods, in terms of defence by magpies, do not suffer higher costs than cuckoo fledglings from only cuckoo broods; at least during the first days after leaving the nest, when they are more vulnerable to predation given that they have limited mobility like fledglings of most altricial species (see references above). Perhaps, it is due to the possibility that when testing defence behaviour of adult magpies towards cuckoo fledglings from mixed broods, one unmarked magpie fledgling might have been in the proximity of the fledgling cuckoo and defence behaviour by the adults was then also directed towards the magpie fledgling.
Survival rate of great spotted cuckoo fledglings was 50 %. This percentage is not significantly lower than the one reported previously for this species (63.2 %; Chi2 (1) = 1.77, P = 0.18, N = 38; Soler et al. 1994), but very similar to the one reported for brown-headed cowbirds (47.6 %; Chi2 (1) = 0.04, P = 0.84, N = 21; Woodward and Woodward 1979) and lower than that reported for the common cuckoo (84.2 %; Chi2 (1) = 12.51, P = 0.004, N = 38; Wyllie 1981). Likely, the difference in survival of great spotted cuckoo fledglings between this and the previous study could be due to the fact that the present study has been made in a rural, mostly non-cultivated area, in which dispersed holm-oak trees predominate, while in the previous study cuckoo fledglings were mainly observed in irrigated agricultural areas near the villages, in which almond tree groves predominate. We do not have exact information about the predator community in each habitat, but we think that such important differences between study areas likely should affect the composition and density of the predator community.
However, contrary to our predictions, survival rate of fledgling great spotted cuckoos did not differ between mixed and only cuckoo broods. This finding is surprising given that our results on feeding behaviour suggest important differences in parental care depending on the presence of magpie young on the nest, and previous assumptions that suggest that parental care provided during this post-fledgling phase is very important for the survival of young (Royama 1966; Sullivan 1989; Grüebler and Naef-Daenzer 2010; Tarwater and Brawn 2010). There are two possible, not exclusive, explanations for this result. The first one is related to the previously mentioned fact that cuckoo fledglings raised in mixed broods soon (usually between 7 and 20 days after leaving the nest) abandon their foster parents to join a group of cuckoo fledglings, which are fed communally by a group of magpies (Soler et al. 1995a). This cuckoo behaviour of abandoning their foster parents to join a group if they are undernourished would allow cuckoo fledglings to avoid starvation. In fact, we have not found any cuckoo fledgling that died due to starvation. In addition, a translocation experiment has demonstrated that fledgling cuckoos are able to obtain food from other magpies different from their foster parents (MS et al. unpublished data). However, it is also possible that undernourished cuckoo fledglings could be more vulnerable to predation.
The second one could be a consequence of the fact that cuckoo fledglings from mixed broods are defended by their magpie foster parents at a higher intensity than cuckoo fledglings from only cuckoo broods, which would decrease the predation rate of cuckoo fledglings from mixed broods during the first days of the fledgling period, the most critical period for juvenile survival (see above).
The only factor that explained significant variation in predation events was nest identity, which is logical given that if one nest is included in a territory prospected by several predators their fledglings will be more prone to be predated than if it is in an area without predators. Furthermore, it is known that fledglings that received more parental care during the nestling period have a greater probability of post-fledgling survival (Grüebler and Naef-Daenzer 2010).
In conclusion, we have found several pieces of evidence showing that magpies might be able to discriminate fledgling great spotted cuckoos soon after leaving the nest when some of their own nestlings survive to fledge, and adjust part of their parental care behaviours (feeding strategies) correspondingly, but they do not discriminate parasitic fledglings when they have been reared alone in the magpie nest. This last result probably is the consequence of the fact that host foster parents (as it occurs with their own young) learn the begging calls of a parasitic chick at the end of the nestling period (see references above), and later, continue feeding it because they have learnt the vocal signature of that fledgling as one of their own nestlings.
There is another plausible alternative explanation that multiple cuckoo fledglings from only cuckoo broods collectively elicit higher levels of parental care than cuckoo fledglings from mixed broods. This possibility would be supported by the fact that cuckoo fledglings joined in cuckoo groups receive a higher per capita feeding rate than solitary cuckoo fledglings (Soler et al. 1995a). However, our results do not show that cuckoo fledglings from only one cuckoo broods were disadvantaged by their foster parents, which supports the idea that magpies might be able to discriminate great spotted cuckoo fledglings when some of their own fledglings survive.
From the brood parasite's point of view, elimination versus acceptance of host nestlings has always been discussed in relation to costs and benefits of sharing the nest during the nestling period (Soler 2002; Kilner et al. 2004; Kilner 2005; Martín-Gálvez et al. 2005; Grim 2006b; Hauber and Moskát 2008; Grim et al. 2009; Soler and de Neve 2013). The result here showing that great spotted cuckoo fledglings are more frequently fed by magpie hosts when reared alone in the nest compared to those reared sharing the nest with host nestlings implies that the risk of discrimination at the fledgling stage probably is another important selection pressure driving the evolution of strategies involved in elimination of all host progeny in brood parasites.
We thank Francisco Espinosa, Francisco Ferri and Juan S. Sanchez for their help with field work. This work was supported by the Spanish Ministerio de Economía y Competitividad/FEDER (research project CGL2011-25634/BOS). We also thank Tomas Grim, the Associate Editor and two anonymous reviewers for constructive comments on a previous version of the manuscript, which considerably improved and clarified the text.
Research has been conducted according to relevant Spanish national (Real Decreto 1201/2005, de 10 de Octubre) and regional (permissions provided yearly by la Consejería de Medio Ambiente de la Junta de Andalucía) guidelines.