Nest desertion as an anti-parasitism strategy in hosts selects for late egg-laying behavior in cuckoos

Summary Studies have shown that brood parasites lay their eggs early in the egg-laying sequence of their hosts, providing them with the advantage of earlier hatching. However, common cuckoos (Cuculus canorus) appear to parasitize the nests of gray bushchat (Saxicola ferreus) during the late egg-laying stage. The bushchat often abandons parasitized nests in the early stages, but not in the late egg-laying stages, thus favoring late egg-laying by cuckoos. In this study, four experiments were conducted to determine whether gray bushchats employ a nest desertion strategy targeted at cuckoo parasitism. The results showed that nest desertion was significantly correlated with parasitism cues and occurred mainly during the hosts’ early egg-laying stage. Our study provides the first experimental evidence that nest desertion is an anti-parasitic strategy used by hosts in response to cuckoos. Additionally, our experiments demonstrated that the nest desertion is influenced by the trade-offs of investments in different egg-laying stages.


INTRODUCTION
2][3] The antagonistic interaction between avian brood parasites and their hosts has emerged as a model system for studying the ''arms race'' theory of coevolution. 4,5According to the natural selection theory, hosts have evolved various anti-parasitic adaptations in response to the high costs of parasitism. 1,6Host responses to parasitized nests are generally categorized into two types: acceptance and rejection (ejection of the parasite egg and nestling, burying of the parasite egg, and nest desertion). 7,80][11][12] However, the mechanisms underlying nest desertion remain poorly understood.
The fitness costs associated with host reproduction due to brood parasitism facilitate the evolution of anti-parasitism adaptations among hosts. 6For example, most hosts exploited by cuckoos (Cuculus spp.) are small passerine birds that have evolved the ability to recognize and reject parasite eggs, owing to the high cost of parasitism. 1 However, ejecting highly mimetic and large parasite eggs from some cuckoo hosts is difficult. 1,6Therefore, in such instances, nest desertion strategies are vital for cuckoo hosts, as nest desertion can lead to favorable reproductive adaptations. 7,13,149][20][21] Observations of naturally parasitized nests have indicated that most hosts exhibit various degrees of desertion of parasitized nests. 14A study on the common redstart (Phoenicurus phoenicurus) demonstrated that the desertion rate of parasitized nests was significantly higher than that of non-parasitized nests. 11Therefore, comprehensive studies on anti-parasitism strategies are needed to elucidate the reasons and adaptive mechanisms underlying nest desertion in cuckoo hosts following parasitic encounters.
Consequently, several studies have explored the causes of nest desertion adaptation in hosts.First, nest desertion by parents is associated with a decrease in the number of eggs in the nest.For example, studies on cowbird hosts have found that the nest desertion rate of parents is positively correlated with a reduction in the number of eggs in the nests. 22This could be related to the reproductive trade-offs associated with parasitized nests. 23,249][30] Studies on the reed warbler (Acrocephalus scirpaceus) have shown that when hosts spot a cuckoo dummy, they abandon their nests. 9Similar observations have been reported in cowbird hosts. 30,31However, other studies have found that encounters with parasites do not increase nest abandonment. 32,33Third, host nest desertion is triggered by non-mimetic parasite eggs.5][36] For example, Western Bonelli warblers (Phylloscopus bonelli) and marsh warblers (Acrocephalus palustris) ejected smaller model eggs but abandoned nests with large model eggs. 37,38However, despite the ejection of parasite eggs, some hosts still abandon their nests at times. 39,40In some cases, the presence of parasite eggs does not lead to nest desertion by the host. 302][43][44] For example, studies on gray fantails (Rhipidura albiscapa) found that stimulation by predators significantly increased nest desertion in hosts compared to controls without predators. 450][51][52] Field observations suggest that the nest desertion behavior of hosts is primarily triggered by cues of parasitism, suggesting that nest desertion is an important adaptation mechanism of anti-parasitism behavior. 11,14,53,54While previous studies have investigated host adaptation to nest desertion from different angles, they have rarely examined the impact of parental investment trade-offs on nest desertion at different stages of egg-laying.
Gray bushchat (Saxicola ferreus) is the dominant host of the common cuckoo (Cuculus canorus) (Table 1, Figure 1). 55Our observations over five breeding seasons revealed that parasitism of the gray bushchat by the common cuckoo tended to occur during the late egg-laying or incubation period, whereas nests parasitized during the early egg-laying stages of the host were more likely to be abandoned (see the Results section).7][58] To date, there have been few studies on the causes and patterns of host nest desertion in parasitic brood systems. 30,31,38,59his study aimed to experimentally demonstrate whether nest desertion adaptation of the gray bushchat host is related to cues of parasitism and to explore nest desertion patterns.We used natural cuckoo parasitism and host nest desertion data from five breeding seasons to calculate the distribution of egg-laying timing and host nest desertion patterns.Second, we determined whether nest desertion patterns were correlated with parasitic events by designing experiments on parasitism with exposure to different cues (see the field experiments section).Based on this, our first conclusion was that if the nest desertion adaptation of the experimental groups was similar to that of natural parasitism (nest desertion behavior occurs mainly in the early laying stages rather than in the later stages), it would suggest that the nest desertion pattern of gray bushchat is related to an investment trade-off by host parents.The second conclusion was that the different rates of nest desertion between parasitized and non-parasitized nests suggest that nest desertion is an anti-parasitic adaptation that is triggered by cues of parasitism.Furthermore, if the ejection rates of Group 1 (white model eggs) and Group 4 (white budgerigar eggs) differed, but the nest desertion rate remained the same, we inferred that the nest desertion adaptation of the gray bushchat may not be related to the cost of ejection.

RESULTS
During the five breeding seasons (March-August) of 2018-2022, 1,116 gray bushchat nests were found, 8.2% of which were parasitized by the common cuckoo.In cases where the number of eggs in the parasitized clutch was determined, we found a significantly higher proportion of parasitism in the late egg-laying (4-5 eggs) than in the early egg-laying stage (1-2 eggs) (16 vs. 50; chi-squared test for given probabilities, c 2 = 17.515, degree of freedom [df] = 1, p = 0.001).Nests parasitized in the early egg-laying stages were more likely to be abandoned by the host than nests in the late egg-laying stages (Bonferroni confidence intervals: 0.099 % P 1 % 0.451 [early stage] vs. 0.014 % P 2 % 0.3 [late stage]; n = 66; Table 2).
We compared the four groups of the parasitic manipulation experiments and found that host adaptation of nest desertion was selected in the early rather than late egg-laying stages (Figure 2).A comparison of the four manipulation experiments in the early egg-laying stage revealed no significant differences among the four groups (c 2 = 0.778, df = 3, p = 0.378).Furthermore, there was a significant difference in nest desertion rates between the three groups of non-parasitized (control) and parasitized nests (natural and experimental parasites) during the early egg-laying stage (Fisher's exact test, p < 0.001).We conducted post-hoc comparisons between the treatments and found that all p values were <0.001 (Table 3).
In all-natural and experimental parasitic experiments, egg ejection occurred mainly in group 4 (Table 2).The comparative study found that even though there was a significant difference in ejection rates between Group 1 and Group 4 (early stage: c 2 = 19.829,df = 1, p < 0.001; late stage: c 2 = 19.49,df = 1, p < 0.001), there was no difference in nest desertion patterns (c 2 = 0.033, df = 1, p = 0.857), and nest desertion only occurred in the early egg-laying stages (Table 2).In Group 5, only one nest was abandoned during the early egg-laying stage, and there was no significant difference in the rate of nest desertion in Group 3 (Fisher's exact test, p = 0.228) (Table 3).

DISCUSSION
Our study demonstrated that the desertion of naturally parasitized nests by gray bushchat hosts was triggered by cues of parasitism from the common cuckoo rather than by disturbances associated with nest visits during the experiment.Additionally, the pattern of parasitized nest desertion by gray bushchats depends on the trade-off of investment at different egg-laying stages, independent of the host's ejection costs of anti-parasitism.Our study demonstrated that nest desertion is an important anti-parasitism strategy for gray bushchats in response to common cuckoo parasitism, and experimental verification showed that this anti-parasitism adaptation exists mainly in the early egg-laying stage.
In the late egg-laying or incubation stage, the gray bushchat rarely abandons its nests, despite the presence of cuckoo parasitism.
In this study, a similar pattern of nest desertion by gray bushchats was elicited in response to both experimentally and naturally parasitized nests, occurring during the early rather than late egg-laying stages.Furthermore, no nest desertion due to nest visits occurred in the nonparasitized nest control group, suggesting that the desertion of parasitized nests by the gray bushchat host is indeed an anti-parasitism strategy, as these responses were elicited by parasitism cues.Similar studies based on dummy stimuli demonstrated that encounters with In the early stage of Group 4, there was one sample of the host rejecting budgerigar eggs before abandoning the nest.parasites cause nest desertion. 9,30Some studies have suggested that host nest desertion behavior is related to the ejection costs of parasite eggs. 34,35,60However, our results do not support the ejection cost hypothesis, as we found that in the late egg-laying period, gray bushchats rarely abandoned larger simulated or non-mimetic cuckoo eggs, whereas budgerigar eggs similar in size to host eggs were abandoned in the early egg-laying stage.Studies have shown that some hosts choose to abandon their nests instead of rejecting the parasite eggs. 39,40In addition, we found that only 16% (n = 50) of the cuckoo eggs were abandoned during the late egg-laying or incubation period, which could be mainly due to excessive parasitic disturbances caused by cuckoos.In four cases, we observed at least six attempts at parasitism by cuckoos before success, suggesting that nest desertion was not due to host ejection. 61otably, the nest desertion behavior of gray bushchats exhibited adaptations for the egg-laying stage.We propose that parental investment at different egg-laying stages influences the host's nest desertion decisions.26][27]62 Similar studies have found  that the desertion rate of nests parasitized by chipping sparrows gradually decreases in the late egg-laying stage, 30 and a significantly higher desertion rate was found before egg laying than after egg laying. 63However, studies on natural parasitism in yellow warblers, the hosts of cowbirds, have revealed their predisposition to bury individual eggs in the nest and abandon their nests after three eggs have been laid. 64Yellow warbler nest desertion behavior might be attributed to factors beyond the parasitic response. 14,64Previous studies have speculated that the cost of egg ejection affects host desertion behavior, suggesting that hosts of highly mimetic parasite eggs are more inclined toward nest desertion; however, this was not supported by comparative studies, 61 possibly because these studies did not consider the influence of reproductive cost trade-offs on host nest desertion behavior.In contrast, our experimental data showed that among the large nonmimetic model eggs, those in the early laying stage were abandoned at a comparable or slightly higher rate than simulated cuckoo eggs.However, in the late egg-laying stage, nest desertion behavior was rarely observed, regardless of the degree of parasite egg mimicry or the possibility of egg ejection.Additionally, our results showed that the degree of simulation and size of parasite eggs influenced host ejection and not nest desertion behavior, suggesting that the nest desertion strategy of gray bushchats is influenced by the degree of parasite egg mimicry in the pre-egg-laying stage, as it determines whether the host will eventually abandon the nest.However, in the late egg-laying stage, the effects of the degree of mimicry became less important as the gray bushchat does not abandon the nest owing to ejection costs, which explains why cuckoo hosts often accept parasite eggs despite recognizing them. 1,6he desertion rate of all experimentally parasitized nests was significantly lower than that of naturally parasitized nests during the early egg-laying stage, particularly in Group 2, which could be attributed to the different parasite encounters during the two processes, 7,11,28,65 as natural parasitism entails diverse cues, and the levels of exposure to these cues are different from those experienced during experimental conditions.For example, multiparasitised nests are more likely to be abandoned, 66 which could be related to multiple nest visits by cuckoos.We also observed that nests with four to five eggs were parasitized only after frequent parasitism attempts and host attacks, eventually leading to nest desertion.However, the reason behind the significantly lower desertion rate of experimentally parasitized nests compared to that of naturally parasitized nests remains unknown.Although our data suggest that hosts use nest desertion as an anti-parasitism strategy stimulated by cues of parasitism and that this pattern of nest desertion correlates with the investment trade-off of parents at different egg-laying stages, caution should still be taken when concluding that nest desertion serves as a host response to parasitism.8][69] Future research should further clarify how cuckoo hosts, such as gray bushchats, identify parasitism cues in the context of the complex roles of both nest parasites and predators.
In conclusion, our study demonstrated that nest desertion is an important anti-parasitic strategy for this cuckoo host.The choice of this adaptation strategy depends on the parents' investment trade-offs at different egg-laying stages of the host, implying that the extent to which parasite eggs are stimulated in the early rather than the late egg-laying stage is important for nest desertion adaptation by the host.Furthermore, the synchronized egg-laying time exhibited by the common cuckoo in this parasitic system might also suggest coevolution with host nest desertion, favoring late egg-laying behavior in cuckoos; however, this requires further verification.To the best of our knowledge, this is the first study to demonstrate that the host's anti-parasitism strategy of nest desertion is related to the investment trade-off of parents at different egg-laying stages.However, the prevalence of this nest desertion pattern in gray bushchat hosts requires further validation by investigating more avian brood parasitism systems and host species.

Limitations of the study
Our study provides the first experimental evidence that nest desertion is an anti-parasitic strategy used by gray bushchats in response to cuckoo parasitism.However, whether the host's pattern of nest abandonment in this study in relation to investment trade-offs varies differently in other parasitic systems needs to be further verified across different geographic populations or species.In addition, future work should also investigate whether the synchronized egg-laying time exhibited by the common cuckoo in this host-parasite system may favor late egglaying behavior in cuckoos.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

SUPPLEMENTAL INFORMATION
Supplemental information can be found online at https://doi.org/10.1016/j.isci.2023.108156.desertion patterns of hosts in response to parasitized nests, we designed different simulated parasitism experiments by dividing the hosts into early and late egg-laying groups (1-2 eggs vs. 4-5 eggs) as follows: In Group 1, as no egg recognition ability has been reported for grey bushchat, the host was first stimulated with non-mimetic white model eggs (both cuckoo and host eggs were pure blue, see Figure 1F) to test the host's recognition ability and nest desertion behavior in response to non-mimetic eggs (i.e., one white model egg was placed directly into the host's nest and checked once every 2-3 d).The experiment concluded after 6 d. 59  (Soler et al. 2015).We recorded the host's behavioral response to the parasitized nest during the experimental cycle.These responses were: (1) Ejecting only the parasite egg, but not deserting the nest, (2) Ejecting both the parasite egg and deserting the ne st, (3) No-ejective but abandoning the nest; (4) No-ejective and not deserting nest.In Group 2, based on the color of the cuckoo eggs, we used mimetic model eggs (blue model eggs) to explore whether the host nest desertion response was related to the degree of egg mimicry, following the same procedure as that used for Group 1.In Group 3, the natural parasite egg-laying process was simulated by placing a cuckoo specimen at a distance of 0.5-1 m from the nest while introducing a model egg into the nest.This approach was inspired by studies suggesting that host nest desertion is related to encounters with parasites. 8,14The experiment was conducted from 6:30 AM to 6:30 PM.Host behaviors such as attacking the specimen (see ESM Videos S1 and S2), ignoring the specimen, or watching the nest for over 1 min were videotaped(Uniscom-T71, 70 3 26 3 12 mm; Mymahdi Technology Co. Ltd., China) and observed using binoculars.The nests were periodically checked.The host nest desertion response in the first three groups of experiments could be related to the cost of egg ejection, as the experimental eggs were model eggs that were similar in size and quality to real cuckoo eggs, but difficult to peck.Thus, an additional experimental group (Group 4) was included in which real white budgerigar (Melopsittacus undulatus) eggs (trade-permitted non-fertilized eggs), which are similar in size and quality to the host eggs, were placed directly into the host nests, and their status was periodically observed.In Group 5, Oriental Turtle-dove specimen was used as an innocuous sympatric control species to verify that nest desertion was triggered in response to the sight of the parasite near the nest.This control involved achieving the same frequency of nest-visiting activity for some early egg-laying (1-2 eggs) non-parasitized nests.This was done to determine whether host nest desertion was associated with a cue of parasitism or an artificial stimulus for nest visitation.

QUANTIFICATION AND STATISTICAL ANALYSIS
Pearson's chi-square test was used to compare the distribution of naturally parasitized nests in the egg-laying stages (early and late egglaying stages) of the host, where samples with indeterminate clutch size and those with three eggs (late egg-laying stage for individuals with a full clutch size of four eggs, and middle egg laying stage for those with a full clutch size of five eggs) were excluded from the analysis.Thereafter, Pearson's chi-square test or Fisher's exact test were used to test for differences in host desertion rates of different parasitic manipulation experiments at the same stage (early or late egg laying), excluding nests with uncertain clutch size or desertion status from the analysis.In addition, we calculated the Bonferroni confidence intervals of nest desertion for each stage(P1: Bonferroni confidence intervals for early egg-laying stage; P2: Bonferroni confidence intervals for late egg-laying stage) in natural and manipulative parasitism per the method described by Byers et al. 71 All statistical analyses were performed using R (R 3 .5.2 ) 72 and 'RStudio', 73 and all tests were two-tailed at a 0.05 significance level, with egg parameters expressed as mean = x (SD = y).

Figure 1 .
Figure 1.The host's nesting habitat and the parasitic process of the cuckoo Landscape of the study site (A), various breeding habitats and nest-sites of the gray bushchat (nesting on cultivated land (B), barren slopes, and near the grassroots (C and D) or rock cover (E) in scrub forests), the egg pecking behavior of the common cuckoo (F), and the eggs of the cuckoo and gray bushchat host (G).The arrow is pointing at the location of the nest and cuckoo egg.

Figure 2 .
Figure 2. Comparison of host nest desertion rates of natural cuckoo-parasitized nests and nests subjected to experimental parasitism at different egglaying stages (A) Schematic diagram of the experiment.(B) Distribution of natural parasitic nests and nest desertion rate.(C) Desertion rate of experimental parasitic nests at different egg-laying stages.In the second chart, Groups 1-4 represent manipulation parasitic experiments.Model eggs are similar in size to true cuckoo eggs, and Budgerigar eggs are similar in size to eggs.In the charts, ''Early stage'' replaces the early egg-laying stage, and ''Late stage'' replaces the late egg-laying stage.P1 represents the Bonferroni confidence interval for the early egg-laying stage; P2 represents the Bonferroni confidence interval for the late egg-laying stage.

Table 1 .
Parameters for parasite eggs (natural and experimental parasitism) and host eggs (mean = x (SD = y))

Table 2 .
Hosts behavioral responses (acceptance, ejection, and nest abandonment) to parasite eggs of manipulation and natural parasitism

Table 3 .
Comparison of host desertion rates of different parasitic manipulation experiments in the same stage (early or late egg-laying)

TABLE d
B Ethical standards d METHOD DETAILS B Field experiments d QUANTIFICATION AND STATISTICAL ANALYSIS