
Citation: | Jörg HOFFMANN, Udo WITTCHEN, Ulrich STACHOW, Gert BERGER. 2013: Identification of habitat requirements of farmland birds based on a hierarchical structured monitoring scheme. Avian Research, 4(4): 265-280. DOI: 10.5122/cbirds.2013.0026 |
Agricultural landscapes are essential for the conservation of biodiversity. Nevertheless, a negative trend continues to be observed in many rural areas for the most prominent indicator species group, the farmland birds. However, clear cause-effect relationships are rarely reported and sometimes difficult to deduce, especially from monitoring data which are based only on the detection of species and counts of the numbers of individuals. Because the identification of habitat preferences is a precondition for farmland bird biodiversity conservation efforts, a monitoring scheme for the simultaneous collection and analysis of bird and land use data was developed and tested. In order to assign the occurrence of bird species to land characteristics at various spatial scales and different land use and crop types, we applied a hierarchical structured sampling design. The spatial scales were 'agri-cultural landscape', 'agricultural landscape types', 'field crops and other habitats' and 'vegetation structures'. These scales were integrated with a novel concept, the 'habitat matrix' (HM). This method was applied to farmland breeding bird abundances on 29 plots, each 1 km2 in size, by the use of the territory mapping method. The same plots were enlarged by a 100 m buffer and the sizes and location of habitats documented. Vegetation height, coverage and density were also recorded for all crop fields in the study area. We propose that this monitoring method facilitates the identification of scale dependent relationships between farmland bird habitat characteristics and bird abundance. This is demonstrated by the farmland bird species Corn Bunting (Emberiza calandra), Skylark (Alauda arvensis), and Whinchat (Saxicola rubetra). The breeding territories of these species reveal large differences within the various spatial scales 'agricultural landscape', 'agricultural landscape types' and 'field crops'. Throughout the breeding season the abundances varied, dependent on the field crop and the development of vegetation structures (height, coverage, and density). HM-analysis led to the identification of specific habitat configurations preferred by individual bird species within the agricultural landscape. These findings indicate that the methodology has the potential to design monitoring schemes for the identification of cause-and-effects of landscape configuration, land use and land use changes on the habitat suitability and abundance of farmland birds.
The interaction between the parasitic adaptations of avian brood parasites and anti-parasitic adaptations of their hosts has long been regarded as a classical model system for studying coevolution (Davies, 2000; Soler, 2014). As a result of this “arms race,” egg recognition behaviors, which recognize and reject parasite eggs, have evolved as highly effective anti-parasitic strategies employed by hosts; thus, they have received much more attention and research than other subjects within avian brood parasitism (Davies, 2011). Accordingly, most studies have explored egg recognition behavior in hosts through experiments on artificial parasitism (e.g., Davies and Brooke, 1989; Moksnes et al., 1991), in which a crucial aspect was to investigate the hosts’ responses to foreign eggs to understand the evolutionary mechanisms of egg recognition (Avilés et al., 2010). Hosts typically reject foreign eggs by puncturing (Moksnes et al., 1991), grasping ejection (Begum et al., 2012), burying the eggs (Guigueno and Sealy, 2010), or deserting the nests (Hosoi and Rothstein, 2000). However, because of the varying ability of hosts to discern morphological differences between parasitic eggs and their own eggs (Kilner, 2006), there is a degree of variability in recognition and rejection capabilities, both within and among host species (Davies and Brooke, 1989; Davies, 2000). Numerous studies have investigated the egg characteristics utilized as cues by hosts to perceive the differences between parasitic eggs and their own eggs. As a result, egg cues such as color (Stoddard and Stevens, 2011; Hanley et al., 2017; Luro et al., 2018), spotting (Spottiswoode and Stevens, 2010; Dainson et al., 2017; Ye et al., 2022a), shape (Underwood and Sealy, 2006b; Guigueno and Sealy, 2012; Zölei et al., 2012), and ultraviolet reflectance (Honza and Polačiková, 2008; Yang et al., 2013; Abernathy and Peer, 2015; Šulc et al., 2015) have been the subjects of the majority of these studies (Samaš et al., 2021). In contrast, much less research has been carried out on the role of other egg characteristics, such as egg size and quantity, in the recognition processes.
Usually, recognition based on egg size can be a prevalent cue among hosts because the egg sizes of many parasites are larger than those of the hosts (Payne, 2005; Yang et al., 2012), making them easy cues for recognition. However, studies supporting this hypothesis are rare (Marchetti, 2000; Tosi-Germán et al., 2020). Only 21 previous studies (Sadam, 2022; Ye et al., 2022b; Liu et al., 2023) have explored host egg recognition based on egg size, most of which involved hosts of parasitic cowbirds, while only four of the remaining eight studies on parasitic cuckoo hosts supported identification cues based on the egg size (Roncalli et al., 2017; Meshcheryagina et al., 2020; Sadam, 2022; Ye et al., 2022b). Furthermore, these studies provide ambiguous evidence to support the hypothesis that egg size recognition evolves as a specific anti-parasitic adaptation against the parasites (Ye et al., 2022b). More studies are needed to reveal the role of egg size in host recognition during the coevolution of the brood parasite–host system, especially for cuckoo–host interactions.
In addition to egg size, the number of parasitic eggs in a nest would affect the host fitness based on the recognition mechanism adopted by the hosts. Hosts can employ two recognition mechanisms to identify parasitic eggs: template recognition (known as true recognition) and discordant recognition (known as false recognition) (Rothstein, 1975; Soler, 2014). The former refers to hosts discriminating against parasite eggs based on a memory template, which yields accurate rejection decisions regardless of the number of parasite eggs in the host nests (Rothstein, 1974, 1978; Moskát et al., 2010). The latter means that the hosts cannot recognize their own eggs but instead make rejection decisions by simply rejecting the egg types that form the minority in the nest (Rothstein, 1974; Moskát et al., 2010). The advantage of the template recognition mechanism lies in its independence from the number of parasite eggs during the recognition process, thus reducing recognition errors, especially in the face of multiple parasitism. Consequently, most host species use this mechanism for egg recognition (Rothstein, 1975; Lotem et al., 1995; Moskát and Hauber, 2007; Lang et al., 2014). The discordant recognition mechanism is relatively simple because learning to recognize their own egg phenotype is unnecessary; therefore, recognition is achieved through inconsistencies in egg quantity. However, a drawback appears when the number of parasitic eggs is larger than that of the host eggs in the same host nest because of multiple parasitism. In this situation, the hosts would mistakenly reject their own eggs because of their smaller quantity. Hence, this recognition mechanism is not observed to be used alone by hosts but generally coexists with the template recognition mechanism in a host population (Moskát et al., 2010; Stevens et al., 2013; Yang et al., 2014b; Wang et al., 2015). Researchers have investigated host responses to the number of parasitic eggs by manipulating the contrast of egg quantity between parasites and hosts in host nests (Lyon, 2007; Yang et al., 2014b).
In this study, we tested the egg recognition behavior of a potential cuckoo host, the Green-backed Tit (Parus monticolus), by considering both egg size and quantity. This species belongs to cavity-nesting birds that have rarely been observed to be parasitized because the previous studies were conducted using artificial nest boxes, which obstruct parasitism and finally lead to a high underestimation of parasitism on natural nests (Grim et al., 2014). Egg size recognition is hypothesized to serve as a recognition cue because this species is a secondary cavity-nesting bird wherein low nest luminance would make the recognition of light-dependent cues, such as egg colors and markings, difficult (Yang et al., 2019; Ye et al., 2022a). This study aimed to determine the roles of egg size and quantity cues in the egg recognition process in this host species. Although no parasitism was observed in the studied species, its natural parasitism status (including multi-parasitism) was underestimated highly, as previously mentioned. Moreover, previous studies have confirmed that egg recognition would be maintained in hosts for a long time (Lahti, 2006; Yang et al., 2014a). Therefore, the current artificial parasitism experiment was still capable of investigating the adaptation of the host toward parasitism.
To test the role of egg size in recognition, we created three egg size levels (smaller than, equal to, and larger than host eggs). This tested the egg recognition capacity of the hosts and revealed the rejection mode because larger eggs are supposed to trigger nest desertion rather than egg ejection due to the limitation of host bill sizes (Underwood and Sealy, 2006a). We hypothesized that (1) hosts would possess the capacity for egg size recognition because they build nests under dim light conditions. Specifically, we expected that parasite eggs that were smaller or larger than the host eggs would be rejected more often by the hosts than equal-sized eggs. We predicted that (2) egg size would affect the rejection mode in that larger parasite eggs would trigger a greater proportion of nest desertion in hosts.
To investigate the effect of egg quantity, we created a wide range of quantity contrasts. This allowed us to investigate the egg recognition mechanism of hosts and to test the potential effect of quantity contrast on egg rejection mode. If the hosts employed the mechanism of discordancy, we predicted that (3) they would reject the eggs of the minority, regardless of whether they were parasite eggs or their own eggs. Conversely, if the hosts used the template mechanism, we expected that (4) they would reject the parasite eggs, regardless of the quantity contrast between the parasite and host eggs.
This study was conducted during the breeding seasons (April to August) of 2020 to 2021 in the Kuankuoshui National Nature Reserve, located in southwestern Guizhou Province of China (28°06′–28°19′ N, 107°02′–107°14′ E), covering an area of 26,231 ha. The reserve spans an elevation range of 650–1762 m, with an average annual temperature of 11.7–15.2 ℃, annual precipitation between 1300 and 1350 mm, and an average annual relative humidity exceeding 82%, making it one of the high-humidity areas in the country (Yang et al., 2011). The Green-backed Tit is a small secondary cavity-nesting bird that breeds in cavities of rocks, walls, eaves, and artificial nest boxes. A typical clutch size is seven eggs with a white background and reddish-brown spots (Ye et al., 2019). Parasitic cuckoos in China are diverse in terms of their number of species and size and utilize many cavity-nesting hosts (Yang et al., 2012, 2024). However, the parasitism status of cavity-nesting species such as tits (Parus spp.) has been critically underestimated because these studies rarely came from natural nest sites but from artificial nest boxes instead (Grim et al., 2014). Therefore, although the studied population of Green-backed Tits was not parasitized, they exhibited a strong ability to recognize parasite eggs (rejection of 100% of non-mimetic foreign eggs and low to intermediate rates for mimetic eggs), implying historical or current interaction with cuckoos (Yang et al., 2019).
During the breeding season, 118 nest boxes (depth, 35 cm; entrance diameter, 4 cm) were regularly inspected to determine the occupancy of Green-backed Tits. Upon discovering the nesting materials, the nest boxes were checked every alternate day to assess the nesting progress of the host. Information such as laying date and clutch size was recorded during the investigation. An egg size recognition experiment was performed during the early incubation stage (the first four days after clutch completion), in which one mimetic model egg made of polymer clay was randomly used to replace one host egg in each observed nest. Three model egg sizes were designed: larger, equal, and smaller, while the color and markings of the model eggs mimicked those of Green-backed Tits, whose eggs are white with reddish spots (Fig. 1). Mimetic rather than non-mimetic eggs were used in the current study because the tits rejected 100% of non-mimetic eggs (Yang et al., 2019), and using non-mimetic eggs as experimental models would conceal the effect of egg size difference. The egg sizes of the Green-backed Tits were 1.44 ± 0.1 cm2 (n = 12); thus, the size of equal-size egg models was determined based on a standard according to their average egg size (i.e., 1.44 cm2). The larger-size models were 2.88 cm2, which was twice as large as that of the hosts and similar to the cuckoo egg sizes in sympatric areas (Yang et al., 2012). The smaller-sized models were 0.97 cm2 because this size is similar to the smallest size of natural bird (Phylloscopus castaniceps) eggs observed in sympatric areas (Yang et al., 2012). Nests were randomly assigned to one of the treatments and artificially parasitized by one model egg (larger, equal, or smaller). The sample sizes for larger, equal, and smaller model eggs were 14, 14, and 17, respectively. A control group that followed the procedure of artificial parasitism, except for the replacement of host eggs with model eggs, was used to test the manipulation disturbance during the experiment. No nest desertion or abnormal phenomena were observed in the host nests in the control group (n = 15). The experimental nests were then monitored for six days (observed 1st, 3rd, and 6th day after artificial parasitism) to assess the hosts’ reactions. Model eggs were considered rejected if they were removed, had peck marks, were buried, or if the nests were deserted. Conversely, model eggs were considered accepted if they remained in the nests, were still being incubated, and had no peck marks.
In the egg quantity recognition experiment, the model eggs were consistent in size with those of the equal-size group in the egg size recognition experiment above but differed in color. The color of the model eggs was immaculate blue, which refers to the color of cuckoo eggs in the sympatric area (Yang et al., 2010), and thus, they were non-mimetic to the host eggs. The experimental procedure was consistent with that of the egg size recognition experiment, except for the number of model eggs. The model eggs replaced a few host eggs in nests, forming a broad range of quantity contrast (i.e., −5 to 5) between parasite and host eggs. For instance, −5 indicated five more host eggs than parasitic eggs, while 5 indicated the opposite. Thus, there were 11 contrasts of egg quantity (−5, −4, −3, −2, −1, 0, 1, 2, 3, 4, and 5; n = 22 for total with n = 2 for each contrast) between parasite and host eggs, with negative numbers indicating more host eggs than parasite eggs, and positive numbers indicating more parasite eggs than host eggs. Furthermore, the clutch sizes in the nests remained constant because the quantity of host eggs replaced by parasitic eggs was consistent. For example, the quantity contrast of −3 for the clutch size of 5 or 7 indicates 4 host eggs vs. 1 parasite egg or 5 host eggs vs. 2 parasite eggs, respectively. We did not change the clutch size in this experiment to avoid the potential side-effect of clutch size manipulation. Each nest received only one quantity contrast manipulation and followed the same monitoring procedure as in the egg size recognition experiment described above. To minimize interference, host eggs that were replaced by parasitic eggs during the experiments were kept in an artificial incubator (Brinsea Mini EX, Brinsea Products Ltd., Sandford, UK) and returned to the host nests after the experiments.
We used generalized linear models (GLMs) to analyze egg size and quantity recognition data. The response variable for the egg size recognition experiment was either the host recognition reaction (reject or accept model eggs) or the rejection mode (ejection or desertion), thus forming two independent models. For both models, the fixed effects included the experimental treatment (three sizes of model eggs), clutch size (2–10), incubation days of the experiment (hereafter, incubation day), and laying date (laying date of the first egg). Post hoc test based on the Tukey method was performed for pairwise comparison. The response variables and fixed effects were the same as in the two models above for the egg quantity recognition experiment, except that the experimental treatment was the egg quantity contrast between the parasite and host eggs. All tests were two-tailed, with a significance level of P < 0.05. Data are expressed as mean ± standard deviation (SD). GLMs and pairwise comparison analyses were performed using SPSS 25.0 for Windows (IBM, Armonk, NY, USA) and the emmeans package in R 4.1.0 for Windows (https://www.r-project.org/), respectively.
The results of the egg size recognition experiment indicated that the rejection rates among larger (57.1%, n = 14), equal (35.7%, n = 14), and smaller (52.9%, n = 17) model eggs decreased from larger to equal model eggs and then increased from equal to smaller model eggs (Fig. 2). However, the GLMs analyses did not reveal significant differences (F = 1.432, P = 0.251; Table 1). Nevertheless, there was a significant difference in the rejection modes among the models of different sizes (GLMs: F = 4.567, P = 0.026; Table 1). All larger-sized model eggs were rejected by nest desertion, and the desertion rates decreased with decreasing egg size, accounting for 40% and 22.2% of the rejection rates in equal- and smaller-sized model eggs, respectively (Fig. 2). Correspondingly, 60% of the equal-sized model eggs and 77.8% of the smaller-sized model eggs were rejected by ejection (Fig. 2), with neither clutch size, incubation day, nor laying date affecting the host rejection rate (GLMs: P = 0.379, 0.298 and 0.654; Table 1) or rejection mode (GLMs: P = 0.366, 0.889 and 0.620; Table 1). Pairwise comparison results indicated that the rejection mode of larger-sized model eggs differed significantly from that of equal- and smaller-sized model eggs (P = 0.023 and P < 0.001, Post-hoc test), while the equal- and smaller-sized model eggs did not have a significant difference (P = 0.672, Post-hoc test).
Response variable | Source | Sum of squares | df | Mean square | F | P |
Recognition reaction | Intercept | 0.397 | 1 | 0.397 | 1.515 | 0.226 |
Treatment | 0.75 | 2 | 0.375 | 1.432 | 0.251 | |
Laying date | 0.053 | 1 | 0.053 | 0.204 | 0.654 | |
Incubate day | 0.291 | 1 | 0.291 | 1.11 | 0.298 | |
Clutch size | 0.207 | 1 | 0.207 | 0.791 | 0.379 | |
Rejection mode | Intercept | 1.931 | 1 | 1.931 | 14.021 | 0.002** |
Treatment | 1.257 | 2 | 0.629 | 4.567 | 0.026* | |
Laying date | 0.035 | 1 | 0.035 | 0.255 | 0.620 | |
Incubate day | 0.003 | 1 | 0.003 | 0.02 | 0.889 | |
Clutch size | 0.147 | 1 | 0.147 | 1.065 | 0.317 | |
Treatment: Egg size contrasts between parasites and hosts (larger, equal, or smaller model eggs); Laying date: laying date of the first egg; * P < 0.05, ** P < 0.01, *** P < 0.001. |
The egg quantity recognition experiment showed that manipulating egg quantity contrast between parasites and host eggs had no significant effect on the host recognition reaction (Table 2). The hosts showed 100% rejection of the non-mimetic model eggs, in that they either ejected the model eggs accurately or deserted the nests without mistakenly rejecting their own eggs (n = 22). Clutch size, incubation day, and laying date had no significant effect on host recognition reactions (Table 2). However, the egg quantity contrast significantly predicted the rejection modes (GLMs: P = 0.002; Table 2), and the probability of nest desertion increased as the quantity of parasitic eggs increased (Fig. 3).
Response variable | Source | Sum of squares | df | Mean square | F | P |
Recognition reaction | Intercept | 0.305 | 1 | 0.305 | – | – |
Laying date | 0 | 1 | 0 | – | – | |
Clutch size | 0 | 1 | 0 | – | – | |
Treatment | 0 | 1 | 0 | – | – | |
Incubate day | 0 | 1 | 0 | – | – | |
Rejection mode | Intercept | 0.074 | 1 | 0.074 | 0.517 | 0.482 |
Laying date | 0.085 | 1 | 0.085 | 0.593 | 0.452 | |
Clutch size | 0.413 | 1 | 0.413 | 2.89 | 0.107 | |
Treatment | 2.023 | 1 | 2.023 | 14.165 | 0.002** | |
Incubate day | 0.163 | 1 | 0.163 | 1.143 | 0.300 | |
The first model does not produce any statistical result because the response variable was a single value (i.e., only rejection was found in the recognition reaction); Treatment: egg quantity contrast between parasites and hosts (–5, –4, –3, –2, –1, 0, 1, 2, 3, 4, 5); Laying date: laying date of the first egg; * P < 0.05, ** P < 0.01, *** P < 0.001. |
The results indicated that the rejection rate of equal-sized model eggs (35.7%) was lower than that of larger (57.1%) or smaller model eggs (53%). Although no statistical significance was observed in the rejection rates among the different sizes of model eggs, certain biologically meaningful patterns may exist. The rejection rate of equal-sized model eggs was 17% lower than that of larger- or smaller-sized model eggs. Therefore, although statistical significance was not observed, our results imply that the hosts might be able to detect an egg size that differs from their own eggs to a certain extent, but egg size may not be the primary recognition cue. Furthermore, the rejection rates of these model eggs might be overestimated compared to natural eggs because of their essence of artificial materials. Nevertheless, this would not affect the outcomes of the comparison above. The rejection modes changed with the model egg size, and the nest desertion rates increased as the model egg size increased. Pairwise comparison results indicated that equal- and smaller-sized model eggs did not differ significantly. The potential reason is that rejection of the larger-sized model eggs by ejection was much more difficult than rejection of the equal- and smaller-sized model eggs by ejection, while rejection of the equal-sized model eggs by ejection was slightly more difficult than rejection of the smaller-sized model eggs by ejection. However, considering the desertion rate decreased obviously from larger-to equal- and then to smaller-sized model eggs, with equal-sized model eggs having a 17.8% higher desertion rate than their smaller-sized counterparts, the results still imply a biologically meaningful pattern of rejection mode with change in model egg size. This supports the findings of previous studies that egg size influences host rejection modes but not overall rejection rates (Martín-Vivaldi et al., 2002; Roncalli et al., 2017). This may be because the larger the parasite eggs, the harder it is for the hosts to puncture or grasp them for ejection, and hosts thus opt to desert the nests to resist parasitism (Antonov et al., 2009).
Research into the effects of egg size on host egg recognition began in the 1980s. However, there are still only 21 studies on this topic, with most focusing on the hosts of parasitic cowbirds (e.g., Mason and Rothstein, 1986; Guigueno et al., 2014; Luro et al., 2018) and only a few involving hosts of parasitic cuckoos (e.g., Antonov et al., 2006; Meshcheryagina et al., 2020; Ye et al., 2022b). Additionally, most of these studies have focused on host species nesting in open or semi-open nests, whereas cavity-nesting hosts have rarely been investigated (Ye et al., 2022b). However, egg size recognition is hypothesized to have evolved in cavity-nesting birds because the ambient light conditions of nests were expected to be too poor for accurate recognition based on vision (Ye et al., 2022b). Size differences may be detected more easily in a dark environment because such detection can be achieved through tactile senses (Soler, 2017). Nevertheless, a recent study indicated that Green-backed Tits were able to persist in their recognition of parasitic eggs based on coloration, even when the luminance of nest boxes was reduced to 4.78 ± 1.31 lux. The hosts became “blind” to the parasite eggs only when the luminance was further reduced to near zero lux (0.35 ± 0.15 lux) (Yang et al., 2022). This reveals that Green-backed Tits possess a significantly strong rejection capacity under poor light conditions. Hosts tended to reject more parasite eggs that were larger or smaller than their own eggs under the normal luminance of nest boxes (38.11 ± 24.02 lux, Yang et al., 2022). This implies that although statistical significance was not found, tactile detection was involved in the rejection rates of the different egg sizes observed in this study.
Animals have evolved to prefer larger offspring due to their higher survival probabilities (Perrins, 1965; Trivers, 1972). This preference has been observed in birds, where parent birds favor larger eggs (Tinbergen, 1951; Perrins, 1965; Vidya, 2018) and prefer to feed larger nestlings (Soler et al., 1995; Bortolato et al., 2019). For parasitized hosts, this preference for larger eggs conflicts with the selective pressure to recognize parasitic eggs based on size, as many parasitic birds lay significantly larger eggs than their hosts (Payne, 2005; Yang et al., 2012). Therefore, Green-backed Tits may have potential conflicting adaptations between preferring larger eggs and rejecting cuckoo eggs. However, if the preference for larger offspring blocks the evolution of egg size recognition, hosts should not reject 21.4% more larger-sized eggs than equal-sized ones. Moreover, the rejection rate of smaller eggs was similar to that of larger eggs (53% vs. 51.7%, respectively). This further implies that Green-backed Tits did not show any sign of preference for larger eggs. We hypothesized that Green-backed Tits might have the capacity for recognition based on egg size contrasts. This capacity may act together with visual cues to detect parasite eggs, probably because the luminance of their natural nests varies with different cavities. In other words, recognition based independently on egg size may only evolve in host species that breed in nests with consistent and extremely low luminance. For example, Chestnut-crowned Warblers (Phylloscopus castaniceps), which are sympatric breeding birds, build nests in environments with very low luminance (close to 0 lux). As a result, they do not recognize non-mimetic eggs by coloration but rely on egg size as an independent cue to recognize parasite eggs (Ye et al., 2022b). Further studies on egg size recognition in Green-backed Tits, after controlling for nest luminance, are needed to test our hypothesis.
Regarding the role of parasite egg quantity in the host egg recognition process, the results showed that Green-backed Tits accurately rejected parasitic eggs, regardless of the differences in the numbers of parasite and host eggs in the nests. This indicates that Green-backed Tits recognize parasite eggs using a template recognition mechanism that is not based on egg quantity contrast between parasites and hosts. Template recognition allows hosts to rely on memory templates for recognition, unaffected by the ratio of parasite eggs to hosts (Rothstein, 1975; Stokke et al., 2007; Moskát et al., 2010). Hosts use either a template or discordant recognition during the process of recognizing parasite eggs. However, previous studies indicated that most hosts with egg-recognition abilities employ only the template recognition mechanism to identify parasitic eggs (Soler, 2017). Even if a few hosts can use discordant recognition to identify parasite eggs, this mechanism does not exist independently but usually operates as an assistant mechanism alongside the dominant mechanism of template recognition (Yang et al., 2014b; Wang et al., 2015). For example, the hosts of the Common Cuckoo (Cuculus canorus), Ashy-throated Parrotbill (Suthora alphonsiana) (Yang et al., 2014b), and Yellow-bellied Prinia (Prinia flaviventris) (Wang et al., 2015) can use both of these recognition mechanisms to identify parasite eggs. Furthermore, we observed that the rejection modes changed with parasite-host egg quantity contrast, with the probability of nest desertion by the hosts increasing as the number of parasite eggs in the host nests increased. This result has not been demonstrated in previous studies, and it implies that Green-backed Tits may be able to distinguish the number of parasitic eggs from their own eggs. Further studies are needed to explore the role of parasite-host egg quantity contrast in host recognition mechanisms and rejection modes. Finally, although this study confirmed that Green-backed Tits use a template recognition mechanism for egg recognition, it is currently unknown whether such recognition mechanisms are innate (Rothstein, 1975; Stokke et al., 2007) or acquired through learning (Lotem et al., 1992); thus, this question remains to be explored in future research.
In summary, this study indicates that (1) Green-backed Tits do not use egg sizes as primary cues to recognize parasite eggs; (2) the role of egg size in the evolution of egg recognition remains unclear and needs further studies controlling for nest luminance; (3) Green-backed Tits use template mechanisms for egg recognition; and (4) the relative sizes and quantity of parasite-host eggs affect the mode of parasite egg rejection by the hosts.
Xu Zhao: Writing – original draft, Investigation. Ping Ye: Writing – original draft, Investigation. Huaxiao Zhou: Investigation. Canchao Yang: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization.
The experiments reported here comply with the current laws of China. Fieldwork was carried out under permission from Kuankuoshui National Nature Reserve, P.R. China. Experimental procedures were in agreement with the Ethical Evaluation Group for Animal Behavior Study (permit no. EEGABS-009).
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
We thank Kuankuoshui National Reserve for its permission and support to carry out this study, and Neng Wu for his assistance to the field work.
Supplementary data to this article can be found online at https://doi.org/10.1016/j.avrs.2024.100216.
Response variable | Source | Sum of squares | df | Mean square | F | P |
Recognition reaction | Intercept | 0.397 | 1 | 0.397 | 1.515 | 0.226 |
Treatment | 0.75 | 2 | 0.375 | 1.432 | 0.251 | |
Laying date | 0.053 | 1 | 0.053 | 0.204 | 0.654 | |
Incubate day | 0.291 | 1 | 0.291 | 1.11 | 0.298 | |
Clutch size | 0.207 | 1 | 0.207 | 0.791 | 0.379 | |
Rejection mode | Intercept | 1.931 | 1 | 1.931 | 14.021 | 0.002** |
Treatment | 1.257 | 2 | 0.629 | 4.567 | 0.026* | |
Laying date | 0.035 | 1 | 0.035 | 0.255 | 0.620 | |
Incubate day | 0.003 | 1 | 0.003 | 0.02 | 0.889 | |
Clutch size | 0.147 | 1 | 0.147 | 1.065 | 0.317 | |
Treatment: Egg size contrasts between parasites and hosts (larger, equal, or smaller model eggs); Laying date: laying date of the first egg; * P < 0.05, ** P < 0.01, *** P < 0.001. |
Response variable | Source | Sum of squares | df | Mean square | F | P |
Recognition reaction | Intercept | 0.305 | 1 | 0.305 | – | – |
Laying date | 0 | 1 | 0 | – | – | |
Clutch size | 0 | 1 | 0 | – | – | |
Treatment | 0 | 1 | 0 | – | – | |
Incubate day | 0 | 1 | 0 | – | – | |
Rejection mode | Intercept | 0.074 | 1 | 0.074 | 0.517 | 0.482 |
Laying date | 0.085 | 1 | 0.085 | 0.593 | 0.452 | |
Clutch size | 0.413 | 1 | 0.413 | 2.89 | 0.107 | |
Treatment | 2.023 | 1 | 2.023 | 14.165 | 0.002** | |
Incubate day | 0.163 | 1 | 0.163 | 1.143 | 0.300 | |
The first model does not produce any statistical result because the response variable was a single value (i.e., only rejection was found in the recognition reaction); Treatment: egg quantity contrast between parasites and hosts (–5, –4, –3, –2, –1, 0, 1, 2, 3, 4, 5); Laying date: laying date of the first egg; * P < 0.05, ** P < 0.01, *** P < 0.001. |