Nest size matters: common cuckoos prefer to parasitize larger nests of Oriental reed warblers

Avian brood parasites leave parental care of their offspring to foster parents. Theory predicts that parasites should select for large host nests when they have sufficient available host nests at a given time. We developed an empirical experimental design to test cognitive ability of female cuckoos in nest size by studying nest choice of common cuckoos (Cuculus canorus) among nests of its Oriental reed warbler (Acrocephalus orientalis) hosts. We presented three groups of experimental nests: 1) nest dyads tied together including one large and one small artificial nest from reed leaves, 2) nest triads tied together used the old modified warbler’s own nests including enlarged, reduced and medium-sized nests, and 3) nest dyads are similar to group 1, but not tied together to elicit parasitism by common cuckoos. We predict that cuckoos prefer larger nest than medium one, the next is smaller nest. Our findings showed that common cuckoo females generally prefer large nests over medium or small sized nests in all three experimental groups. Furthermore, cuckoo parasitism was significantly more common than in previous studies of the same warbler population, implying that larger, higher and more exposed host nests effectively increased the probability of cuckoo parasitism.


Introduction
Avian brood parasites increase their fitness through evolved egg-laying behaviour in the nests of hosts while leaving parental care of their offspring to the hosts (Davies 2000;Soler 2014). Brood parasites always depend on parasitized hosts for the care of their offspring. However, foster parents differ in their food provisioning ability of cuckoo nestlings caused by variation in parental quality (Brooke and Davies 1989;Grim et al. 2017). Therefore, if altricial obligate brood parasites maximize the benefits from the parental effort of their hosts, cuckoos should search for and select host nests with better quality during parasitism (de Neve et al. 2004;Álvarez and Barba 2008;Jelínek et al. 2016).
Birds' nests are used for containing the eggs and/or offspring during breeding (Hansell 2000(Hansell , 2007, the potential signaling function of a high-quality nest being important for successful fledging to avoid hosts from suffering from predation or from falling out of the nest (Álvarez and Barba 2008;Heenan and Seymour 2011;Møller et al. 2014;Møller 2017;Honza et al. 2020). Thus, parasites should search for the most suitable host nest (Davies 2000). How cuckoos find and choose host nests for parasitism is still poorly understood, 1 3 and previous studies suggested that egg laying by cuckoos should match the phenotypes or appearance of their host eggs (Avilés et al. 2006;Cherry et al. 2007;Honza et al. 2014;Li et al. 2016a, b), while other studies challenged these ideas and suggested that cuckoos randomly searched host nests for parasitism Yang et al. 2016b). Recent empirical work showed that when experimental nests in which some have different model eggs were given a similar opportunity for cuckoo parasitism, cuckoos selected these nests for parasitism without use of the content of nests including egg color, shape and size as search cues (Yang et al. 2016a). They randomly selected any one nest for parasitism and neglected the egg categories in the nest. Furthermore, the cuckoo parasitism frequency in their population associated with host nest size tended to be higher in large compared to small nests, they found that cuckoo females more often selected larger Oriental reed warbler nests (Acrocephalus orientalis) over smaller ones of blackbrowed reed warbler (Acrocephalus bistrigiceps) when these two types of nests were available for cuckoo choice (Yang et al. 2017), which could be explained by difference in host species, but also indicated that nest size plays a role in the cuckoo's nest selection for parasitism.
The optimal egg-laying strategy proposes that when there are sufficiently many available nests at a given time, cuckoo females should prefer the nest made by the best quality hosts (Hauber 2007;Parejo and Avilés 2007;Trnka and Prokop 2011;Jelínek et al. 2014). In this scenario, cuckoos may select larger nests associated with increased fitness because larger nests may indicate superior parenting ability (Soler et al. 1998;de Neve et al. 2004;Avilés et al. 2009). A study spanning eight consecutive years showed that large magpie (Pica pica) nests were more likely to be parasitized by great spotted cuckoos (Clamator glandarius) than smaller nests when cuckoos have a larger availability of host nests among which to choose. Therefore, selection of active host nests may arise from nest choice by the great spotted cuckoo (Molina-Morales et al. 2016), probably because nest size correlates positively with parental abilities of magpies (Soler et al. 1995).
Here, we developed a field experimental method to test whether nest-size manipulation affects parasites' cognitive ability in nest preference when there are sufficiently many available different nest-size nests. The common cuckoo (Cuculus canorus) is the most common brood parasite using the Oriental reed warbler as its major host in our study area (Yang et al. 2016a(Yang et al. , 2017Wang et al. 2020a). Different types of artificial experimental nests can be categorized as large, medium or small size, and we divided these nests into two groups (nest dyads and triads) to test our hypotheses. We predicted that cuckoos would lay their eggs in larger host nests in both the two groups if they had evolved the ability to discriminate among different nest sizes.

Study site and study species
This study was performed in Zhalong National Nature Reserve (47° 19′ N, 124° 22′ E), which is located in Heilongjiang, and another study site in Sifangtuozi farm (46° 12′ N, 123° 84′ E), Jilin, China, where is 100 km far from Zhalong. Fieldwork was carried out during the breeding season (June-August) in years 2015-2019, and 2021. Habitats were both primarily reed swamps with several small villages (Yang et al. 2017;Wang et al. 2020b). The Oriental reed warbler females build the nest independently, and the common cuckoo is the main parasitic cuckoo in both study areas, with a high parasitism rate ranging from 34.3 to 65.5% among years in Zhalong for warbler nests (Yang et al. 2017), cuckoos often remove one or two eggs before they parasitize the host nest (Wang et al. 2020a). Furthermore, the appearance of cuckoo eggs resembled the eggs of their hosts well (Yang et al. 2016b;Li et al. 2016a).

Experimental nest manufacture
We designed three experiments to test the nest size hypothesis. For group 1, one large nest (internal diameter, mean ± SD: 86.42 ± 0.91 mm, n = 36) and one small nest (internal diameter, 67.57 ± 1.05 mm, n = 36) were tied together as a combination of nests ( Fig. 1, group 1). We tied two randomly chosen and different-sized nests together to constrain the choice of host nests to the experimental variable of nest size. This constrained the choice between the nest dyads, avoiding an absence of selection caused by distance between nests. This was because if these two experimental nests were located at a far distance, the cuckoo might not find them simultaneously and compare the two different-sized nests, especially when nests were located in dense reeds. These nests were weaved together with a mixture of dry reed leaves and sticks, sewing them with a needle and string. The purpose of this experiment was to control the nest structure, including reed material, shape, duration of nest construction, except for nest size, which was a unique character dividing nests into two categories. This helped us determine the relevance of nest size choice by cuckoos.
For group 2, 'triple nests' were installed in the field including large, medium and small sized old nests as described above. All nests used for this study were modified from natural old Oriental reed warbler nests. We used natural warbler nests as the medium category (internal diameter: 57.88 ± 5.00 mm, n = 33), while large nests were enlarged by 20 mm (internal diameter: 78.74 ± 4.34 mm, n = 33). Similarly, small nests were reduced by approximately 20 mm (internal diameter: 39.18 ± 3.42 mm, n = 33).
For group 3, one large nest (mean ± SD: 85.68 ± 1.04 mm, n = 12) and one small nest (67.29 ± 1.26 mm, n = 12) were at 1.0 m distance to attract cuckoos for parasitism (Fig. 1,  group 3). This treatment was different group 1 and helped make two artificial nests look more natural and similar to the real ones (Yang et al. 2016a(Yang et al. , 2017Wang et al. 2020bWang et al. , 2021. We collected natural old nests that had fledged young or were deserted the previous year. Nests were collected and preserved in cardboard boxes. For the large nest treatment, we selected two old nests that were most similar in appearance (size, shape and nest materials) for our experiment. First, we randomly selected one nest and made a cut in the rim of the nest with scissors. Second, we put one-third of the other nest piece together with the rim cut and sewed it up. Similarly, for small nests, we cut nest material and re-sewed Fig. 1 Example of combination of host nests for eliciting cuckoo parasitism. In group 1 and group 3, a large and b small nests are shown. In group 2, A large, B medium, and C small nests are shown. Arrows shown in red refer to eggs of the common cuckoo while the remainders are quail eggs 1 3 them to keep the internal diameter at approximately 40 mm. For the medium nests, we did not cut them, although we also sewed them with the same color string. We sewed these nests using needle and string to make them look similar in appearance to natural nests. We enlarged and reduced 36 nests in total for our nest experiments. We randomly selected one of three types of nests and sewed them together with string ( Fig. 1, group 2).

Experiments for attracting cuckoo parasitism
We searched systematically for naturally active Oriental reed warbler host nests and monitored reproductive activities of hosts daily. Observed nests were checked every day to confirm the first date of laying an egg. If these target nests were found to have one host egg, one set of the prepared combination of nests were tied to the reeds one meter away from this neighboring active nest, and two common quail (Coturnix coturnix) eggs were inserted into each of the three combinations of nests, respectively (Fig. 1). Here, we controlled that the two quail eggs stayed the same throughout the test period, avoiding that cuckoo females showed a preference bias towards variation in the number of eggs, and the large quail eggs had the advantage that warbler hosts cannot remove them, but the cuckoos can easily remove them.
To increase the risk of brood parasitism, we specifically tied the combination of nests higher (approximately 0.5 m) in the vegetation (the nest height hypothesis, Budnik et al. 2002;Patten and Reinking 2011) than the position of the active host nest. We made them easier to see from the angle of view (nest exposure hypothesis, Moskát and Honza 2000;Clarke et al. 2001), which allowed cuckoos to more easily choose a specific type of nest during parasitism. Cuckoos generally selected to lay their eggs in Oriental reed warbler nests (Yang et al. 2016a(Yang et al. , 2017. Thus, we did not design control nests that simultaneously set up another combination of nests in a random direction and the same nest content and position, as there was little probability for cuckoo parasitism in nests without host activity (Yang et al. 2017).
All experimental and naturally active nests were monitored for up to 6 days to confirm cuckoo parasitism until completion of the clutch and start of incubation. All experimental nests were video recorded and checked once per day. Video devices were retrieved at dusk. For active warbler nests, we did not manipulate, but checked and monitored clutch size and cuckoo parasitism.

Statistical analyses
Binomial tests (two-tailed test) were used to test for a preference of nests during choice by cuckoos (Wang et al. 2021). Differences were considered significant at p < 0.05. Statistical analysis was conducted using IBM SPSS Version 22.0 (IBM Corp., Armonk, NY, USA), and unless otherwise specified, all results are presented as mean ± SD.
There was no case of parasitism among naturally active warbler nests in all groups (0/36, 0/33, 0/42). Six cases of egg laying in large, medium and small nests are shown in Electronic Supplementary Materials Videos S1-S2 in group 1, and Videos S3-S4 in group 2 and Videos S5-S6 in group 3.

Discussion
The main finding of this study was that common cuckoos have the ability to distinguish different nest size, and prefer large host nests for parasitism when they have sufficiently many synchronous available nests. Our findings support the active selection hypothesis suggesting that cuckoos prefer to lay eggs in larger nests providing evidence for the ability to distinguish among host nests varying in size when monitoring available nests. In addition, the frequency of parasitism was higher than in previous studies of the same warbler population, which implies that host nests with a higher and exposed position would effectively increase the possibility of cuckoo parasitism (Yang et al. 2016a(Yang et al. , 2017. Nest size is an indicator of parental care, and it is a likely reason why cuckoos choose large host nests for parasitism. For example, magpie pairs with large nests provide higher quality of food for their nestlings compared to pairs with smaller cuckoo nests (de Neve et al. 2004). The great spotted cuckoo prefers to parasitize large magpie nests because nest owners of such nests can rear nestlings more successfully than control nests (Soler et al. 1995). The chicks in bigger nests great reed warbler (Acrocephalus arundinaceus) showed significantly lower falling-out rate (4 of 29) compared with those remained in smaller Eurasian reed warbler (Acrocephalus scirpaceus) nests (12 of 32), this study suggest that host nest size plays an important role in the suitability of host species and that smaller host nests may pose a high mortality risk to the relatively big cuckoo chicks (Honza et al. 2020). Nest size in the great reed warbler serves as a signal of female quality for investment in reproduction, because enlarged great reed warbler nests significantly increase male feeding effort (Trnka and Prokop 2011;Jelínek et al. 2016). However, no evidence supports the hypothesis that nest size affects the frequency of parasitism in the great reed warbler (Avilés et al. 2009;Jelínek et al. 2015). We showed that cuckoos laid their eggs in larger host nests suggesting that cuckoos have the ability to distinguish among different nest sizes. The combination of different types of experimental nests presented next to each other makes it feasible for cuckoos to choose a nest without consideration of distance, allowing a similar preference for different-sized nests, and the bigger ones were bias chosen. Perhaps this was the reason why our finding was different from the result of Jelínek et al. (2015), which they only provided one experimental nest for cuckoo parasitism; accordingly, the cuckoos have no choice when they come to lay eggs.
When cuckoos approach a combination of nests with a similar angle of view or height, they may be subject to physical constraints, such that it is easier to perch on and deposit eggs in larger rather than smaller nests. In such case, large nests should be easy to find and facilitate egg laying. Of course, this scenario depends on sufficiently many nests being available in the neighborhood. Ma et al. (2018) showed that isolated Oriental reed warbler host nests were more vulnerable to cuckoo parasitism far away from neighbors, cuckoo females having no other choice being forced to select the only isolated nest for parasitism. Our results show that parasitism frequency is higher than in the previous studies (Li et al. 2016a, b;Ma et al. 2018), the conspicuousness of host nests for the likelihood of being parasitized by the cuckoo are vulnerable (Patten and Reinking 2011;Clarke et al. 2001), that is, the higher and more exposed are host nests, the more effectively the increase in the probability of cuckoo parasitism, consistent with the nest height hypothesis (Budnik et al. 2002) and the nest exposure hypothesis (Moskát and Honza 2000;Patten and Reinking 2011).
There may be other reasons for why cuckoos prefer larger nests. For example, there might be a physical body constraint, because many small hosts may have evolved a small sized nest allowing for parasitism by small cuckoo parasites (e.g., Bianchi's warbler Seicercus valentine vs. Asian emerald cuckoo Chrysococcyx maculatus) (for more examples, see Yang et al. 2012). The only evidence is that larger nests in many different species of birds lay larger clutches, and larger clutches being associated with higher quality nest owners (Møller et al. 2014). In this study, two large quail eggs in large nests may also better predict larger final clutch size, two large quail eggs in small nests where there is no free space for other eggs may indicate small size of a complete clutch. This could also result in a higher preference of cuckoos for larger experimental nests. Moreover, Wang et al. Fig. 2 Nest parasitism frequency and sample size of experimental nest types (large and small) in group 1 and group 3, and (large, medium and small) in group 2 exploited by common cuckoos in the field. Probabilities are from statistical tests of the null hypothesis of similar rates of parasitism. Numbers above the bars refer to test sample 1 3 (2020b) revealed that cuckoos prefer to select the naturalactive host nests with small clutch sizes for parasitism [e.g., 1-2 host eggs, 75% (183/245)], and the empirical study also supporting this result (78%, 25/33) (Wang et al. 2020b). In addition, a large clutch size (e.g., 5-6 eggs) also implies more space for cuckoo egg laying. Therefore, nest space may act as an important cue for cuckoo egg laying resulting in strong nest preferences.
In conclusion, the present study provided experimental evidence suggesting that cuckoos preferred to lay their eggs in larger host nests when they have more available nests for use at any given time. Furthermore, the conspicuousness of host nests caused by higher location and full exposure in the field resulting in more cuckoo parasitism, suggesting that host nest exposure could increase the risk of cuckoo parasitism.