Cuckoos use host egg number to choose host nests for parasitism

Abstract

To maximize their offspring success common cuckoo (Cuculus canorus) females should lay their eggs into host nests before incubation has begun. This ensures that the parasite chick hatches before all host chicks and can evict its foster siblings to monopolize host parental care. Many studies have demonstrated that most cuckoo eggs are indeed laid before the onset of host incubation. But cues used by female cuckoos to choose the right nest at the right time remain unclear. Here, we combine field observations with a field experiment to test whether female cuckoos use the number of eggs in the nests of their Oriental reed warbler (Acrocephalus orientalis) hosts to direct their choice. Over 8 years of field observations and 5 years of experiments, cuckoo females placed the majority of their eggs in nests with fewer than three host eggs, i.e. early in the laying sequence. For natural nests, the cuckoos may use information gleaned from the activity and behaviour of the host parents to make their choice. In our sets of experimental nests containing different numbers of model eggs, the vast majority of parasitism events occurred in nests containing a single egg. To our knowledge, this is the first field experiment showing that cuckoos choose host nests for parasitism based on the number of host eggs they contain. It appears that cuckoo females use the egg number to estimate the appropriate host nest stage for timely parasitism.

1. Introduction

Avian interspecific brood parasites lay their eggs in nests of host species, often small passerine birds, and leave parental care of their offspring to the foster parents [1]. This imposes fitness costs on the host, selecting for counter-adaptations by hosts to resist parasitism [2,3]. Such host–parasite interactions provide a model system for studying coevolution [4,5]. The common cuckoo, Cuculus canorus, has evolved behavioural adaptations that include rapid egg laying of less than 10 s that allows them to avoid attack by host parents [6] and timing egg laying within the laying sequence of the hosts so that the parasite chick hatches before the host chicks [7–10].

Early hatching followed by the eviction of all other eggs in the nest grants the chick of the common cuckoo the advantage of uncontested access to nest resources [1,9,11]. However, when the parasite egg is laid too late in the hosts’ laying sequence, the chick is unable to evict the larger chicks and must share the nest. Furthermore, when common cuckoos parasitize the common redstart (Phoenicurus phoenicurus) not all host eggs are evicted from all nests, and cuckoo chicks that cohabit with host nestlings have lower survival and fledging rates than those reared alone [11,12]. Therefore, there is a clear advantage for the cuckoo to lay eggs into nests before the beginning of incubation [13], to ensure that the cuckoo chick hatches before the hosts. Indeed, it should be advantageous for cuckoos to be able to judge the appropriate time for egg laying in each host nest.

When a female cuckoo finds a host nest, she can observe the behaviour of the host parents, for example, whether they are still involved in nest construction and how much time they spend on the nest, whether other cuckoo females are in the vicinity and, when she visits the nest, the number of eggs it contains. Several bird species have elaborate quantification skills [14–19]. Female American coots (Fulica americana) use visual cues to adjust the clutch size and recognize foreign eggs, suggesting that they can count eggs ([20]; but see [21]). Brood-parasitic female cowbirds, under laboratory conditions, prefer to lay eggs into host nests containing a suitable number of eggs [22,23], though a similar effect was not observed for great spotted cuckoos under field conditions [24,25]. Here, using a combination of field observations and a field experiment, we investigate nest choice for parasitism by female common cuckoos and test whether variation in egg number in the nest influences their choice of which host nests to select for their eggs.

2. Material and methods

(a). Study site and study species

This study was performed in reed swamps in the Zhalong National Nature Reserve (46°48′–47°31′ N, 123°51′–124°37′ E) in Heilongjiang, northeast China where the common cuckoo exploits the Oriental reed warbler (Acrocephalus orientalis) as its main host. At this site, the Oriental reed warbler suffers parasitism by the common cuckoo ranging from 34.3% to 65.5% among years (see [26]). Furthermore, the cuckoo eggs observed closely resemble those of the warbler host [27,28]. Field observations were made each breeding season (June to August) from 2012 to 2019 and experiments were carried out from 2015 to 2019.

(b). Field observations

We systematically searched reed beds during the breeding season and targeted warbler nests by monitoring the activities of host parents in 2012–2019. When we found a new warbler nest, we visited it daily and recorded the date of the first egg laid and the number of eggs, both host and cuckoo, present each day. Multiple parasitism occur in this warbler population [29,30], so we recorded the date of appearance of the first and second cuckoo egg if such appeared. After breeding ended, we collected warbler nests for the experimental nests described below.

(c). Field experiment

During each breeding season from 2015 to 2019, once warblers started to lay eggs, a set of four experimental nests from the year before was set up in the same habitat type within about 1 m of an active host nest at a height of about 1.2 and 1.5 m above the ground. This was somewhat higher than natural nests at this study site (0.81 ± 0.24 m (mean ± s.d.), n = 44; [29]) to make them easier for cuckoos to find [31,32] and parasitize [33,34]. Each set of experimental nests included a nest containing 0, 1, 3 and 5 white model eggs (mean size of model eggs: 30.22 mm × 21.64 mm, 11.63 g) (figure 1).

Figure 1.

Example of a set of experimental host nests containing 0, 1, 3 or 5 white model eggs near an active host nest. (Online version in colour.)

Experimental nests were monitored for cuckoo parasitism by daily visits and video recordings for 6 days until the completion of the host clutch in the neighbouring active nest. Video devices (JWD DV-58G, JWD Inc., Shenzhen, China) were installed in the morning and were removed at dusk except on rainy days.

(d). Statistical analysis

We used chi-square tests to determine whether parasitism events were independent of the number of host eggs in the nest. Differences were considered to be significant at the 0.05 level. Statistical analysis was conducted using IBM SPSS version 22.0 (IBM Corp., Armonk, NY, USA).

3. Results

(a). Natural frequency of parasitism

The 245 cases of parasitism we observed over 8 years (2012–2019) were highly significantly heterogeneous across nests with different numbers of eggs (χ² = 241.97, p < 0.001). Most (75%) cuckoo eggs were deposited in nests with one or two eggs (49% and 26%, respectively; figure 2).

Figure 2.

Cumulative frequency of natural parasitism by cuckoos of host nests containing different numbers of eggs from 8 years of observations from 2012 to 2019 at our study site. Numbers above the bars refer to the number of observations out of the total 245.

We found 43 nests that received two cuckoo eggs during the field observations from 2012 to 2019. We discovered 10 nests that already contained two cuckoo eggs, so we cannot know at what exact stage they were laid. For 33 cases of double parasitism, we know the date of the appearance of each of the cuckoo eggs in the host nest. In eight nests, all containing a single host egg, two cuckoo females laid their eggs on the same day. Overall, the second cuckoo egg was laid one day later than the first (mean = 1.09 days, s.d. = 0.84, n = 33). For all doubly parasitized nests, the first cuckoo egg appeared when there was a maximum of two host eggs in the nest (figure 3).

Figure 3.

Frequency of natural parasitism by the first and second cuckoo female of host nests containing different numbers of eggs for the 33 cases of double parasitism we observed. Numbers above the bars refer to the number of observations.

(b). Parasitism of experimental nests with different numbers of eggs

Of the 42 experimental sets of nests we set out from 2015 to 2019, we observed a single event of cuckoo parasitism in 32. Parasitism was highly significantly heterogeneous across the nests containing different numbers of model eggs (χ² = 31.93, p < 0.001), with most cuckoo eggs laid in nests containing a single model egg (figure 4).

Figure 4.

Frequency distribution of parasitism by cuckoos of experimental host nests containing different numbers of model eggs. Each set of experimental nests comprised four types of nests with 0, 1, 3 or 5 eggs. Numbers above the bars refer to the number of observations of parasitism to each nest type.

We observed no cuckoo parasitism in the paired natural active host nests during the experiment, suggesting that our experimental nests were indeed easier to find and to parasitize. Video recordings showed cuckoo female egg-laying behaviour in these experimental sets of nests with 0, 1, 3 or 5 model eggs (electronic supplementary material, videos S1–S4).

4. Discussion

Female cuckoos laid eggs more often in nests that contained fewer host or model eggs both in natural and experimental nests. This bias towards nests containing few eggs, early in the host laying sequence, increases the likelihood that the cuckoo chicks would hatch before most or all of their host nest-mates, allowing them to remove all host young and monopolize resources provided by the foster parents. Our observations on natural nests reinforce previous, similar observations [2,35]. The originality of our study lies in the experiment that elucidates the mechanisms by which cuckoo females make their egg-laying decisions. Indeed, preference for natural nests early in the host laying sequence could result from cuckoos observing and responding to behaviours of the host parents. The strong preference for nests with a single model egg we observed in our experimental nests shows that cuckoos used the information provided by the number of eggs in the nest in absence of any other cues. Thus, cuckoo females were able to distinguish among nests containing different numbers of eggs and use this information to influence their nest choice behaviour.

There are several reasons why cuckoo females should lay their eggs early in the laying cycle of the nests they parasitize. These include avoiding attack by hosts defending their nests, minimizing the opportunities for reliable detection of the cuckoo egg by egg-rejecting hosts and ensuring that their chicks hatch before those of the hosts. Moksnes et al. [6] found that reed warblers (Acrocephalus scirpaceus) are less active at their nest during the first and second laying days than later, and defence attacks on cuckoos become more intense after the laying of the fourth host egg [13]. Thus, early in the host laying sequence hosts may be less vigilant, offering more opportunities for cuckoos to lay an egg without being detected and attacked. Hosts are aggressive in their nest defence against cuckoos [5,36,37], can drive them off, thereby preventing them from laying eggs [38] and even injure or kill them [1]. During this study, in 2014, we even observed a female cuckoo that drowned after being forced into the water below one host nest by her aggressive hosts.

Rejection of foreign eggs, an important defence against cuckoo parasitism, requires that the host either can compare and discriminate against a discordant phenotype or know the phenotype of her own eggs [39], which she learns over the course of laying [40]. Early in the laying cycle, therefore, it makes sense that the host has had less experience with her own eggs and has less means for comparison for identifying a foreign egg. Therefore, laying early in the host's laying sequence could be a way of confounding this host defence mechanism and decreasing the probability of egg rejection (but see [41]). Indeed, because cuckoo females usually remove a host egg before depositing one of their own in a warbler nest [26,42], the host would have little opportunity for comparing egg phenotypes if the cuckoo parasitizes a nest containing a single egg. On the other hand, it may be inadvisable for a cuckoo to lay in a nest before the host has laid a first egg. First, the presence of a host egg in the nest is a strong signal that the nest is active and has not been abandoned. Second, cuckoo eggs that appear before the first host egg may elicit hosts to eject or bury the egg or abandon the nest [1,43], though some cuckoo eggs laid in empty nests are accepted ([13,35,41,44] and this study).

Cuckoo chicks that are reared alone, without competition from host chicks, are more successful than those that have host nest-mates [11], and cuckoo chicks hatched from eggs laid before the start of incubation are usually able to expel all nest-mates, whereas those laid later may not be able to [13]. Overall, eggs laid early in the host laying sequence have higher hatching success and survival than those laid later [13,45]. Therefore, cuckoo females should lay their eggs early in the host laying sequence to ensure their offspring's early hatching and ability to secure a competition-free environment.

What kind of information can cuckoos obtain in order to judge the appropriate time for egg laying? When approaching potential host nests, female cuckoos usually perch at a vantage point and monitor host activities [6,46]. They, thus, gather information on nest location, the status of host nests from the activity of nest building, and even the onset of egg laying and incubation [6,26,42,44,46,47], which they may use to choose when to lay their eggs.

We set out to determine whether, in addition to the information about the host parents, cuckoo females use information about the stage in the host laying sequence given by the egg number in the nest. Our experimental nests, though in the vicinity of an active host nest, were more visible, and easy to parasitize [31,32], being placed higher in the vegetation [33,34], and were not visited by host adults. Therefore, the only information available to the cuckoos for choosing among the four experimental nests was the number of model eggs already present. If cuckoo females did not use this information, we would expect their eggs to be distributed at random among the four nest types, which was not what we found. Our experiment demonstrates clearly that the cuckoo females used the variation in egg number in the nest to make their choice.

5. Conclusion

In summary, this study demonstrates that cuckoos were able to discriminate and made decisions based on the number of eggs in a host nest. Female cuckoos generally preferred to lay their eggs in host nests early in the host laying sequence, and here, we provide a mechanism by which they make their choice. To our knowledge, this is the first field experiment showing that cuckoos differentiate among nests that differ only in the numbers of eggs they contain. Future research should focus on an experimental test of cognitive ability of the common cuckoo and should expand on the mechanism driving the findings of this study.

Ethics

The experiments comply with the current laws of China, where they were performed. Fieldwork was carried out with permission (no. ZL-GZNU-2019–06) from the Zhalong National Nature Reserve, Heilongjiang, China. Experimental procedures were in agreement with the Animal Research Ethics Committee of Hainan Provincial Education Centre for Ecology and Environment, Hainan Normal University (permit no. HNECEE-2012-003).

Data accessibility

All data analysed and videos for this study are available at the Dryad Digital Repository: https://doi.org/10.5061/dryad.47d7wm396 [48].

Authors' contributions

W.L. and L.W. designed the study; L.W. and G.H. carried out field experiments; L.W. and C.Y. performed laboratory and statistical analyses. L.W. wrote the draft manuscript, and W.L. and A.P.M. were involved in the discussion and helped improve the manuscript. All authors approved the final submission.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the National Natural Science Foundation of China (nos 31660617 and 31960105 to L.W., 31672303 to C.Y., 31772453 and 31970427 to W.L.). L.W. was funded by a PhD grant from Guizhou Normal University (no. 0516009), the Initial Fund Key Laboratories of Guizhou Province (grant no. 2011-4005), the Joint Fund of the National Natural Science Foundation of China and the Karst Science Research Center of Guizhou Province (no. U1812401).