Batesian mimicry in brood parasites is often viewed as an evolutionary strategy to mitigate host aggression. Female common cuckoos (Cuculus canorus) exhibit two morphs: the hawk-like grey and the rufous one, potentially maintained by apostatic selection. It was hypothesized that the grey morph's predator-like appearance deters host defences, while the rufous morph benefits from its rarity by evading host attention. Previous research predominantly utilized static cuckoo dummies, lacking insights into real-world interactions. We investigated the effectiveness of the cuckoo morphs in accessing great reed warbler (Acrocephalus arundinaceus) nests under natural conditions. Analysing video-recorded cuckoo attempts, we found no significant difference in nest-access success between the morphs. Both experienced a similar probability of physical attacks when hosts were present, and the rufous morph did not increase success in the nest-access attempts in the absence of hosts. These results fail to support the assumptions of (1) Batesian mimicry that hawk-like mimicry enhances nest access or reduces host aggression, and (2) apostatic selection that the rarity of the rufous morph confers an advantage in the successfully accessing the host nest. Future research should aim to identify stages in the cuckoo's life cycle or host interactions where colour polymorphism provides an evolutionary benefit.

Data collection

Data were collected during the breeding seasons 2020‒2022 within the fishpond area situated between Hodonín (48°51'0"N, 17°07'0"E) and Mutěnice (48°54'0"N, 17°02'0"E), in the Czech Republic. We employed a video-recording setup of one camera focused on the nest from a close distance of <1 metre and another camera installed on a tripod approximately 4–10 metres from the nest, monitoring the nest surroundings. Most of the nests were video-recorded from the day the nest was completed throughout the egg-laying period. In a minority of nests, the filming started anytime during the egg-laying period. The closer-positioned camera was a Carmedien STO-IR rear-view camera (Carmedien, Tschernitz, Germany) with IR illumination connected to a miniature Mini DVR CH-HD0065 digital video recorder (Shenzhen Chu-Tech Co. Ltd., Shenzhen, China) placed in a water-resistant box or a custom HD camera with inbuilt recorders without IR illumination. The cameras were powered by 12 V/100 Ah gel batteries placed on the bank. In 2021, these cameras were replaced with motion sensitive BUNATY micro 4K trail cameras powered by AA batteries and placed about 1 metre from the nest. The camera capturing the wider area around the nests was a Sony HDR-CX 240 powered by a power bank (GoGEN 20000 mAh) with a scanning range of about 10 m from the nest on each side, again depending on reed density. Daily checks were conducted to record the nest status, replace the power bank and SD card. All the recording equipment was camouflaged by green cloth and reed stems to prevent disturbance of nesting birds. The use of simultaneous filming allowed us to analyse cuckoo-host interactions around the nest and at the nest, determining the outcome of cuckoo activity (unsuccessful access or successful access to the nest; and in the latter case further: parasitism, predation or only nest checking event). We considered nest access successful when a cuckoo made physical contact with the host nest. The location of the nest in the reeds sometimes prevented us from detecting cuckoo activity using a more distant camera, so we used only recordings from a closer camera that captured the nest.

Analyses

Firstly, we examined if the hosts feared the hawk-like grey cuckoo more than the rufous morph. We specifically tested if the probability of physical attack by host (binary response variable; yes or no) differed between cuckoo morphs (categorical; grey or rufous). In this model, we additionally tested the effect of the number of host parents present (1 or 2), the clutch size at the time of event (0–5), the time of day (hour), and the ordinal date within the breeding season (1 = 1st January). To account for repeated cuckoo visits to the same nest, we included a random intercept for nest ID in our model. We assessed potential multicollinearity using the variance inflation factor (VIF), which was satisfactory for all predictors (VIF < 2.0). We employed a Bayesian model with Bernoulli distribution and logit link, which effectively solved the problem by setting a weakly informative prior for fixed effects to be the normal distribution of the mean set to 0 and the standard deviation to 10. For this purpose, we used the package brms (version 2.20.4) in R software (version 4.3.2).

Next, we tested whether the rarer rufous morph evades the host attention more often compared to the grey morph. This should be indicated by a higher proportion of rufous morph observed at the nests, where hosts are absent, compared to the nests with hosts present. Again, we employed a Bayesian mixed-effects model with Bernoulli distribution with response variable of cuckoo morph and predictor of the presence of host parents during a cuckoo visit (at least one host present or hosts absent). The additional predictors and model structure have been identical as in the previous model. Further, we aimed to identify the factors that influence the success of cuckoos in accessing host nests (binary response variable; yes or no). We primarily examined the potential effects of the presence of host parents during a cuckoo visit (yes or no), effect of cuckoo morph (grey or rufous) and then included the same additional predictors as in the above models. In this model, there was a problem with a complete separation (all the cuckoos successfully reached the nest if no host was present; n = 29). We effectively solved this problem by employing a Bayesian model and setting a weakly informative prior for fixed effects to be the normal distribution of the mean set to 0 and the standard deviation to 2.

Finally, we used the same Bayesian mixed-effects model to explore the effect of the number of host parents present (1 or 2) on the success of cuckoos to access host nests (binary; yes or no). In this model, we excluded all cases where no parent was present during the cuckoo attempts and applied the same model structure and settings as in the previous model. The only exception was setting a weakly informative prior again to the mean of 0 and standard deviation of 10.