Modelling the nesting-habitat of threatened vulture species in the caucasus: An ecosystem approach to formalising environmental factors in species distribution models

1.   Introduction

Avian interspecific brood parasitism, the laying of eggs in the nest of other species to evade subsequent parental care duties, represents an important threat to host breeding success (Davies, 2000). Brood parasites can reduce the reproductive output of the host in several ways. The nestlings of the Common Cuckoo (Cuculus canorus) and other 'evicting' brood parasites cause complete host reproductive failure as their young usually eliminate all host progeny by evicting or directly killing them (Honza et al., 2007; Spottiswoode and Koorevaar, 2012). Parasitic females also destroy nests unsuitable for parasitism to force hosts to renest, thereby increasing the chances of future parasitism (Zahavi, 1979; Soler et al., 2017). In addition to these direct reproductive costs, brood parasitism can lead to other indirect costs for the host. Prolonged care for the parasitic chick and increased number of replacement nests, for example, may reduce host survival and future reproduction (Koleček et al., 2015). Thus, brood parasitism can, in some cases, be even more costly for host individual fitness than nest predation (Rothstein, 1990; Payne, 2005).

Given that the productivity, the rate at which animals produce their offspring, is a key demographic measure of population growth, reduced productivity due to brood parasitism may also have important population level consequences for host species. These effects on host populations vary depending on the parasite species and the host. Generalist brood parasites, like the cowbirds, may dramatically reduce the growth rates of heavily parasitized host populations and lead to their significant declines (Smith et al., 2002; Kosciuch and Sandercock, 2008), whereas host-specific brood parasites, such as cuckoos, are supposed to have only little population level effects on host species as they occur at low densities, i.e., being much less abundant than their hosts (Payne, 1977; May and Robinson, 1985). However, when the parasitism rate by the Common Cuckoo is extraordinarily high for a long time (64%–66%; Moskát and Honza, 2002), such a host population is usually maintained by immigrants from other, less parasitized or unparasitized populations (Barabás et al., 2004; Krüger, 2007). So, the question is how does brood parasitism affect local, more or less isolated populations of hosts with varying intensity of defence against parasitism? However, while most studies of the Common Cuckoo parasitism have to date focused primarily on specific host anti-parasite behaviours and parasite counter-adaptations, i.e., coevolutionary arms-race between parasites and hosts, the extent to which the Common Cuckoo threatens individual host populations is poorly understood.

The objective of this study was to determine the effect of Common Cuckoo parasitism on the annual productivity, i.e., number of fledglings produced per female within a nesting season, of a regularly parasitized host population. To assess this issue, a local population of one of the major cuckoo hosts in Europe, the Great Reed Warbler (Acrocephalus arundinaceus) (Leisler and Schulze-Hagen, 2011), was monitored for 15 years (2008–2022) in south-western Slovakia, and the factors that might play a role in the annual reproductive success of this host population were evaluated. Specifically, I tested for an effect of year, population size (number of breeding females), laying date, the length of the breeding season, clutch size, polygyny rate, nest failure, successful parasitism rates (i.e., when cuckoo chick successfully evicted host eggs or chicks), local spring temperature and spring total precipitation. I predicted lower annual productivity of the study population in years with higher rates of successful parasitism. Since nest predation is one of the most important factors influencing reproductive success in birds (Ricklefs, 1969; Martin, 1993), I also expected reduced annual productivity with increased nest failure rates due to nest predation and other causes such as nest destruction and nest abandonment.

2.   Materials and methods

2.1   Study site and study species

The study was carried out in a fishpond system near Štúrovo, southwestern Slovakia (47°51′ N, 18°36′ E, 114–116 m a.s.l.) with a total area of 45 ha. The ponds were surrounded by narrow belts of reed beds (maximum width 2–10 m) and scattered willows used by female cuckoos as vantage points for watching potential hosts and locating their nests. Due to intensive pond management, the littoral vegetation was maintained at the similar extent throughout the study period. The most frequent predators of reed passerine nests in the study area were the Little Bittern (Ixobrychus minutus) and the Marsh Harrier (Circus aeruginosus).

The Great Reed Warbler is a medium size (31 g) open-nesting passerine breeding throughout mainland Europe and the West Palearctic (Cramp, 1992). Birds arrive on their breeding grounds in mid-April – early-May and start laying eggs in May, continuing until late July. In central Europe, the Great Reed Warblers breed usually once per year, two broods in one season are very scarce (Trnka and Samaš, 2024), and on average lay clutches of five eggs. Only the female builds the nest and incubates the clutch, however, both the female and male feed and defend the young (Požgayová et al., 2009; Trnka and Grim, 2013). The Great Reed Warbler has a polygynous mating system, with polygynous males providing significantly less parental care to their offspring than monogamous males (Sejberg et al., 2000; Požgayová et al., 2015). The rates of social polygyny range between 8% and 56% (Catchpole et al., 1985; Hasselquist, 1998; Leisler and Wink, 2000). The nests of the Great Reed Warbler suffer relatively high rates of predation and Common Cuckoo parasitism. Rates of parasitism (the percentage of nests in which cuckoo eggs were laid) in European Great Reed Warbler populations range from 0 to 68 % (Moskát et al., 2008; Trnka et al., 2012). There is also known intraspecific nest predation in this species, where lower-ranking females destroy eggs of higher-ranking females to gain increased paternal investment (Trnka et al., 2010).

In the study area, the Great Reed Warbler is the predominant reed-nesting passerine species. Its population comprised 18–56 breeding females during the study period. For specific breeding parameters of the studied population, see Trnka and Samaš (2024).

2.2   Data collection

The fieldwork was conducted in 2008–2022. All birds in the study population were individually marked with one aluminium ring and a unique combination of two or three coloured plastic rings (adults). Great Reed Warbler nests were searched systematically at 4–5 day intervals between May and late July and then checked at 1–2 day intervals where possible to determine the day of clutch completion, the final clutch size and to detect the presence of a cuckoo egg (see also Trnka et al., 2012). As almost all nests were found during nest building or egg-laying, the onset of laying was assessed directly, and only in some cases it was recalculated according to the date of hatching. The brood size and number of fledglings were recorded when the young were 10–12 days old. The social mating status of each male and female was determined based on their captures at or near the nests or on direct observations of colour-ringed birds defending their nests or feeding their young. The rate of polygyny was calculated as the percentage of males mated with two or more females, i.e., number of polygynous males divided by the total number of males in the population. The length of the breeding period was defined as the time between the first egg laid and the last young fledged or last nest failure (see also Trnka and Samaš, 2024). The annual rates of successful cuckoo parasitism were calculated as the number of nests in which cuckoo chick successfully hatched and evicted host eggs or chicks/total number of nests where cuckoo eggs were laid. Analogically, the nest failure rate was calculated as the proportion of nests that failed due to predation, abandonment, destruction (e.g., by wind), flooding, etc. A nest was considered predated if eggs or nestlings disappeared or only eggshells were found in the nest cup. Because of the uncertainty in determining the causes of egg losses in some clutches (see Discussion for details), cases of partial clutch predation were not included in the analyses. Abandoned nests were nests with incomplete or complete clutches with cold eggs that were seen without attending breeders after two consecutive inspections. In the case of nests destroyed by wind or larger animals (e.g., swans, herons, wild boars), the reeds with the nest as well as neighbouring reeds were bent or broken, and the nest itself was often squeezed, leaving at least some nest contents (eggs or dead young) in or under the nest. Flooded nests were nests covered by surface water after heavy rainfall and increased water level in ponds. Annual productivity was calculated as the mean number of fledglings produced per breeding female throughout the breeding season. The meteorological data were obtained from the national climate station (the city Hurbanovo) located 28 km west of the study site.

2.3   Statistical analysis

All statistical analyses were conducted in R v. 4.3.2 (R Core Team, 2023). We used a general linear model with normal distribution and identity link to test the effects of breeding and environmental variables on annual productivity. Continuous predictors included year, total number of breeding females as a proxy of population size, median of first egg laid in breeding season (1 = 1st May), length of breeding season (in days), average clutch size, annual nest failure rate, annual successful parasitism rate, average temperature (℃) from April to June, and total precipitation (mm) from April to June. We standardized all predictors to have a mean of 0 and a standard deviation of 1 because the covariates were on different scales. This also improved the interpretability of the intercept and model overall, ensuring that the coefficients reflect the standardized effect of each predictor.

We presented the outputs of the full and final models. The final model was the one with the lowest AIC_c and was selected using function dredge in the package MuMIn v. 1.47.5 (Bartoń, 2023). Potential collinearity among the covariates was low, and variance inflation factors were < 2 in all cases (Zuur et al., 2010). We performed model residual diagnostics with R package RVAideMemoire v. 0.9-83-7 (Herve, 2023) and did not detect any violation of model assumptions.

3.   Results

A total of 495 breeding pairs (females) of Great Reed Warblers successfully reared 1321 of their own young and 63 young of the Common Cuckoo in the study population over the course of 15 years. Annual productivity of birds ranged from 1.7 to 4.3 (Fig. 1) and was primarily and negatively affected by the successful Common Cuckoo parasitism (Table 1, Fig. 2A). A 21% increase in parasitism rate reduced annual productivity by approximately one host chick. Likewise, annual productivity was negatively affected by nest failure rate, where a 23% increase in nest failure rate resulted in a reduction in annual productivity for one host chick (Table 1, Fig. 2B). A less pronounced but significant effect was also found for the number of breeding females, where a reduction in annual productivity of approximately one chick was associated with an increase in population size of 40 breeding females. Other environmental and breeding parameters had no effect on the annual productivity of Great Reed Warbler females in the study population. Similarly, successfully parasitized nests with a young cuckoo were predated to approximately the same extent as non-parasitized nests with host nestlings regardless of their age (19.1% vs. 21%, respectively, χ² = 0.65, df = 1, p = 0.29).

Figure  1.  Annual productivity (number of fledglings per female per year) over the 2008–2022 period. Numbers at the top of the bars indicate population size (total number of breeding females).

DownLoad: Full-Size Img PowerPoint

Table  1.  Output of full and final general linear models with normal distribution testing effect of breeding and environmental variables on the annual productivity in the Great Reed Warbler. Final model was selected according to the lowest AIC_c.

Model	Parameter	Estimate ± S.E.	t value	P value
Full	Intercept	2.89±0.05	54.35	< 0.001
	Year	−0.04±0.08	−0.55	0.61
	First egg laid	−0.09±0.19	−0.48	0.65
	Number of breeding females	−0.19±0.08	−2.43	0.07
	Clutch size	−0.09±0.18	−0.51	0.64
	Season length	−0.10±0.09	−1.05	0.36
	Polygyny rate	0.00±0.12	0.02	0.99
	Nest failure rate	−0.27±0.06	−4.32	0.01
	Parasitism rate	−0.50±0.08	−6.54	0.003
	Temperature	0.02±0.07	0.33	0.76
	Rainfall	0.04±0.08	0.53	0.62
Final	Intercept	2.89±0.05	59.78	< 0.001
	Number of breeding females	−0.20±0.05	−3.89	0.003
	Nest failure rate	−0.29±0.05	−5.72	< 0.001
	Parasitism rate	−0.55±0.05	−10.57	< 0.001

 | Show Table

DownLoad: CSV

Figure  2.  Effect of (A) parasitism rate, (B) nest failure rate and (C) total number of breeding females on the annual productivity in the Great Reed Warbler.

DownLoad: Full-Size Img PowerPoint

4.   Discussion

Contrary to the general assumption that host-specific brood parasites have only little impact on the productivity of their hosts at the population level because of the small proportion of parasitized nests (Payne, 1977; May and Robinson, 1985), the results of this study showed, for the first time to my knowledge, a significant negative effect of Common Cuckoo parasitism on the annual productivity of a local Great Reed Warbler population. More importantly, successful nest parasitism contributed to a reduction in Great Reed Warbler annual productivity at about the same rate as overall nest failure caused by predation, nest destruction, or nest abandonment. This is not so surprising given that the annual rate of successful parasitism of the Common Cuckoo was as high as 18%–44% in some years in the population studied (this study), and female Great Reed Warblers rarely reared their own chicks after the successful fledging of a Common Cuckoo chick from their first nest in the same season (Trnka and Samaš, 2024).

In addition, lower annual productivity may also be caused by partial or complete predation of host eggs by the Common Cuckoo. According to "mafia" scenario (Zahavi, 1979), these cuckoo predators may probably be females which did not parasitize particular nests (Jelínek et al., 2016). Moreover, Common Cuckoo females usually remove one or two host eggs at laying which they subsequently eat (Davies, 2000) and some or all of host eggs may even be thrown or destroyed by the hosts during the ejection of a cuckoo egg. Such cases of cuckoo predation and host egg losses also contribute to reduced annual host productivity caused directly or indirectly by cuckoo parasitism, thereby strengthening obtained findings. However, without direct observation it is difficult to determine the true reason of the loss of host eggs. In the study population, Great Reed Warblers rejected the parasitic egg in 53.9%–65.7%, most frequently by ejection (42.9%), and less by desertion (22.8%) (Trnka et al., 2012, 2023), and the rate of nest predation on the Great Reed Warbler nests by predators other than the Common Cuckoo averaged about 20% (Trnka and Prokop, 2010). Thus, nest parasitism and nest predation are interconnected and may skew the true impact of nest parasitism on the annual productivity of a host population. Therefore, additional studies focusing specifically on the impact of predation by the Common Cuckoo on the annual productivity of a host population will be very valuable.

However, the fact that brood parasites can limit or regulate population size in hosts has been clearly demonstrated for hosts of the cowbirds (Molothrus spp.) in North America. There have even been implemented cowbird removal programs as an effective management strategy to restore host populations (Clotfelter and Yasukawa, 1999; Kostecke et al., 2005; Ladin et al., 2016).

Nevertheless, despite a relatively high rate of successful parasitism and its negative impact on the annual productivity of the local Great Reed Warbler population, the studied population has been successfully surviving for more than three decades (own data). One reason for this may be the immigration of individuals from another neighbouring populations (see also Barabás et al., 2004). This is partly supported by the relatively high proportion of new unringed individuals in the study population, as evident from the low annual return rate of breeding females, which ranged from only 23%–29.3% (Trnka and Samaš, 2024), and by several recoveries of Great Reed Warbler adults from adjacent breeding sites (unpublished data). These findings suggest that local population dynamics of hosts may be influenced by spatial variation in patch quality where brood parasites may create differences in reproductive success among host populations. According to this model of source-sink dynamics individuals from populations of high-quality habitat, i.e., the source, maintain populations of low-quality habitat, the sink (Yang et al., 2014).

Another possible explanation is that annual productivity of fledglings in the Great Reed Warbler is a subject of density-dependent selection for lower reproductive output at high abundance (Travis et al., 2023), which is also indicated by the results obtained. Such a negative relationship between the increased number of breeding females and annual productivity has been found in several other bird species (Sæther et al., 2021; Monnier-Corbel et al., 2022).

In summary, brood parasitism by the Common Cuckoo significantly reduced the annual productivity of the local Great Reed Warbler population, which was likely maintained by immigrants from other populations and density-dependent selection. Thus, the level of population isolation and the interactions of density-dependent factors such as nest parasitism and nest predation can affect the demography of host populations of the Common Cuckoo. However, further long-term studies assessing the consequences of Common Cuckoo parasitism on host population dynamics are needed to clarify this issue.

Ethics statement

The study was carried out in accordance with Slovak law. Licences and permits for handling and ringing of birds were issued by the Ministry of Environment of the Slovak Republic.

CRediT authorship contribution statement

Alfréd Trnka: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The author declare that he has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

I thank to all the people who helped me in the field. Peter Samaš did the statistical analyses and read an earlier version of the manuscript. I am also grateful to two anonymous reviewers for valuable comments on the manuscript.