Alternative mating strategies in Eurasian Reed Warblers: extra-pair paternity and egg dumping

Abstract Alternative mating strategies, involving mating outside a social pair bond, have been studied intensively during the last few decades. However, we are still far from understanding what factors affect their frequencies in wild bird populations. In 2006–2011 we studied Eurasian Reed Warblers breeding in the nature reserve “Stawy Milickie” in SW Poland, and analysed the frequency of extra-pair paternity (EPP) and intra-specific brood parasitism (IBP). We genotyped a total of 786 nestlings from 259 nests. EPP offspring constituted 13.40% of all young (from 7.63% to 22.82% in different study years), and they were found in 21.62% of the broods (from 14.6% to 32.7% annually). The mean proportion of females with EPP young was 24.2% (from 16.2 to 41.7% annually). The proportion of EPP young in nests containing any EPP offspring ranged from 25% to 100% (on average 60%) and was not related to brood size. EPP young in one nest were sired usually by one or two males (92% of all cases), but we recorded one brood with three EPP fathers and one brood of four, where each young was sired by a different EPP male, and none by a social male. We found only one case of intra-specific brood parasitism (IBP). The brood with IBP constituted 0.39% of all examined broods, while the nestling that hatched from the dumped egg constituted 0.13% of all young whose parents were genotyped. We conclude that extra-pair paternity is a relatively common and regularly occurring alternative mating strategy in our study population of Reed Warblers, while intra-specific brood parasitism is extremely rare.


Introduction
More than a half century ago, Lack (1968) assessed that the overwhelming majority of birds were socially monogamous.However, only much later it was realized that social mating systems did not always coincide with genetic mating systems (Birkhead & Møller 1992).Since the late 1980s, when DNA fingerprinting was first applied, hundreds of studies have provided evidence for mating outside the pair bond.The most common alternative mating strategy is extra-pair paternity (EPP) resulting from extra-pair copulations So far extrapair paternity has been found in more than 300 bird species, and documented in more than 500 studies (Brouwer & Griffith 2019).EPP rates vary among species/populations from 0% to 81.4% (Brouwer & Griffith 2019), and some studies reported considerable inter-annual variation in EPP levels (Wilk et al. 2008;Brouwer et al. 2017 but see Arct et al. 2013).It has been hypothesized that variation in EPP across socially monogamous species can be explained by latitude, population density, migration, generation length, genetic structuring (dispersal distance), or climatic variability (Griffith et al. 2002).However, in a recent review, after accounting for phylogeny, Brouwer and Griffith (2019) did not find support for any of them, in contrast to some earlier findings.The authors suggest that, due to a large heterogeneity within species in both EPP levels and other life-history traits, using species averages might be unreliable.Therefore, more data from different populations of one species are needed to better understand factors affecting variation in extra-pair paternity in avian populations.
Another alternative mating strategy is eggdumping or intra-specific brood parasitism (IBP).It occurs when a female lays her egg into a nest of another female of the same species.Egg-dumping has been investigated using both genetic and nongenetic methods, and found in more than 250 bird species so far (Yom-Tov & Geffen 2017).However, since this last review, new reports from other species have been published (Haraszthy 2019;Wuczyński et al. 2021).
In this paper we report extra-pair paternity and intra-specific brood parasitism levels found across six study years in a central European population of Eurasian Reed Warbler Acrocephalus scirpaceus (hereafter Reed Warbler).Although the species is a common inhabitant of reedbeds across the Palaearctic, and a subject of many studies, data on EPP frequencies have been reported so far only from two populations (Davies et al. 2003;Hoi et al. 2013).To date there have been no reports from studies investigating egg-dumping in the species.

Study species
The Eurasian Reed Warbler is a small passerine (ca.12 g) breeding in Palearctic reedbeds, and wintering in sub-Saharan Africa (Cramp 1992).In Europe Reed Warblers breed from May to August and can reach very high breeding densities (up to more than 200 pairs per 10 ha) (Cramp 1992;Halupka et al. 2014a;Wierucka et al. 2016).In spring males settle in territories and commence singing.After pairing, males cease singing and begin mate-guarding their partners as they build the nest.Cup-shaped nests of the species are attached to reed stems and usually contain 3-5 eggs.Many clutches (on average 55.1% in Europe; Schulze-Hagen et al. 1996) are lost, mostly due to predation.After nest failure females usually re-nest, building a new nest and laying a replacement clutch.Most pairs have two clutches annually, but some females may lay up to five clutches per season (Schulze-Hagen 1991).The frequency of "true" second broods (started after successful fledging of the young from the first nest) varies widely among years and populations (Cramp 1992;Halupka et al. 2021).
The Reed Warbler is classified as a highly socially monogamous species (Schulze-Hagen 1991;Leisler & Catchpole 1992), though some cases of polygyny have been reported from populations in Western Europe (Taillandier 1990;Duckworth 1992;Davies et al. 2003).Males and females usually mate for the whole breeding season, but withinseasonal divorces sometimes occur (Davies et al. 2003).Both parents are engaged in incubation and feeding of the young (Duckworth 1990;Klimczuk et al. 2015;Wojczulanis-Jakubas et al. 2023).

Study population and data collection
The study was carried out in 2006-2011 in the nature reserve Stawy Milickie (Milicz fishponds) in SW Poland (51.5386 N,17.3398 E).The study plot of ca. 3 ha was a part of extensive reedbed up to 150 m wide with a system of bays and canals inside.The dominant plant species of the reedbed was common reed (Phragmites australis), but patches of bittersweets (Solanum dulmacara), sweet flag (Acorus calamus), sedges (Carex spp) and other herbaceous vegetation occurred in some places.Reedbed margins were fringed by cattails (Typha angustifolia).
Each year we monitored the study plot from the time of arrival of the first individuals, and mistnetted and individually marked birds (combinations of three color and one metal ring).In total we individually marked 678 adult Reed Warblers, including 373 males and 305 females.We searched for nests primarily by observing nest-building behaviour, which starts soon after pair formation.Due to frequent mapping of singing males (every 1-2 days), we were able to identify males that stopped singing after attracting of a female.Then we tried to spot a patch with a nest from 6-m wooden towers or a 3-m high portable aluminium ladders.The majority of nests (more than 85%) were discovered at the building stage by observing nest-building and mate-guarding behaviour (total number of nests found: n = 536, on average 89 per season).We identified parental birds at each nest by video-recording (JVC Everio GZ330) during egg-laying, incubation or/and nestling stage.A total of 610 Reed Warblers were identified at 489 nests, including 418 nests at which both parents were colour-marked.
Nests were visited usually every second day (daily during egg-laying and before the expected hatching and fledging dates).After detecting a nest failure we tried to spot the pair and find its replacement Alternative mating strategies in Eurasian Reed Warblers clutch.Due to nest losses many pairs re-nested laying 1-4 clutches in a season (median = 2) (for details see: Halupka et al. 2008Halupka et al. , 2014aHalupka et al. , 2021)).In our study population of Reed Warblers on average 55% of nests fail to produce young (from 31% to 65% of nests annually), and predation is responsible for 77% of all nest failures (Halupka et al. 2014a).So far we have confirmed two species of nest predators: Western Marsh Harrier Circus aeruginosus and Water Rail Rallus aquaticus, However the list of potential predators is much longer, and includes the Hooded Crow Corvus corone, Magpie Pica pica, Bittern Botaurus stellaris, Little Bittern Ixobrychus minutus, Coot Fulica atra, Moorhen Gallinula chloropus, several species of mustelids (Martes sp., Mustela sp), Muskrat Ondatra zibethicus, Water Snake Natrix natrix and rodents (Arvicola terrestris, Micromys minutus) (Halupka et al. 2014a;Hałupka & KLimczuk-Bereziuk 2023).We never recorded polygyny in our study population, though we observed one polygyny attempt (Halupka et al. 2012(Halupka et al. , 2014b)).More details of the study area and research methods are presented elsewhere (Halupka et al. 2021;Płóciennik et al. 2023;Wojczulanis-Jakubas et al. 2023).

Blood sampling
After mist-netting each adult was sexed based on the examination of cloaca protuberance and/or brood patch (Svensson 1992), and a blood sample (10-30 µl) was taken from 666 individuals.A total of 786 nestling were blood-sampled when they were 6-8 days old.Because we did not manage to mistnet all breeding birds, and not all offspring survived until the age suitable for blood-sampling, the final data set for parentage analysis was based on 259 nests containing 786 nestlings (Table I).

Parentage analysis
For the parentage analysis, we used six fluorescently labelled microsatellites (Ase18, Ase25, Ase37, Ase48, Ase 58, Ppi2), following methods used by Davies et al. (2003).The PCR products were fragment-analysed on a DNA sequencer ABI PRISM® 3100-AvantTM Genetic Analyser (Applied Biosystems) and the resulting chromatograms were analysed using ABI Genescan software.Exclusion probabilities were calculated with Cervus 3.0, and paternity analysis was conducted using the GeneMapper software (version 4.0).

Results
Extra-pair young were found in 56 of 259 nests (21.62%), though annual values ranged considerably from 14.63% to 32.65% (Table I).Despite this, differences among years were not significant (Fisher-Freeman-Halton exact, p = 0.371).The mean proportion of EPP young was 13.40%, but varied from 7.63% to 22.82% among years (Table I).These differences were statistically significant (Fisher-Freeman-Halton exact, p = 0.009).All 56 nests containing EPP young belonged to 51 females: two females had EPP young in two broods in one season, another female in two different seasons, and the fourth female had 3 nests with EPP young in two seasons.Because an average female has two clutches annually, we also calculated the proportion of females with any EPP young in a breeding season.The proportion of such females varied from 16.22% to 41.67% annually (mean = 24.17%),and the differences among years were not significant (Fisher-Freeman-Halton exact, p = 0.121)(Table I).
To asses the proportion of EPP young in nests containing any extra-pair offspring, we conducted two analyses based on two different sample sizes.In the first analysis we included all nests containing EPP young (56 nests with 171 young).For the second analysis we selected only broods whose size was the same as original clutch size.Due to various reasons (partial losses of eggs and nestlings, hatching failures) the sample of such broods was much lower (25 nests containing 94 young).Proportion of EPP offspring in all nests with EPP young was 61.98%, while in nests  II.
The proportion of EPP young in nests with nonreduced broods tended to be correlated with clutch size but the trend was not statistically significant (rs = 0.367, p = 0.071, n = 25).
In 60% of non-reduced broods only one extra-pair male sired all EPP offspring, and in 32% of broods we identified two extra-pair males.Furthermore, there were two broods of four young sired by three and four extra-pair males, respectively.In the latter case each nestling was sired by a different extra-pair male, and none by a social male.There was no correlation between brood size and the number of EPP fathers that sired young in the nest (rs = 0.078, p = 0.709, n = 25).
To find out if females may get direct benefits in the form of nest defence from extra-pair males, we calculated nest failures caused by predation for genotyped broods with and without extra-pair young.We excluded 8 nests, where nestlings died of other reasons than predation.The proportion of depredated broods was similar for nests with and without EPP young (12.72% and 17.35%, respectively) and did not differ significantly (Fisher's exact test, p = 0.537, n 1 = 55, n 2 = 196).
In 2007 we recorded one case of egg-dumping.In a brood of three young, one young was sired by an extra-pair male, but also the mother was different than the social mother.The genetic father of the nestling was a male nesting nearby.The brood with intra-specific brood parasitism constituted 0.39% of all examined broods, while the nestling that hatched from the dumped egg constituted 0.13% of all young whose parents were genotyped.

Discussion
Extra-paternity levels found in this study are slightly lower than mean values reported for 255 socially monogamous birds species with biparental care from 386 populations by Brouwer and Griffith (2019).According to these authors on average 19% of offspring are sired by an extra-pair male (in our study: 13.4%), and EPP is found on average in 33% of broods (21% in our study).However, our data still represent values close to the multi-species mean, as in ca.30% of species EPP is rare (<5% of broods contains EPP offspring), while in 13% of species EPP is common (more than 50% of broods contain EPP young) (Brouwer & Griffith 2019).
Extra-pair paternity has been studied previously in two other European populations of Reed Warblers.Davies et al. (2003), who analysed the data from a British population breeding in the nature reserve Wicken Fen, found EPP offspring only in 15% of broods, and they constituted 6% of all young.In Slovakia, EPP offspring were much more numerous (37% of all) and they were present in 57% of broods (Hoi et al. 2013).Thus, frequencies found in our population are between these reported from the UK and Slovakia.
While benefits of extra-pair paternity are obvious for males, they are much less clear for females (reviewed by Brouwer & Griffith 2019).Females have been hypothesised to benefit directly through extra-pair matings by obtaining extra food from extra-pair males, parental care or protection against predators (Gray 1997;Townsend et al. 2010;Santema & Kempenaers 2021;Krams et al. 2022).Indirect benefits may be associated with insurance against infertility (Sheldon 1994), seeking for good or compatible genes, or genetic diversity of their offspring (Westneat et al. 1990;Birkhead & Møller 1992;Kempenaers et al. 1999).With regard to direct benefits we cannot find any support for direct benefits of EPP for females in our population of Reed Warblers.Courtship and incubation feeding have never been found in the species.Despite more than 900 hours of recordings of parental provisioning, we have never observed a non-social male feeding the young at any nest in our study population Alternative mating strategies in Eurasian Reed Warblers (Wojczulanis-Jakubas et al. 2023, Halupka et al. unpubl. data).Though in some populations of other species females gain direct benefits from nest defence behaviour of extra-pair males (Mennerat et al. 2018;Santema et al. 2020;Krams et al. 2022), in our population nest failure rates caused by predation did not differ between nests with and without EPP young.However, females may get indirect genetic benefits, allowing them to enhance their offspring's genetic makeup.A number of studies published in recent years found that the presence of extra-pair young is correlated with genetic similarity between social mates (reviewed by Arct et al. 2015).In this way extra-pair matings may serve to increase offspring viability.Specifically, offspring sired by extra-pair males may have higher heterozygosity and greater survival expectancy (Foerster et al. 2003;Stapleton et al. 2007).In particular, extrapair matings may help to maintain high polymorphism of MHC (major histocompatibility complex) genes, which are crucial for effective immune response (Johnsen et al. 2000;Garvin et al. 2006).We also found that females with EPP young in their broods had a higher MHC similarity with their social mates compared to females without EPP young, suggesting that extra-pair matings could help them to increase genetic diversity of their offspring (Hałupka et al. 2023 preprint).
The proportion of nests containing EPP young does not seem to be associated with breeding density.In two years with the highest breeding density (2010 and 2011; breeding density of 206.7 and 190 pairs/10 ha, respectively) EPP levels were low.
Although we genotyped 786 nestlings from 259 nests during six study years, we found only one case of intraspecific brood parasitism.Such a result is not surprising, considering that egg-dumping has been found mostly in precocial species with large clutch sizes, and often associated with coloniality or cavity nesting (Yom-Tov 2001;Yom-Tov & Geffen 2017).Intraspecific brood parasitism has been described in only 1% of passerine species, and its levels were usually low (Yom-Tov 2001).In many investigated passerines, egg-dumping has not been found at all (Arnold & Owens 2002).
To conclude, it seems that extra-pair paternity is a relatively common and regularly occurring alternative mating strategy in our study population of Reed Warblers, though its frequencies fluctuate considerably from year to year.In contrast, intraspecific brood parasitism is an extremely rare alternative mating strategy.It remains unclear why EPP levels in different populations of the species vary so much.

Table I .
Frequency of nests containing EPP young, frequency of EPP young and females with EPP offspring in different study years.

Table II .
Shares of broods containing certain proportions of EPP young, based on a sample of all broods with EPP young, and non-reduced broods (brood size = clutch size).