Occurrence of extra-pair paternity and intraspecific brood parasitism in the Whiskered Tern Chlidonias hybrida

Abstract

Capsule Patterns of parentage were studied in a colony of Whiskered Terns Chlidonias hybrida in central Poland. We assessed the parentage of 64 chicks from 30 broods using five polymorphic microsatellite loci. The rate of extra-pair paternity was estimated at 8.1% of offspring and 11.8% of broods (the upper 95% CL: 16.9% and 27.1%, respectively), whereas intraspecific brood parasitism was recorded in 4.7% of offspring and 10.0% of broods (the upper 95% cl: 9.9% and 20.7%, respectively). The data obtained on the occurrence of alternative reproductive strategies in Whiskered Tern are consistent with general patterns of parentage found in other colonial larids.

Although most avian species are known to be socially monogamous, molecular parentage studies on birds that have proliferated over the last decades have revealed a high prevalence of alternative reproductive tactics, such as extra-pair paternity and intraspecific brood parasitism. The incidence of both these strategies has been investigated in a number of colonial larid species, but was found to occur at low frequencies in most of the studied taxa, including skuas (e.g. Millar et al. 1994), Larus and Rissa gulls (Gilbert et al. 1998, Helfenstein et al. 2004), and Sterna terns (Griggio et al. 2004). In larids and other seabirds, biparental care to the offspring and synchrony between mates during incubation and chick-rearing are prerequisites to successful breeding. Under these conditions, all seabird species are socially monogamous and individuals are expected to be choosy during mate choice (Lack 1968). Seabirds generally lay small clutches, but their high life expectancy allows individuals to perform many breeding attempts during their life. Consequently, males should not be tolerant of extra-pair copulations of their social mates and extra-pair paternity rates in seabirds should be low (Griffith et al. 2002).

Despite the fact that these predictions have been tested in several seabird taxa, in general, larids remain under-represented in extra-pair paternity studies (Gilbert et al. 1998, Yom-Tov 2001). Accordingly, we are aware of no molecular parentage analyses that have been conducted for the clade of marsh terns from the Chlidonias genus. The aim of this study was to use microsatellite markers to provide an estimation of extra-pair paternity and intraspecific brood parasitism frequency for the Whiskered Tern Chlidonias hybrida, a species of unfavourable Pan-European conservation status (BirdLife International 2004) that is characterized by fluctuating numbers and patchy distribution. So far, the only data on the incidence of alternative reproductive strategies in this species come from the egg morphology analysis, which estimated that intraspecific brood parasitism occurs in ca. 10% of broods (Paillisson et al. 2008).

The study was conducted in a Whiskered Tern colony at the Jeziorsko reservoir (51° 44′ N, 18° 38′ E), central Poland in 2012. All the nests were located on the floating vegetation dominated by Amphibious Bistort Polygonum amphibium. In total, there were 84 clutches recorded in the colony. During the hatching period all the hatchlings were captured, ringed and assigned to their natal nests. Effective localization of the chicks was possible owing to sparse structure of Bistort vegetation in the colony. We could assign correctly all the chicks to their natal nests without application of nest enclosures because: (i) distances between neighbouring nests were relatively long (on average 5.51 ± 0.57 [se] m) and nest density was low (43.5 ± 2.9 [se] nests ha⁻¹); (ii) all the nests were grouped into several clusters separated with deep-water channels that were impassable for hatchlings; (iii) there was considerable between-brood variation in hatching dates (29 June–9 August), resulting in very small numbers of broods being hatched between successive visits in the colony (on average 3.25 ± 0.64 [se] of broods hatched per visit during which at least one hatched brood was recorded); (iv) all chicks were captured and ringed within 2–3 days from hatching, a period when they retain low mobility.

We used a specially-designed nest trap to capture adult birds during incubation. To exclude the possibility of catching birds that were not the social parents of that nest, we only triggered the trap after we had observed the bird incubating for a few minutes. All captured adults were ringed and measured. The following measurements were collected: total head length and bill length (to the nearest 0.1 mm), wing length (to the nearest 1 mm) and body mass (to the nearest 1 g). As Whiskered Terns show sexual size dimorphism, adults were sexed using appropriate discriminant equations developed for Central European populations of the species, which has been demonstrated to yield a classification success rate of 94.6% (Ledwoń 2011).

For the parentage analyses, we randomly selected 30 broods, in which at least one social parent was captured. The mean size of the sampled broods was 2.19 ± 0.13 [se] chicks (range 1–3). In most of the cases we sampled entire broods (80%), but in the remaining broods single eggs were lost during incubation and, thus, could not be sampled for DNA (8.6% of all eggs). In spite of the fact that both parents are known to incubate eggs in the Whiskered Tern, there were only two broods for which both putative parents were captured. In several other broods repeated catching attempts resulted in capturing the same individual multiple times, which was perceived likely to cause clutch desertion and, thus, was abandoned. Consequently, in the remaining broods only a male (n = 15) or a female (n = 13) was captured. All selected broods were evenly distributed across the colony. We collected ca. 50 μl of blood from the brachial vein of each chick (n = 56) and adult (n = 32). If any fertilized egg did not hatch, we collected tissue samples from embryos (n = 8). Samples were stored in ethanol at −4°C until analysis.

Parentage analysis was conducted using five microsatellite loci originally developed for Roseate Tern Sterna dougalli (Sdaat20, Sdaat27, Scaac20; Szczys et al. 2005) and for Ring-billed Gull Larus novaehollandiae scopulinus (RBG18, RBG27; Given et al. 2002), which were successfully cross-amplified and proved polymorphic in the Whiskered Tern (Molecular Ecology Resources Development Consortium et al. 2013). After ethanol removal, DNA was extracted with a Genomic DNA Purification Kit following kit protocol (Fermentas, Thermo Fisher Scientific). PCR amplifications were conducted in a final volume of 25 μl containing 20–50 ng of DNA with a final concentration of 1× for the Ready Mix Taq PCR Reagent Mix (Sigma-Aldrich) and 0.1 μM for each primer. All PCR amplifications followed the steps of: (i) initial denaturation at 95°C for 2 minutes; (ii) 35 cycles of denaturation at 95°C for 30 seconds followed by 30 seconds at annealing temperatures and elongation at 72°C for 30 seconds; (iii) final elongation at 72°C for 5 minutes. Forward primers were end-labelled with FAM fluorescent dye. PCR products were separated using an automatic sequencer ABI/Hitachi 3500 (Applied Biosystems). Allele sizes were scored against GeneScan TM 600 LIZ Standard (Applied Biosystems) using GeneMapper 4.1 software. Deviations from Hardy–Weinberg equilibrium (HWE) were tested with GenePop 4.0.10 (Raymond & Rousset 1995) and departures from linkage equilibrium for each pair of loci were evaluated with f-stat 2.9.3 (Goudet 1995). Frequency of null alleles was estimated with Cervus 2.0 (Marshall et al. 1998).

Table 1. Characterization of five microsatellite loci used for parentage analysis in the Whiskered Tern with data on the number of adult individuals genotyped (n), number of alleles (n_A), size range of alleles, observed heterozygosity (H_o), expected heterozygosity (H_e), frequency of null alleles (NullF), significant departure from HWE (*).

Locus	n	nA	Size range (bp)	Ho	He	NullF
RBG18	32	8	172–200	0.47	0.43	−0.07
RBG27	32	5	188–200	0.53	0.64	0.09
Sdaat20	32	12	164–208	0.88	0.91	0.01
Sdaat27	32	8	235–244	0.91	0.80	−0.07
Scaac20	32	22	113–152	0.72	0.95	0.13*

After Bonferroni correction, the exact test revealed no significant linkage disequilibrium between any pairs of tested loci. Only one locus (Scaac20) deviated from HWE, indicating a heterozygote deficit (Table 1). There was evidence for null alleles occurring in this locus with the frequency of 0.13, which should not, however, affect the parentage analysis greatly (Dakin & Avise 2004). Combined non-exclusion probability for the first parent was 0.041 and for the second parent was 0.007. We repeated PCR amplifications and genotypic analysis to confirm all mismatches between offspring and social parents. Offspring with a minimum of two mismatching loci were considered non-descendant from their social parent. Mismatches at only one locus were considered as resulting from mutations or null alleles. Allelic mismatches between an offspring and its putative mother, or with both putative parents if their genotypes were available, were considered as intraspecific brood parasitism. The occurrence of quasi-parasitism (parasitism by a female who was fertilized by the male attending the parasitized nest) was disregarded, because it has never been reported in larids (Griffith et al. 2004). When allelic mismatches were recorded between an offspring and its putative father, both intraspecific brood parasitism and extra-pair paternity could be involved. We resolved this problem by conducting a sib-ship reconstruction implemented in program Colony v2.0 (Wang 2004). The software uses a group-likelihood approach that considers information from all individuals in order to partition the entire offspring cohort into half-sib families. In the analysis, the error rate of genotyping was set to 0.025 as suggested by Wang (2004). All cases, when an offspring was not sired by its social father but was classified as a half-sib with the putative siblings, were interpreted as extra-pair paternity. By contrast, when an offspring was not sired by its social father and was not classified into half-sib dyad with the putative siblings, a case of intraspecific brood parasitism was inferred.

We recorded four chicks from three broods having genotypes which did not match that of their social fathers (n = 17) at a minimum of two loci. Three non-descendant chicks from two broods were grouped into half-sib dyads with their putative siblings, indicating the occurrence of extra-pair paternity. One remaining non-descendant chick was excluded as a half-sib, which indicated a case of intraspecific brood parasitism. There were also two cases of allelic mismatches at a minimum of two loci between offspring and their social mothers (n = 15) which indicated intraspecific brood parasitism, a conclusion that was confirmed by the half-sib dyad analysis.

Extra-pair paternity was recorded in 8.1% of offspring and 11.8% of broods, whereas intraspecific brood parasitism was recorded in 4.7% of chicks and 10.0% of broods that were molecularly analysed. Given our sample sizes and disregarding any independence in data set, we estimated the maximum level of extra-pair paternity in the studied population of Whiskered Terns at 16.9% of young and 27.1% of brood, which was calculated as the upper 95% confidence limit for the 3/37 chicks and 2/17 broods. The maximum level of intraspecific brood parasitism was estimated with the same method at 9.9% of young and 20.7% of brood (3/64 chicks and 3/30 broods).

Although the frequency of extra-pair paternity has been estimated for a number of colonial larid species, we are aware of no published results for the group of marsh terns from Chlidonias genus. In this study we demonstrated extra-pair paternity for 11.8% of broods and 8.1% of chicks of the Whiskered Tern, one of the four recognized Chlidonias species. These estimations are in accordance with the general patterns observed within the Lari suborder, where the frequency of extra-pair paternity is generally low. Similar estimations have been reported for species such as the South Polar Skua Catharacta maccormicki and Common Gull Larus canus, where extra-pair paternity was recorded, respectively, in 7.7% (Millar et al. 1997) and 8.3% (Bukacińska et al. 1998) of broods. In other larids, the frequency of extra-pair paternity was found to be even lower. No evidence for extra-pair paternity was found in populations of Brown Skua Catharacta antarctica lönnbergi (Millar et al. 1994), Western Gull Larus occidentalis (Gilbert et al. 1998), Black-legged Kittiwake Rissa tridactyla (Helfenstein et al. 2004), and Common Tern Sterna hirundo (Griggio et al. 2004). To date, the Black-headed Gull Chroicocephalus ridibundus is the only larid for which the rate of extra-pair paternity has been demonstrated to be relatively high, with 33% of broods containing at least one extra-pair offspring (Ležalová-Piálková 2011).

Intraspecific brood parasitism has been recorded in several species of gulls and terns (Yom-Tov 2001). The highest rate of intraspecific brood parasitism was found in the Black-headed Gull, where parasitic eggs were observed in up to 34% of nests (Duda et al. 2008). Since the evidence for intraspecific brood parasitism in larids often comes from field observations of abnormally large clutch sizes, the true incidence of this reproductive strategy is difficult to assess. As shown in the Black-headed Gull, the analysis of supernormal clutches strongly underestimates frequency of intraspecific brood parasitism in larid colonies (Ležalová-Piálková 2011). Molecular parentage analyses indicated that intraspecific brood parasitism occurred in 10% of broods in our study population. This estimate is consistent with the results obtained for the French populations of the species where the presence of parasitic eggs was estimated, using the analysis of intra-clutch variation in egg morphology, to occur in 9.2% of clutches (Paillisson et al. 2008). The levels of intraspecific brood parasitism found in the Whiskered Tern might be facilitated by long periods of inattendance at nests by both pair members during the pre-incubation period, as was suggested by Paillisson et al. (2008). Similarly, long male inattendance at the nest resulting in inefficient mate guarding could facilitate occurrence of extra-pair paternity at the reported rate.

ACKNOWLEDGEMENTS

We thank two anonymous reviewers for helpful comments on the earlier draft of the manuscript. The fieldwork was performed by the permissions of the Bioethical Commission and the General Environmental Protection Directorate in Poland.

FUNDING

The study was financially supported by the research grant of the Polish Ministry of Science and Higher Education under the ‘Iuventus Plus’ programme [IP2011 036171].