Journal of Ornithology

, Volume 152, Issue 4, pp 983–989

Mating system, paternity and sex allocation in Eurasian Wrynecks (Jynx torquilla)


    • Institute of Pharmacy and Molecular BiotechnologyHeidelberg University
  • Detlef Becker
    • Museum Heineanum
  • Dirk Tolkmitt
  • Verena Knigge
    • Institute of Pharmacy and Molecular BiotechnologyHeidelberg University
  • Hedi Sauer-Gürth
    • Institute of Pharmacy and Molecular BiotechnologyHeidelberg University
  • Heidi Staudter
    • Institute of Pharmacy and Molecular BiotechnologyHeidelberg University
Original Article

DOI: 10.1007/s10336-011-0684-3

Cite this article as:
Wink, M., Becker, D., Tolkmitt, D. et al. J Ornithol (2011) 152: 983. doi:10.1007/s10336-011-0684-3


This paper provides for the first time an insight into the breeding system of the Eurasian Wryneck (Jynx torquilla). DNA fingerprinting and molecular sexing were used to investigate paternity, mating system and sex allocation in a population near the city of Halberstadt, Germany. Similar to other woodpeckers, social and genetic monogamy is the norm in this species. The abundance of extra-pair paternity is low, and only 0.68% of all young [n = 292] were extra-pair young. In addition, wrynecks are facultatively polygynous, but apparently only sequentially polygynous: among 50 broods, 3 cases of social and genetic bigyny were discovered, where males started a secondary brood with a different partner while nestlings of the primary brood had still not fledged. Bigynous males raised 13 offspring per season, as compared to 6.14 young for monogamous males with a single brood. In first broods, the number of sons was almost the same as the number of daughters (54.9% sons), and in second and replacement broods the proportion of males again did not differ significantly from parity (56.5%). Overall, sex allocation was not significantly correlated with the timing of broods (time of hatching) or the mating status of the female.


Mating systemSequential polygynySex allocationDNA fingerprintingMolecular sexing


Das Paarungssystem des Wendehalses (Jynx torquilla) wird in der vorliegenden Studie erstmals anhand genetischer Methoden untersucht. Von 50 Bruten der Jahre 2003 bis 2005 eines Gebietes bei Halberstadt (Sachsen-Anhalt) stand genetisches Material der Alt- und Jungvögel zur Verfügung. Da die näher untersuchten Bruten ganz ähnliche Gelegegrößen aufwiesen wie die Gesamtpopulation des Gebietes über den Zeitraum 1999 bis 2007, wird davon ausgegangen, dass die gefundenen Ergebnisse zum Paarungssystem und zur Geschlechterverteilung bei den Jungvögeln nicht durch Besonderheiten einzelner Jahre geprägt waren. Monogamie ist im Untersuchungsgebiet das übliche Paarungsmuster des Wendehalses. Wie auch bei anderen paläarktischen Vertretern der Familie Picidae stellen Fälle von Polygamie und von Nachkommen außerhalb des Paarbundes eher die Ausnahme dar. Unter 292 untersuchten Jungvögeln fanden sich lediglich zwei Nachkommen außerhalb des Paarbundes (0,68%). Bei 12% der Bruten wurde mit den genetischen Methoden das Auftreten von Bigynie nachgewiesen; Kontrollfänge der Männchen an den Nistkästen beider Bruten belegen, dass diese auch sozialer Natur war. Da in allen Fällen zwischen dem Brutbeginn des Erst- und des Zweitweibchens Zeiträume von mindestens zwei Wochen lagen, ist von einem Muster sequentieller Bigynie zu sprechen. Möglicherweise werden die Männchen durch den bei der Brut des Erstweibchens zu leistenden Aufwand zeitlich in ihren Möglichkeiten zum Start einer weiteren Brut limitiert. Bigyne Männchen zogen 13 Nachkommen pro Saison auf, während monogame Männchen mit einer Jahresbrut lediglich auf 6 kamen. Das Geschlechterverhältnis der Jungvögel zum Zeitpunkt der Beringung war nahezu ausgeglichen. Weder ein saisonaler Trend noch Unterschiede zwischen den Jahren oder Brutstatus (Erst-, Zweit-, Ersatzbrut) ließen sich feststellen. Der geringe Anteil an männlichen Nachkommen in den Erstbruten bigyner Männchen ließ sich aufgrund der geringen Stichprobengröße statistisch nicht absichern.


The breeding system of woodpeckers in general has been classified as socially and genetically monogamous (Winkler et al. 1995). Whereas many woodpeckers, especially in temperate regions, are solitary species, others have a more social and communal lifestyle. Communal breeding with helper systems is common in woodpeckers of the subtropical and tropical regions (del Hoyo et al. 2002; Winkler et al. 1995). Unlike most other birds, parental care is not shared equally between sexes in woodpeckers; males carry out most of the care, which includes incubation of the nestlings at night. Males help in constructing nest holes, show nest-guarding, and exhibit intensive brood care (Short 1982; Cramp 1985; del Hoyo et al. 2002; Wiebe 2002; Wiebe and Kempenaers 2009). It has been argued that since woodpecker males brood at night, this would make facultative socially polygynous broods (one male breeding with two females) more difficult, only allowing the occurrence of polyandry (where one female breeds with two males in separate nests) (Winkler et al. 1995; Ligon 1999; Wiebe 2002; Andersson 2005). Evidence for facultative social and genetic polyandry has been obtained for the Great Spotted Woodpecker (Dendrocopos major), Middle Spotted woodpecker (D. medius) and Three-toed Woodpecker (Picoides tridactylus) (Kotaka 1998; Pechacek et al. 2005, 2006; Rossmanith et al. 2009; Michalek and Winkler 2001; Michalek and Miettinen 2003). In the Lesser Spotted Woodpecker (D. minor), about 10% of the females are genetically polyandrous and 3% of the males genetically polygynous (Wiktander et al. 2000).

Wrynecks represent a monotypic subfamily (Jynginae) in the woodpecker family (Picidae). Two species are recognised in the genus Jynx. The Eurasian Wryneck (J. torquilla) breeds in Europe and Asia and winters in Africa south of the Sahara and in southern Asia. In Africa, the Rufous-necked Wryneck (J. ruficollis) is a resident or nomadic species showing a patchy breeding distribution in Central, East and Southern Africa (del Hoyo et al. 2002).

Data for the breeding or mating system of wrynecks are not available as far as we know. A few observations (capturing the same ringed adult wryneck in two neighbouring nests at the same time) imply the occurrence of some form of social and/or genetic polygamy (Linkola 1978). In another case, a brood was raised by a single adult, which would be compatible with genetic polygamy but also with the loss of a partner (Kervyn and Xhardez 2006).

Wrynecks return to European breeding sites in April/May. The first brood usually hatches at the end of May/early June. Sometimes a second clutch is laid at the end of June which hatches in July. Wrynecks have a relatively large clutch of 7–12 eggs, which is incubated by both sexes. Average incubation time is 11–12 days. Young are fed by both parents and fledge after 20–22 days. The breeding success is usually quite high (about 70%), so that 3–4 young/brood are produced per pair (del Hoyo et al. 2002).

A new wryneck population was established on a former military training ground near Halberstadt (Germany), where nest boxes had been provided since 1999. We took the opportunity to utilise a growing wryneck population with more than 40 pairs to study the breeding biology of wrynecks, as many aspects of wrynecks have not been investigated so far. In this communication, we report on the genetic mating system for three breeding periods and explore the potential occurrence of extra-pair paternity (EPP) and polygamy, which should be facilitated in such a dense population. In addition, we study sex allocation in young wrynecks in relation to the breeding system of their parents. In this work, molecular sexing was applied in wrynecks for the first time.

Materials and methods

Study area

The study area (11°00 E; 51°54 N) is situated about 5 km north of Halberstadt (state of Sachsen-Anhalt, Germany), and covers approximately 450 ha. The area had been used as a military training ground until 1990. Presently, the area is grazed by sheep and goats to keep the vegetation low. The vegetation consists of arid and semiarid grassland and managed grassland with fruit trees, with the latter covering approximately 5 ha.

Natural breeding sites for hole-nesting birds are nearly completely absent, because trees are either lacking or they are too young and do not have holes. From 1999 onwards up to 92 artificial nest boxes (78 in 2003, 92 in 2004, and 88 in 2005) were placed in the area about 100 m apart (Becker and Tolkmitt 2007, 2008). Wrynecks used the nest boxes from the start and are the dominant hole-nesting species in the study area. In some parts, 100% of the nest boxes are occupied by wrynecks. The overall breeding density is about 1 pair/10 ha. In some areas, density can be as high as 4 pairs/10 ha. The complete wryneck population has been studied thoroughly since 1999, and the number of breeders, the clutch size, the breeding success and the numbers of second and replacement broods have been recorded and documented in locally distributed journals (Becker and Tolkmitt 2007, 2008; Tolkmitt et al. 2009).

Blood samples

Blood samples were collected in the parts of the study site that had the highest densities of breeding pairs, and where adults had been observed visiting more than one nest box during the egg-laying period. We believed that cases of deviation from monogamy would most likely occur here. Adult breeding wrynecks were caught in the nest by hand. Nestlings were taken out of the nest boxes and ringed between 7 and 14 days of age. A leg vein was punctured with a sterile needle (ducd), and about 100 μl of blood were stored in a modified EDTA buffer (Arctander 1988; Wink 2000). Blood samples were stored at 4°C until they were transported to Heidelberg. They were then stored there at −20°C until DNA isolation.

Total DNA was isolated from 100 μl of blood using a standard proteinase K (Merck, Darmstadt) treatment and phenol/chloroform DNA extraction (Sambrook et al. 1989).

Assessment of parentage by DNA fingerprinting

A multilocus DNA fingerprint analysis was carried out to investigate the genetic relationships among and between the sampled wryneck families. For each sample, digestion of total genomic DNA by restriction enzymes (Hae III), agarose gel electrophoresis, and capillary transfer to a nylon membrane (Biodyne B) followed standard protocols established in our laboratory (Swatschek et al. 1994). Nylon membranes were pre-hybridized in a hybridization mixture (5 × SSPE, 0.1% SDS, 1% powdered milk, 5 × Denhardt’s solution; Sambrook et al. 1989) for 2 h at 39°C. Hybridization was carried out with a hybridization mixture containing 10 pmol/ml of the digoxigenated oligonucleotide probe (GGAT)4 (Fresenius) at 39°C overnight. Membranes were washed three times with 6 × SSC for 30 min each. DNA/DNA hybrids were detected by an antibody which was raised against digoxigenin (Boehringer). This antibody was coupled to a phosphatase, which in turn produced a coloured precipitate at the sites of hybridization. After colour reactions were completed, nylon membranes were documented and analysed.

The paternity assignment followed some fundamental rules outlined by Westneat (1990), as we attempted to assign all bands of the nestlings to one of the putative parents. The informative (i.e. polymorphic) bands of the DNA fingerprint (usually between 10 and 20) were visually scored into a data matrix as either absent (“0”) or present (“1”), which was then used to calculate the band-sharing coefficients (BSC) as BSC = C × 2/(A + B) × 100, where C is the number of shared bands, and A and B are the number of bands in the profiles of individuals A and B, respectively (Wetton et al. 1987). A visual inspection made sure that all bands in a fingerprint could be allocated to the parents or the offspring. The frequency distribution of the BSC values of sibling–sibling and parent–offspring comparisons is shown in Fig. 1. As expected, BSC values between siblings are between 60 and 80%, while those for the offspring–parent comparison are a little bit lower (between 50 and 80%). BSC values between the nestlings and their social father were below 40% and nestlings had three or more novel bands in only 2 of 292 cases; in these cases, the BSC values between the nestlings and their social mother were 50–60% for all nestlings. Hence, genetic paternity of the social male was excluded in these two cases. Parents were unrelated and had BSC values of below 30%. Young with one or two nonmatching bands were seen only twice (out of 292 young); however, in these cases, all other bands exactly matched those of the respective father and mother, so EPP can be excluded.
Fig. 1

Frequency distribution of BSC value comparisons between siblings and between parents and offspring. The low BSC values of 3 and 23% derive from comparisons of two EPP young with their social fathers (not their social mothers)

Molecular sexing

Sex identification was conducted following Kahn et al. (1998). PCR was performed with 30–60 ng of template DNA in a 25 μl reaction volume containing 8 pmol of the primer H1272 and 9 pmol of the primer L1237, 0.1 mM of dGTP, dCTP, and dTTP, 0.045 mM dATP, 1 μCi [α-33P]-dATP (Amersham Biosciences), 0.6 units of Taq-Polymerase (Pharmacia Biotech, Freiburg, Germany) and 2.5 μl of 10× amplification buffer (10 mM Tris–Hcl (pH 8.5), 50 mM KCl and 1.5 mM MgCl2). Each reaction was overlaid with two drops of mineral oil. Thermocycling was performed with a Trio Thermo block TB1 (Biometra, Göttingen, Germany). Following the initial 5 min denaturation at 94°C, the program consisted of 31 cycles of 30 s at 94°C, 40 s at 56°C, 40 s at 72°C, and 5 min at 72°C for final elongation. DNA fragments were separated by vertical PAGE (polyacrylamide gel electrophoresis) for 2 h at 65 W using a Base Acer Sequencer (Stratagene). After drying, X-ray films (BioMax MR Film, Kodak) were exposed for 24 h to the denaturing gels. The bands were analysed visually. The presence of two bands was scored as female and one band as male (Kahn et al. 1998).


Differences between means of brood parameters were evaluated using a two-tailed t test. Differences between the sex ratios of young from first, second and replacement clutches, a potential correlation of sex ratio with laying date, and differences between years were analysed using a multi-level logistic regression model with a random intercept (Wilson and Hardy 2002; Demidenko 2004).

Results and discussion

Breeding success

A subsample of about 15–30 broods per year that came from densely populated parts of the study area were monitored for this study in detail. Statistics for brood parameters are based on 80 broods: 55 first and 25 second and replacement broods. Their clutch sizes and breeding successes are summarised in Table 1. The clutch sizes of the first and second/replacement clutches are very similar to those previously published for the complete population: in the years 1999–2007, 212 first, 68 second and 26 replacement clutches were recorded in the population as a whole, with a clutch size of 9.87 ± 1.50 eggs in first clutches and 7.74 ± 1.73 eggs in second clutches; the difference was highly significant (t = 9.13; df = 101.27; P < 0.001) (Becker and Tolkmitt 2007, 2008; Tolkmitt et al. 2009). Therefore, we believe that the results for sex allocation and paternity published here for the first time are representative of the population and are not biased due to natural conditions or other parameters in single years. The clutch sizes of first broods were significantly higher (t test; P < 0.0001) than those in second or replacement clutches: 10.16 versus 8.32 eggs on average. For the 80 broods in the years 2003–2005, the breeding success, as measured by the mean number of fledged young per brood (5.93), was about 58.4% for first broods (compared to clutch size), and was therefore higher than that for second and replacement clutches (breeding success 52%). The difference between the first and later clutches was not significant (P = 0.06). In general, the breeding success of the Wryneck at Halberstadt was higher than the value given in del Hoyo et al. (2002) (3–4 young per brood).
Table 1

Breeding successes and sex allocations in first, second and replacement clutches


Number of brood

Mean ± SD

Clutch size

 First brood


10.16 ± 1.74

 Second + replacement


8.32 ± 2.09

Fledged young

 First brood


5.93 ± 3.69

 Second + replacement


4.33 ± 2.63

Sex allocation (% sons)

 All broods


55.42 ± 18.43

 First brood


54.94 ± 18.41

 Second + replacement


56.52 ± 18.46

Sex allocation

Molecular sexing was carried out successfully, and the results agreed 100% with the sexes of the adult birds that had been deduced from behavioural observations. It was possible to obtain DNA samples that could be analysed from 51 broods. Overall, the ratio of sons to daughters was 55.4% (Table 1). Second broods produced almost 56.5% sons and 45.5% daughters (Table 1) (deviation from parity was n.s.; P = 0.31). Between years the sex ratio did not differ (P = 0.46; logistic regression model). Overall, sex allocation was apparently not influenced by the timing of broods: time of hatching and sex ratios were not correlated according to a multi-level logistic regression model (P = 0.42; n.s.).

Three social and bigynous males (see below) that sired two broods in a season but with different females were detected. Whereas there were 34.3 ± 5.1% sons in the primary brood, the ratio increased to 51.3 ± 8.1% in the secondary brood. The sample size of three instances of bigyny is too small to allow generalisations or speculations about the sex ratio. This question must be addressed in future studies using larger samples. However, if bygynous males have to share their effort between primary and secondary broods or achieve no parental care for one of the broods, there should be differences in the sex ratio, at least under the assumption of some form of sexual dimorphism in the young. In sexual dimorphic species, it is common for food shortages to result in competition between siblings and sex-specific mortality (Newton 1989; Ristow and Wink 2004; van denBurg and Newton 2003; Kenward 2006; Dyrcz and Chicon 2009).


Enough DNA was obtained to carry out multi-locus DNA fingerprinting in 50 broods. Extra-pair paternity (EPP) was apparently a very rare event, and only 2 young were detected in one brood that were sired by a second male. This could have been an instance of polyandry (known to occur in woodpeckers), although the second male could not be captured. The overall EPP rate is thus about 0.68% with respect to the 292 young that were analysed (Table 2). Low EPP rates are common for non-passerine birds with a low or moderate mortality (Wink and Dyrcz 1999).
Table 2

Monogamy, polygamy and extra-pair paternity (EPP) as determined by DNA fingerprinting


Number of broods

Number of young


Number of broods with polygamy (number of young)





4 (20)





2 (19)









2 (0.68%)

6 (39)

The low EPP rates in wrynecks agree with the findings in other woodpeckers, such as D. major and D. medius (del Hoyo et al. 2002; Wiebe and Kempenaers 2009).

Evidence for polygamy

In most nest boxes, only a single pair could be captured, confirming the high rate of monogamy. In most cases of second or replacement clutches, the same partners as in the first clutch could be confirmed by DNA fingerprinting and ring IDs. But there were three exceptions. Two males were captured in 2003 and one in 2004 that were raising young with two different females each, indicating cases of social and genetic bigyny. In these cases, ring numbers and DNA fingerprint analysis both unequivocally showed that the social mothers of the male’s primary and secondary broods were not identical: they were different individuals. The mean clutch size of these six broods (10.33) does not differ from the mean clutch size of all studied broods in the years 2003–2005.

Male 40 was seen in 2003 at nest boxes H1 and H16, which were 200 m apart. The first eggs were laid on 28th May in H1 and on 17th June in H16, respectively. The nestlings hatched on 17th June and 5th July, respectively. Both broods were replacement clutches, and DNA fingerprinting unequivocally showed that male 40 was the genetic father of the nestlings in H1 and H16, whereas the mothers differed.

Male 8 was captured in 2003 at nest boxes M2 and M4, which were 100 m apart. The first egg was found in M2 on 15th May, and the young fledged on 24th June. In M4, which was a replacement clutch, the first egg was observed on 13th June, and nestlings fledged on 24th July. DNA fingerprinting unequivocally showed that male 8 was the genetic father of the nestlings in M2 and M4, but that the mothers differed.

In 2004, male 7 was captured at nest boxes M5 and M7, which were 200 m apart. The first egg in M7 was found on 6th May, and in M5 on 24th May. Young fledged on 19th June and 3rd July, respectively. DNA fingerprinting unequivocally showed that male 7 was the genetic father of the nestlings in M5 and in M7, whereas the mothers differed.

Two of the three males were monitored at the nest box during the egg or nestling stages of both broods. In the third case, the male was caught only during the egg stage of the secondary brood, which started some 20 days after the primary brood had been initiated. The length of the delay and the success of the primary brood (with six fledglings) are reasons to believe that this male also made some contributions to parental care for the primary brood. In all three cases it is thus likely that the males were not only the genetic but also the social fathers of both broods.

In all of the broods with bigynous males, the males started a secondary brood while nestlings of the respective primary broods were still in the nest or just fledging. We did not discover broods with bigynous males serving broods at the same time. Thus, the data would support the concept of sequential polygyny. Our data do not help to answer the old dispute over the question of whether males or females incubate at night (Bussmann 1941; Steinfatt 1941; Ruge 1971) because the broods with bigynous males were apparently started when the nestlings had already hatched. However, the observed consistent pattern of delayed starts of the secondary broods indicates that the incubation duties of males might have interfered with or limited their prospects of polygyny to some extent.

Altogether, 39 young came out of broods with putatively bigynous males (= 13.3%; with respect to 292 young in the subsample of the broods with genetic material), which accounted for 12% of the broods studied. Putative bigynous males raised 13 offspring per season, as compared to 6.14 young for monogamous males with a single brood. A similar finding was reported for polyandrous females of Colaptes auratus, which raised about 10.8 nestlings in a season, as compared with the 5.5 young of monogamous females (Wiebe and Kempenaers 2009). In contrast, in Great Spotted Woodpeckers, polyandrous females exhibited lower breeding success (Michalek and Winkler 2001, Michalek and Miettinen 2003).

Facultative genetic polygamy has been observed in some woodpeckers; in D. minor, about 10% of the females are polyandrous and 3% of the males are polygynous (Wiktander et al. 2000). In Three-toed Woodpeckers (Picoides tridactylus), 7.4% of the 27 females studied were genetically polyandrous. 7.3% of the 55 chicks from 15.4% of the broods were the result of EPP. Similar to most other woodpeckers, Three-toed woodpeckers are mainly socially and genetically monogamous (Pechacek et al. 2005, 2006). Polyandry and polygyny took place in Lesser Spotted Woodpeckers only when the sex ratio in the population was biased (Wiktander et al. 2000).

In conclusion, DNA analysis supports the field data, which indicate that wrynecks are generally socially and genetically monogamous but that facultative and sequential polygamy also occurs: in 12% of the broods, social and genetic bigyny was the likely breeding system. It is apparent that broods with bigynous males were raised sequentially rather than concomitantly. The secondary brood was usually started in the last third or at the end of the primary brood. This species has a relatively high breeding density in the study plot, which may have influenced the degree of polygamy. In other areas with low densities, where close neighbours are less frequent, social and genetic monogamy would be the most likely breeding system.


Blood sampling was carried out with the permission of the former Regierungspräsidium Magdeburg. Dr. Bernd Nicolai (Halberstadt) helped in the field work and with the collection of blood samples. We thank Prof. Dr. M. Neuhäuser (University of Applied Sciences, Remagen) for help with statistics, and two referees for helpful comments.

Copyright information

© Dt. Ornithologen-Gesellschaft e.V. 2011