Sex-related differences in autumn migration timing of adult common sandpipers Actitis hypoleucos (Linnaeus, 1758) (Charadriiformes: Scolopacidae)

Abstract

In the common sandpiper both sexes do not participate equally in brood rearing. The attendance of females progressively declines and by the third week after hatching, most of them leave their broods. To check if this unequal parental care results in different migration phenology of males and females, adult common sandpipers were caught at four ringing sites in Poland during their autumn migration and were sexed using a discriminant equation. The median date of female migration was 6 days earlier than that of males. In the first days of July, males comprised only 22% of common sandpipers caught, while later their percent share increased gradually and in August reached up to 80% of captured birds. However, birds of both sexes were present among migrants during the whole migration period and migration period of both sexes overlapped to a great extent. Most probably birds migrating through the study area came from a vast breeding area, where the period of egg laying differs between its northernmost and southernmost parts by at least 1 month.

Keywords:

• Waders
• migration
• protogyny
• Central Europe

Introduction

Detailed studies on migration of birds are difficult to conduct because of their complexity and different processes which shape migration patterns (Cristol et al. 1999; Jenni & Schaub 2003). Within a species, individuals of a given sex or age and also birds from different populations may reveal different timing or migration strategies (Ydenberg et al. 2005; Meissner 2015; Ożarowska & Zaniewicz 2015; Pinchuk et al. 2016). Different migration timing of males and females may have influence on the onset of other parts of the annual cycle, such as primary moult (Gill et al. 1995; Barshep et al. 2011). Moreover, it may be related to the amount of accumulated energetic reserves, which influences the migration strategy (Pinchuk et al. 2016). Hence, there is growing evidence of sex-based differences in bird migration phenology, also in waders in which sexual dimorphism is usually weakly expressed (e.g. McCloskey & Thompson 2000; Bishop et al. 2004; Meissner 2005, 2015; Ydenberg et al. 2005; Remisiewicz & Wennerberg 2006; Pinchuk et al. 2016).

The common sandpiper Actitis hypoleucos (Linnaeus, 1758) is a wader species widely distributed in Europe and Asia (Cramp & Simmons 1983). Its autumn migration through Europe is well described according to the results of regular counts (e.g. Harengerd et al. 1973; Winkler & Herzig-Straschil 1981; Meissner 1996; Mitrus et al. 1998; Laber 2003; Iwajomo & Hedenström 2011) and analyses of ringing recoveries (Stiefel et al. 1985; Meissner 1997; Cepák et al. 2008; Fransson et al. 2008; Saurola et al. 2013). It is known that adult birds precede juveniles when moving towards the wintering grounds, but migration dates of both age classes overlap to a great extent (Meissner 1996; Balmori 2003). However, there are no studies that examined possible differences in migration timing of adult males and females, because the common sandpiper is a monomorphic species and it is impossible to recognise the sex of the birds by their plumage characteristics (Meissner et al. 2015).

Observations from the breeding grounds suggest that in the early brood-rearing period, both sexes participate equally, and the attendance of males remains around 90% throughout the chick-rearing period. However, the attendance of females declines and by the third week most of them leave their broods (Yalden & Holland 1992). Hence, it might be expected that females migrate towards the wintering grounds earlier than males, but it is unknown how long females stay in the breeding area before they start autumn migration. The only attempt to distinguish migration timing of adult males and females was made in south-eastern Germany, where birds were sexed according to a wing length of 108 mm, which was thought to be a boundary for the wing length distributions of both sexes (Teubert & Kneis 1984). According to the data on birds sexed molecularly, individuals with wing length shorter than 111 mm and longer than 117 mm may be sexed safely as males and females, respectively (Meissner & Krupa 2016). Hence, using only one value of the wing length as a criterion for sexing birds can produce a bias towards one of the sexes. Moreover, a recently developed discriminant function for sexing adult common sandpipers according to linear measurements enables us to recognise the sex of birds caught and measured in previous years (Meissner & Krupa 2016). Thus, this study is aimed to validate the hypothesis of earlier migration of females of the studied species.

Material and methods

Birds were caught in walk-in traps (Busse & Meissner 2015) at four ringing sites in Poland (Figure 1) and aged according to plumage characteristics (Meissner et al. 2015). Only adult birds (older than 1 year) were analysed. Bird trapping started between the beginning of and mid-July and was finished in late September. This period (July–September) covers the majority of the migration of adult common sandpipers through Poland (Meissner 1996). The subsequent division of the study time into 5-day periods (pentades) followed Berthold (1973), and the median test was used to assess differences between median migration dates of males and females.

Figure 1. Location of the ringing sites, where common sandpipers were captured.

Sexing was performed according to the equation given by Meissner and Krupa (2016) based on the wing and tarsus plus toe lengths:

(1) $\begin{array}{l}\text{D}=0.359\text{wing}\text{length}+0.173\text{tarsus}\text{plus}\text{toe}\\ -48.838\end{array}$(1)

This equation allows correct sexing of 77.1% of birds (87.5% of males and 54.1% of females). Hence, it may produce sex bias in the analysed sample. To limit this error, the individuals with discriminant score D < −0.33 are identified as males and those with D > 1.26 as females, which allows us to reduce misclassification to 5% of males and 5% of females (Meissner & Krupa 2016).

In total, 1649 adult common sandpipers were caught and measured; however, only 926 of them were sexed using the equation given above with D value limits (Table I). The percentage of unsexed birds varied between 41 and 50% in the following migration pentades, but these differences were statistically insignificant (G test, G = 4.52, df = 9, p = 0.87; the last four pentades were pooled due to a small sample size).

Table I. The total number of common sandpipers caught at different sites, and the number and percent share of individuals sexed by discriminant function. Study years at each site are given.

Site	Years of work	Total bird number	Number of birds withassigned sex	Percent sexed (%)
Puck Bay	1986–2001	272	147	54.0
Vistula mouth	2008–2015	390	207	53.1
Lisewo Malborskie	2002–2008	403	238	59.1
Pawłowice	2005–2008	583	333	57.1
Total		1648	925	56.1

For assessing difference in median migration dates of males and females only data from Lisewo Malborskie and Pawłowice were taken into account, because the number of walk-in traps at these two ringing sites was stable; hence, it was assumed that the changes in the daily number of trapped birds roughly reflect species migration dynamics (Meissner 2008). Common sandpipers from these two sites comprised 60% of all caught birds and 62% with sex assigned. Hence, it was assumed that the data collected at these two sites reflect the overall migration pattern of the common sandpiper. All statistical procedures were performed using STATISTICA 12 software (StatSoft Inc.).

Results

Migration periods of both sexes overlapped to a great extent, and males and females were present during the whole migration period. However, in the first days of July, males comprised only 23% of common sandpipers sexed molecularly and 31% among birds sexed by discriminant function. Later their percent share increased gradually, and in August between 74% and 94% of birds were males (Figure 2).

Figure 2. The percent share of the common sandpiper females (black) and males (grey) caught in following pendates and in the second half of August. The middle day of each pentade was given. A: birds sexed according to discriminant function; B: bird sexed molecularly. Numbers above the bars indicate the sample size. Data from all ringing sites were pooled.

The median capture dates of females and males differed significantly (median-test, χ² = 15.50, p = 0.001), being earlier in females, with a difference between medians of 6 days (19 July vs. 25 July) (Figure 3).

Figure 3. Migration phenology of adult common sandpiper males (grey) and females (white). Vertical line: median migration date; box: interquartile range; horizontal line: range. Data from Lisewo Malborskie and Pawłowice were pooled.

Discussion

In the case of the common sandpiper, the possibility of sexing birds using linear measurements is limited due to a large overlap of linear measurement distributions of males and females (Meissner & Krupa 2016). Even though only small males and large females could be recognised, we assumed that obtained results showed actual differences in migration phenology of both sexes, especially as biometric variation among the common sandpiper populations originating from different parts of the breeding area has not been reported (Glutz Von Blotzheim et al. 1977; Cramp & Simmons 1983; del Hoyo et al. 1996) and the number of unsexed birds remained stable within the whole migration period.

Our study showed a temporal sex bias among adult common sandpipers captured when migrating through Poland towards wintering grounds. This supports predictions on earlier migration of females, which was described in some Arctic waders from the Scolopacidae family, in which one sex deserts the brood before fledging and migrates towards the wintering grounds earlier than the second sex (Butler & Kaiser 1995; Ydenberg et al. 2005; Barshep et al. 2012). However, migration terms of male and female waders breeding in the Arctic are well separated in time (e.g. Meissner 2005, 2015; Barshep et al. 2012), but the common sandpiper nests in the temperate zone, where the egg laying is less synchronised than in the Arctic (Wallander & Andersson 2003; McKinnon et al. 2012). Moreover, in the temperate zone there is enough time for repeated clutches after breeding failure, which occurs rarely in more northerly latitudes (Holland et al. 1982; Jamieson 2011). The common sandpiper migrating through Central Europe comes from a vast breeding area extending from the Eastern Baltic countries through Sweden and Finland to Northwestern Russia (Stiefel et al. 1985; Fransson et al. 2008; Saurola et al. 2013). The period of egg laying differs between the northernmost and southernmost parts of this area by at least 1 month (Glutz Von Blotzheim et al. 1977). That is why, although the brood attendance by females declines rapidly about 3 weeks before chicks fledge (Yalden & Holland 1992), migration periods of both sexes overlapped to a great extent. Nonetheless, the difference in median capture dates of males and females was 6 days, which is almost a quarter of the pre-fledging period.

In this paper we showed for the first time that in the common sandpiper, females precede males during migration towards their wintering grounds, which was expected due to progressive decline in females’ attendance during brood rearing. Our results show that the difference in migration time of both sexes in this species is less pronounced than that in Arctic breeding waders, and this may be due to the less synchronised period of egg laying in temperate zone. This may concern also other species with unequal parental care inhabiting vast non-Arctic breeding areas.

Acknowledgements

We would like to thank all ringers and volunteers working at WRG KULING ringing stations. Special thanks to Agnieszka Ożarowska for language correction and helpful comments, and to an anonymous reviewer for the suggestions that have greatly improved the first version of the manuscript. Publication of Waterbird Research Group KULING no. 160.