Introduction

In migratory birds, protandry is the earlier arrival of males at the breeding grounds as compared to the females [1]. Different hypotheses (reviewed in [2]) have been formulated in order to explain the evolutionary causes of protandry. Most of them are not mutually exclusive and include direct and indirect fitness correlates. The degree of protandry in birds is an adaptive trait linked to reproductive fitness upon which we may expect selection. Although the ultimate (evolutionary) consequences of protandry have received much attention, knowledge of its proximate causes is still at its beginnings and only a few studies have focused on the mechanisms governing protandry in migratory birds (reviewed in [3]). Based on field observations, four main and not mutually exclusive hypotheses are proposed: (1) males depart earlier from the wintering grounds than females [4], [5]; (2) both sexes have common winter quarters and leave the winter quarters at the same time but males migrate faster than females (e.g. [6]); (3) females overwinter in areas more distant from the breeding grounds than males [7]–[9] and (4) males and females winter in different habitats [4]. Despite these observations it remains open whether protandry has an endogenous component as well. The present study deals with the earlier departure of males from the wintering grounds (hypothesis 1), which is likely to underlie endogenous regulation. Only a few studies were conducted on captive birds [3], [10], [11], all of which revealed an earlier onset of spring migratory activity in males relative to females. However, these studies were conducted either with birds captured in the wild prior to the experiment [3] or with hand-raised birds kept in simulated natural photoperiods [10], [11]. Thus, it cannot be precluded that the differences between males and females in the onset of spring migratory activity were either due to early experience prior to capture or to the photoperiodic cues [12]. Therefore, we aimed to study hand-raised young birds kept under constant photoperiod and otherwise controlled conditions for the whole first year of life because such an experimental design is necessary in order to isolate the intrinsic, endogenous component of protandry. Our model species was the northern wheatear (Oenanthe oenanthe), a widespread Holarctic breeding species with all populations overwintering in sub-Saharan Africa [13], which thus represents one of the most stunning migration systems in passerine birds. In the wheatear protandry is reported at breeding grounds [14], [15] as well as at stopover sites during migration [16], [17]. This species has been shown to reveal endogenous migratory restlessness and concurrent migratory fattening [18]. It is therefore a perfect candidate to address the question whether protandry has an innate background.

Results

We were able to determine the onset of nocturnal activity in spring in 25 out of 31 individuals (12 males and 13 females), and the onset of body mass increase in 37 out of 43 individuals (18 males and 19 females). Table 1 shows the outcome of the GLMMs testing for effects of year, sex and provenance on the onset of nocturnal restlessness and body mass gain in spring. There were effects of year on both variables which were probably due to the different age of the birds at the beginning of the experiments in early autumn. However, this had no influence on the effects of sex on spring behaviour as shown by the outcome of the models (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. GLMMs for effects of sex, year and provenance on onset of nocturnal restlessness and body mass increase in spring.

https://doi.org/10.1371/journal.pone.0031271.t001

In both populations males began their spring nocturnal restlessness on average 17 days (Iceland: 16 days; Norway: 36 days, Figure 1) earlier than females (p = 0.005, Table 1). Additionally, Icelandic birds started their nocturnal restlessness slightly earlier in spring than Norwegian birds (p = 0.002, Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Mean (±s.d.) dates of onset of nocturnal restlessness and of body mass increase in male and female northern wheatears in spring.

Black squares: Icelandic population, white squares: Norwegian population. The numbers above each boxplot indicate the sample sizes. Stars indicate P-level: **P<0.01. For better understanding, we assigned Day 1 of our count (starting on the day of photoperiodic switch) to a standardized date which was defined as the mean date of photoperiodic switch in the two years of experiment. For our data, this date was 26 August.

https://doi.org/10.1371/journal.pone.0031271.g001

Regarding the spring onset of body mass increase, males started significantly earlier only in the Norwegian population (p = 0.008, Table 1), on average 29 days earlier than in females (Figure 1). Nevertheless, Icelandic males started increasing their body mass on average 5 days earlier than females, although this difference was not significant (Table 1).

The sex-specific difference in spring migratory behaviour is not a result of different autumn onsets of migratory traits. We were able to determine the onset of nocturnal restlessness in autumn in 14 out of 22 individuals (7 males and 7 females), and the onset of body mass increase in autumn was calculated in all individuals. In autumn there were no differences between sexes in the onset of nocturnal restlessness (Wilcoxon rank-sum test: W = 10.5, p = 0.794). The increase of body mass started on average 5 days earlier in females than males (GLMM: intercept p<0.001; provenance p = 0.003, sex p = 0.021).

Discussion

Our study shows for the first time that even under constant photoperiodic conditions and without other environmental cues male northern wheatears prepared for and started spring migration earlier than females, thus proving that they are endogenously disposed to depart from their wintering grounds earlier than females.

In general it seems that the degree of protandry was higher in the Norwegian than in the Icelandic population. Populations living further North might respond to completely different time constraints which do not allow a high degree of protandry. This finding supports the theory that the endogenous regulation of migration-related traits, including nocturnal restlessness and body mass changes, is finely regulated and responds to the different needs of different populations [18].

The laboratory observed time lag in onset of spring migratory restlessness between males and females was more pronounced than the observed protandry in wild birds, either at arrival on breeding grounds [14], [15] or during passage at migratory stopover sites [16], [17]. The smaller time lag in wild birds could be attributed to the fine-tuning by photoperiod [19] which was kept constant in our experiment. Additionally, at stopover sites the difference in passage dates may be masked by the passage of different populations with unknown origin and destination. Moreover, other environmental factors may play a role in shaping the innate schedule to the pattern observed in nature, such as adverse weather, predation, parasites, and intra- and interspecific competition during stopover [20].

The observed differences in spring are unlikely to have originated from carry-over effects of the photoperiodic shift in late summer because the onset of nocturnal migratory restlessness did not differ between the sexes. Furthermore, females began increasing their body mass in autumn slightly earlier than males. Therefore, differences observed in spring are based on endogenous mechanisms, possibly due to a different pace of the males' internal clock as compared to females, which might be related to differences in energy turnover between males and females [21].

The finding that males are endogenously pre-disposed for earlier spring departure from wintering sites has important implications for a better understanding of the consequences of climate driven phenological changes in migratory birds [22]. Phenological traits can evolve quickly within just a few generations when the selective pressures are high [23]. A response to strong selection may cause considerable changes in the patterns of sex-differentiated timing of migration, and ultimately impact on the reproductive success of the species [3].

Materials and Methods

Supporting Information

Acknowledgments

We are grateful to Simon Fabian Becker, Rolf Nagel, Ulrike Strauss and Adolf Völk for technical assistance. Zoltan Barta, Paul Bartell, Andre Chiaradia, Mark Brigham, Nigel Richards and two anonymous reviewers provided helpful comments on an earlier version of the manuscript.