In our experiment, we successfully simulated the natural sequence of life-history stages in captive reed buntings. This species shows a high degree of spring arrival protandry and is believed to combine nocturnal and diurnal migration. We obtained the data on the dynamics of nocturnal and diurnal locomotor activity, fat reserves and leukocyte profile from late September up to June in the following year, i.e., from autumn migration through wintering to spring migration and breeding. We found that reed buntings exhibited pronounced nocturnal migratory activity both in autumn and in spring. In general, periods of Zugunruhe in caged birds coincided well with the timing of migration of the species in the wild (Valkama et al. 2014; Noskov et al. 2020): autumn movements progress from mid-September to mid-November and spring movements from February to the beginning of May. Due to the fact that reed buntings also migrate at daytime in the wild (Panov 2011), to explicitly describe the migratory strategy of this species, it was necessary to take into account diurnal activity patterns in experimental conditions. Our data report high morning activity of males just before and during spring period of Zugunruhe in our caged birds (Fig. 1), which testify in favor of innate propensity for diurnal migration in this sex. As previously noted, distinguishing between diurnal migratory activity and other forms of locomotion can be challenging in captivity. In our case, the dynamics of fat reserves in males suggest that morning activity in this sex did not reflect their intensive foraging (we did not record a higher rise of fat reserves compared to females) but rather their readiness for migratory flight expanding over the day. Next, we discuss how our results fit the three non-mutually exclusive hypotheses associated with proximate mechanisms of spring arrival protandry.
Timing of migration (Hypothesis 1: males start their spring migration earlier than females)
The dynamics of spring Zugunruhe did not provide evidence for sex differences in the onset dates of spring migration in reed buntings. However, it is plausible that male reed buntings do start their migration earlier than females in our experiment too. The dynamics of diurnal activity showed that males increased morning activity approximately two weeks earlier before the start of Zugunruhe (Fig. 1). Thus, it is possible that in the nature, males could depart from wintering grounds during the daytime. Sex differences in departure time from wintering grounds are assumed to be the primary mechanism of spring arrival protandry compared to other mechanisms (Briedis et al. 2019). Available data on tagged birds (geolocators, PTT tags, radio-telemetry, mark-recapture method) and a few captive studies (with the timing of Zugunruhe as a proxy for migratory departure) have shown an earlier onset of spring migration by males in 14 out of 18 cases (Morbey and Hedenström 2020).
Latitudinal segregation of sexes in winter (Hypothesis 2: males winter closer to the breeding grounds)
Ringing data suggests that female reed buntings generally winter further south from their breeding areas than males (George 2002; Villarán and Pascual-Parra 2003; Arizaga et al. 2011), a phenomenon also widely reported in other species (Ketterson and Nolan 1976; Woodworth et al. 2016). Therefore, we hypothesized that if males winter at more northern latitudes, they should travel shorter distances and exhibit an overall shorter period of migratory activity compared to females. Indeed, the duration of spring Zugunruhe in captive male reed buntings is approximately 10 days shorter than in females, which translates into earlier termination of nocturnal migratory activity. However, if we consider the increase of diurnal activity in males in February as the start of spring migration, the duration of the spring migratory period would be comparable in both sexes. Thus, our data precludes us from making a firm conclusion on the latitudinal sex segregation hypothesis.
Sex-specific migration speed (Hypothesis 3: males travel faster)
If spring arrival protandry is achieved by faster migration speed, we expected a higher overall migratory activity in captive males compared to females. To support higher activity levels, we also expected males to accumulate more fat reserves, as this allows potentially longer flight bouts, which can contribute to the higher migration speed (Alerstam and Lindström 1990; Schmaljohann 2018). Sex differences in activity levels and energetic conditions can potentially translate into variation in leucocyte profile parameters (e.g. Råberg et al. 1998; Hasselquist et al. 2007).
In reality, our results revealed a more nuanced and complicated picture. Analysis of seasonal patterns of locomotor activity showed that in spring, diurnal activity levels in males were significantly higher than in females (Fig. 1b). In contrast, the nocturnal activity levels in males were much lower than in females, which was especially evident in the middle and terminal portions of the spring migration (Fig. 1a, Online resource; Table S4). Unfortunately, differences in measuring diurnal and nocturnal activity and sex-specific seasonal dynamics of both parameters do not allow for a quantitative comparison of overall migratory activity levels between sexes in our experiment. We can only suggest that a remarkable rise in diurnal locomotor activity in males reflects a genuine increase in total migratory activity levels.
Analysis of circadian activity dynamics in spring confirmed and elaborated the difference between the sexes. During the first hours after lights on, males were on average 5–7 times more active than females (Fig. 2), whereas nocturnal activity in males was significantly lower compared to females. Males increased their activity only shortly before lights on, whereas females expressed rather high Zugunruhe levels during the two-thirds of the night (Fig. 2). Assuming that the circadian pattern of migratory activity might reflect actual behaviour in the wild (Schmaljohann et al. 2015; Ilieva et al. 2023), the observed dynamics of circadian activity suggest that, in the wild, male reed buntings start their migration shortly before sunrise and continue migratory flight some hours during daytime, while females migrate mainly during the second half of the night and cease migration around sunrise. The possibility of continuing the nocturnal migratory flight into the daytime was recently reported in another short-distance migrant, European starling (Sturnus vulgaris), which had been previously assumed to migrate primarily during the day (Vīgants et al. 2023). In captive conditions, a similar explanation was recently proposed for circadian activity dynamics in Sedge warblers (Acrocephalus schoenobaenus) in autumn (Ilieva et al. 2023).
Results on physiological conditions also contrast our initial predictions. Physiological preparations for migration and the level of energetic reserves in particular are assumed to reflect migratory strategy (Dolnik 1975; Alerstam and Lindström 1990; Maggini and Bairlein 2010). We observed that, in spring, females increased their energy reserves to a much higher degree than males (Fig. 3), even though all birds had ad libitum access to food. Probably, such pattern of fattening is associated with predominantly night migration in females. Furthermore, sex-specific dynamics of locomotor activity and fat reserves did not result in sex differences in leukocyte profile parameters during spring migration (Online resource, Fig. S3, Table S7). Thus, the data on reed buntings corroborates most studies published so far (Ewenson et al. 2001; Davis et al. 2004; Norte et al. 2009; Pap et al. 2010; Jakubas et al. 2011). Our results indicated that immune system activity and the level of long-term physiological stress were similar between the sexes, which can be partly explained by mild experimental conditions and unlimited access to food.
The pivotal question is whether and how patterns of migratory activity and fattening found in captive male reed buntings in spring can contribute to their earlier arrival to breeding grounds in the wild. The total speed of migration is affected mainly by the stopover duration and daily travel speed (Schmaljohann 2018). It is proposed that stopover duration (and factors affecting this parameter, e.g., fuel deposition rate) is a major determinant of total migration speed compared to travel speed (Schmaljohann 2018). We propose that by combining night and daytime flights in spring, male reed buntings benefit from both the advantage of nocturnal flight and an efficient start of foraging because they minimize search/settling costs due to the better selection of optimum foraging sites (Alerstam 2009). We suggest that such behaviour reduces the total stopover duration, thereby increasing migration speed. Additionally, during the daytime, males can combine foraging and movement in the migratory direction, which further increases daily travel speed (Alerstam 2009). To behave in such a way, males do not need large energetic reserves, and our data on fat reserves seem to confirm this. On average, higher energetic reserves in females are likely associated with their strategy to migrate predominantly at night and can serve, e.g., as a safety margin in case of landing at a site with poor refuelling opportunities.
To our knowledge, only a single study has reported sex differences in the intensity of spring migratory activity in experimental conditions. In a Nearctic-Neotropical night migrating songbird, the black-throated blue warbler (Setophaga caerulescens), males displayed higher intensity of wing whirring behaviour, suggesting potential sex differences in flight behaviour that could influence migration speed (Deakin et al. 2019). However, no sex difference in the timing and extent of migratory fat deposition during spring migration was recorded in this species. In the wild, sex differences in traits related to faster travel speed, such as flight speed (Tøttrup et al. 2012), refuelling rate (Seewagen et al. 2013; Schmaljohann et al. 2016), or stopover duration, are exceptional (5 of 23 cases studied) (Morbey and Hedenström 2020).
Autumn / spring migrations comparison
Keeping birds from late September to early June provided insights into patterns of locomotor activity and physiological conditions during autumn and spring migrations as well as during wintering. We noted that throughout wintering, both sexes exhibited similar behavior, being inactive at night with a peak in morning activity (Fig. 1, 2), aligning with our expectation of a diurnal lifestyle in winter. We are aware of only a few studies that provide comparative details about migratory locomotor activity for both spring and autumn migration (Bäckman et al. 2017; Sharma et al. 2018; Macías-Torres et al. 2022; Vīgants et al. 2023) and none that specifically aimed to compare seasonal patterns of migratory activity between males and females.
Unlike in spring, male and female reed buntings exhibited rather uniform seasonal dynamics of locomotor activity in autumn, with high nocturnal and relatively low diurnal activity levels (Fig. 1). This suggests that both sexes migrate mainly at night. The data on birds killed at lighthouses collected in the Baltic region (Hansen 1954) also suggested that nocturnal migration of the reed bunting is more pronounced in autumn than in spring. The number of reed buntings found dead in autumn was more than ten times higher than in spring. This was not the case for many other common nocturnal migrants, such as blackbird (Turdus merula) or northern wheatear (Oenanthe oenanthe). Subtle sex variations in seasonal activity patterns were noted, with females exhibiting earlier onset and a slightly more intensive Zugunruhe at the start of migration. This could be possibly linked to the earlier development of the autumn migratory state in young female birds, which was proposed for reed bunting in the wild (Schmitz and Steiner 2006; Lehikoinen et al. 2017) and has been recorded in other long- and short-distance migrants (Lehikoinen et al. 2017).
The circadian dynamics of locomotor activity in autumn were also similar in both sexes (Fig. 2). Males displayed slightly elevated morning activity, suggesting potential daytime migratory flights, albeit less pronounced than in spring. The overall magnitude of the sex-related difference in diurnal activity levels was much reduced compared to spring, which further solidifies the assumption of predominantly nocturnal migration in reed buntings in autumn. Clear sex-related variation in the diurnal/nocturnal activity ratio between spring and autumn further supports the notion that increased diurnal activity in males during spring reflects genuine migratory behaviour.
The fat reserve dynamics in autumn mirrored the activity data, with both sexes gaining reserves in parallel with increased nocturnal migratory activity (Fig. 1a and Fig. 3). However, females tended to accumulate more fat reserves in the first half of autumn (Fig. 3), which is probably associated with their earlier development of the autumn migratory disposition mentioned earlier. These and other results highlight the close association between Zugunruhe levels and the amount of fat reserves in reed buntings. As in spring, we did not find sex differences in leukocyte profile parameters during autumn migration (Online resource, Fig. S3, Table S7). However, we observed elevated total leukocyte counts in autumn, which aligns with previous findings for other bird species (Owen and Moore 2008; Demina et al. 2019). This is probably associated with elevated immune system activity in reed bunting in autumn.