How to recover after sea crossing: the importance of small islands for passerines during spring migration

Abstract

Every year, billions of migratory birds cover thousands of kilometers between their wintering grounds and their breeding areas. Along their migratory route, birds are often forced to break up their journey into several legs, mainly in correspondence with ecological barriers (e.g. seas and deserts), and stopover in areas where they can sleep and rebuild their energy stores. Stopover is a key migratory stage for many songbirds because strategies adopted at this stage may have direct bearing on the success of migration and, consequently, on the fitness of the individual. Although the restoration of energy reserves is likely the main goal for stopover in many cases, our recent studies suggest that there might be other reasons for stopping over, particularly after long non-stop flights. This seems to be the case of millions of songbirds that land on small coastal islands after crossing large waterbodies, although most of these sites do not provide sufficient refueling opportunities. Here, we aim to provide a new perspective on the role of such small islands by reviewing our recent findings on energy management in migratory songbirds. We are convinced that understanding the role of these islands for migratory birds, as well as how birds manage their energy at these sites, represents an important step toward an optimization of conservation strategies, since setting priorities for protection based on food availability alone may miss critical functions of stopover sites.

KEY WORDS:

• sleep
• locomotion
• avian migration
• energetics
• migratory strategies
• Mediterranean islands

INTRODUCTION

Migration is the most energy-demanding life-history stage in the life cycle of migratory birds and it can cost up to 50% of the annual energy budget of an individual (Drent & Piersma 1990). Due to the high energy requirements to cross ecological barriers and the impossibility to carry enough energy reserves to complete their migratory journey in one leg (Klaassen 1996), many bird species are forced to break their migration in several legs, alternating long endurance flights with refueling stopovers (Schmaljohann et al. 2007). Given that the process of energy accumulation is slower than the energy consumption (Alerstam & Lindström 1990; Alerstam 2011), migratory birds can spend up to 80% of the total migratory period at stopover sites (Hedenström & Alerstam 1997). Indeed, this stage requires double the amount of energy that is invested in migratory flights (Wikelski et al. 2003; Schmaljohann et al. 2012) and, therefore, it makes the stopover phase particularly important for a successful migration.

Recovery from fatigue and restoration of energy reserves are often considered as the main purposes of stopover. Indeed, the amount of energy reserves – mostly subcutaneous fat which is the main fuel of migratory flights (McWilliams et al. 2004) – plays a key role in behavioral strategies during stopover, such as departure decision and refueling (Meissner 1998; Dierschke & Delingat 2001; Fusani et al. 2009; Goymann et al. 2010; Cohen et al. 2014; Deppe et al. 2015; Woodworth et al. 2015; Dossman et al. 2016). At stopover sites, as well as when preparing for migration, birds experience a rapid fuel accumulation that results from a drastic increase in food intake (King 1961; Metcalfe & Furness 1984; Bairlein 1985), changes in diet (Bairlein 1990; Dingle 1996; Berthold 2003; Cooper-Mullin & McWilliams 2016) and modifications of metabolic pathways (Dingle 1996; Berthold 2003) that lead to an increased lipogenesis and, consequently to a rapid deposition of fat reserves. After their arrival at the stopover site, migrants need to make decisions that can influence their pace of migration (Nilsson et al. 2013; Gómez et al. 2017; Schmaljohann & Both 2017; Schmaljohann 2018). Because of differences in body condition at departure from previous sites, distance between sites, and weather conditions encountered during flight, the amount of remaining energy reserves is typically not homogenous among birds landing at stopover sites (Pilastro & Spina 1997). Energy management at stopover may vary depending on species (Ferretti et al. 2020b; Maggini et al. 2020), habitat quality of the site (Alerstam 2011; Cohen et al. 2012; McCabe & Olsen 2015; Ferretti et al. 2019a), and migratory period (Nilsson et al. 2013; Horton et al. 2016).

According to theoretical models, birds seek to optimize either total energy investment or overall time of migration (Alerstam & Lindström 1990). Individuals that adopt an energy-minimizing strategy, also called energy minimizers, should seek to reach an optimal amount of energy reserves before leaving the stopover site. However, this may require a considerable amount of time and slow down the pace of migration. On the contrary, time minimizers should aim at completing their migratory journey in the shortest possible time and therefore minimize stopover duration by leaving the stopover site if it does not guarantee a rapid accumulation of energy reserves (Hedenström & Alerstam 1997). The latter strategy seems to be more common during spring than fall (Nilsson et al. 2013), likely due to the high competition for early arrival during the pre-nuptial period (Smith & Moore 2005; Drake et al. 2014). Migrants that arrive with good energy stores will try to maintain their body condition and resume migration in the shortest time (Bairlein 1985; Biebach et al. 1986; Goymann et al. 2010), whereas birds with poor energy reserves will seek to restore their energy reserves by actively exploiting the site (Schmaljohann et al. 2013; Cohen et al. 2014). In the latter case, the interplay between physiological condition and the rate at which individuals accumulate energy stores – the so-called fuel deposition rate (FDR) – is the most important factor affecting behavioral decisions. Migratory birds need to maintain a high FDR in order to reduce stopover duration (Alerstam & Lindström 1990; Jenni & Schaub 2003; Alerstam 2011). Many studies have shown that birds experiencing a high FDR show a better body condition at departure (Carpenter et al. 1983; Lindström & Alerstam 1992; Fransson & Weber 1997; Schmaljohann & Dierschke 2005; Bayly 2006, 2007; Delingat et al. 2008; Schmaljohann et al. 2013). Moreover, FDR decreases with increasing body condition (Schmaljohann & Eikenaar 2017). FDR is affected by a number of environmental factors such as predation risk (Fransson & Weber 1997; Weber et al. 1998), competition (Moore & Yong 1991), weather conditions (Jenni-Eiermann & Jenni 1994; Schaub & Jenni 2001) and abundance of food resources (Bibby & Green 1981; Graber & Graber 1983). Although the restoration of energy reserves is likely the main goal for stopover in many cases, recent studies of our group indicate that there might be other reasons for stopping over, particularly in certain stages of migration, for example soon after crossing large ecological barriers. Here, we will review our recent findings about energy management in migratory passerines after crossing the Mediterranean Sea and provide a new perspective on the role of small islands for migratory birds. We will start by discussing why billions of birds stopover on islands that do not offer sufficient resources for an efficient refueling. Then, we will explore behavioral strategies to manage energy reserves after long flights, as well as how such strategies may vary in response to some physiological and environmental factors. Finally, we will discuss new perspectives on the key role of small islands for migratory birds, to better understand stopover ecology and site exploitation.

Barrier crossing: why do migrants land on small islands during sea crossing?

During migration, large waterbodies are major ecological barriers for billions of terrestrial birds. Due to the lack of landing opportunities encountered during such crossing, migrants engage in exhausting multi-hour flights that in case of nocturnal migrating birds may extend their travel into day time (Grattarola et al. 1999; Deppe et al. 2015; Adamík et al. 2016) and, consequently, may lead to an extreme depletion of energy reserves. In this ecological context, small islands, such as those in the Northern Gulf of Mexico (Moore et al. 1990) and in the Mediterranean Sea (Spina et al. 1993), gain a considerable importance for migrants by providing suitable sites to land. Regardless of their ecological requirements, a wide range of species stops on these islands (Moore et al. 1990; Spina et al. 1993) even when they are not far from the mainland and do not provide much variety of habitat type. Especially during northward migration, Tyrrhenian islands are good sites where to investigate site-exploitation strategies after sea crossing. The fact that most individuals stop despite having enough energy reserves to continue their migratory flight (Pilastro & Spina 1997) seems to indicate reasons for landing other than refueling, which challenges the common view that the main goal of stopover after barrier crossing is energy recovery (Carmi et al. 1992; Klaassen 1996; Aamidor et al. 2011). These islands are usually just a few kilometers from the coast and offer scarce opportunities for foraging and limited freshwater sources, which does not make them the most suitable sites for replenishing energy reserves. A recent study based on mark-recapture data showed that most of the individuals captured on the island of Ponza (40°55ʹN, 12°58ʹE) did not stop over for longer than 1 day and, in general, did not increase their body condition during this period with the exception of species that breed in habitats similar to those found on the island (Maggini et al. 2020). These findings indicate a strong dependency of refuelling on stopover habitat (Cohen et al. 2012; McCabe & Olsen 2015) and raise the question why all other species land at these sites although more suitable habitats are available a few kilometers away. Once landed, individuals that experienced an extreme depletion of energy reserves during the migratory flight will try to improve their condition by intensively refueling (Ferretti et al. 2019a) and carefully managing their energy reserves (Ferretti et al. 2019a, 2019b). However, most of them hardly increase their energy stores at these sites (Maggini et al. 2020) despite considerable investments in food searching (Cohen et al. 2012), which has negative effects on FDR (Fig. 1). For this reason, these individuals leave the site soon after landing (Schmaljohann & Eikenaar 2017), as confirmed by the relatively short time (41.3 hr, Goymann et al. 2010) that lean birds spend on these small islands. On the other hand, some individuals manage to reach these sites in good conditions. Despite they having an amount of energy reserves that would allow them to continue their migratory flight, they decide to stop for few hours (Goymann et al. 2010) before resuming migration on the following night. During their stay, fat individuals spent most of their time asleep (Schwilch et al. 2002; Ferretti et al. 2019b, 2020b), indicating a strong pressure to recover for sleep lost during the endurance migratory flight. The fact that these birds stop before reaching the coast, regardless of the presence of an optimal habitat, and mainly sleep during their permanence on these small islands, suggests a homeostatic regulation of sleep following endurance flight (Rattenborg et al. 2009) that leads to land at the first available area. Moreover, small coastal islands may host a low density of potential predators (Beauchamp 2004) and, for this reason, be optimal sites to stop when sleep recovery is the primary goal. Taken together, these evidences indicate that the use of islands may be less related to refueling than generally thought. Indeed, other physiological aspects seem to explain better than FDR why so many songbirds stopover at these sites after sea crossing.

Fig. 1. Rates of energy accumulation of migrating songbirds that spent at least 1 day on Ponza. Rates of energy accumulation (presented as “Boxplots”) were calculated based on bird’s body mass at last capture subtracted from body mass at first capture relative to lean body mass and time spent at the stopover site. To improve accuracy of this measurement, we used the number of hours between the first and the last capture as “time of permanence”. Lean body mass was estimated following Kelsey et al. (2019). Data from Maggini et al. (2020)

ENERGY MANAGEMENT ON SMALL ISLANDS

The role of locomotion

When migrants with heightened energy demands are concentrated in unfamiliar areas, competition for available food resources can reduce the rate at which they meet nutritional requirements (Carpenter et al. 1983; Moore & Yong 1991). In case birds land at a site that is suitable for refueling, a fast energy recover ybecomes the main goal of stopover (Alerstam & Hedenström 1998), and consequently birds initiate intensive foraging (Schaub & Jenni 2001). However, intense foraging may increase the exposure to potential predators (Lima & Dill 1990), therefore behavioral decisions aiming to optimize energy accumulation during stopover are supposed to be the result of a trade-off between energetic needs and predation risk. Several studies have confirmed this hypothesis (Sih 1980; Lima 1988; Pomeroy et al. 2006) and behavioral decisions can be summarized in a simple rule-of-thumb: migratory birds actively search for areas with abundant food (Cohen et al. 2012), however in case of high predation risks they move to safer stopover sites even though the new site may not guarantee a high FDR (Lima & Dill 1990; Brown & Kotler 2004), as may be the case of small coastal islands. For example, when released in a stopover site characterized by few food sources – a scenario similar to the one on small islands – Red-eyed vireos (Vireo olivaceus) moved further and searched for habitat types that could guarantee higher food availability (Cohen et al. 2012). Nevertheless, there are cases in which moving to different sites might be very hazardous, as for birds with very poor energy reserves. When leaving the site is not an option, the only alternative is to optimize energy consumption by the modulation of energy invested in locomotion. In a recent comparative study conducted on the island of Ponza, we showed species-specific responses to a gradient of food availability among songbirds. By using the amount of diurnal locomotion as a proxy for food searching behavior (Morton 1967; Astheimer et al. 1992; Fokidis et al. 2011), we found that a few species show an intense diurnal activity regardless of food availability, whereas most of the species increased locomotor activity with decreasing amount of food (Ferretti et al. 2019a). Although it appears plausible that the reduction of diurnal locomotion in response to food availability has an energy-saving function (Bäckman et al. 2016; Schofield et al. 2018) – as suggested by the comparison of metabolic data between migrating birds displaying different behaviors (Ferretti et al. 2019b), our current knowledge does not allow to rule out that reduction in locomotion is also functional to reduce potential predation risk. Indeed, birds that find limited food sources are expected to increase their activity in the attempt to improve refueling (Cohen et al. 2012; Ferretti et al. 2019a), although this strategy may increase their exposure to potential predators (Lima 1986; Lind & Cresswell 2006). From this perspective, the habitat homogeneity that characterizes most of small islands may be beneficial for birds after sea crossing. In fact, at these sites the lack of food sources is often counterbalanced by a reduced predation risk (Beauchamp 2004) which potentially allows to invest more time and energy into refueling. A lower predation risk on small islands might be attributed to two main factors: the size of the site and the number of individuals sharing the site. Small islands can support only a low number of resident predators (Del Monte-Luna et al. 2004), mainly due to the competition for food sources out of the migratory season and breeding spots (i.e. nest sites in the case of raptors). Although aerial predators might migrate or increase their home range, this is not possible for mammalians predators whose population is significantly affected by potential limitations due to the size of the island (Del Monte-Luna et al. 2004), an aspect that may reduce the overall predation risk. Migrating birds may also benefit from sharing the site with thousands of individuals. This reduces the risk of being caught thanks to an increased group alertness and a “dilution effect” (Sirot 2006) if attacked by the few predators present on islands. Taken together, the aforementioned reasons seem to make small islands optimal sites to stopover no matter if the priority is to replenish energy reserves or simply rest.

The role of sleep

Locomotion is not the only behavior that can be modulated in order to optimize energy investment. Sleep is a fundamental behavior of all organisms and occupies most of bird’s lives (Toates 1980). Even though the functions of sleep are still poorly understood, we know that animals during sleep can lower their metabolic rate by 15–20% compared to the wake phase (Stahel et al. 1984; Fontvieille et al. 1994; Jung et al. 2011; Markwald et al. 2013). The lower metabolic rate can be related to a small decrease in body temperature, facilitated by physiological mechanisms (i.e. vasoconstriction and piloerection) that reduce the body’s thermal conductance (Berger & Phillips 1995). Therefore, the modulation of sleep based on the need for physiological recovery can be a good strategy to optimize energy expenditure (Allison & Van Twyver 1970; Berger 1975). How birds sleep during migration remains a mystery, although several non-mutually exclusive hypotheses have been formulated (Rattenborg et al. 2004; Rattenborg 2006). Previous field studies on nocturnal migratory songbirds suggest that sleep occurs during migration as a homeostatic process (Borbély & Tobler 2011; Lesku et al. 2011): birds profit from stopover sites after crossing large ecological barriers to recover from sleep loss accumulated during non-stop flights (Bairlein 1985; Biebach et al. 1986; Ferretti et al. 2019b, 2020b). Indeed, birds may sleep either during the day or during the night depending on their amount of energy reserves and on their migratory strategy (Ferretti et al. 2019b, 2020b). Moreover, in migrating Whitethroats (Sylvia communis) the total amount of sleep is positively correlated with changes in energy reserves (Ferretti et al. 2020b), confirming an energy-saving function of sleep. However, the duration of sleep is not the only parameter of this behavior that can be regulated to manage energy expenditure. Another aspect that is important for energy saving is the posture displayed during sleep (Amlaner & Ball 1983). It has been observed in different species that the posture assumed during sleep is often associated with environmental conditions, in particular external temperature. The Black-billed magpie adopts a posture with the beak tucked in the scapular feathers more readily in the coldest winter nights than in warmer ones, suggesting that this posture can reduce heat loss by reducing the exposed surface (Reebs 1986). The use of certain positions depending on environmental factors has been studied also in mallards, where it was seen that the transition to a posture with the head turned backward is favored by climatic conditions such as rain and lower temperatures (Midtgård 1978). An association between the tucked posture and low ambient temperature has been found also in passerines (Wellmann & Downs 2009). The hypothesis of a link between posture preference and environmental temperature has been further supported by a recent study that showed that birds with particular morphology – i.e. larger bills in relation to body size – and/or living at northern latitudes have a higher probability to sleep with their head turned backwards and tucked in the scapular feathers (Pavlovic et al. 2019). The association between sleep posture and energy consumption has been recently confirmed also in nocturnal migratory songbirds during spring migration (Fig. 2) (Ferretti et al. 2019b). Taken together, these results suggest a direct relationship between sleep posture displayed and energy expenditure. Although the amount of energy stores may affect the sleep pattern species-specifically, the strong influence of energetic status on posture preference seems to be a strategy that is shared across migratory songbirds (Ferretti et al. 2020b). Indeed, energy challenged individuals spend more time with their head tucked in than those that have large energy depots. Lean individuals trade-off safety for a lower energy consumption by tucking in their head under the scapular feathers. By contrast, fat individuals prefer to sleep with their head facing forward, which allows a faster reaction to potential predators (Ferretti et al. 2019b). The low energy consumption during tucked-in sleep is reflected also in the variation in energy reserves in both Whitethroats and Garden warblers, with tucked individual showing a lower change in energy reserve (Ferretti et al. 2020b). Although it is not clear whether each posture is associated with a specific sleep architecture, the reduction of metabolic rate in Garden warblers sleeping tucked in seems to be resulting from a reduced conductance achieved by increasing insulation on featherless body part, e.g. the eyes and the bill (Ferretti et al. 2019b). Indeed, birds have been estimated to save an amount of energy equivalent to 9 hr of fueling by simply displaying the tucked posture during the whole night (Ferretti et al. 2020a).

Fig. 2. Energy consumption in sleeping Garden warblers. Energy consumption decreases by almost 14% when birds tuck their head in. If we consider that Garden warblers on Ponza spend 62% of the night sleeping (i.e. 5 hr between 20:00 and 4:00; Ferretti et al. 2020b), this species can save approximately 0.68 kJ/night by spending the whole night tucked in. This amount of energy corresponds to 0.02 g of fat (considering the energy equivalent of fat as 37.6 kJ/g; Jenni & Jenni-Eiermann 1998). Redrawn from Ferretti et al. (2019b)

CONCLUSION

In this review, we discussed the roles of small coastal islands for migrating songbirds and examined recent studies challenging the common view that refueling is the primary goal during stopover at these sites. Indeed, several songbird species landing on these islands likely prioritize recovery from sleep loss experienced during their nocturnal migratory flight and exploit these sites for opportunistic refueling only in case of a drastic depletion of energy reserves. When the habitat does not provide sufficient resources for refueling, lean individuals try to preserve their already poor energy reserves by adopting behavioral strategies (e.g. sleeping with their head tucked into the feathers or modulating locomotor activity) aimed at optimizing energy saving as to allow relocation to more favorable refueling areas the following day. Such behavioral strategies favoring rest and sleep recovery will be facilitated by the scarcity of predators on small coastal islands. Future studies should focus on the interplay between species-specific physiological adaptations, stopover site exploitation, and habitat quality. Indeed, these aspects are determinant for understanding the ecophysiology and energetics of bird migration. This is overwhelmingly important for small islands that may represent critical migratory bottlenecks because of the direct impact that they may have on population dynamics at the breeding grounds (Dunn et al. 2004; Osenkowski et al. 2012).

AUTHOR CONTRIBUTIONS

A. Ferretti, I. Maggini and L. Fusani equally contributed to this study.

ACKNOWLEDGEMENTS

Research of our group presented in this review was made possible by the contributions of many students, personnel and volunteers in the field and laboratory. The ringing station in Ponza is operating within the long-term ringing project ‘Piccole Isole’ coordinated by Dr Fernando Spina, Istituto Superiore per la Protezione e la Ricerca Ambientale.

DISCLOSURE STATEMENT

No potential competing interest was reported by the authors.

Additional information

Funding

This study was supported by start-up funds of the University of Veterinary Medicine, Vienna, and the University of Vienna to L. Fusani, and a “Completion grant” of the University of Vienna to A. Ferretti.