Naive birds rely on endogenously controlled migration strategies to reach their migratory destinations. These strategies include among others a general spatiotemporal pattern of migration (Berthold 2001), whereby the onset of migration and the general trajectory are genetically determined (Gwinner & Wiltschko 1978; Berthold & Helbig 1992; Gwinner 1996; Conklin et al. 2010).
In many bird species, whose breeding and wintering grounds are separated by large ecological barriers, i.e. mountains, deserts and sea, specific migration strategies evolved that minimize the constraints of such a demanding journey. Genetically determined shifts in the migratory direction allow birds to cross an ecological barrier along a route avoiding the most risky parts (Gwinner & Wiltschko 1978; Hake, Kjellén & Alerstam 2001). Despite this genetically determined general spatiotemporal organization, migrants respond to environmental cues on the ground (Erni et al. 2002; Jenni & Schaub 2003) as well as aloft (Liechti 2006; Schmaljohann, Bruderer & Liechti 2008; Schmaljohann, Liechti & Bruderer 2009) and to their current body condition (e.g. Bairlein 1988; Dierschke, Mendel & Schmaljohann 2005). Physically fit migrants are less likely to detour from the principal migratory direction, if a barrier is ahead, whereas lean birds orient away from the barrier (Alerstam 1978; Sandberg 1994; Sandberg & Moore 1996; Sandberg 2003; Deutschlander & Muheim 2009). There are, however, no studies testing the separate and combined effect of wind condition and fuel load in free-flying birds. In migrants, in which a shift in the migratory direction is linked with the crossing of an ecological barrier, the timing of departure in relation to environmental conditions and body condition is particularly important. Ignoring one crucial factor may lead to failure, because wind alone cannot carry a bird and high fuel loads are useless when flying in strong headwind (Liechti 2006). Migrants need to consider both wind conditions and fuel load, because both factors determine a bird’s potential flight range. Additionally, they have to select among alternative directions to set off, if there are cues on the extent of the barrier in different directions. In summary, migrants cannot rely solely on the endogenously controlled spatiotemporal migration strategy to reach their migratory goals, but need complex adaptations for short-term decisions on barrier crossing and detour migration (Gwinner 1996; Thorup & Rabøl 2001; Alerstam 2001; Henningsson & Alerstam 2005). Following the Lagrangian approach by considering the inner state of a migrant (body condition) and the current dynamic environment (wind) allows to study the phenotypic response in relation to the general spatiotemporal pattern of the migration system (Nathan et al. 2008; Shamoun-Baranes et al. 2010).
The northern wheatear’s (Oenanthe oenanthe L.) subspecies leucorhoa exhibits such a migration pattern with a shift in the migratory direction and an ecological barrier to be crossed. When en route from wintering areas in West Africa (Cramp 1988), the birds first fly north towards Western Europe. They eventually shift their migratory direction towards the north-west to cross via a long flight the North Atlantic and finally reach their breeding grounds (Fig. 1). On Helgoland, a small island in the North Sea, large numbers of leucorhoa northern wheatears (leucorhoa wheatear hereafter) rest during spring migration (Dierschke & Delingat 2001). Birds have no visual cues on potential stopover sites in the open sea towards north-west, but they may collect information on the dimension of the sea barrier in any other direction (Fig. 1). Hence, birds departing from Helgoland have two options. They may head directly towards their breeding areas in a north-westerly direction and cross several hundred kilometres of sea. Alternatively, they may depart in any other direction. If they do so, their next stopover opportunities are relatively near (about 50–100 km or 1–3 h of flight, Fig. 1). Departing to the north-west enables fast migration but is otherwise risky (i.e. no cues on potential stopover sites). It is to the advantage of lean birds and birds facing unfavourable weather conditions to circumvent the barrier initially and to follow the longer but safer pathway over land (Fig. 1). The latter birds would then head towards their breeding areas after refuelling and a wait for favourable wind conditions from a new stopover site. We hypothesize that the genetically determined general strategy for shifting the general migration direction from north to north-west to cross the sea barrier is adjusted jointly by body condition and wind conditions through their influence on the birds’ potential flight range. Thus, we expect that fuel load, wind conditions and visual cues will significantly influence the direction in which leucorhoa wheatears depart from Helgoland.
The timing of departure is also crucial in relation to the birds’ decision about their subsequent migration stage. Birds that choose to depart directly towards the north-west and that anticipate crossing the sea are expected to depart early after sunset to fly all night, as they are nocturnal migrants. Birds leaving in other directions may depart at any time in the night because their next stopover sites can be reached by a short flight. We thus hypothesize that body condition and wind conditions at departure influence the timing of nocturnal departure.
To test these hypotheses, we radio tagged leucorhoa wheatears on Helgoland during spring migration. In contrast to cage experiments (Nievergelt & Liechti 2000), radio tagged birds interact freely with their environment and allow determining the exact departure time and tracking the departure direction for the first 12–15 km of birds’ migration off Helgoland. The tracking range was, therefore, about three times that of tracking radar (cf. Schmaljohann et al. 2008).