Although age-related spatial segregation is ubiquitous, the underlying mechanisms are poorly understood. Here, we aim to elucidate the processes behind a previously established age-related foraging distribution of red knots (Calidris canutus canutus) in their main wintering area in West Africa (Banc d’Arguin, Mauritania). Based on 10 years of observations of 1232 uniquely color-ringed individuals of 1 to 18+ years old, we examined whether the observed age-related foraging distribution resulted from 1) spatial differences in mortality or 2) age-related shifts in habitat use. Using multistate capture–recapture modeling, we showed that with age foraging red knots moved away from the shoreline, that is, to areas with fewer surprise attacks by raptors. Considering uncertainties in the subjective gradient in predation danger with increasing distance from shore (as assessed from correlations between vigilance and distance from shore in foraging birds), we applied 2 different danger zone boundaries, at 40 m and 500 m from shore. Between years, red knots had a much higher chance to move from the dangerous nearshore area to the “safe” area beyond (71–78% and 26% for 40-m and 500-m danger zone boundary, respectively), than vice versa (4% and 14%). For neither danger zone boundary value did we find differences in annual mortality for individuals using either dangerous or safe zone, so the move away from the shore with age is attributed to individual careers rather than differential mortality. We argue that longitudinal studies like ours will reveal that ontogenetic shifts in habitat use are more common than so far acknowledged.
Data are provided for: Fig. 2 panels a (avgHUdfs10.txt) and b (vigilance_largescale.txt): vigilance as a function of distance from shore; Fig. 3: Number of individuals that were observed in nearshore dangerous and offshore safe habitat, shown for the 40 m (ch.knots.40m.txt ) and 500 m (ch.knots.500m.txt) danger zone boundaries; for the Multi-state mark-recapture analyses with use of the RMark package, including fig. 4(cr.knots.May2015.txt).