Effects of predator landscapes on the evolutionary ecology of routing, timing and molt by long-distance migrants

Raptors have long been known to prey on avianmigrants (Rudebeck 1950, 1951, Dekker 1980), andseveral studies estimate significant mortality rates formigrants (Lindstro¨m 1989, Kerlinger 1989, Mooreet al. 1990, Sillett and Holmes 2002). These findingsamply justify Alerstam and Lindstro¨m’s (1990) identi-fication of safety as one of the prime selective forcesshaping the evolution of avian migration strategies andbehaviour. However, as shown by a survey of recentvolumes on the ecology and evolution of avianmigration, safety has garnered much less interest fromecologists compared to that paid to time and energy,the other selective factors identified by Alerstam andLindstro¨m. Several contributions in ‘Stopover ecologyof Nearctic-Neotropical Migrants’ (Moore 2000)briefly consider predation risk, but only one chapterof 17 (by Moore et al. 1995) in ‘Ecology andManagement of Neotropical Migratory Birds’ (Martinand Finch 1995) even mentions predators. Neither‘The Ecology of Migrant Birds’ (Rappole 1995) nor theedited volume ‘Birds of two worlds: the ecology andevolution of migration’ (Greenberg and Marra 2005)devote any attention to the effects of predators onmigrants, though one chapter in the latter documentswhat is known of raptor migrations. That things may bestarting to change is evidenced by the inclusion of apaper directly addressing this topic (Lind and Cresswell2006) in the most recent symposium published onmigration issues, in the J. Orn., as well as by thepublication of several field experiments (Schmaljohannand Dierschke 2004, Pomeroy et al. 2006).Here we argue that features of migration such astiming and routing are strongly affected by safetyconcerns, and that careful consideration of such canilluminate aspects of migratory ecology. Phenomenasuch as bad weather may also threaten migrant safety,but our interest here is limited to predators. Our thesisis that both density-mediated and trait-mediated effects(see Lind and Cresswell 2005) have profound andpervasive influences on the evolution of avian migra-tion, with the latter being especially significant. Weconsider a number of continental-scale migratoryphenomena and explain how, in contrast to existingexplanatory hypotheses, these can be viewed as tactics toincrease safety.