McKellar, Ann E., Reudink, Matthew W., Marra, Peter P., Ratcliffe, Laurene M., and Wilson, Scott. 2015. "Climate and density influence annual survival and movement in a migratory songbird." Ecology and Evolution 5 (24):5892-5904. https://doi.org/10.1002/ece3.1854
Assessing the drivers of survival across the annual cycle is important for understanding when and how population limitation occurs in migratory animals. Density-dependent population regulation can occur during breeding and nonbreeding periods, and large-scale climate cycles can also affect survival throughout the annual cycle via their effects on local weather and vegetation productivity. Most studies of survival use mark–recapture techniques to estimate apparent survival, but true survival rates remain obscured due to unknown rates of permanent emigration. This is especially problematic when assessing annual survival of migratory birds, whose movement between breeding attempts, or breeding dispersal, can be substantial. We used a multistate approach to examine drivers of annual survival and one component of breeding dispersal (habitat-specific movements) in a population of American redstarts (Setophaga ruticilla) over 11 years in two adjacent habitat types. Annual survival displayed a curvilinear relation to the Southern Oscillation Index, with lower survival during La Niña and El Niño conditions. Although redstart density had no impact on survival, habitat-specific density influenced local movements between habitat types, with redstarts being less likely to disperse from their previous year's breeding habitat as density within that habitat increased. This finding was strongest in males and may be explained by conspecific attraction influencing settlement decisions. Survival was lowest in young males, but movement was highest in this group, indicating that apparent survival rates were likely biased low due to permanent emigration. Our findings demonstrate the utility of examining breeding dispersal in mark–recapture studies and complement recent work using spatially explicit models of dispersal probability to obtain greater accuracy in survival estimates.
