Evaluating the 50% assumption: Greater than expected first-year survival rates for migratory songbirds

Background/Question/Methods

With improved field methods for data collection and powerful new modeling tools, biologists are beginning to better understand survival rates and the processes that affect them.  However, for migratory animals, especially birds, recent advances in quantifying survival rates have largely focused on adults and ignored survival during the first annual cycle.  This knowledge gap is primarily a product of natal dispersal, as young birds rarely return to breed near their birth site.  As a result, population models, including one by the authors, have traditionally estimated first-year survival as 50% of adult survival.  At the same time, population viability analyses (using this 50% estimate), have found high sensitivity to variation in juvenile survival.   Consequently, more precise and robust estimates of first-year survival are needed to better understand population processes.  In 2002-10 we studied adult and first-year survival of two species of migratory grassland songbirds, the bobolink (Dolichonyx oryzivorus) and Savannah sparrow (Passerculus sandwichensis), in a large agricultural region of Vermont.  We estimated first-year survival with Cormack Jolly-Seber models in Program MARK.   We improved the accuracy of our models by 1) increasing detection rates by searching for natal dispersers within 1.5 km of our study fields (2005-10), and 2) applying molecular sex identification to the entire nestling data set, such that we knew the sex of all nestlings. 

Results/Conclusions

Both species and sexes showed short-distance natal dispersal (BOBO = 1275m ± 1354.3 SD, n=94; SAVS = 1401.3m ± 4037.3 SD, n=71), thereby providing a robust sample with which to model first-year survival.  Across both species and sexes, apparent first-year survival ranged from 0.30 ± 0.11 SE to 0.61 ± 0.13.  Estimates ranged from 57% less than to 52% greater than adult survival.  Savannah sparrow and bobolink first-year survival was estimated to be 0.48 ± 0.10 and 0.43 ± 0.17, respectively.  For Savannah sparrows, female first-year survival was 33% higher than male first-year survival, whereas adult female survival was 21% lower than adult male survival.  For bobolinks, female first-year survival was 33% higher than male first-year survival, whereas adult female survival was 33% lower than adult male survival.  These results have two important implications for population models: 1) first-year survival may be significantly greater than commonly assumed in population models, and 2) sex-specific survival may differ significantly between age classes.