A method for studying post-fledging survival rates using data from ringingrecoveries

We present a method for studying post-fledging survival rates from data on national ringing recoveries. The approach extends the classical two-age-cla ss models of Brownie et al. (1985) to include a third age-class of birds ri nged as nestlings. The models can incorporate age-class-specific and year-s pecific variation in reporting rates and survival rates. Unlike intensive r esighting studies at single sites, this approach does not confound death wi th emigration and provides estimates of survival which are representative o f a larger geographical area. We implemented the models with SURVIV softwar e, using likelihood-ratio tests to assess the significance of possible age and year effects art survival rates or reporting rates. We demonstrate this approach using a set of data an British-ringed Song Thrushes Turdus philom elos. In this case, we were unable to detect any age-class-specific or year -specific variation in reporting rates and, although the survival rates of first-year and adult Song Thrushes varied significantly between years, post -fledgling survival did not show any detectable inter-annual variation. Son g Thrushes ringed as nestlings had a 0.38 chance (95% CI = 0.30-0.45) of su rviving the 63-day post-fledging period (here defined as the period separat ing the average dates of birds ringed as nestlings and birds ringed as inde pendent first-year birds).