On the use of the robust design with transient capture-recapture models

Capture-mark-recapture studies provide a useful mechanism for estimating the components of the population dynamics of birds, especially survival. In such studies, it is important that the population being captured matches the population of interest. In many studies, transients are captured along with the population of interest (e.g. resident breeders). Ignoring that phenomenon produces negatively biased survival estimates, because transients do not return. Although transients cannot be distinguished from residents in the hand, previous methods have adjusted for transience by relying on those captured in more than one year to provide direct estimates of survival rate for residents. Here, we extend that approach by supplementing that cohort of known residents with those previously unmarked birds captured twice or more, at least d days apart, within a season. We provide an ad hoc approach to that extension in detail and outline two more formal approaches. That extension increases the precision of survival estimates.