Abstract
In using a population projection model to help manage the conservation of along-lived species, we usually need a reliable estimate of adult survival. Mark-recapture studies are often used to estimate survival, and typically require the assumption that there is no permanent emigration from the study area. We consider how such a study might be extended to allow for local emigration, that is, movement of individuals into an area peripheral to the study area. In particular, we focus on the question as to how much field effort is required in this peripheral area in order to obtain sufficiently precise estimates of both the survival probability and the probability of local emigration. We consider the use of multi-state, mark-recapture models as a means of providing these estimates and show how to assess the precision of a potential study design by calculating the expected confidence limits associated with the resulting estimates. We considera range of design scenarios for the situation that motivated this work, involving a population of Hector’s dolphins in New Zealand. For this case, it appears that there is little gain in precision once a capture probability of 0.4 is reached in the peripheral area.