Print PageNest survival is often used as a measure of the status of avian populations. It can be indicative of overall health or illustrate differences within a population. Many approaches have been used to characterize nest success, each with limiting assumptions. Two common strategies are the Mayfield method and analysis using Program MARK. The former is widely employed in avian studies but assumes constant nest survival over the entire nesting interval. Program MARK relaxes this biologically unrealistic assumption and facilitates the inclusion of additional covariates to explain variation. More recently, the BUGS (Bayesian analysis Using Gibbs Sampling) language has made complex, hierarchical modeling readily accessible to ecologists. We compare daily survival rates calculated with Program MARK and WinBUGS. We then use random and mixed effects models to explore the variation in nest survival rates for the American Oystercatcher (Haematopus palliatus) in North Carolina.
Results/Conclusions
Preliminary comparison between the two methods (n=494 nests) indicates that WinBUGS survival estimates are lower, both in overall estimate and for individual years. Including year and island as random effects does not appear to improve the intercept model. We expect stronger explanatory power to come from smaller, more organic spatial divisions. Ongoing investigation using WinBUGS includes multiple random effects, lending a hierarchical structure to nest survival estimates.
