Nine Examples T R E N T L . M C D O N A L D , S T E V E N C . A M S T R U P, E R I C V. R E G E H R , A N D B R YA N F. J . M A N LY 9.1 Introduction In this chapter we provide empirical examples of how to use the models described in earlier chapters to analyze real-world data sets. With num- bers rather than symbolic notation, we illustrate some practical aspects of capture­recapture modeling. We provide explicit examples and in- structions on the mechanics of model building, and illustrate how to set up analyses in program MARK. We hope that the examples provided in this chapter will reduce the anxiety that often accompanies attempts at new analytical approaches. Much of this chapter is built around capture­recapture data collected on the European dipper (Cinclus cinclus) (Marzolin 1988). This data set is typical of many capture­recapture studies, and has been analyzed a num- ber of times in the literature (e.g., Lebreton et al. 1992). The dipper data provide a convenient framework for the application of several of the mod- els covered in previous chapters of this book. We start by illustrating how to compute two early and relatively simple capture­recapture models for open populations (Jolly-Seber and Manly-Parr, chapter 3). We then work through a series of more complicated Cormack-Jolly-Seber (chapter 5) analyses of the dipper data using program MARK. We also present an al- ternative parameterization of the CJS model, in which explanatory covari- ates are framed in a general regression context. This general regression approach allows capture­recapture data to be analyzed in a familiar and intuitive framework, facilitates complicated covariate structures, and may be more easily understood by newcomers to capture­ recapture analysis. Next, we treat the dipper data as if they were collected from a closed pop- ulation and perform one of the recent approaches to closed-population assessment (i.e., the Huggins model, chapter 4). We then explain how to as- sess the goodness of fit of a CJS model to the dipper data with parametric bootstrapping. We also estimate the number of dippers using the Horvitz- Thompson population estimator. Finally, we manipulate the dipper data in a way that allows us to illustrate a simple multistate analysis (chapter 7).