Analysis of Capture-Recapture Data
Chapman and Hall/CRC – 2014 – 314 pages
An important first step in studying the demography of wild animals is to identify the animals uniquely through applying markings, such as rings, tags, and bands. Once the animals are encountered again, researchers can study different forms of capture-recapture data to estimate features, such as the mortality and size of the populations. Capture-recapture methods are also used in other areas, including epidemiology and sociology.
With an emphasis on ecology, Analysis of Capture-Recapture Data covers many modern developments of capture-recapture and related models and methods and places them in the historical context of research from the past 100 years. The book presents both classical and Bayesian methods.
A range of real data sets motivates and illustrates the material and many examples illustrate biometry and applied statistics at work. In particular, the authors demonstrate several of the modeling approaches using one substantial data set from a population of great cormorants. The book also discusses which computer programs to use for implementing the models and contains 130 exercises that extend the main material. The data sets, computer programs, and other ancillaries are available at www.capturerecapture.co.uk.
The book is accessible to advanced undergraduate and higher-level students, quantitative ecologists, and statisticians. It helps readers understand model formulation and applications, including the technicalities of model diagnostics and checking.
"Analysis of Capture-Recapture Data is an invaluable companion to the modern theory and practice of capture-recapture modelling. It is a text with multifaceted appeal, ranging in coverage from traditional models to cutting-edge developments, and flowing effortlessly from practical model-fitting advice to advanced technical topics such as parameter redundancy. It is presented throughout in a concise, accessible style that strikes an impeccable balance between illumination of concepts and succinct mathematical detail.
This book is a must-have for all statisticians working with ecological data and is also suitable for ecologists with a mild quantitative bent or as a course companion for students from senior undergraduate years onwards. The text can be used either as a dip-in reference or as a cover-to-cover read. Anyone who completes the latter can feel confident that they are up to date with everything that matters in this vibrant and expanding field."
—Rachel Fewster, Associate Professor, University of Auckland, New Zealand
History and motivation
Introduction to the Cormorant data set
Modelling population dynamics
Model fitting, averaging, and comparison
Estimating the size of closed populations
The Schnabel census
Analysis of Schnabel census data
Accounting for unobserved heterogeneity
Spuriously large estimates, penalized likelihood and elicited priors
Medical and social applications
Testing for closure-mixture estimators
Spatial capture-recapture models
Survival modeling: single-site models
Combining separate mark-recapture and recovery data sets
Joint recapture-recovery models
Survival modeling: multi-site models
Multi-site joint recapture-recovery models
Multi-state models as a unified framework
Extensions to multi-state models
Model selection for multi-site models
The two-parameter occupancy model
Moving from species to individual: abundance-induced heterogeneity
Accounting for spatial information
Covariates and random effects
Use of P-splines
Simultaneous estimation of survival and abundance
Estimating abundance in open populations
Diagnostic goodness-of-fit tests
Absolute goodness-of-fit tests
Using symbolic computation
Parameter redundancy and identifiability
Decomposing the derivative matrix of full rank models
The moderating effect of data
Exhaustive summaries and model taxonomies
Fitting linear Gaussian models
Models which are not linear Gaussian
Bayesian methods for state-space models
Formulation of capture-re-encounter models
Formulation of occupancy models
Integrated population modeling
Normal approximations of component likelihoods
Goodness of fit for integrated population modelling: calibrated simulation
Hierarchical modelling to allow for dependence of data sets
Appendix: Distributions reference
Summary, Further reading, and Exercises appear at the end of each chapter.