Résumé
Good estimates of the genetic parameters of natural populations, such as heritability, are essential for both understanding how genetic variation is maintained and estimating a population's evolutionary potential. Longterm studies on birds are especially amenable for calculating such estimates because of the ease with which pedigrees can be inferred. Recent 'animal model' methodology, originally developed by animal breeders to identify animals of high genetic merit, has been applied to natural bird populations of known pedigree. Animal models are more powerful than traditional analyses such as parent-offspring regression because they use all of the available pedigree information simultaneously. In doing so, they can accommodate common phenomena like selection and inbreeding and are especially suitable for the complex and incomplete pedigrees typical of natural populations. Animal models not only provide a better way of estimating genetic and environmental variance components, they also allow individual phenotypes to be separated into their genetic and environmental components. Here we aim to provide the interested ornithologist with an accessible entry into the vast and sometimes daunting quantitative genetics literature and, in particular, into the literature on the animal model. We outline not only the possibilities offered by the animal model for the accurate estimation of genetic parameters in the wild but also associated potential pitfalls and limitations. On the whole, we aim to provide an accessible and up-to-date overview of the rapidly developing and exciting field of evolutionary genetics applied to long-term studies of wild bird populations.