I. Introduction. Understanding evolutionary processes at one level of organization often requires addressing processes at higher or lower levels of organizations. For example, questions at the family (kinship) level may raise questions about genetic differences among populations. Even when a particular question does not require answers from other levels, it may raise intriguing possibilities about consequences for processes at those other levels. Given limited resources available for tackling problems, evolutionary biologists should benefit from tools that allow one to address problems at several levels using a single data source and a single technical tool.

DNA microsatellites are genetic markers that can be useful in addressing questions at a variety of scales, ranging from the extremely fine-grained to the fairly coarse-grained. More specifically, this genetic tool can help solve problems ranging from individual-specific, such as determining gender (Longmire et al., 1993; Delehanty, 1995) to questions of relatedness and parentage (Amos et al., 1993; McDonald and Potts, 1994; Kellogg et al., 1995; Primmer et al., 1995), the genetic structure of populations (Bowcock et al., 1994; Taylor et al., 1994; Dallas et al., 1995; Estoup et al., 1995; Paetkau and Strobeck, 1995; Gibbs et al., 1996) and up to comparisons among species (Roy et al., 1994b; Garza et al., 1995) Further, it has several technical and analytical advantages that make it superior to genetic markers whose domains are far smaller. It therefore comes close to meriting the rubric "master of all trades". We present an overview of the technique, our assessment of the sorts of problems to which it is well-suited, and provide examples from the literature and our own research that exemplify some of the scales at which microsatellites can provide useful answers. Because several reviews of the technique already exist (Bruford and Wayne, 1993; Schlotterer and Pemberton, 1994; Westneat and Webster, 1994), we will strive to cover the essentials without unnecessary overlap. A recent review of molecular techniques in zoology (Fleischer, 1996) provides an overview of the role of other markers such as minisatellites and mitochondrial DNA, as well as microsatellites.

We also caution against application of inappropriate models to the data. We will stress that careful attention to model assumptions is required; microsatellite data sets may not always meet model assumptions concerning, for example, the balance between drift, mutation and migration. We provide case histories to illustrate some of the potential pitfalls. We hope to stimulate interest in the development of more sophisticated models and empirical tests of assumptions, so that the analytical techniques become maximally consistent with actual patterns observed in natural populations of birds.