Estimation of population size with molecular genetic data

"Population size is a central parameter in ecology, evolutionary biology and conservation. Until the advent of molecular genetic methods, population size was measured through observation and/or capture of individuals. Now, molecular data are frequently used to estimate size. There are several definitions of population size, which differ mainly with respect to the temporal, spatial and genealogical scale of reference. Those interested in the number of individuals usually study the census size, whereas those primarily interested in the genetic consequences of population size generally study the effective size. The various definitions of effective size are largely defined in reference to a Wright/ Fisher or 'ideal' population. We provide an overview of the different definitions of population size and of the methods to estimate them using molecular genetic data, with an emphasis on recently developed methods for estimating effective size. These methods can use genetic samples from single or multiple points in time. They can also estimate effective size in a defined interval or an indeterminate one that ranges from several generations to the entire coalescent history of the genetic sample. We also discuss methods for evaluating changes in size, both long-term changes and recent reductions or bottlenecks. We argue that the scale of reference is crucial in the choice and interpretation of a size estimation method. Specifically, we note that when estimating and reporting a population size investigators should 1) specify precisely the time period and spatial area over which the size is believed to prevail, 2) remember that different genetically-based methods for estimating effective size do apply to specific time periods and time scales, 3) note that none of the three main effective sizes-inbreeding, variance, or eigenvalue-are intrinsically associated with a particular time scale"--Abstract. Eric C. Anderson, John Carlos Garza. "September 2009." Mode of access: World Wide Web. System requirements: Internet browser; Adobe Acrobat Reader. Includes bibliographical references (p. 41-58).