ABSTRACT
AbstractGeneration time is a measure of the pace of life and is used to describe processes in population dynamics and evolution. We show that three commonly used mathematical definitions of generation time in age-structured populations can produce different estimates of up to several years for the same set of life history data. We present and prove a mathematical theorem that reveals a general order relation among the definitions. Furthermore, the exact population growth rate at the time of sampling influences estimates of generation time, which calls for attention. For phylogenetic estimates of divergence times between species, included demographic data should be collected when the population growth rate for each species is most common and typical. In conservation biology, demographic data should be collected during phases of population decline in declining species, contrary to common recommendations to use predisturbance data. The results can be used to improve the International Union for Conservation of Nature's recommendation in parameterizing models for evaluating threat categories of threatened species and to avoid underestimating extinction risk.
