Generation Length: Often Ignored and Sometimes Misunderstood

Traditional mathematical models of population genetics have generally assumed discrete generations, but most species have overlapping generations. It is necessary to estimate generation length with overlapping generations for purposes of comparison with discrete-generation models. In populations with overlapping generations, generation length is the average age of parents at the time their progeny are born. Genetic risks associated with small populations depend on both effective population size and elapsed time in generations. In some cases, populations with larger effective size can lose heterozygosity more quickly than populations with smaller effective size because they have shorter generation length. Therefore, estimates of effective size alone cannot be used to predict the loss of heterozygosity over calendar time. Generation length is also required to estimate effective population size from the rate of loss of heterozygosity, to estimate the time of divergence between species using molecular data, and to use coalescent models for historical demographic analyses. Generation length is often not carefully defined or estimated in these cases. In some papers, authors have used the generation length only for females and have ignored males. A review of the literature indicates that the generation length for males and females can be substantially different in some species. In addition, a variety of proxies have been used in the literature for generation length. Many of these proxies do not provide reliable estimates of generation length. Accurate estimates of generation length require detailed life-history information that is unfortunately not available for many species. Analyses using these proxies for generation length should be treated with some skepticism. Most population genetics textbooks do not define generation length. It is hard to understand why such an important parameter in population genetics has been so ignored.