Divergence time and population size in the lineage leading to modern humans

Theor Popul Biol. 1995 Oct;48(2):198-221. doi: 10.1006/tpbi.1995.1026.


We have developed maximum likelihood (ML) methods for comparisons of nucleotide sequences from unlinked genomic regions. In the case of a single species, the ML method primarily estimates the effective population size (Ne) under both constant size and abrupt expansion conditions. In the case of two or three species, the ML method simultaneously estimates the species divergence time and the effective size of ancestral populations. This allows us to trace the evolutionary history of the human population over the past several million years (my). Available sequences at human autosomal loci indicate Ne = 10,000 in the Late Pleistocene, a figure concordant with the results obtained from mitochondrial DNA sequence and allele-frequency data analysis, and there is no indication of population expansion. The ML analysis of two species shows that humans diverged from chimpanzees 4.6 my ago and that the human and chimpanzee clade diverged from the gorilla 7.2 my ago. Furthermore, the effective population size of humans more than 4.6 my ago is nearly 10 times larger than Ne of modern humans. The effective population size in the human lineage does not seem to have remained constant over the past several million years. The ML model for three species predicts slightly different, but consistent results to those obtained by the two-species analysis.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Animals
  • Base Sequence
  • Evolution, Molecular*
  • Gene Frequency
  • Genetics, Population*
  • Gorilla gorilla / genetics
  • Hominidae / genetics*
  • Humans
  • Likelihood Functions
  • Models, Genetic
  • Pan troglodytes / genetics
  • Phylogeny
  • Population Density*
  • Predictive Value of Tests
  • Species Specificity
  • Time Factors