Molecular loci that fail relative-rate tests are said to be "overdispersed.
" Traditional molecular-clock approaches to estimating divergence times do
not take this into account. In this study, a method was developed to estima
te divergence times using loci that may be overdispersed. The approach was
to replace the traditional Poisson process assumption with a more general s
tationary process assumption. A probability model was developed, and an acc
ompanying computer program was written to find maximum-likelihood estimates
of divergence times under both the Poisson process and the stationary proc
ess assumptions. In simulation, it was shown that confidence intervals unde
r the traditional Poisson assumptions often vastly underestimate the true c
onfidence limits for overdispersed loci. Both models were applied to two da
ta sets: one from land plants, the other from the higher metazoans. In both
cases, the traditional Poisson process model could be rejected with high c
onfidence. Maximum-likelihood analysis of the metazoan data set under the m
ore general stationary process suggested that their radiation occurred well
over a billion years ago, but confidence intervals were extremely wide. It
was also shown that a model consistent with a Cambrian (or nearly Cambrian
) origination of the animal phyla, although significantly less likely than
a much older divergence, fined the data well. It is argued that without an
a priori understanding of the variance in the time between substitutions, m
olecular data sets may be incapable of ever establishing the age of the met
azoan radiation.