TESTING THE GENERATION TIME HYPOTHESIS USING DNA/DNA HYBRIDIZATION BETWEEN ARTIODACTYLS

Heterogeneous rates of molecular change between some mammalian lineages are commonly explained by contrasts in generation time length. Here the generation time hypothesis is tested by comparing the relative rates of molecular change in related artiodactyl taxa differing by their generation time. A demographic model based on allometric relations with the adult body weight is used to estimate the cohort generation time in Bovidae and Cervidae families (Artiodactyla, Mammalia). Two pairs of closely related taxa (two cervids, two bovids) were selected, each showing clear ratios (1.5 to 3.5 times) in their generation time. Rates of genetic change in non-repeated nuclear DNA were estimated by DNA/DNA hybridization experiments performed among these ruminants and a camelid outgroup. Relative rate tests were applied to the two pairs of ingroup taxa differing by their generation time, in order to test if shorter generation time would correspond to higher rate of molecular change. Contradictory statistical results did not show a greater accumulation of nucleotide changes in the lineage leading to the short generation time species. The recorded differences in branch lengths of sister taxa were either conflicting or too small (relative to the contrasted generation times) to reveal a generation time effect. Alternative hypotheses are suggested to explain these preliminary results.