Figure 2

Schematic representation of the workflow of the study. We compiled the fossil record (purple colour) and molecular data (orange colour) at the species level for Carcharhiniformes that were analysed separately and jointly (blue colour). The fossil occurrence data are first analysed under a birth–death with constrained shifts (BDCS-Fossils) model, using PyRate, while jointly modelling the preservation process to estimate speciation (λ), extinction (μ), and preservation (q) rates. This also estimates the times of speciation (Ts) and times of extinction (Te) of all species with a fossil record. The DNA sequences were analysed with Bayesian phylogenetic inferences, using BEAST, while simultaneously estimating the tree topology and divergence times using fossil calibrations retrieved from the fossil compilation. The divergence times provide Ts for extant species that have no fossil record (see Fig. 3 for details on this step). We then combined the Ts and Te estimated with PyRate and the Ts and Te inferred with molecular dating in BEAST to generate the most comprehensive species-level dataset incorporating Ts and Te as well as the age uncertainties. This dataset is used for the subsequent analyses under the BDCS model (BDCS-Combined) to compare diversification dynamics with the fossil-only estimates, and the multivariate birth–death (MBD) model to assess the role of global environmental change, as well as clade competition on speciation and extinction rates.