A survey of palaeontological sampling biases in fishes based on the Phanerozoic record of Great Britain

Fishes represent more than half of all living vertebrate species, but patterns of fish diversity remain little explored in the fossil record. A compendium of fossil occurrences from Great Britain was assembled in order to address a series of questions concerning the palaeontological record of fishes. There are broad similarities between British richness trajectories and those compiled from global data, including an initial peak in the mid-Palaeozoic (Devonian or Carboniferous, depending on the compilation), with a late Palaeozoic trough followed by a sharp rise in diversity in the Late Cretaceous and Paleogene. The British dataset is too small to reveal any significant differences in richness between time bins using subsampling, but a modelling approach based on sampling and geological proxies consistently shows lower-than-predicted richness in the Silurian–Devonian and higher-than-predicted richness in the Late Cretaceous and Eocene. This positive excursion is robust to the exclusion of data from the early Eocene London Clay Lagerstätte. Chondrichthyans (sharks, rays, and ratfishes) and osteichthyans (ray-finned and lobe-finned fishes) show contrasting relationships with geological and sampling proxies, possibly reflecting different taphonomic profiles or idiosyncratic variation in the relative proportion of freshwater and marine deposits over the British Phanerozoic.

► Inferred patterns of fish palaeobiodiversity have changed substantially in the last 170 years.
► The most pronounced feature of fish diversity curves is a Cretaceous–Recent rise.
► The British record suggests this rise cannot be explained by sampling.
► The fish record is not as strongly Lagerstätte-driven as marine tetrapods.
► Unlike most vertebrates, diversity patterns in bony fishes cannot be explained by sampling.