Resilience of Pacific pelagic fish across the Cretaceous/Palaeogene mass extinction

Open-ocean ecosystems experienced profound disruptions to biodiversity and ecological structure during the Cretaceous/Palaeogene mass extinction about 66 million years ago1–3. It has been suggested that during this mass extinction, a collapse of phytoplankton production rippled up the food chain, causing the wholesale loss of consumers and top predators3–5. Pelagic fish represent a key trophic link between primary producers and top predators, and changes in their abundance provide a means to examine trophic relationships during extinctions. Here we analyse accumulation rates of microscopic fish teeth and shark dermal scales (ichthyoliths) in sediments from the Pacific Ocean and Tethys Sea across the Cretaceous/Palaeogene extinction to reconstruct fish abundance. We find geographic di�erences in post-disaster ecosystems. In the Tethys Sea, fish abundance fell abruptly at the Cretaceous/Palaeogene boundary and remained depressed for at least 3 million years. In contrast, fish abundance in the Pacific Ocean remained at or above pre-boundary levels for at least four million years following the mass extinction, despite marked extinctions in primary producers and other zooplankton consumers in this region.We suggest that the mass extinction did not produce a uniformly dead ocean or microbially dominated system. Instead, primary production, at least regionally, supported ecosystems with mid-trophic-level abundances similar to or above those of the Late Cretaceous. TheCretaceous/Palaeogene (K/Pg) event precipitated an 80–95% species-level extinction of calcareous nannoplankton (primary producers) and planktonic foraminifera (primary consumers), decimating part of the base of the open-ocean food web. This loss of productivity is thought to have driven extinction at higher trophic levels2. For instance, ⇠34% extinction at the genus level has been inferred for sharks and rays, with the highest losses among coastal and surface ocean groups6. The K/Pg event also produced a major shift in coastal bony fish functional diversity with particularly large losses among predatory fishes with ecologies similar to modern tuna, billfish and jacks7,8. Complete extinction of mosasaurs, plesiosaurs and ammonites further suggests that the extinction reverberated to the top of the food web3,9. Whereas the response of well-fossilized plankton and megafauna to the K/Pg mass extinction has been well studied1,6,10,11, the ecological and evolutionary response of the trophic link between the two groups, the mid-level consumers such as small-bodied fishes, is relatively unknown (as discussed in refs 7,8). Ichthyoliths have an excellent, but underappreciated, fossil record that spans the K/Pg boundary in the deep ocean12. Teeth are typically small (most abundant in the <150 μm sieve fraction) and so are likely to represent small pelagic species or juveniles, whereas rarer denticles may come from sharks with a range of body sizes. Unlike most microfossils, ichthyoliths are composed of calcium phosphate, which is highly resistant to dissolution13. Ichthyoliths are thus found in nearly all sediment types, including pelagic red clays13. An analysis of stratigraphic ranges of teeth in Pacific red clay suggests that the K/Pg extinction of tooth morphotypes was slight in contrast to the marked extinction of top pelagic predators6,7. A stage-level biostratigraphic compilation of ichthyolith morphological diversity throughout the Pacific Ocean shows extinction of only 5 of 42 morphotypes between the Late Cretaceous and the early Palaeocene (a ⇠12% loss; ref. 13). The low level of extinction of tooth morphotypes suggests that few basic trophic groups of fishes were lost among small pelagic taxa. However, these data indicate little about the magnitude of loss of fish taxa at the boundary, because the samples represent several million years of time-averaging. In addition, in modern fishes tooth shape can evolve rapidly among closely related species, and convergence is common in fishes exploiting similar prey14. We produced high-resolution time series of pelagic fish tooth abundance, in the North Pacific (Ocean Drilling Program (ODP) Site 886), Central Pacific (ODP Site 1209, Shatsky Rise), South Pacific (Deep-Sea Drilling Program (DSDP) Site 596) and the Tethys Sea (Bottaccione Gorge, Gubbio, Italy; Fig. 1). The absolute abundance of fish tooth remains is presented as an ichthyolith mass accumulation rate (MAR). Ichthyolith MAR accounts for changes in the sedimentation rate and density of deep-sea sediments, and provides an approximation for the relative abundance of pelagic fish in the overlying water column (Methods and Supplementary Information and Supplementary Figs 1–15). Our data sets use slightly di erent timescale andMARmetrics, based on the material and lithology of the site (Methods and Supplementary Information and Supplementary Figs 1–15). As a result, although absolute ichthyolith abundances are not equivalent across sites, the patterns and trends are comparable (Fig. 2). Ichthyolith accumulation from the South Pacific Ocean (DSDP Site 596, Fig. 2c) increases across the boundary from an average of 41.8 ichthyoliths cm 2 Myr 1 in the last one million years of the Maastrichtian to 59.6 ichthyoliths cm 2 Myr 1 in the first million years of the Danian (two-sample t-test, P = 0.03; counts of ichthyoliths in the >106 μm fraction). The age model is based on cobalt accumulation rate and strontium isotope chronologies calibrated to the K/Pg boundary, which is placed at a prominent iridium anomaly and impact debris horizon15,16. Teeth are preserved in red clay with a sedimentation rate of ⇠0.25mMyr 1, so it is possible that a brief decline in fish abundance could be masked