Land-sea ecological connectivity during a Jurassic warming event

Knowledge on how climate change affects land-sea ecological connectivity in deep time is scarce. To fill this knowledge gap we have assembled a unique dataset through a Jurassic (early Toarcian) warming event that includes quantitative abundance data from pollen and spores, organic-walled marine plankton and benthic macro-invertebrates, in association with geochemical data derived from the same sampled horizons, from the Cleveland Basin, UK. Using this dataset we: (i) reconstruct the timing of degradation and recovery of land-plants, marine primary producers and benthic fauna in response to this event, and (ii) test for connectivity between changes in land and marine ecosystems. We find a discrepancy between the timing of the response of land-plant and marine ecosystems to the event. Land-plants were the first to be affected by initial warming, but also recovered relatively quickly after the peak of warmth to return to pre-event levels of richness and diversity. Plankton and benthic fauna instead experienced a delayed response to initial warming, but as warming peaked, they suffered a rapid and extreme turnover. Recovery in the shelf sea was also delayed (particularly for the benthos) compared to the vegetation. Ecological connectivity analyses show a strong link between changes in terrestrial and marine ecosystems. The loss of large trees on land contributed to changes in marine plankton, from dinoflagellate-to prasinophyte algal-dominated communities, by enhancing erosion, runoff and nutrient-supply into shallow seas. Eutrophication and changes in primary productivity contributed to the decrease of dissolved oxygen in the water column and in bottom waters, which in turn affected benthic communities. Such cause-effect mechanisms observed in the Cleveland Basin are likely to have occurred in other basins of the Boreal Realm, and in part also in basins of the Sub-Boreal and Tethyan realms. Although palaeolatitudinal and palaeoceanographic gradients may have controlle
This research was funded by a Swedish Research Council grant to S.M.S. [grant no. 2019-04524] and to V.V. [grant no. 2019-04061], and a Natu-ral Environment Research Council (NERC) grant to R.J.T. [grant no. NE/I005641/1].