A review of molluscan biodiversity in the Mediterranean Sea during the Last Interglacial

The Last Interglacial (LIG, 129–116 ka) was characterised by global temperatures marginally higher than pre-industrial levels, and by sea level slightly higher than today. Additionally, the positions of the continents, major oceanographic features, and marine biodiversity were also similar to those of the present day. These features combined make the LIG an excellent analogue to predict how global biodiversity responds to forecasted moderate anthropogenic warming. The Mediterranean Sea is a global warming hotspot with an excellent LIG fossil record and thus an ideal setting to establish conceptual and methodological pathways to use the LIG for predictions. We revised the palaeontological record of molluscs, whose fossilization potential is among the highest of marine invertebrates, from the LIG of the Mediterranean Sea (107 sites, 2278 occurrences, 439 species). We document a longitudinal gradient in species richness that resembles the present-day patterns: the eastern Mediterranean during the LIG was less rich than the western and central sub-basins. Also, as it is today, the Strait of Sicily was an important biogeographic divide. To the west of it, species with a cold and temperate affinity from the Boreal and Lusitanian ecoregions were more abundant, while to the east, species with a warmer affinity belonging to the West African and Mediterranean ecoregions, occurred in higher proportions. At the same time, the so-called “warm guests” (WGs)—tropical Atlantic species that expanded their geographic ranges into the Mediterranean during this interval but regressed into the tropical Atlantic belt during the subsequent glaciation—were more prevalent to the east of the Strait of Sicily, reflecting seawater palaeotemperatures. No WGs were reported from the northern areas of the basin, such as the Gulf of Lions and the Adriatic Sea. Nevertheless, geographical coverage remains uneven, the data are mostly qualitative, small-sized species are under-represented, and the taxonomy of some species is still uncertain. These limitations currently hamper the application of quantitative and modelling approaches, while also highlighting promising directions for future research.