Data from: A decline in molluscan carbonate production driven by the loss of vegetated habitats encoded in the Holocene sedimentary record of the Gulf of Trieste

Carbonate sediments in non-vegetated habitats on the NE Adriatic shelf are dominated by shells of molluscs. However, the rate of carbonate molluscan production prior to the 20th century eutrophication and overfishing on this and other shelves remains unknown because (1) monitoring of ecosystems prior to the 20th century was scarce and (2) ecosystem history inferred from cores is masked by condensation and mixing. Here, based on geochronological dating of four bivalve species, carbonate production during the Holocene is assessed in the Gulf of Trieste, where algal and seagrass habitats underwent a major decline during the 20th century. Assemblages of sand-dwelling Gouldia minima and opportunistic Corbula gibba are time-averaged to > 1,000 years and C. gibba shells are older by >2,000 years than shells of co-occurring G. minima. This age difference is driven by temporally disjunct production of two species coupled with decimeter-scale mixing. Stratigraphic unmixing shows that Corbula gibba declined in abundance during the highstand phase and increased again in the 20th century. In contrast, one of the major contributors to carbonate sands, Gouldia minima, increased in abundance during the highstand phase, but declined to almost zero abundance over the past two centuries. G. minima and herbivorous gastropods associated with macroalgae or seagrasses are abundant in the top-core increments but are rare alive. Although G. minima is not limited to vegetated habitats, it is abundant in such habitats elsewhere in the Mediterranean Sea. This live-dead mismatch reflects the difference between highstand baseline communities (with soft-bottom vegetated zones and hard-bottom Arca beds) and present-day oligophotic communities with organic-loving species. Therefore, the decline in light penetration and the loss of vegetated habitats with high molluscan production traces back to the 19th century. More than 50% of the shells on the seafloor in the Gulf of Trieste reflect inactive production that was sourced by heterozoan carbonate factory in algal or seagrass habitats.Supplement Table 1-core dataAbsolute abundances of molluscs in two cores at Piran 1 and Piran 2, including two surface living assemblages collected by Van Veen grabs at the same sites in 2014 (1 mm mesh size). Complete valves or fragments with umbo preserved were selected from the >1 mm sieve fraction of each increment, identified to species level, and counted. For each species, the higher number of single valves (either right or left) was added to the number of double-valved specimens to get the final count in each increment.Supplement Table 2-Gouldia age dataAmino acid racemization data and calibrated estimates of postmortem age of Gouldia minima collected from the shell bed at Piran 1 and Piran 2.Supplement Table 3-Corbula age dataAmino acid racemization data and calibrated estimates of postmortem age of Corbula gibba and Piran 2.Supplement Table 4-LA dataAbsolute abundances of molluscs in living assemblages collected by Van Veen grabs at 11 stations in 2011.Gouldia age unmixing finalR language source code for analyses of age data and age unmixin