Late Pleistocene–Holocene palynology and paleoceanography of İmralı Basin, Marmara Sea: Pollen-spore, dinoflagellate cyst and other NPP zonations

Abstract

The Marmara Sea is the oceanographic gateway that links the hypersaline Eastern Mediterranean Sea with the hyposaline Black Sea via Straits of Dardanelles and Bosphorus, respectively, and also occupies a transitional vegetation and climatic regime between the Mediterranean and Euxinic phytogeographic ecosystems. Recent palynological studies of long cores from the shallow İmralı Basin perched on the southeastern slope of the Marmara Sea reveal that major paleoecological changes accompanied the isolation and reconnection of the Marmara Sea during the last two Pleistocene glacial–interglacial cycles. Here, we use data primarily from two İmralı Basin cores to determine new quantitative zonations for the pollen-spore, dinoflagellate cyst (dinocyst) and other non-pollen palynomorph (NPP) assemblages in sediments representing the last ∼30 cal ka. A gap in one core is closed using data from three short gravity cores. Twenty-four radiocarbon ages and 20 correlation-ties for sapropel and ash layers from these five cores provide an essentially complete record for the past 30 cal ka. Use of a full suite of palynomorph data (pollen and spores, fungal spores, dinocysts and other phytoplankton, aquatic parasites/saprophytes, zoo-planktonic and zoo-benthic NPPs) shows for the first time that fluctuations in aquatic palynoassemblages during the glacial periods are more complex than simple salinity or temperature responses. The changes involve food web-level turnovers in trait- and size-based food web components in which potential parasites/saprophytes appear critically important.Pollen stratigraphies reveal a significant loss of Euxinic tree taxa by MIS 3c, with only partial recovery in MIS 1. The Euxinic forest elements are replaced by cultivars and other indicators of anthropogenic influences during the MIS 1 that is also marked by a dry interval and corresponding in part with the 4.2 cal ka Eastern Mediterranean Drought Event. The dinocyst zones for the last ∼30 cal ka show a clear turnover from assemblages dominated by brackish Ponto–Caspian taxa to marine assemblages at ∼11.7 cal ka. This phytoplanktonic signal lags slightly behind (a) the vegetation and shift in the other NPP commencing ∼13.2 cal ka, the initial marine incursion at 13.8 cal ka, and the arrival of coccolithophores at ∼13 cal ka. Comparison to published MIS 6 and 5 data reveals higher MIS 2–1 species diversity, reappearance of the MIS 6 marker Impagidinium inaequalis, and the emergences of Pyxidinopsis psilata f. quadrata, cf. Boreadinium breve, and Peridinium spp. of Brenner 2005. The pre-reconnection microbiota is dominated by the unicellular alga Sigmopollis, spikes of colonial Pediastrum algae, the saprophytes/parasites cf. Multiplicisphaeridium and Chytridiomycota. This microbiota is displaced by the post-reconnection assemblage dominated by Botryococcus coenobia, various Platyhelminth egg capsules and ostracod lining-types that persist through the Holocene. A distinctive transitional microbiota with Discocotyle-type turbellarian egg capsules marks the Bølling–Allerød warming, followed at ∼13.3 ka by a marine assemblage with micro-foraminiferal linings and crustacean eggs. The interglacial marine phase of MIS 1 in İmralı Basin has much higher dinocyst species diversity than MIS 5e, reflecting Neolithic impacts (agriculture, land clearance) and expanded shipping trade. Initial marine flooding in the Marmara Sea during both interglacials are marked by sapropel deposition but variations of ravinement surface erosion, and marine deep water residence time in a stratified Marmara Sea are potentially important factors underlying hypoxia/anoxia and sapropel deposition.