Most of the primary productivity in the ocean comes from phytoplankton, and is impacted, among other things, by the amount of nutrients available, as well as by temperature. The Late Miocene and Pliocene were marked by global aridification, linked to the emergence of the large deserts, likely increasing the input of dust and thus nutrients into the ocean. There was also a global decrease in temperature during this period, linked to a decline in atmospheric CO2 concentration. The objective of this study is to quantify the impact of dust and pCO2 levels on primary productivity in the oceans under Late Miocene boundary conditions. New simulations were performed with the coupled ocean-atmosphere model IPSL-CM5A2 and its marine biogeochemistry component PISCES with a Late Miocene paleogeography. Our results show that an increase in dust input produces a quasi-generalized increase in primary productivity, associated with a decrease in nutrient limitation. This increase in productivity also leads to nutrient deficits in some areas. The decrease in pCO2 levels and the associated lower water temperatures lead to a reduction in primary productivity. This decrease is mainly due to a reduction in the supply of nutrients resulting from less intense remineralization. In addition, our results show that change in carbon export resulting from change in dust input and temperature are highly heterogeneous spatially. Simulations combined with sedimentary data suggesting a link between aridification, cooling and the Biogenic Bloom of the Late Miocene and Pliocene.

Modern Ocean is characterized by the formation of deep-water in the North Atlantic (i.e. NADW). This feature has been attributed to the modern geography, in which the Atlantic Ocean is a large basin extending from northern polar latitudes to the Austral Ocean, the latter enabling the connection of the otherwise isolated Atlantic with the Pacific and Indian Oceans. Sedimentary data date the establishment of the NADW between the beginning of the Eocene (∼49 Ma) and the beginning of the Miocene (∼23 Ma). The objective of this study is to quantify the impact of Miocene geography on NADW through new simulations performed with the earth system model IPSL-CM5A2. We specifically focus on the closure of the eastern Tethys seaway (dated between 22 and 14 Ma), which allowed the connection between the Atlantic and Indian Oceans, and on the Greenland ice sheet, whose earliest onset remains open to discussion but for which evidence suggest a possible existence as early as the Eocene. Our results show that the closure of the eastern Tethys seaway does not appear to impact the establishment of NADW, because waters from the Indian Ocean do not reach the NADW formation zone when the seaway is open. Conversely, the existence of an ice sheet over Greenland strengthens the formation of NADW owing to topography induced changes in wind patterns over the North Atlantic, which in turn, results in a larger exchange of water fluxes between the Arctic and the North Atlantic, and in a re-localization of deep-water formation areas.

The Late Miocene Biogenic Bloom (LMBB) is a late Miocene to early Pliocene oceanographic event characterized by high accumulation rates of opal from diatoms and calcite from calcareous nannofossils and planktic foraminifera. This multi-million year event has been recognized in sediment cores from the Pacific, Atlantic and Indian Oceans. The numerous studies discussing the LMBB lead us to believe that this event is omnipresent in all oceans, although this hypothesis need to be tested. Moreover, the origin of this event is still widely discussed. In this study we aim to provide a comprehensive overview of the geographical and temporal aspects of the LMBB by compiling published ocean drilling (DSDP, ODP and IODP) records of sedimentation rates, CaCO\textsubscript{3} and opal and terrigenous accumulation rates that cover the late Miocene and early Pliocene interval. Our data compilation shows that traces of the LMBB are present in many different locations but in a very heterogeneous way, highlighting that the LMBB is not a pervasive event. The compilation in addition shows that the sites where the LMBB is recorded are mainly located in areas with a high productivity regime (i.e. upwelling systems). We suggest that the most likely hypothesis to explain the LMBB is a global increase in upwelling intensity due to an increase in wind strength or an increase in deep water formation, ramping up global thermohaline circulation.