Deciphering latitudinal shifts in coccolith accumulation in the eastern tropical Pacific Ocean through the Pleistocene

Abstract

Abstract Middle and Late Pleistocene coccolithophore assemblages from Ocean Drilling Program (ODP) sites 1241 and 1242 in the northeastern tropical Pacific (north ETP) were analyzed to reconstruct coccolithophore-productivity and surface-water conditions over the last 925 kyr. Stratigraphic control was provided by δ18O data and geomagnetic time scales and the identification of synchronous calcareous nannofossil events, formerly described in mid and low latitudes. Changes in sea surface coccolithophore-productivity were inferred by estimates of the so-called N ratio – an independent proxy on conversion to accumulation rates – and the total coccolith-derived carbonate accumulation rate (CAR). CAR variations follow those of the N ratio but show larger amplitude changes. Fluctuations in the N ratio and total CAR values imply coccolithophore-productivity maxima during glacial times, most likely as a result of a shallower nutri-thermocline, defined as the boundary beneath the sea surface at which nutrient-rich/cool and nutrient-poor/warm waters are separated. Low interglacial values of the N ratio and CAR suggest lower coccolithophore-productivity and a deeper nutri-thermocline (deep stratification of the mixed layer). This result implies that coccolithophore-productivity-controlled CARs and dissolution were driven by the ratio between CaCO3 production and organic matter oxidation, and that effects of dissolution were enhanced when coccolith production fell. The data suggest high coccolithophore-productivity and a shallow nutri-thermocline during MIS 22 through MIS 6, likely due to the reinforcement of northeastern wind-driven advection of nutrient-rich waters, which is compatible with a La Nina-like state. The dominance of Gephyrocapsa caribbeanica during the Mid-Brunhes Event (MBE) is consistent with this eutrophic environment. Conversely, data from MIS 5 suggests a weakening of trade winds in the ETP, which would have reduced the normal upwelling of cool water and warmed SST, and which is compatible with the oligotrophic setting of an El Nino-like state. During MIS 4 through MIS 1, coccolithophore-productivity was not significantly different from MIS 5 and therefore not indicative of a prominent La Nina-like state, as previously reported for the Equatorial Cold Tongue (ECT). Comparison with coccolithophore data from Site 1240 supports a reduction of coccolithophore-productivity and nutri-thermocline deepening from the ECT to the Costa Rica margin and suggests that nutrient availability was the primary control on the distribution of the whole assemblage during the past 925 kyr.