MIXED-LAYER HYDROGRAPHY DURING THE EARLY PLIOCENE

The interval between 4.8 and 3.7 Ma reveals an increase in average SST_Mg/Ca of 1°–1.5°C (Fig. F2B). Comparison with the ¹⁸O_benthic record (Fig. F2D), which indicates no significant trend during this period, shows that this increase in SST_Mg/Ca cannot be correlated to a global warming trend. Therefore, the increase in SST_Mg/Ca between 4.8 and 3.7 Ma represents regional warming. This suggests increased influence of the warm NECC and the ITCZ, as is also indicated by the ¹⁸O_salinity record. This might have been achieved either by the plate tectonic movement of Site 1241 toward a more northerly position (Mix, Tiedemann, Blum, et al., 2003) or to southward movement of the ITCZ/NECC.

A southward shift of the ITCZ during this period is supported by several other studies (Billups et al., 1999; Cannariato and Ravelo, 1997). On the other hand, it has also been suggested that the ITCZ remained at its position or even moved northward because of increasing heat piracy to the North Atlantic due to the progressive closure of the Panamanian Gateway (Tiedemann et al., 1989; Maier-Reimer et al., 1990; Haug and Tiedemann, 1998). This, then, would have strengthened the EUC, leading to a shallowing of the thermocline, as is also shown by Steph et al. (this volume). A shallowing of the thermocline might have led to a shallowing of the habitat depth of G. sacculifer, because G. sacculifer mainly lives in the mixed layer. A possible imprint on the SST reconstruction would have been a cooling trend because the whole upper water column is affected when the thermocline significantly shallows. However, our results do not show such a trend in the SST_Mg/Ca record of G. sacculifer during this period. Therefore, we conclude that the shallowing of the thermocline between 5.4 and 4.0 Ma had no or only minor effect on average SST in the mixed layer at the location of Site 1241.

The time interval between 4.8 and 3.7 Ma is further characterized by two significant cooling events at ~4.4–4.2 and 4.0 Ma (Fig. F2B). The event at 4.0 Ma can be clearly correlated with the ¹⁸O_benthic record and can be identified as MIS Gi24, representing a global cooling event. The event between 4.4 and 4.2 Ma, however, is not present in the ¹⁸O_benthic record and, therefore, possibly either signifies a regional cooling event or just an oscillation on the long-term trend. Possible explanations for this "event" include increasing southeast trade winds and a resulting northward movement of the ITCZ/NECC or increasing influence of cooler thermocline waters, as suggested by the shallowing thermocline and the progressive closure of the Panamanian Gateway.

Because the ¹⁸O_benthic record reveals no significant trend for the time period from 4.8 to 3.6 Ma (Fig. F2D), variations in the ¹⁸O_G.sacculifer and ¹⁸O_salinity records reflect a large portion of variance that has to be ascribed to changes in regional oceanography (Fig. F2C, F2E). The decrease in ¹⁸O_G.sacculifer from –1.5 to –1.8 (Fig. F2C) is explained by an increase in temperature, as is recorded by the Mg/Ca record (Fig. F2B). The ¹⁸O_salinity fluctuations (Fig. F2E) remain relatively constant during this period, except for the interval from 4.4 to 4.2 Ma, when average ¹⁸O_salinity values decrease to 0.4. This suggests that between 4.4 and 4.2 Ma Site 1241 bathed in a less saline water mass. This time interval corresponds with the main phase in the shallowing of the thermocline and is paralleled by a decrease in SST_Mg/Ca, suggesting increased influence of cooler thermocline waters or an increase in southeast trade winds and an intensified EUC. On the other hand, this would not result in decreasing SSS and SST_Mg/Ca does not show any changes that can be correlated with the general shoaling of the thermocline. Less saline waters are associated with the tropical rainbelt (ITCZ). Increasing southeast trade winds would move the ITCZ toward the north. Therefore, it seems that this temporarily freshening is due to a a decrease in temperatures during this interval, changing the precipitation/evaporation ratio, causing less evaporation and lower saline waters. The synchronicity of this event with the critical threshold in the progressive closure of the Panamanian Gateway during this time might suggest a coupling between these events (Haug et al., 2001).