The Atlantic-Pacific Salinity Gradient; A Variable?

Estimates of Glacial salinity change in the tropical-subtropical Atlantic are derived by measuring the δ18O of the surface-dwelling planktonic foraminifer Globigerinoides ruber (white) taken from a suite of box cores collected between 30° North and 30° South and solving the calcite δ18O paleotemperature equation for δ18O of seawater (δw), which varies systematically with salinity. Mg/Ca paleothermometry is used to derive SSTs used to solve the δ18O paleotemperature. The δw values are converted to salinity via the modern salinity-δ18O relationship. At sites where the water depth is less than 3600 meters the δ18O -Mg/Ca technique applied to G. ruber precisely estimates summer (not mean annual) sea surface temperatures and salinities in the Atlantic. In the late glacial the tropical and subtropical SSTs were an average of 2.5°C colder than modern and late Holocene values. The difference between late Holocene and LGM δw values is close to 1‰ at tropical sites located near the equator and therefore comparable to the ice volume effect. At subtropical sites the difference is much less than 1‰, varying from 0.6 to 0.0‰ in both the northern and southern hemisphere. The Holocene-LGM comparison implies that surface salinities in the subtropical gyres of both the north and south Atlantic were lower during the LGM than during the late Holocene. These results highlight a potentially important hydrologic feedback on the salinity structure of the Atlantic.