EVIDENCE FOR CRETACEOUS TROPICAL WARMTH FROM DATA

A compilation of several hundred Cretaceous through Eocene oxygen isotope ratios from open ocean sites between 60° south and 30° north latitude show a systematic, statistically significant difference between values from clay-hosted planktonic foraminifera and those from chalk-hosted foraminifera from similar time intervals at the same sites. Clay-hosted planktonics exhibit significantly lower d¹⁸O values than chalk-hosted specimens, supporting the hypothesis that diagenetic alteration toward higher d¹⁸O values may be most common in chalks. To estimate Cretaceous upper ocean temperatures, we have used clay-hosted foraminifera with glassy shells similar to those of foraminifera caught in plankton tows. In some cases, the excellent preservation of biogenic carbonate is supported by the presence in the same strata of ammonites with aragonitic shells. Oxygen isotopic analysis of a variety of near-surface dwelling species from closely sampled sequences yield d¹⁸O values of ~-3.5 to -4 in the Albian and early Cenomanian (subtropical ODP Sites 1050 and 1052), -3.9 to -4 in the late Cenomanian (tropical DSDP 144), -4.2 to -4.5 in the late Turonian (tropical DSDP 144), and -3.9 to -4.6 in the late Turonian (subpolar DSDP 511). Assuming equilibrium fractionation and a wide range of plausible water d¹⁸O values, almost all the data (including subpolar Site 511) indicate upper ocean temperatures equal to or, in most cases, higher than modern tropical sea surface temperatures. This is true even if it is assumed that Cretaceous upper ocean d¹⁸O everywhere was closer to the global mean ocean value (taken to be -1 to -1.25 per mil SMOW) than it is today, as is suggested by Cretaceous isotope tracer model experiments. The apparent warm Cretaceous equatorial temperatures from Demerara Rise (ODP 144) are especially remarkable if we consider that planktonic foraminifera live in the shallow subsurface and that waters in that region today cool rapidly below the surface because of upwelling.