simple physical arguments involving changes in evaporative cooling with warming (Newell et al., 1978; Newell and Dopplick, 1979) and questions on the mechanisms of tropical temperature change (e.g., Horrell, 1990) have also been utilized to support the notion of tropical temperature stability. However, a number of more comprehensive model experiments suggest that variations within a limited, but significant, range cannot be ruled out (Washington and Meehl, 1984; Manabe and Bryan, 1985; Hansen et al., 1988; Schlesinger, 1989). In combination, the isotopic data and the model studies support the hypothesis that past climate changes should have had a substantial impact on the character and distribution of life within the tropics. The climate stability of the tropics and its implications for the distribution and character of life are addressed here by (1) consideration of the oxygen isotopic records of low latitude temperature variations; (2) discussion of the physical arguments for temperature stability within the tropics; (3) examination of climate model-derived tropical temperatures; (4) examination of model evidence for tropical salinity differences between different time periods in Earth history; (5) consideration of the climate tolerances of tropical organisms; and (6) consideration of a mid-Cretaceous case study in which simulated climate changes in the tropics can be compared with the biological record.
The primary conclusions are that (1) throughout Earth history there has been significant variation in tropical temperature (3 to 5°C differences from the present day) and salinity (several parts per thousand); (2) these variations are large enough to have substantial impact on life; and (3) greater study of the geologic record within the tropics will yield important insights into climate sensitivity and into the biologic response to global change.
The oxygen isotope method of determining paleotemperatures has been widely utilized to study the Cenozoic and the Cretaceous (Savin, 1977). These paleotemperature determinations for the tropics suggest substantial variation.
Isotopic measurements on apparently unaltered planktonic foraminifera from the Shatsky Rise, near the equator during the mid-Cretaceous, yield temperature values of 25 to 27°C (Douglas and Savin, 1975), if an ice-free Earth is assumed. These values can be taken at face value and used to indicate little change in tropical temperatures or slightly lower temperatures than are present (e.g., Horrell, 1990). However, several factors (regional variations, habitat, and selective preservation) must be considered in interpreting isotopic measurements on planktonic foraminifera. First, the isotopically lightest measurement (27°C) is likely to represent the shallowest dwelling foraminifera. Even present day shallow-dwelling foraminifera give isotopic temperatures that are 3 to 5°C cooler than the surface. Further, selective dissolution of the more fragile, shallow-dwelling forms tends to bias estimates in the cold direction (Savin et al., 1975). Consequently, a reasonable interpretation of the isotopic data within the Cretaceous tropics is surface temperatures of 27 to 32°C. The range of possible interpretation is from similar to the present day (28°C) to several degrees higher than at present (Figure 6.1).
Pre-Pleistocene Cenozoic isotopic temperatures are also substantially different from the present day. Shackleton (1984) presents data yielding isotopic paleotemperatures as low as 18°C for the low latitude Pacific from the Maastrichtian to the Late Miocene. Values similar to the present day occurred only in the late Neogene. Early Eocene and Early Miocene values represent tropical ocean sea-surface temperature minima in the Shackleton (1984) analysis. The Early Eocene low-temperature values have received particular attention (Shackleton and Boersma, 1981). Recent synthesis and interpretation of these and other isotopic data (Sloan, 1990) suggest that at a maximum, Early Eocene tropical surface temperatures were about 24°C, about 3 to 5°C lower than present values. Analysis of tropical sea-surface temperatures during the last glacial maximum also contributes to the notion of tropical temperature variation. Early estimates of tropical sea-surface temperatures from oxygen isotopes for ice age