Paleotemperature Records From Alkenones

Certain marine phytoplankton of the class Prymnesiophyceae, most notably the coccolithophorid Emilyania huxleyi, respond to changes in water temperature by altering the molecular composition of their cell membranes. Specifically, as water temperature decreases, they increase the production of unsaturated alkenones (ketones). Cells contain a mixture of long-chain alkenones with 37, 38, or 39 carbon atoms (n-C37 to n-C39), which are either di- or tri-unsaturated (designated for example, as C37.2 and C37.3, respectively). A temperature-dependent unsaturation index U37 is defined as:

where [C37 2] represents the concentration of the di-unsaturated methyl ketone, alkadienone, containing 37 carbon atoms. The index varies from -1 (when all alkenones are C37.4) to +1 (all C37.2). However, as C37.4 is absent in most sediments the index can be simplified to:

so that values are generally positive (-0.2 to 0.98) in Quaternary sediments (Brasell et al., 1986). The importance of these organic biomarkers is that they do not appear to significantly degrade in marine sediments, nor are they influenced by changes in salinity or isotopic composition of the ocean. They thus provide a crucial complement to SlsO and faunal composition studies of marine paleotemperatures, as discussed in what follows.

Studies of the algae E. huxleyi in controlled growth chambers, and of sediments accumulating beneath areas with different SSTs, show a strong signal relating water temperature to U37 (Prahl et al., 1988; Sikes et al., 1991; Rosell-Mele et al., 1995; Miiller et al., 1998). From controlled experiments (Prahl et al., 1988) the Ujy -temperature relationship is:

Others have found that this relationship varies somewhat, but generally the slope is the same over the range of 15-25 °C.

The potential of U37 as an SST paleothermometer is tremendous and is the basis of an emerging new field in paleoclimatology (molecular stratigraphy). A number of studies have now been carried out to reconstruct SSTs from the alkenone record in marine sediments. Almost uniformly, these studies show a smaller temperature difference between Holocene and LGM SSTs than estimates based on faunal composition (using modern analog or transfer functions). However, part of this difference is probably related to the fact that the two approaches are not dealing with precisely the same phenomena. Alkenone-based SSTs are derived from phytoplank-ton that live predominantly in the photic zone, with the bulk of the organisms inhabiting the upper 10 m of the water column. Furthermore, phytoplankton blooms commonly occur rapidly in the spring or early summer and the resulting organic sediment may thus represent a relatively short episode of only a few weeks (Sikes and Keigwin, 1994, 1996). By contrast, the foram-based paleotemperatures are based on a set of different organisms that may reach peak abundances at different times of the year, and live at different depth habitats. Consequently, such estimates are likely to represent a more time and depth-integrated measure of temperature change than that from alkenones. Furthermore, foram assemblage (and S180) data are subject to potential biases due to dissolution effects, whereas alkenone-based paleotemperatures are not (Sikes and Keigwin, 1994). However, in areas with a large annual range of SSTs, any shift in the seasonal timing of maximum phytoplankton productivity could result in a shift in alkenone-based paleotemperatures, even without any real change in oceanographic conditions (Chapman et al., 1996).

By combining alkenone-based paleotemperature reconstructions with other approaches (such as faunal assemblage, or 8lsO-based paleotemperature estimates) important new insights into paleo-oceanographic conditions can be obtained. For example, Zhao et al. (1995) studied sediments from off the northwest coast of Africa and found that paleotemperature minima at different times in the last 80 ka correspond closely to Heinrich events seen in North Atlantic sediments (see Section 6.10.1). These abrupt changes appear to represent times when cold meltwater, produced by ice-rafting events, was transferred southward by the Canary current, caus-

Alkenone Glacial Cycles

FIGURE 6.30 The U^-based temperature reconstruction of SSTs in ODP core 658C from off the northwest coast of Africa (upper panel) compared to percentages of the cold-water foram N. pachyderma (s.) in two cores from the North Atlantic.The strong relationship between Heinrich events and cold water episodes in the North Atlantic with episodes of low SSTs off the African coast suggests a linkage via the cold Canary current carrying cool, low salinity meltwater southward at these times (Zhao et a/, 1995).

FIGURE 6.30 The U^-based temperature reconstruction of SSTs in ODP core 658C from off the northwest coast of Africa (upper panel) compared to percentages of the cold-water foram N. pachyderma (s.) in two cores from the North Atlantic.The strong relationship between Heinrich events and cold water episodes in the North Atlantic with episodes of low SSTs off the African coast suggests a linkage via the cold Canary current carrying cool, low salinity meltwater southward at these times (Zhao et a/, 1995).

ing temperatures to decline by 3-4° C in <100 yr (Fig. 6.30). Meltwater effects were also detected in northeastern Atlantic sediments by Sikes and Keigwin (1996) by comparing both alkenone and 8lsO records. This approach was used to good effect in "backing-out" past changes in salinity from an Indian Ocean sediment record by Ros-tek et al. (1993). They established SSTs using alkenones, then applied that to the 8lsO record to reconstruct paleosalinity. By subtracting the local temperature effect on 8lsO, and knowing the effect of changing ice volume on SlsO, the residual change in 8lsO was interpreted as a record of changing salinity (Fig. 6.31). High salinity (by

Paleoclimate Record Temperature

FIGURE 6.3 I Paleosalinity reconstructed for the site of Indian Ocean core MD 9000963, (South southwest of lndia).This was derived from 8lsO by obtaining paleotemperatures from alkenones.then adjusting the 8lsO record for these changes, plus changes in global ice volume and salinity due to continental ice growth and decay.The residual 8iaO changes are interpreted as a record of paleosalinity. Dotted lines bracket the range of paleosalinity estimates (Rostek et al., 1993).

FIGURE 6.3 I Paleosalinity reconstructed for the site of Indian Ocean core MD 9000963, (South southwest of lndia).This was derived from 8lsO by obtaining paleotemperatures from alkenones.then adjusting the 8lsO record for these changes, plus changes in global ice volume and salinity due to continental ice growth and decay.The residual 8iaO changes are interpreted as a record of paleosalinity. Dotted lines bracket the range of paleosalinity estimates (Rostek et al., 1993).

+ 0.5-l%o) from 160-140 ka and 75-25 ka B.P. resulted from a weaker southwest monsoon (with less rainfall on the subcontinent, and hence less runoff to the Bay of Bengal) and/or a stronger northeastern (counter) monsoon airflow at those times.

Other studies have applied alkenone analysis to high-resolution paleotempera-ture reconstruction of both recent sediments (e.g., to examine ENSO events; Kennedy and Brassell, 1992) as well as to periods of rapid environmental change during glacial Terminations. "Younger Dryas-type" oscillations were found to have occurred during Terminations II and IV, suggesting that similar mechanisms involving rapid reorganization of North Atlantic deepwater formation had ocurred during earlier déglaciation events, as well as the most recent one (Eglinton et al., 1992).

0 0

Post a comment