Antarctic Bottom Water

Globigerina eggeri---

Globorotalia tumida Globorotalia menardii~' Globigerinoides ruber~ Sphaeroidinella dehiscens' Pulleniatina obliguiloculata

Globigerina eggeri---

Globorotalia tumida Globorotalia menardii~' Globigerinoides ruber~ Sphaeroidinella dehiscens' Pulleniatina obliguiloculata

0 1000 2000 3000 4000 5000

CORE DEPTH (M)

6000

FIGURE 6.34 Changes In percentage abundance of several diagnostic planktonic foraminifera in equatorial Atlantic core tops with increasing water depths due to differential dissolution.The more corrosive Antarctica Bottom Water dissolves poorly resistant species (such as Globigerinuides ruber) so that more resistant species (such as Globorotalia tumida) increase in relative abundance (Ruddiman and Heezen, 1967).

studies of coccoliths indicate corresponding problems, with structurally solid, cold-water forms preferentially preserved in thanatocoenoses (Berger, 1973a).

Berger (1973b) suggests that the partially dissolved assemblages be designated taphocoenoses, to distinguish them clearly from assemblages more representative of the original biocoenoses. Clearly, paleoclimatic reconstructions based on taphocoenoses require very careful interpretation. This is particularly so if the rate of dissolution has changed over time as suggested by a number of studies (Chen, 1968; Broecker, 1971; Berger, 1971, 1973b, 1977; Thompson and Saito, 1974; Ku and Oba, 1978). There is evidence that dissolution rates increased during interglacial times in the tropical Pacific and Indian Oceans, resulting in the removal of many less resistant species, and a relative concentration of individuals with a cold-water aspect (Wu and Berger, 19 8 9).20 Conversely, in glacial times, dissolution rates were reduced, giving rise to assemblages of both solution-susceptible and solution-resistant forms. In short, glacial-interglacial changes may be characterized by corresponding dissolution cycles in these areas (Berger, 1973b). Such effects would result in erroneous isotopic paleotemperature estimates as interglacial-age samples would have a higher abundance of cold-water individuals compared to glacial-age samples, thereby reducing the apparent glacial-interglacial temperature range (Berger, 1971; Emiliani, 1977; Berger and Killingley, 1977). Similarly, in foraminiferal assemblage studies (Section 6.4) dissolution cycles may result in taphocoenoses, which lead to quite erroneous paleotemperature estimates. Thus, Berger (1971) and Ruddiman (1977a) urge that all carbonate sediments be considered residual, unless easily dissolved material is present.

One interesting, and potentially very important, aspect of dissolution rate changes over time is the presence of a "deglacial preservation spike" or strati-graphic zone representing a peribd when dissolution rates were markedly reduced (Broecker and Broecker, 1974; W. Berger, 1977). There is evidence that at around 14,000 yr B.P. (and during other terminations) there was a significant worldwide drop in the aragonite compensation depth and the lysocline, lasting for only a relatively short period (perhaps <1000 yr). This resulted in enhanced preservation of carbonate fossils at that time and hence a "spike" of well-preserved foraminifera and pteropods in the sedimentary record (Wu et al., 1990). Indeed, the preservation spike is particularly apparent because dissolution rates appear to have been even greater at -12,000 yr B.P., directly following the time of the dissolution minimum (Berger and Killingley, 1977).

Another significant dissolution signal (the "Brunhes dissolution cycle") is seen in sediments from the equatorial Pacific and Indian Oceans (Wu and Berger, 1989). This is clearly shown in Fig. 6.35 in which the difference in oxygen isotope stratigraphies between records from the Ontong Java Plateau are plotted; one record (V28-238) is from 3120 m and the other (V28-239) is from 3490 m water depth. A generally positive difference in SlsO (of ~0.3%o) between these records is expected because the core from the deeper site has been affected by dissolution (making it

20 In the equatorial Atlantic and Gulf of Mexico, dissolution seems to have increased in glacial times (Gardner, 1975; Luz and Shackleton, 1975).

Core V28 238 Image

FIGURE 6.35 Two oxygen isotope records from the Ontong Java Plateau (eastern Equatorial Pacific) (upper diagram) and the difference between them (V28-239 minus V28-238) (lower diagram).The V28-239 record is from 3490 m andV28-238 from 3120 m water depth. Dissolution has affected the deeper record leading to the preservation of forams with a higher lsO content. In addition, a pronounced episode of enhanced dissolution is seen in the lower diagram, from -300-500 kyr B.P. (Wu and Berger, 1989).

FIGURE 6.35 Two oxygen isotope records from the Ontong Java Plateau (eastern Equatorial Pacific) (upper diagram) and the difference between them (V28-239 minus V28-238) (lower diagram).The V28-239 record is from 3490 m andV28-238 from 3120 m water depth. Dissolution has affected the deeper record leading to the preservation of forams with a higher lsO content. In addition, a pronounced episode of enhanced dissolution is seen in the lower diagram, from -300-500 kyr B.P. (Wu and Berger, 1989).

isotopically heavier) and this effect is enhanced during the interglacials. However, from -300-500 ka B.P. the dissolution effect is systematically greater, indicating some persistent influence on carbonate dissolution. This has also been seen in Indian Ocean sediments (Peterson and Prell, 1985) but the reason is not, as yet, fully understood.

The cause of these rapid shifts in compensation depth is not clear and, in fact, may result from a combination of many factors. Redeposition of carbonate from the continental shelves as sea level rose could have increased oceanic alkalinity and thereby reduced dissolution (W. Berger, 1977). However, as the sea level rose a low salinity upper water layer may have formed (from continental ice-sheet melting) creating a lid on the ocean and preventing vertical mixing (Worthington, 1968). Continued biological activity in the oceans might have led to the accumulation of C02 and hence increased dissolution (Berger et al, 1977). Perhaps it is this signal that is observed following the dissolution minimum. Such a hypothesis has intriguing implications; if an extensive meltwater layer did exist during deglaciation (and if this resulted in a build-up of C02 in the subsurface waters), when ocean mixing was eventually restored an increase in atmospheric C02 concentrations would have ensued, resulting in an enhanced greenhouse effect. This may have been the sequence of feedbacks that contributed to the further decay of ice sheets and to early Holo-cene warmth.

0 0

Post a comment