. .

. . . . - -

28 Environ Scl Techno1 , Vol 27. No. 1, ? 9

,-.: ".9 . . . .

RE(0NSTRUCTING PAST ECEANIC ++’-’* TEMPERATURES

- - - - F R O M MARINE O R G A N I C

BIOGEOCHEMISTRY, CHEMICAL FOSSILS I AND MOLECULAR STRATIGRAPHY 5

Organic biomarkers

Organic geochemists use organic chemical fossils, known as “bio- markers,” to infer biogenic sources for sedimentary organic matter. Be- cause the carbon skeletons of bio- markers are preserved in the sedi- mentary record, they provide an unambiguous link to known natural products biosynthesized by particu- lar organisms. Biomarkers are of in- I terest in petroleum geochemistrv

edimentary rocks preserve a record of the environment that existed at the time th- sediments were laid dowi From t h i s s t ra t igraphi - record, information on pa- leoecosystems and paleoen- vironments may be inferred,

assuming that geoscientists can in- terpret the record. Just as paleontol- ogists use morphological fossils in biostratigraphy, so too can organic geochemists use “chemical fossils”

#

0 SEDIMENTARY

ALKENONE UNSATURA-

TION INDICES CAN BE

USED TO RECONSTRUCT ( I ) in an analogous “molecular stratigraphy” (2). The basis of mo- lecular stratigraphy is relating mo- lecular organic components iso- lated from a sediment to their biological sources. Like morpholog- ical fossils, the interpretation of dis- tributions of organic molecules

because t6ey provide htal , diagnoi- tic information about the ancient bi- ological sources of organic material that has been converted into fossil fuels. They also help to explain the complex biological, chemical, and physical processes by which these conversions occur.

must be constrained by facies and by considerations of variations in ecological distributions of biota and selective preservation.

Molecular stratigraphy is not a new concept in environmental sci- ence: it has been used to trace in-

I I -. .. . . - . . - . .. . -

VALUABLE FOR J A T I N G GLOBAL For example, in petroleums the

puts of xenobiotics into age-dated contemporary aquatic sediments. This provides a means of assessing time-changing inputs of xenobiotics to the environment and, hence, clues about the health of the envi- ronment. For example, concentra- tions of polycyclic aromatic hydro- carbons in diverse sediments have

increased in parallel with fossil fuel combustion over the past century (3). Similarly, sedimentary distribu- tions of polychlorinated biphenyls (PCBs) track increased manufactur- ing, usage and dispersal and, more recently, declining inputs of these chemicals to the environment (4 ) .

late algae (5).~The conversion of dinosterol to its hydrocarbon prod- uct requires a suite of specific geochemical reaction conditions.

Organic geochemical indicators can also be useful in evaluating the role of oxic versus anoxic sedimen- tary environments as a factor in pre- serving the biomarker record. The

Environ. Sci. Technol.. Vol. 27, No. 1, 1993 29

relative abundance of the iso- prenoid hydrocarbons pristane and phytane (Figure 1) may depend on the redox potential and acidity of the sediments in which they form from the common precursor phy- tol-the esterified alkyl side-chain of chlorophyll a. Phytane may be preferentially formed under anoxic conditions, whereas pristane may form primarily in oxic environ- ments (6).

Deducing paleoenvironmental conditions requires a fundamental understanding of present-day pro- cesses that produce biomarkers and that deliver them to and preserve them in the sediment. Implicit in the biomarker concept is the as- sumption that modern environ- ments are comparable to ancient en- vironments; as the saying goes, “The present is the key to the past.”

Within this context, the field of marine organic geochemistry has developed. A major goal of this multifaceted field-which bridges the disciplines of biology, geology, and chemistry (yielding “bio- geochemistry”)-is to characterize source and transformation pro- cesses involving organic matter in the modern ocean to better utilize the biomarker information con- tained in sediments. An excellent example of the close coupling of marine organic biogeochemistry with sedimentary molecular strati- graphy is the attempt to use algal lipids, particularly long-chain alke- nones, as molecular indicators of paleoenvironments, thereby infer- ring information ahout variations in previous global climates.

Alkenone paleothermometry In the late 1970s organic geo-

chemists discovered a series of alk- enones consisting of 37-39 carbon atoms and 2 4 double bonds (Fig- ure 1) in various marine sediments (7, 81. Subsequent work showed that these alkenones were biosyn- thesized by the coccolithophorid Emiliania huxleyi, a ubiquitous uni- cellular marine alga (9).

The relative amounts of di-, tri-, and tetra-unsaturated alkenones produced by the algae vary with temperature, possibly as a biologi- cal mechanism for maintaining cel- lular membrane fluidity under changing habitat temperatures (9 ) . Alkenone biosynthesis appears to be restricted to a few species of the class Prymnesiophyceae, notably the coccolithophorids, of which E. huxleyi is an abundant species. Moreover, alkenone unsaturation

FIGL..- , Structures of organic biomarkers

patterns appear largely unaltered by food web processes (101 or by subse- quent sedimentation and burial in marine sediments ( I 1 ) . The unusual stability of alkenones may he attrih- utable to their trans-configurations ( in contrast to the majority of unsat- urated lipids, which are in the cis- configuration 1121) and t i l their un- usual spacing of double bonds along the carbon chain.

These facts suggested that alk- encme distributions in marine sedi- ments might faithfully record ocean temperatures at the time of their biosynthesis in surface waters and deposition in the sediment. The first demonstration of this potentinl paleothermometer came from a sed- iment core covering the past 500,000 years in the Kane Cap off northwestern Africa.

The core was analyzed for bath alkenones and oxygen-18 (13). The lRO isotope signal (61801 preserved in the carbonate tests of planktonic foraminifera (protozoan zooplank- ton that reside in the upper water column and produce calcium car- bonate skeletons) is an established

paleoceanographic tool that indi- cates ice volume, or the amount of water stored in polar ice caps, and hence is a proxy for oceanic temper- ature.

The temperature-induced varia- tion in alkenone unsaturation was expressed as an alkenone unsatura- tion index, U5,where U:, = [C,,,, -

indicates the number of carbon at- oms in the molecule: number of double bonds). Trends in tempera- ture based on alkenone unsatura- tion in the Kane Gap core were con- sistent with trends in temperature suggested by the “0 data: lower Ut, corresponded to periods of lower temperatures. However, abso- lute temperatures could not be ac- curately assigned to the Uk,data. Fortunately, concurrent work on the organic composition of particu- late matter in the surface ocean pro- vided the necessary calibration.

Suspended particles from several locations contained significant con- centrations of alkenones, and a cor- relation between & and the ocean temperature led to a field-based U$,

G , J [ C 3 m + C3,:3 + C,,,l subscript

30 Environ. Sci. Technoi.. Vol. 27. No. 1. 1993

Plot of alkenone unsaturation index ( U k = [C, 2v[C3, + C,, J), whereC,, , has been omitted because of its low concentrations, against surface mixed-layer (SML) temperature for natural particulate matter samples collected at various geographic locations in the world's oceans

.... ...'

..... ...* .... ..' .... ..* ,.."' Suspended palliculate material

.... 0 Sediment trap material

versus temperature calibration over a temperature range of 8-28 "C (14, Figure 2). Laboratory cultures of E. huxlevi provided a similar calibra- tion (i4,-15).

With the availabilitv of a eener- " ally accepted Ui, temperature cali- bration, sedimentary alkenone un- saturation indices can be used to reconstruct past surface ocean tem- peratures and are valuable for eval- uating global climate change. An understanding of the relationship between past sea surface tempera- tures and climate change is required for the development of predictive models of global atmospheric and oceanic circulation. Present-day alkenone-derived ocean tempera- ture records are available as a basis for assessing past environmental conditions.

For example, wind-driven up- welling of cold, nutrient-rich waters in the Pacific Ocean off the coast of Peru (at 11-15' south latitude) leads to cold and highly productive sur- face waters, producing one of the world's major fisheries for ancho- vies. However, climatic distur- bances every several years (the well-known El Nifio events) lead to markedly reduced upwelling, warmer surface waters, and cata- strophic crashes in the anchovy fishery.

Analysis of the stratigraphic

record underlying the Peruvian up- welling area has shown that lower Ut, values and lower estimated sea temperatures (-19 "C) occurred during strong upwelling periods, whereas higher Ut, values and warmer temperatures (-21 "C) oc- curred during strong El Niiio events (16). A comparable but more de- tailed stratigraphic analysis of an- nual sediment laminations in the Santa Barbara Basin has yielded a remarkable correspondence be- tween alkenone temperatures, El Nifio events, and historical records of ocean temperatures off the south- ern California coast during the twentieth century (17, Figure 3).

Long-term global climatic varia- tions can also be reconstructed from sedimentary alkenone distribu- tions. For example, based on a core taken in the equatorial central Pa- cific Ocean, the estimated tempera- ture during the most recent glacial period, 25,000 years ago, was 25.5 OC compared with 27 "C during the current interglacial period (16). In a 650,000-year-old core from the eastern Atlantic Ocean, alkenone and P O records indicate tempera- ture fluctuations with warmings of 5-6 OC apparently related to glacier loss in the northern hemisphere (19). Moreover, significant short- term (<lo00 years) variations in ocean temperature were apparent

for alkenone data but were absent from the isotopic record.

In both the Pacific and the Atlan- tic reconstructions, ocean tempera- ture variation likely reflects global shifts in wind strength over glacial- to-interglacial transitions. Stronger winds during glacial times would lead to greater upwelling, cooler sea surface temperatures, and lower U,k, values; weaker interglacial winds would lead to reduced up- welling, warmer ocean tempera- tures, and higher Ui7 values. Alke- nones have been detected i n sediments as old as the Cretaceous (-105 million years before the present) (20). However, these hy- drocarbons arise from a source other than E. huxleyi because the fossil record for this organism only dates back to 250,000 years (20).

Interpreting the alkenone temper- ature record is not without caveats. U& temperature calibrations are based primarily on cultures of E. huxleyi, but field evidence suggests that races different from those used in culture may have different tem- perature responses. Other alkenone- producing organisms of the class Prymnesiophyceae might behave differently.

It is also not clear exactly what environmental temperature the alkenones are recording. If the or- ganisms inhabit surface waters, the temperature record may be of sea surface temperatures. However, if the algae reside deeper in the photic zone-say at 100 m-and there are strong temperature gradients with increasing depth, the temperatures would correspond to subsurface temperatures.

Further complications arise if water column @,values are in- consistent with those recorded in the underlying sediment. Such in- consistencies could arise because of confounding effects of varying tem- perature response of the putative source organisms and their distribu- tions within the world's ocean, or because the alkenones were biosyn- thesized elsewhere geographically and under different environmental conditions, then transported later- ally by ocean currents to their site of deposition.

Although unsaturation patterns of alkenones appear to be unaffected by water column and sedimentary diagenesis over relatively short- term geological periods, long-term stability of the U5,record is difficult to verify. Reconstructions of abso- lute sea surface temperatures may be elusive without a more robust

Envimn. Sci. Technol., Vol. 27. NO. 1, 1993 31

5* 1959

3 A 1 9 6 6

yh

e- 1954

191w19

, ,P,

vs (198Z83

S (197273)

M + (1 965)

S (1957/58)

M + (1 953)

M + (1943) S(1940/41) M + (1939)

S (1932)

VS (1925/26)

S (1917) 1

calibration, but the results available to date certainly indicate a convinc- ing relationship between ocean temperatures and the distributions of these sedimentary compounds. The alkenone-derived temperature record may even be recoverable in sediments from which the "0 iso- topic record has been lost because of carbonate dissolution (21).

Other potential applications Paleoenvironmental reconstruc-

tions based on alkenones may not be limited to estimating ocean tem- peratures. Three additional applica- tions are being tested in several geochemical laboratories.

First. initial evidence suggests

that alkenones are remarkably sta- ble in sediments relative to many other biomarkers. If this long-term stability is confirmed, higher con- centrations of alkenones in sedi- ments may correlate to higher pho- tosynthetic production by marine algae in surface seawater. Thus these biomarkers may provide in- formation about variations in cli- matically driven productivity in the past. In both the Santa Barbara Ba- sin and the central Pacific, it is be- lieved that elevated concentrations of alkenones may indeed be found in sediments deposited during times of stronger upwelling and higher ocean productivity.

Second, sea surface temperatures

derived from alkenone unsaturation indices may be used to indirectly estimate seawater salinity. This es- timate is based on the dependence of the "0 isotopic signature of inor- ganic carbonates on both the tem- perature and the salinity of seawa- ter a t the time the carbonates precipitated. Measurement of 6"O in sedimentary carbonates coupled with temperature data makes it pos- sible to calculate salinity. In fact, such an exercise has yielded a paleosalin- ity history of the Black Sea (22). Ap- plication of this concept to the open ocean has yet to be attempted.

Finally, alkenones have been used in a biomarker-based proce- dure for estimating the history of CO, partial pressure in ocean sur- face waters and the atmosphere (23). The 613C of inorganic and or- ganic carbon produced in surface waters by photosynthetic plankton depends both on the 6I3C and con- centration of dissolved CO,, and on the isotopic fractionations that oc- cur during incorporation of dis- solved CO, into inorganic carbonate and organic matter. The 6% of CO, may be estimated from a measure- ment of 6% of foraminiferal tests, whereas the 613C for organic carbon may be obtained by isotopic analysis of alkenones by using newly devel- oped gas chromatography-isotope ratio mass spectrometry (24).

If the two isotopic fractionations have been characterized by previ- ous laboratory experiments, the concentration of dissolved CO, can be readily calculated. Application of Henry's law would yield an esti- mate of atmospheric PCO2. A 6% analysis of alkenones in a sediment core from the Gulf of Mexico has provided a 100,000-year record of atmospheric CO, concentrations that compares favorably with CO, levels obtained from the analysis of gases trapped in ice cores.

These applications are but a few examples of how organic biomark- ers are providing excitement for geoscientists interested in recon- structing the history of the Earth's environment. Future work by bio- geochemists will probably show that other chemical fossils have similar potential. As with the basic premise of determining tempera- tures from alkenone unsaturation, these potential applications will re- quire rigorous testing. With a better understanding of the changes that occurred in the past environment will come a better understanding of the forces leading to global environ- mental change in the future.

32 Environ. Sci. Technol., Vol. 27, No. 1, 1993