Climatic impact on Al, K, Sc and Ti in marine sediments: Evidence from ODP Site 1144, South China Sea

Transcription

1 Geochemical Journal, Vol. 37, pp. 593 to 602, 2003 Climatic impact on Al, K, Sc and Ti in marine sediments: Evidence from ODP Site 1144, South China Sea GANGJIAN WEI, 1 * YING LIU, 1 XIANHUA LI, 1 LEI SHAO 2 and XIRONG LIANG 1 1 Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou , P.R. China 2 Laboratory of Marine Geology, Tongji University, Shanghai , P.R. China (Received June 20, 2002; Accepted May 29, 2003) Al, K, Sc and Ti concentrations of the terrestrial material-dominant sediments from ODP site 1144 were reported. Comparison between the bulk and the acid-leached sediments indicates that about 20~30% of the Al, K and Sc in the bulk sediments are not hosted in terrestrial detritus, rather they are of authigenic origin. However, authigenic Ti is negligible. The results indicate that Ti rather than Al is the best proxy for terrestrial materials. Significant climate controls are displayed in the Al/Ti, K/Ti and Sc/Ti variation patterns both for the bulk and the acid leached sediments. Such variation patterns can be mainly accounted for in terms of climate change in their provenance areas in South China. Elevated Al/Ti, K/Ti and Sc/Ti ratios during interglacial periods indicate that chemical weathering then was stronger than during glacial periods, which might be related to a more humid climate in interglacial periods. INTRODUCTION Marine sediments are one of the most important paleoclimate archives. Chemical compositions of the sediments may well record the climate changes. For example, carbonates, organic materials and biogenic opals are generally used climate proxies. In addition, the biogenically-related elements, such as Ba, and those sensitive to redox conditions, such as Mn, may also reveal paleoclimate change (Dymond et al., 1992; Mangini et al., 1990). However, little attention has been paid to the climatic implications of the elements that are mostly associated with terrestrial materials, such as Al, Ti, Sc, K, etc. Al, Ti and Sc concentrations in seawater are very low (Bruland, 1983), and they are mostly bonded to terrestrial materials in marine sediments (Goldberg and Arrhenius, 1958). Biogenic/nonbiogenic scavenged Ti and Sc may be present in pelagic sediments (Goldberg and Arrhenius, 1958). Also, there exists biogenic scavenged Al in the sediments where biogenic materials are dominant (Murray and Leinen, 1996). However, for sediments in which terrestrial materials are dominant, such authigenic components should be negligible if compared with their fairly high contents in terrestrial detritus. K is one of the major cations in seawater (Bruland, 1983), and it tends to be absorbed by clay minerals in marine sediments (Weaver, 1967). However, most of the K in marine sediments is still associated with terrestrial materials (Goldberg and Arrhenius, 1958; Weaver, 1967). In marine sediments, the bulk of these elements are from terrigenous sources. Some of the climate-controlled processes, such as weathering and transportation processes, would affect their concentrations or ratios. These elemental records may probably provide some paleoclimate information that is different from those revealed by authigenic materials, and may improve our understanding of paleoclimate changes. However, in paleoceanographical studies, Al and Ti are generally used to estimate the abundance of terres- *Corresponding author ( gjwei@gig.ac.cn) 593

2 594 G. Wei et al. Fig. 1. Location of the ODP Site trial materials (Murray and Leinen, 1996; Klump et al., 2000), and paleoclimate records based on these elements are scarcely reported. In the North Slope of the South China Sea, Quaternary sedimentary rates are very high. As a result, terrestrial materials are dominant in the sediments (Wang, 1999). Monsoon climate is prevalent in this area (Wang, 1999). Therefore, some of the paleoclimate records in marine sediments can be linked to changes on nearby continents. We analysis Al, K, Sc and Ti concentrations of the sediments from ODP site 1144, which records very high sedimentary rates (Shipboard Scientific Party, 2000). We attempt to use these elements to obtain insight into their paleoclimate implications and the linkage to paleoclimate changes in South China. MATERIALS AND METHODS The sediments are from ODP Site 1144 in northern South China Sea ( N, E, water depth of 2037 m, Fig. 1). The length of core 1144A is 452.8m (Shipboard Scientific Party, 2000). The analyzed samples were collected from top to section 17H of core 1144A. Sampling interval was about 1.3m, with a time resolution about 1.5 ka. To evaluate the terrigenous background of these sediments, another series of samples were harvested from the upper 400 m section to estimate the chemical components of the terrestrial detritus. The bulk sediment samples were dried at 80 C, ground to powder, and then baked at 670 C for 30 minutes to destroy organic matters. Those samples for terrigenous background measurement were further treated with 1N HCl for 24 hours to remove carbonate, washed three times by de-ionized water and then dried. After each of the acid leaching and water washing, the tubes were centrifugated for 20 minutes on rate of 6000 cycles/min to prevent the losing of the very fine grain materials. Such chemical treatments can remove most of the non-detrital materials, such as carbonate, Fe-Mn oxyhydroxides, absorbed components and sea salt, except for biogenic opal, but result in very small amount of the detritus losing by dissolution (Freydier et al., 2001). According to our chemical treatment, the base of our bulk sediments is absorbed water free, and the base of the acidleached samples is very close to the terrestrial background of the sediments at ODP Site 1144 except for some dilutions by biogenic opal. The dry bulk samples and acid-leached samples were totally digested by HNO 3 +HF. A series of Chinese rock standards GSR-1, GSR-2, GSR-3, GSR-4, GSR-5, GSR-6, and some Chinese and USGS sediment standards, such as GSD-9, GSD-12, MAG-

3 Climatic impact on elements in South China Sea sediments were also digested along with the samples. These Chinese rock standards have different mineralogy, with their major element concentrations spanning a fairly large range (Govindaraju, 1994). They were used as external standards for monitoring measurement of the sediment samples. The Chinese and USGS sediment standards were used for quality control of the analysis. Al, K, Ti and Sc concentrations were measured on a Varian Vista ICP-AES in Guangzhou Institute of Geochemistry, Chinese Academy of Science. Precision for these elements is better than 3%. Repeated analysis of the sediment standards indicates that the accuracy for these elements were better than 5% (Table 3). The paleoclimate patterns and the preliminary age control for sediments of ODP site 1144 were obtained by comparing oxygen isotopes of G. ruber (Buehring et al., 2003) with the SPECMAP stack (Martinson et al., 1987). RESULTS Al 2 O 3, TiO 2, K 2 O and Sc results are listed in Table 1. Average Al 2 O 3, TiO 2, K 2 O and Sc for the acid leached samples are (14.84 ± 1.11) wt%, (0.96 ± 0.04) wt%, (3.26 ± 0.34) wt% and (14.2 ± 1.4) ppm, respectively. By comparison, TiO 2 is very close to that in the Post Archean Average Shales (PAAS, ~1 wt%) (Taylor and Mclennan, 1985). However, Al 2 O 3, K 2 O and Sc are lower than those in PAAS, about 10% lower for K 2 O and Sc, and 20% lower for Al 2 O 3. Because the acid leached samples had organic materials removed by baking and carbonates and absorbed materials removed by acid leached, their concentrations should be best representative of the detrital components of the sediments from ODP The difference between the concentrations of these samples and PAAS indicates that the detrital component in north South China Sea may slightly differ from PAAS. The average concentrations of Al 2 O 3, TiO 2, K 2 O and Sc for the dry bulk sediments are (15.09 ± 1.01) wt%, (0.74 ± 0.03) wt%, (3.01 ± 0.28) wt% and (15.1 ± 1.3) ppm, respectively. Considering that the top 160 mcd sediments of Table 1. Al 2 O 3, K 2 O, TiO 2 and Sc concentrations for the acid-leached samples from the ODP 1144 mcd (m) Al 2 O 3 K 2 O TiO 2 Sc *Note: Units for Al 2 O 3, K 2 O and TiO 2 are wt%, and µg/g for Sc. mcd is the abbreviation of meters composite depth. ODP 1144A contain about 20~30% of non-detrital materials, such as biogenic materials including carbonates, organic materials and biogenic opal, Fe-Mn oxyhydroxides, sea salt and absorbed materials (Shipboard Scientific Party, 2000), the concentrations of Al 2 O 3, TiO 2, K 2 O and Sc of the bulk sediments should be lower than those of acid-

4 596 G. Wei et al. Table 2. Al 2 O 3, K 2 O, TiO 2 and Sc concentrations of the dry bulk sediments from ODP 1144 mcd (m) Al 2 O 3 K 2 O TiO 2 Sc Table 2. (continued) mcd (m) Al 2 O 3 K 2 O TiO 2 Sc leached samples as a result of dilution by the nondetrital materials. TiO 2 and K 2 O concentrations are consistent with this case. However, bulk sediment Al 2 O 3 and Sc concentrations are even higher than acid-leached samples, indicating that the presence of non-detrital components. The element records were transferred to Tinormalized ratios because titanium is the best proxy for terrigenous component in marine sediments, especially for the sediments from South China Sea (Wei et al., 2003). The Ti-normalized ratios could help to evaluate the authigenic enrichment of these elements in marine sediments (Murray et al., 2000). Also, Titanium is conservative during chemical weathering process, and Tinormalized ratios are benefit to find the behavior of the elements during chemical weathering (Nesbitt and Markovics, 1997). Supposed that Al,

5 Climatic impact on elements in South China Sea sediments 597 Table 2. (continued) mcd (m) Al 2 O 3 K 2 O TiO 2 Sc *Note: Units for Al 2 O 3, K 2 O and TiO 2 are wt%, and µg/g for Sc. mcd is the abbreviation of meters composite depth. Sc, K, and Ti are not concentrated in biogenic opal after acid leaching, the Ti-normalized ratios of the acid-leached samples are not modified by biogenic opal dilution. Therefore, they could be the best estimate for the terrestrial background of these ratios in this area. Al/Ti, Sc/Ti and K/Ti variations for both the dry bulk sediments and the acidleached samples resemble that of the SPECMAP (Fig. 2). In contract with the oxygen isotope record of G. ruber, significant shifts in Al/Ti, Sc/Ti and K/Ti ratios generally occurred during major climate changes. Al/Ti, Sc/Ti and K/Ti ratios were higher during interglacial periods, and lower during glacial periods. The magnitude of change in these ratios over glacial/interglacial cycles are 20~30%, which are several times larger than the analysis precision, indicating that such variation patterns are significant for these sediments. Therefore, such element ratios in the sediments from ODP 1144 are climatically sensitive. DISCUSSION Authigenic components Assuming that the entire Al, Ti, K and Sc in the bulk sediments are terrigenous, 20% correction of the dilution effect will increase the average concentrations of Al 2 O 3, TiO 2, K 2 O and Sc in the bulk sediments to 18.9 wt%, 0.92 wt%, 3.76 wt% and 18.9 ppm, respectively. Compared with the average values of these elements in the acid leached samples, TiO 2 concentrations are nearly the same. However, the corrected concentrations for Al 2 O 3, K 2 O and Sc are 27%, 15% and 33% higher than those of acid-leached samples. It is worth to note that acid leaching may dissolve some poorly crystallized clay minerals that are possibly detrital originated, which may cause some lose of these elements in acid leached samples. However, only when Al, K, and Sc but not Ti are highly enriched in these poorly crystallized clay minerals, and the amount of these minerals is large enough, the loss of these minerals by acid leaching could cause significant decrease of the concentrations of Al, K and Sc in acid-leached samples. Such case sounds impossible. Sc is highly compatible element in continental crust, and it seems not concentrate in any definite minerals in clastic sediments (Taylor and McLennan, 1985). Therefore, selective remove of Sc from detrital component of the sediments by acid leaching seems not the case. As Al, K and Ti are all major metals in detrital materials, when significant loss of Al and K occurs by dissolving part of the detri-

6 598 G. Wei et al. Table 3. The measured results of the sediment standards GSD-9, GSD-12 and MAG-1 Sample ID GSD-9 GSD-12 MAG-1 Al 2 O ± 0.14 (10.58) 9.49 ± 0.22 (9.3) ± 0.05 (16.37) K 2 O 1.94 ± 0.03 (1.99) 2.83 ± 0.04 (2.91) 3.57 ± 0.05 (3.55) TiO ± 0.03 (0.92) ± 0.01 (0.25) ± (0.751) Sc ± 0.54 (11.1) 5.39 ± 0.19 (5.1) ± 0.35 (17.2) *Units for Al 2 O 3, K 2 O and TiO 2 are percentage in weight, and µg/g for Sc. The bracketed data are the reference values (Govindaraju, 1994). For each standard sediment samples, at least 3 measurements were taken. The errors are 1σ. Fig. 2. Variations of the Al/Ti, K/Ti and Sc/Ti ratios, the foraminifer oxygen isotopes and the medians of grain sizes. The filled symbols represent the bulk sediments, and the open symbols denote the acid leached sediments. Oxygen isotopes are from Buehring et al. (2003). Horizontal dashed lines mark the significant climate changes. tal materials, Ti should be expected to lose, either. Also, the amount of the lost detrital materials by acid dissolving is small because within our estimation of 20% loss by acid leaching, carbonate accounts for 10% to 15%, and absorbed materials, Fe-oxyhydroxides and other non-detrital may account for several percentage. Therefore, the loss of the detrital originated poorly crystallized clay minerals by acid leaching seems not contribute much to the significant decrease of Al, K and Sc concentrations of the acid leached samples, and non-detrital materials should be the main contribution. This indicates that all the Ti in bulk sediments are hosted in detrital materials, but a significant proportion of the Al, Sc and K in the bulk sediments seems not to be associated with detrital materials. Generally, Al, Ti, K and Sc in marine sediments are associated with detrital materials (Goldberg and Arrhenius, 1958). However, Al can also be scavenged by Fe-oxyhydroxides, biogenic opal and clay minerals in organic-rich sediments (Mackin and Aller, 1984), and Sc tends to enrich in ferromanganese oxide minerals and biogenic apatite in pelagic sediments (Goldberg and Arrhenius, 1958). Biogenic/non-biogenic scavenge for Ti in pelagic sediments also occurs (Goldberg and Arrhenius, 1958; Orians et al., 1990). The above calculation indicates that even in the terrigenous materials-dominated sediments, authigenic enrichment of Al and Sc is significant. However, authigenic enrichment of Ti is negligible. As for K, it tends to be absorbed by clay minerals (Weaver, 1967), resulting in higher K 2 O con-

7 Climatic impact on elements in South China Sea sediments 599 centrations in the bulk sediments. Al and Ti are generally used to estimate the percentage of the terrestrial materials in marine sediments, and to subtract the contribution of the detrital materials when calculating the authigenic components of marine sediments for both pelagic and terrigenous materials-dominated sediments (Murray and Leinen., 1996; Klump et al., 2000). Our results indicate that Ti should be the best indicator for calculation of the detrital components in marine sediments. In contrast, the use of Al may overestimate the terrigenous abundance by 20~30%, resulting in over subtraction for the authigenic materials. Paleoclimate implications Even though authigenic Al, K and Sc are present in the bulk sediments of ODP 1144, the SPECMAP-like variation patterns displayed by the Al/Ti, Sc/Ti and K/Ti ratios are not due to the presence of authigenic components. As shown in Fig. 2, both the bulk sediments and the acid leached samples exhibit similar climate cycles, and the relative variation of these ratios for bulk sediments within glacial/interglacial cycles are nearly the same as those of the acid leached samples. Therefore, the SPECMAP-like patterns of the Al/Ti, Sc/ Ti and K/Ti ratios are controlled by the detrital components, which can be considered as proxies for climate change in the provenance region. Continental climate information was generally extracted from clay minerals in marine sediments (Singer, 1984). However, the presence of authigenic clay minerals as well as erosion and transportation processes may obscure the climate signals within the clay minerals (Singer 1984; Thiry, 2000). In addition, clay mineral records can only be indirectly linked to short-term paleoclimate changes (Thiry, 2000). As for our element ratios, they are comprehensive records of the mixture of clastic minerals, such as clay minerals, feldspar, quartz etc., and authigenic materials. As discussed above, the SPECMAP-like variations of the Al/Ti, Sc/Ti and K/Ti ratios are not controlled by authigenic materials. The transportation processes may be responsible for the sorting of the terrestrial materials and distribution of grain sizes to some extent. The variation patterns of the median of the grain size are shown in Fig. 2. The grain size median values of the ODP 1144 sediments mostly range within 2 µm and 10 µm with some spikes over 10 µm. The variation patterns of the grain size median exhibit relatively higher values during the glacials revealed by foraminifer oxygen isotopes than during interglacials. Generally, many of the elements in marine sediments, such as Al, K, Sc and Ti tend to enrich in fine grain size components (Zhang et al., 2002). The variations of the grain size might take some effect to the concentrations of these elements, for examples, higher contractions during interglacials. However, the grain size patterns do not always match the oxygen isotopes in Fig. 2 as those of the element records. For examples, during oxygen isotope Stage 6 (MIS 6), the median values do not exhibit higher values. As all these elements tend to enrich in fine grain size components, the controls of the grain size to the element ratios sound uncertain. Therefore, the grain size effect seems not the main controls of the SPECMAP-like patterns of Al/Ti, Sc/Ti and K/Ti of the ODP 1144 sediments. We propose that variations in the clastic material supply from the continent may be responsible for such variations. The coastal areas of South China are the most important provenance for the sediments of ODP 1144, and the provenance had not changed significantly during the last 1 Ma (Shao et al., 2001). The variations of the Al/Ti, Sc/Ti and K/Ti ratios of the ODP 1144 sediments may indicate that the elemental components of clastic materials supplied to the South China Sea changed during glacial/interglacial cycles. Generally, the terrestrial materials in marine sediments are derived from weathering products on continents. Chemical weathering is the most important way to fractionate elements between weathering products and parental bedrocks. The elemental components of the weathering products may be different when the parental bedrocks experience different weathering conditions. Al is resistant to fluid leaching, and it tends to enrich in the resi-

8 600 G. Wei et al. due during chemical weathering processes (Nesbitt and Young, 1982). K may be leached when K-feldspar is weathered, but most of the dissolved K could be absorbed and fixed in secondary minerals (Weaver, 1967), which explains the fact that weathering products, i.e., secondary minerals, generally have higher K concentrations than parental bedrocks (Peuraniemi and Pulkkinen, 1993). Ti concentrations in weathering products are slightly lower than in parental rocks, but the concentration changes are not significant (Peuraniemi and Pulkkinen, 1993). Sc is a strong compatible element in igneous rocks. However, very little is known about the Sc concentration change during chemical weathering processes. As indicated by Taylor and McLennan (1985), the behavior of Sc is similar to rare earth elements (REE) in sedimentary rocks. Considering that REEs tend to enrich in weathering residue (Taylor and McLennan, 1985), it is likely that Sc should enrich in weathering residue, too. Therefore, chemical weathering processes could concentrate Al, K and Sc in the weathering products. The strengthening of the chemical weathering may result in higher Al/Ti, Sc/Ti and K/Ti ratios in the weathering products. Higher Al/Ti, Sc/Ti and K/Ti ratios of the ODP 1144 sediments during interglacial periods may indicate that the clastic materials supplied to the north South China Sea may undergo stronger chemical weathering during interglacial periods than during glacial periods. In most areas of South China there covering layer of red clay with depth generally about several tens to over 100 meters (Zhu, 1993). This red clay layer is believed to have formed prior to Neogene and is still developing at present time (Zhu, 1993). This red layer may have made a significant contribution to the clastic materials supplied to the north South China Sea. Therefore, much of the chemical weathering process may be reworking of the red clay other than weathering of the fresh bedrocks. Such processes should be much faster than forming new soils (Thiry, 2000), and they are sufficient to reveal the variations during glacial/interglacial cycles. Stronger chemical weathering is generally associated with warm and humid climates (Nesbitt and Young, 1982). The variations of the Al/Ti, Sc/ Ti and K/Ti ratios in the ODP 1144 sediments indicate that warm and wet climate occurred in South China during interglacial periods, whilst cold and dry climate occurred during glacial periods. Similar conclusions have also been drawn from pollen records at northern South China Sea (Sun and Li, 1999). Monsoon climate dominates in South China (Ding, 1994). As indicated by Wang (1999), winter monsoon was strengthened during glacial periods, and summer monsoon was enhanced during interglacial periods. Summer monsoon controls precipitation in South China (Ding, 1994). Enhanced summer monsoon during interglacial periods brought more precipitation in South China, and boosted chemical weathering of the bedrocks and red clay layer. Thus, the terrestrial detritus in the sediments in northern South China Sea should exhibit stronger chemical weathering signals. The elemental records of the sediments at ODP Site 1144 are consistent with this model, suggesting that they are reliable proxy for the change of monsoon climate in South China. CONCLUSION (1) The comparison of the terrestrial dominion elements, such as Ti, Al, Sc and K in the bulk and the acid-leached sediments of ODP 1144 demonstrate the presence of some authigenic Al, Sc and K components in the bulk sediments. However, authigenic Ti is negligible in the bulk sediments. Considering that Ti and Al are generally used to estimate terrigenous abundance in marine sediments, we conclude that Ti should be the best indicator for estimating terrigenous abundance, whilst the use of Al might result in 20~30% overestimation. (2) Al/Ti, Sc/Ti and K/Ti ratios for both the bulk and the acid-leached sediments display significant SPECMAP-like patterns, implying a climate forcing for these elements. The glacial/interglacial variations of these element ratios are not related to the presence of authigenic materials but to climate change in the continental source region.

9 Climatic impact on elements in South China Sea sediments 601 The sorting of the detrital materials during the transport process may partly account for such variation patterns, but not the main controls. Higher Al/Ti, Sc/Ti and K/Ti ratios during interglacial periods indicate that chemical weathering in South China are stronger than during glacial periods. Enhanced summer monsoon, which would bring more precipitation and result in more humid climate in South China, may be responsible for the elevated Al/Ti, Sc/Ti and K/Ti ratios. Thus, Al/Ti, Sc/Ti and K/Ti ratios in the ODP 1144 sediments could be reliable proxies for the monsoon climates in South China. Acknowledgments All the samples are supply from the Ocean Drilling Project. We thank the two anonymity reviewers for their constructive comments and suggestions, which significantly improved the manuscript. We thank Dr. Zhao J. X. from Queensland University of Australia for help to improve the manuscript. This work is supported by National Natural Science Foundation of China (NSFC Grant No ). REFERENCES Bruland, K. W. (1983) Trace elements in sea water. Chemical Oceanography, Vol. 8, 2nd ed. (Riley, J. P. and Chester, R., eds.), , Academic Press, New York. Buehring, C., Sarnthein, M., Erlenkeuser, H. (2003) Toward a high-resolution stable isotope stratigraphy of the last 1.1 million years: ODP Site 1144, South China Sea. Proc. ODP Sci. Res. (in press). Ding, Y. H. (1994) Monsoons over China. Kluwer Academic Publishers, 432 pp. Dymond, J., Suess, E. and Lyle, M. (1992) Barium in deep-sea sediment: A geochemical proxy for paleoproductivity. Paleoceanography 7, Freydier, R., Michard, A., De Lange, G. and Thomson, J. (2001) Nd isotope compositions of Eastern Mediterranean sediments: tracers of the Nile influence during sapropel S1 formation? Mar. Geol. 177, Goldberg, E. D. and Arrhenius, G. O. S. (1958) Chemistry of Pacific pelagic sediments. Geochim. Cosmochim. Acta 13, Govindaraju, K. (1994) Special Issue of Geostandards Newsletter. Geostandards Newsletter, X VIII, Special issue, 44. Klump, J., Hebbeln, D. and Wefer, G. (2000) The impact of sediment provenance on barium-based productivity estimates. Mar. Geol. 169, Mackin, J. E. and Aller, R. C. (1984) Diagenesis of dissolved aluminum in organic-rich estuarine sediments. Geochim. Cosmochim. Acta 48, Mangini, A., Eisenhauer, A. and Walter, P. (1990) Response of Manganese in the ocean to the climatic cycles in the Quaternary. Paleoceanography 5, Martinson, D. G., Pisias, N. G., Hays, J. D., Imbrie, J., Moore, T. C. and Shackleton, N. J. (1987) Age dating and the orbital theory of the ice ages: Development of a high-resolution 0 to 300,000-year chronostratigraphy. Quaternary Res. 27, Murray, R. W. and Leinen, M. (1996) Scavenged excess aluminum and its relationship to bulk titanium in biogenic sediment from the central equatorial Pacific Ocean. Geochim. Cosmochim. Acta 60, Murray, R. W., Knowlton, C., Leinen, M., Mix, A. C. and Polsky, C. H. (2000) Export production and terrigenous matter in the Central Equatorial Pacific Ocean during interglacial oxygen isotope Stage 11. Global and Planetary Change 24, Nesbitt, H. W. and Markovics, G. (1997) Weathering of grandioritic crust, long-term storage of elements in weathering profiles, and petrogenesis of siliciclastic sediments. Geochim. Cosmochim. Acta 61(8), Nesbitt, H. W. and Young, G. M. (1982) Early Proterozoic climatee and plate motions inferred from major element chemistry of lutites. Nature 299, Orians, K. J., Boyle, E. A. and Bruland, K. W. (1990) Dissolved titanium in the open ocean. Nature 348, Peuraniemi, V. and Pulkkinen, P. (1993) Preglacial weathering crust in Ostrobothnia, western Finland, with special reference to the Raudaskyla occurrence. Chem. Geol. 107, Shao, L., Li, X. H., Wei, G. J., Liu, Y. and Fang, D. Y. (2001) Provenance of a prominent sediment drift on the northern slope of the South China Sea. Science in China (Ser. D) 44(10), Shipboard Scientific Party (2000) Site Proc. ODP, Init Repts. (Wang, P., Prell, W. L., Blum, P. et al., eds.), 184, 1 97 [CD-ROM]. Avaiable from: Ocean Drilling Program, Texas A&M University, College Station TX , U.S.A. Singer, A. (1984) The paleoclimatic interpretation of clay minerals in sediments: A review. Earth Sci. Rev. 21, Sun, X. J. and Li, X. (1999) A pollen record of the last 37 ka in deep sea core from the northern slope of the South China Sea. Mar. Geol. 156,

10 602 G. Wei et al. Taylor, S. R. and McLennan, S. M. (1985) The Continental Crust: Its Composition and Evolution. Blackwell, Oxford, 312 pp. Thiry, M. (2000) Paleoclimatic interpretation of clay minerals in marine deposits: an outlook from the continental origin. Earth Sci. Rev. 49, Wang, P. X. (1999) Response of western Pacific marginal seas to glacial cycles: paleoceanographic and sedimentological feastures. Mar. Geol. 156, Weaver, C. E. (1967) Potassium, illite and the ocean. Geochim. Cosmochim. Acta 31, Wei, G. J., Liu, Y., Li, X. H., Chen, M. H. and Wei, W. C. (2003) High-resolution elemental records from the South China Sea and their paleoproductivity implications. Paleoceanography (in press). Zhang, C. S, Wang, L. J., Li, G. S., Dong, S. S., Yang, J. R. and Wang, X. L. (2002) Grain size effect on multi-element concentrations in sediments from the intertidal flats of Bohai Bay, China. Appl. Geochem. 17, Zhu, X. M. (1993) Red clay and red residuum in South China. Quaternary Sci. No. 1, (in Chinese with English abstract).