Global data compilation of benthic data sets II

In this study we present a global distribution pattern and budget of the minimum flux of particulate organic carbon to the sea floor (J POC alpha). The estimations are based on regionally specific correlations between the diffusive oxygen flux across the sediment-water interface, the total organic carbon content in surface sediments, and the oxygen concentration in bottom waters. For this, we modified the principal equation of Cai and Reimers [1995] as a basic monod reaction rate, applied within 11 regions where in situ measurements of diffusive oxygen uptake exist. By application of the resulting transfer functions to other regions with similar sedimentary conditions and areal interpolation, we calculated a minimum global budget of particulate organic carbon that actually reaches the sea floor of ~0.5 GtC yr**-1 (>1000 m water depth (wd)), whereas approximately 0.002-0.12 GtC yr**-1 is buried in the sediments (0.01-0.4% of surface primary production). Despite the fact that our global budget is in good agreement with previous studies, we found conspicuous differences among the distribution patterns of primary production, calculations based on particle trap collections of the POC flux, and J POC alpha of this study. These deviations, especially located at the southeastern and southwestern Atlantic Ocean, the Greenland and Norwegian Sea and the entire equatorial Pacific Ocean, strongly indicate a considerable influence of lateral particle transport on the vertical link between surface waters and underlying sediments. This observation is supported by sediment trap data. Furthermore, local differences in the availability and quality of the organic matter as well as different transport mechanisms through the water column are discussed.

Chemical and isotopic compositions of water, particulates, plankton, gases, sediments, and minerals from the Kara Sea and lower Ob and Yenisei Rivers

The Kara Sea is an area uniquely suitable for studying processes in the river-sea system. This is a shallow sea, into which two great Siberian rivers, Yenisei and Ob, flow. From 1995 to 2003, the sea was studied by six international expeditions onboard the R/V Akademik Boris Petrov. This publication summarizes the results obtained, within the framework of this project, at the Vernadsky Institute of Geochemistry and Analytical Chemistry, Russian Academy of Sciences. Various hydrogeochemical parameters, concentrations and isotopic composition of organic and carbonate carbon of the sediments, plankton, particulate organic matter, hydrocarbons, and dissolved CO2 were examined throughout the whole sea area at more than 200 sites. The d13C varies from -22 and -24 per mil where Atlantic waters enter the Kara Sea and in the north-eastern part of the water area to -27 per mil in the Yenisei and Ob estuaries. The value of d13C of the plankton is only weakly correlated with the d13C of the organic matter from the sediments and is lower by as much as 3-4 per mil. The paper presents the results obtained from a number of meridional river-sea profiles. It was determined from the relations between the isotopic compositions of plankton and particulate matter that the river waters carry material consisting of 70% detrital-humus matter and 30% planktonogenic material in the river part, and the material contained in the offshore waters consists of 30% terrigenous components, with the contribution of bioproducers amounting to 70%. The carbon isotopic composition of the plankton ranges from -29 to -35 per mil in the riverine part, from -28 to -27 per mil in the estuaries, and from -27.0 to -25 per mil in the marine part. The relative lightness of the carbon isotopic composition of plankton in Arctic waters is explained by the temperature effect, elevated CO2 concentrations, and long-distance CO2 supply to the sea with river waters. The data obtained on the isotopic composition of CO2 in the surface waters of the Kara Sea were used to map the distribution of d13C. The complex of hydrocarbon gases extracted from the waters included methane, C2-C5, and unsaturated C2=-C4= hydrocarbons, for which variations in the concentrations in the waters were studied along river-estuary-sea profiles. The geochemistry of hydrocarbon gases in surface fresh waters is characterized by comparable concentrations of methane (0.3-5 µl/l) and heavier hydrocarbons, including unsaturated ones. Microbiological methane with d13C from -105 to -90 per mil first occurs in the sediments at depths of 40-200 cm. The sediments practically everywhere display traces of methane oxidation in the form of a shift of the d13C of methane toward higher values and the occurrence of autogenic carbonate material, including ikaite, enriched in the light isotope. Ikaite (d13C from -25 to -60 per mil) was found and examined in several profiles. The redox conditions in the sediments varied from normal in the southern part of the sea to highly oxidized along the Novaya Zemlya Trough. Vertical sections through the sediments of the latter exemplify the complete suppression of the biochemical activity of microorganisms. Our data provide insight into the biogeochemistry of the Kara Sea and make it possible to specify the background values needed for ecological control during the future exploration operations and extraction of hydrocarbons in the Kara Sea.

Carbonate preservation of sediment cores from the Great Bahama Bank

The oceanic carbon cycle mainly comprises the production and dissolution/ preservation of carbonate particles in the water column or within the sediment. Carbon dioxide is one of the major controlling factors for the production and dissolution of carbonate. There is a steady exchange between the ocean and atmosphere in order to achieve an equilibrium of CO2; an anthropogenic rise of CO2 in the atmosphere would therefore also increase the amount of CO2 in the ocean. The increased amount of CO2 in the ocean, due to increasing CO2-emissions into the atmosphere since the industrial revolution, has been interpreted as "ocean acidification" (Caldeira and Wickett, 2003). Its alarming effects, such as dissolution and reduced CaCO3 formation, on reefs and other carbonate shell producing organisms form the topic of current discussions (Kolbert, 2006). Decreasing temperatures and increasing pressure and CO2 enhance the dissolution of carbonate particles at the sediment-water interface in the deep sea. Moreover, dissolution processes are dependent of the saturation state of the surrounding water with respect to calcite or aragonite. Significantly increased dissolution has been observed below the aragonite or calcite chemical lysocline; below the aragonite compensation depth (ACD), or calcite compensation depth (CCD), all aragonite or calcite particles, respectively, are dissolved. Aragonite, which is more prone to dissolution than calcite, features a shallower lysocline and compensation depth than calcite. In the 1980's it was suggested that significant dissolution also occurs in the water column or at the sediment-water interface above the lysocline. Unknown quantities of carbonate produced at the sea surface, would be dissolved due to this process. This would affect the calculation of the carbonate production and the entire carbonate budget of the world's ocean. Following this assumption, a number of studies have been carried out to monitor supralysoclinal dissolution at various locations: at Ceara Rise in the western equatorial Atlantic (Martin and Sayles, 1996), in the Arabian Sea (Milliman et al., 1999), in the equatorial Indian Ocean (Peterson and Prell, 1985; Schulte and Bard, 2003), and in the equatorial Pacific (Kimoto et al., 2003). Despite the evidence for supralysoclinal dissolution in some areas of the world's ocean, the question still exists whether dissolution occurs above the lysocline in the entire ocean. The first part of this thesis seeks answers to this question, based on the global budget model of Milliman et al. (1999). As study area the Bahamas and Florida Straits are most suitable because of the high production of carbonate, and because there the depth of the lysocline is the deepest worldwide. To monitor the occurrence of supralysoclinal dissolution, the preservation of aragonitic pteropod shells was determined, using the Limacina inflata Dissolution Index (LDX; Gerhardt and Henrich, 2001). Analyses of the grain-size distribution, the mineralogy, and the foraminifera assemblage revealed further aspects concerning the preservation state of the sediment. All samples located at the Bahamian platform are well preserved. In contrast, the samples from the Florida Straits show dissolution in 800 to 1000 m and below 1500 m water depth. Degradation of organic material and the subsequent release of CO2 probably causes supralysoclinal dissolution. A northward extension of the corrosive Antarctic Intermediate Water (AAIW) flows through the Caribbean Sea into the Gulf of Mexico and might enhance dissolution processes at around 1000 m water depth. The second part of this study deals with the preservation of Pliocene to Holocene carbonate sediments from both the windward and leeward basins adjacent to Great Bahama Bank (Ocean Drilling Program Sites 632, 633, and 1006). Detailed census counts of the sand fraction (250-500 µm) show the general composition of the coarse grained sediment. Further methods used to examine the preservation state of carbonates include the amount of organic carbon and various dissolution indices, such as the LDX and the Fragmentation Index. Carbonate concretions (nodules) have been observed in the sand fraction. They are similar to the concretions or aggregates previously mentioned by Mullins et al. (1980a) and Droxler et al. (1988a), respectively. Nonetheless, a detailed study of such grains has not been made to date, although they form an important part of periplatform sediments. Stable isotopemeasurements of the nodules' matrix confirm previous suggestions that the nodules have formed in situ as a result of early diagenetic processes (Mullins et al., 1980a). The two cores, which are located in Exuma Sound (Sites 632 and 633), at the eastern margin of Great Bahama Bank (GBB), show an increasing amount of nodules with increasing core depth. In Pliocene sediments, the amount of nodules might rise up to 100%. In contrast, nodules only occur within glacial stages in the deeper part of the studied core interval (between 30 and 70 mbsf) at Site 1006 on the western margin of GBB. Above this level the sediment is constantly being flushed by bottom water, that might also contain corrosive AAIW, which would hinder cementation. Fine carbonate particles (<63 µm) form the matrix of the nodules and do therefore not contribute to the fine fraction. At the same time, the amount of the coarse fraction (>63 µm) increases due to the nodule formation. The formation of nodules might therefore significantly alter the grain-size distribution of the sediment. A direct comparison of the amount of nodules with the grain-size distribution shows that core intervals with high amounts of nodules are indeed coarser than the intervals with low amounts of nodules. On the other hand, an initially coarser sediment might facilitate the formation of nodules, as a high porosity and permeability enhances early diagenetic processes (Westphal et al., 1999). This suggestion was also confirmed: the glacial intervals at Site 1006 are interpreted to have already been rather coarse prior to the formation of nodules. This assumption is based on the grain-size distribution in the upper part of the core, which is not yet affected by diagenesis, but also shows coarser sediment during the glacial stages. As expected, the coarser, glacial deposits in the lower part of the core show the highest amounts of nodules. The same effect was observed at Site 632, where turbidites cause distinct coarse layers and reveal higher amounts of nodules than non-turbiditic sequences. Site 633 shows a different pattern: both the amount of nodules and the coarseness of the sediment steadily increase with increasing core depth. Based on these sedimentological findings, the following model has been developed: a grain-size pattern characterised by prominent coarse peaks (as observed at Sites 632 and 1006) is barely altered. The greatest coarsening effect due to the nodule formation will occur in those layers, which have initially been coarser than the adjacent sediment intervals. In this case, the overall trend of the grain-size pattern before and after formation of the nodules is similar to each other. Although the sediment is altered due to diagenetic processes, grain size could be used as a proxy for e.g. changes in the bottom-water current. The other case described in the model is based on a consistent initial grain-size distribution, as observed at Site 633. In this case, the nodule reflects the increasing diagenetic alteration with increasing core depth rather than the initial grain-size pattern. In the latter scenario, the overall grain-size trend is significantly changed which makes grain size unreliable as a proxy for any palaeoenvironmental changes. The results of this study contribute to the understanding of general sedimentation processes in the periplatform realm: the preservation state of surface samples shows the influence of supralysoclinal dissolution due to the degradation of organic matter and due to the presence of corrosive water masses; the composition of the sand fraction shows the alteration of the carbonate sediment due to early diagenetic processes. However, open questions are how and when the alteration processes occur and how geochemical parameters, such as the rise in alkalinity or the amount of strontium, are linked to them. These geochemical parameters might reveal more information about the depth in the sediment column, where dissolution and cementation processes occur.

Degradation rates and productivity signals of organic-walled dinoflagellate cyst species in surface sediment samples

To understand the role of the ocean within the global carbon cycle, detailed information is required on key-processes within the marine carbon cycle; bio-production in the upper ocean, export of the produced material to the deep ocean and the storage of carbon in oceanic sediments. Quantification of these processes requires the separation of signals of net primary production and the rate of organic matter decay as reflected in fossil sediments. This study examines the large differences in degradation rates of organic-walled dinoflagellate cyst species to separate these degradation and productivity signals. For this, accumulation rates of cyst species known to be resistant (R-cysts) or sensitive (S-cysts) to aerobic degradation of 62 sites are compared to mean annual chlorophyll-a, sea-surface temperature, sea-surface salinity, nitrate and phosphate concentrations of the upper waters and deep-water oxygen concentrations. Furthermore, the degradation of sensitive cysts, as expressed by the degradation constant k and reaction time t, has been related to bottom water [O2]. The studied sediments were taken from the Arabian Sea, north-western African Margin (North Atlantic), western-equatorial Atlantic Ocean/Caraibic, south-western African margin (South Atlantic) and Southern Ocean (Atlantic sector). Significant relationships are observed between (a) accumulation rates of R-cysts and upper water chlorophyll-a concentrations, (b) accumulation rates of S-cysts and bottom water [O2] and (c) degradation rates of S-cysts (kt) and bottom water [O2]. Relationships that are extremely weak or are clearly insignificant on all confidence intervals are between (1) S-cyst accumulation rates and chlorophyll-a concentrations, sea-surface temperature (SST), sea-surface salinity (SSS), phosphate concentrations (P) and nitrate concentrations (N), (2) between R-cyst accumulation rates and bottom water [O2], SST, SSS, P and N, and between (3) kt and water depth. Co-variance is present between the parameters N and P, N, P and chlorophyll-a, oxygen and water depth. Correcting for this co-variance does not influence the significance of the relationship given above. The possible applicability of dinoflagellate cyst degradation to estimate past net primary production and deep ocean ventilation is discussed.

Benthic foraminifera of the Red Sea

The distribution of living (Rose Bengal-stained), dead and fossil benthic foraminifera was investigated in six short cores (multicores, 30-32 cm total length) recovered from the central Red Sea. The ecological preferences as well as the relationship between the live and dead/fossil assemblages (preserved down-core) were examined. The sites, located along a W-E profile and between the depth of 366 and 1782 m, extend from the center of the oxygen minimum zone (OMZ, ~200-650 m), through its margin at ~600 m, and down to the well-aerated deep-water environment. Live (Rose-Bengal stained) and coexisting dead foraminifera were studied in the upper 5 cm of each of the sites, and the fossil record was studied down to ~32 cm. Q-mode Principal Component Analysis was used and four distinct foraminiferal fossil assemblages were determined. These assemblages follow different water mass properties. In the center of the OMZ, where the organic carbon content is highest and the oxygen concentration is lowest (<=0.5 ml O2/l), the Bolivina persiensis-Bulimina marginata-Discorbinella rhodiensis assemblage dominates. The slightly more aerated and lower organic-carbon-content seafloor, at the margin of the OMZ, is characterized by the Neouvigerina porrecta-Gyroidinoides cf. G. soldanii assemblage. The transitional environment, between 900-1200 m, with its well-aerated and oligotrophic seafloor, is dominated by the Neouvigerina ampullacea-Cibicides mabahethi assemblage. The deeper water (>1500 m), characterized by the most oxygenated and oligotrophic seafloor conditions, is associated with the Astrononion sp. A-Hanzawaia sp. A assemblage. Throughout the Red Sea extremely high values of temperature and salinity are constant below ~200 m depth, but the flux of organic matter to the sea floor varies considerably with bathymetry and appears to be the main controlling factor governing the distribution pattern of the benthic foraminifera. Comparison between live and the dead/fossil assemblages reveals a large difference between the two. Processes that may control this difference include species-specific high turnover rates, and preferential predation and loss of fragile taxa (either by chemical or microbial processes). Significant variations in the degree of loss of the organic-cemented agglutinants were observed down core. This group is preserved down to 5-10 cm at the shallow OMZ sites and down to greater depths at well-aerated and oligotrophic sites. The lower rate of disintegration of these forms, in the deeper locations of the Red Sea, may be related to low microbial activity. This results in the preservation of increasing numbers of organic-cemented shells down-core.

Stable isotope record of Uvigerina peregrina from the Mediterranean Sea

Based on the glacial to postglacial delta13C differences between endobenthic Uvigerina peregrina species from the Alboran basin and from other mediterranean basins, changes in the fertility of the western part of this basin during the last deglaciation are reconstructed. As a result of particulate organic carbon (POC) rain from the highly productive upwelling cell along the northwestern margin of the Alboran basin, U. peregrina is presently depleted by about 1.6per mil with respect to the measured delta13C values of bottom water SumCO2 and by about 0.9per mil with respect to specimens from other areas of the western Mediterranean or from the Gulf of Cadiz within the Mediterranean Outflow Water. The Uvigerina delta13C difference between the Alboran Sea and the Gulf of Cadiz (Delta delta13C), was close to 0per mil at the beginning of the last deglaciation and during the late glacial time. This suggests that highly fertile systems set in the Alboran Sea near 16 kyr B.P. Two rapid increases in the Delta delta13C offset are recorded near 15 kyr and 11 kyr B.P. Fluctuations around 1.1 to 1.2per mil occurred during the early Holocene, and a maximum was reached near 9 kyr B.P. After 4 kyr the Delta delta13C offset decreased to its present-day average value of 0.9per mil. Changes in the intensity of surficial production cannot account for all the observed fluctuations, especially in the early Holocene time. A strong decrease in the intermediate and deep water ventilation of the Alboran basin may have occurred near 8-9 kyr, in phase with the last stagnant phase in the eastern Mediterranean and the deposition of Sapropel S1. As a result, the redistribution and remineralization at depth of the produced organic matter was incomplete. The POC rain reaching the sediment was locally intensified and caused the lowering of the delta13C values of endobenthic foraminifers such as U. peregrina. The benthic 13C signal suggests that the difference between the Alboran Sea and the Gulf of Cadiz was at its maximum. At the same time, an important modification in the water masses structure may have occurred near 9-8 kyr B.P. The deepening of the permanent pycnocline probably related to a thicker Atlantic jet at a stage of high sea level stand is recorded by the replacement of the right coiling N. pachyderma dominance (coincident with a shallow pycnocline) by the G. inflata dominance (coincident with a deep pycnocline). Diatom abundances were strongly reduced indicating an important modification of the productive system. The glacial-postglacial evolution of productivity within the Alboran basin was therefore more complex than in the adjacent Atlantic Ocean and opposite to the global one which displays a general increase in productivity during glacial time. Although it is the global budget of paleoproductivity that would drive the partitioning of carbon within the ocean, local or regional discrepancies with the global glacial-interglacial model must be addressed. Local winds and regional atmospheric pressure systems, which are the forcing factors for circulation and exchange with the Atlantic, control the fertile systems of the Alboran basin.

Size distribution of Holocene planktic foraminifer assemblages

The size of any organism is influenced by the surrounding ecological conditions. In this study, we investigate the effects of such factors on the size spectra of planktic foraminiferal assemblages from Holocene surface sediments. We analyzed assemblages from 69 Holocene samples, which cover the major physical and chemical gradients of the oceans. On a global scale, the range of sizes in assemblages triples from the poles to the tropics. This general temperature-related size increase is interrupted by smaller sizes at temperatures characteristic of the polar and subtropical fronts, at 2°C and 17°C, respectively, as well as in upwelling areas. On a regional scale, surface water stratification, seasonality and primary productivity are highly correlated with the size patterns. Such environmentally controlled size changes are not only characteristic for entire assemblage, but also for the dominant single species.

(Table 2) Age determination of harp seals (Pagophilus groenlandicus) from the Barents Sea - comparison of methods

Lower jaws (containing the teeth), eyes, and skin samples were collected from harp seals (Pagophilus groenlandicus) in the southeastern Barents Sea for the purpose of comparing age estimates obtained by 3 different methods, the traditional technique of counting growth layer groups (GLGs) in teeth and 2 novel approaches, aspartic acid racemization (AAR) in eye lens nuclei and telomere sequence analyses as a proxy for telomere length. A significant correlation between age estimates obtained using GLGs and AAR was found, whereas no correlation was found between GLGs and telomere length. An AAR rate (k Asp) of 0.00130/year ± 0.00005 SE and a D-enantiomer to L-enantiomer ratio at birth (D/L 0 value) of 0.01933 ± 0.00048 SE were estimated by regression of D/L ratios against GLG ages from 25 animals (12 selected teeth that had high readability and 13 known-aged animals). AAR could prove to be useful, particularly for ageing older animals in species such as harp seals where difficulties in counting GLGs tend to increase with age. Age estimation by telomere length did not show any correlation with GLG ages and is not recommended for harp seals.

Meiofauna and the structure of the benthic community in the Iberian deep sea basin

1. On the cruises 3 and 15 of R.V. "Meteor" 6 grab samples, and 6 hauls with the 6 m Agassiztrawl were taken and at 2 stations the deep sea camera was lowered. This material gave quantitative results on the meiofauna and minimum counts of the macrofauna. 2. The nematodes constitute nearly 95% of the meiofauna, the copepoda only 2%. With increasing sediment depth the density of animals decrease gradually. In the uppermost centimeter of sediment 42.6% of the meiofauna are found while only 3.7% live in layer 6-7 cm. Meiofauna weight ranges from 0.6-5.7 mg/25 m**2 surface i.e. 0.24-2.8 g/m**2. 3. Mean numbers of individuals and weights show standard errors of 20-30 %. As an approximate average values for further considerations the weight of the meiofauna in the area was taken as 1 g/m**2 4. Quantitative information on the macrofauna is derived from the trawls and the photographs for the actinia Chitonanthus abyssorum only, which is found in the rate of 1 individual/36-72 m**2, but seems to be less abundant generally. 5. Animal density does not decrease steadily from nearshore to offshore biocoenoses, i.e. generally with increasing depth. The decrease is more pronounced for macro- than for meiofauna. For the deep sea the weight proportion of macrofauna : meiofauna is of the order of 1 : 1. 6. With the assumption, that adaptation of metabolism to deep sea conditions is similar in macro- and meiofauna total metabolism of invertebrates is ascribed to meiofauna to more than 80%. 7. The structure of the biocoenosis of the deep sea floor is characterized by the meiofauna living on and in the sediment and by the dominance of sediment feeders in the macrofauna. 8. Considering the large numbets and high partition rates of bacteria a comparative large part of the metabolism in the deep sea sediment must be ascribed to bacteria. This favours the hypothesis, that with increasing depth and decreasing addition of organic material to the sediment, the importance of meiofauna and microorganisms for total metabolism increases. 9. Considering the different modes of food transport to the deep sea environment, i.e. sinking of dead particles, transport by vertical migration of organisms, aggregation of organic particles, adsorption of dissoloved organic substance to inorganic particles, and heterotrophy, the sediment may be assumed to contain more food for invertebrates than the water above the bottom. 10. Suspensions feeders of macrofauna are fixed to hard substrates in the sediment surface. Some of them are shown to bend themselves down to the bottom in underwater photographs. This suggests the idea that some deep sea suspension feeders partly depend on food from the sediment surface, on which they feed directly.

Carbon analysis of surface sediments from the Laptev Se

During the past six years organic geochemical, micropaleontological, and sedimentological investigations were carried out within the framework of the multidisciplinary bilateral German-Russian research project ''System Laptev Sea'' and detailed biological investigations within the project ''German-Russian Investigations of the Marginal Seas of the Eurasian Arctic'', In order to understand the Laptev Sea ecosystem and to obtain information about sources and fate of organic carbon, the distribution of phyto- and zooplankton, diatoms, chlorophyll a benthic macrofauna, palynomorphs, grain size, total organic carbon, d13Corg and biomarkers (n-alkanes, fatty acids) were determined. In general, the influence of the major rivers draining into the Laptev Sea, is reflected in the water column as well as in the surface sediments. In both habitats three ecological provinces can be distinguished, i.e., the southeastern Laptev Sea, the central Laptev Sea, and the northern Laptev Sea. Additionally, clear differences between the western and the eastern Laptev Sea occur. The comparison of the different data sets of the water column and the surface sediments provide information about organic carbon sources and pathways in the Laptev Sea shelf and continental slope area.

Sea-surface temperature reconstruction and calibration-function of sediment cores from the Southern Ocean

We provide a new multivariate calibration-function based on South Atlantic modern assemblages of planktonic foraminifera and atlas water column parameters from the Antarctic Circumpolar Current to the Subtropical Gyre and tropical warm waters (i.e., 60°S to 0°S). Therefore, we used a dataset with the abundance pattern of 35 taxonomic groups of planktonic foraminifera in 141 surface sediment samples. Five factors were taken into consideration for the analysis, which account for 93% of the total variance of the original data representing the regional main oceanographic fronts. The new calibration-function F141-35-5 enables the reconstruction of Late Quaternary summer and winter sea-surface temperatures with a statistical error of ~0.5°C. Our function was verified by its application to a sediment core extracted from the western South Atlantic. The downcore reconstruction shows negative anomalies in sea-surface temperatures during the early-mid Holocene and temperatures within the range of modern values during the late Holocene. This pattern is consistent with available reconstructions.

Sea-surface reconstructions of the western Pacific Ocean during the last 5.3 million years

The late Neogene was a time of cryosphere development in the northern hemisphere. The present study was carried out to estimate the sea surface temperature (SST) change during this period based on the quantitative planktonic foraminiferal data of 8 DSDP sites in the western Pacific. Target factor analysis has been applied to the conventional transfer function approach to overcome the no-analog conditions caused by evolutionary faunal changes. By applying this technique through a combination of time-slice and time-series studies, the SST history of the last 5.3 Ma has been reconstructed for the low latitude western Pacific. Although the present data set is close to the statistical limits of factor analysis, the clear presence of sensible variations in individual SST time-series suggests the feasibility and reliability of this method in paleoceanographic studies. The estimated SST curves display the general trend of the temperature fluctuations and reveal three major cool periods in the late Neogene, i.e. the early Pliocene (4.7 3.5 Ma), the late Pliocene (3.1-2.7 Ma), and the latest Pliocene to early Pleistocene (2.2-1.0 Ma). Cool events are reflected in the increase of seasonality and meridional SST gradient in the subtropical area. The latest Pliocene to early Pleistocene cooling is most important in the late Neogene climatic evolution. It differs from the previous cool events in its irreversible, steplike change in SST, which established the glacial climate characteristic of the late Pleistocene. The winter and summer SST decreased by 3.3-5.4°C and 1.0 2.1C in the subtropics, by 0.9°C and 0.6C in the equatorial region, and showed little or no cooling in the tropics. Moreover, this cooling event occurred as a gradual SST decrease during 2.2 1.0 Ma at the warmer subtropical sites, while that at cooler subtropical site was an abrupt SST drop at 2.2 Ma. In contrast, equatorial and tropical western Pacific experienced only minor SST change in the entire late Neogene. In general, subtropics was much more sensitive to climatic forcing than tropics and the cooling events were most extensive in the cooler subtropics. The early Pliocene cool periods can be correlated to the Antarctic ice volume fluctuation, and the latest Pliocene early Pleistocene cooling reflects the climatic evolution during the cryosphere development of the northern hemisphere.