PALEOCEANOGRAPHY, VOL. 26, PA4215, doi: /2011pa002153, 2011

Transcription

1 PALEOCEANOGRAPHY, VOL. 26,, doi: /2011pa002153, 2011 Seasonal variation in the oxygen isotopic composition of different sized planktonic foraminifer Neogloboquadrina pachyderma (sinistral) in the northwestern North Pacific and implications for reconstruction of the paleoenvironment Azumi Kuroyanagi, 1 Hodaka Kawahata, 1 and Hiroshi Nishi 2 Received 11 April 2010; revised 15 September 2011; accepted 20 September 2011; published 29 November [1] The oxygen isotope value (d 18 O) of planktonic foraminifer shells is one of the most commonly used proxies to reconstruct the paleoenvironment. Neogloboquadrina pachyderma (sinistral) (N. pachyderma (sin.)) is a dominant foraminifer in subpolar and polar region that is used extensively for high latitude paleoreconstruction. The present study examined seasonal variation in d 18 OofN. pachyderma (sin.) using sediment trap samples collected over 3.5 years in the northwestern North Pacific Ocean. The vital offset value was about 1, which agrees with previous plankton tow studies. Large shell individuals ( mm) secrete their shells in a shallower and warmer environment than small ones ( mm), and the difference in d 18 O between shell size showed seasonal variation. Although large individuals calcify at a slower rate with slightly heavier d 18 O than small individuals, differences between size classes mainly reflect different habitation in the water column and are related to seasonal changes in stratification. During summer, large individuals would calcify mainly at m depth, while small individuals calcify near the pycnocline at 45 m. During winter, both size classes calcify at or slightly above the weak pycnocline at m. Thus, the shell d 18 O could record the oceanographic environment around pycnocline depth, and size specific differences in d 18 O reflect water column stratification. The annual flux weighted d 18 O corresponded to non flux weighted mean d 18 O value due to the equal contribution of two flux peaks and differences between size classes also related to the stratification. It suggests that fossil d 18 O data could represent the annual mean environment around pycnocline and past stratification. Citation: Kuroyanagi, A., H. Kawahata, and H. Nishi (2011), Seasonal variation in the oxygen isotopic composition of different sized planktonic foraminifer Neogloboquadrina pachyderma (sinistral) in the northwestern North Pacific and implications for reconstruction of the paleoenvironment, Paleoceanography, 26,, doi: /2011pa Introduction [2] Oceanographic conditions at high latitudes play an important role in the global ocean environment [Sarmiento et al., 2004], therefore, reconstruction of the paleoenvironment in this region is valuable. Planktonic foraminifera provide a record of the ocean surface environment through the isotopic and chemical composition of their calcite shells. Shell oxygen isotopic composition (d 18 O) is commonly used to reconstruct the paleoenvironment, including temperature, salinity, and water column structure [e.g., Rohling and Cooke, 1999; de Garidel Thoron et al., 2005]. Foraminifera in sediment trap samples are particularly useful for examining changes 1 Atmosphere and Ocean Research Institute, University of Tokyo, Kashiwa, Japan. 2 Tohoku University Museum, Tohoku University, Sendai, Japan. Copyright 2011 by the American Geophysical Union /11/2011PA in d 18 O over time, and for determining which seasons and depths are most represented. [3] Neogloboquadrina pachyderma (sinistral) (N. pachyderma (sin.)) (Ehrenberg) is a dominant species in subpolar and polar region [e.g., Hemleben et al., 1989] and, therefore, useful in high latitude paleooceanography. Sediment trap d 18 O data of N. pachyderma (sin.) (i.e., left coiling N. pachyderma) and correspondent detailed environmental data have been collected in the Southern Ocean [King and Howard, 2005] and the western North Atlantic [Jonkers et al., 2010]. However, no samples are available from the northwestern North Pacific, and year round records are rare. Furthermore, a vital offset value has not yet been reported for N. pachyderma (sin.) in the western North Pacific. The vital offset value is one of the most important parameters for reconstructing paleotemperatures, and Jonkers et al. [2010] hypothesized that the vital offset of N. pachyderma (sin.) would vary regionally. Although previous tow and sediment studies reported a size effect of N. pachyderma [e.g., 1of10

2 Figure 1. Major oceanographic features in the northwestern North Pacific Ocean. The triangle indicates the study site (Site 50N). Hillaire Marcel et al., 2004; Lončarić et al., 2006], sediment trap samples could provide the seasonal differences in d 18 O among different size classes. The influence of these differences on reconstruction of the paleoenvironment is not well understood. [4] The present study investigates the d 18 OofN. pachyderma (sin.) using sediment trap samples collected in the northwestern North Pacific over 3.5 years, with 2 weeks resolution. The data were used to determine the (1) vital offset value, (2) seasonal variation in differences between two size classes, (3) apparent calcification depths for each size class, and (4) flux weighted isotopic values, and all of this information is required to reconstruct the paleoenvironment. 2. Oceanographic Setting of the Study Sites [5] The subarctic North Pacific is one of the three major high nitrate, low chlorophyll regions of the world. There are four gyres in this region: the Okhotsk Sea Gyre, the Western Subarctic Gyre (WSG), the Bering Sea Gyre (Figure 1), and the Alaskan Gyre. Site 50N is located within the WSG, which is strongly cyclonic and elongated along the Kamchatka Peninsula and Kuril Islands. The subarctic or Oyashio front occurs at the southern edge of the WSG, near 41 N, 165 E [Joyce, 1987], and is characterized by mixing of the warm saline waters of the Kuroshio Extension with the cold and less saline waters of the Oyashio current, forming the subarctic water mass. The WSG is generally characterized by a sharp halocline/pycnocline at the bottom of the surface layer ( 100 m [Andreev et al., 2002]). Dichothermal layer (intermediate cold layer) and mesothermal layer (intermediate warm layer), which are closely associated with the halocline, are also observed at the subsurface in the WSG [Nagata et al., 1992; Andreev et al., 2002]. Cooling in the winter and relatively strong halocline (too fresh at the surface) result in preventing diapycnal vertical mixing and forming a dichothermal layer usually in summer (above the mesothermal layer) [Nagata et al., 1992; Andreev et al., 2002]. The mesothermal layer is formed by advection from the east or the south [Nagata et al., 1992]. During this study, the pycnocline was located between 25 and 65 m, a weak dichothermal layer was found at m, and no mesothermal layer was observed in the upper 305 m (Figure 2) [Behringer and Xue, 2004]. 3. Materials and Methods 3.1. Sampling and Isotopic Analysis [6] A time series sediment trap (McLane Mark 7G 21; opening, 0.5 m 2 ) was deployed at Site 50N (50 01 N, E, Figure 1) at a depth of 3260 m, approximately 2500 m above the seafloor. The sediment trap was moored for 3.5 years, during Each trap sample represents a collection period of or days, depending on the recovery/redeployment schedule (every 6 months). The collection cups were filled with filtered deep sea water containing a 5% formaldehyde solution buffered with sodium borate (ph >8) for poison and preservation. Detailed descriptions of sampling duration and particle flux are provided by Honda et al. [2002]. [7] Planktonic foraminifera were wet sieved into three size fractions (<125 mm, mm, and mm), dried, picked, and identified for faunal abundance analysis [Kuroyanagi et al., 2002, 2008]. Most of N. pachyderma (sin.) had encrusted shells. For isotope analysis, specimens of encrusted N. pachyderma (sin.) were picked again in the size fraction mm and mm from the mm size class. In the largest size fractions ( mm), N. pachyderma (sin.) were not observed. The picked specimens were cleaned in an ultrasonic bath, for one minute each for a total of four times, first with deionized distilled water, next with acetone, and finally the last two times with methanol to remove adhering, fine organic particles. [8] Stable isotope measurements of carbonates were conducted at the National Institute of Advanced Industrial 2of10

3 Figure 2. Potential density (sigma t) of upper 160 m during the sampling period, based on monthly salinity and temperature data from Global Ocean Data Assimilation System (GODAS). It shows little seasonal variation at m water depth (see Figure 3). Science and Technology, Tsukuba, Japan, on Micromass Optima and IsoPrime mass spectrometers equipped with an automated carbonate system. Isotopic data are reported relative to the Vienna PeeDee Belemnite (VPDB) standard, established via the NBS 19 calcite standard (U.S. National Bureau of Standards). The internal precision is ±0.03 and ±0.04 (1s) for d 13 C and d 18 O, respectively Estimation of Isotopic Equilibrium of Calcite [9] To compare the measured d 18 O values of foraminiferal shells with the isotopic equilibrium of calcite, the equation of O Neil et al. [1969] for low temperatures (down to 0 C), modified by Wefer and Berger [1991] for the 0 30 C range, was used to calculate equilibrium values: C ¼ 21:9 3:162ð31:06 þ tþ 0:5 þ W;VPDB ð1þ where t is temperature ( C) and d C and d W,VPDB are the d 18 O values of carbonate and ambient water, respectively, relative to VPDB. [10] d W values were converted from the Vienna Standard Mean Ocean Water (VSMOW) to the VPDB scale using a correction value of 0.20 [Bemis et al., 1998], as follows: W;VPDB ¼ 0:9998 W;VSMOW 0:20 where d W,VSMOW is d 18 O value of water on the VSMOW scale. [11] Although few isotopic data points are available for the western Pacific [Schmidt et al., 1999; Bigg and Rohling, 2000], LeGrande and Schmidt [2006] suggested the regional d 18 O salinity relationships in the North Pacific with root mean squared error of between 16 Geochemical Ocean Sections Study points: W;VSMOW ¼ 0:44 salinity 15:13 [12] Monthly temperature and salinity data at 26 depth levels in the upper 303 m were supplied for November 1997 to January 2001, based on a <1 1 gridded data set, by the Global Ocean Data Assimilation System (GODAS) ð2þ ð3þ 3of10

4 4of10 [Behringer and Xue, 2004], and GODAS depends on continuous real time data from the Global Ocean Observing System. Temperature and salinity data were used to calculate potential density (sigma t), using the equation from United Nations Educational, Scientific, and Cultural Organization [1981]. The hydrographic data were adjusted to sediment trap samples from the following month, to allow for survival and settling time between foraminiferal production in the photic zone and deposition in the trap [e.g., Sautter and Thunell, 1991; Eguchi et al., 1999]. Although settling time of foraminifera has not been well known yet [e.g., Takahashi and Bé, 1984; Honjo and Manganini, 1993; Gyldenfeldt et al., 2000], Honjo and Manganini [1993] reported that there was no offset of arrival time for foraminiferal shell fluxes among at 1000, 2000, and 3400 m depth traps in the North Atlantic. Since the sampling duration is about two weeks, and taking these fast settling speed into consideration, there would be no offset of arrival time between the shell size classes Calculation of Flux Weighted d 18 O Value [13] To determine the effect of the foraminiferal seasonal flux on the fossil record, the flux weighted d 18 O was calculated for each sampling year, using the following equation: flux-weighted 18 O ¼ Xn i¼1 flux small;i 18 O small;i þflux large;i 18 O large;i =total flux where flux small,i and flux large,i are shell flux of small and large individuals, and d 18 O small,i and d 18 O large,i are the d 18 O values of small and large individuals, respectively. We calculated the flux weighted isotopic value for small ( mm), large ( mm), and both small and large ( mm) size fraction. 4. Results 4.1. Seasonal and Size Specific Variation in d 18 O of N. pachyderma (sin.) [14] Shell d 18 O of small ( mm) and large ( mm) N. pachyderma (sin.) exhibited similar seasonal variation, with minimum values during September October and maximum values during April May (Figure 3). d 18 O ranged from 0.58 to 2.53 for small shells and from 0.52 to 2.27 for large shells (Table 1). Large shells generally had more depleted values than small shells, and the differences between size classes averaged 0.2 throughout the study (Table 2). The differences were greater during summer than winter (Figure 3). During December May, Figure 3. Oxygen isotopic values (d 18 O) of small ( mm, circles) and large ( mm, diamonds) N. pachyderma (sin.) and calculated calcite equilibrium d 18 O values at various depths. Each symbol includes the analytical error bars (±0.04 ). Open and shaded bars indicate small and large shell fluxes of N. pachyderma (sin.), respectively. Please note that equilibrium values correspond to water column conditions 1 month later to account for survival and settlement time. ð4þ

5 Table 1. Mean, Minimum, and Maximum Oxygen Isotopic Values (d 18 O) of Small ( mm) and Large ( mm) N. pachyderma (sin.) Collected in Sediment Traps During Small Shell ( mm) d 18 O( ) Large Shell ( mm) d 18 O( ) Period (Year) Minimum Maximum Mean ± SD Minimum Maximum Mean ± SD ± ± ± ± ± ± a ± ± 0.39 a No samples were available for the latter part of when the upper water column was not strongly stratified, the isotopic values of small and large shells were closely correlated (r 2 = 0.81; n = 35), and mean differences between the two size classes were only (Table 2). During June November, when the upper water column was well stratified, the correlation was less close (r 2 = 0.15; n = 23), and the mean differences between size classes were Furthermore, Student s t test also revealed a significant difference in d 18 O between shell size classes during stratified period (P < 0.001), while no significant differences during less stratified period (P > 0.5) Predicted d 18 O Calcite Equilibrium and Shell Fluxes [15] The predicted isotopic equilibrium d 18 O of calcite showed distinct seasonal variation at shallow depths, ranging from 0.28 to 3.26 at 5 m, and from 2.15 to 3.19 at 65 m (Figure 3). The pattern generally reflects that of N. pachyderma (sin.) d 18 O, with a maximum in April and a minimum around October. The predicted calcite equilibrium at greater depths ( m) exhibited seasonal variation of only , remaining around throughout the year. [16] N. pachyderma, Globigerina quinqueloba, Globigerina bulloides, and Globigerinita glutinata made up more than 95% of annual total foraminiferal fluxes (>125 mm) at Site 50N [Kuroyanagi et al., 2008]. N. pachyderma (sin.) accounted for 48%, 36%, 34%, and 33% of the total foraminiferal flux in 1998, 1999, 2000, and 2001, respectively. The shell flux of N. pachyderma (sin.) ranged from 74 to 9427 m 2 d 1 (mean, 1704 m 2 d 1 ), with two maxima in April and late November December (Figure 3). In 1998, a large peak also occurred in September. The autumn peak was associated with the destruction of summer stratification with higher water temperature, and high abundance in spring was associated with a well mixed water column and minimum sea surface temperature (SST), allowing nutrients to be supplied to surface water [Kuroyanagi et al., 2008]. These high shell fluxes of N. pachyderma (sin.) at this site could provide the analysis with higher time resolution than equatorial species. 5. Discussion 5.1. Estimation of the Vital Offset [17] The measured oxygen isotope values of foraminiferal shells often demonstrate disequilibrium from the values for inorganic calcite. The offset is species specific [e.g., Niebler et al., 1999] and is due to vital effects, including effects of symbiont photosynthesis, respiration, carbonate ion concentrations, and ontogenic and gametogenic calcification [Rohling and Cooke, 1999]. In the present study, foraminiferal d 18 O showed seasonal changes of (Figure 3), which were greater than changes predicted for d 18 O of inorganic calcite at deeper than 65 m (<0.3 ), suggesting that d 18 O of N. pachyderma (sin.) reflected environmental conditions above 65 m. During January April, when the upper water column was well mixed, equilibrium values were relatively constant from 5 to 65 m, so effects of foraminiferal vertical distribution on d 18 O should have been minimal during this period (Figure 3). In April, both shell and equilibrium d 18 O showed a maximum peak value, and shell d 18 O exhibited relatively stable value probably due to the high shell flux. Therefore, we defined the differences between measured d 18 OofN. pachyderma (sin.) and predicted equilibrium values in April as a vital offset. [18] Vital offset values were approximately 1 for both size classes throughout the study (Figure 3), except during 2000 ( 0.8 ). Jonkers et al. [2010] suggested 0 offset for N. pachyderma (sin.) (size, mm) collected in sediment traps in the western North Atlantic. Bauch et al. [1997, 2002] reported an offset of about 1, based on plankton tow samples from the Arctic Ocean (>160 mm) and Okhotsk Sea ( mm). Other plankton tow samples gave offset values estimates of 0.9 (encrusted individuals >150 mm in Table 2. Mean, Minimum, and Maximum Differences in Oxygen Isotopic Values (d 18 O) of Small and Large N. pachyderma (sin.) Collected in Sediment Traps During Difference of d 18 O( ) Stratified Period (Jun Nov) Less Stratified Period (Dec May) Period (Year) Minimum Maximum Mean ± SD Minimum Maximum Mean ± SD Minimum Maximum Mean ± SD ± ± ± ± ± ± ± ± ± a ± ± 0.12 a No samples were available for the latter part of of10

6 Figure 4. Differences between calcite equilibrium d 18 O values predicted using the equations of O Neil et al. [1969] and Kim and O Neil [1997]. Note that the differences increase with decreasing temperature. Shaded area shows the temperature range (5 10 C) that Jonkers et al. [2010] used to collect N. pachyderma, and it corresponded to difference. the Arctic Ocean [Simstich et al., 2003]) and 0.7 ( mm off California [Ortiz et al., 1996]). In contrast to plankton tow samples, sediment trap samples include dead individuals and, therefore, take into account ontogenic and gametogenic calcite. Nevertheless, the offset values determined for N. pachyderma (sin.) from sediment trap samples in the present study were closer to some of the previously reported values based on plankton tow samples than to the previous value based on sediment trap samples. [19] Differences among offset values estimated for the same species of foraminifera are partly due to the use of different equations to predict the equilibrium value of inorganic calcite [e.g., King and Howard, 2005]. The equation based on the equation by O Neil et al. [1969] used in the present study was also used by Bauch et al. [1997, 2002] and Simstich et al. [2003]. However, Jonkers et al. [2010] used the equation of Kim and O Neil [1997]. The former equation was based on measurements down to 0 C and gave higher equilibrium values, while the latter was based on measurements at C and gave lower values; the difference exceeded 0.5 at 0 C (Figure 4). Jonkers et al. [2010] collected foraminifera at water temperatures of 5 10 C. In this range, the difference in isotope values estimated by the two equations is (Figure 4). Thus, selection of the calcite equilibrium equation affects the estimated vital offset, particularly at low temperatures. [20] Differences in d 18 O values among genotypes or species of foraminifera could also affect the vital offset. Bauch et al. [2003] found a 0.5 difference between two right coiling (dextral) forms of N. pachyderma, [dextral form and dextral genetic] and [dextral form and sinistral genetic], in the sediment, which are attributed to N. pachyderma (sin.) and N. pachyderma (dextral) as two separate species [Bauch et al., 2003; Darling et al., 2006]. Darling et al. [2006] reported an aberrant sinistral morphotype of the dextral genotype south of 65 N in the North Atlantic, further complicating analyses. Jonkers et al. [2010] sampled 59 N in the North Atlantic (0 offset), while Bauch et al. [1997] and Simstich et al. [2003] concluded tow sampling at more northern area, >81 N and >68 N, respectively (both 1 offset). Although there are no genotype data for N. pachyderma in the northwestern North Pacific, the low right coiling ratio throughout the present study ( % [Kuroyanagi, 2006]) suggests that all individuals were probably the sinistral genetic [Darling et al., 2006]. Thus the N. pachyderma (sin.) examined in the present study and in previous plankton tow studies are assumed to be the sinistral genetic, and the offset value for this (sinistral form and sinistral genetic) would be about 1. Ivanova et al. [1999] reported a 0 offset for N. pachyderma (sin.), maybe dextral genetic, collected offshore Oman. Together, these results suggest that regional differences in the dominant genotype or species (i.e., [sinistral form and sinistral genetic] and [sinistral form and dextal genetic]) could result in different d 18 O offset values. An imprint of isotope offsets between genotypes may also be involved in the data pattern we observe in our case from the western North Pacific where the previous study indicated abrupt SST changes [Ishizaki, 2009] coinciding with increased abundance of dextral forms in the same core [Kuroyanagi et al., 2006]. Thus in the reconstruction of paleotemperature, offset values should Figure 5. The d 18 O and d 13 C values of small ( mm, circles) and large ( mm, diamonds) N. pachyderma (sin.). Effects of calcification rate, respiration, and symbiont photosynthesis (N. pachyderma does not have symbiont) are indicated by arrows. Red and blue lines show regression lines of large (r 2 =0.43)andsmall(r 2 = 0.57) individuals, respectively. 6of10

7 be examined from viewpoints of the equation used as oxygen isotope temperature relationship and the presence of genotypes Differences in d 18 O Between Different Size Classes [21] Shell d 18 O of foraminifera is size dependent [e.g., Bauch et al., 1997; Hillaire Marcel et al., 2004; Lončarić et al., 2006], and the size effect varies among species and regions [e.g., Niebler et al., 1999]. d 18 O differences among size classes are attributable to both environmental factors, such as calcification depth, and intrinsic factors, such as growth/calcification rate, respiration, and photosynthesis by symbionts. d 18 O differences between size classes of N. pachyderma (sin.) in the present study varied seasonally, not a consistent offset (Table 2 and Figure 3); suggesting that the differences were due mainly to oceanographic conditions in the water column rather than to size specific kinetic/metabolic effects. [22] In order to evaluate the kinetic/metabolic effects, we plotted the d 18 O and d 13 C data of all samples (Figure 5). McConnaughey [1989] reported that the isotopic ratio of kinetic effects would deplete both d 18 O and d 13 C in the proportion of 1:3 from isotopic equilibrium, while metabolic effects influence only d 13 C. Spero and Lea [1996] also showed similar ratio using nonsymbiont species G. bulloides, and McConnaughey et al. [1997] proposed it might be related to a kinetic effect. N. pachyderma (sin.) does not have symbionts [Hemleben et al., 1989], thus both calcification rate and respiration were considered as kinetic/metabolic effects in the present study (Figure 5). If respiration mainly affected the difference in d 18 O d 13 C relationship between shell sizes, slope of each regression line should be nearly zero as a respiration effect. However, slope of regression line of large and small shell was 1.43 and 1.26, respectively (Figure 5), and these values are relatively close to the opposite ratio (d 18 O:d 13 C = 3:1) of the expected kinetic effect. It suggests that large shell generally secreted their calcite shell more slowly with slightly higher value for both d 18 O and d 13 C (<0.1 ) than small one probably due to a kinetic effect (Figure 5), and it is accordance with previous tow observation on N. pachyderma (sin.) despite different size range [Bauch et al., 1997]. Bauch et al. [1997] reported that d 18 O values were consistently 0.14 higher in large (>250 mm) N. pachyderma (sin.) than in smaller individuals ( mm) collected in plankton tows. In that case, the difference would be mainly attributable to metabolic factors, not environmental effects 7of10 Figure 6. Oxygen isotopic values (d 18 O) of small ( mm, circles) and large ( mm, diamonds) N. pachyderma (sin.) and offset adjusted calcite equilibrium d 18 O values. Adjusted offset values were 1 except during 2000 ( 0.8 ). Each symbol includes the analytical error bars (±0.04 ). The purple, blue, and red lines exhibit the annual flux weighted value for all ( mm), small ( mm), and large ( mm) shell size classes, respectively. Open and shaded bars indicate small and large shell fluxes of N. pachyderma (sin.), respectively. Please note that equilibrium values correspond to water column conditions 1 month later to account for survival and settlement time.

8 Table 3. Flux Weighted d 18 O Values and Seasonal Peak Shell Fluxes for N. pachyderma (sin.) During Period (Year) All Shell ( mm) Flux Weighted d 18 O( ) Seasonal Peak Fluxes (Shells m 2 d 1 ) Small Shell ( mm) Large Shell ( mm) Spring Peak Autumn Peak Small Shell Large Shell Small Shell Large Shell because of the same sampling depth. Thus it confirmed the minor effect on kinetic effect on shell d 18 O (<0.1 ), and habitation would be the main determining factor for the different d 18 O among size Seasonal Variation in Main Calcification Depth of Different Size Classes [23] As size specific differences in d 18 OofN. pachyderma (sin.) from sediment trap samples primarily reflect differences in environmental conditions (section 5.2), the differences might provide information about the depth at which small and large individuals calcify, which is related to water column stratification (section 4.1 and Table 2). To determine the main calcification depth of large and small individuals, foraminiferal d 18 O values were compared to predicted calcite equilibrium values at each water depth in the upper 303 m of the water column, assuming an offset value of 1 except during 2000 ( 0.8 ) (Figure 6). During June to early November, when the water column was stratified, the calcification depth of large and small shells was estimated to be m and 45 m, respectively (Figure 6). On the other hand, in late November and December, both size reflected the 45 65m environments. Although foraminiferal d 18 O exhibited lighter value than equilibrium values at 5m depth during January March, the amplitude of shell d 18 O was consistent with that of equilibrium d 18 O at 45 65m (Figure 6). This inconsistency seemed to be caused by a delay (1 or 2 months) in foraminiferal d 18 O due to a life span extension in winter. Reproduction cycle of N. pachyderma has not been well known [e.g., Kucera, 2007], however, this species from Antarctic sea ice could be cultured for days at 1 C [Spindler, 1996]. Kimoto et al. [2003] also reported that N. pachyderma from the northern part of Honshu (Japan) were kept in culture for approximately 2 months, and some survived more than 3 months at the lowest temperature conditions (4.7 C). King and Howard [2005] and Jonkers et al. [2010] also observed this inconsistency of N. pachyderma (sin.) from trap samples. Consequently, during December May, when the upper water column was well mixed, d 18 O values of both size classes would correspond to conditions at m. [24] Previous studies reported that the vertical distribution of N. pachyderma (sin.) was closely related to the position of the pycnocline and/or thermocline [Ortiz et al., 1996; Hilbrecht, 1997; Simstich et al., 2003; Kuroyanagi and Kawahata, 2004]. The depth of pycnocline at Site 50N ranges from 25 to 45 m during the stratified period, and a weak pycnocline, if any, occurs around m during December June (Figure 2). Therefore, results of the present study suggest that N. pachyderma (sin.) mainly calcifies near the pycnocline during summer, and that calcification depth remains m in winter, even when there is no pycnocline or only a weak pycnocline at slightly greater depth. During the study period, water temperature ranged between 0.7 and 12.8 C at the surface and between 0.9 and 7.2 C at calcification depth [Behringer and Xue, 2004]. Our results suggest that d 18 O values of N. pachyderma (sin.) in sediments provide a record of temperature near or slightly above the pycnocline, reflecting more stable temperature conditions than species living at shallower depths. Furthermore, the d 18 O difference between shell sizes may record the strength of the pycnocline, and in conjunction with indices that provide surface temperature records, such as d 18 O and/or Mg/Ca values of shallow foraminifera or alkenone SST, d 18 O values of N. pachyderma (sin.) thus allow reconstruction of more detailed water column structure Reconstruction of Paleotemperatures Using Shell Isotope Data [25] Seasonal flux patterns of foraminifera have considerable impact on the estimation of paleotemperature, because foraminifera in the seasonal bloom constitute a substantial portion of the total fossil record. In the present study, N. pachyderma (sin.) exhibited large seasonal peaks in spring and autumn, and the cumulative fluxes of those peaks accounted for more than 80% of the total annual flux. Flux weighted d 18 O values estimated for were represented with seasonal peak fluxes (Table 3) and seasonal d 18 O value (Figure 6). The mean value from August 2000 to June 2001 was used as the annual average , owing to the lack of samples during the latter part of [26] Annual flux weighted d 18 O for all shell size ( mm) ranged from 1.3 to 1.7, and it corresponded to the equilibrium d 18 O value at m depth in August each year (Figure 6). Furthermore, mean annual values of small ( mm) individuals (non flux weighted) are in good agreement with the flux weighted value of small and all size fractions (Tables 1 and 3). This may be caused by compensating peak fluxes: the two seasonal flux peaks were consistent with the annual ranges of minimum and maximum water temperatures (Figure 6 and Table 3). On the other hand, large shell ( mm) fluxes in spring peak showed relatively higher value in ( shells m 2 d 1 ), and it resulted in the relatively higher value ( ) of fluxweighted d 18 O than non flux weighted one (Figure 6 and Tables 1 and 3). It suggests that it would be better to use the small ( mm) or all ( mm) shell size fractions for estimation of paleoenvironment in this case. Since the seasonal flux pattern and d 18 O variability each affect the sedimentary d 18 O signals, they should both be discussed when evaluating the paleoenvironment. 8of10

9 [27] During the experiment period, the difference in fluxweighted d 18 O between shell size classes exhibited the maximum value (0.3 ) in 2000, and it consisted with the actual stratification (Figure 2). Because a major disparity in seasonal d 18 O between shell size were observed in autumn flux peak, difference in flux weighted d 18 O between shell size could also record this water column situation. N. pachyderma (sin.) would calcify their shells around pycnocline depth, therefore, d 18 O data from fossil assemblages could reflect the annual mean environment around pycnocline, and the difference in d 18 O between shell size also record the past stratification situation. The two seasonal flux peaks were related to the structure of the water column [Kuroyanagi et al., 2008], thus changes in water column situation, including the pycnoclines might be one of the most important factors to consider when reconstructing paleotemperatures using foraminiferal d 18 O data. 6. Conclusion [28] Seasonal variation in d 18 O of a representative polar foraminifera, N. pachyderma (sin.), was examined using sediment trap samples collected in the northwestern North Pacific over 3.5 years. The offset value of N. pachyderma (sin.) [sinistral form and sinistral genetic] was around 1, which agrees with previous estimates from plankton tow samples. Although large individuals ( mm) calcify at a slower rate than small one ( mm), kinetic effects have a relatively small impact on shell d 18 O (<0.1 ), and habitation is or probably is the main determining factor for the different d 18 O among size. During summer, when the water column is stratified, large and small individuals appeared to mainly calcify near the pycnocline, at m and 45 m, respectively. During winter, when the water column is not stratified or only weakly stratified, both size classes calcified at or slightly above the pycnocline, at m. Because the seasonal peaks in flux coincide with minimum and maximum water temperatures and contribution of the flux peaks is approximately equal, flux weighted values of all ( mm) and small ( mm) shell size are in good agreement with the mean annual values (non flux weighted) of small ( mm) individuals. Thus, the fossil d 18 O data of N. pachyderma (sin.) most likely record the annual mean oceanographic environment around pycnocline depth, and size specific differences in d 18 O reflect water column stratification. [29] Acknowledgments. We express our appreciation to M. C. Honda and crew of the R/V Mirai who were on board during the cruise. We are also grateful to A. Suzuki, K. Minoshima, and N. Hokanishi for their support in analysis and to S. Obrochta for helpful suggestions. The manuscript has been improved with the help of constructive comments by R. Zahn, D. Bauch, T. de Garidel Thoron, and L. Jonkers. This study was carried out as part of the AORI and MEXT (Grant in Aid for Young Scientists (B), ). References Andreev, A., M. Kusakabe, M. Honda, A. Murata, and C. Saito (2002), Vertical fluxes of nutrients and carbon through the halocline in the western subarctic gyre calculated by mass balance, Deep Sea Res., Part II, 49, , doi: /s (02)00200-x. Bauch, D., J. Carstens, and G. Wefer (1997), Oxygen isotope composition of living Neogloboquadrina pachyderma (sin.) in the Arctic Ocean, Earth Planet. Sci. Lett., 146(1 2), 47 58, doi: /s x(96) Bauch, D., H. Erlenkeuser, G. Winckler, G. Pavlova, and J. Thiede (2002), Carbon isotopes and habitat of polar planktic foraminifera in the Okhotsk Sea: The carbonate ion effect under natural conditions, Mar. Micropaleontol., 45, 83 99, doi: /s (02) Bauch, D., K. Darling, J. Simstich, H. A. Bauch, H. Erlenkeuser, and D. Kroon (2003), Palaeoceanographic implications of genetic variation in living North Atlantic Neogloboquadrina pachyderma, Nature, 424, , doi: /nature Behringer, D. W., and Y. Xue (2004), Evaluation of the global ocean data assimilation system at NCEP: The Pacific Ocean, paper presented at 8th Symposium on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface, Am. Meteorol. Soc., Seattle, Wash., Jan. Bemis, B. E., H. J. Spero, J. Bijma, and D. W. Lea (1998), Reevaluation of the oxygen isotopic composition of planktonic foraminifera: Experimental results and revised paleotemperature equations, Paleoceanography, 13, , doi: /98pa Bigg, G. R., and E. J. Rohling (2000), An oxygen isotope data set for marine water, J. Geophys. Res., 105, , doi: /2000jc Darling, K. F., M. Kucera, D. Kroon, and C. M. Wade (2006), A resolution for the coiling direction paradox in Neogloboquadrina pachyderma, Paleoceanography, 21, PA2011, doi: /2005pa de Garidel Thoron, T., Y. Rosenthal, F. Bassinot, and L. Beaufort (2005), Stable sea surface temperatures in the western Pacific warm pool over the past 1.75 million years, Nature, 433, , doi: /nature Eguchi, N., H. Kawahata, and A. Taira (1999), Seasonal response of planktonic foraminifera to the ocean surface condition: Sediment trap results from the central North Pacific Ocean, J. Oceanogr., 55, , doi: /a: Gyldenfeldt, A. B. v., J. Carstens, and J. Meincke (2000), Estimation of the catchment area of a sediment trap by means of current meters and foraminiferal tests, Deep Sea Res., Part II, 47, , doi: / S (00) Hemleben, C., M. Spindler, and O. R. Anderson (1989), Modern Planktonic Foraminifera, Springer, New York, doi: / Hilbrecht, H. (1997), Morphologic gradation and ecology in Neogloboquadrina pachyderma and N. dutertrei (planktic foraminifera) from core top sediments, Mar. Micropaleontol., 31, 31 43, doi: /s (96) Hillaire Marcel, C., A. de Vernal, L. Polyak, and D. Darby (2004), Sizedependent isotopic composition of planktic foraminifers from Chukchi Sea vs. NW Atlantic sediments Implications for the Holocene paleoceanography of the western Arctic, Quat.Sci.Rev., 23, , doi: /j.quascirev Honda, M. C., K. Imai, Y. Nojiri, F. Hoshi, T. Sugawara, and M. Kusakabe (2002), The biological pump in the northwestern North Pacific based on fluxes and major components of particulate matter obtained by sedimenttrap experiments ( ), Deep Sea Res., Part II, 49, , doi: /s (02) Honjo, S., and S. J. Manganini (1993), Annual biogenic particle fluxes to the interior of the North Atlantic Ocean; studied at 34 N 21 W and 48 N 21 W, Deep Sea Res., Part II, 40, , doi: / (93) K. Ishizaki, Y. (2009), Paleoceanography of the northwestern North Pacific in response to global climatic change during glacial interglacial times, Ph.D. dissertation, Tohoku Univ., Sendai, Japan. Ivanova,E.M.,S.M. H. Conan, F. J. C. Peeters, and S. R. Troelstra (1999), Living Neogloboquadrina pachyderma sin and its distribution in the sediments from Oman and Somalia upwelling areas, Mar. Micropaleontol., 36, , doi: /s (98) Jonkers, L., G. J. A. Brummer, F. J. C. Peeters, H. M. van Aken, and M. F. De Jong (2010), Seasonal stratification, shell flux, and oxygen isotope dynamics of left coiling N. pachyderma and T. quinqueloba in the western subpolar North Atlantic, Paleoceanography, 25, PA2204, doi: / 2009PA Joyce, T. M. (1987), Hydrographic sections across the Kuroshio extension at 165 E, Deep Sea Res., Part A, 34, , doi: / (87) Kim, S. T., and J. R. O Neil (1997), Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates, Geochim. Cosmochim. Acta, 61(16), , doi: /s (97) Kimoto, K., X. Xu, N. Ahagon, H. Nishizawa, and Y. Nakamura (2003), Culturing protocol for living calcareous plankton: Preliminary results of the culturing experiment, Rep. Jpn. Mar. Sci. Tech. Cent., 48, King, A. L., and W. R. Howard (2005), d 18 O seasonality of planktonic foraminifera from Southern Ocean sediment traps: Latitudinal gradients and 9of10

10 implications for paleoclimate reconstructions, Mar. Micropaleontol., 56, 1 24, doi: /j.marmicro Kucera, M. (2007), Planktonic foraminifera as tracers of past oceanic environments, in Proxies in Late Cenozoic Paleocenaography, Dev. in Mar. Geol. Ser., vol. 1, edited by C. Hillaire Marcel and A. de Vernal, pp , Elsevier, Amsterdam, doi: /s (07) Kuroyanagi, A. (2006), Spatial and seasonal distribution of planktonic foraminifera Neogloboquadrina pachyderma in the northwestern North Pacific and reconstruction of paleoceanography off Shimokita, Fossils, 80, Kuroyanagi, A., and H. Kawahata (2004), Vertical distribution of living planktonic foraminifera in the seas around Japan, Mar. Micropaleontol., 53, , doi: /j.marmicro Kuroyanagi, A., H. Kawahata, H. Nishi, and M. C. Honda (2002), Seasonal changes in planktonic foraminifera in the northwestern North Pacific Ocean: Sediment trap experiments from subarctic and subtropical gyres, Deep Sea Res., Part II, 49, , doi: /s (02) Kuroyanagi, A., H. Kawahata, H. Narita, K. Ohkushi, and T. Aramaki (2006), Reconstruction of paleoenvironmental changes based on the planktonic foraminiferal assemblages off Shimokita (Japan) in the northwestern North Pacific, Global Planet. Change, 53, , doi: / j.gloplacha Kuroyanagi, A., H. Kawahata, H. Nishi, and M. C. Honda (2008), Seasonal to interannual changes in planktonic foraminiferal assemblages in the northwestern North Pacific: Sediment trap results encompassing a warm period related to El Niño, Palaeogeogr. Palaeoclimatol. Palaeoecol., 262, , doi: /j.palaeo LeGrande, A. N., and G. A. Schmidt (2006), Global gridded data set of the oxygen isotopic composition in seawater, Geophys.Res.Lett., 33, L12604, doi: /2006gl Lončarić, N., F. J. C. Peeters, D. Kroon, and G. J. A. Brummer (2006), Oxygen isotope ecology of recent planktic foraminifera at the central Walvis Ridge (SE Atlantic), Paleoceanography, 21, PA3009, doi: /2005pa McConnaughey, T. (1989), 13 C and 18 O isotopic disequilibrium in biological carbonates: I. Patterns, Geochim. Cosmochim. Acta, 53, , doi: / (89) McConnaughey, T., J. Burdett, J. F. Whelan, and C. K. Paull (1997), Carbon isotope in biological carbonates: Respiration and photosynthesis, Geochim. Cosmochim. Acta, 61, , doi: /s (96) Nagata, Y., K. Ohtani, and M. Kashiwai (1992), Subarctic gyre in the North Pacific Ocean, Oceanogr. Jpn., 1(3), Niebler, H. S., H. W. Hubberten, and R. Gersonde (1999), Oxygen isotope values of planktic foraminifera: A tool for the reconstruction of surface water stratification, in Use of Proxies in Paleoceanography: Examples From the South Atlantic, edited by G. Fischer and G. Wefer, pp , Springer, New York. O Neil, J. R., R. N. Clayton, and T. K. Mayeda (1969), Oxygen isotope fractionation in divalent metal carbonates, J. Chem. Phys., 51, , doi: / Ortiz, J. D., A. C. Mix, W. Rugh, J. M. Watkins, and R. W. Collier (1996), Deep dwelling planktonic foraminifera of the northeastern Pacific Ocean reveal environmental control of oxygen and carbon isotopic disequilibria, Geochim. Cosmochim. Acta, 60, , doi: /s (96) Rohling, E. J., and S. Cooke (1999), Stable oxygen and carbon isotopes in foraminiferal carbonate shells, in Modern Foraminifera, edited by B. K. Sen Gupta, pp , Kluwer Acad., Boston, Mass. Sarmiento, J. L., N. Gruber, M. A. Brzezinski, and J. P. Dunne (2004), High latitude controls of thermocline nutrients and low latitude biological productivity, Nature, 427, 56 60, doi: /nature Sautter, L. R., and R. C. Thunell (1991), Planktonic foraminiferal response to upwelling and seasonal hydrographic conditions: Sediment trap results from San Pedro Basin, southern California Bight, J. Foraminiferal Res., 21, , doi: /gsjfr Schmidt, G. A., G. R. Bigg, and E. J. Rohling (1999), Global Seawater Oxygen 18 Database, NASA Goddard Inst. of Space Stud., New York. Simstich, J., M. Sarnthein, and H. Erlenkeuser (2003), Paired d 18 O signals of Neogloboquadrina pachyderma (s) and Turborotalita quinqueloba show thermal stratification structure in Nordic Seas, Mar. Micropaleontol., 48, , doi: /s (02) Spero, H. J., and D. W. Lea (1996), Experimental determination of stable isotope variability in Globigerina bulloides: Implications for paleoceanographic reconstructions, Mar. Micropaleontol., 28, , doi: / (96) Spindler, M. (1996), On the salinity tolerance of the planktonic foraminifera Neogloboquadrina pachyderma from Antarctic sea ice, Proc. NIPR Symp. Polar Biol., 9, Takahashi, T., and A. W. H. Bé (1984), Planktonic foraminifera: Factors controlling sinking speeds, Deep Sea Res., Part A, 31, , doi: / (84) United Nations Educational, Scientific, and Cultural Organization (1981), Tenth report of the Joint Panel on Oceanographic Tables and Standards, Sidney, B. C., Canada, September 1980, UNESCO Tech. Pap. Mar. Sci., 36, 25 pp., Paris. Wefer, G., and W. H. Berger (1991), Isotope paleontology: Growth and composition of extant calcareous species, Mar. Geol., 100, , doi: / (91)90234-u. H. Kawahata and A. Kuroyanagi, Atmosphere and Ocean Research Institute, University of Tokyo, Kashiwanoha, Kashiwa, Chiba , Japan. (a kuroyanagi@aori.u tokyo.ac.jp) H. Nishi, Tohoku University Museum, Tohoku University, 6 3 Aramaki Aza Aoba, Aoba, Sendai, Miyagi , Japan. 10 of 10