U–Th dating of carbonate nodules from methane seeps off Joetsu, Eastern Margin of Japan Sea

Earth and Planetary Science Letters 272 (2008) 89–96

Contents lists available at ScienceDirect

Earth and Planetary Science Letters j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e p s l

U–Th dating of carbonate nodules from methane seeps off Joetsu, Eastern Margin of Japan Sea Yumiko Watanabe a,⁎, Shun'ichi Nakai b, Akihiro Hiruta c, Ryo Matsumoto c, Kunio Yoshida d a

Division of Earth and Planetary Sciences, Graduate School of Science, Kyoto University, Kyoto 606-8502, Japan Earthquake Research Institute, University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan Department Earth and Planetary Science, University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan d University Museum, University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan b c

A R T I C L E

I N F O

Article history: Received 8 June 2007 Received in revised form 11 April 2008 Accepted 14 April 2008 Available online 23 April 2008 Editor: H. Elderfield Keywords: U–Th radioactive disequilibrium dating ICP-MS carbonate methane seep

A B S T R A C T We performed U–Th radioactive disequilibrium analyses of carbonate nodules and sediment samples recovered from methane seep sites off Joetsu, of the eastern margin of Japan Sea, to decipher the active period of the methane seep. The carbonates contain 230Th, part of which is located in detritus such as silicate and organics, at the time of precipitation. The initial 230Th renders accurate dating with U–Th radioactive disequilibrium method difficult. We assessed the feasibility of correction using radioactive disequilibrium data of ambient sediment to overcome this difficulty. A (230Th/232Th)–(234U/232Th) isochron drawn by three chips divided from a carbonate nodule (PC05-04-50) passed through data points of local sediments. We conclude that the problem of initial 230Th can be resolved by measurements of local sediments. Results show that carbonate nodules include local sediment as impurities. Furthermore, the results of trace element analyses such as Rb, Zr, Nb, REE, Pb, and Th also support the idea. In all, 18 carbonate samples were dated with correction of initial 230Th using the mean value of local sediment in this study. The U–Th correction ages show 12–35ka with an isochron age of 26 ± 3ka. Results indicate that during the time interval of U–Th ages, from 12ka to 35ka, environmental conditions must have been favorable for enhanced methane flux through sediment. The extensive methane flow period at 20ka accords with the lowest-stand sea level during the last glacial age. Results of this study also suggest that U– Th ages of carbonate are useful as a reliable chronometer with regard to methane seep activation. In order to acquire U–Th ages of carbonate at methane seep sites, however, it is important to evaluate the amount of initial 230Th accurately using the value of sediment. © 2008 Elsevier B.V. All rights reserved.

1. Introduction Methane seeps have been observed in many different settings around the world (e.g., Kulmn et al., 1986; Sakai et al., 1992; Aharon et al., 1992). Methane is a strong greenhouse gas, which might strongly affect global climate (e.g., Sloan et al., 1992; Maslin and Thomas, 2003). For that reason, it is critical to elucidate the activation of methane seeps as follows. (1) When was methane released from the seafloor? (2) Is venting from these methane seeps continuous or discontinuous? (3) What geological factors influence methane seep activity?

⁎ Corresponding author. E-mail addresses: [email protected] (Y. Watanabe), [email protected] (S. Nakai). 0012-821X/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.epsl.2008.04.012

Methane seep sites are often associated with authigenic carbonate such as crusts, chimneys and concretion. Carbonate precipitation is triggered by increased alkalinity during anaerobic oxidation of methane via sulfate reduction, as follows (Aharon et al., 1992; Bohrmann et al., 1998). − − CH4 þ SO2− 4 →HCO3 þ HS þ H2 O:

Therefore, authigenic carbonate is a useful chronometer that reflects the time at which methane seep was active. Several attempts have been made at 14C dating of carbonate samples related to methane seeps (e.g., Paull et al., 1989). Carbonate related to methane seep contains dead carbon (Paull et al., 1989; Aharon et al., 1997; Peckmann et al., 2001; Gulin et al., 2003). Therefore, Paull et al. (1989) have reported the obtained 14C ages as maximum ages. Aharon et al. (1997) used 14C in conjunction with δ13C in order to assess the proportion of oxidized thermogenic methane, and only the U/Th age was used for the determination of carbonate formation. Also, Gulin et al. (2003) assessed the amount of dead carbon using stable carbon isotope ratio in order to calculate more appropriate 14C ages.

90

Y. Watanabe et al. / Earth and Planetary Science Letters 272 (2008) 89–96

Radioactive disequilibrium dating of the 238U decay series has been applied to numerous fields such as paleoclimatology, marine geochemistry, hydrology, archaeology, and volcanology (Bourdon et al., 2003). Under oxidizing conditions, U is soluble in natural water as uranyl ion, whereas Th has extremely low solubility. Carbonate precipitated from water usually contains U and rarely Th (e.g. b 1ppb in coral). Hence, the age of carbonate can be determined by measuring the increase of 230Th decaying from 234U, from zero at precipitation to equilibrium, providing that a U–Th system of the carbonate mineral is kept closed after precipitation (Bourdon et al., 2003). However, only a few attempts have been made to determine the ages of carbonates related to methane seeps using U–Th radioactive disequilibrium. The only studies known to us are those of Lalou et al. (1992), Aharon et al. (1997), and Teichert et al. (2003). Why U–Th dating study of carbonate at methane seep areas has not been carried out arises from the fact that those carbonates often contain non-negligible amounts of initial 230Th derived from detritus such as silicate and organic material. The presence of the initial 230Th hampered accurate age determination of carbonate. Lalou et al. (1992) reported that carbonate cements and chimneys at the Nankai accretionary prism have Th concentration of 0.8–4.5ppm. They reported 230Th–234U ages of the carbonates as from 20 to N 350ka, without making corrections of the initial 230Th. Aharon et al. (1997) determined Th concentrations of carbonate cements in the Gulf of Mexico of 0.2 to 1.0ppm, and presented corrected ages using the (230Th/232Th) activity ratio of local sediment. Teichert et al. (2003) reported that authigenic carbonates at Hydrate Ridge have Th con-

centration of less than 50ppb. They calculated corrected ages using the value of the residue that remained after a sample was leached by acetic acid. Subsequent studies have not confirmed whether local sediments shared the same (230Th/232Th) with carbonate. Aside from the results of Teichert et al., who used less-contaminated carbonate, it remains uncertain whether the corrected ages are accurate. In this study, we tackled the problem using an isochron of a carbonate nodule at methane seep, aiming at testing the feasibility of judging whether the age of carbonate can be corrected accurately by the value of sediment and deciphering the timing of the methane seep based on U–Th dating of carbonate. As a case study, we selected carbonate nodules and sediments recovered near active methane seeps off Joetsu, the eastern margin of Japan Sea. In this paper, two methods were examined to estimate the amount of initial 230Th. The first one is the isochron method. Another method is correction of initial 230Th in a carbonate with the mean value of sediments recovered near the carbonate. Comparing the two approaches, the accuracy of U–Th dating and the geological implications will be discussed. 2. Sample descriptions and geological setting The studied site is located at the 900-m depth: the sedimentary basin off Joetsu, on the Eastern Margin of Japan Sea (Fig. 1A). Offshore of Joetsu is the boundary between the Eurasia Plate and North American Plate. During summer 2004, the R&T/V Umitaka-maru of Tokyo University of Marine Science and Technology investigated a

Fig. 1. (A) Geological setting of Japan. The star represents the methane seep area off Joetsu. (B) Detailed map of study area. Black crosses show the positions of piston core and grab samplings in 2004 cruise.

Y. Watanabe et al. / Earth and Planetary Science Letters 272 (2008) 89–96

small spur off Joetsu, as shown in Fig. 1B. As a result of the 2004 cruise, detailed bathymetric and seismic profiles revealed numerous mounds (20–40-m high and 300–500-m across) and pockmarks (40–70-m deep and 300–500-m across) (Matsumoto et al., 2005). Large pockmarks extend to the direction of NNE–SSW, parallel to the strain-belt between the Eurasian and North American Plates. Methane seeps were detected using high-methane seawater off Joetsu (Ishida et al., 2005) and by a high-resolution echo-sounder survey (Aoyama et al., 2005). A piston core that was taken at PC15 recovered gas hydrate (Matsumoto et al., 2005; Hiruta et al., 2005). The gas hydrate is dominated by methane with minor ethane, whereas the δ13C of gas hydrate methane is − 40 to − 42‰PDB (Matsumoto et al., 2005). According to deep-exploratory drilling near the survey area, deep-seated gases are dominated by thermogenic methane with − 40‰PDB of δ13C (Matsumoto et al., 2005). Therefore, methane seep and ocean floor gas hydrate are likely to be connected with deepthermogenic gas reservoirs off Joetsu. Samples for U–Th dating in this study were recovered by piston cores and grabs sampled during the cruise of 2004 summer. Sampling locations were at PC05, PC15, and G3 for carbonate nodules, and PC04, PC05, and PC11 for sediment (Fig. 1B). A sample list is shown in Table 1. The gray to brown carbonate nodules are a few to ten centimeters in diameter. Powder X-ray analyses show that carbonates from PC05 comprise calcite, and those from PC15 and G03 are aragonite. Carbonate nodules are likely to contain minor amounts of quartz and feldspar as well, as identified by their tiny peaks. Secondary electron microprobe (SEM) and microscopic observation revealed that PC05 samples are microcrystalline calcite, less than 10μm, whereas PC15 and G03 samples consist of a few-hundreds-micrometer sized needleshaped aragonite. Because carbonate nodules contain detritus such as silicate and organics, and because pure carbonate cannot be separated physically from detritus, carbonate nodules were analyzed using total sample digestion. In this report, sediment denotes silty clay, which is present in piston cores and is dark gray to dark brown. X-ray diffraction patterns of sediment recognized quartz and minor amounts of feldspar.

Table 1 Sample list used in this study Sample ID

Location

Depth

Dominant minerals

91

3. Analytical method Watanabe and Nakai (2007) reported an accurate and precise analytical procedure for U–Th radioactive disequilibrium of carbonate rock samples. The procedure utilizes commercially available U and Th reagents as spikes for isotope dilution analysis with a multiple-collector inductively coupled mass spectrometer (MC-ICPMS). For this study, we analyzed carbonate nodules from methane seeps off Joetsu. Sediment samples in same area are also analyzed in order to estimate the amount of initial 230Th in carbonate nodules. Because we must analyze samples with different matrices, we modified the method of Watanabe and Nakai (2007) as follows. 3.1. Sample preparation The sample size was about 100mg for carbonate nodules, and 50mg for sediment. Carbonate samples were digested using 7M HNO3. Brownish-grey residue, observed after acid digestion, was digested sequentially using HF/HClO4 and HCl/H3BO3. Sediment samples were dissolved completely using HF/HClO4/HNO3 and HCl/ H3BO3. Then, sample solutions were dried and dissolved using 7M HNO3 and they were subsequently divided into four aliquots for isotope dilution analyses, isotope measurements, and a quadrupole mass analyzer (PQ3; Thermo Elemental) measurement in these proportions: one-tenth was used for Th isotope dilution analysis; one-twentieth was used for U isotope dilution analysis; one-fifth was used for isotope ratio measurements; also, one-twentieth was used for PQ3. The U and Th spikes, which were prepared using 230 Th–235U depleted reagents (see Watanabe and Nakai, 2007 for details), were added respectively to the two aliquots for abundance measurements. The digested sample was separated and purified for Th using AG1-X8 resin (Bio-Rad Laboratories Inc., see Fukuda and Nakai, 2002, for details). Then U/TEVA resin (Eichrom Technologies Inc.) was used for U purification following the method of Yokoyama et al. (1999). These separations were carried out for both spiked and unspiked aliquots of sample solutions. The Th and U recoveries were greater than 80%. Total blanks through the chemical procedure were b 14pg for U and b 11pg for Th, respectively. The recovered spiked and unspiked solutions were dried and dissolved in 2% HNO3. These isotopic compositions were measured using an MC-ICP-MS (IsoProbe; GV instruments Ltd.).

(cmbsf) PC05-02-60 PC05-04-50(1) PC05-04-50(2) PC05-04-50(3) PC05-04-65 PC05-CC(1) PC05-CC(2) PC05-CC(3) PC15-04-00 PC15-04-30 PC15-06-1(b) PC15-06-1(c) PC15-06-2(1) PC15-06-2(2) PC15-06-2(3) PC15-06-3(1) PC15-06-3(2) G3 PC04-02-100 PC04-03-100 PC04-04-100 PC05-05-100 PC11-03-100 PC11-04-100 PC11-05-100 PC11-06-60

PC05 PC05 PC05 PC05 PC05 PC05 PC05 PC05 PC15 PC15 PC15 PC15 PC15 PC15 PC15 PC15 PC15 G3 PC04 PC04 PC04 PC05 PC11 PC11 PC11 PC11

−67.5 −239.5 −239.5 −239.5 −254.5 −374 −374 −374 − 15.0 − 50.0 −234.5 −234.5 −234.5 −234.5 −234.5 −234.5 −234.5 0 −77 −180 −280 −354 − 54 −154 −253 −313

Calcite Calcite Calcite Calcite Calcite Calcite Calcite Calcite Aragonite Aragonite Aragonite Aragonite Aragonite Aragonite Aragonite Aragonite Aragonite Aragonite Quartz, plagioclase Quartz, plagioclase Quartz, plagioclase Quartz, plagioclase Quartz, plagioclase Quartz, plagioclase Quartz, plagioclase Quartz, plagioclase

3.2. Mass spectrometry For U and Th isotope dilution analysis and isotope analysis, we used a single focusing magnetic Sector MC-ICP-MS, IsoProbe. Samples were introduced into the spectrometer via a Cetac Aridus desolvating nebulizer. Watanabe and Nakai (2007) give a detailed explanation of that device and operating parameters. Using 230Th–235U depleted spike results in a larger error multiplication factor in abundance measurements when the spike/ sample ratio is inadequate, compared with enriched spikes (Fukuda and Nakai, 2002). It is crucial to add an appropriate amount of 230 Th–235U spike to avoid this. We estimated the concentrations of U and Th in rock samples using an ICP-MS with a quadrupole mass analyzer, PQ3, before undertaking dilution analyses. In addition, Rb, Zr, Nb, REE, and Pb were also measured with U and Th. Internal standards of indium (In) and bismuth (Bi) were used in order to correct drift and matrix effects. The results of Th and U concentrations using PQ3 agree with the result obtained using isotope dilution technique. Hereafter, the data of U and Th concentrations represent only those results with isotope dilution technique. The Rb, Zr, Nb, REE, and Pb concentrations are measured with PQ3. Analytical errors of concentration are less than 10% (2σ) for La, Ce,

92

Y. Watanabe et al. / Earth and Planetary Science Letters 272 (2008) 89–96

Table 2 U and Th isotopic compositions of carbonate nodule and sediment samples Sample ID

Carbonate PC05-02-60 PC05-04-50(1) PC05-04-50(2) PC05-04-50(3) PC05-04-65 PC05-CC(1) PC05-CC(2) PC05-CC(3) PC15-04-00 PC15-04-30 PC15-06-1(b) PC15-06-1(c) PC15-06-2(1) PC15-06-2(2) PC15-06-2(3) PC15-06-3(1) PC15-06-3(2) G3

U ± 2σ

Th ± 2σ

(ppb)

(ppb)

3626 ± 5 7585 ± 18 9733 ± 23 6715 ± 9 10,183 ± 15 8935 ± 23 9005 ± 14 8542 ± 23 5510 ± 7 4516 ± 12 7327 ± 7 11,480 ± 16 4314 ± 5 3769 ± 12 4420 ± 13 3418 ± 5 3658 ± 4 2944 ± 7

2629 ± 41 2689 ± 46 2696 ± 46 2660 ± 32 2459 ± 32 1746 ± 18 1759 ± 18 1644 ± 17 2782 ± 32 2707 ± 58 1999 ± 23 2581 ± 27 1542 ± 30 1564 ± 21 1598 ± 19 1926 ± 27 1933 ± 24 1154 ± 19

4376 ± 6 5468 ± 7 3327 ± 3 4581 ± 7 3794 ± 4 3615 ± 4 4981 ± 15 5513 ± 17

8678 ± 160 9689 ± 186 8780 ± 162 9752 ± 198 8815 ± 145 10093 ± 167 9109 ± 148 9301 ± 144

Sediment PC04-02-100 PC04-03-100 PC04-04-100 PC05-05-100 PC11-03-100 PC11-04-100 PC11-05-100 PC11-06-60 a

(234U/232Th) ± 2σa

(230Th/232Th) ± 2σa

(234U/238U) ± 2σa

4.77 ± 0.2 9.96 ± 0.3 12.78 ± 0.4 8.92 ± 0.3 14.61 ± 0.4 18.31 ± 0.5 18.26 ± 0.5 18.41 ± 0.5 6.97 ± 0.2 5.85 ± 0.2 13.19 ± 0.4 15.81 ± 0.4 9.93 ± 0.3 8.64 ± 0.3 9.94 ± 0.3 6.21 ± 0.2 6.75 ± 0.2 8.96 ± 0.3

1.65 ± 0.03 2.82 ± 0.04 3.40 ± 0.05 2.56 ± 0.04 6.83 ± 0.11 4.06 ± 0.06 4.08 ± 0.06 4.12 ± 0.06 2.59 ± 0.04 1.84 ± 0.03 3.22 ± 0.05 3.81 ± 0.06 1.99 ± 0.03 1.90 ± 0.03 2.20 ± 0.03 1.87 ± 0.03 2.00 ± 0.03 3.13 ± 0.05

1.109 ± 0.001 1.132 ± 0.002 1.134 ± 0.009 1.132 ± 0.009 1.130 ± 0.001 1.146 ± 0.002 1.143 ± 0.008 1.135 ± 0.008 1.127 ± 0.002 1.124 ± 0.001 1.153 ± 0.004 1.139 ± 0.002 1.137 ± 0.003 1.149 ± 0.003 1.151 ± 0.003 1.122 ± 0.001 1.143 ± 0.004 1.126 ± 0.011

0.99 ± 0.02 1.07 ± 0.02 1.05 ± 0.02 1.34 ± 0.02 0.93 ± 0.01 0.97 ± 0.01 1.05 ± 0.02 1.16 ± 0.02

1.094 ± 0.010 1.088 ± 0.009 1.011 ± 0.009 1.070 ± 0.009 1.061 ± 0.001 1.040 ± 0.001 1.049 ± 0.002 1.054 ± 0.002

1.72 ± 0.07 1.92 ± 0.07 1.20 ± 0.05 1.57 ± 0.06 1.42 ± 0.06 1.16 ± 0.05 1.79 ± 0.07 1.95 ± 0.07

All ratios are represented as activity ratios.

Pr, Nd, Sm, Zr, Nb, Pb; 20% (2σ) for Eu, Gd, Tb, Dy, Ho Er, Tm, Rb; and 30% (2σ) for Yb, Lu. 4. Results and discussion 4.1. Estimation of initial methane seep sites

230

Th for carbonate nodules at submarine

All U and Th results are shown in Table 2 and are displayed graphically in Fig. 2. The activity ratios were calculated from isotopic abundances using the following decay constants, λ230Th = 9.195 × 10− 6yr− 1, λ234U = 2.835 × 10− 6yr− 1, λ232Th = 4.933 × 10− 11yr− 1, λ238U = 1.551 × 10− 10yr− 1 (De

Bievre et al., 1971; Lounsbury and Durham, 1971; Meadows et al., 1980; Firestone, 1999). Table 1 shows that carbonate nodules sampled from the Joetsu methane seep area have U concentrations of 2.9 to 11.5ppm, and Th concentrations of 1.1 to 2.8ppm. Compared to results of Aharon et al. (1997) and Teichert et al. (2003), carbonate samples at Joetsu have high Th content. Therefore, the initial 230Th cannot be neglected to acquire accurate U–Th ages. Here, the estimation method of initial 230Th is evaluated using the isochron method, as follows. For acquisition of this isochron age, PC05-04-50 was cut into three blocks, then each block was analyzed. The analyzed points form a straight line and the slope of the regression line implies an age of 26 ± 3ka, as calculated using IsoPlot (Ludwig and Titterington,1994). Because analyzed points of sediment, which is composed mainly of silicate material, are also on the isochron for PC05-04-50 (dashed line in Fig. 2), the carbonate nodules contain the sediment as impurities. Furthermore, trace element data show that carbonate nodules contain local sediment as impurities (see Section 4.3). Therefore, it is appropriate for estimation of the amount of initial 230Th to use the mean value of local sediment. Corrected U–Th ages can be calculated as two-point isochron between carbonate samples and the mean of the local sediment (Table 2). 4.2. U–Th corrected age and initial (234U/238U) ratio

Fig. 2. U–Th radioactive disequilibrium diagrams. All ratios are activity ratios. Plots of PC05-04-50 carbonate sample draw an isochron, shown as the dashed line. The isochron represents 26 ± 3 ka, calculated using IsoPlot (Ludwig and Titterington, 1994). The isochron of a carbonate nodule is on plots of local sediments.

Corrected U–Th ages were calculated by the two-point isochron method as described in Section 4.1. Table 3 shows corrected ages and initial (234U/238U) activity ratios, which were calculated using IsoPlot (Ludwig and Titterington, 1994). Initial (234U/238U) activity ratios in carbonate nodules vary from 1.143 to 1.181 (Table 3). The values of present seawater are 1.144 ± 0.002 (Chen et al., 1986), 1.143 ± 0.004 (Henderson et al., 1999), and 1.150 ± 0.002 (Delanghe et al., 2002). In pore water, the value is higher than in seawater, e.g. 1.15–1.38 (Teichert et al., 2003), because of alpha recoil (Kigoshi, 1971). Accordingly, we can conclude that uranium in the carbonate component was derived from a mixture pore water and seawater.

Y. Watanabe et al. / Earth and Planetary Science Letters 272 (2008) 89–96 Table 3 U–Th ages and Sample ID

14

C ages of carbonate nodules Corrected U–Th age ± 2σ

Initial (234U/238U)a ± 2σ

(ka) PC05-02-60 PC05-04-50(1) PC05-04-50(2) PC05-04-50(3) PC05-04-65 PC05-CC(1) PC05-CC(2) PC05-CC(3) PC15-04-00 PC15-04-30 PC15-06-1(b) PC15-06-1(c) PC15-06-2(1) PC15-06-2(2) PC15-06-2(3) PC15-06-3(1) PC15-06-3(2) G3

14

C ageb ± 2σ

Lab code TKa no.

(BP)

21.7 ± 7.1 25.3 ± 2.9 25.2 ± 2.2 24.6 ± 3.2 62.4 ± 2.6 21.3 ± 1.4 21.5 ± 1.5 21.6 ± 1.5 35.8 ± 4.5 21.4 ± 5.4 22 ± 2 23.1 ± 1.7 12.8 ± 2.7 13.6 ± 3.2 15.8 ± 2.7 20.6 ± 4.9 21.5 ± 4.4 35.3 ± 3.5

1.143 ± 0.011 1.157 ± 0.005 1.156 ± 0.012 1.160 ± 0.013 1.166 ± 0.004 1.164 ± 0.003 1.161 ± 0.009 1.152 ± 0.009 1.163 ± 0.008 1.159 ± 0.009 1.177 ± 0.006 1.158 ± 0.004 1.159 ± 0.006 1.178 ± 0.007 1.178 ± 0.006 1.153 ± 0.008 1.181 ± 0.009 1.156 ± 0.015

30,070 ± 380 36,390 ± 700 36,390 ± 700 36,390 ± 700 47,450 ± 1260 34,020 ± 460 34,020 ± 460 34,020 ± 460 37,560 ± 700 39,620 ± 760 – – 26,970 ± 360 26,970 ± 360 26,970 ± 360 32,530 ± 540 32,530 ± 540 38,760 ± 600

TKa-13519 TKa-13586 TKa-13586 TKa-13586 TKa-13559 TKa-13563 TKa-13563 TKa-13563 TKa-13590 TKa-13591 – – TKa-13594 TKa-13594 TKa-13594 TKa-13595 TKa-13595 TKa-13520

a

Initial (234U/238U) is represented as the activity ratio. C ages provided by Hiruta et al. (personal communication). The error is represented as 2-sigma, to compare radiocarbon ages with U–Th ages. b 14

4.3. Results of various elements analyses 4.3.1. Rare earth elements (REE) pattern Fig. 3 shows the North American Shale Composite (NASC)normalized REE patterns of the same samples, using the abundances reported by Haskin et al. (1968). In Fig. 3, the REE patterns for sediment samples are flat and similar to the pattern of NASC. The concentrations of sediments are about 80% of NASC. On the other hand, the REE concentrations of carbonate nodules are 10–30% of NASC because a carbonate has the effect of dilution on REE concentrations. Also, REE has two patterns: one is flat; another is an HREE-rich pattern, which is limited to the depths of PC05 position. The two patterns are likely to arise from the difference of the detritus component in carbonate nodules. 4.3.2. Rb, Zr, Nb, REE, Pb, Th and U contents Table 4 shows the Rb, Zr, Nb, REE and Pb amounts. Concentrations of Rb, Zr, Nb, Pb, and Th in carbonate nodules correlated closely with the NASC-normalized La abundance (Fig. 4). This correlation implies that these elements in carbonate nodules are derived mainly from the silicate impurity, whose composition resembles that of the sediment. Furthermore, concentrations of these elements are diluted by pure carbonate. In other words, carbonate components rarely contain Rb, Zr, Nb, REE, Pb, and Th. 4.4. Comparison of

93

disequilibrium dating is a more reliable geochronometer than 14C dating for evaluating activities of methane seeps. Two possible sources of dead carbon exist: upwelling methane and seawater. Seawater circulates for a few hundred years in the Sea of Japan. Furthermore, the proportion of dead carbon is 0–95%, as described above. Accordingly, the origin of dead carbon in carbonate nodules is not likely to be seawater but rather upwelling methane. Carbonate nodules also contain seawater-derived carbon. Carbon in carbonate nodules is extracted from a mixture of methane and seawater. The fraction of methane is 0–0.95, and that of seawater is 0.05–1. At around 20ka, the proportion of dead carbon reaches to the maximum of 95% (Fig. 5). Moreover, ages of carbonates concentrate at 20ka (Fig. 6). The coincidence at 20ka seems to suggest that the amount of carbonate precipitation depends on upwelling methane flux, although the numbers of samples we analyzed are limited. That is, during a period of large methane flux, abundant carbonate nodules were precipitated from pore water containing dead carbon derived from methane. Conversely, during a period of small methane flux, few carbonate nodules were precipitated from pore water containing carbon derived from seawater. A sediment-corrected U–Th age of PC05-04-65 is older than 14C age and clearly older than the U–Th age of other samples. The 14C age of this sample is also the largest among the analyzed ones in this study. Although the reason is unclear, PC05-04-65 might have higher initial (230Th/232Th) than the value of local sediment, 1.07. In addition, PC05-04-65 only shows lamination among studied samples. Accordingly, it might precipitate by different mechanisms from other samples. It is necessary to analyze more samples to use isochron method to comprehend U–Th systematics in PC05-04-65. 4.5. Geologic implications of the U–Th dates In this study, the number of analyzed samples is limited. We subdivided some nodules and analyzed repeatedly. The subdivided samples show the same age (Table 3). This result means that a nodule precipitated at an active period of methane seep. On the other hand, the U–Th ages of the nodules are not controlled by stratigraphical level (e.g. the depth from seafloor). The structure of sedimentation is likely to have been disturbed by methane venting. Accordingly, we need to analyze more carbonate nodules randomly in order to decipher the whole history of methane seep. Although sampling bias is a concern, the results of U–Th dating may suggest the interpretation as follows. Fig. 6 shows a histogram of corrected U–Th ages for carbonate nodules, of which the error is about 3ka. Carbonate nodules represent

14

C ages with U–Th corrected ages

In Fig. 5, U–Th corrected ages are shown in comparison with 14C ages (Hiruta et al. personal communication). Radiocarbon dates were measured using a tandem accelerator mass spectrometer at MALT, the University of Tokyo. The 14C age is shown as raw data in Table 3, which is not calibrated to the calendar age, because the calibrations does not influence what we will discuss below. Radiocarbon dates are older than U–Th ages because of the effect of dead carbon, with the exception of the PC05-04-65 sample. The proportion of dead carbon, which is shown with dashed lines in Fig. 5, is 0–95%. Some studies have used 14C dating of carbonate samples to asses the age of methane seeps (e.g., Paull et al., 1989). Those studies described that 14C ages are maximum ages of methane seeps because it is undeniable that carbonates at methane seeps reflect fossil methane sources. Our results show clearly that 14C dating implies an age that is older than the U–Th disequilibrium age, and that U–Th

Fig. 3. NASC-normalized REE patterns for carbonate nodules and sediments from methane seeps off Joetsu. The error bar represents 2σ.

94

Y. Watanabe et al. / Earth and Planetary Science Letters 272 (2008) 89–96

Table 4 (a) Concentration of carbonate samples (PC05) from methane seeps off Joetsu Sample

PC05-02-60

PC05-04-50(1)

PC05-04-65

PC05-CC(1)

La (ppm) Ce (ppm) Pr (ppm) Nd (ppm) Sm (ppm) Eu (ppm) Gd (ppm) Tb (ppm) Dy (ppm) Ho (ppm) Er (ppm) Tm (ppm) Yb (ppm) Lu (ppm) Rb (ppm) Zr (ppm) Nb (ppm) Pb (ppm)

8.28 17.4 20.2 7.82 1.65 0.406 1.59 0.240 1.44 0.284 0.829 0.121 0.810 0.120 27.8 42.9 3.14 7.64

7.27 15.1 1.75 6.64 1.39 0.299 1.26 0.192 1.24 0.266 0.810 0.123 0.855 0.130 25.4 34.6 2.29 6.56

7.25 15.4 1.77 6.89 1.47 0.390 1.54 0.242 1.65 0.364 1.20 0.186 1.37 0.225 22.7 39.4 2.80 5.68

4.71 10.1 1.16 4.51 0.977 0.210 1.01 0.158 1.05 0.227 0.698 0.100 0.732 0.109 13.5 20.1 1.76 3.55

(b) Concentration of carbonate samples (PC15 and G3) from methane seeps off Joetsu Sample PC15-04-00 PC15-04-30 PC15-06-1(b) PC15-06-1(c) PC15-06-2(1)

PC15-06-2(2)

PC15-06-2(3)

PC15-06-3(1)

PC15-06-3(2)

G3

La (ppm) Ce (ppm) Pr (ppm) Nd (ppm) Sm (ppm) Eu (ppm) Gd (ppm) Tb (ppm) Dy (ppm) Ho (ppm) Er (ppm) Tm (ppm) Yb (ppm) Lu (ppm) Rb (ppm) Zr (ppm) Nb (ppm) Pb (ppm)

4.42 8.97 1.11 3.97 0.842 0.188 0.764 0.118 0.713 0.141 0.404 0.061 0.434 0.063 14.6 19.0 1.71 3.85

4.47 9.56 1.16 4.41 0.953 0.204 0.796 0.128 0.751 0.152 0.441 0.064 0.436 0.066 12.4 15.6 1.49 3.35

5.16 11.2 1.28 4.74 0.963 0.204 0.845 0.129 0.801 0.159 0.469 0.069 0.480 0.072 18.7 25.1 2.09 5.22

5.13 10.7 1.26 4.70 0.981 0.214 0.878 0.140 0.856 0.173 0.512 0.079 0.560 0.080 17.3 21.7 1.96 4.91

3.32 7.09 0.823 3.16 0.661 0.149 0.629 0.097 0.591 0.117 0.352 0.051 0.362 0.054 12.8 16.3 1.34 3.09

7.00 14.9 1.78 6.76 1.51 0.345 1.47 0.231 1.40 0.281 0.818 0.122 0.844 0.116 23.5 25.2 2.27 6.08

7.58 16.3 1.87 7.08 1.48 0.313 1.35 0.209 1.31 0.267 0.779 0.112 0.760 0.110 23.8 31.8 2.02 6.64

5.88 12.2 1.46 5.54 1.20 0.274 1.20 0.183 1.12 0.220 0.628 0.090 0.629 0.087 17.5 21.6 1.80 4.51

7.02 14.9 1.77 6.90 1.68 0.398 1.81 0.279 1.70 0.345 0.991 0.139 0.967 0.136 21.9 26.8 2.52 5.69

4.73 10.2 1.18 4.52 0.978 0.222 0.933 0.142 0.862 0.170 0.495 0.071 0.480 0.070 16.8 22.7 2.09 4.55

(c) Concentration of sediment samples from methane seeps off Joetsu Sample

PC04-02-100

PC04-03-100

PC04-04-100

PC05-05-100

PC11-03-100

PC11-04-100

PC11-05-100

PC11-06-60

La (ppm) Ce (ppm) Pr (ppm) Nd (ppm) Sm (ppm) Eu (ppm) Gd (ppm) Tb (ppm) Dy (ppm) Ho (ppm) Er (ppm) Tm (ppm) Yb (ppm) Lu (ppm) Rb (ppm) Zr (ppm) Nb (ppm) Pb (ppm)

23.6 50.9 5.60 21.1 4.41 0.998 4.04 0.606 3.51 0.693 2.02 0.296 2.12 0.281 84.0 118 11.8 25.4

27.6 57.9 6.31 24.0 4.83 1.02 4.34 0.636 3.72 0.739 2.11 0.312 2.21 0.299 95.0 161 16.9 23.2

25.9 54.4 6.17 23.2 4.79 1.03 4.27 0.639 3.74 0.742 2.13 0.307 2.21 0.304 93.9 133 12.9 26.0

26.6 55.2 6.33 23.8 4.79 1.06 4.33 0.635 3.71 0.731 2.10 0.310 2.20 0.298 96.9 120 12.9 25.3

23.7 51.6 5.88 22.4 4.40 0.909 3.95 0.592 3.66 0.711 2.07 0.304 2.19 0.305 82.0 94.2 10.9 28.7

29.9 64.7 7.40 28.7 5.92 1.37 5.88 0.871 5.62 1.12 3.28 0.487 3.54 0.512 80.5 121 11.5 26.2

28.6 58.8 6.66 27.0 5.43 1.23 5.27 0.759 4.79 0.965 2.74 0.413 2.94 0.420 86.4 143 13.0 24.0

27.9 58.5 6.80 26.3 5.17 1.16 4.95 0.718 4.52 0.897 2.57 0.381 2.71 0.379 84.8 149 13.1 23.4

corrected ages of 12–35ka. Observation of U–Th ages indicates that carbonate precipitation occurs in distinct intervals rather than continuously. Carbonate precipitation is closely linked to the supply of methane seep (Aharon et al., 1992; Bohrmann et al., 1998). Based on comparison of 14C ages and U–Th corrected ages (Fig. 5), it is likely that the amount of carbonate depends on the methane flux. Although 10 carbonate nodules were analyzed in this study; the analyzed carbonates were sampled randomly. Accordingly, the results were representative for carbonate-forming ages in this area. Therefore, the history of methane venting is as follows; methane seeps initiated about 35ka, then reactivated at 20ka, and subsequently diminished at about 12ka. Moreover, the evidence that carbonate has rarely precipitated recently

in spite of active seepage shows that methane was vented more actively at about 20ka than at present. During the time interval of U–Th ages from 12ka to 35ka, environmental conditions must have been favorable for enhanced methane fluxes through sediment off Joetsu, of the Eastern Margin of Japan Sea. The timing of the most intensive methane seep period, at 20ka, is consistent with lowest-stand sea level. Some studies also show that methane was released episodically to the seafloor during the last glacial period (Kennett et al., 2000; Hinrichs et al., 2003; Uchida et al., 2005; Ohkushi et al., 2005). Uchida et al. (2005) and Ohkushi et al. (2005) present evidence of isotopic data of organics molecules and foraminifera, indicating that the methane release

Y. Watanabe et al. / Earth and Planetary Science Letters 272 (2008) 89–96

95

Fig. 4. La vs. Rb, Zr, Nb, and Th concentrations diagram for carbonate nodules. The error bar represents 2σ.

resulted from the decomposition of gas hydrate around 25–17ka, off Hokkaido, Japan. They suggest that the instability of hydrate is modulated by intermediate water warming and/or lowering of the sea level. The mechanism of methane seep off Joetsu leaves three possibilities: (1) the decomposition of gas hydrate and/or ascent of methane gas which is derived from the decrease of hydraulic pressure

because of low-stand sea level; (2) the decomposition of gas hydrate which is derived from temporal warming of seawater; or (3) ascent of methane gas, which is derived from a tectonic event such as earthquake. Future studies must examine more carbonate nodule samples to determine U–Th ages and carbon and oxygen isotopes, which might identify the origin of methane. Synthetic analysis of

Fig. 5. Comparison of radiocarbon dates with U–Th corrected ages of carbonate nodules at methane seeps off Joetsu. The error bar is 2σ. 14C ages except for that of one sample are older than U–Th correction ages because of the effect of dead carbon. Dashed lines show the proportion of dead carbon, which is 50–99%.

Fig. 6. Histogram of U–Th-corrected ages for carbonate nodules (data from Table 2). Vertical axis shows the number of data, which contain the result of one carbonate nodule measured repeatedly. Numbers above the bar show the number of carbonate nodules.

96

Y. Watanabe et al. / Earth and Planetary Science Letters 272 (2008) 89–96

multi-isotopes will elucidate the mechanisms that form methane seeps in greater detail. 5. Conclusions From the results described above, the following conclusions can be drawn: (1)

Carbonate nodules at methane seeps off Joetsu have U concentrations of 2.9–11.5ppm, and Th concentrations of 1.1–2.8ppm. Therefore, the amount of initial 230Th cannot be neglected to acquire accurate U–Th ages. (2) Isochron method was tested by triple measurements of a carbonate nodule. The results show that the three samples form an isochron. Points of analyzed sediment fall on the isochron, showing that carbonate nodules contain sediment as impurity. Accordingly, to correct for initial 230Th, this study assumes that the carbonate nodules contain Th derived from local sediment. Initial 230Th was subtracted using the mean value of local sediment. Consequently, U–Th correction ages vary from 12 to 35ka; many of them concentrate at about 20ka. (3) We also compared U–Th corrected ages with 14C ages, which were older than U–Th ages. At methane seep sites, it is undeniable that carbonates reflect dead carbon, which makes 14 C ages older, that is derived from methane. Therefore, we can conclude that U–Th disequilibrium dating is a more reliable geochronometer than 14C dating for evaluating activities of methane seeps. The proportion of dead carbon reached 95% at around 20ka. The coincidence at 20ka suggests that the carbonate precipitation depends on the upwelling methane flux. (4) During 12–35ka recorded by U–Th ages of the carbonates, environmental conditions must have been favorable for enhanced methane fluxes through sediment. Furthermore, extensive methane flows at 20ka are consistent with loweststand sea level during the last glacial period. Acknowledgments We deeply appreciate all shipboard scientists of the Joetsu cruise in 2004 and 2005 summer for providing us samples and valuable information. Thanks are extended to Dr. YuVin Sahoo, Dr. Takeshi Hanyu, Dr. Yoshiro Nishio, Dr. Yoshiki Miyata, and Mr. Taehoon Kim for advice related to experimental techniques. We would also like to thank our laboratory members for the maintenance of our laboratory. This study was supported partly by Grants-in-Aid for scientific research from the Japan Society for the Promotion of Science to SN. The Hayashi Memorial Foundation for Female Natural Scientists to YW partly supported this study. References Aharon, P., Roberts, H.H., Snelling, R., 1992. Submarine venting of brines in the deep Gulf of Mexico: observation and geochemistry. Geology 20, 483–486. Aharon, P., Schwarcz, H.P., Roberts, H.H., 1997. Radiometric dating of submarine hydrocarbon seeps in the Gulf of Mexico. Geol. Soc. Amer. Bull. 109 (5), 568–579. Aoyama, C., Matsumoto, R., Okuda, Y., Ishida, Y., Hiruta, A., Sunamura, M., Numanami, H., Tomaru, H., Snyder, G.T., Komatsubara, J., Takeuchi, R., Hiromatsu, M., Aoyama, D., Koike, Y., Takeda, S., Hayashi, T., Hamada, H., Kawada, Y., 2005. Acoustical survey of methane plumes using the quantitative echo sounder in the eastern margin of the Sea of Japan. Proceedings of the Fifth International Conference on Gas Hydrate, pp. 790–795. Bohrmann, G., Greinert, J., Suess, E., Torres, M., 1998. Authigenic carbonates from the Cascadia subduction zone and their relation to gas hydrate stability. Geology 26 (7), 647–650. Bourdon, B., Henderson, G.M., Lundstrom, C.C., Turner, S.P., 2003. Uranium-series Geochemistry (Reviews in Mineralogy & Geochemistry Vol. 52). Mineralogical Society of America, Washington D.C. Chen, J.H., Edwards, R.L., Wasserburg, G.J., 1986. 238U, 234Uand 232Th in seawater. Earth Planet. Sci. Lett. 80, 241–251.

De Bievre, P., Lauer, K.F., Leduigon, Y., Moret, H., Muschenborn, G., Spaepen, J., Spernol, A., Vaninbroukx, R., Verdingh, V., 1971. The Half-life of 234-U, Proceedings of the International Conference on Chemical and Nuclear Data, Measurements and Applications, Cantarbury. Institute of Civil Engineers, London. Delanghe, D., Bard, E., Hamelin, B., 2002. New TIMS constraints on the uranium-238 and uranium-234 in seawaters from the main ocean basins and the Mediterranean Sea. Mar. Chem. 80, 79–93. Firestone, R.B., 1999. Table of Isotopes, 8th ed. Oxford Science Publications, Oxford, pp. 206–211. Fukuda, S., Nakai, S., 2002. 38U/230Th disequilibrium measurement for volcanic standard rock samples using a multiple-collector ICPMS. Geochem. J. 36, 465–473. Gulin, S.B., Polikarpov, G.G., Egorov, V.N., 2003. The age of microbial carbonate structures grown at methane seeps in the Black Sea with an implication of dating of the seeping methane. Mar. Chem. 84, 67–72. Haskin, L.A., Haskin, M.A., Frey, F.A., Wildman, T.R., 1968. In: Ahrens, L.H. (Ed.), Relative and Absolute Terrestrial Abundances of the Rare Earths. Origin and distribution of the elements, vol. 1. Pergamon, Oxford, pp. 889–911. Henderson, G.M., Slowey, N.C., Haddad, G.A., 1999. Fluid flow through carbonate platforms: constraints from 234U/238U and Cl− in Bahamas pore-waters. Earth Planet. Sci. Lett. 169, 99–111. Hinrichs, K.U., Hmelo, L.R., Sylva, S.P., 2003. Molecular fossil record of elevated methane levels in Late Pleistocene coastal waters. Science 299, 1214–1217. Hiruta, A., Matsumoto, R., Ishida, Y., Tomaru, H., Snyder, G., Aoyama, C., Hiromatsu, M., 2005. Formation of gas hydrate and carbonate nodules around active seeps thermogenic methane at the eastern margin of Japan Sea, EOS, Transactions AGU86. Fall Meeting Supplement, Abstract OS43A-613. Ishida, Y., Matsumoto, R., Hiruta, A., Aoyama, C., Tomaru, H., Hiromatsu, M., 2005. High concentration of methane and magnificent gas plumes over gas hydrate field in the eastern margin of Japan Sea, EOS, Transactions AGU86. Fall Meeting Supplement, Abstract OS43A-0614. Kennett, J.P., Cannariato, K.G., Hendy, I.L., Behl, R.J., 2000. Carbon isotopic evidence for methane hydrate instability during Quaternary Interstitials. Science 288, 128–133. Kigoshi, K., 1971. Alpha-recoil thorium-234: dissolution into water and the uniraum234/uranium-238 disequilibrium in nature. Science 173, 47–48. Kulmn, L.D., Suess, E., Moore, J.C., Carson, B., Lewis, B.T., Ritger, S.D., Kadko, D.C., Thornburg, T.M., Embley, R.W., Rugh, W.D., Massoth, G.J., Langseth, M.G., Cochrane, G.R., Scamman, R.L., 1986. Oregon subduction zone: venting, fauna, and carbonates. Science 231, 561–566. Lalou, C., Fontugne, M., Lallemand, S.E., Lauriat-Rage, A., 1992. Calyptogena-cemented rocks and concretions from the eastern part of Nankai accretionary prism: age and geochemistry of uranium. Earth Planet. Sci. Lett. 109, 419–429. Lounsbury, M., Durham, R.W., 1971. The alpha-half-life of 234U, Proceedings of the International Conference on Chemical and Nuclear Data, Measurements and Applications, Cantarbury. Institute of Civil Engineers, London. Ludwig, K.R., Titterington, D.M., 1994. Calculation of 230Th/U isochrons, ages and errors. Geochim. Cosmochim. Acta 58, 5031–5042. Maslin, M.A., Thomas, E., 2003. Balancing the deglacial global carbon budget: the hydrate factor. Quat. Sci. Rev. 22, 1729–1736. Matsumoto, R., Tomaru, H., Hiruta, A., Ishida, Y., Takeuchi, R., Snyder, G., Kotani, R., Kotani, R., Okuda, Y., Sato, M., Numanami, H., Aoyama, C., Hiromatsu, M., Lu, H., Matduda, N., Lu, Z., Takeuchi, E., Goto, T., Machiyama, H., Toh, H., Komatsubara, J., 2005. Gas hydrate layer and prominent flares of gas plumes in Naoetsu Basin, Eastern Margin of Japan Sea, EOS, Transactions AGU86. Fall Meeting Supplement, Abstract OS41C-04. Meadows, J.W., Armani, R.J., Callis, E.L., Esling, A.M., 1980. Half-life of 230Th. Phys. Rev. 22C, 750–754. Ohkushi, K., Ahagon, N., Uchida, M., Shibata, Y., 2005. Foraminiferal isotope anomalies from northwestern Pacific marginal sediments. Geochem., Geophys., Geosyst. 6 (4). Paull, C.K., Martens, C.S., Chanton, J.P., Neumann, A.C., Coston, J., Jull, A.J.T., Toolin, L.J., 1989. Old carbon in living organisms and young CaCO3 cements from abyssal brine seeps. Nature 342, 166–168. Peckmann, J., Reimer, A., Luth, U., Luth, C., Hansen, B.T., Heinicke, C., Hoefs, J., Reitner, J., 2001. Methane-derived carbonates and authigenic pyrite from the northwestern Black Sea. Mar. Geol. 177, 129–150. Sakai, H., Gamo, T., Ogawa, Y., Boulegue, J., 1992. Stable isotopic ratios and origins of the carbonates associated with cold seepage at the eastern Nankai Trough. Earth Planet. Sci. Lett. 109, 391–404. Sloan, L.C., Walker, J.C.G., Moore Jr., T.C., Rea, D.K., Zachos, J.C., 1992. Possible methaneinduced polar warming in the early Eocene. Nature 357, 320–322. Teichert, B.M.A., Eisenhauer, A., Bohrmann, G., Haase-Schramm, A., Bock, B., Linke, P., 2003. U/Th systematics and ages of authigenic carbonates from Hydrate Ridge, Cascadia Margin: recorders of fluid flow variations. Geochim. Cosmochim. Acta 67, 3845–3857. Uchida, M., Shibata, Y., Ohkushi, K., Ahagon, N., Hoshiba, M., 2005. Episodic methane release events from Last Glacial marginal sediments in the western North pacific. Geochem., Geophys., Geosyst. 5 (8). Watanabe, Y., Nakai, S., 2007. Accurate U–Th radioactive disequilibrium analyses of carbonate rock samples using commercially available U and Th reagents and MCICP-MS. Microchim. Acta 156, 289–295. Yokoyama, T., Makishima, A., Nakamura, E., 1999. Separation of thorium and uranium from silicate rock samples using two commercial extraction chromatographic resins. Anal. Chem. 71, 135–141.