Water pH and temperature in Lake Biwa from MBT ′ / CBT indices during the last 280 000 years

We generated a 280 000 yr record of water pH and temperature in Lake Biwa, central Japan, by analysing the methylation index (MBT ) and cyclisation ratio (CBT) of branched tetraethers in sediments from piston and borehole cores. Our aim was to understand the responses of precipitation and air temperature in central Japan to the East Asian monsoon variability on orbital timescales. Because the water pH in Lake Biwa is determined by phosphorus and alkali cation inputs, the record of water pH should indicate the changes in precipitation and temperature in central Japan. Comparison with a pollen assemblage in a Lake Biwa core suggests that lake water pH was determined by summer temperature in the low-eccentricity period before 55 ka, while it was determined by summer precipitation in the higheccentricity period after 55 ka. From 130 to 55 ka, the variation in lake pH (summer precipitation) lagged behind that in summer temperature by several thousand years. This perspective is consistent with the conclusions of previous studies (Igarashi and Oba, 2006; Yamamoto, 2009), in that the temperature variation preceded the precipitation variation in central Japan.


Introduction
The East Asian monsoon governs the climate of East Asia (Wang et al., 2003), and East Asian monsoon variability on orbital timescales has been the topic of many studies, which have revealed that it has responded to precession; however, the timing of monsoon variability continues to be debated.Kutzbach (1981) hypothesised that the Asian monsoon responds to insolation changes at low latitudes, which are regulated by precession.According to this hypothesis, the summer monsoon is maximal when Northern Hemisphere summer insolation is maximal in the precession cycle.Indeed, oxygen isotope records from cave stalagmites in China have demonstrated that summer monsoon variability was pronounced at the precession cycle and maximal at the July-August precession (e.g.Wang et al., 2001Wang et al., , 2008;;Yuan et al., 2004;Dykoski et al., 2005).However, some proxy records are not consistent with this hypothesis.Clemens and Prell (2003) reported that Indian summer monsoon variability showed both precession and obliquity cycles and was maximal at the November perihelion on the precession band.The pollen record in the north-western Pacific off central Japan shows that the East Asian monsoon has been strongest at the October-November perihelion in a precession cycle (Heusser and Morley, 1985;Igarashi and Oba, 2006).Thus, the conclusions have varied according to the proxy record used.
The MBT/CBT index has been also applied to lake sediments for palaeotemperature reconstruction (e.g.Niemann et al., 2012), but most applications in lake sediments were unsuccessful (e.g.Tierney and Russel, 2009;Tyler et al., 2010;Zink et al., 2010;Wang et al., 2012).Tierney and Russel (2009) argued that the unrealistic estimated MAATs are attributable to in situ production of branched GDGTs in lake water.Tierney et al. (2010) noted that the correlation between MBT/CBT from sediments and MAAT for 46 lakes in East Africa differed from that of the global soil set and proposed a calibration applicable in lake environments.Ajioka et al. (2014) investigated the distribution of GDGTs in soils and river and lake sediments in the Lake Biwa drainage basin and showed that the distribution of branched GDGTs in the lake sediments was different from that in the catchment soils, suggesting in situ production of branched GDGTs in the lake.They also found, in contrast to the conclusion of Tierney et al. (2010), that the relationships among soil pH, MAAT, and MBT /CBT in soils are not different from those of lake water pH, temperature, and MBT /CBT in lake sediments, implying that the soil calibration is applicable without modification to the study of lake sediments to obtain lake water temperature and pH.
In this study, we investigated branched GDGTs in sediments from borehole BIW08-B and piston core BIW07-6 in Lake Biwa, central Japan, to reconstruct lake water pH and temperature during the last 280 000 yr.We then evaluated the variability of the East Asian summer and winter monsoons based on estimated summer precipitation and winter lake water temperature.

Environmental setting of Lake Biwa
Lake Biwa in central Japan is at an elevation of 84 m and is surrounded by mountains ca.1000 m high.With an area of 674 km 2 and a watershed area of 3850 km 2 , Lake Biwa is the largest lake in Japan (Fig. 1).More than 118 rivers flow into the lake, and the Seta River discharges from it.The climate of the area is affected by the East Asian monsoon (Yoshino, 1965): summer monsoon brings warm and humid conditions and winter monsoon brings snowfall to the northern part of the area and dryness to the southern part.

Samples
Piston core BIW07-6 (18.42 m long) was taken from the central part of Lake Biwa (35 • 13 59.02 N, 136 • 02 51.89 E; water depth of 55 m) in 2007 (Fig. 1; Takemura et al., 2010).The sediments in the core consisted of homogenous dark-grey silty clay (Fig. 2).The age-depth model for core BIW07-6 was generated from 6 well-dated, widespread ash layers and 13 accelerator mass spectrometry (AMS) 14 C ages of plant wood fossils (partly reported in Kitagawa et al., 2010) analysed at 3 AMS facilities during the past decade: the Center for Chronological Research at Nagoya University (Laboratory code NUTA2; Nakamura et al., 2000Nakamura et al., , 2004)), the W. M. Keck Carbon Cycle Accelerator Mass Spectrometry Laboratory (UCIAMS; Southon et al., 2004a, b), and the Korean Institute of Geoscience and Mineral  Resources (KIGAM; Hong et al., 2010) (Table 1).The 14 C ages were converted to calendar ages using the IntCal13 data set (Reimer et al., 2013).The average sedimentation rate was found to be 0.4 m ky −1 (Fig. 3).The sediment was stored at 4 • C for 0.5 yr.Then, 137 samples (1 cm thick) were collected from 0.30 m (0.6 ka) to 18.29 m (46 ka) and immediately freeze-dried.The average sampling interval was ∼ 0.3 ky.Borehole core BIW08-B (100.3 m long) was collected from its site (35 • 13 41.15N, 136 • 03 21.19 E; water depth of 53 m) in 2008 (Fig. 1).The sediments consisted of darkgrey massive silty clay from 0 to 89 m, sandy silt containing abundant sand and plant debris from 89 to 99 m, and darkgrey massive silty clay from 99 to 100.3 m (Fig. 2; Takemura et al., 2010).The age-depth model for core BIW08-B was generated from 17 dated widespread ash layers, 2 AMS 14 C age of plant wood fossils analysed at KIGAM (Hong et al., 2010), and 9 dated anhysteretic remanent magnetisation (ARM) events (Table 2).The 14 C ages were converted to calendar ages using the IntCal13 data set (Reimer et al., 2013).Ages of the ARM events were determined through the age model of core BIW07-6, based on the correlation of the ARM variations observed in Lake Biwa sediments (Hayashida et al., 2007).The average sedimentation rate was found to be 0.3 m ky −1 (Fig. 3).The sediment was stored at 4 • C for 0.5 yr.Then, 152 samples (2.5 cm thick) were collected from 13 m (43 ka) to 88 m (280 ka), and the samples were immediately freeze-dried.The average sampling interval was ∼ 1.9 ky.    1 and 2).Error bars indicate the 95 % confidence intervals (2σ ).
The conditions were: flow rate of 0.2 mL min −1 , isocratic with 99 % hexane and 1 % propan-2-ol (5 min) followed by a linear gradient to 1.8 % propan-2-ol over 45 min.Detection was achieved using atmospheric pressure chemical ionisation (APCI) MS in positive-ion mode.The spectrometer was run in full-scan mode (m/z 500-1500).Compounds were assigned from a comparison of mass spectra and retention times with GDGT standards (from the main phospholipids of Thermoplasma acidophilum via acid hydrolysis) and values from a previous study (Hopmans et al., 2000).Quantification was achieved by integrating the summed peak areas in the (M + H) + and (M + H + 1) + chromatograms and comparing them with the peak area from an internal standard C 46 glycerol trialkyl glycerol tetraether (GTGT; Patwardhan and Thompson, 1999) The methane index (MI) was calculated following Zhang et al. (2011): The pH and MAAT were calculated according to the following equations based on the data set of Lake Biwa watershed soils (Ajioka et al., 2014): The standard deviations of CBT and MBT of replicated analyses averaged 0.034 and 0.015, respectively, in this study.The average root mean square errors for the calibration of pH and MAAT are 0.5 and 3.5 • C, respectively.

Previous results of GDGTs in lake surface and river
sediments and soils in the Lake Biwa drainage basin Ajioka et al. (2014) investigated the distributions of GDGTs in lake surface and river sediments and soils in the Lake Biwa drainage basin.MBT and CBT values are lower in lake surface sediments than in the soils from the catchment areas (Fig. 4).Among lake surface sediments, coarse-grained sediments have higher values (Fig. 4).River sediments have intermediate values between lake surfaces sediments and soils (Fig. 4).Based on this result, Ajioka et al. (2014) suggested that branched GDGTs in lake surface sediments are mostly produced in lake water, but there are some contributions of soil GDGTs in coarse-grained sediments in the lake.
The relationships between CBT and soil pH and between MBT /CBT and mean annual temperature in soil samples in the Lake Biwa drainage basin are not different from those in the global soil set (Peterse et al., 2012).If global and Lake Biwa soil calibrations are applied, the estimated pH values by CBT fit the winter pH values of lake water, but the application of lake calibration (Tierney et al., 2010) was not successful (Fig. 5).If global and Lake Biwa soil calibrations are applied, the temperatures estimated by MBT /CBT fit the winter temperatures of lake water.Ajioka et al. (2014) thus suggested that the branched GDGTs in Lake Biwa sediments are produced in winter, CBT and MBT reflect winter pH and temperature of lake water, and soil calibration is applicable to Lake Biwa sediments.

Results
The concentrations of isoprenoid and branched GDGTs varied between 0.01 and 2.55 and between 0.05 and 12.58 µg g −1 , respectively (Fig. 6).CBT-based pH ranged from 6.0 to 7.6 (Fig. 6).The measured pH value in the surface water of Lake Biwa has an annual average of 8.1 and is lowest (∼ 7.6) in winter and highest (∼ 8.8) in summer, whereas the pH of the bottom water at a depth of 59 m ranges from 7.2 to 7.7 with an average value of ∼ 7.5 (off Minami-Hira: 35 • 11 39 N, 135 • 59 39 E; data from the Lake Biwa Environmental Research Institute, http://www.pref.shiga.lg.jp/biwako/koai/hakusyo/index.html).On the other hand, the measured pH of surface soils in the drainage basin ranged from 3.3 to 8.0 with an average of 5.0 (Ajioka et al., 2014), which is lower than that in lake water.The CBT-based pH values of the core sediments are lower than those in the present lake water.CBT-based pH was maximal at 279, 249, 210, 160, 133, 117, 78, 45, 25, and 6 ka and minimal at 264, 220, 178, 149, 123, 95, 52, 29, and 17 ka.MBT /CBT-based temperature ranged from 4 to 11 • C and was lower during glacials and higher during interglacials (Fig. 6).The modern measured temperature of surface water in Lake Biwa has an annual average of 16.9 • C and is lowest (∼ 7.6 • C) in winter and highest (∼ 27.9 • C) in summer, whereas the temperature of bottom water at a depth of 59 m ranges from 7.2 to 8.2 • C with an annual average of ∼ 7.7 • C (off Minami-Hira; data from the Lake Biwa Environmental Research Institute, http: //www.pref.shiga.lg.jp/biwako/koai/hakusyo/index.html).

Factors controlling water pH in Lake Biwa
Lake water pH depends on the geology of the drainage basin, evaporation, the photosynthesis of phytoplankton and submerged plants, the respiration of organisms, and the decomposition of organic matter by microbes (Wetzel, 2001).In volcanic regions, lake water receives strong mineral acids, particularly sulfuric acid, which decreases pH to less than 4 (Wetzel, 2001).There is, however, no active volcano in the Lake Biwa watershed.Thus, this factor is not important for the pH of Lake Biwa water.In contrast, Ca 2+ supplied from limestone increases lake water pH (Wetzel, 2001).In the Lake Biwa drainage basin, limestone is exposed only in the Mt.Ibuki area, and its contribution toward controlling lake water pH should be minor.However, if chemical weathering is enhanced, this may increase the flux of Ca 2+ from the silicate rocks in the drainage basin.This is one of the potential factors affecting lake water pH in Lake Biwa.Evaporation of lake water increases water pH (Wetzel, 2001), but pollen records in Lake Biwa cores suggest moist environments throughout the last 430 ky (Miyoshi et al., 1999), so evaporation has not been a factor controlling lake water pH.Consumption of CO 2 by the photosynthesis of phytoplankton and submerged plants increases lake water pH (e.g.Talling, 1976).On the other hand, regen-eration of CO 2 by the respiration of organisms and degradation of organic matter by microbes decreases lake water pH (Wetzel, 2001).Because of the balance of the CO 2 budget, the lake's surface water pH values range from 7.6 in winter to 8.8 in summer (data from the Lake Biwa Environmental Research Institute, http://www.pref.shiga.lg.jp/ biwako/koai/hakusyo/index.html).A tripling of the concentration of chlorophyll from winter to spring in Lake Biwa increases lake water pH by 0.85 (data from the Lake Biwa Environmental Research Institute, http://www.pref.shiga.lg.jp/ biwako/koai/hakusyo/index.html).These observations suggest that photosynthesis in the lake water is the second potential factor controlling water pH in Lake Biwa.
In Lake Biwa, photosynthesis is controlled mainly by the phosphorus concentration in the water (Ishida et al., 1982;Tezuka, 1985).The anthropogenic eutrophication of Lake Biwa induced high primary production, resulting in an increase in the pH of the lake water by more than 1 from the 1960s to the 1970s (Nakayama, 1981).The phosphorus concentration in the lake is determined by the inflow of phosphorus from the catchment soils, which is governed by precipitation in the watershed (Kunimatsu, 1993).At present, the East Asian summer monsoon brings most of the annual precipitation to the study area, exceptionally at high elevations in the northern part where snowfall brought by the East Asian winter monsoon is relatively important (http://www.jma.go.jp/jma/index.html).Therefore, summer precipitation in the watershed is a factor that controls photosynthesis and, consequently, the pH of the lake water.The phosphorus concentration may also be governed by air temperature because the dissolution of silicate depends on temperature in chemical weathering processes (White and Blum, 1995).Thus, both higher precipitation and higher temperature potentially increase the inflow of phosphorus to the lake, enhancing primary production, and thus raise the lake water pH.Increases in precipitation and temperature intensify chemical weathering.The chemical weathering increases the inflow of Ca 2+ and phosphorus to the lake, with both raising lake water pH.
It is not clear whether branched GDGTs are produced in the water column or the sediment surface in Lake Biwa (Ajioka et al., 2014).If branched GDGTs are produced in the water column, the main production depth of branched GDGTs is unclear.Recently, Buckles et al. (2014) found a high abundance of intact branched GDGTs in the suspended particulates in the oxygen-deficient bottom water in Lake Challa.They suggested that the oxygen-deficient environment (O 2 < 2 mg L −1 ) provides favourable conditions for branched-GDGT production.In Lake Biwa, such conditions (O 2 < 2 mg L −1 ) exist in the bottom water above the sediment surface from autumn to early winter (http://www.pref.shiga.lg.jp/biwako/koai/hakusyo/index.html).It is thus likely that the branched GDGTs were produced mainly in the bottom water in Lake Biwa.However, this is merely speculation by analogy with the case of Lake Challa.This speculation should be tested using the original data of suspended and sinking particles in Lake Biwa.

Changes in lake water pH during the last 280 ky
The CBT-based pH during the last 280 ky showed a glacialinterglacial pattern with higher pH in interglacial periods (Fig. 6).The variation pattern in CBT-based pH was consistent with that of Cryptomeria pollen abundance and lagged behind that of the pollen temperature index, T p (= 100 × T w /[T c +T w ], where T w is the sum of temperate taxa and T c is the sum of subalpine taxa), in cores BIW95-4 and Takashima-oki (BT) in Lake Biwa from 130 to 55 ka (Fig. 7; Hayashi et al., 2010a, b).Igarashi and Oba (2006) assumed that the abundance of Cryptomeria pollen could be used as a proxy for summer precipitation because the natural distribution of Cryptomeria corresponds to an area where precipitation is greater than 2000 mm per year.The correspondence between CBT-based pH and Cryptomeria pollen abundance indicates that both proxy records are robust records of precipitation.
The T p is used as a proxy of summer air temperature (Igarashi and Oba, 2006).Igarashi and Oba (2006) reported that variation in the T p value preceded the abundance of Cryptomeria pollen in core MD01-2421 off central Japan in the western North Pacific, indicating that the variation in summer air temperature preceded that in summer precipitation by several thousand years.Yamamoto (2009) argued that this time lag was caused by the latitudinal shift of the Baiu Front (an early summer rain front).Our new record indicates that CBT-based pH lagged behind the T p record, implying that summer precipitation lagged behind summer air temperature, as suggested by pollen studies (Igarashi and Oba, 2006;Hayashi et al., 2010a, b).
However, the variation pattern in CBT-based pH was not consistent with that of Cryptomeria pollen abundance but with the T p value in core BIW95-4/BT from 55 ka to the present (Fig. 7; Hayashi et al., 2010a, b).This correspondence suggests that the lake water pH was determined by summer temperature rather than summer precipitation in this period.Nakagawa et al. (2008) suggested that the response of the East Asian monsoon to summer insolation was different between low-and high-eccentricity periods, and that the response was not clear in the low-eccentricity (< 0.024) periods.The eccentricity was lower in the period from 55 ka to the present than the period from 130 to 55 ka.We speculate that higher-amplitude variations in summer insolation in the latter period induced higher-amplitude precipitation variations, which resulted in conditions in which precipitation controlled lake water pH.In contrast, lower-amplitude variations in the former period induced lower-amplitude precipitation variations.Alternatively, the other climatic forcing such as greenhouse gas forcing was relatively important, and temperature variation was a major factor controlling the variation in lake water pH.
The age uncertainty (2σ values (95 % confidence level)) of core BIW08-B ranges from 5 to 11 ky in MIS (marine isotope stage) 5 and 6 because of the large error of radiogenic ages of tephras (Table 2).However, the pollen composition (Tp and Cryptomeria abundance) in cores BIW95-4/BT nearby the  (Hayashi et al., 2010a, b); the δ 18 O of stalagmites in Sanbao and Hulu caves in China (Wang et al., 2001(Wang et al., , 2008)); and eccentricity during the last 150,000 years.Red, blue, and green bars indicate age controls with 95 % confidence intervals of volcanic ashes, the 14 C of plant debris and total organic carbon, and ARM events, respectively.The names of tephras that appeared commonly in BIW07-6/BIW08-B and BIW95-4/BT are shown in the panel.
study site has a variation consistent with that in marine core MD01-2421 from off the coast of central Japan in the western North Pacific.The age-depth model of core MD01-2421 was established by oxygen isotope stratigraphy using benthic foraminifera isotopes (Oba et al., 2006).The assumed synchronous vegetation change in central Japan, namely the correspondence of pollen assemblages between MD01-2421 and Lake Biwa cores, is evidence that the uncertainty of the age-depth models of Lake Biwa cores in MIS 5 and 6 was smaller than that indicated by the dating error of each tephra, and the ages of sediments were precise enough to discuss variation on orbital timescales.
Though caution must be exercised, due to this limitation of the age-depth model, comparison with Chinese stalagmite records revealed a regional perspective on the East Asian summer monsoon variability.In the period from 55 ka to the present, the variation in CBT-based pH was synchronous with that of δ 18 O in stalagmites in China (Fig. 7; Wang et al., 2001Wang et al., , 2008)).This correspondence suggests that summer temperature in central Japan varied synchronously with the δ 18 O of the precipitated water in China.
However, in the period from 143 to 55 ka, the variation in CBT-based pH lagged behind that of the δ 18 O in stalagmites in China by several thousand years (Fig. 7).The mismatch between Japanese precipitation and Chinese stalagmite δ 18 O records was reported by Yamamoto (2009) and Clemens et al. (2010).There are two possible interpretations of this phenomenon.First, precipitation in Japan and China during early summer is determined principally by the position of the Baiu Front that develops at the atmospheric boundary between warm, moist air masses flowing from the south and cold air masses from the north (Yoshino, 1965).The latitudinal shift of the Baiu Front in response to orbital forcing may have resulted in the difference in the response of precipitation between central Japan and China (Yamamoto, 2009).Second, synchronous variations in stalagmite δ 18 O from Hulu Cave (Wang et al., 2001) and Sanbao Cave (Wang et al., 2008) in central China and from Dongge Cave (Yuan et al., 2004) in southern China suggest that the stalagmite δ 18 O is not a simple index of local precipitation (Yamamoto, 2009).The negative shift of stalagmite δ 18 O from Chinese caves is also synchronous with the rapid increase in air temperature in central Japan at glacial terminations, suggesting that the negative shift of stalagmite δ 18 O was related in part to temperature increases in air masses derived from the summer monsoon.Heavy rains, which occur more frequently in warmer air masses, induce isotopically lighter water (Yamamoto, 2009).Clemens et al. (2010) argued, however, that winter precipitation contributed to stalagmite δ 18 O.This question remains open.Our new result confirms the mismatch between the precipitation records in central Japan and Chinese stalagmite δ 18 O records.This is an interesting topic for studies of the Asian monsoon.
In this study, we assume that the branched GDGTs produced in the lake contributed to sedimentary branched GDGTs in the study site.However, the spike of high CBT (low estimated pH) at 13 ka is exceptionally affected by the contribution of soil GDGTs (Fig. 6).Ohira et al. (2013) reported a spike of abundant lignin at the same layer in core BIW08-B.Because of low-CBT-based pH (3.3 to 8.0 with an average of 5.0; Ajioka et al., 2014) due to high CBT in the soils, the high CBT (low estimated pH) at 13 ka associated with high lignin abundance likely reflected the soil GDGT contribution.

Changes in MBT /CBT -based winter lake water
temperature during the last 280 ky MBT /CBT-based winter temperature showed a glacialinterglacial pattern (Fig. 6).The variation is consistent with the winter air temperature estimated from the pollen assemblage in Lake Biwa (Nakagawa et al., 2008).However, the MBT /CBT-based temperatures (3 to 10 • C) are more realistic than the pollen-based temperatures (−10 to 5 • C; Nakagawa et al., 2008).The winter temperature in Lake Biwa is controlled by winter cooling (data from the Lake Biwa Environmental Research Institute, http://www.pref.shiga.lg.jp/biwako/koai/hakusyo/index.html;Japan Meteorological Agency, available at http://www.jma.go.jp/jma/ index.html)and reflects the intensity of the Asian winter monsoon.The rise of the MBT /CBT-based winter lake water temperature lagged behind the rise of the Tp-based summer air temperature by ∼ 10 ky in Lake Biwa cores (Fig. 7).This suggests warmer summers and colder winters in MIS 5e, with a larger seasonal temperature variation in this period.

Changes in the microbial ecosystem of Lake Biwa during the last 280 ky
The BIT index was higher than 0.8 throughout the core (Fig. 6).Because most branched GDGTs are produced in the lake (Ajioka et al., 2014), this high BIT implies high bacterial production in Lake Biwa.The BIT record indicates negative spikes (< 0.9) at 240, 210, 114 and 6-0 ka, which correspond to the maximal peak of crenarchaeol concentration (Fig. 6).Because crenarchaeol is specific to Thaumarchaeota (Sinninghe Damsté et al., 2002), the production of Thaumarchaeota was enhanced in those periods.Although the reason that thaumarchaeotal production was enhanced in these specific periods is unclear, changes in the nutrient conditions are a possible cause of such an ecological change.

Conclusions
We analysed the MBT /CBT indices in sediment cores retrieved from Lake Biwa to reconstruct changes in summer precipitation and temperature and in winter temperature in central Japan during the last 280 ky.Comparison with pollen assemblage in Lake Biwa cores suggests that lake water pH was determined by summer temperature in low-eccentricity periods, while it was determined by summer precipitation in high-eccentricity periods.From 130 to 55 ka, variation in lake pH (summer precipitation) lagged behind that in summer temperature by several thousand years.Thaumarchaeotal production was enhanced in specific periods in interglacials.

Figure 1 .
Figure 1.Map showing the study sites.

Figure 3 .
Figure3.Age-depth models of cores BIW07-6 and BIW08-B.The red, green, and blue dots indicate volcanic ashes, 14 C of plant debris and total organic carbon, and ARM events, respectively (see Tables1 and 2).Error bars indicate the 95 % confidence intervals (2σ ).

Figure 6 .
Figure 6.Variations in the concentrations of total isoprenoid and branched GDGTs and crenarchaeol, CBT-based pH, MBT /CBT-based temperature and BIT and MI in cores BIW07-6 and BIW08-B during the last 280,000 years.Red, blue, and green bars indicate age controls with 95% confidence intervals of volcanic ashes, the 14 C of plant debris and total organic carbon, and ARM events, respectively.

Figure 7 .
Figure7.Variations in CBT-based pH and MBT /CBT-based temperature from cores BIW07-6 and BIW08-B (this study); Tp and Cryptomeria (%) from cores BIW95-4 and BT in Lake Biwa(Hayashi et al., 2010a, b); the δ 18 O of stalagmites in Sanbao and Hulu caves in China(Wang et al., 2001(Wang et al., , 2008)); and eccentricity during the last 150,000 years.Red, blue, and green bars indicate age controls with 95 % confidence intervals of volcanic ashes, the 14 C of plant debris and total organic carbon, and ARM events, respectively.The names of tephras that appeared commonly in BIW07-6/BIW08-B and BIW95-4/BT are shown in the panel.

Table 2 .
Age controls of core BIW08-B.