4.1 Lake Maruwanminami-ike
4.1.1 Sedimentary sequences and the timing of the transition of marine to brackish lake and to freshwater lake environment
Large reservoir effect of 1,230 cal BP of the MwS4C-01 sediment core (Fig. 3) is probably attributed the mixing of glacially eroded marine sediment from surroundings of the lake, because the lake area has been marine environment in the past༎ In addition, Takano et al. (2015) reported that the subglacial weathering of silicate and aluminosilicate minerals in bedforms supplied significant levels of minerals, nitrogen, and relic carbon to subglacial meltwaters into the lake. They contain old and/or dead carbon, and thus cause large reservoir effect. The mean sedimentation rate was linearly calculated to be 0.389 mm/yr.
The transition periods from coastal marine to brackish lake, and to freshwater lake of Lakes Skallen-oike (Matsumoto et al. 2010) and Oyako-ike (Matsumoto et al. 2014) in the Soya Coast region are determined mainly based on the distribution of chlorobactene and/or bacteriopheophorbides derived from green sulfur bacteria (Pfennig 1967; Borrego and Garcia-Gil. 1994; Squier et al. 2002). Green sulfur bacteria require sulfide as an electron donor for photosynthesis, indicating the presence of a stratified water column with a chemocline and an anoxic (sulfidic) layer at the bottom of photic zone (Tani et al. 2009). In this period, the lake being isolated from the coastal marine environment and becoming stratified as the isolated marine water was overlain by freshwater supplied from meltwater to the lake surface.
Furthermore, those environmental transitions and changes in primary producers associated with the glacial-isostatic uplift are supported by the other biogeochemical evidence from δ15N, molecular signatures (DGGE-16S rRNA, denaturing gradient gel electrophoresis with 16S ribosomal RNA gene analysis), and the major elements of sedimentary facies (e.g., Al2O3-SiO2-Fe2O3 diagrams; Takano et al. 2012). The ongoing retreat of glaciers during and after the MHH, and ongoing isostatic uplift of the Soya Coast region accounts for this change as in the case of Lakes Skallen-oike and Oyako-ike.
In Lake Maruwanminami-ike sediment core (MwS4C-01), chlorobactene was found in depths of 134.6-65.6 cm with the most striking peak at a depth of 67.9 cm (Fig. 6). The MwS4C-01 sediment core in depths of approximately 70-0 cm was composed of mainly organic mud with lamination, while depths of 147-70 cm was mainly composed of diatomite with lamination (Fig. 3). TC, TOC, TN, TIC and TS contents, and TOC/TN and TOC/TS weight ratios in the core increased largely from a depth of approximately 70 cm to the surface (Fig. 4). This implies a change from marine to brackish/freshwater sediments in the core, since TS contents and TOC/TS ratios of marine sediments are generally much lower than those of freshwater lake sediments (Berner and Raiswell 1984; Sampei et al. 1997a, b). Marine diatoms such as Fragilariopsis nana and Navicula glaciei (Whitkowski et al. 2004) were found in depths of 70.2 and 65.6 cm, along with freshwater diatom of Psammothidium papilio (Kopalova et al. 2013) at a depth of 65.6 cm (Table 2). Marine diatoms are often dominated in brackish lakes in Vestfold Hills (Roberts and McMinn 1999; Verleyen et al. 2003). It is most likely, therefore, that the transition period of brackish lake environment (BLA) from coastal marine environment (CME) in the lake was between the depths of 72.5 (2,590 cal BP)-65.6 cm (2,500 cal BP, Fig. 3, Fig. 6).
Glacio-isostatic uplift rate of Lake Maruwaminami-ike was calculated by following equation. Lake altitude (mm asl)/(transition age of brackish lake (cal BP)–reservoir effect (cal BP)) = 11,200/(2,590-1,230)= 8.24 mm/yr. To our best knowledge, it is much higher than those of Lakes Skallen-oike (2.8 mm/yr, Matsumoto et al. 2010), Oyako-ike (2.2 mm/y, Matsumoto et al. 2014) and glacio-isostatic uplift rate within Skallen and Skarvesnes (3.2 mm / y, r2 = 0.98; Takano et al., 2012). A first-order linear model based on the radiocarbon dating from L. elliptica (data from Miura et al. 1998b) showed 3.9 mm/y (r2 = 0.68) by the compilation (the supplementary figure 3 in the Takano et al. 2012).
4.1.2 Coastal marine environment (CME, 147-72.5 cm, 5,010-2,590 cal BP)
The MwS4C-01sediment core in depths of 147-72.5 cm was mainly composed of diatomite with lamination (Fig. 3). TOC and TN contents are proxy of biomass and biological production in lake sediment cores (e.g. Matsumoto et al. 2003, 2010, 2012). TOC and TN contents ranging from 0.041 to 2.60% with an average of 1.53±0.56% (standard deviation) and from 0.11 to 0.0.47% with an average of 0.24±0.08% in the CME (147-72.5 cm depth), respectively, were considerably low as compared with those in the BLA (72.5-65.6 cm depth) and freshwater lake environment (FLE, 65.6-0 cm) as discussed below (Table 3, Fig. 4). These results showed low primary productivity in the CME as in the case of Lakes Skallen-oike (Matsumoto et al. 2010) and Oyako-ike (Matsumoto et al. 2014).
The compositions of n-C17 alkane and diploptene and long-chain n-alkanoic acids (C20-C36) in depths of 143.8-72.5 cm were very low less than 10%, while short-chain n-alkanoic acid (C12-C19) in these depths were relatively high with near constant of 70% (Fig. 5, cf. Additional files Tables S2-S3). C28 sterols (24-methylcholetsa-5,22-dien-3β-ol, 24-methylcholest-5-en-3β-ol, 24-methyl-5α-cholestan- 3β-ol) were very high (58.9-82.6%) in depths of 143.8-74.8 cm of the CME (Fig, 5, cf. Addiotional file Table S4). C28 sterols (mainly 24-methylcholest-5-en-3β-ol) are often major sterols of diatoms (Volkman et al. 1998; Matsumoto et al. 2003, Rampen et al. 2010) reflecting probably dominance of diatom assemblage in the CME. Marine diatom assemblage of Amphora oligotraphenta, Diadesmis perpusilla, Nabicula gregaria, N. glaciei, Fragilariopsis curta and F. nana (Gersonde and Zielinski 2000; Cramer et al. 2003; Pike et al. 2009) were distributed in these depths (Table 2).
Small amounts of chlorophyll a, pheophytin a, pyrophephytin a, were detected in depths of 143.8-72.5 cm (Fig. 6). They are commonly distributed in photosynthetic organisms (Tani et al. 2009; Matsumoto et al. 2010). Interestingly, chlorobactene was distributed in depths of 134.6-74.5 cm at small amounts (Fig. 6, cf. Additional file Table S5). This result strongly suggests the presence of green sulfur bacteria in anoxic conditions of the euphotic bottom water because the coastal marine depression water is probably stagnant during the glacio-isostatic uplift of the basin.
4.1.3 Stratified brackish lake environment (BLE, 72.5-65.6 cm, 2,590-2,500 cal BP)
TOC and TN contents in the BLE (72.5-65.6 cm depth) were 2.39±2.20% and 0.29±0.19%, respectively, which are considerably higher than those in the CME (Table 4). Small amounts of photosynthetic pigments of chlorophyll a, pheophytin a, pyropheophytin a, pheophorbide b and pyropheophorbide b were found in these depths. Of special interest is the occurrence of chlorobactene with very high concentration (3,830 µg/L) at a depth of 67.9 cm (Fig. 6). The chlorobactene can be interpreted by the presence of green sulfur bacteria as an indicator for the presence of a stratified water column with a chemocline and an anoxic (sulfidic) layer at the bottom of photic zone (Tani et al. 2009; Matsumoto et al. 2010).
Green sulfur bacteria are presumably contributed relatively high TOC and TN contents. In this period, Lake Maruwanminami-ike was isolated from the sea and becoming stratified as the isolated marine water was overlain by freshwater supplied from glacial meltwater to the lake surface as in the case of Lakes Skallen-oike (Matsumoto et al. 2010) and Oyako-ike (Matsumoto et al. 2014). Diatoms of Amphora oligotraphenta, Diadesmis perpusilla, Nabicula gregaria N. glaciei, Fragilariopsis curta and F. nana (Whitkowski et al. 2004) were distributed in this layer as in the case of the CME (Table 2). Verleyen et al. (2017) reported that marine diatom species could survive in saline conditions in Lake Kobatchi-ike in Skarvsnes for hundreds of years. Also, marine diatoms are often dominated in brackish lakes in Vestfold Hills (Roberts and McMinn 1999; Verleyen et al. 2003).
4.1.4 Freshwater lake environment (FLE, 65.6-0 cm, 2,500-1230 cal BP)
The MwS4C-01sediment core in depths of 65.6-0 cm was mainly composed of organic mud with lamination (Fig. 3). TOC and TN contents in the FLE (65.6-0 cm depth) were 5.75±4.75% and 0.67 ±0.50%, respectively, which are much higher than those in the CME and the BLE. Biological production in the FLE is much higher than those in the CME as in the case of Lake Skallen-oike (Matsumoto et al. 2010), Lake Oyako-ike (Matsumoto et al. 2014), Pup Lagoon (Verleyen et al. 2004) and Lake Reid (Hodgson et al. 2005) in the Larsemann Hills, East Antarctica. TIC% in the FLE/CME ratio is very high with 10.3 (Table 4) which can be explained by the presence of carbonate (e.g. Takano et al. 2015). TS% and TOC/TS ratio of the FLE/CME value were 1.40 and 3.81 times high, respectively, reflecting typical freshwater environment (Table 4; Berner and Raiswell 1984; Sampei et al. 1997a, b).
Abundance of n-C17 alkane in depths of 40.3-8.1 cm in the MwS4C-01 sediment core (Fig. 5) may be attributed to cyanobacteria and green algae because it is often most predominant n-alkane in these organisms (Weete, 1976; Matsumoto, 1993). High percentages of long-chain n-alkanoic acids (9.8-38.9%) in depths of 63.3-3.5 cm of the FLE as compared with those in depths of 143.8-72.5 cm in the CME (147-72.5 cm) may reflect the contribution of green algae such as Cosmarium clepsidora and Staurstrum sp. (Table 2, Matsumoto et al. 2010). Branched (iso and anteiso) alkanoic acids were predominant in depths of 49.5-8.1 cm in the FLE as compared with those in depths of 143.8-72.5 cm in the CME reflect the abundance of bacteria (Fig. 5; O’Learry et al. 1982; Reddy et al. 2002, 2003a,b). High percentages of diploptene in depths of 49.5-8.1 cm are consistent with high bacterial contribution (Fig. 5; Prahl et al. 1992; Elvert et al. 2001).
C29 sterols (mainly 24-ethylcholest-5-en-3β-ol) showed high percentage in the FLE (52.7-68.1%) and low percentages in the CME (less than 24%, Fig. 5). Although 24-ethylcholest-5-en-3β-ol is commonly used as a biomarker of vascular plants in the mid and lower latitudes, this sterol is widely distributed in Antarctic lake waters and sediments including the Soya Coast region as most predominant sterol in spite of the absence of vascular plants (Matsumoto et al. 1982, 1993, 2006; Matsumoto 1993; Volkman et al. 1998). Predominance of 24-ethylcholest-5-en-3β-ol in the FLE is probably derived from green algae and cyanobacteria (Matsumoto et al. 1982, 1993; Matsumoto, 1993) which are much abundant than those of diatoms. C27 sterols (mainly cholest-5en-3β-ol) are derived from microalgae including diatoms (Matsumoto et al. 1982, 2006; Volkman et al. 1998; Rampen et al. 2010), though C27 sterols in the FLE (63.3-3.5 cm depth) are somewhat higher than those in the CME (143.8-70.2 cm depth). It is much likely that the contribution of green algae is greater than diatoms (Fig. 5).
Ubiquitous photosynthetic pigments (chlorophyll a, pheophytin a and pyropheophytin a), pheophorbide b, and pyropheophorbide b) and a carotenoid (zeaxanthin) were found in the FLE (Fig. 6). These pigments are derived from Cyanophyceae of Leptolyngbya spp., Chlorophyceae of Cosmarium clepsydra and Straurstrum sp. and Bacillariophyceae of Amphora oligotraphenta, Diadesmis perpusilla, Navicula gregaria, Fragilariopsis curta and F. nana in the FLE (Table 2).
Very high TOC and TN contents in the FLE are mainly due to the contribution cyanobacteria and green algae rather than diatoms as evidenced by the abundance of 24-ethylcholest 5-en-3β-ol.
4.2 Lake Maruwan-oike
4.2.1 Sedimentary sequences and the timing of the transition of marine to brackish lake and to freshwater lake environments
Very large reservoir effect of 2,230 cal BP in the Mw4C-01 sediment core is probably attributed to the glacially eroded marine sediments deposited in the past and glacial erosion of bed rocks as in the case of L. Maruwanminami-ike. Average sedimentation rate was calculated to be 0.649 mm/y, although sedimentation rates were gradually decreased from the core bottom to the surface (Fig. 7). The sedimentation rate of the Mw4C-01sediment core is, however, 1.67 times higher than that of the MwS4C-01 sediment core from L. Maruwanminami-ike. Glacially eroded sediments of Lake Maruwan-oike are much greater than those of L. Maruwanminami-ike.
The transition period of CME, BLA and FLE are not clear in the visual observation of sedimentary facies of the Mw4C-01 sediment core. Vertical distribution of TOC, TN and TS contents (Fig. 8) showed no increasing trends with the transition from the CME, the BLA and the FLE as observed in Lake Maruwanminami-ike, and previously reported in Lakes Skallen-oike and Oyako-ike (Matsumoto et al. 2010, 2014; Takano et al. 2012).
Seto et al. (2002) reported the transitions from the CME to the BLE and to the FLE using Mw-4 sediment core (length 187 cm, water depth 9.8 m) from Lake Maruwan-oike. The vertical profile of the core is considerably different from our Mw4C-01 core. The CME (187-68 cm) was composed of diatomite mud with benthic forminifera Trochammina antarctica. The BLE was characterized by cyanobacterial mud with fine (<1 mm) laminated black-dark olive sediment formed in the anoxic conditions of depths of 68-60 cm with bottom ages of 3,430 yr BP without calibration of marine sediment age (-1,300 yr) and reservoir effects (no surface age datum). The FLE was composed of cyanobacterial mud with fine lamina (<1 mm) containing aquatic moss (Bryum pseudotriquetrum, Imura et al. 2003).
Here we estimated the transition period by the occurrence of bacteriopheophorbide c and bacteriopheophorbide d in depths of 107.0-79.4 and 47.2-28.8 cm with two large peaks (Fig. 10, cf. Additional file Table S8). It is most probable that first bacteriopheophorbide c and bacteriopheophorbide d peaks are due to green sulfur bacteria in the anoxic photic zone of the CME as in the case of Lake Maruwanminami-ike as discussed above. Diatom assemblage revealed the zonation of marine, brackish and freshwater taxa in depths of 226-35, 35-23 and 23-0 cm, respectively (Fig. 11). Indeed, Navicula cancellate (Ognjanova-Rumenova and Buczko 2015) and N. phyllepta (Vanelslander et al. 2009) are marine and/or brackish species, and Halamphora vyvermaniana (Van de Vijver et al. 2014) is distributed in freshwater environment (Fig. 11). These diatom assemblages support strongly changes from the CME to the BLE. The δ13C values in depths of 28.8-19.6 cm were relatively high ranging from -22.3 to -12.9‰ with an average of -16.8±4.2‰ (Table 5) which reflect probably cyanobacteria and mosses in the FLE as in the case of Lake Skallen-oike (Matsumoto et al. 2010). The transition periods from the CME to the BLA and to the FLE are, therefore, occurred in the second peak of bacteriopheophorbide c and bacteriopheophorbide d in depths of 47.2-28.8 cm in the lake. It is consistent with the result of Takano et al. (2015). They reported that the transition ages from marine to freshwater lake are between 3,382 and 3,560 cal BP (median age 3,371 cal BP) in depths of 49-46 cm based on vertical profile of major elements (SiO2, Fe2O3, Al2O3, TiO2, CaO, K2O) and microflora differences of DGGE (denaturing gradient-gel electrophoresis) in Lake Maruwan-oike sediment core (Mw5S, core length 156 cm).
Glacio-isistatic uplift rate of Lake Maruwan-oike was calculated by the same method of Lake Maruwanminami-ike. 8,000 mm /(3,190 cal BP-2,220 cal BP) = 8.25 mm/y. This result is consistent with the result of Lake Maruwanminami-ike (8.24 mm/y). Crustal uplift of these adjacent lakes occurred at the same time. They are much higher than those of Lakes Skallen-oike (2.8 mm/y; Matsumoto et al. 2010) and Oyako-ike (2.2 mm/y; Matsumoto et al. 2014). The differences of these crustal uplift rates were discussed below (4.3).
4.2.2 Coastal marine environment (226-47.2 cm, 5,700-3,190 cal BP)
The soft X ray photograph of the Mw4C-01 sediment core in depths of 226-47.2 cm (Fig. 7) showed mainly light gray with lamination or globular structures, and consistent with low TOC and TN contents. High TIC peak was found in depths of 182.9-162.2 cm which is probably contribution of sea shelf such as bivalve fossils in Rundvagshetta region (Fig. 8; Hirakawa and Sawagaki, 1998; Takano et al. 2015).
TOC and TN contents ranging from 0.82 to 3.82% with an average of 2.17±0.56% and from 0.22 to 0.67% with an average of 0.43±0.10% in the CME (224.3-47.2 cm depth), respectively, were similar to those in the BLA (47.2-28.8 cm depth), but considerably high as compared with those in the FLE (28.8-0 cm depth) as discussed below (Table 6, Fig. 8). These results reflect the low primary production in the FLE which are very different from those in the Lake Maruwanminami-ike discussed above.
Antarctic marine sediments influenced by phytoplankton have δ13C values lower than -22% (Boutton 1991). In the Mw4C-01 sediment core, δ13C and δ15N values in depths of 224.3-47.2 cm ranged from -23.7 to -18.8‰ with an average of 21.1±1.1 (standard deviation, and from 1.3 to 7.0 with an average of 4.2±1.3‰, respectively (Fig. 9, Table 5). The δ13C values were slightly higher than that of present Antarctic marine sediments, while δ15N values were somewhat lower than the range of diatom-bound δ15N in the Holocene sediment cores from the Antarctic Ocean (Robinson and Sigman 2008). These results suggest the contribution of terrestrial organic matter inputs as in the case of Lake Skallen-oike (Matsumoto et al. 2010).
Chlorophyll a, pheophytin a and pyropheophytin a are commonly distributed in photosynthetic organisms (Tani et al. 2009; Matsumoto et al. 2010). Cis-diatoxanthin is distributed in green algae, diatom and cyanobacteria, but pheophytin b and pyropheophytin b are distributed in green algae, brown algae and vascular plants. These pigments are come from green algae, because no brown algae and vascular plants are distributed in the studied area. Cis-alloxanthin is a typical carotenoid of Cryptophyta (Jeffrey et al. 1997; Tani et al. 2009; Matsumoto et al. 2010).
The presence of bacteriopheophorbide c and bacteriopheophorbide d in depths of 107.0-79.4 cm (Fig. 10; 4,700-4,070 cal BP) strongly suggests the presence of green sulfur bacteria in the photic anoxic environment caused by stagnant bottom seawater in the depression of shallower coastal marine environment due to glacio-isostatic uplift.
Marine diatoms of Thalassiosira australis, Fragilariopsis curtca, F. nana/cyrindrus, Navicula directa and N. glaciei, etc. were distributed in depths of 226-49.6 cm of the Mw4C-01 sediment core (Fig. 11). These diatom species are generally related to sea ice and ice edge environment (Cremer et al. 2003; Pike et al. 2009). The marine environment is covered with sea ice. T. australis and T. antarctica are associated with relatively open water of marginal ice edge environments (Taylor et al. 1997; Zielinski and Gersonde 1997).
4.2.3 Stratified brackish lake environment (BLE, 47.2-28.8 cm, 3,190-2,890 cal BP)
Laminated organic mud sediment (47.2-35 cm) and cyanobacterial sediment (35-28.5 cm) were distributed in this layer. TOC and TN contents ranged from 1.34 to 3.06% with an average of 1.95±0.57% and from 0.22 to 0.50% with an average of 0.32±0.09%, respectively (Table 6). Biological production and biomass in the BLE are somewhat lower than those in the CME.
The δ13C and δ15N values in depths of 47.2-28.8 cm of the Mw4C-01 sediment core ranged from -21.5 to -12.9‰ with an average of -19.2±3.2‰ and from 4.6 to 6.1‰ with an average of 5.6±0.6‰, respectively (Fig. 9, Table 5). These values reflect probably the contribution of terrestrial organic matter and green sulfur bacteria in the BLE, because δ13C values of green sulfur bacteria (Prosthecochloris sp.) are -19.3 - -22.9‰ with an average of -21.5±1.2‰ (Zyekun et al. 2009) and not distinguish with those of terrestrial organic matter input. Ubiquitous photosynthetic pigments of chlorophyll a, pheophytin a and pyropheophytin a in the BLE were very small, but considerable amounts of pheophorbide b and pyropheophorbide b were detected in the BLE (Fig. 10). Cyanobacteria and diatoms are important primary producer in this layer, but red algae are absent in the lake. The occurrence of bacteriopheophorbide c and bacteriopheophorbide d reflects probably the presence of green sulfur bacteria as an indicator for the formation of a stratified water column with a chemocline and an anoxic (sulfidic) layer at the bottom of photic zone. The L.ake Maruwan-oike basin being isolated from the sea and becoming stratified as the isolated marine water was overlain by freshwater supplied from meltwater to the lake surface as in the case of Lake Maruwanminami-ike as discussed above.
4.2.4 Freshwater lake environment (FLE, 28.8-0 cm, 2,890-2,240 cal BP)
This sediment layer is composed of cyanobacterial sediment with mosses (28.5 cm-surface). Soft X-ray of these layer were dark gray of cyanobacterial and moss structures (Fig. 7). TOC contents ranging from 0.12 to 6.49% with an average of 1.36±1.66% in the FLE (28.8-0 cm), were considerably lower than those in the CME (224.3-47.2 cm) and the BLA (47.2-28.8 cm, Table 6), and much lower than those in the FLE of L. Maruwanminami-ike (9.11±6.84%) as discussed above and those in Lake Skallen-oike (11.7±3.3%, Matsumoto et al. 2010) and Lake Oyako (6.84±3.33%, Matsumoto et al. 2014). These results showed very low biological production and biomass in the FLE mainly due to low distribution of photosynthetic organisms of cyanobacteria, green algae, diatoms and mosses with continuously supplied large amounts of glacially eroded bedforms as evidenced by large reservoir effect. Very low TOC and TN contents in depths of 15.0-1.2 cm may be significant contribution of glacially eroded sediments (Fig. 8).
High TIC peak in depths of 30-28 cm is probably contribution of carbonate as in the case of Lake Maruwanminami-ike discussed above (Fig. 8). It is likely that very low TS contents in the FLE are attributed to the influence of marine sediments in the past and/or eroded metamorphic base rocks.
The δ 13C and δ15N values in depths of 28.8-1.2 cm of the sediment core ranged from -30.1 to -12.9‰ with an average of -20.1±5.6‰ and from 2.5 to 7.9‰ with an average of 5.5±0.6‰, respectively (Fig. 9, Table 5). These values varied largely and reflect probably cyanobacteria and mosses with complex sedimentation.
Chlorophyll a and their derivatives and carotenoids were almost absent in the FLE (Fig. 10). The diatom assemblage in the FLE is composed mainly of Antarctic endemic freshwater diatom taxa, Humidophila australis and Psammothidium papilio in the FLE of Lake Maruwan-oike (Fig. 11, Sabbe et al. 2003). These two diatom species are major diatoms in the FLE of Lake Maruwanminami-ike as discussed above (Table 2).
4.3 High glacio-isostatic uplift rates of Rundvågshetta lakes
Here we discuss high glacio-isostatic uplift rates of 8.24 and 8.25 mm/yr in Lakes Maruwanminami-ike and Maruwan-oike, respectively. They are much higher than those in raised beach deposits from the northwest side (N940114-3A. 3.7 mm/yr; d940112-3B, 2.5 mm/yr) of Lake Maruwan-oike and from the northwest sites of some hundreds of meters from the lake (c940112-3, 2.9 mm/yr; M940112-4, 2.9 mm/yr; Fig. 2, Table 7). They were calculated by altitude and corrected marine age (-1,300 yr) data of Hirakawa and Sawagaki (1998). Furthermore, they are much higher than our calculated datum of 2.3 mm/yr of Mw5S sediment core from Lake Maruwan-oike reported by Takano et al. (2015). Their big discrepancy is ascribed to the difference of age pattern of our Mw4C-01core and Mw5S core. Takano et al. (2015) reported ages of the core top (0-2 cm) and 6-8 cm depths of the Mw5S core were 1,350 and 3,950 cal BP, respectively, whereas ages of the core top (0-2.4 cm) and 20.7-23.1 cm depths of our Mw4C-01 core were 2,240 and 2,785 cal BP, respectively (Table 3). Takano et al. (2015) explained the occurrence of hiatus at the core top of the Mw5S, while no hiatus was found in our Mw4C-01core.
It is much likely that the sedimentation in Lake Maruwan-oike is very different in the coring sites, because the Mw4C-01 and Mw5S sediment cores were taken in water depths of 15.7 m and 20.2 m (Takano et al. 2015), respectively. Besides, vertical profile of Mw-4 sediment core from Lake Maruwan-oike (water depth 9.8 m; Seto et al. 2002) was considerably different from the Mw4C-01sediment core. The maximum depth of Lake Maruwan-oike is 37 m (Imura et al. 2003) and thus coring depths are slopes of the lake. Slumping and reworks may be occurred in a certain place and form hiatus in the slopes. If core top of the Mw5S was omitted and extraporated age of core top was obtained ca.2,500 cal BP. Glacio-isostatic uplift rate can be, therefore, calculated to be 8,000/(3,560-2,500) = 7.5 mm/yr which is consistent with our results. These glacio-isostatic uplift rates of Lakes Maruwanminami-ike and Maruwan-oike are much higher than other isolated basins and raised beaches in the Soya Coast region (Table 7).
In order to explain these difference of the uplift rate estimation, further researches using new sediment cores from the deepest sites of the Rundvågshetta lakes other than slopes. It is possible that glacio-isostatic uplift rates of raised beach sediments and isolated basins are different in the Rundvågshetta lakes, because lake sediments are supplied from glacially eroded marine sediments and bed rocks evidenced by very high reservoir effects. Further studies on chronological models and high-precision GPS survey (e.g. Ohzono et al. 2006; Argus et al., 2014) will have to be required with re-assessments of relic carbon flux from sub-glacial meltwater for refining the calendar age reconstruction (e.g. Sawagaki and Hirakawa, 1997; Wingham et al. 2006; Takano et al. 2015, cf. insight from 10Be/9Be profiles from Sproson et al., 2021).
4.4 Paleoenvironmental change in the Soya Coast region and in East Antarctica
Since the LGM glacial retreat and subsequent crustal uplift has been occurred in Antarctica including the Soya Coast region. Kawamata et al. (2020) studied 10Be and 26Al radio isotopes for ground surface exposure dating and demonstrated that the ice sheet completely covered the highest peak of Skarvsnes (400 m asl.) prior to ca. 9 ka and retreated eastward by at least 10 km during the early to middle Holocene (ca. 9 to 5 ka BP). The timing of the abrupt ice-sheet thinning and retreat is consistent with the intrusion of modified Circumpolar Deep Water (mCDW)into deep submarine valleys in the Lützow-Holm Bay.
In the previous study on raised beach deposits at higher altitudes greater than 20 m asl are reported in the Ongul Islands (35.0 m), Langhovde (27.4 m), Skarvsnes (39.0 m), Skallen (32.3 m), Skallevikhalsen (23.9 m) and Rundvågshetta (23.0 m, Omoto, 1977) in the Soya Coast region as well as Cape Hinode (30-35 m) in the Prince Olav Coast (Yoshida and Moriwaki, 1979, Fig. 1), but no further scientific confirmation including dating was done for these higher altitude samples until today. Fossil marine organisms mainly molluscs, foramiminifers and serpuloid tubes in beach deposits are distributed below 20 m asl (Hayashi and Yoshida, 1994; Igarashi et al. 1995a,b; Hirakawa and Sawagaki,1988; Miura et al. 1998a,b, 2002).
Miura et al. (1998a,b, 2002) demonstrated based on the two groups formed in Late Pleistocene (30,000-46,000 yBP), and in the middle Holocene (3,000-7,200 yBP) in the Ongul Islands, Langhovde and Skavsnes. They concluded that East Antarctic Ice Sheet (EAIS) retreated from the northern Soya Coast and a transgression occurred prior to the LGM, and the EAIS had not re-advanced over the beach of the northern Soya Coast region even during the LGM, as evidenced by the existence of 30,360-46,000 yr BP in situ fossil shells. Holocene transgression and subsequent retreat had been occurred in the northern and southern Soya Coast regions during the Holocene. Miura et al. (1998a,b) showed high sea level had reached at least 20 m asl without taking isostatic rebound into consideration (Fig. 1, see Table 7).
Table 7 summarizes typical RSL results, transition ages and glacio-isostatic uplift rates during the Holocene in the Soya Coast region and some related areas in East Antarctica. We use mainly AMS 14C dating data and some selected β ray dating data of 14C ages which are consistent with AMS 14C dating data. Because older fossils occur very close proximity to younger fossils in beach deposits of the Ongul Islands and Langhovde north, certain 14C ages determined by β ray dating method, which requires relatively large amounts of samples, were obtained by the mixture of old and younger fossils especially approximately 10-28 kyr BP samples (Hayashi and Yoshida, 1994; Igarashi et al. 1995b; Miura et al.1998b).
Verleyen et al. (2017) provided new data on deglaciation history and develop new RSL curves along an 80 km transect in the Lützow Holm Bay region. The minimum radiocarbon age for regional deglaciation is ca.11,240 cal. yr BP (no marine age subtraction of -1,300 yr) on West Ongul Island with progressively younger deglaciation ages (from Langhovde to Skarvsnes and Skallen) approaching the main regional ice outflow at Shirase Glacier (Fig. 1, Table 7). AMS 14C dates of fossils in raised marine deposits and data from isolation basins have further refined this record similar RSL curves for the Ongul Islands, Langhovde and Skallen, but a strikingly different RSL curve for Skarvsnes that sediment cores from an isolation basin in combination with 14C dates of marine fossils deposited in the sill revealed a RSL high stand of 32.7 m around 4,400 cal BP (Verleyen et al. 2017). The difference in RSL high stand of Langhovde is tentatively linked to neotectonic processes such as reactivation of a local fault in Skarvsnes (Verleyen et al. 2017). The reconfirmation of the evidence of marine fossils at higher altitude and reactivation of a local fault will be necessary.
Transition ages from marine to lake environments of 11 sediment cores in 8 lakes (isolated basins) were discussed for the Ongul Islands (Lakes Yumi-ike and Oike), Skarvsnes (Lakes Oyako-ike, Mago-ike and Kobati-ike), Skallen (Lake Skallen-oike) and Rundvagshetta (Lakes Maruwanminami-ike and Maruwan-oike; Fig. 1, Table 7). The RSL and transition age of each lakes are as follows: Yumi-ike (10 m asl. 4,370 yr BP), Oike (13, 5,720), Oyako-ike (2.37, 1,060), Oyako-ike (2.37, 1,080), Mago-ike (1.5, 1,380), Kobati-ike (28, 2,120), Skallen-oike (10, 3,440), Skallen O-ike (9.64, 2,940), Maruwanminami-ike (11.2, 1,380), Maruwan-oike (8.0, 970), Maruwan-oike (8.0, 3,471; Table 7). Altitudes (m asl) of all raised beach data are correlated with transition ages with a correlation coefficient of r2 = 0.291 (Fig. 12A). Transition age of Lake Kobati-ike was small in spite of high altitude. Maruwanminami-ike and Maruwan-oike were high reservoir effect as discussed above. The transition ages of all raised beaches and isolated basins (Table 7) ranged from 970 to 9,290 yr BP with an average of 3,660±1,520 yr BP (standard deviation) suggesting that major warm periods in the Soya Coast region are the middle Holocene so-called the MHH. It is consistent with that the warm period of the Lützow-Holm Bay region is 4,800-3,000 yr BP (Verleyen et al. 2011).
Holocene RSL, transition ages and crustal uplift rates in East Antarctica have been reported in the Vestfold Hills, Larsemann Hills, Windmill Islands, Bunger Hills and Scott Coast (Table 7). Zwartz et al. (1998) studied Organic L. (3.5 m asl, 3,465 yr BP), Watts L. (4.3, 3,630), Highway L. (7.7, 4,755), Druzhby L. (8.0, 5,720), Anderson L. (8.4, 5,505), Ace L. (8.8, 4,550) in Vestfold Hills. They subtracted 1,300 years for marine reservoir correction. We showed here mean ages. Verleyen et al. (2020) summarized Larsemann Hills (isolated basin; 8 m asl, 7,000-7,600 yr BP), Windmill Islands (raised beach; 30, 6,900), Bunger Hills (raised beach; 8, 5,000-5,600 yr BP; Figurmoye L. (isolated basin; 11.2, 5,000-8,000) and Scott Coast (raised beach; 32, 6,600; Table 7). These transition ages are somewhat greater than those in the Soya Coast region. According to the warm periods of Verleyen et al. (2011), Larsemann Hills (7,300-5,400 yr BP) and Bunger Hils (5,500-2,000 yr BP) are in the range of the transition ages, while those in Vestfold Hills (8,000-5,500 yr BP) are somewhat higher than the transition ages, but Windmill Islands (4,000-2,000 yr BP) is much lower than the transition ages. Further studies are required for the relationships between Antarctic regional warm periods and the transition ages.
4.5 Glacio-isostatic uplift rates
Omoto (1977) studied raised beach deposits and found glacio-isostatic uplift rates of the Ongul Islands, Langhovde and Skarvsnes less than 6,000 yr BP are 2.7 mm/yr, no data and 2.6 mm/yr, respectively. Crustal uplift rates of submerged ice-free areas to RSL seems to be at least 2.5 mm/yr on the average during the last 6,000 years (Yoshida and Moriwaki, 1979, Table 7).
Miura et al. (1998a,b, 2002) studied the stratigraphy of raised beaches in the sediment layers of excavated trenches and provide sample elevation and AMS 14C age, but no crustal uplift rates are given. We calculated marine reservoir corrected age (-1,300 yr) and glacio-isostatic uplift rates based on data of Miura et al. (1998a,b, 2002, Table 7). Older group formed in Late Pleistocene (30,000-46,000 yBP) are found in East Ongul Island and part of Langhovde north. Their glacio-isostatic uplift rates were very low less than 0.35 mm/yr. Younger groups formed in the middle Holocene of Langhovde north, Langhovde south and Skarvsnes were ranging from 0.72 to 3.02 mm/yr with an average of 1.97±0.71(standard deviation)mm/yr, 0.39 to 3.36 mm/yr with an average of 1.47±1.2 mm/yr and from 1.25 to 2.81 mm/yr with an average of 2.10± 0.49 mm/yr, respectively (Table 7).
Glacio-isostatic uplift rates of lakes and ponds of isolated basins were studied in the Soya Coast region. Here we calculated crustal uplift rates of Lakes Yumi-ike, O-ike, Mago-ike and Kobachi-ike using data of Verleyen et al. (2017) and its supplementary tables (Table 7).
Yumi-ike: 10,000 mm/(4,830 cal BP - 460 cal BP) = 2.3 mm/yr
O-ike: 13,000 mm/(5,720 cal BP – 0 cal BP) = 2.3 mm/yr
Mago-ike: 1,500 mm/(1,380 cal BP – 0 cal BP) = 1.1 mm/yr
Kobachi-ike: 28,000 mm/(2,800 cal BP – 680 cal BP) = 13.2 mm/yr
Glacio-isostatic uplift rates of Lakes Yumi-ike (10 m asl) and O-ike (13 m asl) in the Ongul Islands were 2.3 mm/yr and 2.3 mm/yr, respectively. Those of Lake Oyako-ike (2.37 m asl) in Skarvsnes are 3.2 mm/yr (Takano et al. 2012) and 2.2 mm/yr (Matsumoto et al. (2014). Lakes Mago-ike (1.5 m asl) and Kobachi-ike (28 m asl) in Skarvsnes were 1.1 and 13.2 mm/yr, respectively. Very high crustal uplift rate of Lake Kobachi-ike is not clear, though it is linked with high crustal uplift rate of raised beach data (7.4 mm/yr; Table 7).
Glacio-isostatic uplift rates of Lake Skallen-oike (10 or 9.64 m asl) are 2.8 mm/y (Matsumoto et al. 2010) and 3.2 mm/y (Takano et al. 2012). Calculated crustal uplift rate of L. eliptica in raised breach deposit from L. Skallen-oike (12 m asl, 4,720 yr BP; Igarashi et al. 1995a,b) is 2.5 mm/y which is similar to those of lake sediment core results (Table 7). These uplift rates are consistent with present uplift rates estimated by bedrock GPS (2.3±0.3 mm/yr) and Very Long Baseline Interferometry at Syowa Station in East Ongul Is. (VLBI, 4.6±2.2 mm/y; Kaminuma 2008).
Glacio-isostatic uplift rates of raised beaches and isolated basins in the Soya Coast region ranged from 0.19 to 4.40 with an average of 2.0±0.92 mm/yr except for 4 higher values (7.4, 13.2,8.2, 8.3 mm/yr, Table 7). The glacio-isostatic uplift rates of the Soya Coast region are similar to those estimated in the Lambert Glacier region (1.8-1.9 mm/yr; Verleyen et al. 2005), Vestfold Hills (1.0-1.9 mm/yr; Zwart et al. 1998), Larsemann Hills (1-1-1.14 mm/yr; Verleyen et al. 2005), Bunger Hills (2.2-1.4 mm/yr; Adamson and Colhoun, 1992; Verkulich et al. 2002), but lower than Scott Coast (4.8 mm/y; Hall et al. 2004; Table 7).
The altitudes of raised beaches in the Soya Coast region are correlated with transition ages as discussed above (Fig. 1). The AMS 14C ages of altitudes less than 20 m are mainly obtained from marine fossils of L. elliptica in the sediment layers of excavated trenches (Miura et al. 1998a,b, 2002). Their ages reflect probably coastline in the past because glacial retreat cause rising beaches continuously. Marine organism of L. elliptica fossilized due to changing from marine to terrestrial environment. Glacio-isostatic uplift rates of raised beaches are correlated with altitudes with a correlation coefficient of r2 = 0.724 (Fig. 12B). Increasing crustal uplift rates with altitudes reflect continuous crustal uplifts in the Soya Coast region.