An early Holocene sea-level jump and delta initiation



[1] Early Holocene sea-level change controlled the evolution of clastic coastal depositional systems. Radiocarbon-dated borehole cores obtained from three incised-valley-fill systems in Asia (Changjiang, Song Hong, and Kiso River) record very similar depositional histories, especially between about 9000 and 8500 cal BP. Sedimentary facies changes from estuarine sand and mud to shelf or prodelta mud suggest that the marine influence in the incised valleys increased during this period. In addition, large decreases in sediment accumulation rates occurred. A sea-level jump causes an estuarine system and its depocenter to move rapidly landward. It is possible that the final collapse of the Laurentide Ice Sheet, accompanied by catastrophic drainage of glacial lakes, at approximately 8500 cal BP caused such a jump. The jump was followed immediately by a period of decelerated sea-level rise that promoted delta initiation.

1. Introduction

[2] Ages and elevations of fossil coral reefs and siliciclastic sediments have shown that the sea-level rise since the last glacial maximum (LGM) was not monotonic [e.g., Fairbanks, 1989; Bard et al., 1996; Hanebuth et al., 2000]. Meltwater pulse (MWP)-IA and MWP-IB (∼14 000 and 11 300 cal BP) are well recognized catastrophic sea-level-rise events, and a relationship between these events and abrupt climate change recorded in ice, lake, and marine cores is discussed [Lehman and Keigwin, 1992]. Moreover, studies on sea-level change and its rate during the early Holocene [Blanchon and Shaw, 1995; Törnqvist et al., 2004] are particularly important, because it is likely that they are closely linked to the most prominent Holocene climatic cooling event, which occurred at 8200 cal BP [Alley et al., 1997], and the final deglaciation and melting of the Laurentide Ice Sheet [Andrews et al., 1995; Kerwin, 1996], which was accompanied by catastrophic release of freshwater from glacial lakes [Barber et al., 1999; Teller et al., 2002; Clarke et al., 2004].

[3] The early Holocene sea-level change also controlled the evolution of clastic coastal depositional systems. Stanley and Warne [1994], for example, argue that a deceleration of sea-level rise after about 8000–7000 14C y BP promoted the formation of the world's modern deltas. Many of these deltas are underlain by estuaries, which are common transgressive depositional systems that develop when the rise in sea level overtakes the sediment supply [Boyd et al., 1992]. However, the transition from estuary to delta in response to sea-level change is not well understood.

[4] These two types of depositional systems are well recorded in incised-valley-fill deposits. An incised valley is highly likely to preserve fill deposits, which accumulate continuously during a sea-level rise, and through the following period of stable sea level [Dalrymple et al., 1994], in part because it occupies a low topographic position [Demarest and Kraft, 1987], and in part because it receives a relatively large sediment discharge from the river. Here we focus on incised-valley-fill deposits to investigate the relationship between incised-valley-fill successions and Holocene sea-level change, especially between 9000 and 8000 cal BP.

2. Materials and Methods

[5] Three incised-valley-fill systems in Asia, the Changjiang (Yangtze River), Song Hong (Red River), and Kiso River, were studied (Figure 1). All three rivers have a large sediment discharge owing to monsoon-driven heavy rainfall in the catchment and the young mountainous catchment geology, leading to the formation of thick sedimentary strata at their mouths. Moreover, the part of Asia in which they are located is far from the site of the continental ice sheets of the LGM and therefore has been less influenced by the crustal readjustment induced by glacial isostasy.

Figure 1.

Location map of the study area.

[6] Borehole cores have been retrieved from the subaerial and subaqueous deltas located above these incised valleys: CM97 and HQ98 (Changjiang) [Hori et al., 2001a, 2001b], ND-1 (Song Hong) [Tanabe et al., 2003], and KZ [Yamaguchi et al., 2003], and two additional borehole cores [Masuda and Iwabuchi, 2003] (Kiso River). Here, we call the two borehole cores described by Masuda and Iwabuchi [2003] Ise Bay 1 and 2 (IB1 and IB2, respectively). Only IB1 and IB2 were obtained from the subaqueous delta.

[7] Sedimentary facies in each core have been described in detail in the references cited above according to their lithology, color, sedimentary structures, grain size, and fossils. All radiocarbon ages were determined on molluscan shells and plant materials by using the accelerator mass spectrometry (AMS) method. Although calibrated ages have been already reported from the Song Hong and the Kiso River, we have recalculated calibrated ages according to the updated CALIB program (M. Stuiver et al., CALIB radiocarbon calibration (Calib Rev 5.0.1), 2005, available at by considering the regional marine reservoir effect, ΔR. For the calculation of ages from molluscan shells, ΔR was regarded for the Changjiang as −96 ± 60 yr obtained from Tsingtao [Southon et al., 2002], for the Song Hong as −25 ± 20 yr from the South China Sea [Southon et al., 2002], for the Kiso River as 82 ± 33 yr from the Pacific coast of central Japan [Shishikura et al., 2007], and the marine carbon component as 100%.

3. Changes in Sedimentary Facies and Accumulation Rates Between About 9000 and 8500 cal BP

[8] The sedimentary facies of all of the borehole cores changed considerably between about 9000 and 8500 cal BP: in CM97, at 9400–8000 cal BP; in HQ98, around 9000 cal BP; in ND-1, at 8900–7700 cal BP; in KZ, at 9200–8500 cal BP; and in IB2, after 9400 cal BP. In the Changjiang cores, estuarine facies are overlain by shelf mud/prodelta facies (Figure 2). These estuarine facies are characterized by sand–mud couplets, indicating a strong tidal influence. In contrast, shelf mud/prodelta facies consist predominantly of silty clay containing shell fragments and abundant foraminifers [Hori et al., 2001b]. In the Song Hong core, massive clay, interpreted as sea-floor or prodelta sediments, overlies laminated silty clay [Tanabe et al., 2003]. An alternation of sand and mud is overlain by massive clay to silt in the KZ core (Kiso River) [Yamaguchi et al., 2003]. Moreover, diatom analysis has shown that the marine influence was strong during the deposition of the massive clay and silt. An intermediate sand and mud unit is also overlain by a silt unit in the IB1 and IB2 cores [Masuda and Iwabuchi, 2003]. These results show that in each case, the marine influence on the sedimentary environment of the incised valleys increased between 9000 and 8500 cal BP.

Figure 2.

Simplified geological column and age–depth plot for the CM97 core, collected from the Changjiang incised valley. Accumulation rates were estimated from the age–depth plot.

[9] Sediment accumulation rates can be determined from age–depth plots of core sediments (Figure 2). Average depositional rates at all sites were high (ca. 3–20 m/y) during 11 000–9000 cal BP (Figure 3). In contrast, a large decrease in the rate of deposition occurred at all sites nearly simultaneously, at approximately 9000–8500 cal BP. For example, accumulation rates at CM97 dropped significantly from 6.8 to 1.1 m/ka during the period. Similarly, a seismic section that includes the IB1 and IB2 cores suggests that an erosional surface or thin, condensed horizon, indicating slow or no deposition, separates the intermediate sand and mud and the overlying silt unit [Masuda and Iwabuchi, 2003]. This timing correlates well with the sedimentary facies transition described above.

Figure 3.

Age–altitude plots of four incised-valley-fill deposits and uplift-corrected corals from the Kwambu drillhole, Huon Peninsula [Chappell and Polach, 1991; Ota and Chappell, 1999]. Notably, a large decrease in the sediment accumulation rates is observed in all incised-valley-fill systems at approximately 9000–8500 cal BP, shown by the vertical shaded bars.

[10] After this decrease, accumulation rates again increased clearly and were accompanied by delta progradation. The Changjiang delta front passed the HQ98 and CM97 sites around 5000 and 1000 cal BP, respectively, an age difference that is explained by the locations of the borehole sites: the HQ98 site is approximately 130 km landward of the CM97 site [Hori et al., 2001a]. A similar spatial relationship is also observed between the KZ core on the subaerial Kiso River delta and the IB1/IB2 cores, collected about 30 km seaward of the KZ core, but as the delta front has not yet passed the IB1 and IB2 sites, accumulation rates continue to be slow at these sites.

4. Discussion

[11] Both sedimentary facies and sediment accumulation rates at the three sites changed markedly at approximately 9000–8500 cal BP (Changjiang, 9400–8000 cal BP; Song Hong, 8900–7700 cal BP; Kiso, 9400–8500 cal BP). Moreover, similar changes also occurred between 8800 and 8500 cal BP near the mouth of the Mekong River, observed in the DT1 core [Ta et al., 2005]. Other factors affecting incised-valley-fill systems, such as climate, water and sediment discharge, river-bed gradient, tides and waves, and incised-valley shape, vary greatly among the three incised-valley-fill systems studied. Therefore, sea-level change caused by glacial eustasy must be responsible for this near simultaneous change in sedimentary facies and accumulation rate among the sites. We infer that an abrupt sea-level jump around 9000–8500 cal BP caused each estuarine system and its depocenter to move rapidly landward (Figure 4). As a result, the marine influence at the borehole sites increased, as in a shelf environment, with very slow or no sediment accumulation. The abrupt sea-level jump also affected the sediment stacking of these coastal depositional systems. A clear flooding surface at the base of the present deltaic succession has been found by high-resolution seismic survey within a Holocene succession offshore of the Changjiang mouth [Liu et al., 2007b] and in the northern Yellow Sea along with an abrupt facies change in micropaleontological data [Liu et al., 2007a].

Figure 4.

Schematic model of the response of a river-mouth depositional system to sea-level rise. (1) Transgression. Aggradation with a large accumulation rate occurs in an estuary during a sea-level rise. However, the shoreline retreats because the increasing accommodation surpasses the infilling of the estuary with sediments supplied by the river. (2) Rapid transgression. A sea-level jump backsteps the estuarine system, and the marine influence at the borehole site increases, as in a shelf environment, with slow or no sediment deposition. (3) Regression. The following decelerated sea-level rise leads to the initiation of delta formation because sediment supply overtakes accommodation. The accumulation rate at the borehole site again becomes high as the delta front approaches the site.

[12] The rate of sea-level rise was much larger than the sediment accumulation rates at the borehole sites just before the jump. It is estimated, for example, that the rate was over approximately 20 m/ka based on the accumulation rates at the ND-1 site (Song Hong).

[13] The jump probably occurred within a very short time (less than several hundred years) and was followed immediately by a large deceleration of sea-level rise, because delta progradation beginning by at least 8100 cal BP has been observed in a borehole core retrieved from 50 km landward of the ND-1 site in the Song Hong delta [Hori et al., 2004]. This finding is consistent with a period of slow or no sea-level rise centered on ca. 7700 cal BP at Singapore [Bird et al., 2007] and the timing of worldwide delta initiation proposed by Stanley and Warne [1994]. The slow sea-level rise led to regression and the transition from an estuary to a delta, that is, delta initiation or progradation, as illustrated in Figure 4.

[14] The final collapse of the Laurentide Ice Sheet, accompanied by the catastrophic draining of glacial lakes Agassiz and Ojibway, at about 8500 cal BP (or 7700 14C y BP) likely was related to the jump. If these lakes, which had a volume of 163 000 km3, drained in 1 year, sea level would have risen about 0.4 m in a single year, that is, a geologically instantaneous rise of sea level [Teller et al., 2002]. This momentary sea-level rise may have been of sufficient magnitude to abruptly backstep incised-valley-fill systems with low longitudinal gradients.

[15] The jump preceded the third meltwater pulse at 7600 cal BP, estimated from relic reef demise in the Caribbean–Atlantic region [Blanchon and Shaw, 1995], although some uncertainty in the dates cannot be excluded. The abrupt beginning of a deepening-upward shelf environment at ca. 8400 cal BP has been reported in the Yellow Sea [Liu et al., 2007a], and this event may be related to the jump. Rapid sea-level rise at 8400 14C y BP in New Zealand following a stillstand [Gibb, 1986] and at 8200 14C y BP in the central Great Barrier Reef following a sea-level fall [Larcombe et al., 1995] may also correlate with the jump. Moreover, the rate of sea-level rise estimated at Singapore is high between 9000 and 8500 cal BP, though radiocarbon data from this period are scarce relative to those from the following period [Bird et al., 2007]. Unfortunately but interestingly, age–depth plots of corals at Huon Peninsula (Figures 1 and 3) show a small interstice at approximately 9000 cal BP (∼20 m depth) because of a gap in the borehole core [Chappell and Polach, 1991]. Although the cause of the gap has not been described and discussed in the reference, we infer that coral communities gave up as the rapid sea-level jump occurred. More observational data that constrain the magnitude of sea-level changes during this period are thus required.