Coastal vegetation both mitigates the damage inflicted by marine disasters
on coastal areas and plays an important role in the global carbon cycle
(i.e., blue carbon). Nevertheless, detailed records of changes in coastal
vegetation composition and diversity in the Holocene, coupled with climate
change and river evolution, remain unclear. To explore vegetation dynamics
and their influencing factors on the coastal area of the Bohai Sea (BS)
during the Holocene, we present high-resolution pollen and sediment grain
size data obtained from a sediment core of the BS. The results reveal that
two rapid and abrupt changes in salt marsh vegetation are linked with the
river system changes. Within each event, a recurring pattern – starting with a decline in Cyperaceae, followed by an increase in
Coastal areas, where cities, populations and industries are clustered, are playing an increasingly critical role in trade globalization (Hemavathi et al., 2019). Because they are located between marine ecosystems and terrestrial ecosystems, coastal areas are prone to many natural hazards, such as flooding, storms and tsunamis (Hou and Hou, 2020). Coastal vegetation, which acts as a natural barrier, is widely distributed in coastal areas and could effectively mitigate the damage caused by marine disasters to the economy and environment of coastal areas (Zhang et al., 2018). Moreover,
despite their relatively small global extent (between 0.5 and
Climatic fluctuation, post-glacial sea level rise and changes in river
discharge provoked dramatic habitat changes along coastal areas during the
Late Pleistocene and Holocene (Neumann et al., 2010; Cohen et al., 2012;
Pessenda et al., 2012; França et al., 2015). Presently, the relationship
of sea level change and coastal vegetation (especially mangrove) evolution
has been widely studied by many researchers (e.g., Engelhart et al., 2007;
González and Dupont, 2009; França et al., 2012; Woodroffe et al., 2015; Hendy et al., 2016). In contrast, studies on the
long-term dynamics of coastal vegetation, coupled with climate change and
river evolution, are sparse. During the Holocene, the global rivers
delivered large amounts of material to the ocean, the total suspended
sediment delivered by all rivers to the ocean was approximately
In the coastal areas of the Bohai Sea (BS), vegetation is dominated by warm temperate deciduous broadleaved forests and shrub grasslands (Wang et al., 1993). The Yellow River (YR), as one of the largest rivers in the world in terms of sediment discharge (Milliman and Meade, 1983), transports large amounts of sediment into the BS every year; hence, it has developed a delta complex in the west coastal region of the BS since 7000 BP (He et al., 2019). Deposition of the Yellow River delta (YRD) complex resulted in the formation of a vast area of floodplain and estuarine wetland (Xue et al., 1995; Cui et al., 2009; W. Z. Liu et al., 2009). Based on the study of coastal vegetation of the BS, it is helpful to understand the spatial and temporal drivers of ecological variability, and thus of the vegetation–climate and vegetation–river relationships, especially wetland dynamics. However, there have been few studies investigating the vegetation dynamics and their response to climate and river variables in the Bohai region.
Pollen records have been useful in terms of reconstructing vegetation
dynamics and environmental changes associated with climatic changes in the
geological record (Bao et al., 2007; Cohen et al., 2008; Giraldo-Giraldo et
al., 2018). In this study, we carried out a detailed investigation of core
sediments from Laizhou Bay in the BS. We analyzed pollen and grain size proxies
at a high resolution and refined the chronology of the core by using
The BS, a shallow inland sea in China, is connected with the Yellow Sea
via the narrow Bohai Strait (Fig. 1). The main rivers flowing into the
BS are the YR, Haihe River, Luanhe River and Liaohe River. Among these, the
YR is the largest and is the main source of sediments in this region. Over
the past 2000 years, the YR has annually provided approximately
Geographic map of the Bohai Sea, with locations of the core CJ06-435 site (red circle) and other sites referred to in this study (purple circles). Core references are as follows: H9601 and H9602 (Saito et al., 2000), ZK1 (Li et al., 2013), ZK228 (Xue et al.,1988), HB-1 (J. Liu et al., 2009), and GYDY (Liu et al., 2014).
The tidal current plays a critical role in the transportation and
distribution of sediments in the BS. The tidal currents of the modern BS are
dominated by semi-diurnal tides. The velocity of tidal currents varies from
20 to 80 cm s
The wind waves off the YRD are dominated by the East Asian monsoon and show significant seasonal variations. The prevailing northerly winds are much stronger in winter than the dominant southerly winds in summer. Strong winter winds cause strong wind waves and thus strong bottom shear stresses that readily erode seabed sediment into water (Yang et al., 2011; Wang et al., 2014; Zhou et al., 2017).
The circulation of the BS is weak, and the mean flow velocity is small. In winter, the predominant extension of the Yellow Sea Warm Current (YSWC) intrudes and crosses the Bohai Strait, moving westward along the central part of the BS, and splits into two branches. One branch moves toward the northeast to form a clockwise gyre (Liaoxi Coastal Current, LXCC), and the other veers southward and then turns eastward along the southern coast to form a counterclockwise gyre (Lubei Coastal Current, LBCC). In summer the YSWC disappears in the BS, and eddies generated in the BS are stronger than in winter. During this time, the central eddy is missing, the eddy in Laizhou Bay is more pronounced, and the coastal current along the southern and western coastlines of the BS is established (Fig. 2; Liu et al., 2015; Yang et al., 2016).
Vegetation map around the Bohai Sea and the ocean current in the Bohai Sea during the summer
The Bohai region lies in a zone of warm temperate monsoonal climate with
distinctive seasons. The annual mean air temperature is 9.5–13.1
The regional vegetation is dominated by warm temperature deciduous
broadleaved forests and shrub grasslands. Currently, natural vegetation only
remains in the mountain areas because of widespread anthropogenic activities
(e.g., cultivation and farming). The predominant deciduous broadleaved
species belong to
Core CJ06-435 was collected in Laizhou Bay in August 2007 by the R/V
Isotopes
AMS radiocarbon dates from core CJ06-435 and one tie point corresponding to
the deepest onset of
A total of 127 samples were selected for pollen analyses. Each sample was
oven-dried at 60
The percentage of each pollen type was calculated from the total sum of pollen and spores. The pollen diagram was produced using Tilia software, and the pollen assemblage zones were divided based on the results of a constrained cluster analysis (CONISS) within Tilia (Grimm, 1987).
Analysis of sediment grain size was performed at 2.0 cm intervals throughout
the core using a Malvern Mastersizer 2000 instrument at the laboratory of
the First Institute of Oceanography. The chemical procedure of grain size
experimental pretreatment was consistent with the procedures described by
Chen et al. (2019a). A solution of 30 %
Measurements of
Lithology, grain size, color reflectance
Often, the combined data of
The results of AMS radiocarbon dating are shown in Table 1 and Fig. 3. Three samples above the 20 cm depth were not included in the age model because their
Percentage diagram of the principal pollen taxa from core CJ06-435. Pollen zonation is based on CONISS results.
The grain size parameters and component percentages show distinct variations. The mean grain size and the median grain size both show high values at depths of 271–160, 135–83, and 19–0 cm; lower values at a depth of 83–34 cm; and the lowest values at depths of 160–135 cm and 34–19 cm. There was a smaller proportion of clay in the lower profile (271–160 cm) than in the upper profile (160–0 cm, except for the two sections of 160–135 and 34–19 cm). The sequences of silt and sand contents showed a strong inverse association. There were high proportions of silt and low proportions of sand at depths of 271–222, 180–160, 135–83, 40–34, and 19–0 cm; lower proportions of silt and higher proportions of sand at depths of 222–180 and 83–40 cm; and the lowest proportions of silt and the highest proportions of sand occurred at depths of 160–135 and 34–19 cm (Fig. 3).
A total of 71 pollen taxa were identified, among which
Concentration diagram of the principal pollen taxa from core CJ06-435.
The palynological zone 1 was characterized by abundant broadleaved trees
pollen, dominated by
Palynological zone 2 was divided into four subzones.
In contrast, subzone 2d (41–30 cm) was marked by a sudden decrease in the
pollen of
This zone was characterized by the transition from dominance by the pollen
of arboreal taxa to non-arboreal types. The percentage of
In sediment core CJ06–435, both
Spatial distribution of modern pollen percentage (solid black circle, %) and concentration (open red circle, grains per gram) in Laizhou Bay, Bohai Sea (modified from Yang et al., 2016).
Previous studies have revealed that
Herb pollen, especially Chenopodiaceae, also occupies an important position
in core CJ06-435 (Fig. 4). The spatial distribution of herb pollen in
surface sediment of Laizhou Bay suggests that a higher percentage and
concentration occur in the nearshore area close to the YR estuary and the
southwestern part of Laizhou Bay and that a low percentage and concentration are
found in the eastern part of Laizhou Bay (Fig. 6c and d). The YR is the
main sediment source of the BS. The annual mean sediment load of the YR was
It is worth noting that the composition of fossil pollen in sediment depends
not only on the composition of the vegetation from which the pollen
originates but also on pollen dispersion, deposition and preservation.
In addition, because different pollen types are not equally well preserved
(Havinga, 1967; Cheddadi and Rossignol-Strick, 1995), bias originating from
poor preservation should be eliminated before using the net content of
pollen grains to reconstruct paleovegetation. In this study, the pollen
concentration ranged from 62 to 6050 grains g
Previous research suggests that the sedimentation mechanisms of pollen and spores in marine water is similar to that of sediment with clay and fine silt grain size (Heusser, 1988). A recent investigation into the surface sediment from the BS shows high pollen concentration in sediments with a high proportion of fine particles such as clay and silty clay, while showing low pollen concentration in sediments with a high proportion of coarser sand particles (Yang et al., 2019). Yang et al. (2019) attributed the low pollen concentration in areas with a high sand content of the BS to the strong hydrodynamic suspension and screening for sediments and pollen. We conclude that the low pollen concentrations in the two sections (160–135 and 34–19 cm), correlated with high sand content, could be attributed to the hydrodynamic conditions rather than degradation.
The most important geological events on the northern Chinese coast after 7000 BP were the shift of YR channel and the formation of the YRD. The YR was easily plugged and breached, and therefore its lower reaches migrated because of its huge sediment load. The shifting of the lower reaches of the YR led to the formation of a new delta superlobe (He et al., 2019). Based on a study of cheniers and historical documents, nine YRD superlobes have been proposed by Xue and Cheng (1989) and Xue (1993) on the western shore of the BS. Among these, superlobe 1 (7000–5000 BP, He et al., 2019), superlobe 7 (11–1048 CE, Xue, 1993), and superlobe 10 (since 1855 CE) are positioned near the core area in this study. Information about some of these superlobe formations is recorded in core CJ06-435.
As shown in Fig. 8, herb percentage suddenly changes at 160 and 34 cm. Herb pollen in the sediment of Laizhou Bay is mainly derived from the coastal wetlands of the western BS. At 6000–7000 BP and in 1855 CE, the YR emptied into the BS after a natural course shift, forming two huge delta superlobes in the western part of the BS (Saito et al., 2000). Wetland plants are the most important vegetation type in the YRD (Jiang et al., 2013). The development of YRD wetland would change the amounts of herb pollen that was transported to the study site. In addition, the formation of the YRD caused the coastline to move closer to the position of the CJ06-435 core. Since most herb plants are small in size, their pollen grains are unable to disperse broadly (Chen et al., 2019b). The migration of the coastline would change the availability of herb pollen to the study site, and hence lead to variations in the amount of pollen. Therefore, combined with the age data, we conclude that the abrupt change of herb pollen percentage at 160 and 34 cm in core CJ06-435 is related to the formation of the YRD superlobe 1 and superlobe 10.
Compared with the pollen percentage, the pollen concentration can be
interpreted in different ways. Namely, the percentage of different types of
pollen is relative, whereas the pollen concentration is absolute, and it can
directly reflect the amounts of pollen that were transported to the study
area (Luo et al., 2013). It is crucial that a correct interpretation of
pollen data is based on a percentage diagram as well as concentration. In
core CJ06-435, the concentrations of herbs – especially Chenopodiaceae and
Sediment grain size provides direct information about changes of the sediment
source and the sedimentary environment (Friedman and Sanders, 1978; Wu et
al., 2015). The characteristics of grain size can be expressed by the grain
size distribution curve, and usually the mean or median diameter is used
(Xu, 1999). In this study, the value of mean grain size (Mz) showed that two
major grain size boundaries occur at depths of 34 and 19 cm, separating a
middle sedimentary unit (34–19 cm) that contains coarser sediment from the
lower and upper sedimentary units that contain finer sediment (Fig. 7d). The sand content of the upper, middle, and lower layers was 11.2 %, 33.6 %, and 9 %, respectively (Fig. 7a); the silt content of these
layers was 69 %, 58.6 %, and 76.1 %, respectively (Fig. 7b); and the clay content of these layers was 19.8 %, 7.8 %, and 14.8 %,
respectively (Fig. 7c). On the basis of
The sediment of Laizhou Bay mainly comes from the YR and other small rivers located in the southern part of Laizhou Bay (Zhang et al., 2017; Gao et al., 2018). Prior to 1855 CE, when the YR entered the Yellow Sea, the sediment contribution to the BS from other small rivers was relatively large. The fine fraction suspension sediment that was derived from other small rivers favors the hypothesis of fine sediment accumulation in core CJ06-435 during this period.
When the YR reentered the BS after 1855 CE, the dispersal of YR material contributed substantially to the sedimentation of the BS. It was reported that more than 80 % of the YR sediment discharges into the BS during the summer period (Bi et al., 2011). Owing to the barrier effect of the tidal shear front and the weak river flow, most of the river-delivered sediment is deposited on the offshore delta within 15 km of the river mouth (Wang et al., 2007; Bi et al., 2010). Only a small part of the fine clay fraction is transported by the coastal currents over long distances and deposited across or along the shore in summer (Wu et al., 2015). During the winter (October to March) season, the much stronger winter monsoon generates large waves, resulting in intensive sediment resuspension in the coastal region owing to the enhanced bottom shear stress (Yang et al., 2011; Bi et al., 2011). The resuspended sediment is transported southeastward along the coast of Laizhou Bay by the monsoon-enhanced coastal currents passing through the location of the sediment core CJ06-435. Therefore, after 1855 CE, the sediment of core CJ06-435 mainly included the fine fraction of the suspended sediment dispersed from the YR mouth, the resuspended sediment from the coastal area off the YR delta in the winter, and the locally resuspended sediment.
The accumulation of YR-suspended sediment during the summer season in Laizhou Bay was closely associated with the sediment dispersion pattern off the active delta lobe (Xing et al., 2016). The estuary of the YR, during most of the period 1855–1976 CE, was north of the modern YRD, and suspended sediment from the YR was transported northeastward to Bohai Bay and the central Bohai basin. The contribution of YR-suspended sediment to the sedimentation of core CJ06-435 was smaller, and the resuspended sediment became a dominant material source. During the winter, the large waves generated by the strong winds may result in intensive resuspension of the seabed sediment and lead to part of the coarse sediment in the YR mouth and Laizhou Bay being transported to the study area, which induces an evident increase in mean grain size and a decrease in the amount of fine sediment. After 1976 CE, the lower channel of the YR shifted to the Qingshuigou course in Laizhou Bay; the suspended sediments derived from the YR estuary were primarily driven southward and southeastward along the coast, leading an increasing transportation of most of the YR-suspended sediment into Laizhou Bay (Qiao et al., 2010). As a result, the dispersal of river-laden sediment contributed substantially to the sedimentation of core CJ06-435, with a fine sediment layer being formed in the upper part of the core.
The results inferred from our grain size data on the migrations of the YR lower channel since 1855 CE and their effects on the sedimentary environments of the adjacent BS are in accordance with the results of other studies from Laizhou Bay (Wu et al., 2015) and central BS (Hu et al., 2011). Based on the records of sediment core collected from Laizhou Bay, Wu et al. (2015) found that when the YR mouth approached the core location, the sediment became finer; otherwise, the active resuspension resulted in the accumulation of coarser sediment owing to strong hydrodynamics. The grain size results from the central mud areas of the BS also point to the conclusion that the sediment supply from the YR to the central BS was cut off because of the shift of the YR terminal course from the Diaokou source in outer Bohai Bay to the Qingshuigou course in Laizhou Bay in 1976, resulting in a significant increase in the proportion of sand in sediment of the central BS (Hu et al., 2011).
It is worth noting that the variation of grain size characteristics in the period of 6570–5000 BP is very similar to that after 1855 CE. As shown in Fig. 7d, the shift of Mz in the period of 6570–5000 BP also began with a significantly increased of Mz at 6570 BP (160 cm) when the YR flowed into the BS in northern Shandong province. This similar variation of grain size in the period of 6570–5000 BP (superlobe 1) and after 1855 CE (superlobe 10) implies that a similar YR channel shifting occurred during these two periods. However, further research is needed to reveal how the deltaic and neritic sea sedimentary environment was impacted by the river system.
Two high-amplitude salt marsh vegetation shifts are displayed in the herb
pollen record during 6570–5000 BP (superlobe 1) and after 1855 CE
(superlobe 10), indicating rapid oscillations of environmental conditions in
the coastal area of BS. Within single intervals of the YR superlobe, a
recurrent and directional alternation of herb pollen taxa is observed in the
following order: the shift of herb pollen data began with an abrupt decrease
of Cyperaceae pollen, followed by a steep increase of Chenopodiaceae and
Cyperaceae, Chenopodiaceae, and
In the studied sequence, the YR flowed into the BS toward northern Shandong Province after a course shift. The lower river was initially braided upon relocation, as characterized by unchannelized river flow. At this initial stage, the river-derived sediment was largely accumulated in the floodplain and/or among the antecedent rivers owing to the lack of channelization (Wu et al., 2017), filling the coast of the lake, the interfluvial lowlands of the paleo-river, and the supralittoral zone, etc. This caused the destruction of hydrophytes and phreatophyte sedges in these areas. This process is indicated in our records by the significant decrease in the amount of Cyperaceae pollen percentages for superlobe 1 and superlobe 10 (Fig. 8b).
Eventually, natural channel adjustments resulted in the coalescence of
multiple channels into a single channel (Wu et al., 2017). A large amount of
river-derived sediment was deposited at the mouth of the YR, causing the
progradation of the YRD. Because of the strong influence of the tides, the
intertidal zone in the YRD was originally bare beach. Along with the seaward
expansion of the newly formed beach, the influence of tides was weakening on
the original bare beach wetland and the salinity of the original beach
wetland began to decrease (Zhang et al., 2007). Pioneer species of salt
marshes, e.g.,
Based on the age model, the record between 3000 BP and 1855 CE of core
CJ06-435 is somewhat confused because the CSR was extremely low (about
0.005 cm yr
During the period from 8500 to 6500 BP (palynological zone 1, 271–156 cm), the palynofloral assemblages are mainly composed of the pollen of broadleaved trees, such as
During the period from 6500–5900 BP (palynological zone 2a, 156–128 cm),
a climatic cooling period is identified by an increase of conifers (
During the period from 5900–3500 BP (palynological zone 2b, 128–63 cm),
climate cooling and drying is observed by a reduction of broadleaved trees
such as
Between 3500 and 3000 BP (63–56 cm), a warm climatic phase occurred, as
suggested by an increase in the amount of pollen from broadleaved trees
(
From 3000 BP to 1855 CE (56–30 cm), as mentioned in the first paragraph
of this section, the CSR was extremely low during the period from 3000 BP
to 1855 CE (56–30 cm). We suggest that there might be some deposition
hiatus and only tentatively discuss this section of the pollen record. This
section begins with a relatively high percentage of broadleaved trees
(
After 1855 CE (palynological zone 3, 30–0 cm), a significant decline in
broadleaved tree (
Many previous studies of northern and northeastern China have used the
Comparison of relevant Holocene temperature records with solar irradiance and El Niño–Southern Oscillation (ENSO) proxy records derived from the equatorial Pacific.
The pollen-based record of temperature evolution from core CJ06-435 is
broadly in phase with published high-resolution sea surface temperature
record from core YS01 (Fig. 9e) of the Yellow Sea, suggesting at least a
local pattern of temperature variations during the Holocene. To investigate
whether the temperature pattern was a local characteristic of the BS and
Yellow Sea area or whether it was rather a regional pattern of East Asia
as a whole, the pollen records of core CJ06-435 were compared with
recently-published and relatively well-dated sequences from northern China,
northwestern China, and the Tibetan Plateau (see Fig. 9a for site locations),
including the sedimentary pollen-based temperature record from the Narenxia
Peat within the Kanas Lake, northwestern China (Fig. 9f; Feng et al., 2017); the
Insolation has been widely accepted as an important factor in Holocene
climate variation. The covariation of Northern Hemisphere extratropical
(30
We compared the pollen-based temperature record of core CJ06-435 (Fig. 9c
and d) with the frequency of El Niño events reconstructed from the
Botryococcene concentration in the El Junco Lake sediment (Fig. 9i; Zhang
et al., 2014) and the ENSO variability reconstructed from
In addition, radiative forcing by greenhouse gases (GHGs) rose 0.5 W m
In summary, the temperature characterizations of a cool Middle Holocene and a relatively warm Late Holocene revealed by the East Asian records could be linked with the change of insolation, ENSO activity and GHG forcing. The cooler Middle Holocene may be related to a combination of the decreasing summer insolation, weak El Niño activity and relatively low GHG radiative forcing during this interval. Along with strengthened ENSO activity and enhanced GHG forcing in the late Holocene, there was increased temperature.
Through the palynological and grain size reconstruction of coastal area vegetation and environment in core CJ06-435, we were able to identify specific responses of plant communities to climatic (temperature, precipitation), hydrological and anthropogenic impacts. Our data elucidate the pattern and mechanisms driving coastal salt marsh succession at decade-to-century timescales. Two intervals of expanded salt marsh vegetation correspond to the formation of YR delta superlobes, indicating that soil development and salinity gradients are the main factors determining the vegetation dynamics of coastal wetland. Our pollen-based temperature index revealed a warm early Holocene (8500–6500 BP), a subsequent cool stage between 6500 and 3500 BP, and a slightly warming episode after 3500 BP. The reliability of the record, especially the cooler Middle Holocene, is further supported by several other temperature records from East Asia. We suggest that changes in insolation, ENSO activity and GHG forcing could have played an important role in the temperature evolution in East Asia.
The co-authors declare that all data included in this study are available upon request by contact with the corresponding author (email: jinxiachen@fio.org.cn).
CJ wrote the manuscript. SX and LY revised the manuscript. QS provided many constructive suggestions for the manuscript. YangS provided the pollen data of surface sediment. LJ use pollen data of core CJ06-435 as base for quantitative climate reconstruction. YanS, LH, LX and LC provided financial support for the collection of samples and obtained samples.
The authors declare that they have no conflict of interest.
We thank the crew of the R/V
This research has been supported by the National Natural Science Foundation of China (grant nos. 41420104005, U1606401, and 41576054), the National Program on Global Change and Air–Sea Interaction (grant no. GASI-GEOGE-03), and the Taishan Scholar Program of Shandong.
This paper was edited by Julie Loisel and reviewed by two anonymous referees.