Estuaries and Coasts DOI 10.1007/s12237-016-0156-3

Mangrove Development and Its Response to Asian Monsoon in the Yingluo Bay (SW China) over the Last 2000 years Xianwei Meng 1,2 & Peng Xia 1 & Zhen Li 3 & Dezhen Meng 1

Received: 8 March 2016 / Revised: 22 August 2016 / Accepted: 23 August 2016 # Coastal and Estuarine Research Federation 2016

Abstract The characteristic location of mangroves at the interface between marine and terrestrial habitats makes them vulnerable to regional environmental change. The monsoonal climate of tropical Asia, where mangrove forests are broadly distributed along the coastal zone, differs from that of the other tropical regions. Therefore, mangrove development in tropical Asia is likely related to the evolution of the Asian monsoon and related environmental changes. As a case study, mangrove development was retrospectively studied using the contribution of mangrove-derived organic matter (CMOM), combined with the total pollen concentration of mangrove species, in bulk sediments of the Yingluo Bay, northern Beibu Gulf (SW China) over the last 2000 years. The results clearly show that mangrove development (flourishing or deteriorating) is primarily controlled by the changes in the air temperature and rainfall, which are controlled by the Asian monsoon. The three mangrove flourishing stages (~2.0– 1.6 cal. ka BP, ~1.4–0.8 cal. ka BP, and from ~140 years BP to present) result from warmer and wetter climates during the intensified Asian summer monsoon. In contrast, the two deteriorating stages (~1.6–1.4 cal. ka BP and from ~0.8 cal. ka BP Communicated by Alberto Vieira Borges * Xianwei Meng [email protected] * Peng Xia [email protected]

1

First Institute of Oceanography, State Oceanic Administration, Qingdao 266061, China

2

Function Laboratory for Marine Geology, National Oceanography Laboratory, Qingdao 266061, China

3

School of Earth and Ocean Sciences, University of Victoria, Victoria, British Columbia V8W, Canada

to 140 years BP) result from colder and dryer climates during the intensified Asian winter monsoon. Unexpectedly, the relative sea-level changes, regional seawater salinity, and tidal current, as well as human activity have not imposed notable effects on extension, flourishing/deterioration, and succession of the mangrove forest over the last 2000 years in the Yingluo Bay. Keywords Mangrove development . Asian monsoon . Mangrove-derived organic matter . Yingluo Bay . Last 2000 years

Introduction In addition to protecting coastal areas from waves and storms, mangrove ecosystems play an important role in the carbon, nitrogen, phosphorus, and sulfur cycles. Their characteristic location at the interface between marine and terrestrial habitats makes mangroves vulnerable to climatic change (Padmalal et al. 2011; Rogers et al. 2013). To predict the fate of mangrove ecosystems under future climatic changes, an appropriate strategy is to retrospectively study mangrove development using indicators preserved in sediments collected from the subtropical and tropical sites, where mangrove forests occupy or fringe the coast (Wooller et al. 2003, 2007; Versteegh et al. 2004; Monacci et al. 2009). Tropical or subtropical Asia differs from other tropical regions in relative sea-level changes, monsoonal climate, and stronger human activity; therefore, the effects of these agents on mangrove ecosystems are unique. The impact of changes in the Asian monsoon and relative sea level during the Holocene on the reef coral in the southwestern Leizhou Peninsula near the Yingluo Bay, northern Beibu Gulf (SW China) has been reported (Yu et al. 2004). The mangrove

Estuaries and Coasts

development in the Yingluo Bay and its link with change in air temperature over the last 150 years using contribution of mangrove-derived organic matter (CMOM) and the total pollen concentration of mangrove species in the upper sections of two core sediments (YLW02 and YLW03) drilled in the Yingluo Bay, with length of 49 and 45 cm, respectively, have been studied by Xia et al. (2015). In this study, we also make use of the CMOM but in the entire cored sediments of YLW02, with length of 150 cm, to further address the question of how mangrove forests respond to sea-level change, evolution of the Asian monsoon, and its induced change in the regional environments, including air temperature, rainfall, seawater salinity, dynamical condition, and human activity over the last 2000 years in the Yingluo Bay.

Materials and Analysis Methods Study Site and Sampling Yingluo Bay (21° 28′ N, 109° 43′ E), a core area in the Shankou Mangrove National Nature Reserve (SMNNR), is located in the coastal zone of northern Beibu Gulf, SW China (Fig. 1a), adjacent to the Leizhou Peninsula. The landform of the bay consists of mangrove flats, bare flats, and tidal creeks with sediments characterized by silty clay and sandysilty clay (Fig. 1b).

The climate is characterized as a tropical monsoonal climate with a pronounced maritime influence. The mean annual temperature is 22.4 °C with a maximum of 37.4 °C and a minimum of −0.8 °C. The average annual rainfall is 1815 mm, 80–85 % of which falls during the summer rainy season (April–September). The annual mean evaporation is 1800 mm with an average relative humidity of 81.8 %. The tide is irregularly diurnal with a mean and maximum tidal range of 2.53 and 6.75 m, respectively. Only affected by monsoonal rainfall, the salinity of the tidal water is relatively stable, ranging from 20 to 23 ‰ (Fan et al. 2005). The regional vegetation is typical of tropical forests and includes evergreen monsoon forests and semi-evergreen monsoon forests. However, much of these natural forests have been cleared for human settlements and agricultural cultivation or left to secondary vegetation, including tropical shrubs and grasses. Over 90 % of the 25.51 km coast along the Yingluo Bay is covered by mangrove vegetation (Li et al. 2008). The community distribution is characterized by the zonal transition of the mangrove community from dike (Excoecaria agallocha) to upper beach (Bruguiera gymnorrhiza and Rhizophora stylosa) to middle beach (Aegiceras corniculatum and Kandelia candel) to lower beach (Avicennia marina). In this study, a short sediment core (YLW02) with a length of 150 cm was collected from the middle beach of the Yingluo Bay, which is occupied by an A. corniculatum and K. candel community (Fig. 1). The core sediment was cut into sub-

21.56°

Guangxi Province (China)

N

21.54°

Shatian peninsula 21.52°

b

SNNC

21.50°

YLW02 Yingluo Bay

Beibu Gulf

21.48°

Sand flat 21.46°

a

Beibu Gulf

109.66° Mangrove forest

Mud flat

Fig. 1 A sketch map showing (a) locations of the Shankou Mangrove Natural Nature Reserve (SMNNR) and the Yingluo Bay and (b) the distributions of geomorphic types and mangrove forests, as well as

109.68° Sand flat

b 109.70°

109.72°

109.74°

Sampling site

109.76°

21.44° 109.78° E

SMNNR

sampling site of cored sediment in the Yingluo Bay and location of Station of National Natural Conservation (SNNC). Black arrows indicate the Beibu Gulf circulation

Estuaries and Coasts

samples at intervals of 2 cm, each of which were divided into four aliquots for grain size, organic carbon stable isotope (δ13Corg), total organic carbon (TOC), and total nitrogen (TN) analysis, as well as identification of the mangrove pollen for the upper section of the core, from 87 cm to the surface. Analytical Methods

relative difference in parts per thousand (‰) from the V-PDB standard. The repeatability of δ13C was better than 0.2 ‰. The TOC and TN concentrations were determined in the Key Laboratory of Marine Sedimentology & Environmental Geology, State Oceanic Administration, using a Vario EL-III Elemental Analyzer. Replicate analyses of one standard sample (GSD-9) provided a precision of ±0.02 wt% for TOC and ±0.005 wt% for TN.

Chronology Five peat samples at depths of 51, 61, 91, 99, and 149 cm were selected to measure Δ14C using accelerator mass spectrometry (AMS) at the Beta Analytic Radiocarbon Dating Laboratory. The measured ages were calculated using the 14 C half-life (5570 years) and referring to the modern standard, which was 95 % the 14C activity of the National Institute of Standards and Technology (NIST) Oxalic Acid (SRM 4990C). Quoted errors represent one relative standard deviation statistic (68 % probability) counting error based on the combined measurements of the sample background and modern reference standards. Measured 13C/12C ratios (δ13C) were calculated relative to the PDB-1 standard. The measured ages were corrected to the apparent radiocarbon ages for the isotopic fractions using δ13C. Except for the age at a depth of 51 cm, which is modern, the other four apparent radiocarbon ages were calibrated into calendar ages by intercepting them with calibration curves, which were supplied by the Beta Analytic Radiocarbon Dating Laboratory. The modern ages from the surface to 49 cm were determined by Xia et al. (2015) using the 210Pb method. Grain Size Analysis For the grain size analysis, aliquots of the sub-samples were treated with 10 % HCl and 15 % H2O2 to remove carbonate and organic matter. They were then analyzed using a Malvern 2000 laser-diffraction analyzer in the Key Laboratory of Marine Sedimentology & Environmental Geology, State Oceanic Administration.

Identification of Mangrove Pollen To validate the CMOM as a proxy of mangrove development, only the upper core section (from the surface to 87 cm) was used to identify the mangrove pollen. Aliquots of the section sub-samples were treated chemically according to a standard procedure (Li et al. 2008). Before treatment, three tablets with exotic Lycopodium (approximately 36,000 grains) spores were added to each sample to facilitate the determination of the pollen concentration. HCl (15 %) was used to dissolve the carbonates, and KOH (15 %) was employed to remove the humic components. The pollen and spores were then concentrated using heavy liquid flotation with CdI2 (specific gravity of 2.1) to separate them from undigested mineral detritus. Acetolysis was used to remove unwanted matter such as cellulose and humic debris. The residues were evenly mounted on glass slides with glycerin jelly. Pollen and spore types in each sample were counted under an optical microscope at ×400 magnification, except where difficulty in identification required a higher magnification. Pollen grains (200–300, excluding spores) were identified and counted in each sample. The proportion of pollen and spores was calculated on the basis of the total number of identified grains. The total pollen concentration of the mangrove species was then determined as the sum of each species (Li et al. 2008).

Results Age-Depth Model

Analysis of δ13Corg, TOC, and TN Aliquots of the sub-samples were air-dried and finely ground after the removal of shells and living biomass. They were then treated overnight with 1 M HCl at room temperature to remove the carbonates, followed by washing three times with distilled water and oven-drying at 40 °C (Wang et al. 2008). The organic carbon-stable isotope composition was determined at the Key Laboratory of Plant-Soil Interaction, Ministry of Education, China Agricultural University, using a Delta Plus XP mass spectrometer in the continuous flow mode. The 13C-to-12C ratios are reported in δ-notation as the

From the plot of excess 210Pb (210Pbex) versus compaction depth, the 210 Pb ex at a depth of 49 cm equals zero. Therefore, the ages of each sampling depth from the surface to 49 cm were calculated via a fitted exponential equation (y = 2.06e−0.16x), and therefore, the age at a depth of 51 cm was determined to be 140 years BP (Xia et al. 2015). Combined with four other calendar ages (Table 1), the sedimentation rates of each interval between the neighboring aged depths were calculated (Fig. 2), and thus, the ages at each sampling depth were determined via linear interpolation.

Estuaries and Coasts Results of AMS14C measured and calibrated ages

Table 1

Sample no.

Δ14C/‰

δ13C/‰

Depth/cm

1σ

Δ14C age/year BP

1σ

calibrated age/ cal. year BP

YLW02–26

51

−24.6

10.48

3.73

−85

30

Modern

YLW02–32 YLW02–46

61 91

−24.4 −22.5

−82.67 −127.13

3.44 3.04

705 1090

28 24

700 975

YLW02–50 YLW02–75

99 149

−23.4 −26.5

−131.57 −225.17

3.26 3.73

1360 2050

26 30

1290 1995

Vertical Distributions of Grain Size, TOC, TN, TOC/TN, and δ13Corg Based on the vertical distribution of the grain size, the core sediment of YLW02 can be divided into five lithological portions (Fig. 3a): the section from the bottom (150 cm) to 127 cm is characterized by the coarsest sediments, which have an average median grain size of 1.8 (φ); the section from 127 cm to 89 cm, where the sediments become finer with an average median grain size of 3.3 (φ); the section from 89 cm to 51 cm, where the sediments become coarser with an average median grain size of 3.0 (φ); the section from 51 to 30 cm, where the sediments are the finest with an average median grain size of 7.2 (φ); and the final section, from 30 cm to the surface, which is characterized by finer sediments, with an average median grain size of 4.9 (φ), than depth from 150 to 51 cm. Following the grain size distribution, the TOC and TN concentrations were lower in the coarser sediments than in the finer sediments. Moving through the core sediment section from the bottom to the surface, the average TOC concentrations were 0.33, 0.69, 0.49, 1.98, and 0.99 % (Fig. 3b), while Sedimentation rate (cm/yr) 0

0.1

0.2

0.3

0.4

0

30

Depth (cm)

140 cal. yr BP

60

700 cal yr BP

975 cal yr BP

90

1290 cal. yr BP

the average TN concentrations were 0.038, 0.036, 0.035, 0.131, and 0.057 % (Fig. 3c). The corresponding average calculated C/N (atom) ratios were 13.9, 22.9, 15.4, 17.6, and 20.1, respectively (Fig. 3d). The vertical distribution of δ13Corg was also divided into five portions; however, they are not in good agreement with those of the grain size. For the five sections, 150–127, 127– 103, 103–73, 73–30, and 30–0 cm (the surface), the average δ13Corg varied with values of −27.8, −26.7, −27.7, −27.1, and −28.3 ‰, respectively (Fig. 3e). Vertical Zonation of Mangrove Pollen Assemblages Just as in the current mangrove species in the Yingluo Bay, six mangrove pollen fossils were identified in sediments of the upper section (from 87 cm to the surface), including A. corniculatum, K. candel, R. stylosa, B. gymnorrhiza, A. marina, and E. agallocha, with a decreasing average concentration order of 14.5, 7.7, 5.2, 4.1, 1.3, and 0.3 %, respectively (Fig. 4). The cluster analysis based on mangrove pollen concentration automatically gives three zones of mangrove assemblages: 87–69, 69–43, and 43–0 cm (the surface). However, each zone had a pollen concentration order of the six mangrove species similar to that of the entire section, predominated by A. corniculatum and K. candel, with the other species pollen accounting for less than 10 %, especially A. marina and E. agallocha. In addition, variations in the concentrations of A. corniculatum, K. candel, R. stylosa, and B. gymnorrhiza from zone 1 to zone 2 to zone 3 follow the same pattern from low to the lowest to the highest values (Fig. 4). This implies that the zonation of mangrove pollen primarily represents a synchronous variation in the pollen concentration of each species rather than a change in the pollen assemblage. Quantitative Partition of Buried Organic Matters According to Their Sources

120

1995 cal. yr BP

150 0

500

1000

1500

2000

2500

Potential Sources of Organic Matter in the Core Sediments

Age (cal. yr BP)

Fig. 2 Plot of age versus depth showing one calculated 210Pb and four calibrated AMS 14C ages at depths of 51, 61, 91, 99, and 149 cm, respectively, as well as corresponding sedimentation rates

In general, the organic matter (OM) stored in sediments in the mangrove ecosystem originates from autochonous and allochthonous inputs (Saintilan et al. 2013). The autochonous input

Estuaries and Coasts TOC (%)

Sediment type Median grain size ( ) 0.2 0.4 0.6 0.8

1 0

0

2

a

10

4

6

TN (%)

8 0 0.5 1 1.5 2 2.5 0

b

C:N (atom)

0.06 0.12 0.18 0

c

13

Corg (%0 vs. PDB-1) CMOM (%)

10 20 30 40 -30

d

e (5)

-28

-26

-24 0 20 40 60 80 100 0

f

24

(5)

20

50

30 40

79 Silt

Clay

50

Depth (cm)

104

(4)

187

(4)

60

660

(3)

70

786

80

877

(3)

90

968

100 110

Age (cal. yr BP)

0

1304 Sand

(2)

1445

(2)

120

1587

130

1728

140

(1)

(1)

1869

150

1995 100 80 60 40 20 0

COMT (%)

Fig. 3 Variations in (a) median grain size, (b) organic carbon isotope composition, (c) total organic carbon concentration, (d) total nitrogen concentration, (e) calculated molar ratio of total organic carbon to total nitrogen, and (f) calculated contributions of mangrove-derived organic

matter (CMOM) and contribution of organic matter from terrestrial input. Left bracketed numbers represent stages of variation in sediment characteristics, and right bracketed numbers represent stages of variation in CMOM

includes mangrove production (leaf and root detritus) (Robertson and Alongi 2016) and microbes (bacteria) (Bouillon et al. 2004). The allochthonous input includes terrestrial matter and marine phytoplankton (Bouillon et al. 2003). However, the contribution of organic carbon (OC) from the bacteria accounts only for less than 1 % of the total organic carbon (TOC) in mangrove sediments (Bouillon et al. 2004) and, therefore, should be ignored for the purpose of OM source discrimination, especially for coastal mangrove ecosystems such as the Yingluo Bay, where the terrestrial refractory OM input is much greater than that of island and calcitecoastal mangrove ecosystems. Therefore, the OM source end members in the bulk sediments of the Yingluo Bay were reasonably regarded as terrestrial and mangrove-derived inputs and marine phytoplankton.

addition to leaf detritus (Duarte et al. 2005; Bouillon et al. 2008; Alongi 2011, 2014); however, it is more labile than leaf detritus and therefore has a higher potential for losses of carbon from the mangrove soil via leaching and saprophytic decay (Alongi 2011; Robertson and Alongi 2016). Even if there is less root detritus preserved in sediments, its δ13Corg value is only higher than that of leaf detritus by 1 %–2 ‰ (Badeck et al. 2005; Wei et al. 2008) and is still within the range of δ13Corg values of leaf detritus. Therefore, we reasonably selected senescent leaves to represent the mangrove-derived end member.

Selection of Substitutes for Potential Source End Members the Mangrove-Derived End Member The mangrovederived inputs in mangrove sediments, the leaf, and root detritus of different species usually have different δ13Corg and C/ N (atom) values. Therefore, the variability in the proportions of leaf and root detritus, as well as the mangrove species in core sediments, should be considered when determining the δ13Corg and C/N (atom) values of the mangrove end member. Recent studies have demonstrated that root detritus is also a major contributor to carbon stocks in mangrove soils in

The Terrestrial End Member For the Yingluo Bay, the potential terrigenous organic matter (TOM) input in the sediment should include fluvial, eolian, and eroded coastal matter. The fluvial matter is derived from the adjacent sea transported by tidal currents because no river directly enters the bay. The eolian and eroded coastal material can be ignored due to the tropical climate and the protection of mangrove forests from coastal erosion. The δ13Corg and C/N (atom) values of fluvial TOM primarily depend on the type of plants (C3 or C4) in the river basins, which are characterized by distinguishable δ13Corg and C/N values, and may be altered due to the seasonal effect on the proportion of C3 and C4 plants in fluvial suspended matter. For a tropical region, such as the SW China, C4 plants predominate in the all river basins due to its hot climate (Yin and Li 1997; Rao et al. 2010). In addition,

Estuaries and Coasts

Fig. 4 Variations in the total and respective pollen concentrations of mangrove species from 87 cm to the surface. The numbers (1–3) represent the zones of synchronous co-variation in pollen concentration of mangrove species rather than those of change in pollen assemblage (see text)

all rivers flowing into the northern Beibu Gulf are small ones, which are quite different from large rivers, and more than 80 % of the organic matter discharged into the sea occurs in the summer flooding period; therefore, the geochemical compositions, including the δ13Corg and C/N (atom) values of the fluvial end member, do not change considerably with the seasons and can be characterized by flooding plain sediments (Otteseh et al. 1989). The Phytoplankton End Member For an intertidal mangrove ecosystem, the phytoplankton (algae) is presumed to be another source of organic carbon; however, they are more easily used by bacteria and eaten by the benthic invertebrate community than mangrove litter in sediments (Bouillon et al. 2003). Consequently, the OC from marine phytoplankton preserved in sediments is usually much smaller than those of

mangrove-derived and terrestrial inputs for most coastal mangrove ecosystems (Bouillon et al. 2003), such as that studied in the Yingluo Bay (see below). Determination of the δ13Corg and C/N (Atom) Values of the Potential Source End Members The δ13Corg and C/N (atom) values of 16 senescent leaf samples from five pioneer mangrove species (A. corniculatum, K. candel, R. stylosa, B. gymnorrhiza, and A. marina) in the Yingluo Bay have typical δ13Corg and C/N (atom) characteristics of C3 plants (Rao et al. 1994; Gonneea et al. 2004), with average δ13Corg and C/N (atom) values of −28.8 ± 0.9 ‰ and 40.2 ± 11.2 (n = 16), respectively; however, a slight difference is seen among the five mangrove pioneer species (Xia et al. 2015). To test whether this difference is sufficient to affect the δ13Corg and C/

Estuaries and Coasts

Determination of OM True Sources of the Core Sediments As described above, we have shown that the three OM potential source end members have distinguishing δ13Corg and C/N (atom) values and can therefore be used to indicate the true provenances of organic matter in cored sediments collected from the Yingluo Bay using a plot of C/N (atom) versus δ13Corg. All core sediment samples were plotted on a diagram depicting the δ13C and C/N (atom) values of mangrovederived (mangrove leafs), terrestrial input (overbank sediments), and marine phytoplankton (marine algae). Note that, with the exception of three samples that were sub-sampled from the bottom layer (146–150 cm), the other samples are arrayed between the mangrove leaf and river input end members (Fig. 5). Therefore, they can be simply regarded as mixtures of two sources, mangrove-derived and terrestrial organic matter. Since the OM from marine phytoplankton is more easily decomposed by bacteria and eaten by benthic invertebrate community than mangrove-derived and terrigenous OM (Bouillon et al. 2003), the preserved OM from marine phytoplankton is too little to be discriminated in the cored sediments from the Yingluo Bay.

80 75 70 65 60 55 50

C:N (atom)

N (atom) values of the mangrove end member at different depths, the δ13Corg and C/N (atom) values of the upper section (from the surface to 87 cm) of the sediment core were calculated based on the pollen percentage concentrations and the leaf δ13Corg and C/N (atom) values of the five pioneer mangrove species. The calculated result shows that the average δ13Corg and C/N (atom) values in the upper section are −28.8 ± 0.3 ‰ and 39.5 ± 3.0 (n = 44), approximately equal to those of the mangrove senescent leaf samples, implying that the effect of different mangrove species on the δ13Corg and C/N (atom) values of the mangrove end member at different depths can be reasonably ignored and that the average δ13Corg and C/N (atom) values (−28.8 % and 40.2) of the senescent leaf samples can be regarded as those of the mangrove end member. The terrestrial end member represented by six flooding plain sediment samples from river basins in the northern Beibu Gulf (Guangxi province) and marine phytoplankton have average δ13Corg and C/N (atom) values of −24.3 ± 0.7 ‰ and 12.6 ± 2.0 and −16.1 ± 0.8 ‰ and 6.5 ± 0.1, respectively, and have been successfully used in estimations of OM contributions to bulk mangrove sediments on the Guangxi coast from different sources (Xia et al. 2015, Meng et al. 2016).

Mangrove leaves

45 40 35 30

Core sediment samples

25 20 15

Overbank sediments

10 5 0 -32

Marine algae -30

-28

-26

-24

-22

-20

-18

-16

-14

-12

13

Corg (%0 , VPDB)

Fig. 5 Plot of C/N (molar) vs. δ13Corg for discriminating sources of preserved organic matter in cored sediments showing potential end members of mangrove-derived (leafs), terrestrial (overbank sediments), and phytoplanktonic organic matter (marine algae) and indicating that samples of cored sediments were closely characterized by mixtures of mangrove-derived and terrigenous matter. Average end member values of C/N (atom) and δ 13 C of mangrove-derived, terrestrial, and phytoplanktonic organic matter are cited from Xia et al. (2015)

input (COMT) to the sediment can be calculated using the carbon isotopic mass balance equation (Hedges and Parker 1976; Goñi et al. 1997; Hu et al. 2006; Ramaswamy et al. 2008). Because CMOM and COMT were calculated using only the δ13Corg values measured for the bulk sediments, their vertical distributions must be in agreement with that of the δ13Corg values of the core sediments (Fig. 3e, f). Moving upwards from 143 to 127, 127–103, 103–73, and 73–30 cm to 30–0 cm (the surface), the average CMOM varies from 86, 62, 79, and 67 % to 94 %, respectively, while the average COMT varies from 14, 38, 21, and 34 % to 16 %, respectively (Fig. 3f). This indicates that the mangrove forest is the predominant source of organic matter for sediments in the Yingluo Bay.

Discussion

Quantitative Partitioning of Organic Matter in Sediments According to Their Sources

Validation of CMOM as a Proxy for Mangrove Development

The three samples that could not be explained in terms of the two sources, mangrove-derived and terrestrial organic matter, are related to other unknown sources and were therefore omitted. The relative CMOM and the contributions of terrestrial

The calculated CMOM in the total mixed-source organic matter of the bulk sediments not only depends on the mangrove forest growth status (flourishing or deteriorating) but is also restricted by the contribution of organic matter from other

Estuaries and Coasts

sources such as COMT. Therefore, CMOM should be validated as a proxy for mangrove development. In this study, CMOM was validated as a proxy for mangrove development using its correlation with the total pollen concentration of the mangrove species, which has been proven to be one of the most straightforward and effective proxies and is widely used to trace the historical development of mangrove forests (Versteegh et al. 2004; Wooller et al. 2007; Monacci et al. 2009), in particular for the mangrove forest in the Yingluo Bay (Li et al. 2008). For the upper section (87–0 cm) of the core sediments, the vertical distribution of CMOM is approximately parallel to that of the total pollen concentration of the mangrove species (Fig. 6a, b). For the sections of 87–70, 70–37, and 37–0 cm, the average CMOM values are 82, 67, and 90 %, respectively, corresponding to the variation in the average total mangrove pollen concentration of 23, 16, and 53 %, respectively. The correlation coefficient between CMOM and the total pollen concentration of the mangrove species was calculated to be 0.52 (Fig. 7), which is greater than the lower confidence limit at a significance level of 0.01. This indicates that there is a significant positive linear correlation between CMOM and the total pollen concentration of the mangrove species for the core; therefore, CMOM can be used to trace mangrove development to a certain extent.

Mangrove Development (Flourishing/Deteriorating) Over the Last 2000 years As described above, the mangrove pollen zonation of the upper section from 87 cm to the surface (from 940 years BP to the present) was due only to synchronous increase/decrease in the pollen concentrations of the mangrove species rather than the change in the pollen assemblage. This indicates that a mangrove succession did not occur at least in the last 940 years. Therefore, we can discuss the mangrove flourishing/ deterioration but its succession over the last 2000 years according to the temporal variation in CMOM combined with the variation in the total concentration of mangrove pollen over the last 940 years. The mangrove development in the Yingluo Bay over the last 2000 years is divided into five stages (Fig. 8a, b). The CMOM values were higher during the periods of ~2.0– 1.6 cal. ka BP, 1.4–0.8 cal. ka BP, and 140 years BP–present (Fig. 8a) with higher total concentrations of mangrove species (Fig. 8b), indicating that the mangrove forest was flourishing. Alternatively, the CMOM values were lower in the periods of 1.6–1.4 cal. ka BP and 0.8 cal. ka BP–140 years BP with lower total concentrations of mangrove species (Fig. 8b), indicating that the mangrove forest was deteriorating during these periods.

Influence of Environmental Changes on Mangrove Development The factors affecting mangrove development are usually divided into two categories. One is the so-called natural agents: climate and its associated regional environmental changes. The other is human activity. Influence of Climate Change The climatic factors affecting mangrove development include regional relative sea-level changes and changes in precipitation and temperature, which are controlled by Asian monsoon for tropical Asian, as well as the periodic El Niño/Southern Oscillation (ENSO) and casual storms. To recognize the effects of these climatic factors on mangrove development from historical records in sediments depends not only on the time scale but also on the time resolution of the sub-sampling interval. For the YLW02 core, the sub-sampling interval for the analysis of all parameters is 2 cm, spanning approximately 27 years and clearly overstepping the bounds of ENSO (2– 8 years) (Moy et al. 2002; Loubere et al. 2013) and storms (casual events); therefore, we cannot recognize interannual mangrove development and its link with ENSO and storm events. In addition, previous studies have suggested that ENSO is a modulator of the Asian summer monsoon (ASM) (Ju and Slingo 1995; Zhao et al. 2016), and therefore, its historical marker should be concealed in the interannual record of the ASM. Consequently, we only can link the mangrove development in the Yingluo Bay with the relative sea level rise, air temperature, and rainfall, which are controlled by Asian monsoon. Influence of Relative Sea Level Changes In general, relative sea-level changes can result in seaward or landward extension of forest habitats and the succession of floral compositions. However, these consequences are primarily controlled by the balance between changes in the relative sea level and the mangrove substrate, which primarily depends on the regional sedimentation rate (Ellison 2008, and references herein; Soares 2009). Previous studies on worldwide mangrove development over the entire Holocene have shown that the landward migration of mangrove habitats and the succession of mangrove communities only occur at times when the relative sea level rose quickly and the regional sedimentation rate did not keep pace with the increase during the transitions from the early to middle to late Holocene (Parkinson et al. 1994; Gilman et al. 2007, 2008; Li et al. 2012). While the relative sea level gradually rose at rate of 1–2 mm/year or was relatively stable over the late Holocene, most mangrove forests developed in situ (Urrego et al. 2013; Cohen et al. 2016). Different from previously studied mangrove coasts, where the sea level has continually risen since the Mid-Holocene, the sea level of the

Estuaries and Coasts

Total pollen (%)

70 60 50 40 30 20 10 0

(3)

(2)

923

877

832

786

723

660

424

187

146

104

92

79

65

50

37

24

0

12

Age (cal. yr BP)

Fig. 6 Comparison between the vertical distribution of (a) CMOM and (b) the total pollen concentration of the mangrove species (total pollen) for the section from 87 cm to the surface showing a perfect co-variation between CMOM and the total pollen. The upper bracketed numbers represent the stages of co-variation between CMOM and the total pollen

(1)

b

CMOM (%)

100 80 60 40

a

20 0

5

10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

Depth (cm)

Beibu Gulf, where the Yingluo Bay is located, has gradually descended at an average rate of 0.25 mm/year since ~6 ka BP (Fig. 8d) (Tanabe et al. 2003; Stattegger et al. 2013), suggesting that the sea level in the Ying Bay from 2000 year BP to 140 year BP has been relatively stable. Meanwhile, the average sedimentation rate in the Yingluo Bay during this period has been very slow (aver. ~0.6 mm/year, Fig. 8c). Therefore, on a millennial time scale, the mangrove forests in the Yingluo Bay would not have migrated seawards. The pollen assemblage of the upper section of the cored sediments from 87 cm to the surface does not indicate a succession in the mangrove community since 940 year BP (Fig. 4), further implying that the slight sea-level drop and the slow sediment accumulation have not resulted in an appreciable effect on mangrove migration and succession in the Yingluo Bay.

Since the global sea level rises at ~150 years BP, the relative sea level has risen at an average rate of 2.7 mm/year (Fig. 8d) (State of Oceanic Administration, China 2010), and the compaction-corrected sedimentation rate in the Yingluo Bay over the last 140 years has varied from 2.2 to 3.1 mm/year with an average value of 2.4 mm/year (Fig. 8c) (Xia et al. 2015), making it clear that the increase in the substrate height of the mangrove habitat in the Yingluo Bay has approximately kept pace with the relative sea level rise and that the mangrove forests have not been affected by the relative sea level rise over the last 140 years. A study of the landscape evolution of the Guangxi coastal mangrove based on Landsat images over the last 40 years showed no migration of the mangrove forest in the Yingluo Bay (Jia et al. 2015), further testifying that the relative sea level rise has had no effect on mangrove extension.

100

CMOM (%)

80 60 40

Y=0.56X+61.24 r=0.52 (r0.01=0.38, n-2=42)

20 0 0

10

20

30

40

50

60

70

Total mangrove pollen concentration (%) Fig. 7 Scatter diagram of CMOM vs. total pollen concentration of mangrove species for upper section of YLW02 core sediment showing a significant correlation between CMOM and total pollen concentration of mangrove species

Influences of Air Temperature and Rainfall Under the conditions of a relatively stable sea level, air temperature and rainfall are the primary climate components affecting mangrove forest growth (Ellison 2006; Gilman et al. 2008). For the Yingluo Bay, the rainfall is controlled by Asian summer monsoon, which can be indicated by oxygen isotopic composition of stalagmite (δ18Ostalag.) in SW China (Wang et al. 2005): the more negative δ18Ostalag value indicates intensified ASM, and vice versa. The parallelism of more/less negative δ18Ostalag.to warmer/colder periods (Fig. 8e) implies the intensified/weakened ASM couples with weakened/ intensified Asian winter monsoon (AWM), which is characterized by change in air temperature. Therefore, the influences of air temperature and rainfall on mangrove development in the Yingluo Bay can be attributed to those of AWM and ASM and can be discussed via the linkage of δ18Ostalag with stages

Estuaries and Coasts Mangrove pollen (%) SR (cm/yr)

0 100

Stage5

a

18O stalag.

Sea level (m)

(%0 )

Md ( )

20 40 60 80 0 0.1 0.2 0.3 0.4 0 0.2 0.4 0.6 0.8 -6.6 -7.0 -7.4 -7.8 -8.2

b

c

d

1.2 1.6 2.0 2.4 2.8 3.2 8 AWP

CMOM (%)

0 20 40 60 80 100 0

e

R = 2.7 mm/yr

f

6

4

2

0

g

200 400 500

LIA

300 R = -0.25 mm/yr

Stage4

700 800 900 1100

MWP

1000 Stage3

1200

1500

Stage2

1600 1700 1800

Stage1

1900

DACP

1400

HDWP

1300

Sea level at 140 yr BP

Age (Cal. yr BP)

600

2000 Weak

ASM

Strong

Fig. 8 Linkage of stages of mangrove development (flourishing/ deteriorating) indicated by (a) CMOM and (b) total pollen concentration of mangrove species with regional environment changes in (c) the sedimentation rate, (d) relative sea-level change, and (e) Asian summer monsoon indicated by δ18O value of the Dongge stalagmite in SW China (Wang et al. 2005) and warmer/colder stages (the Han Dynasty, Medieval, and Anthropocene Warm Periods, the Dark Ages Cold Period, and the Little Ice Age, abbreviated as HDWP, MWP, AWP, DACP, and

LIA, respectively) as well as with the hydrological conditions indicated by (f) sorting coefficient of sediments and (g) median value of grain size. The bold black curve in (a) is a three-point smoothed line. R = −0.25 mm/ year (Tanabe et al. 2003; Stattegger et al. 2013) and R = 2.7 mm/year (State of Oceanic Administration, China 2010) are rates of relative sea-level change (drop and rise) from 2000 to 140 years BP and from 140 years BP to the present, respectively

of mangrove development, which is indicated by CMOM and total concentration of mangrove pollen. Comparing the mangrove development (Fig. 8a, b) with 18 δ Ostalag (Fig. 8e) and the five periods of Chinese or northern hemispheric climate change, three flourishing stages (~2.0– 1.6 cal. ka BP, ~1.4–0.8 cal. ka BP, and 140 years BP–present) were observed to approximately correspond to the more negative δ18O values and thus a greater rainfall. They also exactly correspond to warmer climate periods (the Han Dynasty, Medieval, and Anthropocene warm periods). By contrast, the two deteriorating stages (~1.6–1.4 cal. ka BP and ~0.8 cal. ka BP–140 years BP) approximately correspond to the less negative δ18O values and, therefore, a lower rainfall. These stages exactly correspond to colder climate periods (the Dark Ages Cold Period and the Little Ice Age). This suggests that the warmer and wetter climate, induced by the intensified ASM, had facilitated flourishing of mangrove frests. In contrast, the colder and dryer climate, induced by intensified AWM, had resulted in deteriorating of mangrove forests over the last 2000 years BP in the Yingluo Bay.

therefore, primarily affects the structure and thus zonal distribution of mangrove communities. Because no river enters the Yingluo Bay, the seawater salinity is primarily modulated by rainfall. Therefore, the seawater salinity in the Yingluo Bay less changes than that in estuary, and mangrove forests were less influenced by seawater salinity. The fact that zones of mangrove pollen primarily represent the synchronous increase/decrease in the pollen concentration of each species rather than the change in the pollen assemblage over the last 940 years BP (Fig. 4) indicates that the salinity has not been a key factor for mangrove development in the Yingluo Bay.

Influence of the Regional Environments Influence of Seawater Salinity Seawater salinity is an optional climatic factor for individual mangrove species and,

Influence of Tide-Associated Dynamics As no river enters the Yingluo Bay, the allochthonous matter (primarily composed of terrestrial input) deposited here is transported by tidal floods from the adjacent sea and altered by the ebb tidal current. Therefore, the change in the dynamical sedimentary environment must have been related to this to-and-fro current, which can be approximately characterized by the pollen concentration of the mangrove species and the grain size of the sediments. Usually, if the individual pollen concentration of some mangrove species is much lower (<10 %), the pollen of these mangrove species is considered to have not been deposited in situ but rather transported from other sites and, therefore, can

Estuaries and Coasts

be used as an indicator of the hydrological conditions (Li et al. 2008, 2012). However, the use of this rule to deduce the hydrological conditions must take into account the flourishing/ deteriorating status of the mangrove forest. Because the coring site is occupied by A. corniculatum, this pollen is always a dominating component of all the mangrove species even if its concentration varied from ~10 % to less than 10 % to much higher than 10 %, while the pollen concentrations of the other species were below 10 %. Combined with the current spatial zonal distribution of E. agallocha, R . s t y l o s a , an d B . g ym n o r rhi z a, K . ca n de l , an d A. corniculatum and A. marina from dike to upper beach to middle beach and to lower beach, respectively, we have deduced that A. corniculatum pollen would be deposited in situ and the others would be allochthonous, transported by wind (E. agallocha) from the dike and by tidal currents from the upper beach (R. stylosa and B. gymnorrhiza), from the middle beach (K. candel), and from the lower beach (A. marina) respectively. However, this does not reflect the changes in the strength of the tidal current, but rather the flourishing/ deterioration of the mangrove forest because the concentrations of E. agallocha, R. stylosa, B. gymnorrhiza, and K. candel kept a synchronous increasing/decreasing pace with that of A. corniculatum. Along the entire cored sediment, the weaker sediment sorting, characterized by its larger sorting coefficient (>1) (Fig. 8f), indicates that the tidal current was relatively weak and that the regional hydrological conditions in the Yingluo Bay were therefore appropriate for mangrove development as a whole over the last 2000 years. However, the sorting coefficients and median values of the grain size at individual stages of mangrove development are still different from each other (Fig. 8g, f), implying that the hydrological conditions at the different stages were not unchangeable. However, the mangrove flourishing/deteriorating stage did not correspond to weaker/stronger hydrological conditions one to one. For example, stage 5 and stage 1 were both flourishing; however, they corresponded to stronger and weaker hydrological conditions, respectively. This implies that a slight change in the hydrological conditions might not be a determinative factor for mangrove development. Influence of Human Activity Since entering the Anthropocene (~150 years BP), human activity has been a necessarily considered agent affecting mangrove development (Abuodha and Kairo 2001; FAO 2007; Choudhury et al. 2014). This effect should be divided into two categories. One is a negative effect, the so-called anthropogenic disturbance, including industrial and agricultural production, pollution, and reclamation for construction of ports, croplands, and shrimp ponds along the coast (Ellison and Farnsworth 1996; Sakho et al. 2011; Maiti and Chowdhury

2013). The other is a positive effect, including protection and artificial-planting actions. The Yingluo Bay is located far from cities, large villages, and river basins, and thus, populationrelated negative disturbances are small (Fan 1995). In addition, there is no artificially planted mangrove forest (Jia et al. 2015) and the natural mangrove forest has only been protected since SMNNR was established in 1992; the positive effect is therefore also weak. In fact, the flourishing of the mangrove forest over the last 150 years can be firmly attributed to the warm climate (Xia et al. 2015).

Conclusions The organic matter in bulk sediments of the YLW02 core was quantitatively partitioned into mangrove-derived and terrestrial input using a diagram of organic carbon-stable isotopes versus molar TOC/TN ratios along with a mass balance equation of the carbon isotopes. The CMOM in the bulk sediments was validated as a proxy for mangrove development due to its significant correlation with the total pollen concentration of the mangrove species in the upper section from 87 cm to the surface. The temporal variations in CMOM over tha last 2000 years and the total pollen concentration of the mangrove species since 940 years BP indicated that the mangrove forest in the Yingluo Bay experienced three flourishing stages, ~2.0– 1.6 cal. ka BP, ~1.4–0.8 cal. ka BP, and 140 years BP–present, and two deteriorating stages, ~1.6–1.4 cal. ka BP and ~0.8– 140 years BP. Of the potential factors that affect mangrove development, relative sea level changes, regional seawater salinity, and tidal current, as well as human activity have not imposed notable effects on mangrove extension flourishing/ deterioration and succession. The air temperature and rainfall, controlled by Asian monsoon, have been the primary factors affecting the flourishing/deterioration of the mangrove forest over the past 2000 years in the Yingluo Bay. Acknowledgments The authors are grateful for the critical and constructive comments from the three reviewers. This work was supported by the National Key Basic Research Program of China (973 Program) on Global Climate Change (Grant No. 2010CB951203) and the National Natural Science Foundation of China (Grant Nos. 41376075, 41576061, 41576067, and 41206057).

References Abuodha, P.A.W., and J.G. Kairo. 2001. Human-induced stresses on mangrove swamps along the Kenyan coast. Hydrobiologia 458: 255–265. Alongi, D.M. 2011. The energetics of mangrove forests. Austral Ecology 36: 18–19. Alongi, D.M. 2014. Carbon cycling and storage in mangrove forests. Annual Review of Marine Science 6: 195–219.

Estuaries and Coasts Badeck, F.W., G. Tcherkz, S. Nogués, C. Piel, and J. Ghashghaie. 2005. Post-photosynthetic fractionation of stable carbon isotopes between plant organs—a widespread phenomena. Rapid Communications in Mass Spectrometry 19: 1381–1391. Bouillon, S., F. Dahdouh-Guebas, A.V.V.S. Rao, N. Koedam, and F. Dehairs. 2003. Sources of organic carbon in mangrove sediments: variability and possible ecological implications. Hydrobiologia 495: 33–39. Bouillon, S., T. Moens, N. Koedam, F. Dahdouh-Guebas, W. Baeyens, and F. Dehairs. 2004. Variability in the origin of carbon substrates for bacterial communities in mangrove sediments. FEMS Microbiology Ecology 49: 171–179. Bouillon, S., A.V. Borges, E. Castañeda-Moya, K. Diele, T. Dittmar, N.C. Duke, E. Kristensen, S.Y. Lee, C. Marchand, J.J. Middelburg, V.H. Rivera-Monroy, T.J. Smith III, and R.R. Twilley. 2008. Mangrove production and carbon sinks: a revision of global budget estimates. Global Biogeochemical Cycles 22: 1–12. Choudhury, A.K., M. Das, P. Philip, and P. Bhadury. 2014. An assessment of the implications of seasonal precipitation and anthropogenic influences on a mangrove ecosystem using phytoplankton as proxies. Estuaries and Coasts 38: 854–872. Cohen, M.C.L., R.J. Lara, E. Cuevas, E.M. Oliveras, and L.D.S. Sternberg. 2016. Effects of sea-level rise and climatic changes on mangroves from southwestern littoral of Puerto Rico during the middle and late Holocene. Catena 143: 187–200. Duarte, C.M., J.J. Middleburg, and N. Caraco. 2005. Major role in marine vegetation on the ocean carbon cycle. Biogeosciences 1: 173–180. Ellison, J.C. 2006. How South Pacific mangrove may response to predicted climate change and sea level rise. Advances in Global Change. Research 2: 289–300. Ellison, J.C. 2008. Long-term retrospection on mangrove development using sediment cores and pollen analysis: a review. Aquatic Botany 89: 93–104. Ellison, A.M., and E.J. Farnsworth. 1996. Anthropogenic disturbance of Caribbean mangrove ecosystems: past impacts, present trends, and future predictions. Biotropica 28: 549–565. Fan, H.Q. 1995. Mangrove resources, human disturbances and rehabilitation action in China. Chinese Biodiversity 3: 49–54. Fan, H.Q., G.H. Chen, B.Y. He, and Z.C. Mo. 2005. Coastal wetland and management of Shankou Mangroves, 35–89 . Beijing: China Ocean Press.in Chinese with English abstract FAO. 2007. The world’s mangrove 1980–2005. FAO Forestry Papers, Rome, p. 9. Gilman, E., J. Ellison, and R. Coleman. 2007. Assessment of mangrove response to projected relative sea-level rise and recent historical reconstruction of shoreline position. Environmental Monitoring and Assessment 124: 105–130. Gilman, E.L., J. Ellison, N.C. Duke, and C. Field. 2008. Threats to mangroves from climate change and adaption options: a review. Aquatic Botany 89: 237–250. Goñi, M.A., K.C. Ruttenberg, and T.I. Eglinton. 1997. Sources and contribution of terrigenous organic carbon to surface sediments in the Gulf of Mexico. Nature 389: 275–278. Gonneea, M.E., A. Paytan, and J.A. Herrera-Silveira. 2004. Tracing organic matter sources and carbon burial in mangrove sediments over the past 160 years. Estuarine, Coastal and Shelf Science 61: 211– 227. Hedges, J.I., and P.L. Parker. 1976. Land-derived organic matter in surface sediments from the Gulf of Mexico. Geochimica et Cosmochimica Acta 40: 1019–1029. Hu, J., P. Peng, G. Jia, B. Mai, and G. Zhang. 2006. Distribution and sources of organic carbon, nitrogen and their isotopes in sediments of the subtropical Pearl River estuary and adjacent shelf, southern China. Marine Chemistry 98: 274–285. Jia, M., Z. Wang, Y. Zhang, C. Ren, and K. Song. 2015. Landsat-based estimation of mangrove forest loss and restoration in Guangxi

province, China, influenced by human and natural factors. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 8: 311–323. Ju, J., and J. Slingo. 1995. The Asian summer monsoon and ENSO. Quarterly Journal of the Royal Meteorological Society 121: 1133– 1168. Li Z, Zhang Z, Li J‚ Li Z, Liu L, Fan H, Li G (2008) Pollen distribution in surface sediments of a mangrove system, Yingluo Bay, Guangxi, China. Review of Palaeobotany and Palynology 152:21–31. Li, Z., Y. Saito, L. Mao, T. Tamura, Z. Li, B. Song, Y. Zhang, A. Lu, S. Sieng, and J. Li. 2012. Mid-Holocene mangrove succession and its response to sea-level change in the upper Mekong River delta, Cambodia. Quaternary Research 78: 386–399. Loubere, P., W. Creamer, and J. Hass. 2013. Evolution of El NinoSouthern Oscillation in the late Holocene and isolation driven change in the tropical annual SST cycle. Global and Planetary Change 100: 129–144. Maiti, S.K., and A. Chowdhury. 2013. Effects of anthropogenic pollution on mangrove biodiversity: a review. Journal of Environment Protection 4: 1428–1434. Meng, X., P. Xia, Z. Li, and D. Meng. 2016. Mangrove degradation and response to anhtopogenic disturbance in the Mangwei Sea (SW China) since 1926 AD: mangrove-derived OM and pollen. Organic Geochemistry 98: 166–175. Monacci, N.M., U. Meier-Grünhagen, B.P. Finney, H. Behling, and M.J. Wooller. 2009. Mangrove ecosystem changes during the Holocene a t S p a n i s h L o o k o u t C a y, B e l i z e . P a l a e o g e o g r a p h y, Palaeoclimatology, Palaeoecology 280: 37–46. Moy, C.M., G.O. Seltzer, D.T. Rodbell, and D.M. Anderson. 2002. Variability of EI Niño/Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420: 162–165. Otteseh, R.T., J. Bogen, B. Bøliviken, and T. Volden. 1989. Overbank sediments: a representative sample medium for regional geochemical mapping. Journal of Geochemical Exploration 32: 257–277. Padmalal, D., K.P.N. Kumaran, K.M. Nair, R.B. Baijulal, B. Limaye, S. Vishnu, and L. Mohan. 2011. Evolution of the coastal wetland systems of SW India during the Holocene: evidence from marine and terrestrial archives of Kollam coast, Kerala. Quaternary International 237: 123–139. Parkinson, R.W., R.D. Delaune, and J.R. White. 1994. Holocene sealevel rise and the fate of mangrove forest within the wider Caribbean region. Journal of Coast. Research 10: 1077–1086. Ramaswamy, V., B. Gaye, P.V. Shirodkar, P.S. Rao, A.R. Chivas, D. Wheeler, and S. Thwin. 2008. Distribution and sources of organic carbon, nitrogen and their isotopic signatures in sediments from the Ayeyarwady (Irrawaddy) continental shelf, northern Andaman Sea. Marine Chemistry 111: 137–150. Rao, R.G., A.F. Woitchik, L. Goeyens, A.V. Riet, J. Kazungu, and F. Dehairs. 1994. Carbon, nitrogen contents and stable carbon isotope abundance in mangrove leaves from an East-Atlantic coastal lagoon (Kenya. Aquatic Botany 7: 175–183. Rao, Z., Z. Zhu, G. Jia, F. Chen, B. Loukas, J. Zhang, and M. Qiang. 2010. Relationship between climatic conditions and the relative abundance of modern C3 and C4 plants in three regions around North Pacific. Chinese Science Bulletin 55: 1931–1936. Robertson, A., and D.M. Alongi. 2016. Massive turnover rates of fine root detrital carbon in tropical Australian mangroves. Oceanologia 180: 841–851. Rogers, K.G., S.L. Goodbred, and D.R. Mondal. 2013. Monsoon sedimentation on the ‘abandoned’ tide-influenced Ganges–Brahmaputra delta plain. Estuarine, Coastal and Shelf Science 131: 297–309. Saintilan, N., K. Rogers, D. Mazumder, and C. Woodroffe. 2013. Allochthonous and autochthonous contributions to carbon accumulation and carbon store in southeastern Australian coastal wetlands. Estuarine, Coastal and Shelf. Research 128: 84–92.

Estuaries and Coasts Sakho, I., W. Mesnage, J. Deloffre, R. Lafite, I. Niang, and G. Faye. 2011. The influence of natural and anthropogenic factors on mangrove dynamics over 60 years: the Somone estuary, Senegal. Estuarine, Coastal and Shelf. Sciences 94: 93–101. Soares, M.L.G. 2009. A conceptual model for the response of mangrove forests to sea level rise. Journal of Coastal Research 56: 267–271. State of Oceanic Administration, China. 2010. www.soa.gov.cn. Stattegger, K., R. Tjallingii, Y. Saito, M. Michelli, N.T. Thanh, and A. Wetzel. 2013. Mid to late Holocene sea-level reconstruction of Southeast Vietnam using beachrock and beach-ridge deposits. Global and Planetary Change 110: 214–222. Tanabe, S., K. Hori, Y. Saito, S. Haruyama, V.P. Vu, and A. Kitamura. 2003. Song Hong (Red River) delta evolution related to millenniumscale Holocene sea-level changes. Quaternary Science Reviews 22: 2345–2361. Urrego, L.E., A. Correa-Metrio, C. Gonzάlez, A.R. Castaño, and T. Yokoyama. 2013. Contrasting responses of two Caribbean mangroves to sea-level rise in the Guajira Peninsula (Colombian Caribbean). Palaeogeography, Palaeoclimatolgy, Palaeoecology 370: 92–102. Versteegh, G.J.M., E. Schefuß, L. Dupont, F. Marret, J.S.S. Damsté, and J.H.F. Jansen. 2004. Taraxerol and Rhizophora pollen as proxies for tracking past mangrove ecosystems. Geochimica et Cosmochimica Acta 68: 411–422. Wang, Y., H. Cheng, R.L. Edwards, Y. He, X. Kong, Z. An, J. Wu, M.J. Kelly, C.A. Dykoski, and X. Li. 2005. The Holocene Asian monsoon: links to solar changes and North Atlantic climate. Science 308: 854–857.

Wang, G., X. Feng, J. Han, L. Zhou, W. Tan, and F. Su. 2008. Paleovegetation reconstruction using δ13C of soil organic matter. Biogeosciences 5: 1325–1337. Wei, L., C. Yan, G. Wu, X. Guo, and B. Ye. 2008. Variation of δ13C in Aegiceras corniculatum seedling induced by cadmium application. Ecotoxicology 17: 480–484. Wooller, M., B. Smallwood, U. Scharler, M. Jacobson, and M. Fogel. 2003. A taphonomic study of δ13C and δ15N values in Rhizophora mangle leaves for a multi-proxy approach to mangrove palaeoecology. Organic Geochemistry 34: 1259–1295. Wooller, M.J., R. Morgan, S. Fowell, H. Behling, and M. Fogel. 2007. A multi-proxy peat record of Holocene mangrove palaeoecology from twin cays, Belize. The Holocene 17: 1129–1139. Xia, P., X. Meng, Z. Li, A. Feng, P. Yin, and Y. Zhang. 2015. Mangrove development and its response to environmental change in Yingluo Bay (SW China) during the last 150 year: stable carbon isotopes and mangrove pollen. Organic Geochemistry 85: 32–41. Yin, L., and M. Li. 1997. A study on the geographic distribution and ecology of C4 plants in China:I. C4 plant distribution in China and their relation with regional climatic condition. Acta Ecologica Sinica 17: 350–363 .in Chinese with English abstract Yu, K.F., J.X. Zhao, T.S. Liu, G.J. Wei, P.X. Wang, and K.D. Collerson. 2004. High-frequency winter cooling and reef coral mortality during the Holocene climatic optimum. Earth and Planetary Science Letters 224: 143–155. Zhao, K., Y. Wang, R.L. Edwards, H. Cheng, D. Liu, X. Kong, and Y. Ning. 2016. Contribution of ENSO variability to east Asian summer m o n s o o n i n t h e l a t e H o l o c e n e . P a l a e o g e o g r a p h y, Palaeoclimatology, Palaeoecology 449: 510–519.