Content uploaded by Dongling Li
Author content
All content in this area was uploaded by Dongling Li on Oct 28, 2024
Content may be subject to copyright.
Diatoms as indicators of environmental change in coastal areas: a
case study in Lianjiang coast of East China Sea
Tong Li, Jihui Zhang, Dongling Li, Chengxu Zhou, Chenxi Liu, Hao Xu, Bing Song, and Longbin Sha
View online: https://doi.org/10.1007/s13131-024-2292-0
Articles you may be interested in
Characteristics and mechanisms of the intraseasonal variability of sea surface salinity in the southeastern Arabian Sea during
2015–2020
Acta Oceanologica Sinica. 2023, 42(5), 25-34 https://doi.org/10.1007/s13131-022-2074-5
Validation and application of soil moisture active passive sea surface salinity observation over the Changjiang River Estuary
Acta Oceanologica Sinica. 2020, 39(4), 1-8 https://doi.org/10.1007/s13131-020-1542-z
Long-term nutrient variation trends and their potential impact on phytoplankton in the southern Yellow Sea, China
Acta Oceanologica Sinica. 2022, 41(6), 54-67 https://doi.org/10.1007/s13131-022-2031-3
Synthesizing high-resolution satellite salinity data based on multi-fractal fusion
Acta Oceanologica Sinica. 2024, 43(7), 112-124 https://doi.org/10.1007/s13131-023-2209-3
Spatial-temporal dynamics of biogenic silica in the southern Yellow Sea
Acta Oceanologica Sinica. 2019, 38(12), 101-110 https://doi.org/10.1007/s13131-019-1516-1
Coupling virio- and bacterioplankton populations with environmental variable changes in the Bohai Sea
Acta Oceanologica Sinica. 2020, 39(6), 72-83 https://doi.org/10.1007/s13131-020-1591-3
关注微信公众号，获得更多资讯信息
Diatoms as indicators of environmental change in coastal areas:
a case study in Lianjiang coast of East China Sea
Tong Li1, 2, Jihui Zhang2, Dongling Li1, 2*, Chengxu Zhou3, Chenxi Liu1, 2, Hao Xu1, 2, Bing Song4,
Longbin Sha1, 2
1 Donghai Academy, Ningbo University, Ningbo 315211, China
2 Department of Geography and Spatial Information Techniques, Ningbo University, Ningbo 315211, China
3 College of Food Science and Engineering, Ningbo University, Ningbo 315211, China
4 State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese
Academy of Sciences, Nanjing 210008, China
Received 26 September 2023; accepted 24 December 2023
©Chinese Society for Oceanography and Springer-Verlag GmbH Germany, part of Springer Nature 2024
Abstract
Owing to the significant differences in environmental characteristics and explanatory factors among estuarine
and coastal regions, research on diatom transfer functions and database establishment remains incomplete. This
study analysed diatoms in surface sediment samples and a sediment core from the Lianjiang coast of the East
China Sea, together with environmental variables. Principal component analysis of the environmental variables
showed that sea surface salinity (SSS) and sea surface temperature were the most important factors controlling
hydrological conditions in the Lianjiang coastal area, whereas canonical correspondence analysis indicated that
SSS and pH were the main environmental factors affecting diatom distribution. Based on the modern diatom
species–environmental variable database, we developed a diatom-based SSS transfer function to quantitatively
reconstruct the variability in SSS between 1984 and 2021 for sediment core HK3 from the Lianjiang coastal area.
The agreement between the reconstructed SSS and instrument SSS data from 1984 to 2021 suggests that diatom-
based SSS reconstruction is reliable for studying past SSS variability in the Lianjiang coastal area. Three low SSS
events in AD 2019, 2013, and 1999, together with an increased relative concentration of freshwater diatom species
and coarser sediment grain sizes, corresponded to two super-typhoon events and a catastrophic flooding event in
Lianjiang County. Thus, a diatom-based SSS transfer function for reconstructing past SSS variability in the
estuarine and coastal areas of the East China Sea can be further used to reflect the paleoenvironmental events in
this region.
Key words: diatom, transfer function, multivariate statistical analysis, environmental variable, sea surface salinity
Citation: Li Tong, Zhang Jihui, Li Dongling, Zhou Chengxu, Liu Chenxi, Xu Hao, Song Bing, Sha Longbin. 2024. Diatoms as indicators of
environmental change in coastal areas: a case study in Lianjiang coast of East China Sea. Acta Oceanologica Sinica, 43(8): 47–57, doi:
10.1007/s13131-024-2292-0
1 Introduction
Coastal estuarine areas are among the most dynamic and
complex environments on Earth (Rovira et al., 2012). They func-
tion as transition zones among marine, river, and terrestrial en-
vironments and exhibit large spatiotemporal environmental
gradients (Azhikodan and Yokoyama, 2015; Nwe et al., 2021).
Coastal sediments have long been recognised as reliable records
of marine environmental conditions (López-Belzunce et al., 2020;
Triantaphyllou et al., 2009), sea-level changes (Wang et al., 2013;
Yu et al., 2023a), freshwater inputs (Espinosa et al., 2022; Fayó
et al., 2018), and extreme weather events (Benito et al., 2015;
Nakanishi et al., 2022). Variations in freshwater input from rivers
and extreme events such as storm surges can strongly disturb the
water environment in coastal areas. These disturbances can be
partially recorded in the physical and chemical indicators of sedi-
ments as well as in microfossil assemblages. For example, Fayó
et al. (2018) used the alluvial sediments of the delta of the Color-
ado River to identify the historic changes in floodplain hydrology.
Nakanishi et al. (2022) used diatoms and chemical analyses to
reveal the history of extreme wave events in the coastal wetlands
of central Hidaka. The Lianjiang coast is located on the western
side of the East China Sea and is a typical coastal area with thick
sediment accumulations that record river flow dynamics, ocean
currents, and extreme events (Huh and Su, 1999). Thus, this area
is critical for reconstructing the evolution of the coastal environ-
ment and climate.
Coastal and estuarine areas are highly productive regions,
with phytoplankton being the primary producer in the food web
(Nwe et al., 2021; Saifullah et al., 2019). Diatoms, one of the most
important phytoplankton groups, can survive under conditions
of high turbidity (Lionard et al., 2008; Mendes et al., 2009).
Among the biological proxies preserved in sediments, diatoms
are valuable indicators for tracing environmental changes on
various timescales (Espinosa et al., 2022). They can tolerate and
adapt to a wide range of environmental conditions and respond
rapidly to physical, chemical, and biological variations in aquatic
ecosystems; moreover, their siliceous frustules are well-pre-
served in sediments and have a well-defined taxonomy (Chen
Foundation item: The National Natural Science Foundation of China under contract Nos 42376236 and 42176226.
*Corresponding author, E-mail: lidongling@nbu.edu.cn
Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57
https://doi.org/10.1007/s13131-024-2292-0
http://www.aosocean.com
E-mail: ocean2@hyxb.org.cn
et al., 2020a; Espinosa et al., 2022; Gregersen et al., 2023). After
death, planktonic and benthic diatoms sink to the sediment sur-
face and mix with in situ and displaced species (Chen et al.,
2019). This process reflects the local environmental conditions
and provides insights into the dynamics of runoff inputs and
ocean currents. Therefore, identifying the ecological variables
that regulate the seasonal and interannual succession of diatom
communities is valuable for monitoring and evaluating regional
environmental changes (Mendes et al., 2009).
Statistical inference models based on modern species–envir-
onment relationships effectively utilise ecological data derived
from biological assemblages, enabling the quantitative estima-
tion of critical environmental parameters (Hassan et al., 2009).
Thus, transfer functions are reliable for generating high-resolu-
tion quantitative estimates of paleoenvironmental conditions
(Hassan et al., 2009). Over the past decades, numerous diatom-
based transfer functions have been developed to infer a wide
range of environmental variables in lake and marine systems.
Fruitful results have been obtained in the study of the relation-
ship between diatoms and environmental variables in freshwater
ecosystems, such as temperature (Wang et al., 2014; Szczerba
et al., 2023), water depth (Chen et al., 2020b), and eutrophication
(Yang et al., 2008; Chen et al., 2022; Wang et al., 2012). Diatom-
based environmental transfer function datasets have been estab-
lished in different regions, such as Europe (Bennion et al., 2001),
Asia (Yu et al., 2023b), and South America (Gomes et al., 2014).
The study of marine diatom transfer functions has also made sig-
nificant progress in the reconstruction of the sea level (Zong and
Horton, 1999), sea surface temperature (De Sève, 1999; Li et al.,
2017), and sea ice extent (Sha et al., 2015). Although the marine
system includes estuarine and coastal parts, the environmental
characteristics of the land-sea interface area are more distinctive
than those of marine and lakes, involving a wider range of envir-
onmental factors, which make the development of transfer func-
tions more challenging (Hassan et al., 2009). Some research has
been conducted in estuaries and coastal areas. For example, Hor-
ton et al. (2006) developed the first diatom-based transfer func-
tion for the east coast of North America and used it to recon-
struct sea level changes. Hassan et al. (2009) built a two-compon-
ent Weighted Average (WA) Partial Least Squares (PLS) calibra-
tion model to infer salinity in three estuaries along the northeast-
ern coast of Argentina. However, owing to the significant differ-
ences in environmental characteristics and explanatory factors
among different estuarine and coastal regions, research on diat-
om transfer functions and database establishment remains in-
complete in the East China Sea.
The objectives of this study were to (1) identify the major en-
vironmental variables affecting diatom assemblages in the Lianji-
ang coastal area of the East China Sea, (2) establish estuarine and
coastal diatom-based transfer functions, and (3) test the reliabil-
ity of the transfer functions by reconstructing past environmental
events from a sediment core. Our results provide new ecological
information on the relationships between diatoms and environ-
mental parameters and their spatiotemporal distribution charac-
teristics, which may contribute to a better understanding of estu-
arine and coastal ecosystems affected by rivers.
2 Materials and methods
2.1 Environmental setting of the Lianjiang coast
The study area is located on the eastern coast of Lianjiang
County, Fujian Province, Southeast China (26.15°–26.30°N,
119.60°–119.80°E) (Fig. 1). This area has a typical mid-subtropic-
al maritime monsoon climate with warm, humid weather and
abundant rainfall (Lin et al., 2020; Peng et al., 2021). The average
annual rainfall is 1 551 mm, with the wet season (March to Se-
ptember) receiving approximately 80% of the precipitation and
the dry season (October to February) receiving the remaining ap-
proximately 20%. Typhoons are the primary weather hazards in
this region. Lianjiang County experiences an average of 5.5–5.7
tropical cyclones per year (Peng et al., 2021).
Both the Aojiang River and Minjiang River entered the study
area from the west (Fig. 1b). The Aojiang River is a medium-sized
river, 137 km long, with a watershed area of 2 655 km2, and an an-
nual freshwater discharge of 2.728 × 109 m3 (Lei et al., 2021;
Zhang, 2014). The Minjiang River is the largest river in Fujian
E4
E5
Y1
Y4
Y5
Y6
X4
X5
X6
W1
W2
W3
W4
W5
W6
E2
E4
E5
Y1
Y2
Y3
Y4
Y5
Y6
X1
X2
X3
X4
X5
X6
W1
W2
W3
W4
W5
W6
M1
M2 M3
E2
E5
Y3 X3
X4
X5
X6
W1
W2
W3
W4
W5
W6
October 2020
January 2021
April 2021
Hua ngqi Pe ninsula
Lia njian g County
119.6° 119.8°E119.7°
26.2°
estuary area
nearshore area offshore area
119° 120°E
26.0°
26.5°
Lianjiang
County
East China Sea
a
b
26.3°
HK3
central area 02 km
N
N
Aojiang River
Minjiang River
Minjiang River
Aojiang River
Fig. 1. Location of the study area in Lianjiang County. a. Topography and distribution of rivers in the study area. The Aojiang River is
in the northern part of the figure and the Minjiang River is in the southern part; the Lianjiang coast is influenced by both rivers. b. Loc-
ations of surface sediment samples and a sediment core HK3.
48 Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57
Province and flows into the southern part of the study area. It is
577 km long with a drainage area of 60 992 km2 and an average
annual freshwater discharge of 6.20 × 1010 m3 (Fan et al., 2021).
Both the Aojiang River and Minjiang River have strongly tidal es-
tuaries and are characterized by a regular semi-daily tidal range
of 0.74 m to 4.5 m, with an average of 3.8 m (Peng et al., 2021).
The study area is controlled by the Zhe-Min Coastal Current;
it supplies abundant nutrients to the coastal bays of Fujian in
winter, together with the nutrient input maximum from the Aoji-
ang River and Minjiang River, which have a large impact on the
local aquatic environment (Xu et al., 2020). Lianjiang County was
the second most important county in China for fish and shellfish
products over the last three decades (Lei et al., 2021; Yang et al.,
2022). The main fish species cultured were Larimichthys crocea,
Pagrosomus major, Sciaenops ocellatus, and Lateolabrax ja-
ponicus. The main cultured shellfish are Ruditapes philippinar-
um on sandy beaches and Sinonovacula constricta on clay
beaches, which are seeded in March and harvested in September
(Lei et al., 2021; Zhou et al., 2022).
2.2 Sample collection
A total of 52 surface sediment samples were collected from
the Lianjiang coast in October 2020, January 2021, and April 2021.
The sampling locations are shown in Fig. 1b. The environmental
parameters of each seawater sample, including sea surface tem-
perature (SST), sea surface salinity (SSS), pH, redox potential,
electrical conductivity, dissolved oxygen, total dissolved solids,
and turbidity were measured using an in situ multiparameter wa-
ter quality instrument (HORIBA U52G, Japan) at the time of
sample collection (Table S1). The depths of the 52 sampling sites
ranged from 1.5 m to 17 m, with an average depth of 8 m.
Sediment core HK3 was collected in October 2021 from the
tidal flats of the study area (26.25°N, 119.66°E) using a mudflat
sampler. The core measured 100 cm in length. Twenty samples
were extracted at 5 cm intervals from the core HK3.
2.3 Laboratory measurements
Diatom analysis was conducted on 15–16 mg of dried sedi-
ment per sample following the preparation methods described
by Håkansson (1984). Diatoms were counted and identified us-
ing a Motic BA410E microscope at 1 000× magnification. Diat-
oms were identified at the species or species group level follow-
ing the standard taxonomic literature for marine diatoms (Guo
and Qian, 2003; Jin et al., 1982; Krammer and Lange-Bertalot,
1986, 1988, 1991a, 1991b). At least 200 diatom valves were coun-
ted in random transects of most of the samples.
Diatom fluxes were calculated using the following equation
(Battarbee et al., 2001):
A=N×S
n×a×m,(1)
where A is the diatom concentration (valves/g), N is the number
of diatoms counted, S is the area of the Petri dish, n is the num-
ber of fields of vision counted, a is the area of one field of vision,
and m is the dry weight of the sample (g).
The Shannon-Weaver diversity index (SW index) was used to
reflect the biodiversity of diatoms in the samples, and the for-
mula is as follows (Shannon and Weaver, 1949):
H′=−
S
∑
i=(Ni
N)log(Ni
N),(2)
where H' is the SW index, S is the number of diatom species iden-
tified, Ni is the number of the ith diatom species, and N is the total
number of diatoms identified.
Grain size determination was performed on 72 samples (52
surface samples and 20 core samples) using a Beckman Coulter
laser diffraction particle size analyser (LS13320, USA) with a
measurement range of 0.04–2 000 µm. Samples were firstly thor-
oughly mixed and dried at 40℃ for 24 h and then a subsample of
about 0.2 g was taken from each sample. This was followed by the
addition of 5 mL HCl (10%) to eliminate carbonates, 5 mL H2O2
(10%) to remove organic matter, and 5 mL (NaPO3)6 to promote
dispersion before testing (Jiang et al., 2020).
2.4 Age model of core HK3
The chronological framework of core HK3 was established by
210Pb dating. After drying at low temperatures, the samples were
disaggregated using a mortar and pestle to produce a uniform
grain size. The activity of 210Pbex at each level was measured us-
ing a high-purity Ge Gamma Spectrometer (GWL-120-15, USA) at
the Nanjing Institute of Geography and Limnology, Chinese
Academy of Sciences. To provide the best possible new insights
into the sedimentary processes of the Lianjiang coast sediment
accumulation from the decay-corrected 210Pbex profiles, cores
were processed using the clay-normalisation procedure. Initial
210Pbex was recalculated by normalising to average clay based on
the percent of the clay-sized sediments (<4 μm) at each sample
(Sun et al., 2017, 2020).
The sediment accumulation rates were calculated using the
constant initial concentration (CIC) model (Sanchez-Cabeza and
Ruiz-Fernández, 2012):
[Pb(m)] = [ Pb()]e−λt,(3)
where [210Pb(m)] is the specific activity of [210Pbex] (Bq/kg) at
depth m, [210Pb(0)] is the surface sediment specific activity (Bq/
kg), and λ is the decay constant of [210Pb] (0.031 14 a−1).
2.5 Multivariate statistical analysis
Diatom assemblages and their relationships with environ-
mental variables were examined using multivariate statistical
analysis to determine the principal variables and detect similarit-
ies among the diatom samples (Chen et al., 2016).
Principal component analysis (PCA) has the advantage of
weight determination because it classifies the original data into
several comprehensive variables with characterisation signific-
ance using correlation coefficients to accurately reflect the core
information for the evaluation indicators (Abdi and Williams,
2010).
Canonical correspondence analysis (CCA) extracts the best
synthetic gradients from data on biological communities and en-
vironmental variables and intuitively shows the characteristics of
the relationship between these variables and biological taxa
(Klami et al., 2013). To determine the distribution of the surface
sedimentary diatom species, a detrended correspondence ana-
lysis (DCA) was required to determine the gradient length of the
ordination axes before selecting the appropriate model (Chen
et al., 2016). To identify the primary environmental factors influ-
encing the distribution of surface sediment diatoms in the study
area, sites with fewer than 200 diatoms were eliminated.
PCA and CCA were conducted using the CANOCO (version 5)
software (Ter Braak and Smilauer, 2012). The transfer function
was calculated using the C2 software developed by Juggins
Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57 49
(2007). The derivation ability of the model was evaluated accord-
ing to the root-mean-square error of prediction (RMSEPJack) and
squared correlation (R2Jack).
2.6 Modern SSS data
The modern SSS data used in this study were obtained from
the World Meteorological Organization climate explorer open
dataset (http://climexp.knmi.nl/start.cgi), which was averaged
month by month from 1900 to 2018. Monthly gridded SSS data for
1984–2018 were selected and converted to annual data for com-
parison with the diatom-based SSS reconstruction of core HK3.
3 Results and discussion
3.1 Establishing a diatom–environmental variables dataset
3.1.1 Diatom concentration and diversity in the surface samples
A total of 92 diatom species belonging to 36 genera were iden-
tified in surface sediment samples from the Lianjiang coastal
area. The dominant species are: Actinocyclus kuetzingii, Actino-
cyclus octonarius, Amphora coffeaeformis, Cyclotella striata,
Paralia sulcata, Planothidium delicatulum, Pleurosigma angu-
latum, Surirella armoricana, Thalassionema nitzschioides,
Thalassiosira leptopus, and Tryblioptychus cocconeisformis (Fig.
S1). Among them, Planothidium delicatulum and Aulacoseira
granulata are freshwater diatoms (Hartley et al., 1996; Hustedt,
1985); Actinocyclus octonarius, Cyclotella striata, S. armoricana,
and Paralia sulcata are commonly found in brackish water in es-
tuarine areas (Jiang et al., 2004; Prelle et al., 2019; Ran and Jiang,
2005). Some marine diatom species, such as Thalassionema nitz-
schioides, exhibit strong tolerance to SST and SSS (Jousé et al.,
1971).
Surface diatom concentrations and SW index values are
shown in Fig. 2. In October 2020, diatom concentrations were rel-
atively high in the estuary and offshore areas, with the highest
value at site W3 in the offshore area (1.65 × 106 valves/g), fol-
lowed by site E4 (1.52 × 106 valves/g), which also had the highest
SW index value. The nearshore areas had the lowest diatom con-
centrations, with site Y4 showing an almost complete absence of
diatoms and the lowest SW index.
In January 2021, the diatom concentration peaked in the off-
shore areas, showing significant variation compared to the other
zones. The highest values occurred at site W1 (2.64 × 106
valves/g) in the offshore area and the lowest concentrations oc-
curred at the nearshore sites. The area with a low SW index ex-
panded substantially in January compared to October, forming a
roughly north-south distribution that roughly divides the study
area into three parts. The SW index was relatively high in the es-
tuarine areas, northern part of the central area, and offshore
areas and relatively low in the nearshore areas, southern part of
the central area, and Minjiang River estuary area.
Generally, low diatom concentrations were recorded in April
2021. The SW index exhibited relatively minor variations across
the entire study area, with slightly lower values in the estuarine
and nearshore regions than in the central and offshore areas.
3.1.2 PCA of modern environmental variables
PCA of all environmental variables, including SST, SSS, pH,
redox potential, electrical conductivity, dissolved oxygen, total
dissolved solids, turbidity, and sedimentary mean grain size, was
performed to reduce the dimensionality of the dataset and de-
termine the major environmental gradients (Abdi and Williams,
2010). The eigenvalues of the first two principal components
(PC1 and PC2) were all greater than 1 and the cumulative propor-
tion of these components was 72.4%, therefore, they were used to
explain the main environmental variables.
The factor loading matrix expressed the loading (or degree of
influence) of each variable on each principal component. Var-
imax rotation was applied to the factor-loading matrix and the
coefficients of the principal components are shown in Fig. 3. PC1
was strongly related to SSS (52.2%) and PC2 was strongly related
to SST (50.6%). Therefore, we concluded that SSS and SST have
the greatest influence on the aquatic environment in the study
area.
3.1.3 Diatom species–environmental variables relationships
The gradient length of the first axis in the DCA was 2.68 SD,
indicating that the diatom data had a nonlinear unimodal distri-
bution (Ter Braak et al., 1988). This implied that the optimal or
highest concentration of diatoms occurred within a specific
E2
E4
E5
Y1
Y2
Y3
Y4
Y5
Y6
X1
X2
X3
X4
X5
X6
W1
W2
W3
W4W5
W6
M1 M2 M3
E4
E5
Y1
Y4
Y5
Y6
X4
X5
X6
W1
W2
W3
W4
W5
W6
a
E2
E5
Y3 X3
X4
X5
X6
W1
W2
W3
W4
W5
W6
E4
E5
Y1
Y4
Y5
Y6
X4
X5
X6
W1
W2
W3
W4
W5
W6
E2
E4
E5
Y1
Y2
Y3
Y4
Y5
Y6
X1
X2
X3
X4
X5
X6
W1
W2
W3
W4
W5
W6
M1 M2 M3
E2
E5
Y3 X3
X4
X5
X6
W1
W2
W3
W4
W5
W6
119.6° 119.7° 119.8°E
26.3°
26.2°
119.6° 119.7° 119.8°E
119.6° 119.7° 119.8°E
119.6° 119.7° 119.8°E
119.6° 119.7° 119.8°E
119.6° 119.7° 119.8°E
October
October
f
e
d
b c
January April
January April
Diatom concentration/
(106 valves·g−1)
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
Shannon-Weaver
diversity index
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
N
26.3°
26.2°
N26.3°
26.2°
N26.3°
26.2°
N
26.3°
26.2°
N26.3°
26.2°
N
Fig. 2. Spatial distribution of diatom concentrations and the Shannon-Weaver diversity index of diatoms in surface sediment samples
from the Lianjiang coastal area. a and d refer to October, b and e to January, c and f to April.
50 Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57
range of habitat gradients. Therefore, the unimodal ordination
technique of CCA was used to investigate the relationships
between the environmental variables and diatom species from
the surface samples. The Variance Inflation Factor (VIF) for each
environmental variable was used to determine whether it inde-
pendently affected the distribution of diatom combinations. The
VIF results showed that the SSS, electrical conductivity, and total
dissolved solids did not pass the test (VIF > 20). The strong correl-
ation among these three factors indicates a large overlap in the
provision of information on the aquatic environment (Blaine Mc-
Cleskey et al., 2023). After excluding electrical conductivity and
total dissolved solids, the remaining environmental variables
(SST, SSS, pH, redox potential, dissolved oxygen, turbidity, and
sedimentary mean grain size) passed the VIF and Monte Carlo
permutations tests (999 unrestricted permutations, p < 0.05).
The CCA results showed that the pseudo-canonical correla-
tions of CCA1 and CCA2 were 0.879 and 0.757, respectively, with
the first two axes accounting for 74.67% of the total variances.
This indicates that most of the constrained variance can be ex-
plained by the two axes.
The contributions of each factor to the CCA axis are shown in
Fig. 4. The primary environmental factor influencing CCA1 was
pH, which accounted for 79.8% of the total variance. The sedi-
mentary mean grain size contributed 69.2% to CCA1, making it
the second most influential factor. SSS had the highest contribu-
tion rate (75.4%) to CCA2. Consequently, pH and SSS were re-
garded as the most significant environmental factors affecting the
distribution of diatom assemblages in the surface sediments of
the study area, with the sedimentary mean grain size being the
second most important factor.
The distribution of diatom species can be separated into three
spatial zones based on the CCA findings of the diatom–environ-
mental factors (Fig. 4). Zone Ⅰ is dominated by diatoms such as
Planothidium delicatulum, Achnanthes suchlandtii, Amphora
coffeaeformis, Achnanthes laterostrata, and Navicula spectabilis.
These species were positively correlated with sedimentary mean
grain size and turbidity and negatively correlated with pH and
SSS. This indicates that they tend to accumulate in areas with
coarse-grained sediments, high turbidity, low alkalinity, and high
SSS. The nearshore and southern parts of the intertidal zone were
the main areas of coarse-grained sediments. Freshwater diatoms
transported by the Aojiang and Minjiang rivers enter this area
and because of directional sorting, only a small portion of the
smaller freshwater diatoms are retained in this area. The rela-
tionship between freshwater diatoms and coarse-grained sedi-
ments reflects the direct environmental effects on the deposition
of marine microorganisms. Zone Ⅱ was characterised by Aulaco-
seira granulata, Gomphonema parvulum, Cymbella affinis, and
Fragilaria capucina. Compared to ZoneⅠ, these species were
negatively correlated with SSS but were uncorrelated with the
sedimentary mean grain size. Zone Ⅲ was characterised by mar-
ine diatom species that occur within the study region, including
Actinocyclus octonarius, Cyclotella striata, Diploneis bombus,
Nitzschia sociabilis, Paralia sulcata, Surirella armoricana,
Thalassionema nitzschioides, Thalassiosira eccentrica, and Trybli-
optychus cocconeisformis. They are concentrated at the centre of
Fig. 4 in the positive direction of the SSS.
3.2 Age model for sediment core HK3
The calculated 210Pbex activity varied from (54.3 ± 9.4) Bq/kg
to (161.9 ± 11.45) Bq/kg in core HK3 (Fig. 5). Based on the char-
acteristics of the calculated 210Pbex activity, a CIC model was used
to calculate the sedimentation rates of core HK3, which were de-
termined to be 2.56 cm/a. These results are supported by Li et al.
(2009) and Liu et al. (2009). The 210Pb chronology indicated dates
of AD 1884 at a depth of 95–96 cm and AD 2021 at a depth of 0–1 cm.
W5
PCA1
PCA2
E5 E 4
Y6
Y5
W1
X6
X4 X5
E5 Y3
W2
E2 W 4
W3
W5
W6 W 6
M2
W3
W4
W2
X5
X6
X3
W5
X4
W1
W6
W1
X4
X5
W4
X6
W3
W2
X2
X3
X1
E2
Y4
Y5
Y6
M3
Y3
Y2
Y4
Y1
M1
Y1
E5
E4
10
5
0
−5
−5 0 5 10
April 2021
January 2021
October 2020
SST
ORPDO
Tur
MD
pH
SSS
C
TDS
Fig. 3. Summary of the results of the Principal Correspondence
Analysis (PCA) of the environmental variables: variable loadings
(arrows) and sample scores (coloured symbols) on PC1 and PC2.
The solid arrows represent the primary environmental factors
that have the maximum load on PCA1 and PCA2, respectively,
and the dashed arrows represent secondary impact factors. The
angle between arrows indicates the correlation between indi-
vidual environmental variables. SST: sea surface temperature;
ORP: redox potential; C: conductivity; Tur: turbidity; DO: dis-
solved oxygen; TDS: total dissolved solids; SSS: sea surface salin-
ity; MD: sediment mean grain size.
Tnit
Dbom
Pdel
Asuc
Agr a
Acof
Ni t soc
Psul
Pang
Sar m
Fcap
Akue
Aoct Tl ep
Tecc
Tcoc
Cst r
Alat
Npla
Abia
Ncoc
Ncon Nspe
Dam p
Csc u
Cpla
Gpar
Ⅰ
Ⅱ
Ⅲ
Hamp
Aspp
Ahun
Sunl
Dpap
Nf l u
Caf f
Pvi r Ds ub
Ael l
Ar up
Pbl a
−1.0 1.00 0.5−0.5
0
0.5
−0.5
SST
pH
SSS
ORP
MD
DO
Tur
CCA1
CCA2
Fig. 4. Canonical correspondence analysis (CCA) biplot of envir-
onmental variables and diatoms species. See Table S2 for abbre-
viations. Red symbols: diatoms associated with coarse-grained
sediments (Zone Ⅰ); green symbols: main freshwater diatom
species (Zone Ⅱ); blue symbols: predominantly marine diatoms
(Zone Ⅲ). SST: sea surface temperature; ORP: redox potential;
Tur: turbidity; DO: dissolved oxygen; SSS: sea surface salinity;
MD: sediment mean grain size.
Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57 51
3.3 Development of diatom-based transfer functions
The CCA results showed that pH and SSS were the primary
environmental factors influencing diatom distribution, whereas
the PCA results indicated that SSS was the most influential factor
in the study area. Similarly, the relative explanatory power of SSS
as a predictor of diatom assemblage composition can be estim-
ated by calculating the ratio of the eigenvalue of the first con-
strained axis (λ1) with SSS as a single explanatory variable with
the first unconstrained axis (λ2). The ratio λ1/λ2 is 2.468 (>1.0),
indicating that SSS is the main determinant of diatom distribu-
tion in the training set (Ter Braak and Colin Prentice, 1988)
(Table 1). Previous studies have also indicated that SSS is one of
the most important factors controlling diatom distribution in es-
tuarine environments (Hassan et al., 2007, 2009; Nwe et al., 2021;
Sarker et al., 2020). Therefore, a diatom-based SSS transfer func-
tion was developed to reconstruct the past changes in coastal en-
vironments.
Four numerical reconstruction methods were used to define
the optimal diatom-based SSS transfer function (Table S3). The
PLS method with 3 and 5 components resulted in a high R2Jack
(0.32 and 0.36), along with lower RMSEPJack (1.34 and 1.31) and
maximum biasJack (4.56 and 3.91) (Table S3). The WA-PLS method
with 4 and 5 components also resulted in a high R2Jack (0.36 and
0.37), as well as low RMSEPJack (1.29 and 1.29) and maximum bi-
asJack (3.02 and 3.11) (Table S3). Additionally, a plot of recon-
structed SSS versus observed SSS showed a strong linear correla-
tion, with randomly scattered residuals (Fig. 6). Hence, these four
numerical reconstruction methods can be employed to obtain di-
atom-based SSS in the Lianjiang coastal area.
3.4 Reliability of the diatom-based SSS reconstruction
The dominant species in core HK3 were similar to those ob-
served in surface sediment samples, including Achnanthes such-
landtii, Actinocyclus octonarius, Amphora coffeaeformis, Aulaco-
seira granulata, Cocconeis scutellum, Cyclotella striata, Cymbella
affinis, Gomphonema parvulum, Nitzschia sociabilis, Planothidi-
um delicatulum, Parakia sulcata, and Thalassionema nitzs-
chioides. This suggests a continuity of environmental change
from the past to the present in this area, which enables us to re-
construct paleoenvironmental conditions using the transfer func-
tion.
To determine the best model and test the reliability of the di-
atom-based SSS transfer function, the diatom data from core HK3
were used to quantitatively reconstruct SSS changes using the
four models screened above and the reconstructed SSS were
compared with modern SSS data. The results showed that the
PLS method (Figs 7c and d) was superior to the WA-PLA method
(Figs 7a and b) in terms of reconstruction, with a high correlation
coefficient and more significant p-values than modern SSS data.
In contrast, the PLS model with three components (Fig. 7c) had a
higher correlation coefficient (0.517) and more significant p-
value (0.001). Thus, the reconstructed SSS in core HK3 using the
PLS model with three components was found to best match the
actual SSS variations in the study area and was the most effective
model for SSS reconstruction in this area.
The reconstructed SSS shows three low-SSS events, occurring
in AD 2019, 2013, and 1999 (Fig. 8), coinciding with remarkably
low diatom concentrations and SW index, accompanied by an in-
crease in the relative abundance of freshwater diatoms (Figs 4
and 8; Zones Ⅰ and Ⅱ) and a decrease in the relative abund-
ance of marine diatoms (Figs 4 and 8; Zone Ⅲ). Abrupt low SSS
and diatom concentrations, as well as abrupt increases in fresh-
water diatoms to the marine environment, are probably related
to typhoons and floods (Yang et al., 2023). In 2018, Super
Typhoon Maria made direct landfall in Lianjiang County, with
extraordinarily heavy rainfall of more than 200 mm (Liu et al.,
2022). In 2013, Super Typhoon Soulik made landfall along the
Huangqi Peninsula coast in Lianjiang County, bringing heavy
rainfall to Fujian, and the Minjiang River caused excessive flood-
ing. Additionally, the low SSS in 1999 corresponded to a cata-
strophic flood event in 1998, when a major flood occurred in the
Minjiang River, with the maximum inflow at the Minjiang Estu-
ary Power Station reaching 37 000 m3/s, exceeding the historical
maximum. These rapid, high-amplitude increases in freshwater
inflows were likely to significantly reduce the SSS, and affect the
abundance and structure of the phytoplankton communities in
estuarine areas (Qiu et al., 2019; Yang et al., 2023).
In addition, three low-SSS events in AD 2019, 2013, and 1999
y = −82.27 ln(x) + 424.08
R² = 0.586 8
Depth/cm
Calibrated 210Pbex activity/(Bq·kg−1)
40 80 120 160 200
0
10
20
30
40
50
60
70
80
90
100
Fig. 5. Vertical profiles of 210Pbex activity, clay content, and cal-
culated 210Pbex activity in core HK3. Error bars consider counting
statistics uncertainties at 2σ.
Table 1. Results of the λ1/λ2 test of each environmental variable
Variable λ1λ2λ1/λ2
SSS 0.072 0.029 2.468
SST 0.025 0.085 0.292
pH 0.092 0.111 0.825
ORP 0.054 0.063 0.857
Tur 0.031 0.082 0.377
DO 0.038 0.091 0.415
MD 0.053 0.114 0.469
Note: SSS: sea surface salinity; SST: sea surface temperature; ORP:
redox potential; Tur: turbidity; DO: dissolved oxygen; MD: sediment
mean grain size.
52 Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
Residual error
Residual error
Observed SSS
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
22.8
24.0
25.2
26.4
27.6
28.8
30.0
31.2
32.4
Observed SSS
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
Observed SSS
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
Observed SSS
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
Observed SSS
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
Observed SSS
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
Observed SSS
22.8 24.0 25.2 26.4 27.6 28.8 30.0 31.2 32.4
Observed SSS
Reconstructed SSS
22.8
24.0
25.2
26.4
27.6
28.8
30.0
31.2
32.4
Reconstructed SSS
22.8
24.0
25.2
26.4
27.6
28.8
30.0
31.2
32.4
Reconstructed SSS
22.8
24.0
25.2
26.4
27.6
28.8
30.0
31.2
32.4
Reconstructed SSS
b
a d
c
f
eh
g
4.8
3.6
2.4
1.2
0
−1.2
−2.4
−3.6
−4.8
3.6
2.4
1.2
0
−1.2
−2.4
−3.6
Residual error Residual error
3.6
2.4
1.2
0
−1.2
−2.4
−3.6
4
3
2
1
0
−1
−2
−3
−4
Fig. 6. Plots of observed versus predicted values and observed versus residual (predicted minus observed) values for four transfer
function models derived for SSS. a and b: PLS with 3 components model; c and d: PLS with 5 components model; e and f: WA-PLS with
4 components model; g and h: WA-PLS with 5 components model.
WA-PLS component 4 method
correlation coefficient: 0.348
p value: 0.037
Reconstructed SSS
SSS of the CE dataset
WA-PLS component 5 method
correlation coefficient: 0.313
p value: 0.064
PLS component 3 method
correlation coefficient: 0.517
p value: 0.001
PLS component 5 method
correlation coefficient: 0.437
p value: 0.008
a
d
c
b
34.2
34.1
34.0
33.9
33.8
33.7
33.6
33.5
33.4
25 26 27 28 29 30 31
Reconstructed SSS
SSS of the CE dataset
34.2
34.1
34.0
33.9
33.8
33.7
33.6
33.5
33.4
25 26 27 28 29 30 31
Reconstructed SSS
SSS of the CE dataset
34.2
34.1
34.0
33.9
33.8
33.7
33.6
33.5
33.4
25 26 27 28 29 30 31
Reconstructed SSS
SSS of the CE dataset
34.2
34.1
34.0
33.9
33.8
33.7
33.6
33.5
33.4
25 26 27 28 29 30 31
Fig. 7. Correlations between the diatom-based reconstructed SSS for core HK3 and the modern SSS data. a. WA-PLS with 4 compon-
ents model. b. WA-PLS with 5 components model. c. PLS with 3 components model. d. PLS with 5 components model.
Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57 53
corresponded to coarser sedimentary grain sizes in core HK3
(Fig. 8). A study of the sedimentary processes revealed that the
sediment discharged by rivers increases under the influence of
typhoons and is then deposited in the estuary, accompanied by
an expansion of the area of coarse-grained sediments (Yang
et al., 2023). Rainfall triggered by a typhoon can cause a dramatic
increase in river runoff to the sea, which is several tens of times
higher than normal runoff (Zhao et al., 2008), whereas fine-
grained suspended sediments remain in suspension and previ-
ously deposited fine-grained sediments within the coastal zone
are resuspended and redistributed (Zang et al., 2018). This res-
ults in the coarsening of the sedimentary grain size and an in-
crease in sand content (Lou et al., 2016).
The analysis of the correlation between the reconstructed SSS
and the dominant diatom species showed a strong negative cor-
relation (−0.858, p < 0.01) between salinity and the relative con-
centration of Planothidium delicatulum. Planothidium delicatu-
lum is a freshwater diatom species commonly found in coastal
areas with low salinity (Hartley et al., 1996). A study on the distri-
bution of diatoms in the surface sediments of Qinzhou Bay and
Zhenzhu Bay in Guangxi revealed that Planothidium delicatu-
lum is mainly concentrated near the estuaries of the bays and is
almost absent in the open sea, which can effectively indicate low
sea surface salinity (Huang, 2017; Huang and Huang, 2016). Plan-
othidium delicatulum is mainly found in the Aojiang and Minji-
ang river estuaries and rarely occurs offshore, also indicating its
sensitivity to freshwater runoff. In addition, the relative concen-
trations of Surirella armoricana and Cyclotella striata showed a
positive correlation with the reconstructed SSS, with a correla-
tion of 0.6 (p < 0.01) and 0.51 (p < 0.05). Surirella armoricana and
Cyclotella striata are marine benthic species that prefer brackish
environments (Shang et al., 2023). Because the change in water
salinity in estuarine coastal areas over a short period is closely re-
lated to the dilution effect of runoff injection, the variations in
Planothidium delicatulum, Surirella armoricana, and Cyclotella
striata are likely to be important indicators of river action intensity
in estuarine coastal areas, indicating extreme events such as
typhoons, rainstorms, and floods.
4 Conclusions
To quantify the relationship between coastal sediment diat-
oms and environmental variables in the Lianjiang River estuar-
ine coastal areas and construct diatom-environment transfer
functions, 52 surface sediment diatom samples and nine types of
environmental factors were collected. The dominant diatom spe-
cies in this region were Actinocylus octonarius, Amphora coffeae-
formis, Actinocyclus kuetzingii, Cyclotella striata, Paralia sulcata,
Pleurosigma angulatum, Planothidium delicatulum, Surirella ar-
moricana. The diatom concentrations and SW index were relat-
ively high in the estuary and offshore areas and lower in the
nearshore areas.
The PCA results showed that the SSS and SST had the highest
contribution rates to the PC1 and PC2 axes. CCA results indic-
ated that pH and SSS were the most important environmental
factors affecting the distribution of diatom assemblages in the
surface sediments of the study area. The inferred SSS, based on
PLS with 3 components model, best matched the actual SSS vari-
ation in the study area. This model was used to quantitatively re-
construct the SSS changes from 1984 to 2021 and the results were
in good agreement with the measured SSS. In particular, the low
SSS events in 1999, 2013, and 2019 were consistent with an in-
crease in the relative concentration of freshwater diatoms and
the coarsening of sediment grain size records, corresponding to
two super-typhoon events and an extreme flood event in Lianji-
ang County. This transfer function is potentially useful for recon-
structing past SSS in estuarine coastal regions with applications
in future paleoceanographic studies.
Acknowledgements
We thank the editor and two anonymous reviewers for valu-
able comments to improve the manuscript.
1984
1986
1988
1990
1992
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
2019
2021
2927 31 3533
Reconstructed
SSS
0 0.1 0.2 0.3 0.4
Diatom concentration/
(10⁶ valves·g−1)
3.0 3.5 4.0 4.5
Shannon-Weaver
diversity index
4.4 5.2 6.0 6.8
Sediment mean
grain size (Φ)
0 0.2 0.4 0.6
Freshwater
diatom abundance/%
0 0.2 0.4 0.6
Marine
diatom abundance/%
Super Typhoon Maria
Super Typhoon Soulik
catastrophic flood
Yea r (AD)
Fig. 8. Time series of reconstructed sea surface salinity (grey intervals are error values), diatom concentration, Shannon-Weaver di-
versity index, sediment mean grain size, and relative abundance of freshwater and marine diatom species in core HK3. Note that ab-
rupt decreases in the SSS occurred in 2019, 2013, and 1999.
54 Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57
References
Abdi H, Williams L J. 2010. Principal component analysis. WIREs
Computational Statistics, 2(4): 433–459, doi: 10.1002/wics.101
Azhikodan G, Yokoyama K. 2015. Temporal and spatial variation of
mixing and movement of suspended sediment in the Macrotid-
al Chikugo River Estuary. Journal of Coastal Research, 31(3):
680–689, doi: 10.2112/JCOASTRES-D-14-00097.1
Battarbee R W, Jones V J, Flower R J, et al. 2001. Diatoms. In: Smol J,
Birks H, Last W, eds. Tracking Environmental Change Using
Lake Sediments Volume 3: Terrestrial, Algal, and Siliceous In-
dicators. Dordrecht: Springer, 155–202
Benito G, Macklin M G, Zielhofer C, et al. 2015. Holocene flooding
and climate change in the Mediterranean. CATENA, 130: 13–33,
doi: 10.1016/j.catena.2014.11.014
Bennion H, Appleby P G, Phillips G L. 2001. Reconstructing nutrient
histories in the Norfolk Broads, UK: implications for the role of
diatom-total phosphorus transfer functions in shallow lake
management. Journal of Paleolimnology, 26(2): 181–204, doi:
10.1023/A:1011137625746
Blaine McCleskey R, Cravotta III C A, Miller M P, et al. 2023. Salinity
and total dissolved solids measurements for natural waters: An
overview and a new salinity method based on specific conduct-
ance and water type. Applied Geochemistry, 154: 105684, doi:
10.1016/j.apgeochem.2023.105684
Chen Min, Li Yunhai, Qi Hongshuai, et al. 2019. The influence of sea-
son and Typhoon Morakot on the distribution of diatoms in
surface sediments on the inner shelf of the East China Sea. Mar-
ine Micropaleontology, 146: 59–74, doi: 10.1016/j.marmicro.
2019.01.003
Chen Xu, Liang Jia, Zeng Linghan, et al. 2022. Heterogeneity in diat-
om diversity response to decadal scale eutrophication in flood-
plain lakes of the middle Yangtze reaches. Journal of Environ-
mental Management, 322: 116164, doi: 10.1016/j.jenvman.
2022.116164
Chen Xu, McGowan S, Bu Zhaojun, et al. 2020b. Diatom-based wa-
ter-table reconstruction in Sphagnum peatlands of northeast-
ern China. Water Research, 174: 115648, doi: 10.1016/j.watres.
2020.115648
Chen Min, Qi Hongshuai, Intasen W, et al. 2020a. Distributions of di-
atoms in surface sediments from the Chanthaburi coast, Gulf of
Thailand, and correlations with environmental factors. Region-
al Studies in Marine Science, 34: 100991, doi: 10.1016/j.rsma.
2019.100991
Chen Xiang, Zhou Weiqi, Pickett S T A, et al. 2016. Diatoms are bet-
ter indicators of urban stream conditions: A case study in
Beijing, China. Ecological Indicators, 60: 265–274, doi: 10.1016/
j.ecolind.2015.06.039
De Sève M A. 1999. Transfer function between surface sediment diat-
om assemblages and sea-surface temperature and salinity of
the Labrador Sea. Marine Micropaleontology, 36(4): 249–267,
doi: 10.1016/S0377-8398(99)00005-5
Espinosa M A, Fayó R, Vélez-Agudelo C. 2022. Diatom-based pa-
leoenvironmental reconstruction from the coast of Northern
Patagonia, Argentina. Journal of South American Earth Sci-
ences, 116: 103874, doi: 10.1016/j.jsames.2022.103874
Fan Jiayu, Jian Xing, Shang Fei, et al. 2021. Underestimated heavy
metal pollution of the Minjiang River, SE China: Evidence from
spatial and seasonal monitoring of suspended-load sediments.
Science of the Total Environment, 760: 142586, doi: 10.1016/j.
scitotenv.2020.142586
Fayó R, Espinosa M A, Vélez-Agudelo C A, et al. 2018. Diatom-based
reconstruction of Holocene hydrological changes along the
Colorado River floodplain (northern Patagonia, Argentina).
Journal of Paleolimnology, 60(3): 427–443, doi: 10.1007/s10933-
018-0031-2
Gomes D F, Albuquerque A L S, Torgan L C, et al. 2014. Assessment
of a diatom-based transfer function for the reconstruction of
lake-level changes in Boqueirão Lake, Brazilian Nordeste. Pa-
laeogeography, Palaeoclimatology, Palaeoecology, 415: 105–
116, doi: 10.1016/j.palaeo.2014.07.009
Gregersen R, Howarth J D, Atalah J, et al. 2023. Paleo-diatom records
reveal ecological change not detected using traditional meas-
ures of lake eutrophication. Science of the Total Environment,
867: 161414, doi: 10.1016/j.scitotenv.2023.161414
Guo Yujie, Qian Shuben. 2003. Flora algarum marinarum sinicarum
(in Chinese), Volume 5, Diatom Phylum, Book 1, Central Out-
line. Beijing: Science Press, 1–493
Håkansson H. 1984. The recent diatom succession of Lake
Havgårdssjön, South Sweden. In: Proceedings of the Seventh
International Diatom Symposium. Philadelphia: Otto Koeltz,
411–429
Hartley B, Barber H G, Carter J R, et al. 1996. An Atlas of British Diat-
oms. Bristol: Biopress Ltd, 1–601
Hassan G S, Espinosa M A, Isla F I. 2007. Dead diatom assemblages
in surface sediments from a low impacted estuary: the
Quequén Salado river, Argentina. Hydrobiologia, 579(1):
257–270, doi: 10.1007/s10750-006-0407-6
Hassan G S, Espinosa M A, Isla F I. 2009. Diatom-based inference
model for paleosalinity reconstructions in estuaries along the
northeastern coast of Argentina. Palaeogeography, Palaeocli-
matology, Palaeoecology, 275(1–4): 77–91, doi: 10.1016/j.pa-
laeo.2009.02.020
Horton B P, Corbett R, Culver S J, et al. 2006. Modern saltmarsh diat-
om distributions of the Outer Banks, North Carolina, and the
development of a transfer function for high resolution recon-
structions of sea level. Estuarine, Coastal and Shelf Science,
69(3–4): 381–394, doi: 10.1016/j.ecss.2006.05.007
Huang Yue. 2017. Distribution of the surface sediment diatoms in the
outer bay of Qinzhou bay of Guangxi. Marine Sciences (in
Chinese), 41(1): 96–103, doi: 10.11759/hykx20150927001
Huang Yue, Huang Yuanhui. 2016. Characterastics of surface sedi-
ments diatom distribution in Zhenzhu Bay of Guangxi. Ad-
vances in Marine Science (in Chinese), 34(3): 411–420, doi: 10.
3969/j.issn.1671-6647.0000.00.011
Huh Chih-An, Su Chih-Chieh. 1999. Sedimentation dynamics in the
East China Sea elucidated from 210Pb, 137Cs and 239, 240Pu. Mar-
ine Geology, 160(1-2): 183–196, doi: 10.1016/S0025-3227(99)
00020-1
Hustedt F. 1985. The Pennate Diatoms. Koenigstein: Koeltz Scientif-
ic Books, 1–918
Jiang Yamei, Saito Y, Ta T K O, et al. 2020. Spatial and seasonal vari-
ability in grain size, magnetic susceptibility, and organic ele-
mental geochemistry of channel-bed sediments from the
Mekong Delta, Vietnam: Implications for hydro-sedimentary
dynamic processes. Marine Geology, 420: 106089, doi: 10.1016/
j.margeo.2019.106089
Jiang Hui, Zheng Yulong, Ran Lihua, et al. 2004. Diatoms from the
surface sediments of the South China Sea and their relation-
ships to modern hydrography. Marine Micropaleontology,
53(3-4): 279–292, doi: 10.1016/j.marmicro.2004.06.005
Jin Dexiang, Cheng Zhaodi, Lin Junmin, et al. 1982. Chinese marine
benthic diatoms (Volume 1) (in Chinese). Beijing: China Ocean
Press, 17–236
Jousé A P, Kozlova O G, Muhina V V. 1971. Distribution of diatoms in
the surface layer of sediment from the Pacific Ocean. In: Fun-
nell B M, Riedel W R, eds. The Micropalaeontology of Oceans.
London: Cambridge University Press, 263–269
Juggins S. 2007. C2 Version 1.5 User guide. Software for ecological
and palaeoecological data analysis and visualisation. New-
castle upon Tyne: Newcastle University, 1–73
Klami A, Virtanen S, Kaski S. 2013. Bayesian Canonical correlation
analysis. The Journal of Machine Learning Research, 14(1):
965–1003
Krammer K, Lange-Bertalot H. 1986. Bacillariophyceae 1. Teil: Na-
viculaceae. In: Ettl H, Gerloff J, Heynig H, et al, eds. Süsswasserflora
von Mitteleuropa, Band 2/1. New York: Gustav Fisher Verlag,
1–876
Krammer K, Lange-Bertalot H. 1988. Bacillariophyceae 2. Teil: Bacil-
lariaceae, epithemiaceae, surirellaceae. In: Ettl H, Gerloff J,
Heynig H, Mollenhauer D, eds. Susswasserflora von Mitteleur-
opa, Band 2/2. Jena: Gustav Fisher Verlag
Krammer K, Lange-Bertalot H. 1991a. Bacillariophyceae 3. Teil:
Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57 55
Centrales, fragilariaceae, eunotiaceae. In: Ettl H, Gerloff J,
Heynig H, Mollenhauer D, eds. Süsswasserflora von Mitteleur-
opa 2/3. Jena: Gustav Fisher Verlag, 1–576
Krammer K, Lange-Bertalot H. 1991b. Bacillariophyceae 4. Teil:
Achnanthaceae, kritische ergänzungen zu navicula (Lineolatae)
und gomphonema, gesamtliteraturverzeichnis Teil 1-4. In: Ettl
H, Gerloff J, Heynig H, Mollenhauer D, eds. Süsswasserflora
von Mitteleuropa 2/4. Jena: Gustav Fischer Verlag.
López-Belzunce M, Blázquez A M, Carmona P, et al. 2020. Multi
proxy analysis for reconstructing the late Holocene evolution of
a Mediterranean Coastal Lagoon: Environmental variables
within foraminiferal assemblages. CATENA, 187: 104333, doi:
10.1016/j.catena.2019.104333
Lei Jiajun, Yang Liyang, Zhu Zhuoyi. 2021. Testing the effects of
coastal culture on particulate organic matter using absorption
and fluorescence spectroscopy. Journal of Cleaner Production,
325: 129203, doi: 10.1016/j.jclepro.2021.129203
Li Dongmei, Liu Guangshan, Li Chao, et al. 2009. Radionuclide dis-
tribution in sediments and sedimentary rates in seas surround-
ing Xiamen. Journal of Oceanography in Taiwan Strait (in
Chinese), 28(3): 336–342
Li Dongling, Sha Longbin, Li Jialin, et al. 2017. Summer sea-surface
temperatures and climatic events in Vaigat Strait, West Green-
land, during the Last 5000 Years. Sustainability, 9(5): 704, doi:
10.3390/su9050704
Lin Xiaohong, Yin Siyu, Wu Wei, et al. 2020. Genetic diagnosis for
heavy typhoon rainfall attenuated by Fujian landfall. Tropical
Cyclone Research and Review, 9(3): 178–184, doi: 10.1016/j.tcrr.
2020.08.001
Lionard M, Muylaert K, Hanoutti A, et al. 2008. Inter-annual variabil-
ity in phytoplankton summer blooms in the freshwater tidal
reaches of the Schelde estuary (Belgium). Estuarine, Coastal
and Shelf Science, 79(4): 694–700, doi: 10.1016/j.ecss.2008.06.
013
Liu Shenfa, Shi Xuefa, Liu Yanguang, et al. 2009. Sedimentation rate
of mud area in the East China Sea inner continental shelf. Mar-
ine Geology & Quaternary Geology (in Chinese), 29(6): 1–7, doi:
10.3724/SP.J.1140.2009.06001
Liu Jingli, Zhang Han, Zhong Rui, et al. 2022. Impacts of wave feed-
backs and planetary boundary layer parameterization schemes
on air-sea coupled simulations: A case study for Typhoon Maria
in 2018. Atmospheric Research, 278: 106344, doi: 10.1016/j.at-
mosres.2022.106344
Lou Sha, Huang Wenrui, Liu Shuguang, et al. 2016. Hurricane im-
pacts on turbidity and sediment in the Rookery Bay National
Estuarine Research Reserve, Florida, USA. International Journ-
al of Sediment Research, 31(4): 330–340, doi: 10.1016/j.ijsrc.
2016.06.006
Mendes S, Fernández-Gómez M J, Resende P, et al. 2009. Spatio-
temporal structure of diatom assemblages in a temperate estu-
ary. A STATICO analysis. Estuarine, Coastal and Shelf Science,
84(4): 637–644, doi: 10.1016/j.ecss.2009.08.003
Nakanishi R, Ashi J, Miyairi Y, et al. 2022. Holocene coastal evolution,
past tsunamis, and extreme wave event reconstructions using
sediment cores obtained from the central coast of Hidaka,
Hokkaido, Japan. Marine Geology, 443: 106663, doi: 10.1016/j.
margeo.2021.106663
Nwe L W, Azhikodan G, Yokoyama K, et al. 2021. Spatio-temporal
distribution of diatoms and dinoflagellates in the macrotidal
Tanintharyi River estuary, Myanmar. Regional Studies in Mar-
ine Science, 42: 101634, doi: 10.1016/j.rsma.2021.101634
Peng Tong, Zhu Zhuoyi, Du Jinzhou, et al. 2021. Effects of nutrient-
rich submarine groundwater discharge on marine aquaculture:
A case in Lianjiang, East China Sea. Science of The Total Envir-
onment, 786: 147388, doi: 10.1016/j.scitotenv.2021.147388
Prelle L R, Graiff A, Gründling-Pfaff S, et al. 2019. Photosynthesis and
respiration of baltic sea benthic diatoms to changing environ-
mental conditions and growth responses of selected species as
affected by an adjacent peatland (Hütelmoor). Frontiers in Mi-
crobiology, 10: 1500, doi: 10.3389/fmicb.2019.01500
Qiu Dajun, Zhong Yu, Chen Yongqiang, et al. 2019. Short-term
phytoplankton dynamics during typhoon season in and near
the Pearl River Estuary, South China Sea. Journal of Geophysic-
al Research: Biogeosciences, 124(2): 274–292, doi: 10.1029/
2018JG004672
Ran Lihua, Jiang Hui. 2005. Distributions of the surface sediment di-
atoms from the south China sea and their palaeoceanographic
significance. Acta Micropalaeontologica Sinica, 22(1): 97–106
Rovira L, Trobajo R, Ibáñez C. 2012. The use of diatom assemblages
as ecological indicators in highly stratified estuaries and evalu-
ation of existing diatom indices. Marine Pollution Bulletin,
64(3): 500–511, doi: 10.1016/j.marpolbul.2012.01.005
Saifullah A S M, Kamal A H M, Idris M H, et al. 2019. Community
composition and diversity of phytoplankton in relation to envir-
onmental variables and seasonality in a tropical mangrove es-
tuary. Regional Studies in Marine Science, 32: 100826, doi: 10.
1016/j.rsma.2019.100826
Sanchez-Cabeza J A, Ruiz-Fernández A C. 2012. 210Pb sediment ra-
diochronology: An integrated formulation and classification of
dating models. Geochimica et Cosmochimica Acta, 82: 183–
200, doi: 10.1016/j.gca.2010.12.024
Sarker S, Yadav A K, Shahadat Hossain M, et al. 2020. The drivers of
diatom in subtropical coastal waters: A Bayesian modelling ap-
proach. Journal of Sea Research, 163: 101915, doi: 10.1016/j.
seares.2020.101915
Sha Longbin, Jiang Hui, Liu Yanguang, et al. 2015. Palaeo-sea-ice
changes on the North Icelandic shelf during the last millenni-
um: Evidence from diatom records. Science China Earth Sci-
ences, 58(6): 962–970, doi: 10.1007/s11430-015-5061-2
Shang Zhiwen, Li Jianfen, Freund H, et al. 2023. Quantitative rela-
tionship between surface sedimentary diatoms and water depth
in North-Central Bohai Bay, China. China Geology, 6(1): 61–69,
doi: 10.31035/cg2022040
Shannon C E, Weaver W. 1949. The Mathematical Theory of Com-
munication. Urbana: The University of Illinois Press, 1–117
Sun Xueshi, Fan Dejiang, Liao Huijie, et al. 2020. Variation in sedi-
mentary 210Pb over the last 60 years in the Yangtze River Estu-
ary: New insight to the sedimentary processes. Marine Geology,
427: 106240, doi: 10.1016/j.margeo.2020.106240
Sun Xueshi, Fan Dejiang, Tian Yuan, et al. 2017. Normalization of ex-
cess 210Pb with grain size in the sediment cores from the
Yangtze River Estuary and adjacent areas: Implications for sedi-
mentary processes. The Holocene, 28(4): 545–557, doi: 10.1177/
0959683617735591
Szczerba A, Rzodkiewicz M, Tylmann W. 2023. Modern diatom as-
semblages and their association with meteorological condi-
tions in two lakes in northeastern Poland. Ecological Indicators,
147: 110028, doi: 10.1016/j.ecolind.2023.110028
Ter Braak C J F, Colin Prentice I. 1988. A theory of gradient analysis.
Advances in Ecological Research, 18: 271–317, doi: 10.1016/
S0065-2504(08)60183-X
Ter Braak C J F, Smilauer P. 2012. Canoco Reference Manual and
User’s Guide: Software for Ordination, Version 5.0. Ithaca: Mi-
crocomputer Power.
Triantaphyllou M V, Ziveri P, Gogou A, et al. 2009. Late Glacial–Holo-
cene climate variability at the south-eastern margin of the Ae-
gean Sea. Marine Geology, 266(1–4): 182–197, doi: 10.1016/j.
margeo.2009.08.005
Wang Rong, Dearing J A, Langdon P G, et al. 2012. Flickering gives
early warning signals of a critical transition to a eutrophic lake
state. Nature, 492(7429): 419–422, doi: 10.1038/nature11655
Wang Zhanghua, Jones B G, Chen Ting, et al. 2013. A raised OIS 3 sea
level recorded in coastal sediments, southern Changjiang delta
plain, China. Quaternary Research, 79(3): 424–438, doi: 10.
1016/j.yqres.2013.03.002
Wang Qian, Yang Xiangdong, John Anderson N, et al. 2014. Diatom
response to climate forcing of a deep, alpine lake (Lugu Hu,
Yunnan, SW China) during the Last Glacial Maximum and its
implications for understanding regional monsoon variability.
Quaternary Science Reviews, 86: 1–12, doi: 10.1016/j.quascirev.
2013.12.024
Xu Zhimeng, Li Yifan, Lu Yanhong, et al. 2020. Impacts of the Zhe-
56 Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57
Min Coastal Current on the biogeographic pattern of microbial
eukaryotic communities. Progress in Oceanography, 183:
102309, doi: 10.1016/j.pocean.2020.102309
Yang Liyang, Chen Yu, Lei Jiajun, et al. 2022. Effects of coastal
aquaculture on sediment organic matter: Assessed with mul-
tiple spectral and isotopic indices. Water Research, 223: 118951,
doi: 10.1016/j.watres.2022.118951
Yang Xiangdong, John Anderson N, Dong Xuhui, et al. 2008. Surface
sediment diatom assemblages and epilimnetic total phosphor-
us in large, shallow lakes of the Yangtze floodplain: their rela-
tionships and implications for assessing long-term eutrophica-
tion. Freshwater Biology, 53(7): 1273–1290, doi: 10.1111/j.1365-
2427.2007.01921.x
Yang Shangshang, Li Yunhai, Lin Yunpeng, et al. 2023. Evolution of
sedimentary dynamic process/pattern in the Quanzhou Bay
under impact of Typhoon Matmo (2014). Regional Studies in
Marine Science, 62: 102974, doi: 10.1016/j.rsma.2023.102974
Yu Fengling, Li Nannan, Tian Ganghua, et al. 2023a. A re-evaluation
of Holocene relative sea-level change along the Fujian coast,
southeastern China. Palaeogeography, Palaeoclimatology, Pa-
laeoecology, 622: 111577,doi: 10.1016/j.palaeo.2023.111577
Yu Siwei, Wang Junbo, Rühland K M, et al. 2023b. Spatial distribu-
tion of surface-sediment diatom assemblages from 45 Tibetan
Plateau lakes and the development of a salinity transfer func-
tion. Ecological Indicators, 155: 110952, doi: 10.1016/j.ecolind.
2023.110952
Zang Zhengchen, George Xue Z, Bao Shaowu, et al. 2018. Numerical
study of sediment dynamics during hurricane Gustav. Ocean
Modelling, 126: 29–42, doi: 10.1016/j.ocemod.2018.04.002
Zhang Rijun. 2014. Construction of digital Aojiang watershed. Ap-
plied Mechanics and Materials, 687–691: 2157–2160, doi: 10.
4028/www.scientific.net/AMM.687-691.2157
Zhao Hui, Tang Danling, Wang Yuqing. 2008. Comparison of phyto-
plankton blooms triggered by two typhoons with different in-
tensities and translation speeds in the South China Sea. Marine
Ecology Progress Series, 365: 57–65, doi: 10.3354/meps07488
Zhou Min, Fang Futao, Zeng Cong, et al. 2022. Community competi-
tion is the microorganism feedback to sedimentary carbon de-
gradation process in aquaculture tidal flats. Frontiers in Mar-
ine Science, 9: 880120., doi: 10.3389/fmars.2022.880120
Zong Yongqiang, Horton B P. 1999. Diatom-based tidal-level trans-
fer functions as an aid in reconstructing Quaternary history of
sea-level movements in the UK. Journal of Quaternary Science,
14(2): 153–167, doi: 10.1002/(SICI)1099-1417(199903)14:2<153::
AID-JQS425>3.0.CO;2-6
Supplementary information:
Table S1. Environmental variables in surface sampling sites.
Table S2. Full names and abbreviations of the main diatom species in the Lianjiang coastal area.
R
Jack
Table S3. Results of method testing for transfer function. Maximum bias (Max BiasJack), coefficient of determination between ob-
served and predicted values , and root mean squared error of prediction, based on the leave-one-out jack-knifing (RMSEPJack) for
the reconstructed SSS in seven reconstruction procedures. WA: weighted averaging regression and calibration, calibration, WA(tol):
weighted averaging with tolerance downweighting, PLS: partial least squares, WA-PLS: weighted averaging with partial least squares,
IKM: Imbrie and Kipp Model. Both the inverse and classical deshrinking regressions were used in the WA and WA(tol) reconstruction
procedures. The tests showed that PLS with three and five components and WA-PLS with four and five components were the most reli-
able (values in bold).
Figure S1. Dominant and typical diatoms on the Lianjiang coast.
The supplementary information is available online at https:// doi.org/10.1007/s13131-024-2292-0 and http://www.aosocean.
com/. The supplementary information is published as submitted, without typesetting or editing. The responsibility for scientific accur-
acy and content remains entirely with the authors.
Li Tong et al. Acta Oceanol. Sin., 2024, Vol. 43, No. 8, P. 47–57 57
