Journal of Asian Earth Sciences 114 (2015) 447456

Contents lists available at ScienceDirect

Journal of Asian Earth Sciences

journal homepage: www.elsevier.com/locate/jseaes

Late Holocene environment of subalpine northeastern Taiwan from

pollen and diatom analysis of lake sediments
Liang-Chi Wang a,b, Hermann Behling b, Shuh-Ji Kao c, Hong-Chun Li d, Kandasamy Selvaraj c,
Meng-Long Hsieh e, Yuan-Pin Chang f,
a

Collection Management Department, National Taiwan Museum, Taipei 100, Taiwan

Department of Palynology and Climate Dynamics, Albrecht-von-Haller Institute for Plant Sciences, University of Gttingen, Gttingen 37073, Germany
State Key Laboratory of Marine Environmental Science, Xiamen University, Xiamen 361005, China
d
Department of Geosciences, National Taiwan University, Taipei 106, Taiwan
e
Department of Earth and Environmental Sciences, National Chung Cheng University, Chiayi 621, Taiwan
f
Department of Oceanography, National Sun Yat-sen University, Kaohsiung 804, Taiwan
b
c

a r t i c l e

i n f o

Article history:
Received 23 September 2014
Received in revised form 20 March 2015
Accepted 24 March 2015
Available online 31 March 2015
Keywords:
Typhoon
East Asian summer monsoon
ENSO
Precipitation
Vegetation history
Paleolimnology

a b s t r a c t
We used multi-decadal pollen and diatom records from sediment core TFL-1 from Tsuifong Lake to reconstruct the vegetation dynamics and hydroclimate in northeastern Taiwan during the past 3500 cal BP.
Coarse grained sediments in association with higher percentages of wetland pollen (Cyperaceae) and
upper conifer pollen (Tsuga and Pinus) in the lower part of the core indicate low lake levels and a relatively cold/dry climate between 3500 and 2030 cal BP, reecting a decline of the East Asian summer monsoon (EASM). Muddy sediments coupled with reduction of wetland pollen represent the rise of lake
levels, implying the re-strengthening of the EASM during the past 2000 years. Paleotemperature was
inferred from the variation of pollen origin from the upper and lower mountain forest, indicating the global temperature anomalies of the Medieval Warm Period (MWP) and the Little Ice Age (LIA). In comparison to the main climate forces in the North Pacic, we suggest that the long-term climatic trend in
Taiwan was controlled by variations in EASM intensity, while increased precipitation over the past
2000 years may also be linked to warmer sea surface temperature (SST) of the western Pacic Warm
Pool (WPWP) and increased El Nio-Southern Oscillation (ENSO) events, which increased typhoon intensity. Higher diatom-inferred pH during 29302030 cal BP and the LIA suggest strong hydrological disturbances, reecting more typhoons passing over Taiwan. The frequent typhoon events could be linked by
an abrupt shift of typhoon track, due to the reduction of the WPWP and expansion of the Northwestern
Pacic High, which move typhoons in a more westerly direction.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
The East Asian summer monsoon (EASM) and typhoon systems
of the West Pacic have signicant effects on the regions hydroclimate. The summer monsoon precipitation is an important factor
for food production and maintaining ecosystem biodiversity. Its
anomaly often induces extreme ooding or droughts in the worlds
most populated area (Zhang et al., 2011). Typhoons are one of the
most extreme natural hazards and a major cause of forest ecosystem disturbances (Ito, 2010), catastrophic landslides (Tsou et al.,
2011), and organic carbon transportation at tectonically active
margins (Kao and Liu, 1997; Goldsmith et al., 2008) across the
regions. A clear understanding of historical EASM and typhoon
Corresponding author. Tel.: +886 75252000x5161.
E-mail address: yuanpin.chang@mail.nsysu.edu.tw (Y.-P. Chang).
http://dx.doi.org/10.1016/j.jseaes.2015.03.037
1367-9120/ 2015 Elsevier Ltd. All rights reserved.

activities and regional precipitation-related variations is urgently

required to improve future predictions of climate change.
Paleo-EASM studies from stalagmite, and marine and lacustrine
sediment records revealed that EASM strength closely followed
changes in long-term boreal insolation change, showing a continuously decreasing trend in association with drier climate in the lowlatitude regions of East Asia since the late Holocene (4500 cal BP)
(Laskar et al., 2004; Selvaraj et al., 2007; Wang et al., 2014a; Yuan
et al., 2004). Some monsoonal records derived from subtropical
and tropical regions suggested that the return of EASM rainfall during the past 2000 years cannot be explained by Holocene summer
insolation in the northern Hemisphere (Zhao et al., 2013). The spatial variations in precipitation represent the characteristics of
regional climate, which are mainly affected by monsoon variations.
In recent years, studies on typhoon have focused on the
variations in typhoons intensity, frequency and track on the

448

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

intra-seasonal to inter-decadal scales using observational datasets

over a 30 year period (Tu et al., 2009). Typhoon strength is generally linked to equatorial sea surface temperature (SST) (Lin et al.,
2009), and the shift of typhoon tracks in the West Pacic is affected
by the El Nio-Southern Oscillation (ENSO) phenomenon (Tu et al.,
2009). However, studies on paleo-typhoons as well as their
inuences on the local hydrology and forest ecosystems are rare,
especially those from the late Holocene period (Chen et al., 2012;
Wang et al., 2014b).
The Medieval Warm Period (MWP) and Little Ice Age (LIA) are
multi-centennial-scale temperature anomalies with global signatures during the past 1000 years. The MWP is characterized by
relative warmth while the LIA is characterized by pronounced cooling (Mann et al., 2009). Although there are several regional differences, the temporal ranges of the MWP and LIA could be dened as
950650 cal BP and 55050 cal BP, respectively (IPCC, 2013).
However, during both intervals, precipitation changes there were
signicant regional variations in precipitation changes (Chen
et al., 2015; Hurrell, 1995).
Since Taiwan, is an island located near the North Pacic/Eurasia
boundary, it is close to the turning track for most West Pacic
typhoons, which is strongly inuenced by EASM (Fig. 1).
Paleoclimate records from lacustrine sediments have been shown
to be useful in detecting EASM trends and typhoon activities
(Chen et al., 2012; Wang et al., 2014b). Due to the high erosion rate
and frequent occurrences of typhoons and earthquakes, undisturbed high-resolution late Holocene records are very rare in subalpine Taiwan (Chen et al., 2009; Liew and Huang, 1994). Selvaraj
et al. (2012) rst reconstructed the high-resolution climate variations using elemental and isotopic data from the sediments of
Emerald Peat Lake (also known as Tsuifong Lake, TFL) in NE
Taiwan. The intense precipitation during 37302030 cal BP that
was inferred using a geochemical proxy from the TFL record is contrary to most published late Holocene palynological

reconstructions of Taiwan (Lee and Liew, 2010; Liew et al., 2014,

2006; Liew and Huang, 1994; Wang et al., 2014a). Further research
is required to understand the contradictions between pollen and
geochemical proxies, which will allow for an improved understanding of late Holocene precipitation patterns for Taiwan and
potential links to typhoons and the EASM.
Pollen records are considered one of the most useful indicators
for determining the decadal impacts of climate change on land
ecosystems (Tinner and Lotter, 2001). Diatom analysis is also a
constructive tool for establishing hydrological conditions.
Because of methodological limitations in previous studies, the vegetation history and climate pattern in Taiwan during the late
Holocene period remains incomplete and controversial. Taking this
uncertainness and importance of climate trends in Taiwan into
account, we used pollen and diatom records from well-dated TFL
lake sediments to reconstruct reliable vegetation history and
hydrological changes for the last 3.5 ka. We compared our results
with the established paleoclimatological proxies in the North
Pacic to examine whether variations in EASM strength and
typhoon activities are relative to known climate forces during the
Holocene epoch.

2. Study site
The TFL (24300 N, 121360 E, 1840 m above sea level (a.s.l.)) is a
subalpine, hydrological-closed, seepage lake in northeastern
Taiwan (Fig. 1). The lake is recharged mainly by precipitation, runoff and groundwater input from small catchments. The climatic
condition of the area is mainly affected by subtropical monsoon
system with a mean annual temperature of 12.5 C (ranges from
3.3 to 17.9 C), and a precipitation of 362 mm (ranges from 117
to 788 mm) based on records from the Taipinsan weather station
(24300 N, 121310 E, 1810 m a.s.l.). Bedrock components of this lake

Fig. 1. Important climate features of Taiwan and the location of Tsuifong Lake (TFL). Locations (black squares) are denoted as follow: (1) TFL; (2) marine sediment core
17,940; (3) Hani peat bog; (4) Lake Qinghai; (5) marine sediment core MD982141. EASM = East Asian summer monsoon.

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

449

Fig. 2. This diagram denotes TFL-1 core sampling site (solid star) and vegetation distributions within dry and wet seasons in TFL basin. Two pie charts represent the water
coverages, rock fragments and the four vegetation zones during dry and wet periods, respectively. The vegetation maps are redrawn according to the TFL survey in April 2006
(dry season) and in August 2006 (wet season) by Mao (2006).

basin are mainly compose of slate, embedded with some metamorphic sandstone and conglomerate layers (Integrated Geological
Data Inquiry System, Central Geological Survey, MOEA, Taiwan).
The vegetation type on northeastern Taiwan is strongly inuenced by altitudinal changes. Trees growing in the upper mountain
needle-leaf forest (corresponding to upper mountain forest, UMF)
at height above 2200 m a.s.l. are dominated by Tsuga chinensis
var. formosana and Pinus taiwanensis. At altitudes between 1200
and 2200 m a.s.l., Chamaecyparis formosensis, T. chinensis var. formosana, Cyclobalanopsis morii, Alnus formosana, Trochodendron aralioides, and Hydrangea angustipetala are the main components of
montane mixed needle-broad-leaf forest (corresponding to mountain forest, MF). At altitudes below 1200 m a.s.l., Machilus thunbergii, Castanopsis cuspidata var. carlesii, Engelhardtia roxburgiana,
Lagerstroemia subcostata, Mallotus japonicus, Myrica rubra, Morus
australis, and Ardisia sieboldii are dominated in lower mountain
evergreen broad-leaved forest (corresponding to the lower mountain forest, LMF). Vegetation type in the study area is assigned to be
mountain mixed needle-broad-leaf forest (MF), although the primary forest is reclaimed by yellow cypress (C. formosensis and
Cryptomeria japonica) during AD 19261982.
The TFL basin is covered mainly by lake water, rock fragments,
moss, shrubs, Poaceae and aquatic plants during the dry seasons.
Otherwise, the rock fragments and moss areas are totally submerged with higher lake levels during the wet seasons (Mao,
2006) (Fig. 2). In the shrub zone, domination of Pourthiaea villosa
var. parvifolia is less impacted by uctuated seasonal water table.
Miscanthus sinensis and Haloragis micrantha are the main species
of the Poaceae zone, which is presents at lake borders during wet
seasons. Carex phacota, Festuca ovina, Polygonum thunbergii,
Polygonum praetermissum are characteristic species of the aquatic
zone, and are submerged more than 50% of the duration of wet seasons (Fig. 2). Moss and rock fragment areas are represented merely
at lake borders during the dry periods.
3. Materials and methods
3.1. Field coring and chronological framework
The sediment core, TFL-1, was collected from the central region
of the TFL at November 2004. A continuous 168-cm long lacustrine
record was retrieved by using a 2 m long stainless gravity corer.
Sediments were subsampled at 1 cm interval and freeze-dried
before further analysis. The lithology of TFL-1 is mainly composed

of coarse to ne sand between 168 and 100 cm. Sediments of the

top 100 cm of core consist of ne mud. More details of the sampling information and lithology are all presented in previous studies published by Selvaraj et al. (2012).
For the establishment of the chronology, seven radiocarbon dating points with plant remains were obtained using accelerator
mass spectrometry (AMS) done by Rafter Radiocarbon Laboratory
located in New Zealand (Selvaraj et al., 2012). The AMS radiocarbon dates were calibrated to calendar year before present (cal
BP) by using CALIB program 6.0 with IntCal13 calibration curve
(Reimer, 2013). For maximizing the information of all available
dates, the age-depth model were carried out by Bayesian analysis
through usage of the Bacon program for R (Blaauw and Christen,
2011).
3.2. Laboratory analyses, data treatments, and diatom inferred-pH
Pollen samples were taken at 2-5 cm intervals. Around 0.5 g dry
sediment was processed using the standard method, including
chemical treatment with 10% HCl, 40% HF, acetolysis, and cleaned
with an ultrasonic bath on a nylon sieve in 5 lm mesh size, then
mounting with glycerin jelly onto the slide for microscopic observations. On table with a known number of Lycopodium spores was
added into samples before chemical treatments for calculating the
pollen concentration (grains per gram). Pollen grains were then
identied by referring to pollen and spore ora for Taiwan
(Huang, 1981, 1972). Totally, 300 pollen grains were counted of
each sample. Percentages of pollen, fern spores, and green algae,
Botryococcus, are computed based on the sum of terrestrial pollen
grains, in which wetland pollens were excluded. Based on modern
vegetation distributions of the study area (Fig. 1), we grouped
arboreal pollen spectra to UMF, MF, and LMF, respectively, in this
study. The non-arboreal pollen spectra was attributed into two
subgroups referring to the ecological habitats of the TFL basin
(Fig. 2) as terrestrial herbs (corresponding to the Poaceae zone),
and wetland plants (corresponding to the aquatic zone).
In regards to the diatom analysis, sediments were also subsampled with intervals of 25 cm along the core. 0.5 g dry subsamples
were then treated with 10% HCl, 30% H2O2, and digested fully
within a water bath (at 90 C). After rinsing with distilled water,
the processed samples then were mounted on slides by using
mounting media (Wako). Diatom concentration (valve numbers
in per gram sediment) was estimated with a known volume of processed solution sample. The observational area of the coverslip was

450

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

examined under microscope as well as the required number of diatom specimen was counted (300 grains) (Friel et al., 2014).
Taxonomic identication and ecological interpretation are based
primarily on Freshwater Diatom Flora of Taiwan (Wu et al.,
2011) and collecting references (Charles, 1985; Krammer and
Lange-Bertalot, 1986; Wang et al., 2010). All diatom taxa were
subgrouped according to the inferring pH values (alkaliphilous,
circumneutral, acidophilous).
Based on a previous study examining diatom inferred pH
(DI-pH) in the TFL (Wang et al., 2013), we estimated the DI-pH of
core TFL-1 by using a transfer function derived from the combined
pH training set of European Diatom Database Initiative (EDDI)
(Juggins, 2001). The EDDI transfer function for pH reconstruction
is developed by using locally-weighted weighted averaging
method (LWWA). Validation of this approach had been undertaken
before proceeding. The web verication program of EDDI-set
(http://craticula.ncl.ac.uk/Eddi/jsp/verify.jsp) was executed to verify taxonomy against an EDDI transfer function. Our result prove
that the sum of fossil diatom at each level is more than 75%, suggesting the high reliability of reconstructed DI-pH of this study
(Birks and Birks, 1998).
3.3. Numerical methods for pollen data
Numerical analyses for pollen data were based on the terrestrial
pollen percentages by excluding the wetland pollen taxa. In order
to reduce the effect causing by rare taxa which has less inuence
on explaining varied vegetation, only pollen taxa values larger than
1% were included. The pollen-based biostratigraphy zones were
identied using CONISS cluster analysis within the computer programs Tilia and Tiliagraph (Grimm, 1993). Pollen and diatom diagrams were then constructed using the Tilia, Tiliagraph and C2
programs (version 1.5 for Windows) (Grimm, 1993; Juggins, 2005).
Multivariable analysis was conducted to examine the links
between important environmental variables with the pollen data.
Multivariable analysis is undertaken with the program CANOCO
version 4.5 (ter Braak and Smilauer, 2002). Pollen percentages
were square-root-transformed to stabilize variance and to smooth
the impact of dominate taxa (Jantz and Behling, 2012). Detrended
correspondence analysis (DCA) was carried out to dissolve the
underlying linearity of the pollen data in this study. The rst axis
of DCA with shorter gradient length of 0.96 standard deviations
was obtained, indicating that the principle component analysis
(PCA) was appropriate to explore the relationship between pollen
assemblages and environmental variables (Ammann et al., 2000;
ter Braak and Smilauer, 2002). Several proxies were picked, including percentages of wetland pollen and fern spores, published TOC,
TN, C/N ratio, d13Corg and, d15N (Selvaraj et al., 2012), as supplementary variables for PCA analysis to enrich the environmental
interpretation (Abdi and Williams, 2010). The values of all supplementary variables were denoted with log10(x + 1) and transformed
to standardize the variances (Cao et al., 2014).
4. Results
4.1. Chronology
Seven radiocarbon dates were used to establish the age-depth
model based on Bayesian analysis for the core TFL-1 (Fig. 3). The
radiocarbon date of plant remains in the surface sediment represented a negative age of 800 BP, which reected the inuence
of modern nuclear weapon testing. Thus, the chronology of surface
sediment was assumed as present time (54 cal BP) (Selvaraj et al.,
2012). The Bayesian age-depth analysis yielded a nearly linear
interpolation model (Fig. 3), covering the period from 3500 cal

Fig. 3. Age-depth model of core TFL-1 bases on seven AMS dates and surface
sample (54 cal BP) by using Bayesian software Bacon program for R (Blaauw and
Christen, 2011). Radiocarbon dates are calibrated with 2r errors of condence level.
Gray shades indicate models 95% probability intervals used in this study.

BP to present with an average sedimentation rate of 0.05 cm/yr.

The samples below 59 cm of core depth showed a slower sedimentation rate of approximately 0.04 cm/yr, while the samples
above 59 cm had a higher sedimentation rate with an average of
0.07 cm/yr.
4.2. Biostratigraphy trends
Pollen grains were preserved in all samples (Fig. 4). Diatoms
appeared in sufcient abundance for counting at depths above
136 cm (Fig. 5). The compositions and concentrations of pollen
and diatom assemblages show variations consistent with the
lithology. To clarify the analysis, ve stratigraphic zones based
on pollen cluster analysis were used for further discussion (Fig. 4).
The pollen concentration was low in zone TFL1 (168143 cm,
35002930 cal BP). UMF and MF taxa were dominant. Tsuga (6
10%), Pinus (39%), Chamaecyparis (410%), Quercus (2236%) and
Alnus (1625%) were the main tree pollen components. Poaceae
pollen is relatively dominant and represents the main terrestrial
herb with an average of approximately 14%. This zone is characterized by high percentages of Cyperaceae (2135%), the wetland pollen, and fern spores (5383%). Green algae Botryococcus was also
observed (17%) (Fig. 4).
In zone TFL2 (143100 cm, 29302030 cal BP), the pollen concentration rose gradually, and MF pollen was dominant (>60%) in
this zone. Otherwise, Alnus pollen (1731%) became more pronounced compared to zone 1. Poaceae pollen (1122%) increased
slightly toward the end of the zone. The percentage of
Cyperaceae (1236%) showed large uctuations. Fern spores
decreased markedly from 60% to <20%. The Green algae,
Batryococcus (25%), had a lower percentage value than in the previous zone, and the diatom concentration was extremely low
(Fig. 4). The acidophilus diatoms, including Aulacoseira distance
(4870%), Frustulia crassinervia (1428%), and Pinnularia gibba (2
17%) dominated the ora. Among the circumneutral diatoms,
Pinulari virdis (216%) was relatively high (Fig. 5). Based on the diatom data, estimated DI-pH varied between 5.46 and 5.96 (Fig. 6).
Zone TFL3 (10069 cm, 20301100 cal BP) was characterized by
mud deposits with higher pollen concentrations compared with
previous zones. The percentages of MF pollen remained high
(60%) with few uctuations. The decreased trend of Cyperaceae
percentage (1126%) was noted above 81 cm of core depth
(1500 cal BP), while fern spores increased slightly (2231%)

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

451

Fig. 4. Lithology and percent pollen diagram with biostratigraphic zones determined with constrained cluster analysis (CONISS) for core TFL-1. Only taxa with percentage
larger than 1% are plotted. Percentages of all taxa are based on the sum of terrestrial pollen. Secondary y-axis presents calibrated age. UMF = upper mountain forest,
MF = mountain forest, LMF = lower mountain forest.

Fig. 5. Percentages of diatom taxa respect to pollen zones in TFL-1. Only the main taxa with >2% in percentage are plotted.

(Fig. 4). The diatom concentration was higher than in zone 2.

Acidophilus diatoms increased up to >82% in TFL3, and Pinnularia
gibba (721%) appeared more frequently in this zone. Eunitia praerupta (611%) was the main species of circumneutral diatoms
(Fig. 5). Somewhat, the estimated DI-pH was somehow lower than
in zone TFL2, and the average value was 5.44 (Fig. 6).
The pollen concentration in zone TFL4 (6940 cm, 1100560 cal
BP) remained high compared with previous zones. The concentration of LMF pollen was greater than that for UMF pollen. Haloragis

(27%) was seen more frequently among the terrestrial herbs. The
percentages of the wetland pollen Cyperaceae (913%) were lower
than in TFL3 (Fig. 4). The diatom concentration increased gradually
and reached a maximum at the end of the zone. The acidophilus
diatom Aulacoseira distance (6084%) was the most abundant species (Fig. 5). The average estimated DI-pH increased slightly to 5.57
(Fig. 6).
The pollen concentration in zone TFL5 (400 cm, 560 to 54 cal
BP) remained high. UMF pollen showed an increased trend with an

452

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

Fig. 6. Fluctuations of PCA derived rst axis based on pollen data (pollen-PC1), diatom-inferred pH (DI-pH), total organic carbon (TOC), C/N ratio, d13C, and d15N with
respecting to pollen zones (TFL). Dashed lines represent the mean value of each proxy.

abrupt decrease above 10 cm of core depth (150 cal BP). Pollen of

the cold-resistant conifers Tsuga (19%) and Pinus (312%) were
more abundant in this zone, while deciduous Quercus (1424%)
declined. Among the herbs, Haloragis pollen (213%) increased
gradually and peaked at a depth of 18 cm (240 cal BP). Fern
spores (1729%) increased slightly. The green algae, Botryococcus
(19%), had the highest occurrence in the uppermost samples
(Fig. 4). This zone was also characterized by higher diatom concentrations, with alkaliphious diatoms, including Encyonopsis microcephala (341%) and Pseudostaurosira elliptica (15%) dominating
several depths. Occurrences of circumneutral diatoms were relatively frequent in this zone (Fig. 5), and the calculated DI-pH was
much higher than TFL4 with values between 5.85 and 6.26 (Fig. 6).
4.3. Ordination of the palynological records
Ordination results of the principle component analysis (PCA)
conrmed that strong gradients along the rst two axes explained
35% of total variance within the pollen data (Fig. 6). The terrestrial
herb Haloragis and some rare taxa, such as Myrica, were located on
the positive side of PCA axis 1(PC1). MF and LMF taxa, especially
Alnus, Chamaecyparis, Mallotus were found on the positive side of
PCA axis 2 (PC2), while the conifer taxa which grow in higher latitude regions, such as Tsuga and Pinus, were located on the negative
side of PC2.
The geochemical proxies (TOC, TN, C/N ratio, d13C, d15N)
(Selvaraj et al., 2012) and palynological data (wetland pollen, fern
spores) were plotted along two PCA axes as supplementary data
(Fig. 6). Supplementary variables TN, TOC, C/N ratio, d13C were
plotted on the right side of the PC1, indicating positive correlation,
while the percentages of wetland pollen and fern spores along with
d15N were plotted on the left side of PC1, indicating a negative
correlation.
5. Interpretation and discussion
5.1. Pollen-PCA ordinations
The environmental inference of PC1 is interpreted to reveal the
major factors affecting the pollen assemblages. PC1 has an

eigenvalue of 0.23 with a strong positive correlation with

Haloragis (Fig. 7). H. micrantha is the only known Haloragis species
in Taiwan, and is widely distributed at low to medium altitudes
(Huang, 1994). Therefore, Haloragis was taken as an indicator for
wetter areas (Black et al., 2006), and the pollen was well preserved
in the surface sediments in seasonally submerged ponds (Chen
et al., 2009). During the wet season (high lake level period),
Haloragis (Poaceae zone) has a higher coverage ratio in TFL basin.
Conversely, the coverage ratio of wetlands (aquatic zones) was
reduced (Fig. 2). Thus, we assumed PC1 as an indicator of lake level,
reecting precipitation within the study area. The negative correlation between PC1 and wetland pollen supports this assumption.
PC2 had a relatively low eigenvalue of 0.12 showing strong
negative correlated with Tsuga and Pinus, which were the
dominant taxa of UMF. Thus, PC2 may reect the expansion of
high-latitude forest, correlate to the variations in temperature.
5.2. Sedimentary proxy tracked paleoenvironmental history
The microfossil proxies of both allochthonous (pollen) and
autochthonous (diatom) origins associated with geochemical proxies (Selvaraj et al., 2012) were used for paleoenvironmental reconstruction of the study site (Fig. 6). The proxies reected signals of
local hydrological changes and regional climatic patterns in northeastern Taiwan during the past 3500 years.
The 35002930 cal BP period (zone TFL1) was characterized by
the dominance of wetland pollen and sandy sediments. The coarse
sediments consisted mainly of rock fragments and soils washed in
from the catchment. This interpretation was inferred from high
d15N, high d13C, low C/N ratio, and high percentages of fern spores
(Selvaraj et al., 2012; Wang et al., 2013). The diluted microfossil
and total organic carbon (TOC) concentrations from the record also
supported this nding. For a small lake situated in a closed basin,
the lake level variations can be inferred from the changes in sediment composition and the littoral vegetation distributions reconstructed from pollen analysis. For example, when the lake level is
lower, littoral vegetation and the sandy layer move toward the lake
center (Dearing, 1997). Thus, the sandy sediments accompany the
wetland pollen can indicate a shallower lake TFL landscape, and
reects drier climate conditions. The cold-resistant conifers Tsuga

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

453

Fig. 7. This diagram presents PCA results for 41 samples of pollen data. Red arrows on the left denote the 30 pollen taxa with a minimum value of >1% along with the rst axis
(PC1) and the second axis (PC2). Please note that only the names of major pollen taxa (>2% in at least one samples for each species) are shown. Sediment samples and
supplementary data (black arrows) along with two PCA axes are also presented. (For interpretation of the references to color in this gure legend, the reader is referred to the
web version of this article.)

and Pinus show higher percentages during this period, indicating a

dry and cold climate.
For the period of 29302030 cal BP (zone TFL2), better microfossil preservation and higher TOC than in zone TFL1 suggested less
dilution caused by inorganic input. The higher C/N ratio, lower
d15N, and lower d13C also implied a reduction in mineral content
in the sediments. However, the higher percentage of the circumneutral diatom Pinnularia viridis, which dominates mineral-rich
water (Wang et al., 2010), suggested that littoral rock fragment
input from the watershed might have increased and can reect a
low lake level for TFL during this period. Rapid uctuations of pollen-PC1 indicated frequent changes in lake level, with agreement
from the relatively higher percentages of the terrestrial/freshwater
diatom Frustulia crassinervia (Wu et al., 2011). The frequent
changes in lake level may correlate to increased typhoon frequency, which is revealed by the consistency of the timing of terrace deposition in the Lanyang and Erhjen drainage basins (Liew
and Hsieh, 2000). The UMF pollen decreased, suggesting that the
temperature was not as cold as in the previous period. The climate
was cool and dry, but extreme precipitation events occurred.
The important characteristics of the following period, 2030
1100 cal BP (zone TFL3), were the formation of mud deposits and
the gradual decrease in Cyperaceae pollen associated with higher
pollen-PC1. All the results pointed to higher lake levels during this
period. The higher microfossil concentrations and increased TOC
and C/N ratios all indicated lower littoral rock fragment inputs
because of the higher lake level. The increase in the acidophilus
diatom Pinnularia gibba, which favors freshwater with low mineral
content, also conrmed this assumption. Compared to the previous
period, consistently low DI-pH values reect stable hydrological
conditions, while the higher LMF implied a warm climate.
The characteristic between 1100 and 560 cal BP (zone TFL4,
corresponding to the (MWP)) is the rapid increase in pollen-PC1.
This indicates a marked rise in lake levels, and reects intense precipitation. Although the intense precipitation was also in agreement with the slight increase in estimated DI-pH, an increase in
a shade-resistant species of Quercus suggests a stable environment.
When combined with higher LMF pollen, we conclude that the climate was warmer and humid with stable precipitation rates.
Between 560 and 54 cal BP (zone TFL5, corresponding to a
period since LIA toward present) was characterized by both high
pollen-PC1 and DI-pH values. Based on the TFL lacustrine sediment
records, Wang et al. (2013) found an increase in DI-pH values along
with low TOC, high d13C, and low d15N during the LIA, indicating

intense soil input and erosion in relation to frequent typhoons.

Homogeneous changes in sedimentary data in our study enhanced
the reliability of the inferred environmental changes, suggesting
extreme precipitation in this interval. The increased UMF pollen
indicated that climate conditions during the LIA were cold and
humid. Since 130 cal BP, pollen-PC1 and DI-pH values declined
markedly, which can be related to anthropogenic activities, including deforestation of the near mountains since 1906 (Lin, 1996) and
acid rain (Wang et al., 2013). The high abundance of the green
microalgae Botryococcus in the uppermost sediment indicated lake
eutrophication, reecting intense human impacts during the past
century (Zhao et al., 2007).
5.3. Long-term climate pattern and abrupt hydrological oscillations
We used the UMF and LMF ratio, pollen-PC1, and DI-pH proxies
to indicate temperature, rainfall, and typhoon activities to compare
with the main climate forces of East Asia, including EASM (Hong
et al., 2005; Ji et al., 2005; Wang et al., 1999), SST of WPWP
(Stott et al., 2004) and ENSO events (Moy et al., 2002) (Fig. 8).
Multiple proxy data inferred a cold-to-cool and dry climate in
TFL during 37002040 cal BP, corresponding to a decrease in monsoonal precipitation, which was supported by numerous long-term
climate records in Taiwan. For example, the sedimentation hiatus
in Retreat Lake during 45002100 cal BP (Selvaraj et al., 2011,
2007; Wang et al., 2014a), and higher percentages of Poaceae pollen in Yuanyang Lake during 38002300 cal BP (Chen and Wu,
1999), indicating a reduction of moisture in northeastern Taiwan.
In central Taiwan, a drier climatic scenario is indicated by the
dominance of Tsuga and Pinus pollen in Chitsai Lake during
37302030 cal BP (Liew and Huang, 1994), and by the charcoal
deposits in an outcrop in Hohuanshan Mountain with radiocarbon
ages of 3770 and 2400 cal BP (Wenske et al., 2011).
The paleotemperature record of the past 2000 years matched
well with the global surface temperature anomalies, namely a warmer MWP, and a cooler LIA period (Fig. 8), while the paleo-precipitation increased gradually, especially during the past thousand
years. The climate interpretation inferred from the TFL record
inferred is conrmed by the climate simulation from twelve
palynological records in northern Taiwan (Wang et al., 2011). The
increased monsoonal rainfall during the late Holocene, although
against the generally weakening Holocene trend, was also supported by several EASM proxies (Fig. 8). Wang et al. (1999) found
decline of sea surface salinity at the northern margin of the

454

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

Fig. 8. Summary for the sedimentary proxies inferred paleotemperature (ratio of upper montane forest and lower montane forest, UMF/LMF, pollen-PC2), precipitation
(pollen-PC1), and typhoon activities (DI-pH) in TFL comparing to the proxies of major climate forces of East Asia (Fig. 1), including East Asian summer monsoon (EASM), sea
surface temperature (SST) of western Pacic Warm Pool (WPWP) (Stott et al., 2004), and El Nio-Southern Oscillation (ENSO) intensity (Moy et al., 2002). Records associated
with EASM intensity are sea surface salinity (SSS) of marine sediment core 17,940 (Wang et al., 1999), d13C of plant cellulose of Hani peat bog (Hong et al., 2005), and sediment
redness of Lake Qinghai (Ji et al., 2005). Thick red lines are smooth curves by using locally weighted scatterplot smoothing method in C2 program version 1.5 for Windows
(Juggins, 2005). Please note that the x-axis is reversed for UMF/LMF, 17,940 SSS, and Hani peat d13C. Right column shows the late Holocene climate conditions in associating
with periods of global temperature changes, including the MWP, and the LIA inferred by multi-proxies records of core TFL-1 (Chen et al., 2015; Yin et al., 2010). (For
interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

South China Sea, indicating events of enhanced precipitation during the past 2000 years. During this period, the smaller d13C of peat
cellulose at Hani peat bog (Hong et al., 2005) and increased sediment redness at Lake Qinghai (Ji et al., 2005) also indicated the
stronger EASM activity or the moister climate in the monsoon
region. The reverse trend in EASM during the past 2000 years
might be linked to the movement of the position of ITCZ and
ENSO variability, which were forced by changes in austral summer
insolation and increases of greenhouse gas (Zhao et al., 2013).
Typhoon activities strongly affect regional and local hydrological patterns in East Asia. The paleo-precipitation record seems
positively correlated to the variations in the SST of WPWP and
ENSO intensity, except for the periods with frequent typhoon
activities, which were inferred by higher DI-pH (Fig. 8). SSTs of
WPWP and ENSO intensities control the intensity, frequency, and
tracks of typhoons originating from WPWP (Wang and Chan,
2002). Based on recent observational data, the warm SST anomalies
over the equatorial western and central Pacic tend to induce
typhoon enhancement, namely, the weakening of the subtropical
high, strengthening of the EASM, and enhanced low-level vorticity
(Tu et al., 2009). Therefore, we suggest that the increase in precipitation in northeastern Taiwan during the past 2000 years can also
be associated with enhanced typhoon activities in the western
Pacic due to the warmer SST of WPWP and coupled with intensied ENSO events.
Based on grain size records in lake sediments deposited in ood
events from the Ilan Plain in NE Taiwan, Chen et al. (2012) stated
that super typhoons chiey occurred in Taiwan during the La
Nia-like stage because typhoon tracks are highly inuenced by
the ENSO. Yin et al. (2010) also found that more typhoons affected
southern China during La Nia years even though the amount of
heavy rain days decreased. According to recent meteorological
records, the westward shift in typhoon tracks is relative to the
shrinking of WPWP and expansion of the Northwestern Pacic
High, which pushes typhoons westerly toward southern China
and Taiwan (Ho et al., 2004). Therefore, the frequent typhoons
and abrupt precipitation anomalies in NE Taiwan during

29302030 cal BP and the LIA may correlate to a track shift in

typhoon activities, causing an asymmetric distribution in local precipitation, superimposed onto the regional hydrological pattern.
6. Conclusions
The past 3500 year pollen and diatom records derived from
Tsuifong Lake (TFL) sediments in northeastern Taiwan revealed
clearly that the climate in NE Taiwan is strongly linked to variations with the EASM, SST of WPWP, and ENSO intensity. The TFL
results support the following conclusions:
1. During 35002030 cal BP, the low lake level referred proxy values (pollen-PC1) and the dominance of Cypeaceae indicated a
shallow lake level with widespread sedges, and reected a dry
climatic condition and a weakening of the EASM. This result
agrees with all pollen records from subtropical Taiwan.
2. The variations between UMF and LMF indicate that the uctuated paleotemperature matches well with global climate
events, including the MWP and the LIA. The wetter climatic
conditions recovered during the late Holocene, possibly caused
by the reversed EASM strength and enhanced typhoon activities
of the past 2000 years.
3. During 29302030 cal BP and the LIA, frequent typhoon events
were observed inferred from high DI-pH values. The abrupt precipitation changes might be linked to the cooler SST of WPWP
and weakening ENSO, which favored the western movement
of typhoon tracks toward Taiwan and southern China.

Acknowledgments
The authors appreciate the nancial support provided by a
Grant from the National Science Council (NSC 100-2611-M-110011-, 101-2611-M-110-003-, 102-2917-I-564-058), Ministry of
Science and Technology (MOST 103-2811-M-194-010), the Open

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

Fund of Key Laboratory of Chinese Academy of Geological Sciences

(KERDC201301). This research was also funded by Asia-Pacic
Ocean Research Center locates in National Sun Yat-sen
University, Taiwan. The constructive commends contributed by
two anonymous reviewers are also appreciated.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jseaes.2015.03.
037. These data include Google maps of the most important areas
described in this article.
References
Abdi, H., Williams, L.J., 2010. Principal component analysis. Wiley Interdiscipl. Rev.:
Comput. Stat. 2, 433459. http://dx.doi.org/10.1002/wics.101.
Ammann, B., Birks, H.J.B., Brooks, S.J., Eicher, U., von Grafenstein, U., Hofmann, W.,
Lemdahl, G., Schwander, J., Tobolski, K., Wick, L., 2000. Quantication of biotic
responses to rapid climatic changes around the Younger Dryas a synthesis.
Palaeogeogr., Palaeoclimatol., Palaeoecol. 159, 313347.
Birks, H.J.B., Birks, H.J.B., 1998. D.G. Frey and E.S. Deevey review 1: numerical tools
in palaeolimnology progress, potentialities, and problems. J. Paleolimnol. 20,
307332. http://dx.doi.org/10.1023/A:1008038808690.
Blaauw, M., Christen, J.A., 2011. Flexible paleoclimate age-depth models using an
autoregressive gamma process. Bayesian Anal. 6, 457474. http://dx.doi.org/
10.1214/ba/1339616472.
Black, M.P., Mooney, S.D., Martin, H.A., 2006. A >43,000-year vegetation and re
history from Lake Baraba, New South Wales, Australia. Quatern. Sci. Rev. 25,
30033016. http://dx.doi.org/10.1016/j.quascirev.2006.04.006.
Cao, Y., Zhang, E., Langdon, P.G., Liu, E., Shen, J., 2014. Chironomid-inferred
environmental change over the past 1400 years in the shallow, eutrophic Taibai
Lake (south-east China): separating impacts of climate and human activity. The
Holocene 24, 581590. http://dx.doi.org/10.1177/0959683614522308.
Charles, D.F., 1985. Relationships between surface sediment diatom assemblages
and lake water characteristics in Adirondack Lakes. Ecology 66, 9941011.
Chen, H.-F., Wen, S.-Y., Song, S.-R., Yang, T.-N., Lee, T.-Q., Lin, S.-F., Hsu, S.-C., Wei, K.Y., Chang, P.-Y., Yu, P.-S., 2012. Strengthening of paleo-typhoon and autumn
rainfall in Taiwan corresponding to the Southern Oscillation at late Holocene. J.
Quatern. Sci. 27, 964972. http://dx.doi.org/10.1002/jqs.2590.
Chen, J., Chen, F., Feng, S., Huang, W., Liu, J., Zhou, A., 2015. Hydroclimatic changes in
China and surroundings during the medieval climate anomaly and little ice age:
spatial patterns and possible mechanisms. Quatern. Sci. Rev. 107, 98111.
http://dx.doi.org/10.1016/j.quascirev.2014.10.012.
Chen, S.-H., Wu, J.-T., 1999. Paleolimnological environment indicated by the diatom
and pollen assemblages in an alpine lake in Taiwan. J. Paleolimnol. 22, 149158.
Chen, S.-H., Wu, J.-T., Yang, T.-N., Chuang, P.-P., Huang, S.-Y., Wang, Y.-S., 2009. Late
Holocene paleoenvironmental changes in subtropical Taiwan inferred from
pollen and diatoms in lake sediments. J. Paleolimnol. 41, 315327.
Dearing, J.A., 1997. Sedimentary indicators of lake-level changes in the humid
temperate zone: a critical review. J. Paleolimnol. 18, 114. http://dx.doi.org/
10.1023/A:1007916210820.
Friel, C.E., Finkelstein, S.A., Davis, A.M., 2014. Relative importance of hydrological
and climatic controls on Holocene paleoenvironments inferred using diatom
and pollen records from a lake in the central Hudson Bay Lowlands, Canada. The
Holocene 24, 295306. http://dx.doi.org/10.1177/0959683613518587.
Goldsmith, S.T., Carey, A.E., Lyons, W.B., Kao, S.-J., Lee, T.-Y., Chen, J., 2008. Extreme
storm events, landscape denudation, and carbon sequestration: Typhoon
Mindulle, Choshui River, Taiwan. Geology 36, 483486. http://dx.doi.org/
10.1130/G24624A.1.
Grimm, E., 1993. TILIA: A Pollen Program for Analysis and Display. Illinois State
Museum, Springeld, IL.
Ho, C.-H., Baik, J.-J., Kim, J.-H., Gong, D.-Y., Sui, C.-H., 2004. Interdecadal changes in
summertime typhoon tracks. J. Clim. 17, 17671776. http://dx.doi.org/10.1175/
1520-0442(2004)017<1767:ICISTT>2.0.CO;2.
Hong, Y.T., Hong, B., Lin, Q.H., Shibata, Y., Hirota, M., Zhu, Y.X., Leng, X.T., Wang, Y.,
Wang, H., Yi, L., 2005. Inverse phase oscillations between the East Asian and
Indian Ocean summer monsoons during the last 12 000 years and paleo-El
Nio. Earth Planet. Sci. Lett. 231, 337346. http://dx.doi.org/10.1016/
j.epsl.2004.12.025.
Huang, T.-C., 1972. Pollen Flora of Taiwan. National Taiwan University Botany
Department Press, Taipei, p 297.
Huang, T.-C., 1981. Spore Flora of Taiwan. National Taiwan University Botany
Department Press, Taipei, p. 111.
Huang, T.-C., 1994. Flora of Taiwan. National Taiwan University Botany Department
Press, Taipei, p. 654.
Hurrell, J.W., 1995. Decadal trends in the North Atlantic Oscillation: regional
temperatures and precipitation. Science 269, 676679. http://dx.doi.org/
10.1126/science.269.5224.676.
IPCC, 2013. Climate change 2013: the physical science basis. In: Stocker, T.F., Qin, D.,
Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V.,

455

Midgley, P.M. (Eds.), Contribution of Working Group I to the Fifth Assessment

Report of the Intergovernmental Panel on Climate Change. Cambridge
University Press, Cambridge, United Kingdom and New York.
Ito, A., 2010. Evaluation of the impacts of defoliation by tropical cyclones on a
Japanese forests carbon budget using ux data and a process-based model. J.
Geophys.
Res.:
Biogeosci.
115,
G04013.
http://dx.doi.org/10.1029/
2010JG001314.
Jantz, N., Behling, H., 2012. A Holocene environmental record reecting vegetation,
climate, and re variability at the Pramo of Quimsacocha, southwestern
Ecuadorian Andes. Veg. Hist. Archaeobot. 21, 169185. http://dx.doi.org/
10.1007/s00334-011-0327-x.
Ji, J., Shen, J., Balsam, W., Chen, J., Liu, L., Liu, X., 2005. Asian monsoon oscillations in
the northeastern Qinghai-Tibet Plateau since the late glacial as interpreted from
visible reectance of Qinghai Lake sediments. Earth Planet. Sci. Lett. 233, 6170.
http://dx.doi.org/10.1016/j.epsl.2005.02.025.
Juggins, S., 2001. The European Diatom Database User Guide, Version 1.0. University
of Newcastle, UK.
Juggins, S., 2005. C2 Version 1.5: Software for Ecological and Palaeoecological Data
Analysis and Visualisation. University of Newcastle, Newcastle.
Kao, S.-J., Liu, K.-K., 1997. Fluxes of dissolved and nonfossil particulate organic
carbon from an Oceania small river (Lanyang Hsi) in Taiwan. Biogeochemistry
39, 255269. http://dx.doi.org/10.1023/A:1005864605382.
Krammer, K., Lange-Bertalot, H., 1986. Sswasserora von Mitteleuropa.:
Bacillariophyceae, pp. 14.
Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A.C.M., Levrard, B., 2004. A
long-term numerical solution for the insolation quantities of the Earth. Astron.
Astrophys. 428, 261285. http://dx.doi.org/10.1051/0004-6361:20041335.
Lee, C.-Y., Liew, P.-M., 2010. Late quaternary vegetation and climate changes
inferred from a pollen record of Dongyuan Lake in southern Taiwan.
Palaeogeogr., Palaeoclimatol., Palaeoecol. 287, 5866. http://dx.doi.org/
10.1016/j.palaeo.2010.01.015.
Liew, P.-M., Hsieh, M.-L., 2000. Late Holocene (2 ka) sea level, river discharge and
climate interrelationship in the Taiwan region. J. Asian Earth Sci. 18, 499505.
http://dx.doi.org/10.1016/S1367-9120(99)00081-4.
Liew, P.-M., Huang, S.-Y., 1994. A 5000 year pollen record from Chitsai Lake in
Central Taiwan. Terr., Atmos. Ocean. Sci. 5, 411420.
Liew, P.-M., Lee, C.-Y., Kuo, C.-M., 2006. Holocene thermal optimal and climate
variability of East Asian monsoon inferred from forest reconstruction of a
subalpine pollen sequence, Taiwan. Earth Planet. Sci. Lett. 250, 596605. http://
dx.doi.org/10.1016/j.epsl.2006.08.002.
Liew, P.-M., Wu, M.-H., Lee, C.-Y., Chang, C.-L., Lee, T.-Q., 2014. Recent 4000 years of
climatic trends based on pollen records from lakes and a bog in Taiwan.
Quatern. Int. 105112. http://dx.doi.org/10.1016/j.quaint.2014.05.018.
Lin, I.-I., Pun, I.-F., Wu, C.-C., 2009. Upper-ocean thermal structure and the western
North Pacic category 5 Typhoons. Part II: dependence on translation speed.
Mon.
Weather
Rev.
137,
37443757.
http://dx.doi.org/10.1175/
2009MWR2713.1.
Lin, Q.C., 1996. The Development History of Taiping Mountain. Fulunxiaozhu, Ilan
County (in Chinese).
Mann, M.E., Zhang, Z., Rutherford, S., Bradley, R.S., Hughes, M.K., Shindell, D.,
Ammann, C., Faluvegi, G., Ni, F., 2009. Global signatures and dynamical origins
of the little ice age and medieval climate anomaly. Science 326, 12561260.
http://dx.doi.org/10.1126/science.1177303.
Mao, J.-J., 2006. Investigation on the biological resources and Habitat of Tsuei-Feng
Lake. Research and Conservation of Forestry Bureau No. 94-10, p. 51 (in
Chinese).
Moy, C.M., Seltzer, G.O., Rodbell, D.T., Anderson, D.M., 2002. Variability of El Nio/
Southern Oscillation activity at millennial timescales during the Holocene
epoch. Nature 420, 162165. http://dx.doi.org/10.1038/nature01194.
Reimer, P., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0
50,000 years cal BP. Radiocarbon 55, 18691887. http://dx.doi.org/10.2458/
azu_js_rc.55.16947.
Selvaraj, K., Arthur Chen, C.-T., Lou, J.-Y., Kotlia, B.S., 2011. Holocene weak summer
East Asian monsoon intervals in Taiwan and plausible mechanisms. Quatern.
Int. 229, 5766. http://dx.doi.org/10.1016/j.quaint.2010.01.015.
Selvaraj, K., Chen, C.T., Lou, J.Y., 2007. Holocene East Asian monsoon variability:
links to solar and tropical Pacic forcing. Geophys. Res. Lett. 34, L01703.
Selvaraj, K., Wei, K.-Y., Liu, K.-K., Kao, S.-J., 2012. Late Holocene monsoon climate of
northeastern Taiwan inferred from elemental (C, N) and isotopic (d13C, d15N)
data in lake sediments. Quatern. Sci. Rev. 37, 4860. http://dx.doi.org/10.1016/
j.quascirev.2012.01.009.
Stott, L., Cannariato, K., Thunell, R., Haug, G.H., Koutavas, A., Lund, S., 2004. Decline
of surface temperature and salinity in the western tropical Pacic Ocean in the
Holocene epoch. Nature 431, 5659. http://dx.doi.org/10.1038/nature02903.
ter Braak, C.J.F., Smilauer, P., 2002. Canoco 4.5: Reference Manual and Canodraw for
Windows. Users Guide: Software Form Canonical Community Ordination
(version 4.5). Microcomputer Power, Ithaca, New York.
Tinner, W., Lotter, A.F., 2001. Central European vegetation response to abrupt
climate change at 8.2 ka. Geology 29, 551554. http://dx.doi.org/10.1130/00917613(2001)029<0551:CEVRTA>2.0.CO;2.
Tsou, C.-Y., Feng, Z.-Y., Chigira, M., 2011. Catastrophic landslide induced by Typhoon
Morakot, Shiaolin, Taiwan. Geomorphology 127, 166178. http://dx.doi.org/
10.1016/j.geomorph.2010.12.013.
Tu, J.-Y., Chou, C., Chu, P.-S., 2009. The abrupt shift of typhoon activity in the vicinity
of Taiwan and its association with western North Pacic-East Asian climate
change. J. Clim. 22, 36173628. http://dx.doi.org/10.1175/2009JCLI2411.1.

456

L.-C. Wang et al. / Journal of Asian Earth Sciences 114 (2015) 447456

Wang, B., Chan, J.C.L., 2002. How strong ENSO events affect tropical storm activity
over the western North Pacic. J. Clim. 15, 16431658. http://dx.doi.org/
10.1175/1520-0442(2002)015<1643:HSEEAT>2.0.CO;2.
Wang, L.-C., Behling, H., Chen, Y.-M., Huang, M.-S., Chen, A.C.-T., Lou, J.-Y., Chang, Y.-P.,
Li, H.-C., 2014a. Holocene monsoonal climate changes tracked by multiproxy
approach from a lacustrine sediment core of the subalpine Retreat Lake in Taiwan.
Quatern. Int. 333, 6976. http://dx.doi.org/10.1016/j.quaint.2014.02.014.
Wang, L.-C., Behling, H., Lee, T.-Q., Li, H.-C., Huh, C.-A., Shiau, L.-J., Chang, Y.-P.,
2014b. Late Holocene environmental reconstructions and their implications on
ood events, typhoon, and agricultural activities in NE Taiwan. Clim. Past 10,
18571869. http://dx.doi.org/10.5194/cp-10-1857-2014.
Wang, L.-C., Behling, H., Lee, T.-Q., Li, H.-C., Huh, C.-A., Shiau, L.-J., Chen, S.-H., Wu, J.T., 2013. Increased precipitation during the little ice age in northern Taiwan
inferred from diatoms and geochemistry in a sediment core from a subalpine
lake. J. Paleolimnol. 49, 619631. http://dx.doi.org/10.1007/s10933-013-9679-9.
Wang, L.-C., Lee, T.-Q., Chen, S.-H., Wu, J.-T., 2010. Diatoms in Liyu Lake, Eastern
Taiwan. Taiwania 55, 228242.
Wang, L.-C., Wu, J.-T., Lee, T.-Q., Lee, P.-F., Chen, S.-H., 2011. Climate changes
inferred from integrated multi-site pollen data in northern Taiwan. J. Asian
Earth Sci. 40, 11641170. http://dx.doi.org/10.1016/j.jseaes.2010.06.003.
Wang, L., Sarnthein, M., Erlenkeuser, H., Grootes, P.M., Grimalt, J.O., Pelejero, C.,
Linck, G., 1999. Holocene variations in Asian monsoon moisture: a bidecadal
sediment record from the South China Sea. Geophys. Res. Lett. 26, 28892892.

Wenske, D., Bse, M., Frechen, M., Lthgens, C., 2011. Late Holocene mobilisation of
loess-like sediments in Hohuan Shan, high mountains of Taiwan. Quatern. Int.
234, 174181. http://dx.doi.org/10.1016/j.quaint.2009.10.034.
Wu, J.-T., Babu, B., Chou, C.-L., Saraswathi, S.J., 2011. Freshwater Diatom Flora of
Taiwan. Biodiversity Research Center, Academia Sinica, Taipei, Taiwan, p. 747.
Yin, Y., Gemmer, M., Luo, Y., Wang, Y., 2010. Tropical cyclones and heavy rainfall in
Fujian Province, China. Quatern. Int. 226, 122128. http://dx.doi.org/10.1016/
j.quaint.2010.03.015.
Yuan, D., Cheng, H., Edwards, R.L., Dykoski, C.A., Kelly, M.J., Zhang, M., Qing, J., Lin, Y.,
Wang, Y., Wu, J., Dorale, J.A., An, Z., Cai, Y., 2004. Timing, duration, and
transitions of the last interglacial Asian monsoon. Science 304, 575578. http://
dx.doi.org/10.1126/science.1091220.
Zhang, J., Chen, F., Holmes, J.A., Li, H., Guo, X., Wang, J., Li, S., L, Y., Zhao, Y., Qiang,
M., 2011. Holocene monsoon climate documented by oxygen and carbon
isotopes from lake sediments and peat bogs in China: a review and synthesis.
Quatern. Sci. Rev. 30, 19731987. http://dx.doi.org/10.1016/j.quascirev.2011.04.
023.
Zhao, C., Chang, Y.-P., Chen, M.-T., Liu, Z., 2013. Possible reverse trend in Asian
summer monsoon strength during the late Holocene. J. Asian Earth Sci. 69, 102
112. http://dx.doi.org/10.1016/j.jseaes.2012.09.028.
Zhao, Y., Yu, Z., Chen, F., Ito, E., Zhao, C., 2007. Holocene vegetation and climate
history at Hurleg Lake in the Qaidam Basin, northwest China. Rev. Palaeobot.
Palynol. 145, 275288.