Study of modern pollen distribution in the

Review of Palaeobotany and Palynology 256 (2018) 50–62

Contents lists available at ScienceDirect

Review of Palaeobotany and Palynology journal homepage: www.elsevier.com/locate/revpalbo

Study of modern pollen distribution in the northeastern Indian Ocean and their application to paleoenvironment reconstruction Chuanxiu Luo a,⁎, Chixin Chen b, Rong Xiang a,⁎, Weiming Jiang c,d, Jianguo Liu a, Jun Lu a, Xiang Su a, Qiang Zhang a, Yiping Yang a, Mingxi Yang e a

CAS Key Laboratory of Ocean and Marginal Sea Geology, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou, China Guangzhou Marine Geological Survey, Guangzhou, China School of Geographical Sciences of Guangzhou University, Guangzhou, China d School of Earth and Ocean Sciences, University of Victoria, Victoria, BC V8W 2Y2, Canada e Department of Biology, University of Kentucky, Lexington, KY 40506, USA b c

a r t i c l e

i n f o

Article history: Received 4 February 2018 Received in revised form 22 May 2018 Accepted 29 May 2018 Available online 02 June 2018 Keywords: Pollen Northeastern Indian Ocean Pollen terrestrial area Climate Sea level change

a b s t r a c t Learning pollen transport mechanisms is the basis for correct interpretation of fossil pollen data. Firstly, 33 seabed surface sediment samples from the northeastern Indian Ocean were selected for pollen analysis. Utilizing the 90°E Indian Ridge as a dividing line, more Trilete spores (mainly Microlepia)and herbaceous pollen were present to the west or northwest of the line, possibly originating in Sri Lanka and India. More Polypodiaceae spores and tree pollen (Pinus excluded) were present to the east or southeast of the boundary, possibly originating in Sumatra Island. Based on the PCA analysis, the ocean current from the southeast to the northwest and the distance from land masses around the study area affect the percentage of pollen. Then, based on a fossil pollen analysis on the YDY05 core from the northeastern Indian Ocean, we determined that since 16 ka BP, the pollen terrestrial area might change from Sumatra Island to Sri Lanka Island and India, probably due to the climate and sea level change. © 2018 Elsevier B.V. All rights reserved.

1. Introduction Seabed sediments in the northeastern part of the Indian Ocean have preserved comprehensive information of the typical tropical oceans and ancient marine environment chronicling South Asian monsoon evolution. This area provides a uniquely important opportunity for the high-resolution study of global climate change. Pollen is from terrestrial vegetation, the occurrence of pollen in marine sediments reflects the vegetation ecology and climatic conditions on the adjacent terrain. Therefore, conducting research on modern pollen dissemination in the northeastern Indian Ocean and applying it in Quaternary environmental reconstruction work serves three purposes; to reveal the characteristics of pollen distribution under a variety of dynamic effects in the study area, to reconstruct ancient terrestrial vegetation succession in the surrounding area, and to track the evolution of ancient historical processes of the marine environment. The distribution and dissemination of pollen in the northeastern Indian Ocean are closely related to its environmental conditions of hydrothermal power. The study area of the present study is located near Southeast Asia; therefore, it is possible to collect abundant pollen from ⁎ Corresponding authors. E-mail addresses: [email protected], (C. Luo), [email protected] (R. Xiang).

https://doi.org/10.1016/j.revpalbo.2018.05.007 0034-6667/© 2018 Elsevier B.V. All rights reserved.

South Asia, Indochina, and the Indonesian islands. The terrestrial pollen feature of South and Southeast Asia were studied previously (Leipe et al., 2014a, 2014b; Liang et al., 2015; Riedel et al., 2015; Chauhan et al., 2015; Rawat et al., 2015; Demske et al., 2009; Premathilake and Seneviratne, 2015). The northeastern Indian Ocean is the area of strongest material exchange and most active deposition in the world today, it can provide experimental conditions of rare tropical deep-sea area edge for the study of ancient climate change. Learning pollen transport mechanisms are the basis for correct interpretation of pollen data from the perspective of the ancient environment. This is even more important for studies on pollen in marine sediments due to the highly complex controlling factors of pollen distribution, such as ocean currents, atmospheric circulation, sediment type, as well as the mechanisms that allowed pollen grains to sink to the surface sediment in the water column, the large seabed topography, and micro landforms. Therefore, we first need to study the process of distribution and dissemination of modern marine pollen. In recent years in the Western Pacific and the South China Sea, research on pollen in marine surface sediments has been gradually conducted; however, no similar studies have been conducted in the northeast Indian Ocean. From the beginning of the 1990s, some studies have discussed the evolutionary role of the Indian monsoon and paleoclimate using foraminiferans and other indicators in the Indian

C. Luo et al. / Review of Palaeobotany and Palynology 256 (2018) 50–62

Ocean (Murgese and Deckker, 2005). John et al. (1992) discussed the composition of the foraminifera fauna group in marine sediments controlled by the southwest monsoon in the northwest Arabian Sea using wind, hydro, and marine remote sensing data. Clemens et al. (1991, 1996) reconstructed the strength variance of the Indian summer monsoon in evolutionary history from 530 million years ago, using the abundance of the foraminifera, Globigerina bulloides as an indicator of upwelling in the Arabian Sea, combined with deep-sea oxygen isotopes. Prell (1984a, 1984b) studied Indian summer monsoon change on the scale of orbit in the northern Arabian Sea, using G. bulloides as an indicator of upwelling for time series analysis. In addition to the studies on foraminifera, research of radiolarians (Gupta, 2003) and Dinoflagellate in the northeastern Indian Ocean has been conducted (Marret and Vernal, 1997); however, few studies have focused on pollen present in drilling cores (Sander and Patrick, 2003). The longest pollen record spans the last 200 thousand years in the neighboring Arabian Sea and the Indian subcontinent, with a resolution of 5700 years. When combined with oxygen isotope stratigraphy, it is evident that the climate in marine isotope stages (MIS) 2, 4, and 6 is cold and dry with weakening summer wind and stronger winter winds, while the climate in MIS 1, 3, and 5 is warm and humid (Prabhu et al., 2004) with increased precipitation brought by an enhanced summer monsoon. The remaining drilling cores in the Arabian Sea are almost not more than twenty thousand years of age. Pollen deposition research of these cores shows that there were major changes in monsoon intensity occurred during the last glacial–interglacial cycles (Bentaleb et al., 1997). Pollen results of the drilling core SO90-56KA from the Makran coastal area in the Arabian Sea indicated that the end of the Holocene wet period occurred between 4700 and 4200 BP, correlating with the weakening of the Indian monsoon flux (Ivory and Lézine, 2009). However, in previous studies in the Indian Ocean, conclusions on pollen analysis in drilling cores lack modern pollen processes as a reference (Prabhu et al., 2004; Ivory and Lézine, 2009), indicating that modern information on the mechanisms of pollen transport from land to ocean is not included for comparison. Thus it is difficult to prove that pollen grains stored in ocean sequences can represent vegetation of adjacent lands. Previous studies on the Indian monsoon were mostly based on the presence of foraminifera that reflects upwelling, which in turn indirectly reflects the Indian paloemonsoon. Conversely, pollen in deepsea sediments can provide information on vegetation along the coast; apply direct evidence of environmental change on the continent for comparison with oceanic environmental change, especially for the reconstruction of the paloemonsoon evolutionary history. Sediment pollen of the northeastern Indian Ocean is more easily influenced by the winter monsoon due to its close proximity to the Asian continent. Previous studies on the Indian monsoon focus mostly on the summer monsoon, with few studies on the winter monsoon. The pollen index probably has much information about the winter monsoon. To date, some paleoenvironmental study in this area is very rare due to sampling limitations. It is especially rare to attempt paleovegetation and paleoclimate reconstruction using pollen analysis. The northeastern Indian Ocean has been selected as the present study area. Thirty-three seafloor surface sediment samples were collected for pollen analysis to reveal the surface pollen distribution (Fig. 1). These samples were used to explore the modern marine pollen dissemination process. Then we will apply this process as a reference for paleoclimate reconstruction of core YDY05 in the northeastern Indian Ocean and discuss the historical process of the India monsoon. 2. Regional setting 2.1. Submarine topography The main topographical characteristics of the northeastern Indian Ocean are the countless volcanic guyots, submerged seamounts, and

51

two subaerial islands. The northeastern Indian Ocean is separated into all types of physiographic divisions (Supplementary Fig. 1) comprising the North West Shelf, Wombat Plateau, Exmouth Plateau, Cuvier Abyssal Plain, Argo Abyssal Plain, Gascoyne Abyssal Plain, and Platypus Spur (Taneja et al., 2015). 2.2. Atmospheric circulation and climate The oceanography of the Indian Ocean is characterized by tropical ocean and monsoons. The climate can be divided into four climate zones. The northeastern part of the Indian Ocean has a monsoon climate. In the summer (May to October), a strong southwest monsoon flows from the Indian Ocean to the Asian continent, carrying pollen and spores from the far African continent to the northeastern part of the Indian Ocean. In the winter (October to April), a north and northeast monsoon flows from the Asian continent to the Indian Ocean, carrying abundant tropical and subtropical pollen and spores from the Southeast and South Asian continent to the northeastern Indian Ocean (Supplementary Fig. 2). The basin-wide surface wind field is dominated by an alternation between the winter (Supplementary Fig. 2a) and summer monsoons (Supplementary Fig. 2c) (Hastenrath, 2007). In winter the flow from the north recurves near the equator to meet the South Indian Ocean trade winds, whereas in summer the southern trade winds recurve near the equator to flow into South Asia. During the short transition between monsoons (Supplementary Fig. 3b and d), there are strong westerlies in the equatorial zone of the Indian Ocean (Hastenrath, 2007). 2.3. Oceanic circulation There is a complex system of surface currents in the eastern Indian Ocean (Supplementary Fig. 3) that moves according to the monsoon and exhibits an obvious seasonality (e.g., Wijffels and Sprintall, 2002; Wijffels et al., 1996; Gordon, 2005). From December to March (NW monsoon), the South Java Current (SJC), originating in the Equatorial Counter Current (ECC), flows to the southeast to meet the Leeuwin Current (LC) (e.g., Tomczak and Godfrey, 1994). There is a high volume of runoff from Sumatra and Java because there are high precipitation rates during this season. There is a buildup of a low salinity “tongue” in the SJC because of the advection of fresher water from the Java Sea through the Sunda Strait (e.g., Wijffels et al., 1996). The establishment of an active upwelling system off Java and southwest Sumatra is led by the prevailing wind direction during the SE monsoon (July–October), and the upwelling is characterized by lower sea surface temperature (SST) and higher sea surface salt (e.g., Tomczak and Godfrey, 1994). It is rather weak due to the existence of a “barrier layer” (Sprintall and Tomczak, 1992; Du et al., 2005; Qu et al., 2005) (e.g., Wijffels et al., 1996). The SJC flows northwest between July and October in contrast to the boreal winter, contributing to the South Equatorial Current (SEC) instead of the LC (e.g., Quadfasel and Cresswell, 1992). The SJC reverses every three months and is characterized by a high directional variability (e.g., Wijffels et al., 1996; Hessler et al., 2013). 2.4. Vegetation 2.4.1. Vegetation of Southeast Asia 1) Tropical rain forests and mountain rain forests Tropical rain forests in Southeast Asia are mainly distributed in lowlands near the equator (normally below 600 m to 1000 m), also known as the equatorial rain forest. Most rain forests are located in the Kalimantan Islands, Sumatra, Sulawesi, Irian Jaya, and the islands of the Malay Peninsula. They are only partially developed on the mainland, mainly distributed in the south of Vietnam, Cambodia, Thailand, and Myanmar. Lianas, or vines, are particularly lush in the rainforest of Southeast Asia. There are nine genera in the palm family. Of these, the Calamus genus is most widely distributed, with

52


C

YDY05

B

A

a YDY05

Depth (cm) 250

200

150

100

50

0 0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

hO P Fig. 1. a. Map of location of surface sediment sample and AB, BC transects n in the northeastern Indian Ocean. b. Age model of core YDY05.


72 known species of Calamus in peninsular Malaysia. Of the rain forest trees, Dipterocarpaceae is the characteristic family, known as the “king of the rainforest.” More than 50% of the tall trees of the upper rain forest is of the Dipterocarpaceae family (Supplementary Fig. 4) (Yan, 1984). In tropical rainforest regions, low land dipterocarp forests occur at an altitude of less than 350 min Peninsular Malaysia, while hilly dipterocarp forests occur at an altitude of 350 to 850 m. Finally, upper dipterocarp forests occur at an altitude of 850 to 1000 m. This change in altitude also marks the transition of a staggered community between lowland rain forest and mountain rain forest. Mountain rain forest typically occurs at an altitude of 1000 to 2500 m, and the subalpine forest begins above 2500 m (Supplementary Fig. 4) (Yan, 1984). 2) Tropical monsoon forest and mountain monsoon forest In Southeast Asia, the equatorial rainforest transitions into the seasonal forest in both the northern and southern hemispheres. Monsoon forests are widely distributed on Indochina north of the equator, as well as on some islands south of the equator, such as East Java and the Lesser Sunda Islands. Tectona grandis (Verbenaceae family) is the representative species in various types of monsoon forests in Southeast Asia, widely distributed in Myanmar, Indochina, middle and southern Vietnam, northeastern Thailand, Laos, and the Indonesian islands. Dipterocarpaceae is the main species of monsoon forest. Bamboo (Poaceae family) is lush in the seasonal rain forest; there are many bamboo species in Southeast Asia, including Bambusa, Gigantochloa, Dendrocalamus, etc. Mountain monsoon forest is mainly distributed at an altitude of 1000–2600 m. Pine trees are often dominant in the composition of monsoon mountain forest; this trend is more obvious with increasing altitude. Pinus merkusii, Pinus kesiya, and Pinus insularis are dominant pines. Pinus merkussi is mainly widely distributed in Myanmar, Thailand, Vietnam, Laos, Cambodia, the Philippines, and Indonesia. Pinus insularis is less widespread, mainly distributed in Thailand, Vietnam, and the Philippines. Pinus merkuisi often grows among Dipterocarpaceae species, while Pinus insularis often grows among Quercus or Castanea (Supplementary Fig. 4) (Yan, 1984). 2.4.2. Vegetation in South Asia 1) Vegetation of India India has forests, scrub/shrublands, and grasslands of natural and semi-natural systems (Roy et al., 2015). Tropical evergreen forests are mainly distributed in the northeast region, the Western Ghats, and the Andaman and Nicobar Islands, whereas tropical semi-evergreen forests are distributed as a transition zone between moist deciduous forests and evergreen forests. Tropical moist deciduous forests occur in strips along the eastern side of the Western Ghats, along with the foothills of the Himalayas, and on the northwestern hills and the Chota Nagpur Plateau. Tropical dry deciduous forests are distributed on both sides of the Tropic of Cancer, predominantly comprising teak (Tectona grandis) and sal (Shorea robusta). The tropical thorn forests occurring in western India generally belong to thorny leguminous species. Subtropical forests comprising both dry evergreen forests and broad-leaved hill forests could be distributed in both the eastern and western Himalayas. Temperate broad-leaved forests are distributed in the eastern Himalayas between 1500 m and 3000 m elevation and the Nilgiris, the upper reaches of the Western Ghats. Temperate mixed forests include both coniferous and broad-leaved species and are distributed in the eastern and western Himalayas (Supplementary Fig. 4). Subalpine forests extend up to the treeline throughout the Himalayas and are replaced by alpine meadows, which are dry and moist. Subalpine forests are dominated by Abeis, Picea, Betula, and Rhododendron genera. Mangroves are mainly distributed in the river deltas along the coasts, such as the Sunderbans. Scrub/shrub areas and small saplings and trees are distributed in northern India, areas of southern India and the central highlands. Grasslands are both

53

Table 1 Location of 33 surface sediment samples in the northeastern Indian Ocean. Sample number

Longitude

Latitude

Depth (m)

Position (cm)

15I306 15I301 15I318 15I502 15I209 12I210 I611 11YDY406 12I407 15I206 12I206 HFA04 HF02 HF06 11YDY409 HFA03 12I201 15I202 11YDY411 I42110 11YDY03 15I118 11YDY413 YDY02 11YDY415 11YDY415 15I107 11YDY502 15I101 11YDY505 HFA01 HFA01–2 I601

79.9785 80.00025 80.08608333 80.94163333 82.96876667 84.49583333 84.547267 85.03055 86.0025 86.02708333 86.49416667 87.44554 87.62572833 87.902896 88.0368 88.1674 89.00726667 89.98835 90.05141667 90.17848394 91.005763 91.61493333 92.03035 93.55861 94.00751667 94.00751667 94.62295 96.22585 96.38866667 98.59796667 88°0.928′E 88°0.928′E 89°29.9761′E

3.49215 5.9985 2.4545 1.955533333 0.02255 5.994133333 5.99727 0.00915 0.003116667 0.026266667 6.038633333 8.12876 7.01055833 9.0703083 0.019566667 7.30527 6.001266667 0.002583333 0.009016667 4.857256842 9.9978133 2.518916667 0.03795 9.212757 0.074183333 0.074183333 2.99755 2.019366667 5.966483333 5.076733333 6°31.261′N 6°31.261′N 6°00.1467′N

4989 5172 4316 4836 4352 3950 3945 4517 4505 4511 3933 3644 no data no data 4495 3719 3828 4082 4318 3675 3470 4218 4508 1306 4477 4477 4762 4654 5223 4819 3824 3824 3224

Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface Surface

primary and secondary formations occurring in the plains, in abandoned shifting cultivation lands, along with the slopes in the Himalayas and along the coasts of western India (Roy et al., 2015) (Supplementary Fig. 4). 2) Vegetation in Sri Lanka Vegetation on Sri Lanka consists of tropical rainforest in the southwest lowlands, tropical mountain rain forest at an altitude of 1000–1800 m and tropical forest at higher altitudes. Natural forest accounted for 30.89% of the national land area in 1992. The most forest is arid monsoon forest, with a small area of tropical rainforest. The major vegetation type in the dry zone of Sri Lanka is dry semievergreen forest, which includes the dry peat ssp. Manilkara hexandra, Bauhinia racemosa, and Grewia tiliaefolia as the most common taxa. The most common species of forest plantation are Tectona grandis, followed by eucalyptus (mainly Eucalyptus grandis), Pinus caribaea, and Swietenia mahogani. 3. Materials and methods 3.1. Sampling 3.1.1. Surface sediment sampling The samples for pollen analysis were collected mainly between 2010 and 2015 by the northeastern Indian Ocean oceanography cruises of integrated science experiments organized by the Micropaleontology Research Group of the South China Sea Institute of Oceanology, Chinese Academy of Sciences (CAS). Only the mud-like portion of the sediment to a depth of 2 cm was obtained for the laboratory for analysis to ensure that the samples were representative of modern sediments. The sediment samples were retrieved using a box or grab sampling collection method. A total of 33 samples from the northeastern part of the Indian Ocean were selected for pollen analysis (Fig. 1, Table 1). The samples were taken from deep basins, ridges, slopes, and straits. The water depths of samples varied from 3224 to 5223 m, with the

54


Table 2 AMS 14C and calibrated calendar ages and calculated sedimentation rate at cores YDY05 in the northeastern Indian Ocean. Depth (cm)

Lab number

Materials

δ13C (‰)

14

7 25 53–54 122–123 237

391109 391110 391111 391112 391113

Mixed planktonic foraminifera Mixed planktonic foraminifera Mixed planktonic foraminifera Mixed planktonic foraminifera Mixed planktonic foraminifera

1 0.9 1.5 1 0.8

5190 ± 30 9490 ± 40 13,730 ± 50 23,350 ± 100 N43,500

C ages (a BP)

exception of sample YDY02. These samples were located on a slope with a shallower water depth of 1302 m, with a longitudinal range of 79°E to 99°E, and a latitudinal range of 0° to 10°N. The number of samples used in the present study and their distribution is a good representation of the northeastern Indian Ocean surface sediment (Fig. 1). 3.1.2. Core sediment sampling The Gravity piston cores were collected in a No. 1 R/V Experiment boat in 2012 and 2010 by the Chinese Academy of Sciences Institute of the South China Sea. The length of core YDY05 is 240 cm. The samples of the YDY05 core were taken from the west slope of the East Indian Range. The YDY05 core was located at a longitude of 90.32°E and latitude of 9.99°N. The carbon isotope analyses of benthic foraminifera oxygen and AMS14C planktonic foraminifera chronostratigraphic tests were compared with establish a high-resolution chronostratigraphic sequence of nearly 43.5 ka years for the YDY05 core (Fig. 1c). The YDY05 core was separated and described by sampling at 1-cm intervals in the laboratory and analyzed by pollen from 119 samples with 450 a resolution. Planktonic foraminifera from 5 horizons in core YDY05 was picked for AMS 14 C dating at the Beta Analytic Inc., USA, and about 15 mg of shells were picked from the N150 μm fraction for the measurements. All the following ages are reported as calendar years before present (ka, Table 2).

Calendar ages (a BP)

Sedimentation rate (cm/ka)

Error ages (a BP)

5440 10,200 15,855 27,210

3.78 (7–25 cm) 4.95 (25–53 cm) 6.08 (53–122 cm) 7.06 (122–237 cm)

30 40 50 100

the distribution patterns of pollen types. The percentage values for pollen and spores are relative. In contrast, it is absolute for the pollen concentration, and it provides a basis for comparing the amounts of pollen or spores of a given species. Combining the two parameters, we can obtain powerful clues about the source of the pollen or spores. The concentration was calculated in two steps. First, the relative proportions of pollen or spores were calculated, and then the number of Lycopodium spores was used as a reference to normalize the values: R ¼ ½27; 637=Lycopodium number per slide ½pollen or spore number per slide=total weight per sample where R is the concentration of pollen or spores. After determining the number of pollen grains or spores in each sediment sample and

3.2. Laboratory analysis Samples of 4 g to 10 g were processed for pollen. We used a hydrofluoric acid treatment to remove siliceous materials from the samples. Then, a heavy liquid (HBr, KI, and Zn) with a specific gravity of 2.0 (d = 2) was used to separate the spores and pollen by flotation. The spores and pollen were concentrated in the supernatant. Then, the supernatant was kept in other tubes, and the pollen and spore were kept in the remaining materials in the tubes after centrifugation. If there were many siliceous materials in the remaining material, then an HF acid treatment was used to remove siliceous materials. Finally, the remaining material mixed with pollen and spore was filtered through a 7-μm nylon sieve with an ultrasonic treatment to make the slides clearer. To calculate the pollen concentrations, we added Lycopodium spore tablets with a known concentration (27,637/grains) before analyzing the samples. Maher (1972) calculated the confidence intervals that could be expressed as the significance of the pollen percentages. Since pollen content in the ocean samples is normally lower than that of the continent samples, the presence of more than 100 pollen and spore grains per sample is the standard for a statistically significant sample. Counting pollen grains until the added Lycopodium spores total more than 1000 is the standard for samples with a low pollen content. 3.3. Calculations of pollen percentages and concentration Using the total number of pollen grains from terrestrial seed plants as a base value, we calculated the percentage of each type of spore and pollen in both modern and sentimental samples. The distribution pattern of the pollen was obtained using distribution maps of the percentages of surface pollen. This distribution pattern reflects the pollen transmission route and mechanism, the distance from the shore, and

Fig. 2. Total pollen concentration of the northeastern Indian Ocean.


calculating the pollen or spore percentages and concentrations, the values were plotted on a map of the study area. 3.4. PCA Principal component analysis (PCA) uses an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values of linearly uncorrelated variables that are called principal components. PCA is a statistical ordination procedure, and it calculates the distances between samples in a way that approximates their Euclidean distances in a multidimensional space (Lamb, 1984). The influence of rare pollen types is increased, and the weight of the more abundant ones is reduced by square-root-transformed pollen proportions in these ordinations (Prentice, 1980; Ter Braak and Prentice, 1988; Minckley and Whitlock, 2000). We used C2 software (Juggins, n. d.) to perform the principal components analysis based on the pollen percentages of 33 surface sediment samples collected in the study area. 4. Results 4.1. Distribution of total pollen concentration and percentage of surface sediments The pollen and spore composition of the surface sediments were complex. We identified 66 different pollen and spore taxa. The spores included Trilete spores (mainly Microlepia), Cyathea, Pteris, Polypodiaceae, Cibotiumbarometz, Lygodium, Hicriopteris, Selaginella,

55

and Hymenophyllaceae. Herbaceous pollen included Chenopodiaceae, Artemisia sp., Compositae, Ericaceae, Poaceae, Polygonum sp., Cyperaceae, Caryophyllaceae, Parkeriaceae, Potamogetonaceae, Myriophyllum sp., Mimosaceae, and Orchidaceae. The tree pollen included Pinus sp., Palmae, Podocarpus sp., Dacrydium sp., Tsuga sp., Picea, Abies sp., Cupressaceae, Keteleeria sp., Pieca sp., Taxodiaceae, Larix sp., Tiliaceae, Alnus sp., Betula sp., Juglans sp., Pterolobium sp., Quercus sp., Liquidambar sp., Magnoliaceae, Meliaceae, Myrica sp., Myrtaceae, Ulmaceae, etc. There are very few pollen grains in the pollen diagram of the tropical rain forest because there are a lot of insects in the tropical rain forest and most of the pollen is transported by insects, thus pollen cannot be scatted onto the air. So there is very few pollen in the tropical air, pollen is, even more, few in the tropical ocean because of long distance transporting from far away, this result is consisted with from Sun et al. (1999) and van der Kaars (2001). Thus spores are relatively abundant in the tropical pollen diagram; so much as spore can be taken as an index of winter monsoon in the northern South China Sea (Sun et al., 1999). The highest measured total concentration was 1124 pollen grains/g (Fig. 2). The samples with the highest pollen concentration were taken from the north-west of middle Indian Ocean basin at the equator. Samples with concentrations above 73 grains/g were taken mainly from the north of the basin with water depths of less than 3000 m. Their distribution was substantially on the west of the 90°E Indian Ridge. Samples with concentrations of less than 73 grains/g were taken from the middle Indian Ocean basin near the equator with water depths of 3000 to

Trilete spore ) (mainly

Fig. 3. Pollen percentage of the main taxa of the northeastern Indian Ocean.

56


Trilete spore (mainly

)

Fig. 4. Pollen concentration of the main taxa of the northeastern Indian Ocean.

5000 m, near the Strait of 10°N and the Strait of Malacca. The percentages of Trilete spores (mainly Microlepia) and Pteris were the highest among the trilete spores, they are probably mainly from Sri Lanka. The percentage of Polypodiaceae spores was the highest among the monolete spores, they are probably mainly from both Sumatra and Sri Lanka. The reason might have a close relationship with the distribution of their concertation. Fern spores had the highest percentage compared with herbs and trees. Herb pollen constituted the second high percentage. The overall percentage of herb pollen was higher than that of tree plants (the maximum percentage of Chenopodiaceae reached 60%, mainly distributed north of the middle Indian Ocean basin, with a water depth of 3000 to 5000 m, which is located north of the equator and west of the 90°E Indian Ridge). The percentage of tree plant pollen was the lowest compared with fern and herb pollen (the highest percentage of Pinus pollen reached 53.8%, they are distributed at the equator and north of the study area, mainly from Sri Lanka and India). 4.2. Pollen percentages and concentrations of the main taxa of the surface sediments Pollen percentages and concentrations of the main taxa have a different distribution. The concentration of fern spores was higher than that of the tree or herb pollen. The concentration of Polypodiaceae was found to be the highest among fern spores, with values of 434

grains/g. The concentration of Trilete spores (mainly Microlepia) was also high with a concentration of 159 grains/g. The concentration of Pinus was lower only in comparison to Polypodiaceae, with a concentration of 179 grains/g. Generally, the concentration of fern spore was the highest, the concentration of tree pollen was second highest, and the concentration of herb pollen was the lowest. When comparing the percentage of spores, tree pollen, and herb pollen, fern spores had the highest (among fern spores, the Microlepia genus has the highest percentage, followed by Polypodiaceae), tree plants had the lowest percentage (but the percentage of Pinus was up to 53.8%), and the overall percentage of herb pollen was higher than that of tree plants (the percentage of Chenopodiaceae reached up to 60%).

4.2.1. The spatial distribution of spores Trilete spores probably originated from Sri Lanka. Samples with the highest percentages (39% to 93%) of Trilete spores (mainly Microlepia) were distributed at the equator and the north part of the study area, the Strait of Malacca, near Smetana Island and Sri Lanka (Fig. 3). Samples with a high Trilete spores (mainly Microlepia) concentration (46.5 to 158.8 grains/g) were mainly distributed in the north part of the study area, BC transect, at the equator near Sri Lanka (Fig. 4). The distribution of Trilete spores (mainly Microlepia) spore percentages was similar to that of its spore concentrations.


The samples with a Polypodiaceae (monolete spore) percentage of 28% to 54% were mainly distributed at the equator and the BC transect, Strait of Malacca, near Smetana Island and Sri Lanka (Fig. 3). Samples with a high Polypodiaceae concentration (41 to 434 grains/g) were mainly distributed at the equator, BC transect, near Sri Lanka and Sumatra Island (Fig. 4). The distribution of Polypodiaceae spore percentages was similar to that of its spore concentrations. This distribution suggested that the monolete spores mainly originated from Sumatra Island and Sri Lanka, and were transported mainly by wind (Luo et al., 2015), as well as ocean currents (Supplementary Fig. 3). 4.2.2. The spatial distribution of trees Pine forest is often dominant in the composition of mountain monsoon forests in Southeast Asia. The percentage of Pinus pollen was lower than that of pine vegetation on the continent, with the Pinus pollen percentage ranging from 33% to 54%. Samples with a Pinus pollen percentage of N20% were distributed on the equatorial and northern part of the study area, BC transect (Fig. 3). The current study found that pollen concentration of Pinus was relatively high compared with other pollen. The distribution of Pinus pollen percentages was similar to that of its pollen concentrations, with a sample concentrated at 178 grains/g mainly distributed at the equator and samples concentrated at 15 to 178 grains/g mainly distributed on the 90°E range (Fig. 4). The distribution of Pinus pollen showed long-distance transport of high concentrations of Pinus pollen, mainly from Sri Lanka and Smetana Island. The major transport mechanisms would be the summer monsoon from the west (Supplementary Fig. 2b), as well as the ocean currents associated with this monsoon (Supplementary Fig. 3). Pollen percentages of tropical and subtropical broad-leaved trees (e.g., Castanopsis spp., evergreen Quercus spp., Myrtaceae spp.) were generally low. Samples with a relatively high Castanopsis pollen percentage (N 12%) were mainly distributed in the eastern part of the study area, near Smetana Island (Fig. 3). The samples with a low pollen percentage (b12%) were distributed in basins near the middle of the study area. The concentration of Castanopsis was the lowest (the maximum pollen concentration is 61 grains/g). Samples with a pollen concentration of N8.2 grains/g were mainly distributed at the equator. Samples with a pollen concentration of 3.4 to 8.2 grains/g were mainly distributed on the northwest–southeast transect, extending across 90°E (Fig. 4). The distribution data suggested that samples with a high concentration of Castanopsis pollen were mainly distributed in the middle of the study area (N 3.4 grains/g). Thus, the distribution of Castanopsis pollen percentages was opposite to that of its pollen concentrations. These results suggested that the Castanopsis pollen probably came from the Smetana Island, indicating that transport was mainly driven by the winter monsoon ocean circulation from the southeast along the northwest–southeast transect (Supplementary Fig. 3). Three samples with a Myrtaceae pollen percentage range of 0.4% to 3.6% were distributed in the northeast of the study area (Fig. 3). The other samples with a pollen percentage of b 0.4% were distributed at the equator and in the middle of the study area. The concentration of Myrtaceae pollen was relatively low. A pollen concentration of N 1.7 grains/g occurred in three samples. One sample contained a pollen concentration of 4.1 grains/g, which was distributed at the equator near Sri Lanka Island. Samples with a pollen concentration of 1.7 to 4.1 grains/g were mainly distributed in the north part of the study area (Fig. 4). Myrtaceae pollen concentration in the other samples was very low, with a concentration of b0.1 grains/g. The distribution of Myrtaceae pollen shows that this pollen originates mainly from Sri Lanka Island and other islands, with a few samples transported to deep basins. The main transport mechanisms seem to be ocean currents. Most percentages of Ulmaceae pollen range from 1.35% to 3.58%; these samples were distributed east and north of the study area, and few pollen grains of this pollen type were found south of the study area. Samples with Ulmaceae pollen concentrations ranging from 1 to 5.2 grains/g were found north and east of the study area (Figs. 3, 4).

57

From this distribution, the Ulmaceae pollen appears to have originated from Sumatra Island and the Indochina Peninsula. 4.2.3. The spatial distribution of herbs The percentage of Chenopodiaceae was low, with the highest value reaching 65%. Samples with Chenopodiaceae pollen percentages N 5.3% occurred at the equator and north of the study area (Fig. 3), with a sample near the west coast of Sri Lanka. Samples with a relatively high pollen concentration of 4.12 to 25.6 grains/g were also distributed at the equator and north of the study area, BC transect (Fig. 4). Thus, the distribution of Chenopodiaceae pollen percentages was similar to that of its pollen concentrations. This distribution suggests that the Chenopodiaceae pollen originated from Sri Lanka and Smetana Island. A sample with the highest Artemisia pollen concentration of 6.9 grains/g was distributed south of Sri Lanka and north of 0°N. Samples with a concentration of 1.229 to 4.122 grains/g were distributed in the basin west of the 90°E range, except one sample which was distributed on the transect of the equator and 90°E range. The sample of 18.3% Artemisia pollen was distributed south of Sri Lanka and north of 0°N, while samples of 3.8% to 10% Artemisia pollen were distributed at the equator and the Strait of Malacca near Sumatra Island and few grains of this pollen type were found in the middle of the study area. The Artemisia pollen percentage was 0% in more than half of the samples. Artemisia pollen was scattered more south than north (Figs. 7, 8). From this distribution, Artemisia pollen appears to have mainly originated from Sumatra Island and Sri Lanka and has been transported by eastbound ocean currents (Supplementary Fig. 3). 4.3. Pollen in the east–west and northwest–southeast transects of surface sediments In order to reflect the relationship between surface pollen and the topography and monsoon circulation of the northeastern Indian Ocean, two transects, AB and BC, were selected (Fig. 1b). Transect AB runs east–west parallel with the equator, crossing deep basins and passes through the 90°E Indian Ridge. Transect BC begins near Sumatra Island and crosses the 90°E Indian Ridge from southeast to northwest. In relation to the 90°E Indian Ridge, the AB transect extends roughly from the western basin to the eastern basin over the Ridge, while the BC transect extends roughly from the south-eastern basin to the north-western basin over the Ridge. 4.3.1. Characteristics of pollen percentages along the east–west AB transect Sample 11YDY411 is on 90°E longitude line. On the east–west transect, 90°E longitude line is a boundary; pollen character on the western side is different from that of the eastern side. There are more Trilete spore (mainly Microlepia) and herb pollen west of the boundary. Such as in sample 15I209, there is a high percentage (58.3%) of Trilete spore (mainly Microlepia) in this sample, probably because 15I209 is the farthest sample west of the equator and the Trilete spore (mainly Microlepia) is from Sri Lanka. There is more Polypodiaceae east of the boundary, as well as more trees (except Pinus), such as Quercus, Liquidambar, and Mysica, possibly from Sumatra Island. Pinus pollen is from both Sumatra Island and Sri Lanka Island, but Pinus pollen can be transported far away, so the percentage is higher offshore. The percentage of herbs, such as Chenopodiaceae and Artemisia, is higher west of the boundary because the ocean currents and monsoon go mainly eastward at the equator (Supplementary Figs. 2, 3, Fig. 3). It is worth mentioning the high percentage (60%) of Chenopodiaceae in sample 12I407. The reason for this high percentage is currently unknown (Fig. 5a). 4.3.2. Characteristics of pollen percentages along the northwest–southeast BC transect In the northwest–southeast transect, sample I42110 is roughly on the 90°E longitude line. The 90°E longitude line is a boundary; pollen character of the north-western part differs from that of the south-


GXG

15I209

GYG

11YDY406

GZG

12I407

G[G

15I206

G\G

11YDY409

G]G

11YDY411

G^G

15I202

G_G

11YDY413

G`G

11YDY415

GXWG

ea

st

W

YW

[W

]W

W

YW

YW W

YW

[W

j

h

j

]W W

n t w w { { x j j s t j

s

[W

n e pl

t

am

es

S

w GWG

um { O be r G G t P w G

58

]W W

W

W

W

W

YW W

W

W

W

YW W

W

W

W

W

GWG

GXG

HF06

GYG

HFA04

GZG

HFA03

G[G

HF02

G\G

12I201

G]G

I42110

G^G

15I118

G_G

s G o j ¡ n w w w w k { w j { x j j t t | j

j

11YDY415

G`G

ea

w

t

t

S

or

es

9G

N

hw

VR K T[ { O S HK G X G t P G

a

s st-

ou

th

W

YW

[W

]W

_W

XWW

W

YW

[W W

W

W

W

YW

W

W

W

YW

W

W

W

W

W

W

W

W

W

W

W

W

W

W

b

Fig. 5. a. Pollen diagram for surface samples on the east–west transect in the northeastern Indian Ocean. Western samples are shown at the top of the diagram, and eastern samples are at the bottom; b. Pollen diagram for the surface samples on the northwest–southeast transect in the northeastern Indian Ocean. Northwestern samples are shown at the top of the diagram, and southeastern samples are at the bottom.

eastern part. There are more Trilete spores (mainly Microlepia) and herb pollen in the north-western part of the boundary. More Polypodiaceae and tree pollen (such as Castanopisis, Mysica, and Myrtaceae, except Pinus) occur in the south-eastern of the boundary. The Polypodiaceae spores and tree pollen are possibly from Sumatra Island (Fig. 5b). 4.4. PCA of surface sediment samples 4.4.1. PCA analysis based on Axis 1 (component 1) The PCA for the percentage of the 33 surface sediment samples from the northeastern Indian Ocean shows that the eigenvalue of Axis 1

(component 1) is highest (0.397), Axis 2 (component 2) is the second highest (0.237), and Axis 3 (component 3) is the third highest (0.154) (Table 3). The first two axes or components explain 63.4% of the variation within the dataset (Table 3), so we just need to show the Biplot of the first two axes and discuss component 1. Based on species scores comp. 1 vs. comp. 2, taxa with scores on component 1 from high to low are (Fig. 6a): Castanopsis, Pinus, Polypodiaceae, Artemisia, Concentricystes, Quercus, Chenopodiaceae, Castane, Compositae, Orchidaceae, etc. Samples with scores on component 1 from high to low are (Fig. 6a): 15I107 (94.62, 2.99), 15I502 (80.94, 1.96), I601 (89°29.98′E, 6°00.15′N), 11YDY406, 15I202,

C. Luo et al. / Review of Palaeobotany and Palynology 256 (2018) 50–62 Table 3 Eigenvalues of the main five axes of the PCA. #

Code

Name

Eigenvalue

1 2 3 4 5

Axis 1 Axis 2 Axis 3 Axis 4 Axis 5

Eigenvalue for axis 1 Eigenvalue for axis 2 Eigenvalue for axis 3 Eigenvalue for axis 4 Eigenvalue for axis 5

0.396794 0.237097 0.153896 0.0949236 0.0492284

11YDY411, 12I407 (86.0025, 0.00312). These samples are mainly distributed on the east and south margins of the study area. In the opposite direction of Axis 1 (component 1), Trilete spore (mainly Microlepia) has the lowest score on component 1, samples with scores on component 2 from low to high are (Fig. 6b): HF06, HFA04, 15I209, HFA01-2, HFA01, 12I206, 12I201. These samples are mainly distributed northwest near the BC transect. Thus, it seems the components 1 that can mainly affect pollen percentage in the study area are the ocean current that flows from southeast to westnorth. This is consistent with high pollen percentages of

j

t WX H z H zWY

\UW [UW ZUW

{GG O GtP

YUW XUW j h w j

WUW

Castanopsis, Polypodiaceae, Concentricystes, and Quercus with a high score in component 1 that is found in the southeastern part of the study area, whereas high percentages of Trilete spore (mainly Microlepia) with a low score in component 1 that is found in the northwestern part of the study area. Since the SJC flows to northwest between July and October (e.g. Quadfasel and Cresswell, 1992, Supplementary Fig. 3) in the study area, thus the component 1 that can mainly affect pollen percentage are probably the ocean currents driven by summer monsoon.

4.4.2. PCA analysis based on Axis 2 (component 2) The PCA for the percentage of the 33 surface sediment samples from the northeastern Indian Ocean shows that the eigenvalue of Axis 2 (component 2) is the second highest (0.24). Based on species scores comp. 1 vs. comp. 2, taxa with scores on component 2 from high to low are (Fig. 6a): Castanopsis, Trilete spore (mainly Microlepia), Hicriopteris, Cupressaceae, Potamogetonaceae, Ulmaceae, Myrtaceae, Palmae, and ciborium. These species are mainly distributed in the north and south sides in the northwest–southeast transect (Fig. 6b). Samples with scores on component 2 from high to low are (Fig. 6c): 15I107, HF06, 11YDY505, HFA04, HFA03, 15I209, HFA01-2, 12I201, 12I206. They are distributed on the east, north, and southwest margins of the study area. Thus, component 2 could possibly be the distance of the samples from the lands around the study area (Fig. 1). In the opposite axis, Polypodiaceae has the lowest score on component 2, it is mainly distributed in the middle of the northwest– southeast transect (Fig. 5b). Samples with scores on component 2 from low to high are (Fig. 6c): I601, 15I202, 15I318, I611, 11YDY406, I42110, 11YDY413. They are mainly distributed on the northwest– southeast transect, far away from lands. This coincides with the result that component 2 might be the distance between the samples from the lands (Fig. 6b).

TXUW

4.5. Pollen diagram in the YDY05 core in the northeastern Indian Ocean

w

TYUW

a

w

TZUW T]UW T\UW T[UW TZUW TYUW TXUW WUW

XUW

YUW

ZUW

[UW

tGWXGHGzGGHGzWX

ZUW

t WX H z H zWY

59

X\pXW^

YUW

XUW

omW] XXk\W\ omhW[

WUW

X\pYW` omhWZ omhWXTY omhWX X\pZW] XXkWZ X\pXX_ XXk[X\ XXk[W` X\pXWX kWY XXk[XX X\p\WY XXk\WY XYpYXW XXk[XZ XYp[W^

p[YXXW

X\pZX_ XXk[W] p]XX X\pYWY p]WX

TXUW TYUW

TXUW

WUW

b XUW

YUW

tGWXGHGzGGHGzWX Fig6 Fig. 6. a. Species scores comp. 1 vs. comp. 2; b. Sample scores comp. 1 vs. comp. 2.

According to the pollen and AMS 14C results of the YDY05 core, the pollen diagram has been divided into four distinct zones, A, B, C, and D, from bottom to top based on the Tilia and Coniss analysis, and a description of each zone is given below (Fig. 7b). Zone A covers the time span of 45-36 ka (Calendar ages, a BP) (239–182 cm): Polypodiaceae, Pinus, Castanopsis, evergreen Quercus, and Concentricystes are dominant. This zone is characterized by the dominance of fern spores (35.42%), tree pollen (54.71%), and Concentricystes (9.86%), with an average pollen concentration of 736 grains/g. Zone B covers the time span of 36–26 ka (182–116 cm): Polypodiaceae, Pinus, Castanopsis, evergreen Quercus, and Concentricystes are dominant. This zone is characterized by the dominance of fern spores (46.61%), tree pollen (43.99%), and herbaceous pollen (0.29%), Concentricystes (9.12%), with an average pollen concentration of 619 grains/g. Zone C covers the time span of 26–16 ka (116–54 cm): Polypodiaceae, Pinus, Castanopsis, evergreen Quercus, and Concentricystes are dominant. This zone is characterized by the dominance of fern spores (49.65%), tree pollen (42.98%), and herbaceous pollen (0.99%), Concentricystes (6.38%), with an average pollen concentration of 285 grains/g. Zone D covers the time span of 16–0 ka (54-0 cm): Trilete spore (mainly Microlepia), Polypodiaceae, Lygodium, Myrica, Castanopsis, evergreen Quercus, Ulmaceae, Tsuga, Taxodiaceae, Myrtaceae, Artemisia, Liquidambar, and Concentricystes are dominant. This zone is characterized by the dominance of fern spores (67.19%), tree pollen (15.4%), and herbaceous pollen (3.44%), Concentricystes (13.97%), with an average pollen concentration of 92 grains/g.

60

C. Luo et al. / Review of Palaeobotany and Palynology 256 (2018) 50–62 m

{G

X]

Y]

Z]

O j

w k { w r { j t t t i h y l k Gx j Gx j s j y w ¡ z l | j

z j h j l n j w j w

o

j

o

s G

w

w

j

k { O G G t G

P

¡

V

P

o

Pollen zones

W \ XW X\ YW Y\ ZW Z\ [W [\ \W \\ ]W ]\ ^W ^\ _W _\ `W `\ XWW XW\ XXW XX\ XYW XY\ XZW XZ\ X[W X[\ X\W X\\ X]W X]\ X^W X^\ X_W X_\ X`W X`\ YWW YW\ YXW YX\ YYW YY\ YZW YZ\ Y[W

jvupzz

D C B A YW [W ]W

YW

YW

YW [W ]W

YW [W

YW

YW [W

YW [W ]W _W

YW

YW

YW

YW

YW

YW

YW [W ]W _W

\WW XWWW X\WW

YW

[W

]W

_W

XWW XYW X[W X]W

{GGG

Fig. 7. Pollen diagrams of core YDY05.

5. Discussions 5.1. Comparison surface sediment pollen with vegetation in Southeast Asia and South Asia Normally, we think that pollen concentration is also driven by the sedimentation rate. A higher pollen concentration might point to a lower sedimentation rate. However, there is no close relationship between the total pollen concentration of surface sediment and the sediment deposition rate in the South China Sea. This might be due to the close relationship of the deposition rate with the sediment sources transported by air, currents, and rivers into the South China Sea, which generally contain volcanic, biogenic, and terrigenous material. Among them, volcanic, biogenic (e.g., coral and foraminifera) have less relationship with pollen (Luo et al., 2016). The relationship between the total pollen concentration and the sediment deposition rate was also observed in other areas by Mudie (1982), Heusser (1983) and Hooghiemstra et al. (1986), who noted that pollen concentrations decrease with increasing distance from the shore in surface sediment samples from the continental shelf of eastern Canada, the United States and West Africa. Thus, pollen concentration reliably indicates the pollen source of the parent plants, instead of the deposition rate. From the distribution of the total pollen concentration, it can be seen that the pollen in the study area is mainly from Sri Lanka, while other sources include the islands of Sumatra and the Malay Peninsula. For example, samples with a high Trilete spores (mainly Microlepia) concentration (46.5 to 158.8 grains/g) were mainly distributed in the north part of the study area, BC transect, at the equator near Sri Lanka (Fig. 4), so its source is probably Sri Lanka (Fig. 4). Samples with a high Polypodiaceae (monolete spores) concentration (41 to 434 grains/g) were mainly distributed at the equator, BC transect, near Sumatra Island and Sri Lanka (Fig. 4), so the monolete spores mainly originated from Sumatra Island and Sri Lanka. Among trees, samples with a high concentration of Castanopsis pollen were mainly distributed in the middle of the study area (N 3.4 grains/g), but samples with a relatively high Castanopsis pollen percentage (N 12%) were mainly distributed in the eastern part of the study area, near Smetana Island (Fig. 3). These results suggested that the Castanopsis pollen came from the Smetana Island. Among herbs, samples with a relatively high pollen concentration of 4.12 to 25.6 grains/g were also distributed at the equator and north of the study area, BC transect (Fig. 4). This distribution suggests that the Chenopodiaceae pollen originated from Sri Lanka and Smetana Island. Tropical rain forests in Southeast Asia are mainly distributed in lowlands near the equator (normally below 600 m to 1000 m). Most rain forests are located in the Kalimantan Islands, Sumatra, Sulawesi, Irian Jaya, and the islands of the Malay Peninsula. Dipterocarpaceae is known as the “king of the rainforest”, but Dipterocarpaceae pollen has not been found in the present study. Maybe Dipterocarpaceae pollen is low representative or not transported far away. As a typical type of tropical

rainforest, Dipterocarpaceae has very low pollen expression, for example, the percentage of Dipterocarpaceae pollen in pollen assemblage of topsoil is less than 5% (Sun et al., 2001), and the percentage of the surface sediments samples of the tropical deep sea is even about 1% (van der Kaars, 2001). A small amount of Dipterocarpaceae pollen combined with other abundant tree pollen can indicate the existence of tropical rain forest. One reason may be that the pollen ingredients from the rainforest are mostly insect-borne; pollen is mainly spread by insects collecting near in different plants, and pollen is not spread into the air; vegetation that produces wind-borne pollen grows only at the higher elevations, so the percentage of tree pollen of the rainforest component is relatively low. Second, for the past 1400 a, rainforest on one of the main pollen resources, Kalimantan Island, has been affected by increased human activity (Gusti et al., 2001), especially in the relatively low-lying and flat terrain of northern Kalimantan Island, and there is mostly secondary forest in the region; pristine virgin rainforest is relatively less, it lies in central Kalimantan Island at a relatively high altitude, while ferns can grow easily in the hot and humid environment of the secondary forest. Meanwhile, tropical rainforests have a multilayer structure; there are many epiphytes (mostly ferns) under the tall trees (Zhu et al., 2003). Thus, there are mainly fern spores in the northern part of the island of Kalimantan, with low amounts of tree pollen. Most ferns are distributed in humid places. Their spores are usually larger in size than pollen grains of trees and herbs and hence are transported mainly by water (Wang, 2009). This is probably also because of the lightweight of fern spores; they can be carried far away from the seashore by flowing water, so they are especially representative in sediments, and the concentration of spores is high (Zheng et al., 2011). In the present study, extremely large proportions of fern spores including tree ferns (mainly Cyathea) in the pollen assemblages do not imply vegetation with ferns prevailing in the source area. Pine trees are often dominant in the composition of monsoon mountain forest; this trend is more obvious with increasing altitude. While Pinus insularis often grows among Quercus or Castanea in Southeast Asia. This is coherent with the results from AB, BC transects and PCA, indicating that pollen of Pinus, Quercus or Castanopsis are from eastern islands of the study area. Natural and semi-natural systems in India are classified into forests, scrub/shrublands, and grasslands on the basis of the extent of green cover (Roy et al., 2015). In Sri Lanka, the most forest is arid monsoon forest, with a small area of tropical rainforest. The major vegetation type in the dry zone of Sri Lanka is a dry semi-evergreen forest. Compared with larger area of closed forest in Southeast Asia, the vegetation in South Asia (India and Sri Lanka) are dryer and has more secondary forest and non-forest area (Supplementary Fig. 4), this might the reason that there are more Trilete spore (mainly Microlepia) and herb pollen west of the boundary (90°E longitude line) based on pollen analysis results along the east–west AB and northwest–southeast BC transects.


61

5.2. Terrestrial vegetation changes of the YDY05 core

Acknowledgments

In YDY05, zones A, B and C cover the time span of 45–16 ka (239–54 cm): the dominant pollen is from trees (mainly Pinus, Castanopsis and evergreen Quercus) and monolete spores, with very few herbaceous pollen; this is similar to the modern pollen diagram of seabed surface sediment samples near Sumatra Island, with a higher percentage of cold-loving Pinus pollen, indicating the pollen terrestrial area might be Sumatra Island, where Pine trees are often dominant in the composition of monsoon mountain forest, with a cool and humid climate (Fig. 7b). Zones D cover the time span of 16 ka-present (54–0 cm): the percentage of Trilete spores (mainly Microlepia) spores increased sharply, and the dominant pollen come from a larger variety of trees (Myrica, Castanopsis, evergreen Quercus, Ulmaceae, Tsuga, Taxodiaceae, Myrtaceae, and Liquidambar) (Fig. 7); this is similar to the modern pollen diagram of Sri Lanka Island, with more herbaceous and trilete spores (Fig. 3), indicating the pollen terrestrial area might be Sri Lanka Island and India, with a warm and dry climate. The change of the terrestrial area of pollen probably has a relationship with the climate and sea level change in the northeast Indian Ocean. Before 16 ka, due to lower sea level, pollen and spore (mainly Pinus, Castanopsis and evergreen Quercus and monolete spores) might come mainly from nearby Sumatra Island, based on the distribution data of surface sediments in the study area; there was a shorter distance from the YDY05 core to the river estuary at that time, and it received a large share of pollen and spore. After 16 ka, due to higher sea level, probably with the other reason of the intensive summer monsoon (Supplementary Fig. 2c), there was a larger distance from the YDY05 core to the Sumatra Island, more herb pollen, tree pollen (except Pinus, Castanopsis and evergreen Quercus) and spore, Trilete spores (mainly Microlepia) might come from Sri Lanka Island and India, based on the distribution data of surface sediments in the study area. The climate change during in 16 ka had obvious difference compared with the upper and lower layers. Thus it can be estimated that LGM might occur in 16 ka at the location of the core YDY05. In the previous study, the percentage of fern in the southern South China Sea was used as a proxy for the summer monsoon (Sun and Li, 1997). In the present study, it was found that in the YDY05 core from the northeastern Indian Ocean, the percentage of fern spores also increased in the interglacial period with the intensive summer monsoon. At approximately 11 ka, the Concentricystes (Freshwater algae fossils) were predominant, indicating that the runoff of the rivers, the Brahmaputra River and the Ganga River, was the largest.

This work was funded by the National Natural Science Foundation of China (grants NSFC 41376058, 41676047, 91228207), research program of Guangzhou Science Technology and Innovation Commission (Contract Number 201510010043), and by the National Program on Global Change and Air-Sea Interaction (GASI-03-04-01-03), and by the Strategic Priority Research Program of the Chinese Academy of Sciences with grant no. XDA11030104.

6. Conclusions (1) In the northeastern Indian Ocean, the trilete spores and herbaceous pollen in the surface sediment are primarily from Sri Lanka and India, whereas monolete spores and tree pollen are probably from Smetana Island; the exception is Pinus pollen, which is from Sri Lanka and India. (2) Based on the PCA analysis, two main factors can affect the percentage of pollen in the study area: component 1 is the northwestward ocean currents (probably driven by the summer monsoon), and component 2 is possibly the distance of the samples from the lands around the study area. (3) Based on a fossil pollen analysis on YDY05 core from the northeastern Indian Ocean, the pollen terrestrial area might change from Sumatra Island to Sri Lanka Island and India since 16 ka, the change of the pollen terrestrial area probably have a relationship with the climate and sea level change in the northeast Indian Ocean.

Supplementary data to this article can be found online at https://doi. org/10.1016/j.revpalbo.2018.05.007.

References Bentaleb, I., Caratini, C., Fontugne, M., Morzadec-Kerfourn, M.T., Pascal, J.P., Tissot, C., 1997. Monsoon regime variations during the late Holocene in the southwestern India. In: Dalfes, N., Kukla, G., Weiss, H. (Eds.), Third Millennium BC Social Collapse and Climate Change, NATO ASI Series. Series I, Global Environmental Change. Chauhan, M.S., Pokharia, A.K., Srivastava, R.K., 2015. Late Quaternary vegetation history, climatic variability and human activity in the Central Ganga Plain, deduced by pollen proxy records from Karela Jheel, India. Quat. Int. 371, 144–156. Clemens, S.C., Prell, W., Murray, D., Shimmield, G., Weedon, G., 1991. Forcing mechanisms of the Indian Ocean monsoon. Nature 353, 720–725. Clemens, S., Murray, D., Prell, W., 1996. Nonstationary phase of the Plio-Pleistocene Asian monsoon. Science 274, 943–948. Cutler, A.N., Swallow, J.C., 1984. Surface currents of the Indian Ocean. (To 25°S, 100°E). I. O. S. Technical Rept, p. 187. Demske, D., Tarasov, P.E., Wünnemann, B., Riedel, F., 2009. Palynological investigation of profile Tso Kar. PANGAEA, In supplement to: Demske, D et al. (2009): Late glacial and Holocene vegetation, Indian monsoon and westerly circulation in the Trans-Himalaya recorded in the lacustrine pollen sequence from Tso Kar, Ladakh, NW India. Palaeogeogr. Palaeoclimatol. Palaeoecol. 279 (3-4), 172–185. Du, Y., Qu, T., Meyers, G., Masumoto, Y., Sasaki, H., 2005. Seasonal heat budget in the mixed layer of the southeastern tropical Indian Ocean in a high-resolution ocean general circulation model. J. Geophys. Res. 110, C04012. Gordon, A., 2005. Oceanography of the Indonesian seas and their throughflow. Oceanography 18 (4), 13. Gupta, S.M., 2003. Orbital frequencies in radiolarian assemblages of the central Indian Ocean: implications on the Indian summer monsoon. Palaeogeogr. Palaeoclimatol. Palaeoecol. 197, 97–112. Gusti, A., Kershaw, A.P., van der Kaars, S., 2001. A Late Pleistocene and Holocene pollen and charcoal record from peat swamp forest, Lake Sentarum Wildlife Reserve, West Kalimantan, Indonesia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 171 (3), 213–228. Harris, P.T., et al., 2014. Geomorphology of the oceans. Mar. Geol. 352, 4–24. Hastenrath, S., 2007. Circulation mechanisms of climate anomalies in East Africa and the equatorial Indian Ocean. Dyn. Atmos. Oceans 43, 25–35. Hessler, I., Young, M., Holzwarth, U., Mohtadi, M., Luckge, A., Hermann, B., 2013. Imprint of eastern Indian Ocean surface oceanography on modern organic-walled dinoflagellate cyst assemblages. Mar. Micropaleontol. 101, 89–105. Heusser, L.E., 1983. Palynology and paleoecology of post-glacial sediments in an anoxic basin, Saanich Inlet, British Columbia. Can. J. Earth Sci. 20, 873–885. Hooghiemstra, H., Agwu, C.O.C., Beug, H.J., 1986. Pollen and spore distribution in recent marine sediments: a record of NW-African seasonal wind patterns and vegetation belts. “Meteor” Forsch. Ergebn. C 40, 87–135. Ivory, S.J., Lézine, A.-M., 2009. Climate and environmental change at the end of the Holocene Humid Period: a pollen record off Pakistan. C. R. Geosci. 341, 760–769. John, C., Brock, C.R. McClain, Anderson, D.M., Prell, W.L., Hay, W.W., 1992. Southwest monsoon circulation and environments of recent planktonic foraminifera in the northwestern Arabian Sea. Paleoceanography 7 (6), 799–813. Juggins, S., d. School of Geography, Politics and Sociology, Newcastle Universityhttp:// www.staff.ncl.ac.uk/stephen.juggins n.d. Lamb, H.F., 1984. Modern pollen spectra from Labrador and their use in reconstructing holocene vegetational history. J. Ecol. 72 (1), 37–59. Leipe, C., Demske, D., Tarasov, P.E., HIMPAC Project Members, 2014a. A Holocene pollen record from the northwestern Himalayan lake TsoMoriri: implications for palaeoclimatic and archaeological research. Quat. Int. 348, 93–112. Leipe, C., Demske, D., Tarasov, P.E., Wünnemann, B., Riedel, F., HIMPAC Project Members, 2014b. Potential of pollen and non-pollen palynomorph records from TsoMoriri (Trans-Himalaya, NW India) for reconstructing Holocene limnology and humanenvironmental interactions. Quat. Int. 348, 113–129. Liang, F.Y., Brook, G.A., Kotlia, B.S., Railsback, L., Hardt, B., Cheng, H.R., Edwards, L., Kandasamy, S., 2015. Panigarh cave stalagmite evidence of climate change in the Indian Central Himalaya since AD 1256: monsoon breaks and winter southern jet depressions. Quat. Sci. Rev. 124, 145–161. Luo, C., Chen, M., Xiang, R., Liu, J., Zhang, L., Lu, J., 2015. Comparison of modern pollen distribution between the northern and southern parts of the South China Sea. Int. J. Biometeorol. 59, 397–415. Luo, C., Jiang, C., Yang, M., Chen, M., Xiang, R., Zhang, L., Liu, J., Pan, A., 2016. Transportation modes of pollen in surface waters in the South China Seaand their environmental significance. Rev. Palaeobot. Palynol. 225, 95–105. Maher, L.J., 1972. Absolute pollen diagram of Redrock Lake, Boulder Co., Colorado. Quat. Res. 2, 531–553. Marret, F., Vernal, A.D., 1997. Dinoflagellate cyst distribution in of the southern Indian surface sediments ocean. Mar. Micropaleontol. 29, 367–392.

62


Minckley, T., Whitlock, C., 2000. Spatial variation of modern pollen in Oregon and southern Washington, USA. Rev. Palaeobot. Palynol. 112, 97–123. Molinari, R.L., Olson, D., Reverdin, G., 1990. Surface current distributions in the tropical Indian Ocean derived from compilations of surface buoy trajectories. J. Geophys. Res. 95, 7217–7238. Mudie, P.J., 1982. Pollen distribution in recent marine sediments, eastern Canada. Can. J. Earth Sci. 19, 729–747. Murgese, D.S., Deckker, P.D., 2005. The distribution of deep-sea benthic foraminifera in core tops from the eastern Indian Ocean. Mar. Micropaleontol. 56, 25–49. Prabhu, C.N., Shankar, R., Anupama, K., Taieb, M., Bonnefille, R., Vidal, L., Prasad, S., 2004. A 200-ka pollen and oxygen isotopic record from two sediment cores from the eastern Arabian Sea. Palaeogeogr. Palaeoclimatol. Palaeoecol. 214, 309–321. Prell, W.L., 1984a. Monsoonal climate of the Arabian Sea during the Late Quaternary: a response to changing solar radiation. In: Berger, A.L., Imbrie, J., Hays, J., Kukla, G., Saltzman, B. (Eds.), Milankovitch and Climate. D. Riedel, Hingham, pp. 349–366. Prell, W.L., 1984b. Variation of monsoonal upwelling: a response to changing solar radiation. In: Hansen, J., Takahashi, T. (Eds.), Climate Processes and Climate Sensitivity. Premathilake, R., Seneviratne, S., 2015. Cultural implication based on pollen from the ancient mortuary complex in Sri Lanka. J. Archaeol. Sci. 53, 559–569. Prentice, I.C., 1980. Multidimensional scaling as a research tool in quaternary palynology: A review of theory and methods. Rev. Palaeobot. Palynol. 31 (1-2), 71–104. Qu, T., Du, Y., Meyers, G., Ishida, A., Wang, D., 2005. Connecting the tropical Pacific with Indian Ocean through South China Sea. Geophys. Res. Lett. 32, L24609. Quadfasel, D., Cresswell, G.R., 1992. A note on the seasonal variability of the South Java Current. J. Geophys. Res. Atmos. 97 (C3), 3685–3688. Rawat, S., Gupta, A.K., Sangode, S.J., Srivastava, Priyeshu, Nainwal, H.C., 2015. Late Pleistocenee Holocene vegetation and Indian summer monsoon record from the Lahaul, northwest Himalaya, India. Quat. Sci. Rev. 114, 167–181. Riedel, N., Stebich, M., Anoop, A., Basavaiah, N., Menzel, P., Prasad, S., Sachse, D., Sarkar, S., Wiesner, M., 2015. Modern pollen vegetation relationships in a dry deciduous monsoon forest: a case study from Lonar Crater Lake, central India. Quat. Int. 371, 268–279. Roy, P.S., et al., 2015. New vegetation type map of India prepared using satellite remote sensing: comparison with global vegetation maps and utilities. Int. J. Appl. Earth Obs. Geoinf. 39, 142–159. Sander, V.D.K., Patrick, D.D., 2003. Pollen distribution in marine surface sediments offshore Western Australia. Rev. Palaeobot. Palynol. 124 (1–2), 113–129. Schott, F.A., Mc Creary Jr., J.P., 2001. The monsoon circulation of the Indian Ocean. Prog. Oceanogr. 51, 1–123.

Shenoi, S.S.C., Saji, P.K., Almeida, A.M., 1999. Near-surface circulation and kinetic energy in the tropical Indian Ocean derived from Lagrangian drifters. J. Mar. Res. 57, 885–907. Sprintall, J., Tomczak, M., 2004. On the formation of central water and thermocline ventilation in the southern hemisphere. J. Geophys. Res.-Atmos. 109 (D22), 827–848. Sun, X., Li, X., 1997. Different dynamics and routes of modern pollen transport in the northern and southern parts of the South China Sea. Sci. China Ser. D 41 (6), 494–498 (in Chinese). Sun, X.J., Chen, X.D., Luo, Y.L., Li, X., 1999. An insight into paleovegetation on the emerged shelf at the last glaciation: pollen data of the South China Sea. Acta Bot. Sin. 41 (9), 1016–1023. Sun, X., Li, X., Luo, Y., 2001. Vegetation and climate on the sunda shelf of the South China Sea during the last Glactiation—pollen results from station 17962. Acta Bot. Sin. 44 (6), 746–752. Taneja, R., O'Neill, C., Lackie, M., Rushmer, T., Schmidt, P., Jourdan, F., 2015. 40Ar/39Ar geochronology and the paleoposition of Christmas Island (Australia), Northeast Indian Ocean. Gondwana Res. 28, 391–406. Ter Braak, C.J.F., Prentice, I.C., 1988. A theory of gradient analysis. Adv. Ecol. Res. 18, 271–317. Tomczak, M., Godfrey, S., 1994. Regional oceanography: an introduction. Pergamon Press. van der Kaars, S., 2001. Pollen distribution in marine sediments from the south-eastern Indonesian waters. Palaeogeogr. Palaeoclimatol. Palaeoecol. 171 (3), 341–361. Wang, P.X., 2009. Global monsoon in a geological perspective. Sci. Bull. 5, 535–556. Wijffels, S., Sprintall, J., Fieux, M., Nan, B., 2002. The JADE and WOCE I10/IR6 Throughflow sections in the southeast Indian Ocean. Part 1: water mass distribution and variability. Deep-Sea Res. II Top. Stud. Oceanogr. 49 (7), 1341–1362. Wijffels, S.E., Toole, J.M., Bryden, H.L., Fine, R.A., Jenkins, W.J., John, L.B., 1996. The water masses and circulation at 10°N in the Pacific. Deep-Sea Res. I Oceanogr. Res. Pap. 43 (4), 501–544. Yan, C.C., 1984. Types and distribution patterns of vegetation in Southeastern Asia. Chin. J. Ecol. 05. Zheng, Z., Yang, S., Deng, Y., Huang, K., Wei, J., Berne, S., Suc, J.P., 2011. Pollen record of the past 60 ka BP in the Middle Okinawa Trough: terrestrial provenance and reconstruction of the paleoenvironment. Palaeogeogr. Palaeoclimatol. Palaeoecol. 307, 285–300. Zhu, H., Yan, L., Qin, H., 2003. Floristic Composition and Characteristics of Vietnamese Flora. Acta Sci. Natur. Univ. Sunyatseni 42 (6), 98–102 (in Chinese).