Impact of monsoon-driven circulation on phytoplankton assemblages near fringing reefs along the east coast of Hainan Island, China

Impact of monsoon-driven circulation on phytoplankton assemblages near fringing reefs along the east coast of Hainan Island, China

Author's Accepted Manuscript Impact of monsoon–driven circulation on phytoplankton assemblages near fringing reefs along the east coast of Hainan Isl...

1MB Sizes 0 Downloads 38 Views

Author's Accepted Manuscript

Impact of monsoon–driven circulation on phytoplankton assemblages near fringing reefs along the east coast of Hainan Island, China Y. Li, D.R. Wang, J. Su, J. Zhang

www.elsevier.com/locate/dsr2

PII: DOI: Reference:

S0967-0645(13)00155-0 http://dx.doi.org/10.1016/j.dsr2.2013.04.016 DSRII3374

To appear in:

Deep-Sea Research II

Cite this article as: Y. Li, D.R. Wang, J. Su, J. Zhang, Impact of monsoon–driven circulation on phytoplankton assemblages near fringing reefs along the east coast of Hainan Island, China, Deep-Sea Research II, http://dx.doi.org/10.1016/j. dsr2.2013.04.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Impact of monsoon–driven circulation on phytoplankton assemblages near fringing reefs along the east coast of Hainan Island, China Y. Lia,*, D.R. Wangb, J. Suc and J. Zhanga a

State Key Laboratory of Estuarine and Coastal Research, East China Normal University, 3663 Zhongshan Road North, Shanghai 200062, China

b

Hainan Provincial Marine Development Plan and Design Research Institute, 15 Longkun bei Road, Haikou 570203, China

c

Institute of Oceanography, Centre for Marine and Climate Research, University of Hamburg, 53 Bundesstrasse, Hamburg 20146, Germany

*Corresponding author. Y. Li's E-mail: [email protected]

Abstract Monsoonal hydrodynamic prevails over east coast of Hainan Island induced by southwest monsoon (SWM) and northeast monsoon (NEM) which drives coastal Ekman divergence/convergence cycle and the reversal of Guangdong coastal current (GCC) between the sGCC in the SWM season and nGCC in the NEM season. We report the control of such hydrodynamics on biological properties such as phytoplankton assemblages in east coast of Hainan Island.

Physico-chemical and biological

observations were carried out in two oceanographic cruises along east coast of Hainan Island during SWM period (July-August) of 2008 and NEM period (March- April) of 2009. Results indicated that phytoplankton assemblages in coastal regions (fringing reefs and coastal shelf) changed dramatically accompanying with the reverse of monsoonal hydrodynamic processes, with chain-forming diatoms (mainly, Pseudo-nitzschia spp. and Thalassionema nitzschioides) dominating during SWM cruise when coastal Ekman divergence and the sGCC were prevailed, but the pelagic N. scintillans and T. erythraeum dominating during NEM cruise when coastal Ekman convergence and the nGCC were prevailed. Furthermore, phytoplankton assemblages in fringing reefs along coastline were somewhat different from ones of coastal shelf, as fringing reefs is just located at dynamic boundary of offshore (or onshore) Ekman transport processes. Offshore diffusion of pelagic cells (such as T. erythraeum) driven by offshore Ekman transport process led to the lower abundance of T. erythraeum in fringing reefs than ones in coastal shelf during SWM cruise; on the contrary, onshore aggregation of pelagic cells (such as N. scintillans and T. erythraeum) driven by onshore Ekman transport process lead to higher abundances of N. scintillans and T. erythraeum in fringing reefs than ones in coastal shelf during NEM cruise, especially, N. scintillans formed bloom in fringing reefs. Last, we suggested that hydrodynamic processes must be taken into account in scientific management of fringing coral reefs health of east coast of Hainan Island, especially during northeast monsoon season when blooming specie cells (such as N. scintillans) could be introduced from eutrophic South China mainland coast to east coast of Hainan Island and piled to high-abundance at fringing reefs by monsoonal hydrodynamics.

Key words: Phytoplankton; Monsoon; Ekman; coastal current; coral; Pseudo-nitzschia sp.; Noctiluca scintillans; Trichodesmium erythraeum

1. Introduction Hydrodynamic process is an important factor to be considered in marine phytoplankton ecology research. On one side, the vertical-dimension hydrodynamic processes affect local phytoplankton assemblages by building habitat; for example turbulence facilitates nutrients availability from bottom water for phytoplankton growth and is benefit to diatom reproduction, but stratification conversely impedes ventilation and is fit to the reproduction of the shear-sensitive dinoflagellates (Smayda, 2002). On the other side, the horizontal-dimension hydrodynamic processes lead to the exchange of pelagic phytoplankton population between local and distant regions during watermass transport (Keafer et al., 2005; Dela-Cruz et al., 2003). Therefore, phytoplankton assemblages in sea water are balanced between vertical-dimension hydrodynamics (creating habitats) and the horizontal-dimension hydrodynamics (exchanging phytoplankton cells). Furthermore, hydrodynamic conditions in monsoon-controlled sea regions of the world are strongly and periodically perturbed by annual cycles of prevailing wind direction, consequently, phytoplankton assemblages usually present the distinct and predictable oscillations (Tan et al., 2006; Wiggert et al., 2002; Shalapyonok et al., 2001). The South China Sea (SCS) is under the persistent influence of the East Asian Monsoon i.e. strong NEM (northeast monsoon, winds speed ~9 m s-1) during October to April, and weak SWM (southwest monsoon, winds speed ~6 m s-1) during May to September (Tseng et al., 2005). At the northern boundary of the SCS, the coastlines of South China mainland and the east coast of Hainan Island oriented NE-SW, parallel to the prevailing wind directions; consequently, annual transition of prevailing wind directions significantly induce monsoonal hydrodynamic phenomena: the reverse of flow direction of Guangdong Coastal Current (GCC) (i.e. nGCC during the NEM and sGCC during SWM) (Figure 1A), as well as the coastal Ekman divergence/convergence cycle (Hu et al., 2000; Yang et al., 2002). To our knowledge, research about the control of monsoonal hydrodynamic on phytoplankton assemblages at the northern boundary of the SCS is mainly concentrated at the South China mainland coast. Southwesterly movement of the nGCC guided the progression of red tides cases from northeast to southwest along the South China mainland coast (Yin et al., 1999). Moreover, the monsoon-driven coastal Ekman divergence/convergence cycle along South China mainland coast powerfully controls the recurrent red tides in water bodies near to Hong Kong (Yin, 2003; Yin et al., 2001; Yin, 2002). Several studies reported the important role of monsoonal hydrodynamics on the regulation of phytoplankton assemblages along South China mainland coast (Wang et al., 2008; Tang et al., 2003; Lee, 2001). However, systematic research of monsoonal hydrodynamics regulated phytoplankton assemblages is still a fundamental requirement along east coast of Hainan Island. Available literatures about phytoplankton research are only a single sampling at east coast of Hainan Island as a part of the whole north South China Sea investigation (Sun et al., 2007; Ning et al., 2004). Considering the widely distribution of fringing reefs at east coast of Hainan Island, it is important for the coral reefs health management to carry out systematic research for phytoplankton assemblages in east coast of Hainan Island. This study presents the results of phytoplankton observations in east coast of Hainan Island respectively in the SWM season, 2008 and the NEM season, 2009. The results of this study have led to a better understanding how the monsoonal hydrodynamics affect the constructions of phytoplankton assemblages near fringing reefs in east coast of Hainan Island.

2. Materials and methods 2.1. Study area, sampling stations and cruises The east coast of Hainan Island faces to the broad continental shelf of the north SCS, with fringing reefs well develop along its coastline (Figure 1B). Two lagoons systems, Bamen lagoon and Boao lagoon, collect small-scale insular runoffs into the SCS, respectively from Wenchang river (i.e. Wenchanghe, with water discharge 9.09 m3 s−1), Wenjiao river (i. e. Wenjiaohe, with water discharge 11.6 m3 s−1) and Wanquan river (i.e. Wanquanhe, with water discharge 166 m3 s-1). Study area covers Bamen lagoon and Boao lagoon, fringing coral reefs, coastal shelf and offshore shelf (Figure 1B-C). Sampling is done along the salinity gradients of the lagoons. Sampling stations in fringing coral reefs covered the typical fringing reefs platforms or the reefs edges. Sampling stations in shelf composed of 25 gridding stations. Multidisciplinary

investigation were repeatedly conducted respectively in boreal summer from 25 July to 13 August, 2008 when SWM prevailed (hereafter referred to SWM cruise), and in early spring from 27 March to 14 April, 2009 when NEM prevailed (hereafter referred to NEM cruise). The SWM cruise was carried out smoothly however the NEM cruise was interrupted at the beginning of cruise by the severe sea condition (rainy and high wind). 2.2. Samples collection and analysis Hydrographic parameters (temperature and salinity) at lagoons and fringing coral reefs stations were measured using 3400i multi-parameter field meters (WTW, Weilheim, Germany). Vertical profiles of temperature and salinity of 25 hydrographic stations at shelf were examined with a Seabird SBE25 CTD. Surface water samples at lagoons and fringing coral reefs stations were collected with 5 L Niskin bottle, and the near-bottom water samples were either retrieved when the water depth were greater than 3 m. The discrete water samples of 15 biological/chemical stations at shelf were taken at three to five depths between the surface and 100 m depth using 10 L Niskin bottle. Subsamples for nutrients were filtered through 0.45μm cellulose acetate filters pre-cleaned with hydrochloric acid (pH=2) and rinsed with Milli-Q water before use. Filtrates were fixed using saturated HgCl2 and stored in the dark (Li, 2010; Liu et al., 2011). Subsamples (1 L) for phytoplankton enumeration were fixed with formaldehyde (2 % final concentration) immediately after collection. In the laboratory, concentrations of dissolved inorganic nitrogen (DIN= NO3- + NO2- + NH4+), reactive phosphorus and silica were measured using an autoanalyzer (Model: Skalar SANplus) by the chemical term of programs (Li, 2010; Liu et al., 2011). Phytoplankton samples were processed through a series of setting and siphoning steps to obtain a 30-50 ml concentrate which were counted and identified in a 1 ml Sedgwick-Rafter chamber (Mcalice, 1971) using a Leica phase contrast inverted microscope at 200 or 400 magnifications. Observing that cell abundance widely varied among several orders of magnitude among different habitats, >30 cells per sample were counted (for the oligotrophic water samples from the shelf) and >400 cells per sample were counted (for the nutrient-rich water samples in lagoons or coastal zone) to estimate cell abundances along with an error of counting less than 10%. Identification of phytoplankton was done to the species level whenever possible, and when not, they were identified to genus, using the taxonomic keys by Jin et al. (1985, 1992) and Tomas (1997). Only diatoms, dinoflagellates and cyanobacteria, three algal divisions are reported in this paper, which were widely distributed in whole study area from lagoon to offshore shelf however other algal divisions such as chlorophyta and euglenophyta etc. are not involved as they almost occurred only in lagoons samples. Pseudo-nitzschia spp. abundance was quite high in samples of SWM cruise, but it was difficult to identify to the species level, and mainly consisted of two species Pseudo-nitzschia sp1 and Pseudo-nitzschia sp2 respectively similar to P. delicatissima and P. pungens which were often reported as dominant species in the SCS (Sun et al., 2007; Le et al., 2006). 2.3. Statistics analysis Compositions

of

diatoms

and

dinoflagellates

communities

were

quite

manifold,

which

produced serious

difficulties for judging the similar degree among different diatoms (or dinoflagellates) communities. So, Canonical Discriminant Analysis (CDA) was used in this paper. Linear functions of quantitative variables yielded by CDA allowed maximal separation of two or more groups of individuals, while minimizing variation within groups (Manly, 2005). Cell abundances in genera level other than species level were used as variables in CDA, because not all of diatoms (or dinoflagellates) could be identified to the species in our study. In light of cell abundances widely varied among several orders of magnitude, so all of cell abundances data applying to CDA had been logarithmically transformed using equation logl0(x+1) in order to meet the assumptions of CDA. The general form of non-standardization of discriminant function built by CDA is as follow:

⎧ di1 = b01 + b11 Ai1 + b 21 Ai 2 + " + bj1 Aij + " + bs1 Ais ⎪ di 2 = b 02 + b12 Ai1 + b 22 Ai 2 + " + bj 2 Aij + " + bs 2 Ais ⎪ ⎪" ⎨ ⎪ dik = b 0 k + b1kAi1 + b 2 kAi 2 + " + bjkAij + " + bskAis ⎪" ⎪ ⎩ dir = b 0 r + b1rAi1 + b 2 rAi 2 + " + bjrAij + " + bsrAis

(1)

Where i is sample (i=1-n); j is genera (j=1-s); k is function (k=1-r); Aij is abundance; b0k is constant; bjk is coefficient; dik is canonical discriminant score.

3. Results 3.1. Monsoonal hydrodynamic conditions The coastal Ekman divergence/convergence cycle occurring in east coast of Hainan Island could be distinctly discriminated according to the thermohaline status of shelf water during two cruises. The slackness of both isothermal and isohaline in the surface thermohaline field (Figure 2 A and C) along with the uprising of the high-density bottom water in coastal water column during the SWM cruise (Figure 2B and D) indicated the existence of Ekman divergence phenomena (i.e. offshore diffusion of surface water and the coastal upwelling). The opposite phenomena was occurred at shelf water during the NEM cruise, where the denseness of both isothermal and isohaline in the surface thermohaline fields (Figure 2 E and G) along with the sinking of the low-density surface water in coastal water column (Figure 2F and H) indicated the existence of Ekman convergence phenomena (i. e. onshore aggregation of surface water and the coastal downwelling). The high-temperature and high-salinity sGCC from low altitude ocean (see Figure 1A) was difficult to be identified in the surface thermohaline fields of shelf water during the SWM cruise, because it was smoothened by the low- temperature coastal upwelling water (see Figure 2 A). However, the low-temperature and low-salinity nGCC from high altitude South China mainland coast (see Figure 1A) was significant in the surface thermohaline fields of shelf water during the NEM cruise, with the nGCC constrained on the shore by the Ekman convergence and formed steep thermohaline fronts with shelf water (see Figure 2 E and G). Seasonal stratification phenomena of tropical South China Sea (Su, 2004) also was observed at the offshore water column during our study. Summer stratification occurs in the SWM cruise (see Figure 2B and D) and spring transitional stratification at the NEM cruise (see Figure 2F and H). The temperature variation of the offshore stratified water column reached to 11°C during the SWM cruise and 2.5°C during the NEM cruise, which suggested that effect of winter vertical mixing remained in the spring transitional stratification during the NEM cruise. Briefly the general hydrological phenomena along the north shelf of South China Sea, including the monsoonal hydrodynamic processes (coastal Ekman divergence/convergence cycle and coastal current sGCC/nGCC reverse) and offshore water column structures (summer stratification and spring transitional stratification), were also observed at the open shelf of eastern Hainan Island coast during two cruises of this study. 3.2. Water-type division and nutrients conditions The whole investigated water bodies during the SWM cruise was divided into 5 water-types (Table 1, Figure 3A): lagoon waters (group a), water overlying fringing reefs (group b), coastal upwelling water (group c), non bottom waters (group d) and bottom waters (group e) of offshore shelf. The whole investigated water bodies during the NEM cruise was also divided into 5 water-types: lagoon waters (group f), water overlying fringing reefs (group g), coastal nGCC water (group h), upper mixed-layer water (group i) and lower mixed-layer water (group j) of offshore shelf. Lagoon waters (group a, group f) in both the cruises were found nutrient rich (Table 1, Figure 3B to D). Nutrients conditions in offshore shelf water column were well stratified during the SWM cruise, with oligotrophic status at non bottom

water (group d) but nutrients rich in bottom water (group e); however, nutrients conditions were vertically uniform during the NEM cruise with relative low nutrients concentrations observed in both of the upper mixed-layer water (group i) and the lower mixed-layer water (group j). Nutrients concentrations of the coastal water bodies were always found consistent between the reefs overlying water and coastal shelf water, with nutrient low in coastal water bodies (group b and c) during the SWM cruise but nutrient rich in coastal water bodies (group g and h) during the NEM cruise. 3.3. Diatoms/dinoflagellates assemblages during the SWM cruise 3.3.1. Cell abundance of diatoms/dinoflagellates Diatom cells were sparse in non bottom water of offshore shelf (group d), with regional average abundance of 0.03×103 cells L-1; however, diatom cells were flourishing in the rest 4 water-types, with regional average abundance respectively 84.9×103 cells L-1 of lagoon waters (group a), 34.2×103 cells L-1 of reefs overlying water (group b), 62.7×103 cells L-1 of upwelling water (group c) and 4.3×103 cells L-1 of offshore bottom water (group e) (Table 2, Figure 4A). Dinoflagellate cells were abundant in lagoon waters (group a), with regional average abundance of 25.1×103 cells L-1; however, dinoflagellate cells were sparse in the whole open sea water (group b, c, d, e), with regional average abundance all lower than 103 cells L-1 (Table 2, Figure 4B). As a result, the ratio between diatom and dinoflagellate abundances were very high (>50) in all of reefs overlying water (group b), upwelling water (group c), and bottom water of offshore shelf (group e) (Table 2). 3.3.2. Community structure of diatoms/dinoflagellates The CDA analysis for diatom (or dinoflagellate) communities of 5 water-types yielded 4 discriminant functions (Table 3). The main discriminant function (Function 1, explained for 86.1% of the total variance) of CDA analysis for diatom communities divided diatom communities of 5 water-types into 2 major clusters: diatom community of lagoon waters (group a) was one cluster, and ones of the rest 4 water-types (group b, c, d, e) was another cluster (Figure 5A). The main discriminant function (Function 1, explained for 86.7% of the total variance) of CDA analysis for dinoflagellate communities also divided dinoflagellate communities of 5 water-types into 2 major clusters: dinoflagellate community of lagoon waters (group a) was one cluster, and ones of the rest 4 water-types (group b, c, d, e) was another cluster (Figure 5B). Diatom community of lagoon waters (group a) was dominated by estuarine species (Table 4), and the main genera included 5 genera: Melosira, Cyclotella, Leptocylindrus, Cylindrotheca and Skeletonema, which together contributed for 91.4% of total diatom abundances of lagoon waters. Diatom communities all of the rest 4 water-types (group b, c, d, e) were dominated by chain-forming diatoms, and the main chain-forming diatom genera included 9 genera: Pseudo-nitzschia, Thalassionema, Chaetoceros, Thalassiosira, Thalassiothrix, Rhizosolenia, Climacodium, Cerataulina and Eucampia, which together contributed for above 82% of total diatom abundances of each of 4 water type (group b, c, d, e). All of 3 chain-forming diatoms genera Thalassionema, Thalassiothrix and Cerataulina identified only one species, respectively T. nitzschioides, T. frauenfeldii and C. pelagica. All of 5 chain-forming diatoms genera Pseudo-nitzschia, Chaetoceros, Rhizosolenia, Climacodium and Eucampia could identified more than one species, but dominant species all of which were almost consistent among the 4 water-types (group b, c, d, e), respectively Pseudo-nitzschia sp1, C. diversus, R. bergonii, C. frauenfeldianum and E. cornuta. Dinoflagellate community of lagoon waters (group a) was dominated by estuarine species (see Table 4), and the main genera included 3 genera: Glenodinium, Gymnodinium and Gonyaulax, which together contributed for 91.5% of total dinoflagellate abundance in lagoon waters. Dinoflagellate communities in the rest 4 water-types (group b, c, d, e) were dominated by marine dinoflagellates, and the main genera included 4 genera: Protoperidinium, Ceratium, Gyrodinium and Prorocentrum, which together contributed for 91.7% of total dinoflagellate abundance in reefs overlying water (group b), 95.5% in upwelling water (group c), 88.0% in non bottom water (group d) and 61.1% in bottom water (group e).

In short, both of diatom and dinoflagellate communities in lagoons were flourishing and dominated by estuarine species. Diatom communities in whole open sea area from fringing reefs to offshore shelf were dominated by chain-forming diatoms, with high abundance occurring at reefs overlying water, coastal upwelling water and offshore bottom water, but sparse in offshore non bottom water. Dinoflagellate communities in whole open sea area were sparse and dominated by marine dinoflagellates. 3.4. Diatoms/dinoflagellates assemblages during the NEM cruise 3.4.1. Cell abundances of diatoms/dinoflagellates Diatom cells were relatively abundant in both lagoon waters (group f) and reefs overlying water (group g), with regional average abundance respectively reaching to 11.9×103 cells L-1 and 0.8×103 cells L-1; similarly, dinoflagellate cells were also abundant in both lagoon waters (group f) and reefs overlying water (group g), with regional average abundance respectively reaching to 1.6×103 cells L-1 and 8.9×103 cells L-1. However, both of diatom and dinoflagellate cells were sparse in the whole shelf water involving the nGCC water (group h), offshore shelf water (group i and j), with regional average abundance all lower than 103 cells L-1 (Figure 6A, Table 5). As a result, the ratio between diatom and dinoflagellate abundances were low in the whole open sea water (group g, h, i, j) (Table 5). 3.4.2. Community structures of diatom/dinoflagellate The CDA analysis for diatom (or dinoflagellate) communities of 5 water-types yielded 4 discriminant functions (Table 6). The main discriminant function (Function 1 and Function 2, respectively explained for 75.2% and 17.6% of the total variance) of CDA analysis for diatom communities divided diatom communities of 5 water-types into 3 major clusters: diatom community of lagoon waters (group f) was independently one cluster, ones both of nGCC water (group h) and lower mixed-layer water (group j) was one cluster, and ones both of reefs overlying water (group g) and upper mixed-layer water (group i) was one cluster (Figure 7A). The main discriminant function (Function 1 and Function 2, respectively explained for 53.5 % and 35.9% of the total variance) of CDA analysis for dinoflagellate communities also divided dinoflagellate communities of 5 water-types into 3 major clusters: dinoflagellate community of lagoon waters (group f) was independently one cluster, ones both of reefs overlying water (group g) and nGCC water (group h) was one cluster, ones in offshore water column (group i and j) was one cluster (Figure 7B). Diatom community of lagoon waters (group f) was dominated by estuarine species (Table 7), and the main estuarine genera included 3 genera: Cylindrotheca, Cyclotella and Melosira, which together could contributed for 66.0% of total diatom abundance in lagoon waters. Diatom communities both of nGCC water (group h) and lower mixed-layer water (group j) were dominated by benthic diatoms, and the main benthic diatom genera included 9 genera: Pleurosigma, Paralia, Navicula, Bacillaria, Coscinodiscus, Diploneis, Biddulphia, Surirella and Actinoptychus, which together contributed for about 84.8% of total diatom abundance in nGCC water (group h) and 66.7% in lower mixed-layer water (group j). Last, diatom communities both of reefs overlying water (group g) and upper mixed-layer water (group i) were dominated by chain-forming diatoms. The main chain-forming diatom genera in upper mixed-layer water (group i) included 5 genera: Chaetoceros, Hemiaulus, Thalassiothrix, Thalassiosira and Rhizosolenia, which together contributed for 100% of total diatom abundance. The main chain-forming diatom genera in coral overlying water (group g) included 3 genera: Pseudo-nitzschia, Thalassionema and Thalassiosira, which together contributed for 47.7% of total diatom abundance. Dinoflagellate community of lagoon waters (group f) was dominated by estuarine species (see Table 7), and the main estuarine species dinoflagellate genera included Glenodinium and Peridinium, which together could contributed for 99.0% of

total dinoflagellate abundances in lagoon waters (group f). Dinoflagellate communities both of reefs overlying water (group g) and nGCC water (group h) were dominated by marine dinoflagellates, and the main dinoflagellate genera involved 2 genera Noctiluca and Prorocentrum, which together contributed for 94.6% of total dinoflagellate abundances in reefs overlying water (group g) and 61.2 %

in nGCC water (group h). Noctiluca could identify only the big cell (200-600 μm)

heterotrophic dinoflagellate N. scintillans, which could form the discontinuous and pink bloom patches in fringing reefs, with average abundance of 2,077 cells L-1. Prorocentrum could identified more than one species, with P. triestinum as the dominant species in reefs overlying water (group g) and P. micans as the dominant species in nGCC water (group h). Last, dinoflagellate communities in offshore water column (group i and j) were dominated by marine dinoflagellates, and the main dinoflagellates genera included 4 genera: Ceratium, Protoperidinium, Alexandrium and Gyrodinium, which together contributed for 74.4% of total dinoflagellate abundances in upper mixed-layer water (group i), and 76.9% in lower mixed-layer water (group j). In short, both of diatom and dinoflagellate communities in lagoons were flourishing and dominated by estuarine species. Diatom communities in whole open sea area from fringing reefs to offshore shelf were relatively low-abundance, with chain-forming diatoms dominating in both reefs overlying water and offshore upper mixed-layer water, but benthic diatoms dominating in both coastal nGCC water and offshore lower mixed-layer water. Dinoflagellate communities in both reefs overlying water and coastal nGCC water was highly in abundance and could detected big cell heterotrophic dinoflagellate N. scintillans, but ones of offshore shelf water were low-abundance and failed to detected N. scintillans cells. 3.5. Cyanobacteria communities In contrast with diatom/dinoflagellate communities, compositions of cyanobacteria communities were quite simple during both of two cruises. Cyanobacteria communities in lagoons always were composed of freshwater species mainly including Merismopedia, Oscillatoria and Anabaena, in which Merismopedia always was the main dominating genera, and respectively reached to 335.2×103 cells L-1 in the SWM cruise and 19.2×103 cells L-1 in the NEM cruise (Table 8). Cyanobacteria communities in the whole open water bodies (from fringing reefs to offshore shelf) during both of two cruises always observed oceanic species Trichodesmium erythraeum. T. erythraeum mainly occurred at upwelling water (group c) and non bottom water of offshore shelf (group d) during SWM cruise, with regional average abundance respectively reaching to 116 trichomes L-1 and 82 trichomes L-1; however, T. erythraeum mainly occurred at reefs overlying water (group g) and nGCC water (group h) during the NEM cruise, with regional average abundance respectively reaching to 1036 trichomes L-1 and 12 trichomes L-1 (Figure 8, Table 8).

4. Discussion 4.1. Monsoonal hydrodynamic processes during two cruises and its relationship with phytoplankton assemblages The general monsoonal hydrodynamic phenomena (coastal Ekman divergence/convergence cycle and coastal current sGCC/nGCC reverse) in the north shelf of South China Sea (Hu et al., 2000; Yang et al., 2002) were observed in east coast of Hainan Island during two cruises of this study. Each of these monsoonal hydrodynamic phenomena has specific feature of its own in relationship with phytoplankton (Table 9). The vertical-dimension coastal Ekman upwelling occurring in eastern Hainan Island coast during SWM cruise could nourish the flourishing chain-forming diatoms, because it is a very common phenomenon of the world that coastal upwelling ecosystems often triggers the thriving of chain-forming diatoms but suppresses the reproduction of shear-sensitive dinoflagellates (Kudela et al., 2005; Estrada & Blasco, 1985; Silva et al., 2009; Smayda & Reynolds, 2003). On the contrary, coastal Ekman downwelling occurring in NEM cruise was not benefit for the prosperity of phytoplankton, as it can not obtain nutrients supply from deep water just as upwelling performed (Ryan et al., 2009; Gonzalez-Rodriguez et al., 1992); however, coastal downwelling caused coastal water bodies possess the relatively long residence time, so as to be benefit to the

reproduction of phytoplankton by use of local inputs of nutrients (i. e. river-borne nutrients) (Yin, 2003). The horizontal-dimension offshore Ekman transport process occurring in eastern Hainan Island coast during SWM cruise can entrain coastal pelagic algal to offshore, but onshore Ekman transport process occurring in NEM cruise in turn can entrain offshore pelagic algal to coastal region of eastern Hainan Island coast and further aggregate along coastline which just located at dynamic boundary of onshore Ekman transport process, just as ones in other sea regions of world (Pickett & Paduan, 2003; Yin, 2003). Last, the large-scale coastal current sGCC occurring in SWM cruise can transport the oceanic phytoplankton assemblages (i. e. abundant T. erythraeum and sparse diatoms/dinoflagellates) from the oligotrophic SCS basin (Ning et al., 2004; Chen et al., 2003) to east coast of Hainan Island (see Figure 1A), but the nGCC occurring in NEM cruise in turn can transport the coastal phytoplankton assemblages (often blooms) of the eutrophic South China mainland coast (Lu & Hodgkiss, 2004; Wang et al., 2008; Zhang & Su, 2006) to east coast of Hainan Island. 4.2. Construction of monsoonal hydrodynamic processes on phytoplankton assemblages in coastal shelf Monsoonal hydrodynamic processes almost adequately controlled hydrodynamic conditions of coastal shelf in eastern Hainan Island coast during SWM (or NEM) cruise, with coastal Ekman upwelling (or downwelling) controlling vertical habitats of coastal shelf, and the combination of the sGCC (or nGCC) and offshore (or onshore) Ekman transport process controlling the horizontal-dimension flow fields (Table 10, Fig. 9-10). So, taking into consideration together relationships between monsoonal hydrodynamic processes of coastal shelf and phytoplankton (see Table 9) well explains construction mechanism of phytoplankton assemblage in coastal shelf. Phytoplankton assemblage features in coastal shelf during SWM cruise were that diatom community was thriving (average 5

10 cell L-1) with chain-forming diatoms dominating, but dinoflagellate community was failing (average 102 cell L-1); oceanic species cyanobacteria T. erythraeum widespread appeared in relatively high abundance (average 116 trichomes L-1) (Table 10). Taking into consideration relationships between monsoonal hydrodynamic processes of coastal shelf during SWM cruise and diatoms (see Table 9), the flourishing chain-forming diatoms occurring at coastal shelf during SWM cruise obviously should be nourished by the vertical-dimension upwelling habitat and almost had nothing to do with the horizontal transport processes (i. e. the sGCC and offshore Ekman transport). Indeed, the coastal upwelling ecosystem of the SWM cruise was quite typical. On one side, the nutrient-enriched shelf bottom water (DIN~8.3 μM, DIP~0.2 µM and H4SiO4~9.8 µM) (see Figure 3B-D) could provide the continuous nutrients supply to coastal upwelling, although nutrients concentration of upwelling water were at the oligotrophic level (DIN~3.0 μM, DIP~0.1 µM and H4SiO4~3.3 µM) due to the rapid uptake and transformation by the chain-forming diatoms. on the other side, the main members of chain-forming diatoms in coastal upwelling of the SWM cruise included Pseudo-nitzschia spp., Thalassionema nitzschioides, Chaetoceros spp., Thalassiosira spp., Thalassiothrix frauenfeldii, Rhizosolenia spp., etc., which were quite similar to ones often observed in other upwelling ecosystems of the world (Silva et al., 2009; Lassiter et al., 2006; Pitcher & Nelson, 2006; Kudela et al., 2005). Furthermore, oceanic species T. erythraeum appeared in coastal shelf of the SWM cruise obviously should be carried from the SCS basin by the sGCC. The trichomes abundance (average 116 trichomes L-1) of T. erythraeum in coastal shelf was close to one (200-280 trichomes L-1) of the summer SCS basin (Chen et al., 2003), which suggested that the losses of pelagic T. erythraeum was relatively small in the long-term transport from the SCS basin to the coastal shelf of east coast of Hainan Island. Last, taking into consideration relationships between monsoonal hydrodynamic processes of coastal shelf during SWM cruise and dinoflagellates (see Fig. 9), it was reasonable that dinoflagellate cells were very sparse in coastal shelf. Phytoplankton assemblage features in coastal shelf during NEM cruise were that both of diatom and dinoflagellate communities were low abundance (average 102 cells L-1), but the big cell (200-600 µm) heterotrophic dinoflagellate N. scintillans was detected with average abundance of 111 cells L-1; oceanic species T. erythraeum widespread appeared with relatively low abundance of 12 trichomes L-1 (Table 10). Taking into consideration relationships between monsoonal hydrodynamic processes of coastal shelf during NEM cruise and phytoplankton (see Table 9), the heterotrophic dinoflagellate N. scintillans cells occurring at coastal shelf during NEM cruise obviously could be entrained by the nGCC from the South China

mainland coast and almost had nothing to do with both of coastal downwelling and onshore Ekman transport, because coastal downwelling was not benefit the proliferation of local phytoplankton so as not to provide sufficient food for heterotrophic dinoflagellate N. scintillans, and the onshore Ekman transport process can not entrain N. scintillans cells from offshore to coast as N. scintillans was absent in offshore shelf (see Table 7). Indeed, N. scintillans is a common species at the eutrophic South China mainland coast where the flourishing phytoplankton cells nourished by sufficient nutrients provides the preferred food source for the proliferation of heterotrophic N. scintillans (Qi et al., 1993; Yin et al., 1999; Wang et al., 2008). In fact, just at February 2009, one month before the NEM cruise (March 2009), the extensive N. scintillans bloom was observed at the Pearl River estuary (see Figure 1A) along the South China mainland coast (South China Sea Bulletin, State Oceanic Administration of China, 2009). Furthermore, oceanic species T. erythraeum appeared in coastal shelf of the NEM cruise could be carried not only from offshore shelf of eastern Hainan Island coast by onshore Ekman transport, but also from the South China mainland coast by the nGCC, because the onshore Ekman transport process occurring at the whole north SCS during the NEM season can entrain the pelagic T. erythraeum of oceanic basin to both South China mainland coast and eastern Hainan Island coast. Last, it was reasonable that both of diatom and dinolagellate (other than N. scintillans) cells were very sparse in coastal shelf, as N. scintillans can exert predation pressure on small sized diatom/dinoflagellate cells (Fukuda & Endoh, 2006; Nakamura, 1998). In all, monsoonal hydrodynamic processes almost adequately controlled hydrodynamic conditions of coastal shelf in eastern Hainan Island coast during two cruises, and the opposite monsoonal hydrodynamic processes between the SWM and NEM cruises constructed the different phytoplankton assemblages in coastal shelf. Monsoonal hydrodynamic processes during SWM cruise constructed a phytoplankton assemblage characterized by the flourishing chain-forming diatoms and T. erythraeum but sparse dinoflagellate, mainly via the nourishment of chain-forming diatoms by coastal Ekman upwelling, and the entrainment of pelagic T. erythraeum by the large-scale coastal current sGCC from the low-latitude SCS basin. Monsoonal hydrodynamic processes during NEM cruise constructed a phytoplankton assemblage mainly containing pelagic cells (i. e. N. scintillans and T. erythraeum), mainly via the entrainment of N. scintillans by the large-scale coastal current nGCC from the high-latitude South China mainland coast, and the entrainment of T. erythraeum by onshore Ekman transport process from offshore shelf of eastern Hainan Island coast. 4.3. Construction of monsoonal hydrodynamic processes on phytoplankton assemblages in fringing reefs Monsoonal hydrodynamic processes almost adequately controlled hydrodynamic conditions of fringing reefs just as ones of coastal shelf, because the impact of insular runoffs on fringing reefs were obviously quite limited during both of two cruises, with physico-chemical variable (i. e. temperature, salinity and nutrients) of reefs overlying water almost consistent to ones of coastal shelf water (see Figure 3A to D). Furthermore, fringing reefs along the coastline just located at dynamic boundary of coastal Ekman divergence (or convergence) (Table 10, Fig. 9-10); so, phytoplankton assemblage features in fringing reefs should be somewhat different to ones in coastal shelf, although the construction of phytoplankton assemblage in fringing reefs was mainly controlled by monsoonal hydrodynamics just as ones in coastal shelf (see Table 9). Phytoplankton assemblage features in fringing reefs during SWM cruise were that diatom/dinoflagellate communities were quite similar to ones of coastal shelf, with chain-forming diatoms thriving (average 104 cell L-1) but dinoflagellates cells were sparse (average 102 cell L-1); however, abundance of T. erythraeum was far lower than ones of coastal shelf (Table 10). Fringing reefs was just located at the dynamic boundary of coastal Ekman divergence during SWM cruise, so, both of offshore Ekman diffusion of surface water and the next coastal Ekman upwelling were more significant in fringing reefs than ones in coastal shelf. As a result, offshore diffusion of surface water lead to the sGCC (entraining the pelagic T. erythraeum) far away from fringing reefs, and coastal upwelling entrained chain-forming diatoms of coastal shelf into fringing reefs; consequently, structure features of phytoplankton assemblage in fringing reefs during SWM cruise were characterized by the flourishing chain-forming diatoms and the sparse dinoflagellates and T. erythraeum, so as to be somewhat different to ones of coastal shelf. Phytoplankton assemblage in fringing reefs during NEM cruise was somewhat different to ones of coastal shelf (Table 10).

On one side, the abundances of pelagic algal cells (both of N. scintillans and T. erythraeum) steep increased from coastal shelf (respectively 111 cells L-1 and 12 trichomes L-1) to fringing shelf (respectively 2,077 cells L-1 and 1,036 trichomes L-1); on the other side, diatoms abundance (average 103 cells L-1) and dinoflagellates (other than N. scintillans) abundance (average 104 cells L-1) in fringing reefs were clearly higher than ones in coast shelf. Fringing reefs was just located at the dynamic boundary of coastal Ekman convergence during NEM cruise, so, both of onshore Ekman aggregation of surface water and the next coastal Ekman downwelling were more significant in fringing reefs than ones in coastal shelf. As a result, onshore aggregation of surface water lead to the nGCC (entraining the pelagic N. scintillans and T. erythraeum) pile at fringing reefs so as to be fit to the aggregation of pelagic cells along the coastline, and coastal downwelling caused reefs overlying water to possess the relatively long residence time so as to be benefit to the reproduction of phytoplankton by use of local inputs of nutrients (Yin, 2003); consequently, structure features of phytoplankton assemblage in fringing reefs during NEM cruise were characterized by the high abundance of pelagic cells (such as N. scintillans and T. erythraeum) and the relatively more abundant diatoms/dinoflagellate communities in fringing coral reefs than ones of coastal shelf, so as to be somewhat different to ones of coastal shelf. In all, monsoonal hydrodynamic processes also adequately constructed phytoplankton assemblages of fringing reefs just as ones of coastal shelf, but phytoplankton assemblages of fringing reefs were somewhat different from ones of coastal shelf, as fringing reefs along the coastline just located at dynamic boundary of coastal Ekman divergence (or convergence). Difference of phytoplankton assemblage between fringing reefs and coastal shelf mainly embodied that the pelagic T. erythraeum exhibited lower abundance in fringing reefs than ones in coastal shelf during SWM cruise, because offshore Ekman transport entrained pelagic T. erythraeum far away from coastline; the pelagic N. scintillans and T. erythraeum exhibited higher abundance in fringing reefs than ones in coastal shelf during NEM cruise, because onshore Ekman transport aggregated pelagic N. scintillans and T. erythraeum at fringing reefs along coastline. 4.4. Construction of monsoonal hydrodynamic processes on phytoplankton assemblages in offshore shelf Monsoonal hydrodynamic processes only partially controlled hydrodynamic conditions of offshore shelf in eastern Hainan Island coast during SWM (or NEM) cruise, mainly by offshore (or onshore) Ekman transport processes affecting horizontal flow fields of offshore shelf (Table 10, Fig. 9-10). So, construction of monsoonal hydrodynamic processes on phytoplankton assemblages in offshore shelf was obviously limited. Phytoplankton assemblage features in offshore shelf during SWM cruise were that phytoplankton assemblages in the euphotic layer possessed sparse diatoms/dinoflagellates (average 101 cells L-1) but relatively abundant diazotroph T. erythraeum (82 trichomes L-1); however, phytoplankton assemblages in bottom water abnormally possessed the high abundance of chain-forming diatoms (average 103 cells L-1) (Table 10). Phytoplankton assemblage (sparse diatoms/dinoflagellates and abundant T. erythraeum) in the euphotic layer of offshore shelf was obviously compatible with its oligotrophic status (DIP~0.07 µM and H4SiO4~2.6 µM) (see Figure 3B-D) creating by summer stratification of water column. Furthermore, in views of relationships between monsoonal hydrodynamics process (offshore Ekman transport) of offshore shelf and phytoplankton (see Table 9), the formation mechanism of the abnormal high abundance of chain-forming diatoms in bottom water of offshore shelf should be related to offshore Ekman transport process. That is, the lush chain-forming diatoms of coastal shelf was entrained to offshore shelf by offshore Ekman transport, and then sank out of upper layer and accumulated at bottom water, so as to form the high abundance of chain-forming diatoms at bottom water (see Fig. 9). Indeed, the main members of chain-forming diatoms of offshore bottom water were quite similar to ones of coastal upwelling (see Table 4), which further confirmed the consanguinity of chain-forming diatoms between coastal and offshore. Phytoplankton assemblage in offshore shelf during NEM cruise contained sparse diatoms/dinoflagellates (average 102 cells L-1) and T. erythraeum, which was just similar to previous observation results obtained in the SCS basin (Ning et al., 2004; Chen et al., 2003). Moreover, diatom community in the most water bodies of offshore shelf was dominated by benthic diatoms (Figure 10), which perhaps resulted from resuspension of benthic diatom by the remained winter vertical turbulence in spring

water column. So, the construction of phytoplankton assemblages of offshore shelf completely controlled by spring water column structure, and almost had nothing to do with onshore Ekman transport process. In all, only the horizontal-dimension offshore (or onshore) Ekman transport processes could affected hydrodynamic conditions of offshore shelf, so the control of monsoonal hydrodynamic processes on phytoplankton assemblage construction of offshore shelf were quite limited. During SWM cruise, offshore Ekman transport entrained coastal chain-forming diatoms to offshore shelf, so as to affect phytoplankton assemblage construction of offshore shelf; however, during NEM cruise, monsoonal hydrodynamics failed to involve the construction of phytoplankton assemblages of offshore shelf.

5. Conclusion (1) Monsoonal hydrodynamic processes (coastal Ekman divergence/convergence cycle and coastal current sGCC/nGCC reverse) almost adequately constructed phytoplankton assemblages of coastal region (including fringing reefs and coastal shelf) in eastern Hainan Island coast during both two cruises. The opposite monsoonal hydrodynamic processes between the SWM and NEM cruises constructed the different phytoplankton assemblages in coastal region, with chain-forming diatoms dominating during SWM cruise but the pelagic N. scintillans and T. erythraeum dominating during NEM cruise. (2) Furthermore, phytoplankton assemblages in fringing reefs were somewhat different from ones of coastal shelf, because fringing reefs is along coastline and located at dynamic boundary of offshore (or onshore) Ekman transport processes. So, the pelagic T. erythraeum exhibited lower abundance in fringing reefs than ones in coastal shelf during SWM cruise, because the offshore Ekman transport process entrained pelagic T. erythraeum far away from coastline; the pelagic N. scintillans and T. erythraeum exhibited higher abundance in fringing reefs than ones in coastal shelf during NEM cruise (especially, N. scintillans formed bloom in fringing reefs), because onshore Ekman transport aggregated pelagic N. scintillans and T. erythraeum at fringing reefs along coastline. (3) The monsoonal hydrodynamic processes only partially affected hydrodynamic conditions of offshore shelf, so the control of monsoonal hydrodynamics on phytoplankton assemblage construction of offshore shelf was also quite limited.

Acknowledgements This work was funded by the Ministry of Science and Technology of China through contract Nos. 2007DFB20380 (LANCET, Land-Sea Interactions along Coastal Ecosystems of Tropical China: Hainan), by Natural Science Foundation of China through contract Nos. 40830850 (Impact of Anthropogenic Perturbation and Dynamic Processes on the Sustainability of Coral Reefs from East Coast of Hainan). We thank G.S. Zhang, Z.Y. Zhu, C. Feng, J. Song, H.Y. Bao, and other colleagues from State Key Laboratory of Estuarine and Coastal Research of East China Normal University, and Marine Biogeochemistry Laboratory of Ocean University of China for the assistance in field observations, and thank L. Li, H.Q. Gu for the process of the wind field data. (Jin et al., 1985; Jin et al., 1992; Tomas, 1997; Chen et al., 2003)

Reference:

(Pickett & Paduan, 2003)

Chen, Y. L., Chen, H. Y. & Lin, Y. H. (2003) Distribution and downward flux of Trichodesmium in the South China Sea as influenced by the transport from the Kuroshio Current. Marine Ecology Progress Series, 259, 47-57. Dela-Cruz, J., Middleton, J. H. & Suthers, I. M. (2003) Population growth and transport of the red tide dinoflagellate, Noctiluca scintillans, in the coastal waters off Sydney Australia, using cell diameter as a tracer. Limnology and Oceanography, 48, 656-674. Estrada, M. & Blasco, D. (1985) Phytoplankton assemblages in coastal upwelling areas. International Symposium on the Upwelling Areas off Western Africa (ed. by M.R. Bas C., Rubies P.), pp 379-402, Barcelona. Fukuda, Y. & Endoh, H. (2006) New details from the complete life cycle of the red-tide dinoflagellate Noctiluca scintillans (Ehrenberg) McCartney. European journal of protistology, 42, 209-219. Gonzalez-Rodriguez, E., Valentin, J. L., Andre, D. L. & Jacob, S. A. (1992) Upwelling and downwelling at Cabo Frio (Brazil): comparison of biomass and primary production responses. Journal of plankton research, 14, 289-306. Hu, J. Y., Kawamura, H., Hong, H. & Qi, Y. (2000) A review on the currents in the South China Sea: seasonal circulation, South China Sea Warm Current and Kuroshio intrusion. Journal of oceanography, 56, 607-624. Jin, D. X., Chen, J. H. & Huang, K. G. (1985) Planktonic Diatom of the China Sea, edn. Shanghai Science Technology Publisher, Shanghai. Jin, D. X., Cheng, Z. D., Lin, J. M. & Liu, S. C. (1992) Marine Benthic Diatoms in China, edn. China Ocean Press, Beijing. Keafer, B. A., Churchill, J. H., Mcgillicuddy, J. D. J. & Anderson, D. M. (2005) Bloom development and transport of toxic Alexandrium fundyense populations within a coastal plume in the Gulf of Maine. Deep Sea Research Part II: Topical Studies in Oceanography, 52, 2674-2697. Kudela, R., Pitcher, G., Probyn, T., Figueiras, F., Moita, T. & Trainer, V. (2005) Harmful algal blooms in coastal upwelling systems. Oceanography, 18, 184-197. Lassiter, A. M., Wilkerson, F. P., Dugdale, R. C. & Hogue, V. E. (2006) Phytoplankton assemblages in the CoOP-WEST coastal upwelling area. Deep Sea Research Part II: Topical Studies in Oceanography, 53, 3063-3077. Le, F. F., Sun, J., Ning, X., Song, S., Cai, Y. M. & Liu, C. (2006) Phytoplankton in the northern South China Sea in summer 2004. Oceanologia et Limnologia Sinica, 37, 238-248. Lee, J. H. Y., Qu, B. (2001) Hydrodynamic modelling and algal bloom tracking in Hong Kong`s coastal waters in March-April 1998. Hydraulics for Maritime Engineering, Proceedings of the 29th IAHR Congress (ed. by G. Li), Beijing. Li, R. H. (2010) Nutrient dynamics along coastal ecosystems of east Hainan. pp 162. Ocean University of China, Qingdao. Liu, S. M., Li, R. H., Zhang, G. L., Wang, D. R., Du, J. Z., Herbeck, L. S., Zhang, J. & Ren, J. L. (2011) The impact of anthropogenic activities on nutrient dynamics in the tropical Wenchanghe and Wenjiaohe Estuary and Lagoon system in East Hainan, China. Marine Chemistry, 125, 49-68. Lu, S. & Hodgkiss, I. J. (2004) Harmful algal bloom causative collected from Hong Kong waters. Hydrobiologia, 512, 231-238. Manly, B. F. J. (2005) Multivariate Statistical Methods: A Primer, edn. Chapman and Hall, London. Mcalice, B. J. (1971) Phytoplankton sampling with the Sedgwick-Rafter cell. Limnology and Oceanography, 16, 19-28. Nakamura, Y. (1998) Biomass, feeding and production of Noctiluca scintillans in the Seto Inland Sea, Japan. Journal of plankton research, 20, 2213. Ning, X., Chai, F., Xue, H., Cai, Y., Liu, C. & Shi, J. (2004) Physical-biological oceanographic coupling influencing phytoplankton and primary production in the South China Sea. Journal of Geophysical Research, 109, C10005. Pickett, M. H. & Paduan, J. D. (2003) Ekman transport and pumping in the California Current based on the US Navy's high-resolution atmospheric model (COAMPS). Journal of Geophysical Research, 108, 3327. Pitcher, G. C. & Nelson, G. (2006) Characteristics of the surface boundary layer important to the development of red tide on the southern Namaqua shelf of the Benguela upwelling system. Limnology and Oceanography, 51, 2660-2674. Qi, Y., Zhang, Z., Hong, Y., Lu, S., Zhu, C. & Li, Y. (1993) Occurrence of red tides on the coasts of China. Toxic Phytoplankton Bloom in the Sea (ed. by S.Y. Smayda T.J.), pp 43-46. Elsevier Science Publishers, New York. Ryan, J. P., Fischer, A. M., Kudela, R. M., Gower, J. F. R., King, S. A., Marin Iii, R. & Chavez, F. P. (2009) Influences of upwelling and downwelling winds on red tide bloom dynamics in Monterey Bay, California. Continental Shelf Research, 29, 785-795. Shalapyonok, A., Olson, R. J. & Shalapyonok, L. S. (2001) Arabian Sea phytoplankton during Southwest and Northeast monsoons 1995: composition, size structure and biomass from individual cell properties measured by flow cytometry. Deep Sea Research Part II: Topical Studies in Oceanography, 48, 1231-1261. Silva, A., Palma, S., Oliveira, P. B. & Moita, M. T. (2009) Composition and interannual variability of phytoplankton in a coastal upwelling region (Lisbon Bay, Portugal). Journal of Sea Research, 62, 238-249. Smayda, T. J. (2002) Turbulence, watermass stratification and harmful algal blooms: an alternative view and frontal zones as" pelagic seed banks". Harmful Algae, 1, 95-112. Smayda, T. J. & Reynolds, C. S. (2003) Strategies of marine dinoflagellate survival and some rules of assembly. Journal of Sea Research, 49, 95-106. Su, J. L. (2004) Overview of the South China Sea circulation and its influence on the coastal physical oceanography outside the Pearl River Estuary. Continental Shelf Research, 24, 1745-1760. Sun, J., Song, S. Q., Le, F. F., Wang, D., Dai, M. H. & Ning, X. R. (2007) Phytoplankton in northern South China Sea in the winter of 2004. ACTA OCEANOLOGICA SINICA, 29, 132-145. Tan, C. K., Ishizaka, J., Matsumura, S., Yusoff, F. M. & Mohame, M. H. (2006) Seasonal variability of SeaWiFS chlorophyll a in the Malacca Straits in relation to Asian monsoon. Continental Shelf Research, 26, 168-178. Tang, D., Kester, D. R., Ni, L. H., Qi, Y. & Kawamura, H. (2003) In situ and satellite observations of a harmful algal bloom and water condition at the Pearl River estuary in late autumn 1998. Harmful Algae, 2, 89-99. Tomas, C. R. (1997) Identifying Marine Phytoplankton, edn. Academic Press, San Diego. Tseng, C. M., Wong, G. T. F., Lin, I. I., Wu, C. R. & Liu, K. K. (2005) A unique seasonal pattern in phytoplankton biomass in low-latitude waters in the South China Sea. Geophysical Research Letters, 32, L08608. Wang, S. F., Tang, D., He, F., Fukuyo, Y. & Azanza, R. (2008) Occurrences of harmful algal blooms (HABs) associated with ocean environments in the South China Sea. Hydrobiologia, 596, 79-93. Wiggert, J. D., Murtugudde, R. G. & Mcclain, C. R. (2002) Processes controlling interannual variations in wintertime (Northeast

Monsoon) primary productivity in the central Arabian Sea. Deep Sea Research Part II: Topical Studies in Oceanography, 49, 2319-2343. Yang, H., Liu, Q., Liu, Z., Wang, D. & Liu, X. (2002) A general circulation model study of the dynamics of the upper ocean circulation of the South China Sea. Journal of Geophysical Research, 107, 3085. Yin, K. (2002) Monsoonal influence on seasonal variations in nutrients and phytoplankton biomass in coastal waters of Hong Kong in the vicinity of the Pearl River estuary. Marine Ecology Progress Series, 245, 111-122. Yin, K., Qian, P. Y., Wu, M. C. S., Chen, J. C., Huang, L. M., Song, X. Y. & Jian, W. J. (2001) Shift from P to N limitation of phytoplankton biomass across the Pearl River estuarine Plume during summer. Marine Ecology Progress Series, 221, 17-28. Yin, K. D. (2003) Influence of monsoons and oceanographic processes on red tides in Hong Kong waters. Marine Ecology Progress Series, 262, 27-41. Yin, K. D., Harrison, P. J., Chen, J., Huang, W. & Qian, P. Y. (1999) Red tides during spring 1998 in Hong Kong: is El Nino responsible? Marine Ecology Progress Series, 187, 289-294. Zhang, J. & Su, J. L. (2006) Nutrient dynamics of the China Seas: The Bohai Sea, Yellow Sea, East China Sea and South China Sea. The Sea: The Global Coastal Ocean Interdisciplinary Regional Studies and Syntheses, 14, 637–671.

Table 1 The division of water-types at study area, and the average concentrations (μM) of nutrients (DIN, DIP, and H4SiO4) at each water-type. Study area Variable SWM cruise Groups Water-typ e DIN (μM) DIP (μM) H4SiO4 (μM) n

Lagoons

Fringing reefs

Offshore shelf - upper layer

Coastal shelf

Offshore shelf - lower layer

a

b

c

d

e

Lagoon waters

Reefs overlying water

Upwelling water

Non-bottom water

Bottom water

15.7±14.4 0.2±0.2

2.3±0.5 0.1±0.1

3.0±1.8 0.1±0.1

3.3±2.1 0.07±0.04

8.3±4.9 0.2±0.2

109.0±95.7

2.8±1.5

3.3±2.8

2.6±0.9

9.8±3.9

22

13

16

26

13

h

i

j

nGCC water

Upper mixed-layer

Lower mixed-layer

4.1±0.9 0.2±0.1

1.0±0.6 0.1±0.02

1.6±0.9 0.1±0.03

4.8±0.3

3.1±0.9

3.7±1.4

12

11

23

NEM cruise Groups f g Water-typ Lagoon waters Reefs overlying water e DIN (μM) 51.6±42.3 3.3±1.2 DIP (μM) 0.6±0.3 0.2±0.2 H4SiO4 88.2±79.3 7.8±6.2 (μM) n 17 23 n: Number of samples employed in each group.

Table 2 Average abundances (103 cells L-1) of diatoms (or dinoflagellates) in each water-type during the SWM cruise, and the abundance ratio between diatoms and dinoflagelaltes.

Items Diatom abundances (103 cells L-1) Dinoflagellate abundances (103 cells L-1) Abundance ratio between diatoms and dinoflagellates

Group a 84.9±115.9 25.1±43.8 3.4

Group b 34.2±109.1 0.2±0.3 171.0

Water-types Group c 62.7±120.5 0.3±0.2 209.0

Group d 0.03±0.03 0.08±0.09 0.4

Group e 4.3±8.5 0.08±0.1 53.8

Table 3 Significance of the CDAs respectively for diatom communities and dinoflagellate communities during SWM cruise. Variable Diatom communities

Dinoflagellate communities

Functions 1 2 3 4

Eigenvalue

% of Variance

Cumulative %

55.142 3.974 2.818 2.084

86.1 6.2 4.4 3.3

86.1 92.3 96.7 100.0

Canonical Correlation 0.991 0.894 0.859 0.822

1 10.906 86.7 86.7 0.957 2 0.718 5.7 92.4 0.646 3 0.537 4.3 96.7 0.591 4 0.416 3.3 100.0 0.542 Note: the significance of individual function can be inferred from a large eigenvalue or a small P value.

P <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001

Table 4 Diatoms/dinoflagellates lists during the SWM cruise, and the average abundances (cell L-1) of each diatom (or dinoflagellate) species in each water-type. Diatoms/dinoflagellates genera were arranged according to the correlative degree with discriminant functions (see Table 3). Water-types Division Functions Genera Species Group a Group b Group c Group d Group e

Diatoms

1 1 1 1 1

Cyclotella Leptocylindrus Cylindrotheca Skeletonema Melosira

1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3 3 3 4 4

Synedra Campyloneis Fragilaria Amphora Tabellaria Striatella Licmophora Nitzschia Rhabdonema Diploneis Achnanthes Hemiaulus Guinardia Bacteriastrum Campylodiscus Cocconeis Planktoniella Pleurosigma Paralia Navicula Thalassionema Pseudo-nitzschia

4

Climacodium

4

Chaetoceros

4 4

Cerataulina Rhizosolenia

4 4 4

Thalassiothrix Thalassiosira Eucampia

4

Biddulphia

4 4 4 4 4 4 4

Asterionellopsis Lauderia Bacillaria Rhaphoneis Corethron Triceratium Asteromphalus

4 4

Coscinodiscus Surirella

(Continus Table 4)

C. stylorum L. danicus C. closterium S. costatum M. granulata M. moniliformis Synedra spp. C. grevillei F. capucina Amphora spp. T. fenestrata S. unipunctata Licmophora spp. Nitzschia spp. R. adriaticum Diploneis spp. Achnanthes spp. H. sinensis G. striata Bacteriastrum spp. Campylodiscus spp. Cocconeis spp. P. sol Pleurosigma spp. P. sulcata Navicula spp. T. nitzschioides Pseudo-nitzschia sp1* Pseudo- nitzschia sp2** C. biconcavum C. frauenfeldianum C. compressus C. danicus C. debilis C. distans C. diversus C. laevis C. lorenzianus C. messanensis C. peruvianus C. radicans Chaetoceros spp. C. pelagica R. alata f. gracillima R. alata f. indica R. bergonii R. calcar-avis R. cylindrus R. fragilissima R. imbricata R. setigera R. styliformis var. longispina T. frauenfeldii Thalassiosira spp. E. cornuta E. zodiacus B. mobiliensis B. reticulata A. glacialis L. annulata B. paxillifera Rhaphoneis spp. C. hystrix T. formosum A. arachne A. cleveanus A. elegans A. heptactis Coscinodiscus spp. S. hybrida S. qluminensis S. spiralis

32,620 26,411 11,727 813 5,967 16 46 27 23 11 11 11 3 32 155 1 49 123 101 6,060 425 1 5 75 5 11 9 4 1 7 28 1 1 2 11 4 65 4 30

5 32 8 46 68 8 18 23 5 3 2 2 6 17 35 1,095 27,823 165 18 8 15 2 35 3,858 11 20 31 12 378 12 2 34 2 2 6 2 5 168 2 3 18 74 142 5 5

9 33 3 9 15 4 13 7,301 47,884 3,186 31 109 36 100 36 1,338 5 26 1 233 153 1 1 95 5 28 3 521 911 30 25 43 79 325 25 8 50 1 3 1 1 1 13 4

4 3 3 7 8 3 1 1 -

3 2 5 29 40 11 777 2,643 317 2 6 6 3 3 2 6 2 103 238 40 3 17 2 3 17 -

Division

Functions

Genera

Species

Group a 4,735 G. gymnodinium Glenodinium Dinoflagellates 1 G. abbreviatum Gymnodinium 1 G. breve 16,346 G. sanguineum 1,358 G. spinifera Gonyaulax 1 11 D. acuta Dinophysis 1 342 S. trochoidea Scrippsiella 1 11 P. quinquecorne Peridinium 1 A. bidentata Amphisolenia 2 Triposolenia spp. Triposolenia 2 A. catenella Alexandrium 2 C. gourretii Ceratocorys 2 C. horrida 346 Protoperidinium Protoperidinium spp. 3 4 P. gracile Prorocentrum 3 33 P. lima P. mexicanum 9 P. micans 4 P. sigmoides C. deflexum Ceratium 3 1,556 C. furca 4 C. fusus C. horridum C. teres C. trichoceros Gyrodinium spp. Gyrodinium 4 316 O. sceptrum Oxytoxum 4 O. scolopax P. bipes Podolampas 4 P. palmipes *similar to Pseudo-nitzschia delicatissima; **similar to Pseudo-nitzschia pungens

Group b 2 8 6 2 122 3 8 8 2 2 51 3 3 5 2

Water-types Group c Group d 1 2 1 1 1 2 1 1 11 138 1 1 4 3 2 21 5 1 5 1 2 3 49 63 1 1 1 2 1 4

Group e 8 5 2 17 18 2 2 2 28 -

Table 5 Average abundances (103 cells L-1) of diatoms (or dinoflagellates) in each water-type during the NEM cruise, and the abundance ratio between diatoms and dinoflagelaltes. Items Diatom abundances (103 cells L-1) Dinoflagellate abundances (103 cells L-1) Abundance ratio between diatoms and dinoflagellates

Group f 11.9±10.8 1.6±1.6 7.4

Group g 0.8±1.3 8.9±19.8 0.1

Water-types Group h Group i 0.3±0.2 0.04±0.03 0.3±0.3 0.1±0.1 1.0 0.4

Group j 0.5±0.2 0.08±0.08 6.3

Table 6 Significance of the CDAs respectively for diatom communities and dinoflagellate communities during NEM cruise. Variable Diatom communities

Dinoflagellate communities

Function 1 2 3 4

Eigenvalue

% of Variance

Cumulative %

19.613 4.585 1.523 0.364

75.2 17.6 5.8 1.4

75.2 92.8 98.6 100.0

Canonical Correlation 0.975 0.906 0.777 0.516

1 4.045 53.5 53.5 0.895 2 2.713 35.9 89.3 0.855 3 0.686 9.1 98.4 0.638 4 0.120 1.6 100.0 0.327 Note: the significance of individual function can be inferred from a large eigenvalue or a small P value.

P <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001

Table 7 Diatoms/dinoflagellates lists during the NEM cruise, and the average abundances (cell L-1) of each diatom (or dinoflagellate) species in each water-type. Diatoms/dinoflagellates genera were arranged according to the correlative degree

with discriminant functions (see Table 6). Division

Functions

Diatoms

1 1 1

Cylindrotheca Cyclotella Chaetoceros

1 1 1 2 2 2

Melosira Cocconeis Cymbella Pleurosigma Thalassiothrix Biddulphia

2

Surirella

3 3

Diploneis Rhizosolenia

3 3 3 3 3 3 3 4

Navicula Pseudo-nitzschia Licmophora Campylodiscus Leptocylindrus Asterionellopsis Rhaphoneis Asteromphalus

4 4 4 4 4

Thalassiosira Bacillaria Climacodium Thalassionema Hemiaulus

4 4 4 4 4 4 4 1 1

Coscinodiscus Bacteriastrum Eucampia Paralia Nitzschia Streptotheca Actinoptychus Glenodinium Ceratium

1 1

Peridinium Oxytoxum

2 2

Noctiluca Prorocentrum

2 3 3 4 4 4 4 4

Scrippsiella Gyrodinium Podolampas Alexandrium Dinophysis Gymnodinium Amphisolenia Protoperidinium

Dinoflagellates

Genera

*similar to Pseudo-nitzschia delicatissima;

Species C. closterium C. stylorum C. debilis C. distans C. lorenzianus Chaetoceros spp. M. granulata Cocconeis spp. Cymbella spp. Pleurosigma spp. T. frauenfeldii B. aurita B. mobiliensis B. sinensis S. gemma S. qluminensis S. spiralis Diploneis spp. R. alata f. gracillima R. alata f. indica R. bergonii R. fragilissima Navicula spp. P. delicatissima* Licmophora spp. Campylodiscus spp. L. danicus A. glacialis Rhaphoneis spp. A. cleveanus A. heptactis Thalassiosira spp. B. paxillifera C. frauenfeldianum T. nitzschioides H. hauckii H. membranacus Coscinodiscus spp. Bacteriastrum spp. E. zodiacus P. sulcata Nitzschia spp. Streptotheca spp. A. senarius G. gymnodinium C. deflexum C. furca C. fusus C. horridum C. massiliense C. teres C. trichoceros C. tripos P. quinquecorne O. milneri O. scolopax N. scintillans P. dentatum P. micans P. triestinum S. trochoidea Gyrodinium spp. P. palmipes A. catenella D. caudata G. sanguineum A. bidentata Protoperidinium spp.

Group f 1,825 436 1,028 480 810 1,021 5,608 2 35 349 8 14 14 21 15 27 7 113 1 46 38 15 2 31,689 7 402 1 12 15 315

Group g 305 9 12 1 1 17 27 354 4 2 3 14 1 9 23 17 4 22 1 70 6 1 4 147 2,077 137 6,248 40 133 22 13 2 44

Water-types Group h Group i 83 9 15 2 7 5 3 12 13 2 2 2 33 5 10 30 5 8 5 20 37 2 2 18 47 16 18 4 2 7 4 5 3 2 4 3 111 2 2 13 63 2 2 22 2 9 24 15 5 10 2 13 3

Group j 160 30 3 8 9 3 15 3 2 3 3 48 5 1 23 3 24 45 3 3 21 1 3 34 2 1 2 18 10 1 1 1 2 2 3 3 3 2 15 10

Table 8 Cyanobacteria lists during two cruises, and the average abundances (cell L-1) of each cyanobacteria species in each water-type. Genera SWM cruise Anabaena Merismoped ia Oscillatoria Trichodesmi um

Species Anabaena spp. M. glauca M. elegans Oscillatoria spp. T. erythraeum

Water-types Group a

Group b

Group c

Group d

Group e

164 315,665

-

-

-

-

19,593 55 -

sparse

*9,300 (116)

*6,585 (82)

sparse

Group h

Group i -

-

Group j -

M. elegans 10,927 Oscillatoria Oscillatoria spp. 35 Trichodesmi T. erythraeum - *82,852 (1,036) *933 (12) um * Average abundance (trichomes L-1) of T. erythraeum was given in brackets.

Sparse

sparse

NEM cruise Merismoped ia

M. glauca

Group f 8,329

Group g -

Table 9 Monsoonal hydrodynamic processes during two cruises and its essential relationship with phytoplankton. Vertical-dimension Coastal Ekman upwelling SWM cruise Relationship with phytoplankton

References

NEM cruise Relationship with phytoplankton

References

Monsoonal hydrodynamic processes Horizontal-dimension Offshore Ekman transport Coastal current sGCC

Nourishing the flourishing chain-forming diatoms, but suppressing the proliferation of shear-sensitive dinoflagellates;

Entraining coastal pelagic algal to offshore regions;

Entraining oceanic phytoplankton assemblage (abundant T. erythraeum) from the oligotrophic SCS basin;

(Kudela et al., 2005; Estrada & Blasco, 1985; Silva et al., 2009; Smayda & Reynolds, 2003)

(Pickett & Paduan, 2003; Yin K. D., 2003)

(Ning et al., 2004; Chen et al., 2003)

Coastal Ekman downwelling (1) Failing to obtain nutrients of bottom water so as not to be benefit for the thriving of phytoplankton; (2) Causing coastal water bodies possess the long residence time so as to be benefit to phytoplankton growth by use of local inputs of nutrients (i. e. river-borne nutrients).

Onshore Ekman transport (1) Entraining offshore pelagic algal to coastal regions; (2) Aggregating pelagic algal cells along coastline which just located at the dynamic boundary of Onshore Ekman transport process;

Coastal current nGCC Entraining coastal phytoplankton assemblage (often blooms) from the eutrophic South China mainland coast;

(Ryan et al., 2009; Yin K. D., 2003; Gonzalez-Rodriguez et al., 1992)

(Pickett & Paduan, 2003; Yin K. D., 2003)

(Wang et al., 2008; Zhang & Su, 2006; Lu & Hodgkiss, 2004)

Table 10 Monsoon hydrodynamics conditions and phytoplankton assemblage features in open sea areas of the east Hainan Island coast respectively during the SWM and NEM cruise. Variable SWM cruise Mosoonal hydrodynamics Phytoplankton assemblages Diatoms community Dinoflagellates community Trichodesmium erythraeum NEM cruise Mosoonal hydrodynamics

Fringing coral reefs

Open sea areas of the east Hainan Island coast Coastal shelf

Dynamics boundary of coastal Ekman divergence Chain-froming, ~104 cells L-1 (i. e. Pseudo-nitzschia spp.) ~102 cells L-1 Sparse Dynamics boundary of coastal Ekman convergence

Offshore shelf

----Coastsl Ekman upwelling------------------Offshore Ekman transport of surface water-------------------Coastal current sGCC-----Chain-froming, ~105 cells L-1 (i. e. Pseudo-nitzschia spp.) ~102 cells L-1 116 trichomes L-1

*Chain-froming, ~103 cells L-1 (i. e. Pseudo-nitzschia spp.) **~102 cells L-1 ***82 trichomes L-1

---Coastsl Ekman downwelling------------------Onshore Ekman transport of surface water--------------------Coastal current nGCC-------

Phytoplankton assemblages Diatoms community Chain-froming, ~103 cells L-1 Bethic, ~102 cells L-1 *Bethic, ~102 cells L-1 Dinoflagellates community Noctiluca bloom, ~104 cells L-1 Noctiluca-present, ~102 cells L-1 **~102 cells L-1 Trichodesmium erythraeum 1036 trichomes L-1 12 trichomes L-1 ***Sparse * Lower layer of offshore water bottom; ** Whole offshore water column; *** Upper layer of offshore water bottom

Fig. 1. (A) The seasonally reversed Guangdong Coastal Current (GCC) between the sGCC in the SWM season and nGCC in the NEM season along the north shelf of the South China Sea (SCS) (after Hu et al., 2000). (B) Locations of sampling stations in shelf. (C) Locations of sampling stations in lagoons and fringing reefs (C). The year-round vector plots of the six-hourly wind velocities from June 2008 to June 2009 were detected by using the six-hourly wind vector data of NASA’s Quick Scatterometer (QuikSCAT) which is available online at http://podaac.jpl.nasa.gov/quikscat.

Fig. 2. The temperature (℃) and salinity (psu) structures of shelf water, respectively during the SWM cruise (A-D) and the NEM cruise (E-H). The station names see Fig. 1B.

Fig. 3. The T-S diagrams (A), DIN concentrations (μM) (B), DIP concentrations (μM) (C), and H4SiO4 concentrations (μM) (D) of 90 water samples in 5 water-types (group a, b, c, d, e) during the SWM cruise, and of 90 water samples in 5 water-types (group f, g, h, i, j) during the NEM cruise.

Fig. 4. Diatoms abundances (cells L-1) (A) and dinoflagellates abundances (cells L-1) (B) of 90 water samples in 5 water-types (group a, b, c, d, e) during the SWM cruise.

Fig. 5. Canonical Discriminant Analysis for diatoms communities (A) and dinoflagellates communities (B) of 5 water-types (group a, b, c, d, e) during the SWM cruise, based on abundance (cell L-1) respectively of 44 diatoms genera and 16 dinoflagellates genera (see Table 4) using variables.

Fig. 6. Diatoms abundances (cells L-1) (A) and dinoflagellates abundances (cells L-1) (B) of 90 water samples in 5 water-types (group f, g, h, i, j) during the NEM cruise.

Fig. 7. Canonical Discriminant Analysis for diatoms communities (A) and dinoflagellates communities (B) of 5 water-types (group f, g, h, i, j) during the NEM cruise, based on abundance (cell L-1) respectively of 32 diatoms genera and 14 dinoflagellates genera (see Table 7) using variables.

Fig. 8. Cyanobacteria abundances (cells L-1) of 180 water samples in 10 water-types (group a, b, c, d, e, f, g, h, i, j) during two cruises.

Fig. 9. Conceptual diagram about the construction of monsoonal hydrodynamic on phytoplankton assemblages near fringing reefs at the east coast of Hainan Island during the SWM cruise. Coastal Ekman upwelling nourished the diatoms/dinoflagellates assemblages dominated by chain-forming diatoms at the east coast of Hainan Island, and the sGCC entrained oceanic phytoplankton assemblage (abundanct T. erythraeum but sparse diatoms/dinoflagellates) of the low-altitude oligotrophic ocean to the east coast of Hainan Island. Furthermore, offshore Ekman transport process entrained the coastal phytoplankton assemblages to offshore shelf. This added up to phytoplankton assemblage of fringing reefs which was characteristics of abundance chain-forming diatoms but sparse dinoflagellates and T. erythraeum.

Fig. 10. Conceptual diagram about the construction of monsoonal hydrodynamic on phytoplankton assemblages near fringing reefs at the east coast of Hainan Island during the NEM cruise. The nGCC entrained coastal phytoplankton assemblages (big celled N. scintillans and sparse diatoms) of the eutrophic South China mainland coast to the east coast of Hainan Island, and onshore

Ekman

transport

process

entrained

offshore

phytoplankton

assemblage

(sparse

T.

erythraeum

and

diatoms/dinoflagellates) to the east coast of Hainan Island. Furthermore, the pelagic N. scintillans cells and T. erythraeum trichomes onshore aggregated and formed high abundances at fringing reefs along the coastline, so as to form the N. scintillans blooms.

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Fig. 10