Science of the Total Environment 698 (2020) 134289

Contents lists available at ScienceDirect

Science of the Total Environment

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

The diversity and biogeography of microeukaryotes in the euphotic zone

of the northwestern Pacific Ocean
Peng-Fei Wu a, Dong-Xu Li a, Ling-Fen Kong a, Yuan-Yuan Li a, Hao Zhang a, Zhang-Xian Xie a,
Lin Lin a, Da-Zhi Wang a,b,⁎
a
State Key Laboratory of Marine Environmental Science/College of the Environment and Ecology, Xiamen University, Xiamen, 361005, China
b
Key Laboratory of Marine Ecology & Environmental Sciences, Institute of Oceanology, Chinese Academy of Sciences, Qingdao, 266071, China

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Microeukaryotic communities in the

northwestern Pacific were investigated.
• The microeukaryotes were dominated
by Metazoa and Dinoflagellata.
• The microeukaryotic biogeography
were driven mainly by environmental
factors.
• Temperature was the most important
environmental factor.
• Bacteria presented biotic potential of
shaping microeukaryotic biogeography.

a r t i c l e i n f o a b s t r a c t

Article history: Microeukaryotes are the key ecosystem drivers mediating marine productivity, the food web and biogeochemical
Received 18 July 2019 cycles. The northwestern Pacific Ocean (NWPO), as one of the world's largest oligotrophic regions, remains largely
Received in revised form 12 August 2019 unexplored regarding diversity and biogeography of microeukaryotes. Here, we investigated the community com-
Accepted 3 September 2019
position and geographical distribution of microeukaryotes collected from the euphotic zone of three different re-
Available online 04 September 2019
gions in the NWPO using high-throughput sequencing of the 18S rRNA gene and quantified the contributions of
Editor: Lotfi Aleya environmental factors on the distributions of microeukaryotes. The relative abundance of different group taxa, ex-
cept for Ciliophora, presented distinct patterns in each region, and Metazoa and Dinoflagellata dominated the com-
Keywords: munity, contributing approximately half of reads abundance. Spatial and environmental factors explained 66.01% of
The northwestern Pacific Ocean community variation in the NWPO. Temperature was the most important environmental factor significantly corre-
Microeukaryotes lated with community structure. Bacterial biomass was also significantly correlated with microeukaryotic distribu-
Biogeography tion, especially for Dinoflagellata and Diatomea. Network analysis showed strong correlations between
18S rRNA gene sequencing microeukaryotic groups and free-living bacteria and different bacterial taxa were correlated with specific
Environmental factors
microeukaryotic groups, indicating that their interactions enabled microeukaryotic groups to adapt to diverse envi-
ronments. This study provides a first glance at the diversity and geographical distribution of microeukaryotes in the
NWPO and sheds light on the biotic and abiotic factors in shaping the microeukaryotic community in the ocean.
© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).

⁎ Corresponding author: Key Laboratory of Marine Ecology & Environmental Sciences, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China.
E-mail address: [email protected] (D.-Z. Wang).

https://doi.org/10.1016/j.scitotenv.2019.134289
0048-9697/© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
2 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289

1. Introduction subarctic is characterized by high nutrient levels and low temperatures,

and the confluence of them shapes the Kuroshio-Oyashio Transition
Microeukaryotes (protists, fungi and small zooplankton) are vital to Area which is also called the Mixed water region (Yasuda, 2003).
marine ecosystem, regulating marine productivity, the food web and These currents contribute a gradient of physics, chemistry and biology
biogeochemical cycles of various biogenic elements (Worden et al., to the NWPO (Qiu, 2001) which provides an ideal area to study the in-
2015). Recently, more and more attention has been paid to protists for teraction between microorganisms and environmental factors. Previous
their high diversity and diverse trophic modes (Caron et al., 2008; studies show that the different regions are dominated by different
Caron et al., 2012; Caron et al., 2016). In the epipelagic ocean, photosyn- groups of phytoplankton (Selph et al., 2005; Kok et al., 2014; Kataoka
thetic protists, like diatoms, dinoflagellates and certain haptophytes et al., 2017). Recent studies focusing on either bacteria or
(coccolithophores), are the major contributors of primary production picoeukaryotes show that temperature shapes community structures
and construct the basement of the ecosystem, while high trophic pro- (Xia et al., 2017; Li et al., 2018; Wang et al., 2019). However, so far,
tists, such as ciliates and flagellates, consuming bacteria and small pro- we know little about the diversity and geographic distribution of
tists are prayed by zooplankton (Sherr et al., 2007). Moreover, diverse microeukaryotes and their interactions with biotic and abiotic factors
mixotrophic protists also play important roles in balancing the ecosys- in this area.
tem in the oligotrophic ocean (Unrein et al., 2014). This food-web trans- In this study, we conducted a comprehensive study of the diversity
fer can fix CO2 into particle organic carbon (POC) and sink down into the of microeukaryotes in three different regions in the NWPO using high-
deep ocean to maintain the balance of the ocean system and carbon throughput sequencing of the 18S rRNA gene. We assessed the contri-
cycle (Worden et al., 2015). However, each group of microeukaryotes bution of spatial and environmental factors to shaping the geographic
makes their own contribution to the balance of the ecosystem and dom- patterns of microeukaryotes. Furthermore, we assessed the biotic
inates in distinct habitats (Grossmann et al., 2016). Recent studies dem- factors by constructing the co-occurrence network among
onstrate that microeukaryotic community structure is driven by biotic microeukaryotes and free-living bacteria. The purpose of this study
(e.g., prey availability, top-down grazing and bacterial effects) and abi- was to unveil eukaryotic microbial community structure in different
otic (e.g., spatial factors, light, temperature, salinity and nutrients) fac- habitats and to explore biotic and abiotic factors shaping community
tors (Sherr et al., 2007; Caron et al., 2016). Moreover, individual structure and distribution of microeukaryotes in the NWPO.
microeukaryotic taxa have distinct capabilities to adapt to different en-
vironmental factors, such as nutrients and temperature (Palenik, 2015;
Rao et al., 2018). Understanding microeukaryotic diversity and group- 2. Materials and methods
specific interactions with environmental factors in different regions is
critical to unveiling the ecological processes and mechanisms involved 2.1. Sampling and sample preparation
in maintaining the stability and function of the ecosystem (Naeem and
Li, 1997). The survey was carried out in the northwestern Pacific Ocean
High-throughput sequencing methods have enhanced our ability to (NWPO) from Mar. 30th to May. 6th 2015. Seawater samples were col-
assess the biodiversity of microeukaryotes and identify their ecological lected with Niskin bottles from the sea surface and the deep chlorophyll
significance in the ocean (Bik et al., 2012). Large-scale and multi- maximum (DCM) layer at nine sites except for site P5 which was sam-
sample datasets show that distribution patterns of protists are distinct pled just one layer due to the strong mixing in the upper 30 m. Each
between global ocean areas and different habitats (de Vargas et al., sample was pre-filtered by 200 μm Bolting Cloth and filtered through
2015; Grossmann et al., 2016). Growing evidence supports a long- the 1.6 μm GF/A membrane (142 mm diameter, Waterman) and then
tailed species abundance curve with a large number of rare species pre- a 0.2 μm PC membrane (142 mm diameter, Millipore) by peristaltic
senting in most ecosystems (Sogin et al., 2006; Fuhrman, 2009; Pedros- pump (Flojet). 50 L to 500 L of seawater were filtered from each layer
Alio, 2012; Lynch and Neufeld, 2015), however, microbial geographic depending on microeukaryotes biomass: in the oligotrophic Kuroshio
patterns are shaped by different processes, like dispersal limitation region, 500 L seawater was filtered due to the extremely low biomass,
and environmental stresses (Foissner et al., 2006; Hanson et al., 2012; while in the coastal region, 50 L seawater was filtered due to the high
Sul et al., 2013). Global-scale investigation proves that microbial com- biomass. Finally, 17 microeukaryotic samples (1.6–200 μm) were col-
munity compositions are driven by spatial and environmental factors, lected and each sample contained two replicates. Meanwhile, 17 free-
especially temperature (Sunagawa et al., 2015; Villar et al., 2015). Spa- living bacterial samples (0.2–1.6 μm) were collected for network analy-
tial factors together with environmental factors also play important sis of microeukaryotic groups and bacteria. All samples were immedi-
roles in structuring the microeukaryotic community (Zhang et al., ately frozen with liquid nitrogen and stored at −80 °C until processed.
2018a; Zhang et al., 2018b). However, biotic factors are less frequently
mentioned in previous microeukaryotic studies. Bacterial activity, such
as material exchange, chemical communication and algicidal activity, 2.2. Environmental parameters measurement
should also be considered as important factors shaping microeukaryotic
community structure in complex oceanic conditions (Caron et al., Temperature, salinity and oxygen concentration were obtained from
2016). Direct associations between bacteria and protists have been CTD (conductivity-temperature-depth) profiler (SeaBird Electronics,
proved by co-culture studies (Amin et al., 2015; Cruz-Lopez and Inc., Bellevue, WA, United States). Seawater samples were collected
Maske, 2016). Co-occurrence patterns in different marine regions for nutrient analysis by filtering through GF/F membrane (47 mm diam-
can provide us a profound understanding of microbe-microbe interac- eter, Waterman). The nutrient concentrations were analyzed photo-
tions in natural environments (Fuhrman, 2009), and specific bacterial metrically using an autoanalyzer (Model: SkalarSANplus). The
groups are tightly associated with phytoplankton bloom (Needham analytical precision of NO− − 3− −
2 , NO3 , PO4 , and SiO3 were 0.1, 0.1, 0.05,
and Fuhrman, 2016). However, the interactions between and 0.2 μM. Bacterial abundance samples were collected by pre-
microeukaryotes and bacteria in different natural habitats are still not filtering through a 1.6 μm GF/A membrane. Triplicate 2 mL seawater
well understood. samples with 2% glutaraldehyde in 2 mL tubes were stored at −20 °C
In the northwestern Pacific Ocean (NWPO), the environment is until processed. Bacterial cells were stained with SYBR Green I in anhy-
shaped by the Kuroshio and Oyashio currents which are driven by drous dimethyl sulfoxide and incubated in the dark for 15 min. Bacterial
wind and possess different physical-chemical features (Qiu, 2001). abundance was measured by a BD FACSAria Flow Cytometer (Becton
The Kuroshio Current carries oligotrophic water at high temperature Dickinson, USA) following the protocol described previously (Marie
from the equator, while the Oyashio Current originating from the et al., 2001).
 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289 3

2.3. DNA extraction and sequencing 2011), and rarefaction curves were drawn by R (version 3.5.2). For
beta-diversity analysis, non-metric multidimensional scaling (NMDS)
DNA of microeukaryotic samples (N1.6 μm) was extracted by im- analysis was operated in R based on Bray-Curtis similarity by ‘vegan’
proved extracting protocols (Yuan et al., 2015), owing to the complex package. The Bray-Curtis similarity was derived via the vegdist function
cell covering of dinoflagellates and diatoms. Prokaryotic samples exe- in ‘Vegan’ package. An analysis of similarity (ANOSIM) was used to sta-
cuted DNA were extracted by FastDNA SPIN extraction kit for soil (MP tistically test for significant differences in microeukaryotic communities
Biomedicals, Santa Ana, CA), according to the manufacturer's in the three regions and two layers. Furthermore, the nonparametric
instructions. Mann-Whitney U test and Kruskal-Wallis H test were calculated by
The V4–V5 hypervariable region of eukaryotic 18S rDNA was am- ‘stats’ package to test significant differences of microeukaryotic taxa
plified with Ek-NSF573 and Ek-NSR951 (378 bp) (Mangot et al., among regions and layers.
2013). This primer couple was selected by an in silico approach Mantel tests were run in R to determine correlations between en-
with a suitable overlap to conform to the sequencing accuracy of vironmental factors and the microeukaryotic community (based on
Miseq platform (Bradley et al., 2016). All PCR reactions were carried Bray-Cutis similarity). For environmental parameters (z-score-
out in 30 μL reactions with 15 μL of Phusion® High-Fidelity PCR Mas- transformed), Euclidean distance matrices were calculated via
ter Mix (New England Biolabs); 0.2 μM of forward and reverse the R base dist function. We quantified the relative effects of envi-
primers, and about 10 ng template DNA. Thermal cycling consisted ronmental and spatial factors in shaping the microeukaryotic com-
of initial denaturation at 98 °C for 1 min, followed by 30 cycles of de- munity with variation partitioning analysis (VPA) based on
naturation at 98 °C for 10s, annealing at 50 °C for 30s, and elongation redundancy analysis (RDA). A set of spatial variables was generated
at 72 °C for 60s, and finally, 72 °C for 5 min. A negative PCR control using principal coordinates of neighbor matrices (PCNM) analysis
with no template DNA was included for the reactions. All amplicons (Borcard and Legendre, 2002) based on the longitude and latitude
were then sequenced on a single run using the Illumina MiSeq of the sampling sites. Then, RDA was used to partition the variation
2x300bp platform. of the community composition between the extracted PCNM spatial
The V3–V4 hypervariable region of prokaryotic 16S rDNA was am- variables and environmental variables by ‘vegan’ package. VPA was
plified with 341F (5′CCTACGGGRBGCASCAG-3′) and 806R (5′GGAC performed using the “varpart” function of the vegan package which
TACNNGGGTATCTAAT-3′) using the same protocol as eukaryotes. The allows the total variation to be decomposed into fractions that indi-
primers can amplify both bacterial and archaea but with a bias for bac- cate the importance of pure environmental variables, pure spatial
teria (Yu et al., 2005). All amplicons were sequenced on a single run variables, shared fraction and unexplained variation.
using the Illumina HiSeq 2 × 250 bp platform. In order to demonstrate the relationship between eukaryotic groups
and the environmental parameters, sparse partial least square (sPLS)
2.4. Sequence assembly, clustering and annotation was used as implemented in the R package ‘mixOmics’ (Lê Cao et al.,
2008). The sPLS was applied in regression mode, which could model a
Raw data were first separated to each sample by barcodes and then causal relationship between the lineages and the environmental traits,
the barcodes and primer sequences were removed. Separated raw data that is, PLS could predict environmental traits from lineage abundances
were merged into raw tags using FLASH (V1.2.7, http://ccb.jhu.edu/ by calculating correlation between two matrices constructed with eu-
software/FLASH/) (Magoc and Salzberg, 2011). Raw tags were filtered karyotic relative abundance and environmental parameters. This ap-
by quality filters processed using QIIME (V1.7.0, http://qiime.org/ proach enabled us to identify high correlations between certain
index.html) (Caporaso et al., 2010). Filtered tags were grouped into op- lineages and environmental traits but without taking into account the
erational taxonomic units (OTUs) at 0.97 similarities with removal of global structure of the planktonic community.
chimera using USEARCH (version 7.1 http://drive5.com/uparse/) Network analysis was conducted based on Spearman's correla-
(Edgar, 2013). OTUs from 16s and 18s rDNA data were annotated tions calculated using the rcorr function in the ‘Hmisc’ package. Ro-
using the RDP classifier (Cole et al., 2009) confronted against the Silva bust correlations were considered if Spearman's correlation
(release 128) (Quast et al., 2013) respectively by a confidence threshold coefficient (ρ) was N0.6 for positive or b −0.6 for negative and statis-
of 0.7. tically significant (p b 0.05) (Barberán et al., 2011). Positive correla-
For analyses on higher levels of taxonomic group, the following taxa tion could be cooperation, symbiosis or parasitism, and negative
were selected: Alveolata. others, Ciliophora, Chloroplastida, correlation means predation or competition. Networks were visual-
Chrysophyceae, Cryptophyceae, Diatomea, Dictyochophyceae, ized using Cytoscape (version 3.7.1).
Dinoflagellata, Haptophyta, Rhizaria, Stramenopiles. others, Syndiniales,
Fungi, Metazoa. The group ‘Alveolata. others’ excluded Ciliophora,
Dinoflagellata and Syndiniales. The group ‘Stramenopiles. others’ ex- 3. Results
cluded Chrysophyceae, Diatomea and Dictyochophyceae. The OTUs am-
biguously annotated or unclassified by one of the given taxonomic 3.1. Overview of the survey area
groups were shown as ‘others’. Except for the multicellular groups of
Metazoa, Fungi and others, the remaining groups were clustered into The survey was conducted in the NWPO from Mar. 30th to May. 6th
protists. 2015. The environmental parameters of the investigation area are
shown in Table 1. Three regions: the coastal region (CR, including C5
2.5. Statistical analysis and P5 sites), the Kuroshio region (KR, including K1, K2, K3 and B9
sites) and the mixed water region (MWR, including A4, A6 and A8
To ensure inter-sample comparability for our taxonomic diversity sites) were classified based on current flows (Fig. 1) and environmental
estimates and following statistical analyses, we utilized the QIIME soft- factors (Table 1). Concentrations of nitrogen and phosphorus in the sur-
ware to subsample by randomly reducing the number of reads in each face layer were much lower than that in the DCM layer, and the highest
sample to the lowest number (19,793 sequence reads) of reads in any was observed in the DCM layer of the coastal C5 site. Concentration of
individual sample. Finally, the dataset retained 653,169 sequence chlorophyll a and bacterial abundance were higher in the DCM layer, ex-
reads for the entire community. cept for bacterial abundance at A6 site. Temperature and salinity in the
For alpha-diversity analysis, community diversity parameters CR were obviously lower than that in the other two regions, but bacte-
(Shannon index) were calculated using the mothur software (http:// rial abundance and concentrations of chlorophyll a and oxygen were
www.mothur.org/wiki/Schloss_SOP#Alpha_diversity) (Schloss et al., higher.
4 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289

Fig. 1. Sampling sites and schematic illustration of the Kuroshio current and the Oyashio current.

3.2. Diversity and distribution of microeukaryotes layers (Fig. S1, S2), indicating that the community difference was
greater between regions than between water layers.
2,188,792 high-quality sequences were obtained from 33 samples The relative abundances of microeukaryotic groups in the three re-
and were clustered into 3036 operational taxonomic units (OTUs). The gions exhibited different patterns (Fig. 3A). Protists accounted for 64%
rarefaction curves were roughly saturated for all samples (Fig. 2A). In of total sequences, comprising 80%, 71% and 47% in the CR, KR and
α diversity analysis, the Shannon index showed the diversity of each MWR, respectively. Metazoa and Dionflagellata were the most abun-
sample in the NWPO varied from 3.62 to 5.23 in the KR, from 2.94 to dant groups in the investigation areas, and they accounted for more
4.55 in the MWR and from 2.38 to 3.72 in the CR (Table S1). Based on than half of the proportion of the community (35% and 23% of total se-
Mann-Whitney U test, the Shannon index of samples in the KR was sig- quence reads), however, their relative abundance differed among the
nificantly higher than that in the MWR (P b 0.001) and CR (P b 0.01) three regions. Dinoflagellates were most abundant (44%) in the CR,
(Fig. 2B). In β diversity analysis, nonmetric multidimensional scaling while Metazoa had the highest proportion (52%) in the MWR.
analyses (NMDS) indicated that the community shifted across the Dinoflagellata and Metazoa dominated in the KR and presented compa-
three regions by sorting of sites (Fig. 2C). Analysis of similarity rable proportions (27% and 26% respectively). Protistan composition of
(ANOSIM) (Table 2) also showed that the communities of the three re- each site was relatively stable in the CR and KR but fluctuated in the
gions were significantly (P b 0.01) separated (R = 0.596). However, MWR (Fig.3B). Dinoflagellata was the most abundant protist group in
community dissimilarities between the two layers were 0.19 in the the CR and KR, but Haptophyta was more abundant in the KR and
NWPO. Of the 15 most abundant taxa at class level, more taxa groups MWR than that in the CR. Diatomea contributed a little to the protist
presented significant difference among different regions than two composition in the KR but its contribution was higher in the other two

Table 1
Sampling sites locations and environmental parameters.

Region Station Latitude Longitude Depth (m) Temperature Salinity Oxygen Chlorophyll BA NO2 PO4 SiO4 NO3 (μM)
(oN) (°E) (°C) (PSU) (mg/l) (μg/L) (/L) (nM) (μM) (μM)

CR C5-S 34 124 5 13.43 32.91 8.63 0.19 1,289,806 6 0.04 0.43 0.07
C5-D 34 124 30 9.31 32.90 9.33 2.69 1,813,807 301 0.24 4.50 5.11
P5-S 28.29 124.78 5 17.51 33.73 9.23 5.80 1,995,198 25 0.07 3.36 0.03
KR K1-S 25 130 5 23.24 34.89 4.67 0.01 256,543 UD UD UD UD
K1-D 25 130 60 21.97 34.87 6.29 0.93 768,711 24 0.01 1.18 0.07
K2-S 25 134 5 24.58 35.19 4.79 0.02 302,876 15 0.11 9.92 0.22
K2-D 25 134 85 20.19 34.90 6.56 0.29 608,536 59 0.13 14.10 0.55
K3-S 26.8 136.73 5 23.25 34.77 5.12 0.01 576,059 12 0.01 0.32 0.08
K3-D 26.8 136.73 80 19.97 34.87 6.67 0.52 804,833 34 0.15 1.31 1.70
B9-S 30 147 5 20.26 34.82 4.77 0.06 176,553 14 0.14 22.38 0.13
B9-D 30 147 60 18.00 34.83 7.31 0.66 552,371 17 0.13 20.90 0.12
MWR A4-S 34 147.83 5 17.66 34.67 6.77 0.72 197,351 23 0.07 4.25 0.03
A4-D 34 147.83 35 17.51 34.76 7.43 3.96 387,288 67 0.09 2.88 0.26
A6-S 34 150 5 17.73 34.86 7.01 1.24 345,482 14 0.07 2.95 0.17
A6-D 34 150 45 17.43 34.77 7.33 2.30 278,962 119 0.12 2.52 0.73
A8-S 34 152 5 18.00 34.92 6.04 0.11 310,029 39 0.07 2.89 0.13
A8-D 34 152 65 17.14 34.77 7.23 1.05 1,013,229 295 0.10 1.91 1.44

S, surface; D, DCM layer; BA, bacterial abundance; PSU, practical salinity unit; Chla, chlorophyll a. UD means the value is under the detection-limit.
 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289 5

Sobs index of OUT level

BA
Latitude

NMDS2
Number of reads sampled

T
Shannon index

Longitude

CR KR MWR
NMDS1

Fig. 2. Alpha and beta diversity analysis of eukaryotic composition among three regions. (A) rarefaction curves of similarity-based operational taxonomic unit (OTUs) at 97% sequence
identity threshold, (B) difference significance of average Shannon index among three regions, (C) Non-metric multidimensional scaling ordinations (NMDS) for microeukaryote
community and significantly correlated environmental parameters for sorting of sites (arrows). T: temperature; S: salinity; BA: bacterial abundance.

regions. Ciliophora contributed a similar proportion (4%) in the three Mantel test indicated that microeukaryotic community structure in
regions. the NWPO was significantly correlated with almost all environmental
factors (p b 0.01) except for NO− −
2 and SiO3 (Table 3). Microeukaryotic
community was significantly correlated with temperature and salinity
3.3. Correlations of microeukaryotic communities with spatial and environ- in the CR, while it was significantly correlated with all environmental
mental factors factors except for nutrients in the KR. Microeukaryotic community in
the MWR was significantly correlated with all the environmental factors
Spearman correlation comparing Bray-Cutis community similarity except for salinity and presented high correlation with nutrients.
with geographic distance between samples presented significant nega- sPLS method was used to calculate the correlation between each
tive correlations for the microeukaryotic community in the NWPO and microeukaryotic group distribution and each environmental factor in
KR with the correlation coefficients of −0.4997 (P b 0.01) and the NWPO. Heatmap results showed that the highest positive correla-
−0.3444 (P b 0.01), respectively (Fig. 4). However, in the CR and tion was observed between Diatomea and NO− 3 concentration (R =
MWR, the correlation was not significant. Variation partitioning analysis 0.73) (Fig.6). Metazoa and Rhizaria were positively correlated with lat-
showed that spatial and environmental factors together explained itude and longitude, while Dinoflagellata, Dictyochophyceae and
66.01% of the entire microeukaryotic community variation and 12.63% Syndiniales presented negative correlations, suggesting that distribu-
of variation was jointly explained by both factors (Fig. 5A). The spatial tions of different microeukaryotic groups had different geographic pat-
factors (transformed into five spatial variables principal coordinates of terns. Temperature and salinity presented similar correlation patterns
neighbor matrices) explained 18.62% of pure variation (P b 0.01), with different microeukaryotic groups. Chrysophyceae, Syndiniales
while environmental factors explained 34.76% (P b 0.01). Within envi- and Dictyochophyceae preferred high temperature and salinity while
ronmental factors, variation explained by abiotic factors (33.17% of Diatomea favored low temperature and salinity. Others presented
pure variation) was higher than that explained by biotic factors (2.40% high correlations with nutrients, such as nitrate, nitrite and phosphate.
of pure variation) (Fig. 5B). Diatomea and Dinoflagellata displayed positive correlations with con-
centrations of chlorophyll and oxygen, indicating that they are the
Table 2
main photosynthetic groups contributing to oxygen generation in the
Analysis of similarities (ANOSIM) of microeukaryotic communities among different layers ocean. Dinoflagellata (R = 0.65) and Diatomea (R = 0.45) presented
and regions. highly positive correlations with bacterial abundance, while Metazoa
Layers Regions
exhibited a negative correlation.

R P R P
3.4. Network analysis of microeukaryotic groups and free-living bacteria
Total 0.190⁎⁎ 0.002 Total 0.596⁎⁎ 0.001
CR 0.500 0.335 sur 0.686⁎⁎ 0.001
16 s rRNA gene sequencing results of free-living bacteria (0.2–1.6
KR 0.314⁎ 0.019 DCM 0.726⁎⁎ 0.001
MWR 0.654⁎⁎ 0.004 μm) collected from the surface and DCM layers at each site indicated dif-
ferent patterns of bacterial structure in the three regions (Fig. S3).
The ANOSIM results are calculated using the Bray-Curtis similarity data estimated from
999 permutations.
Spearman correlation analysis denoted strong correlations between
⁎ P b 0.05. microeukaryotic groups and free-living bacteria (Fig. 7). 36 orders of
⁎⁎ P b 0.01. bacteria were strongly correlated with 14 microeukaryotic groups, and
6 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289

Fig. 3. Community composition of eukaryotic groups and protist groups (A) relative abundance of eukaryotic groups in three regions, (B) relative abundance of protists. Protists groups are
all the groups except Fungi, Metazoa and others.

16 of them belonged to Proteobacteria. Almost all microeukaryotic Massana et al., 2015; de Vargas et al., 2015) and our results also revealed
groups were correlated with bacteria except for Cryptophyceae, and the different diversity and structure of microeukaryotes among the
each group was associated with specific bacterial assemblage. Bacterial three regions (Fig. 2, Table 2). Metazoa and Dionflagellate were the
taxa were the most associated with Syndiniales (14 orders). most abundant eukaryotic groups in our study, which was consistent
Thiotrichales and Xanthomonadales were positively correlated with with a previous study on global ocean scale (de Vargas et al., 2015). Al-
most microeukaryotic groups (five groups). Interestingly, some bacteria though we used a 200 μm sieve to remove large zooplankton, Metazoa
were specifically correlated with Diatomea, Chloroplastida, Rhizaria and still accounted for most sequence reads of all microeukaryotes and par-
Ciliophora. ticularly dominated in the MWR. A previous study shows that elongated
species of small-size and eggs, spores or larvae of large-size zooplankton
4. Discussion can pass through the 200 μm pores and contribute to the assemblage
(Liu et al., 2017). The Oyashio current is a productive current carrying
4.1. Microeukaryotic distribution in the NWPO a high biomass of phytoplankton from the subarctic and contributes to
a high biomass of zooplankton as latitude increases in the western Pa-
Studies have shown that microeukaryotes present different commu- cific Ocean (Taniguchi, 1972; Sakurai, 2007; Sun and Wang, 2017).
nity compositions in different habitats (Grossmann et al., 2016; Dinoflagellata was the second most abundant eukaryotic group in the
 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289 7

Fig. 4. Spearman's correlations between the Bray–Curtis similarity of microeukaryotic community and geographical distance between sampling sites in (A) NWPO, (B) CR, (C) KR and
(D) MWR. n is the number of comparison and P values are indicated.

NWPO and dominated in the CR and KR. It is also reported that dinofla- factor that strengthens the distance-decay relationship and eukaryotes
gellates dominate the phytoplankton community in these two regions are more likely to be affected by environmental factors than dispersal
(Kok et al., 2014; Liu et al., 2016). Previous studies show that diatoms limitation (Hanson et al., 2012; Wu et al., 2017). In our study, environ-
prefer to live in coastal areas with high nutrient levels and low temper- mental factors contributed more than spatial factors (34.76% compared
ature (Smetacek, 2012), while haptophytes dominate in open ocean and with 18% of pure variation) (Fig. 5), indicating that environmental selec-
oligotrophic ocean (Not et al., 2008; Unrein et al., 2014). Our study dem- tion plays a more important role in shaping the microeukaryotic com-
onstrated these preferences: Diatomea constituted a high proportion in munity than dispersal limitation in the NWPO.
the CR and MWR, while Haptophyta was abundant in the KR and MWR. Studies demonstrate that both environmental factors and geograph-
In addition, other groups also showed different distributions across ical distance play important roles in driving community structure on a
three regions. Overall, these results indicated that microeukaryotes pre- small scale (Horner-Devine et al., 2004; Martiny et al., 2006). However,
sented habitat-specific distributions in the NWPO. our study showed that the factors shaping the microeukaryotic commu-
nity in the KR were significantly different from that in the MWR
4.2. Factors regulating geographical distribution of microeukaryotes (Table 3, Fig. 4). Community dissimilarity in the KR was significantly
correlated with geographical distance, however, the distance-decay re-
Dispersal limitation and environmental heterogeneity are the main lationship in the KR was weaker than in the MWR (Fig. 4), suggesting
factors shaping the distributions of microorganisms, and they generate that the distance impact in the KR was decreased by other factors.
a negative correlation between community similarity and geographic Meanwhile, all environmental factors showed a weak correlation with
distance (distance-decay relationship) (Green and Bohannan, 2006; the microeukaryotic community in the KR. Hanson et al. (2012) report
Hanson et al., 2012). Previous studies show that dispersal limitation that dispersal may counteract microbial compositional differentiation
caused by geographical distance is one of the important factors in shap- and weaken the distance–decay relationship. It was postulated that
ing the microeukaryotic community in the ocean (Zhang et al., 2018a; the Kuroshio Current imposes strong dispersal activity on microorgan-
Zhang et al., 2018b). We found that the microeukaryotic community isms and contributes to shaping the community in the KR. In contrast,
presented a distance-decay relationship in each region which was the driving factors in the MWR presented another scenario where al-
strengthened as distance increased (Fig. 4). Limitation of microbial dis- most all environmental factors had significant correlations with the
persal strengthened by increasing geographic distances has been re- community, except for salinity (Table 3). These results indicate the im-
ported (Nekola and White, 1999; Martiny et al., 2006). However, portance of local environmental conditions in shaping the
some studies show that environmental selection is also an important microeukaryotic community structure on a small scale.
8 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289

composition and diversity and is also a stronger driver than other envi-
ronmental factors in shaping microbial community composition by
global investigation (Sunagawa et al., 2015). It is reported that temper-
ature is the main factor driving picoeukaryotes and bacteria in the
NWPO (Li et al., 2018; Wang et al., 2019). Studies indicate that more
species diversity is present in low latitude ocean because high temper-
ature leads to increasing productivity, metabolic rate and even specia-
tion (Evans and Gaston, 2005; Fuhrman et al., 2008). In our study, we
found that α diversities of the community were significantly higher in
the warm KR than in the other two regions (Fig. 2B, table S1). For differ-
ent taxa groups, temperature also showed high correlation with more
groups than other factors. Among these correlations, temperature pre-
sented the highest negative correlation with Diatomea. A previous
study shows that diatoms contribute more to phytoplankton commu-
nity in the polar area than in the tropical and subtropical ocean
(Malviya et al., 2016). A long-term study also demonstrates that dia-
toms prefer low temperature (Xiao et al., 2018), indicating that diatoms
have a special strategy to adapt to low temperature. However,
Chloroplastida, Syndiniales and Dictyochophyceae presented an oppo-
site preference for temperature, while the remaining groups showed
weak correlations with temperature. These results indicated that the
adaptive capability of different groups to temperature is an important
factor to determine the community distribution.
Inorganic nutrients are essential for growth and development of
microorganisms and are thought to be important factors in shaping
the phytoplankton community, and different microorganisms adapt
to their optimal growth concentrations (Gregg and Casey, 2007;
Follows and Dutkiewicz, 2011). Our results showed that only NO− 3
and PO− 4 were significantly correlated with the microeukaryotic
community in the NWPO (Table 3). Distributions of phytoplankton
groups exhibit distinct patterns under different nutrient limitations
on a global scale for their specific strategies to utilize nutrients
(Palenik, 2015). In our study, Diatomea was highly correlated with
Fig. 5. Variation partitioning, showing the effects of spatial, environmental on the all nutrients, especially with NO− 3 . A previous study indicated that di-
community composition of microeukaryotes in the NWPO. The percentage of variation atoms dominate in the early spring bloom, and they generally adapt
were explained by each fraction, including pure, shared explained and unexplained
to turbulent environments characterized by high levels of nutrients
variability. ANOVA permutation tests were calculated on the pure variation. ** P b 0.01. *
P b 0.05.
(Liess et al., 2009). Therefore, nutrients played an important role in
shaping the communities of diatoms in the NWPO. A large-scale
meta-analysis indicates that salinity is one of major determinants
across different habitats (Lozupone and Knight, 2007). Our result
showed that salinity presented a significant correlation with total
4.3. Environmental factors shaping the microeukaryotic community community. Dinoflagellata and Diatomea presented a negative cor-
relation with salinity, while Haptophyta presented a positive corre-
We used Mantel test to determine the impact of each environmental lation. The former two groups are predominant in coastal ocean
factor on shaping the microeukaryotic community. Among the factors, (Liu et al., 2016) which is easily affected by river input which con-
temperature had the highest impact on community (Table 3). Temper- tributes to a lower salinity. Meanwhile, haptophytes prefer to live
ature is known as an important factor that may alter community in the open ocean, indicating that the salinity tolerance of
microeukaryotes also impacts their distribution.
All microeukaryotic groups were more or less regulated by environ-
mental factors, except for Ciliophora, Chloroplastida, Cryptophyceae
and Fungi. Ciliophora are a free-living group with cilium that distribute
Table 3
Mantel test for the correlation between community and environmental factors. in various habitats and are adaptive to environment (Foissner et al.,
2009), which explains its even distribution in the NWPO (Fig. 3).
Total CR KR MWR
Chloroplastida are an ancient eukaryotic group with smallest genome
Env_distance 0.363⁎⁎ 0.622 0.201⁎ 0.659⁎⁎ and perform simple function to adapt different environments (Leliaert
Temperature 0.416⁎⁎ 0.636⁎ 0.232⁎ 0.717⁎⁎ et al., 2012). Cryptophyceae are mixotrophic, motile and low-light
Oxygen 0.321⁎⁎ 0.109 0.299⁎⁎ 0.564⁎⁎
adapted with low metabolic rate, and their growth is less influenced
Salinity 0.312⁎⁎ 0.156⁎ 0.168⁎ 0.267
BA 0.310⁎⁎ 0.045 0.223⁎ 0.319⁎ by environmental changes (Morgan and Kalff, 1975; Klaveness, 1989;
Chlorophyll a 0.193⁎⁎ 0.119 0.324⁎⁎ 0.260⁎ Litchman and Klausmeier, 2008). Fungi are the main decomposers in
NO−3 0.321⁎⁎ 0.679 0.036 0.572⁎⁎ the ocean which are more likely to be correlated with other organisms
NO− 0.103 0.668 −0.030 0.505⁎⁎
2
than environmental factors (Hyde et al., 1998; Richards et al., 2012).
PO43−
0.192⁎⁎ 0.646 0.090 0.715⁎⁎
SiO4− −0.079 0.242 0.110 0.241⁎ Overall, the microeukaryotic community was more or less regulated
4
by environmental factors. However, the response of each group to envi-
Env_distance, Euclidean distance of all environmental variables between sampling sites;
BA, bacterial abundance. The significances are tested based on 999 permutations.
ronmental factors was different due to their specific adaptation strate-
⁎ P b 0.05. gies. The molecular mechanisms involved in adaptation to varying
⁎⁎ P b 0.01. environmental conditions needs further study.
 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289 9

Fig. 6. Clustered heat map depicting correlations between eukaryotic groups and environmental parameters. Strong correlation is highlighted (R N 0.4 or b −0.4). BA: bacterial abundance.

4.4. Interaction between microeukaryotic groups and free-living bacteria contributing the highest proportion of protists with a percentage rang-
ing from 22.4% to 55.9% across the three regions. It is reported that more
Although spatial and environmental factors explained 66.01% of var- than half (58%) of dinoflagellates are heterotrophic species in the ocean
iation in total, 33.99% of variation was still unexplained and might be and pray on other small organisms like bacteria and even small protists
caused by unmeasured environmental and ecological factors. Studies in- (Gómez, 2012), and they also need vitamin B1 and B12 produced by
dicate that the interactions, for example, mutualism, cross-feeding, bacteria to support high community abundance (Tang et al., 2010). In
competition, parasitism, predation and allelopathy among microbes our study, dinoflagellates exhibited the highest correlation with bacte-
are also responsible for community structure (Lima-Mendez et al., rial abundance, indicating a strong reliance on biomass of bacteria. We
2015; Caron et al., 2016). Recent studies demonstrate that phytoplank- also found that Dinoflagellata were positively correlated with several
ton exhibits a stronger relationship with bacteria than with environ- different bacterial groups (Fig. 7), suggesting that dinoflagellates pre-
mental factors (Pearman et al., 2016; Needham and Fuhrman, 2016). ferred to pray on bacteria and/or utilize their metabolites with a wide
In our study, bacterial abundance was significantly correlated with the selection.
microeukaryotic distribution in the NWPO (Table 3) and highly corre- Diatoms contribute 20% of global primary productivity and produce
lated with specific groups (Fig. 6). Previous studies reveal that the large quantities of dissolved organic matter (DOM) (Armbrust, 2009;
growth of diatoms and dinoflagellates relies on bacteria through impor- Myklestad, 2000). Interactions between diatoms and bacteria are well
tant materials exchange (Aziz et al., 2010; Tang et al., 2010; Amin et al., studied, and heterotrophic bacteria can utilize the DOM secreted by di-
2015; Cruz-Lopez and Maske, 2016). We also observed that these two atoms, while diatoms also need the heteroauxin or vitamins from bacte-
groups were positively correlated with bacterial abundance, indicating ria to support growth (Amin et al., 2012). Our results showed that the
that bacteria played important roles in shaping distribution of specific distributions of diatoms presented high correlations with bacterial
microeukaryotic groups in the natural environment across different abundance, suggesting the interaction between diatoms and bacteria.
habitats. Previous studies show that phytoplankton exhibit high selec- Furthermore, we observed specifically negative correlations of
tion on bacterial species (Hendrik et al., 2002; Sapp et al., 2007), Diatomea with Sphingomonadales, Order_II and Caldilineales (Fig. 7).
which was also presented in our study, that each group of A previous study shows that bacteria can present algicidal activity spe-
microeukaryotes interacted with a specific group-assembly of bacteria. cific for diatoms (Paul and Pohnert, 2011). These results suggested that
Dinoflagellates are an important group of protists with a wide distri- diatoms required high abundance of bacteria to support growth, how-
bution and strong adaptive ability (Gómez, 2012). The relative abun- ever, and were easily affected by specific bacterial groups.
dance of dinoflagellates reaches 40% of protists on average, ranging Co-occurring bacteria stimulating or inhibiting colony formation of
from 18% to 67% on the global scale (Le Bescot et al., 2016). Our results the choanoflagellate Salpingoeca rosetta has been reported (Woznica
showed that Dinoflagellata was the most abundant protist in the NWPO, et al., 2016), which may influence the development of multicellularity
10 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289

Fig. 7. Network depicting correlations between eukaryotic groups and bacterial groups. Red edges denote a strong positive correlation (r N 0.6) while blue edges characterize a strong
negative correlation (r b −0.6).

in the super-groups such as Metazoa and Fungi. We also found that sev- Data availability
eral orders of bacteria were significantly correlated with these two
groups. Overall, each microeukaryotic group showed a high correlation All raw sequence data from this study were deposited in the
with specific groups of free-living bacteria, suggesting a special associa- GenBank's Sequence Read Archive (SRA) database (http://www.ncbi.
tion among different groups of microbes. However, field-based studies nlm.nih.gov/) under the accession number SRP (BioProject Accession
complemented with laboratory simulations are required to understand PRJNA476806 and BioSample accession 18 s for SAMN06246921–
the various influences of associated bacteria on microeukaryotic physi- SAMN06246956, 16s for SAMN10963348- SAMN10963364).
ology, such as those related to toxin production, organic matter produc-
tion and recycling, and algal bloom formation. Declaration of Competing Interest

5. Conclusion The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
Our results indicated that both diversity and composition of the ence the work reported in this paper.
microeukaryotic community differed among the three different regions
of the NWPO. A distance-decay relationship for microeukaryotes was Acknowledgements
observed in the NWPO and was strengthened by environmental selec-
tion. Geographical distance had a significant impact on shaping commu- We thank the captain and crew of the R/V Dongfanghong II. This work
nity structure on a large scale, while environmental factors played was partially supported by research grants from the National Natural
important roles on a small scale, especially for dynamic environments. Science Foundation of China (Project no. 41425021), and the Ministry
Temperature was the main abiotic factor regulating diversity and com- of Science and Technology of the People's Republic of China (Project
position of the microeukaryotic community in the NWPO, and in each no. 2015CB954003). Da-Zhi Wang was also supported by the Ten Thou-
region. Bacterial abundance was significantly correlated with distribu- sand Talents Program for leading talents in science and technological
tion of the microeukaryotic community, especially with Dinoflagellata innovation.
and Diatomea. However, they presented different interactions with
free-living bacteria revealed by co-occurrence network analysis. Dino- Appendix A. Supplementary data
flagellates preferred to pray on bacteria and utilize their metabolites
with a wide selection. Diatoms were sensitive to environmental factors Supplementary data to this article can be found online at https://doi.
and were vulnerable to negative effects caused by specific bacterial org/10.1016/j.scitotenv.2019.134289.
groups. Although interactions between the microeukaryotic community
and biotic and abiotic factors across different habitats were unveiled in
our study, metabolic activity influenced by different factors and utiliza- References
tion mechanisms of exchanging materials still need to be explored in fu- Amin, S.A., Parker, M.S., Armbrust, E.V., 2012. Interactions between diatoms and bacteria.
ture study. Microbiol. Mol. Biol. Rev. 76, 667–684.
 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289 11

Amin, S.A., Hmelo, L.R., van Tol, H.M., Durham, B.P., Carlson, L.T., Heal, K.R., et al., 2015. In- Leliaert, F., Smith, D.R., Moreau, H., Herron, M.D., Verbruggen, H., Delwiche, C.F., et al.,
teraction and signalling between a cosmopolitan phytoplankton and associated bac- 2012. Phylogeny and molecular evolution of the Green algae. Crit. Rev. Plant Sci. 31,
teria. Nature 522, 98–101. 1–46.
Armbrust, E.V., 2009. The life of diatoms in the world's oceans. Nature 459, 185–192. Li, Y.Y., Chen, X.H., Xie, Z.X., Li, D.X., Wu, P.F., Kong, L.F., et al., 2018. Bacterial diversity and
Aziz, R.K., Moustafa, A., Evans, A.N., Kulis, D.M., Hackett, J.D., Erdner, D.L., et al., 2010. nitrogen utilization strategies in the upper layer of the northwestern Pacific Ocean.
Transcriptome profiling of a toxic dinoflagellate reveals a gene-rich Protist and a Front. Microbiol. 9, 797.
potential impact on gene expression due to bacterial presence. PLoS One 5, Liess, A., Lange, K., Schulz, F., Piggott, J.J., Matthaei, C.D., Townsend, C.R., 2009. Light, nutri-
e9688. ents and grazing interact to determine diatom species richness via changes to pro-
Barberán, A., Bates, S.T., Casamayor, E.O., Fierer, N., 2011. Using network analysis to ex- ductivity, nutrient state and grazer activity. J. Ecol. 97, 326–336.
plore co-occurrence patterns in soil microbial communities. The ISME Journal 6, 343. Lima-Mendez, G., Faust, K., Henry, N., Decelle, J., Colin, S., Carcillo, F., et al., 2015. Determi-
Bik, H.M., Porazinska, D.L., Creer, S., Caporaso, J.G., Knight, R., Thomas, W.K., 2012. Se- nants of community structure in the global plankton interactome. Science 348.
quencing our way towards understanding global eukaryotic biodiversity. Trends Litchman, E., Klausmeier, C.A., 2008. Trait-based community ecology of phytoplankton.
Ecol. Evol. 27, 233–243. Annu. Rev. Ecol. Evol. Syst. 39 (1), 615–639.
Borcard, D., Legendre, P., 2002. All-scale spatial analysis of ecological data by means of Liu, X., Xiao, W., Landry, M.W.R., Chiang, K.-P., Wang, L., Huang, B., 2016. Responses of
principal coordinates of neighbour matrices. Ecol. Model. 153, 51–68. phytoplankton communities to environmental variability in the East China Sea. Eco-
Bradley, I.M., Pinto, A.J., Guest, J.S., 2016. Design and evaluation of Illumina MiSeq- systems 19, 832–849.
compatible, 18S rRNA gene-specific primers for improved characterization of mixed Liu, L., Liu, M., Wilkinson, D.M., Chen, H., Yu, X., Yang, J., 2017. DNA metabarcoding reveals
phototrophic communities. Appl. Environ. Microbiol. 82, 5878–5891. that 200-mum-size-fractionated filtering is unable to discriminate between plank-
Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., et al., tonic microbial and large eukaryotes. Mol. Ecol. Resour. 17, 991–1002.
2010. QIIME allows analysis of high-throughput community sequencing data. Nat Lozupone, C.A., Knight, R., 2007. Global patterns in bacterial diversity. Proc. Natl. Acad. Sci.
Meth 7, 335–336. 104, 11436–11440.
Caron, D.A., Worden, A.Z., Countway, P.D., Demir, E., Heidelberg, K.B., 2008. Protists are Lynch, M.D., Neufeld, J.D., 2015. Ecology and exploration of the rare biosphere. Nat. Rev.
microbes too: a perspective. The ISME Journal 3, 4–12. Microbiol. 13, 217–229.
Caron, D.A., Countway, P.D., Jones, A.C., Kim, D.Y., Schnetzer, A., 2012. Marine protistan di- Magoc, T., Salzberg, S.L., 2011. FLASH: fast length adjustment of short reads to improve
versity. Annu. Rev. Mar. Sci. 4, 467–493. genome assemblies. Bioinformatics 27, 2957–2963.
Caron, D.A., Alexander, H., Allen, A.E., Archibald, J.M., Armbrust, E.V., Bachy, C., et al., 2016. Malviya, S., Scalco, E., Audic, S., Vincent, F., Veluchamy, A., Poulain, J., et al., 2016. Insights
Probing the evolution, ecology and physiology of marine protists using transcripto- into global diatom distribution and diversity in the world's ocean. Proc. Natl. Acad.
mics. Nat. Rev. Microbiol. 15. Sci. 113, E1516–E1525.
Cole, J.R., Wang, Q., Cardenas, E., Fish, J., Chai, B., Farris, R.J., et al., 2009. The ribosomal da- Mangot, J.F., Domaizon, I., Taib, N., Marouni, N., Duffaud, E., Bronner, G., et al., 2013. Short-
tabase project: improved alignments and new tools for rRNA analysis. Nucleic Acids term dynamics of diversity patterns: evidence of continual reassembly within lacus-
Res. 37, D141–D145. trine small eukaryotes. Environ. Microbiol. 15, 1745–1758.
Cruz-Lopez, R., Maske, H., 2016. The vitamin B1 and B12 required by the marine dinofla- Marie, D., Partensky, F., Vaulot, D., Brussaard, C., 2001. Enumeration of Phytoplankton,
gellate Lingulodinium polyedrum can be provided by its associated bacterial commu- Bacteria, and Viruses in Marine Samples. Current Protocols in Cytometry. John
nity in culture. Front. Microbiol. 7, 560. Wiley & Sons, Inc.
de Vargas, C., Audic, S., Henry, N., Decelle, J., Mahé, F., Logares, R., et al., 2015. Eukaryotic Martiny, J.B., Bohannan, B.J., Brown, J.H., Colwell, R.K., Fuhrman, J.A., Green, J.L., et al., 2006.
plankton diversity in the sunlit ocean. Science 348. Microbial biogeography: putting microorganisms on the map. Nat. Rev. Microbiol. 4,
Edgar, R.C., 2013. UPARSE: highly accurate OTU sequences from microbial amplicon reads. 102–112.
Nat. Meth. 10, 996–998. Massana, R., Gobet, A., Audic, S., Bass, D., Bittner, L., Boutte, C., et al., 2015. Marine protist
Evans, K.L., Gaston, K.J., 2005. Can the evolutionary-rates hypothesis explain species- diversity in European coastal waters and sediments as revealed by high-throughput
energy relationships? Funct. Ecol. 19, 899–915. sequencing. Environ. Microbiol. 17, 4035–4049.
Foissner, W., Salzburg, U., Biologie, F.O., 2006. Biogeography and dispersal of micro- Morgan, K., Kalff, J., 1975. The winter dark survival of an algal flagellate — Cryptomonas
organisms: a review emphasizing protists. Acta Protozool. 45, 111–136. erosa (Skuja). SIL Proceedings, 1922–2010, pp. 2734–2740 19(4).
Foissner, W., Chao, A., Katz, L.A., 2009. Diversity and geographic distribution of cili- Myklestad, S.M., 2000. Dissolved organic carbon from phytoplankton. In: Wangersky,
ates (Protista: Ciliophora). In: Foissner, W., Hawksworth, D.L. (Eds.), Protist Di- P.J. (Ed.), Marine Chemistry. Springer Berlin Heidelberg, Berlin, Heidelberg,
versity and Geographical Distribution. Springer, Netherlands: Dordrecht, pp. 111–148.
pp. 111–129. Naeem, S., Li, S., 1997. Biodiversity enhances ecosystem reliability. Nature 390, 507–509.
Follows, M.J., Dutkiewicz, S., 2011. Modeling diverse communities of marine microbes. Needham, D.M., Fuhrman, J.A., 2016. Pronounced daily succession of phytoplankton, ar-
Annu. Rev. Mar. Sci. 3, 427–451. chaea and bacteria following a spring bloom. Nat. Microbiol. 1, 16005.
Fuhrman, J.A., 2009. Microbial community structure and its functional implications. Na- Nekola, J.C., White, P.S., 1999. The distance decay of similarity in biogeography and ecol-
ture 459, 193–199. ogy. J. Biogeogr. 26, 867–878.
Fuhrman, J.A., Steele, J.A., Hewson, I., Schwalbach, M.S., Brown, M.V., Green, J.L., et al., Not, F., Latasa, M., Scharek, R., Viprey, M., Karleskind, P., Balagué, V., et al., 2008. Protistan
2008. A latitudinal diversity gradient in planktonic marine bacteria. Proc. Natl. assemblages across the Indian Ocean, with a specific emphasis on the picoeukaryotes.
Acad. Sci. 105, 7774–7778. Deep-Sea Res. I Oceanogr. Res. Pap. 55, 1456–1473.
Gómez, F., 2012. A quantitative review of the lifestyle, habitat and trophic diversity of di- Palenik, B., 2015. Molecular mechanisms by which marine phytoplankton respond to
noflagellates (Dinoflagellata, Alveolata). Syst. Biodivers. 10, 267–275. their dynamic chemical environment. Annu. Rev. Mar. Sci. 7, 325–340.
Green, J., Bohannan, B.J.M., 2006. Spatial scaling of microbial biodiversity. Trends Ecol. Paul, C., Pohnert, G., 2011. Interactions of the Algicidal bacterium Kordia algicida
Evol. 21, 501–507. with diatoms: regulated protease excretion for specific algal lysis. PLoS One 6,
Gregg, W.W., Casey, N.W., 2007. Modeling coccolithophores in the global oceans. Deep- e21032.
Sea Res. II Top. Stud. Oceanogr. 54, 447–477. Pearman, J.K., Casas, L., Merle, T., Michell, C., Irigoien, X., 2016. Bacterial and protist com-
Grossmann, L., Jensen, M., Heider, D., Jost, S., Glücksman, E., Hartikainen, H., et al., 2016. munity changes during a phytoplankton bloom. Limnol. Oceanogr. 61, 198–213.
Protistan community analysis: key findings of a large-scale molecular sampling. Pedros-Alio, C., 2012. The rare bacterial biosphere. Annu. Rev. Mar. Sci. 4, 449–466.
The ISME Journal 10, 2269–2279. Qiu, B., 2001. Kuroshio and Oyashio currents. In: Steele, J.H. (Ed.), Encyclopedia of Ocean
Hanson, C.A., Fuhrman, J.A., Horner-Devine, M.C., Martiny, J.B., 2012. Beyond biogeo- Sciences. Academic Press, San. Diego, pp. 1413–1425.
graphic patterns: processes shaping the microbial landscape. Nat. Rev. Microbiol. Quast, C., Pruesse, E., Yilmaz, P., Gerken, J., Schweer, T., Yarza, P., et al., 2013. The SILVA ri-
10, 497–506. bosomal RNA gene database project: improved data processing and web-based tools.
Hendrik, S., Ben, A., Harry, W., Gerard, M., 2002. Genetic diversity of ‘satellite’ bacteria Nucleic Acids Res. 41, D590–D596.
present in cultures of marine diatoms. FEMS Microbiol. Ecol. 42, 25–35. Rao, K., Zhang, X., Yi, X.J., Li, Z.S., Wang, P., Huang, G.W., et al., 2018. Interactive effects of
Horner-Devine, M.C., Lage, M., Hughes, J.B., Bohannan, B.J., 2004. A taxa-area relationship environmental factors on phytoplankton communities and benthic nutrient interac-
for bacteria. Nature 432, 750–753. tions in a shallow lake and adjoining rivers in China. Sci. Total Environ. 619-620,
Hyde, K.D., Jones, E.B.G., Leaño, E., Pointing, S.B., Poonyth, A.D., Vrijmoed, L.L.P., 1998. Role 1661–1672.
of Fungi in Marine Ecosystems. vol. 7 pp. 1147–1161. Richards, T.A., Jones, M.D.M., Leonard, G., Bass, D., 2012. Marine Fungi: Their Ecology and
Kataoka, T., Yamaguchi, H., Sato, M., Watanabe, T., Taniuchi, Y., Kuwata, A., et al., 2017. Molecular Diversity. vol. 4 pp. 495–522.
Seasonal and geographical distribution of near-surface small photosynthetic eukary- Sakurai, Y., 2007. An overview of the Oyashio ecosystem. Deep-Sea Res. II Top. Stud.
otes in the western North Pacific determined by pyrosequencing of 18S rDNA. FEMS Oceanogr. 54, 2526–2542.
Microbiol. Ecol. 93. Sapp, M., Schwaderer, A.S., Wiltshire, K.H., Hoppe, H.G., Gerdts, G., Wichels, A., 2007.
Klaveness, D., 1989. Biology and ecology of the Cryptophyceae: status and challenges. Species-specific bacterial communities in the Phycosphere of microalgae? Microb.
Biol. Oceanogr. 6 (3–4), 257–270. Ecol. 53, 683–699.
Kok, S.P., Tsuchiya, K., Komatsu, K., Toda, T., Kurosawa, N., 2014. The protistan micro- Schloss, P.D., Gevers, D., Westcott, S.L., 2011. Reducing the effects of PCR amplification and
plankton community along the Kuroshio current revealed by 18S rRNA gene clone sequencing artifacts on 16S rRNA-based studies. PLoS One 6, e27310.
analysis: a case study of the differences in distribution interplay with ecological var- Selph, K.E., Shacat, J., Landry, M.W.R., 2005. Microbial community composition and
iability. Plankton and Benthos Research 9, 71–82. growth rates in the NW Pacific during spring 2002. Geochem. Geophys. Geosyst. 6
Le Bescot, N., Mahe, F., Audic, S., Dimier, C., Garet, M.J., Poulain, J., et al., 2016. Global pat- (n/a-n/a).
terns of pelagic dinoflagellate diversity across protist size classes unveiled by Sherr, B.F., Sherr, E.B., Caron, D.A., Vaulot, D., Worden, A.Z., 2007. Oceanic Protists. Ocean-
metabarcoding. Environ. Microbiol. 18, 609–626. ography 20, 130–134.
Lê Cao, K.A., Rossouw, D., Robert-Granie, C., Besse, P., 2008. A sparse PLS for variable selec- Smetacek, V., 2012. Making sense of ocean biota: How evolution and biodiversity of land
tion when integrating omics data. Stat. Appl. Genet. Mol. Biol. 7, 35. organisms differ from the plankton. J. Biosci. 37, 589–607.
12 P.-F. Wu et al. / Science of the Total Environment 698 (2020) 134289

Sogin, M.L., Morrison, H.G., Huber, J.A., Welch, D.M., Huse, S.M., Neal, P.R., et al., 2006. Mi- Wu, W., Lu, H.P., Sastri, A., Yeh, Y.C., Gong, G.C., Chou, W.C., et al., 2017. Contrasting the
crobial diversity in the deep sea and the underexplored “rare biosphere”. Proc. Natl. relative importance of species sorting and dispersal limitation in shaping marine bac-
Acad. Sci. 103, 12115–12120. terial versus protist communities. The ISEM Journal 12, 485.
Sul, W.J., Oliver, T.A., Ducklow, H.W., Amaral-Zettler, L.A., Sogin, M.L., 2013. Marine bacte- Xia, X., Partensky, F., Garczarek, L., Suzuki, K., Guo, C., Yan Cheung, S., et al., 2017.
ria exhibit a bipolar distribution. Proc. Natl. Acad. Sci. 110, 2342–2347. Phylogeography and pigment type diversity of Synechococcus cyanobacteria in sur-
Sun, D., Wang, C., 2017. Latitudinal distribution of zooplankton communities in the West- face waters of the northwestern Pacific Ocean. Environ. Microbiol. 19, 142–158.
ern Pacific along 160°E during summer 2014. J. Mar. Syst. 169, 52–60. Xiao, W., Liu, X., Irwin, A.J., Laws, E.A., Wang, L., Chen, B., et al., 2018. Warming and eutro-
Sunagawa, S., Coelho, L.P., Chaffron, S., Kultima, J.R., Labadie, K., Salazar, G., et al., 2015. phication combine to restructure diatoms and dinoflagellates. Water Res. 128,
Structure and function of the global ocean microbiome. Science 348. 206–216.
Tang, Y.Z., Koch, F., Gobler, C.J., 2010. Most harmful algal bloom species are vitamin B1 and Yasuda, I., 2003. Hydrographic structure and variability in the Kuroshio-Oyashio transi-
B12 auxotrophs. Proc. Natl. Acad. Sci. 107, 20756–20761. tion area. J. Oceanogr. 59, 389–402.
Taniguchi, A., 1972. Primary production in the Oyashio region with special reference to Yu, Y., Lee, C., Kim, J., Hwang, S., 2005. Group-specific primer and probe sets to detect me-
the subsurface chlorophyll maximum layer and phytoplanktonzooplankton relation- thanogenic communities using quantitative real-time polymerase chain reaction.
ships. Biological Oceanography of the Northern North Pacific Ocean 231–243. Biotechnol. Bioeng. 89, 670–679.
Unrein, F., Gasol, J.M., Not, F., Forn, I., Massana, R., 2014. Mixotrophic haptophytes are key Yuan, J., Li, M., Lin, S., 2015. An improved DNA extraction method for efficient and quan-
bacterial grazers in oligotrophic coastal waters. ISME J. 8, 164–176. titative recovery of phytoplankton diversity in natural assemblages. PLoS One 10,
Villar, E., Farrant, G.K., Follows, M., Garczarek, L., Speich, S., Audic, S., et al., 2015. Environ- e0133060.
mental characteristics of Agulhas rings affect interocean plankton transport. Science Zhang, H., Huang, X., Huang, L., Bao, F., Xiong, S., Wang, K., et al., 2018a. Microeukaryotic
348. biogeography in the typical subtropical coastal waters with multiple environmental
Wang, F., Xie, Y., Wu, W., Sun, P., Wang, L., Huang, B., 2019. Picoeukaryotic diversity and gradients. Sci. Total Environ. 635, 618–628.
activity in the northwestern Pacific Ocean based on rDNA and rRNA high-throughput Zhang, W., Pan, Y., Yang, J., Chen, H., Holohan, B., Vaudrey, J., et al., 2018b. The diversity
sequencing. Front. Microbiol. 9, 3259. and biogeography of abundant and rare intertidal marine microeukaryotes explained
Worden, A.Z., Follows, M.J., Giovannoni, S.J., Wilken, S., Zimmerman, A.E., Keeling, P.J., by environment and dispersal limitation. Environ. Microbiol. 20, 462–476.
2015. Rethinking the marine carbon cycle: factoring in the multifarious lifestyles of
microbes. Science 347, 1257594.
Woznica, A., Cantley, A.M., Beemelmanns, C., Freinkman, E., Clardy, J., King, N., 2016. Bac-
terial lipids activate, synergize, and inhibit a developmental switch in
choanoflagellates. Proc. Natl. Acad. Sci. 113, 7894–7899.