Harmful Algae 37 (2014) 142–152

Genetic structure of Pseudo-nitzschia pungens (Bacillariophyceae)populations: Implications of a global diversification of the diatom

Hong Chang Lim a, Po Teen Lim b, Sing Tung Teng a, Stephen S. Bates c, Chui Pin Leaw b,*a Faculty of Resource Science and Technology, Universiti Malaysia Sarawak, 94300 Kota Samarahan, Sarawak, Malaysiab Bachok Marine Research Station, Institute of Ocean and Earth Sciences, University of Malaya, 16310 Bachok, Kelantan, Malaysiac Fisheries and Oceans Canada, Gulf Fisheries Centre, PO Box 5030, Moncton, NB, Canada E1C 9B6

A R T I C L E I N F O

Article history:

Received 11 March 2014

Received in revised form 3 June 2014

Accepted 3 June 2014

Keywords:

Diversification

ITS secondary structure

Phylogeography

Population structure

Pseudo-nitzschia pungens

A B S T R A C T

Pseudo-nitzschia pungens is a planktonic marine diatom known to be widespread in tropical and

temperate coastal waters. We examined the population genetic structure of tropical Southeast Asian

populations of P. pungens and compared it with those of northern and southern temperate populations.

The secondary structures of the nuclear encoded internal transcribed spacer (ITS) region of 164 strains of

P. pungens were modeled and analyzed. The tree revealed three ITS entities: clade I (comprised of P.

pungens var. pungens) was distributed mainly in northern temperate waters; clade II (comprised of both

P. pungens var. pungens and var. cingulata) was mainly from the NE Pacific; and clade III (comprised of

both P. pungens var. pungens and var. aveirensis) was restricted to tropical and warm-temperate waters.

Hybrids of both P. pungens var. pungens and var. cingulata co-occurred in clades I and II. Sixty haplotypes

were revealed from the sequences of 164 strains. Haplotype diversity inferred from the median-joining

network was in accordance with phylogenetic analysis, further supporting the grouping of the P. pungens

haplogroups. Our results revealed limited gene flow between P. pungens from tropical and temperate

waters, and significant population structure, as estimated by an analysis of molecular variance (AMOVA),

with 75% of the total ITS variation found among populations (KST = 0.75). This study suggests that

distinct environmental clines, such as ocean thermohaline circulation, have a potential for fragmenting

and dispersing global populations of P. pungens. Formation of the Isthmus of Panama, in particular, is

speculated to play a role in this allopatric differentiation in P. pungens populations worldwide.

� 2014 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Harmful Algae

jo u rn al h om epag e: ww w.els evier .c o m/lo cat e/ha l

1. Introduction

Phytoplankton, as primary producers, have always been animportant food source. Marine phytoplankton also play a role inthe carbon cycle by absorbing atmospheric CO2 and removing it toocean depths (Falkowski et al., 2000). However, certain algalgroups are known to be harmful when they bloom in theenvironment, causing a nuisance to other organisms. Certainspecies of the genus Pseudo-nitzschia are responsible for amnesicshellfish poisoning (ASP). The toxin responsible for this illness isdomoic acid (DA). This neurotoxin affects not only humans, butalso marine mammals and sea birds (reviewed by Lelong et al.,2012; Trainer et al., 2012). Outbreaks of toxic Pseudo-nitzschia

blooms, with levels of DA in shellfish above the regulatory limit,have resulted in the closure of shellfish farming and harvesting in

* Corresponding author. Tel.: +6097785001; fax: +6097785006.

E-mail addresses: cpleaw@um.edu.my, chuipinleaw@gmail.com (C.P. Leaw).

http://dx.doi.org/10.1016/j.hal.2014.06.004

1568-9883/� 2014 Elsevier B.V. All rights reserved.

several countries (Kotaki et al., 1996; Amzil et al., 2001; Vale andSampayo, 2001; Trainer et al., 2007).

Pseudo-nitzschia pungens (Cleve) Hasle was first described in1993 (Hasle, 1993). In 1998, P. pungens var. cingulata Villac wasdescribed, characterized by apparent morphological differences invalve structure and cingular bands (Villac and Fryxell, 1998). Adecade later, P. pungens var. aveirensis Lundholm, Churro, Carreiraand Calado was described (Churro et al., 2009). The originallydescribed P. pungens was then designated as the nominal variety (P.

pungens var. pungens). Its taxonomic classification did not change.Pseudo-nitzschia pungens is one of the most studied species owingits wide distribution (Hasle, 2002; Casteleyn et al., 2010) and theability of some strains to produce trace levels of DA (Rhodes et al.,1996; Trainer et al., 1998; Calu et al., 2009; Moschandreou et al.,2012). However, most P. pungens strains have been reported to benon-toxic (Bates et al., 1998; Villac and Fryxell, 1998; Li et al.,2005; Churro et al., 2009; Lim et al., 2010, 2012a; Quijano-Scheggiaet al., 2010). Pseudo-nitzschia pungens is heterothallic (Chepurnovet al., 2005), and natural hybridization can take place in theenvironment (Casteleyn et al., 2009a; Holtermann et al., 2010).

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152 143

Identification of Pseudo-nitzschia species requires well-trainedtaxonomists and skills in electron microscopy. Molecular toolswith specific gene markers were later developed to aid speciesidentification and these have been implemented for monitoringpurposes (Miller and Scholin, 1996; Vrieling et al., 1996; Cho et al.,2001; Turrell et al., 2008). Evolutionary and phylogenetic insightsinto P. pungens have been inferred from the nuclear encoded largesubunit (LSU) ribosomal DNA gene (Lundholm et al., 2002; Orsiniet al., 2002), the internal transcribed spacer (ITS) region (Casteleynet al., 2008; Churro et al., 2009; Quijano-Scheggia et al., 2010), thesecondary structure of the ITS2 transcript (Lim et al., 2012a), andthe chloroplast-encoded RUBISCO (rbcL) gene (Casteleyn et al.,2009a; Lundholm et al., 2012).

The spatial and temporal genetic composition of P. pungens

populations has been investigated widely. Its cosmopolitanism hasprovided an incentive for it to be a model species in the genus. Thepopulation genetic structure of P. pungens in the North Sea wasassessed using specific microsatellite markers (Evans and Hayes,2004), and findings supported a single largely unstructuredpopulation (Evans et al., 2005; Casteleyn et al., 2009b), laterreferred to as clade I sensu Casteleyn et al. (2008, 2010). On theother hand, microsatellite data revealed two distinct geneticpopulations of P. pungens from the Washington coast (Adams et al.,2009). This was further supported by Casteleyn et al. (2009a), whoshowed that one of the populations was actually composed of theoffspring from the hybridization between P. pungens var. pungens

and var. cingulata. In a limited number of taxa analyzed, thetropical and warm-temperate water populations split out as adistinct clade, i.e. clade III sensu Casteleyn et al. (2008, 2010)(Churro et al., 2009). The recognition of varieties based onmorphology was unclear (e.g. Villac and Fryxell, 1998). As well,molecular phylogenetic reconstruction, particularly for var. pun-

gens, did not clearly resolve sub-specific boundaries, as isolateswere distributed throughout the phylogenetic tree (Casteleynet al., 2008, 2010; Churro et al., 2009). Generally, Clade I sensu

Casteleyn et al. (2010) comprises var. pungens from the Northernand Southern hemispheres; clade II has only been found in NEPacific and comprises both var. pungens and var. cingulata, theoffspring of both varieties co-occurring in clades I and II; and cladeIII comprises var. aveirensis, and is found only in the tropical towarm-temperate zones.

In this study, we examined the genetic structure of P. pungens ofthe tropical/warm-temperate lineage (clade III), with the inclusionof tropical Southeast Asian populations. The species is widelydistributed along the coasts of Malaysia (the Peninsular andMalaysian Borneo), where it has been documented at 17 locations(Teng et al., 2013). We analyzed sequence data from 47 strains of P.

pungens isolated from the Malaysian tropics and compared them toother populations distributed globally. Genetic differentiation wasassessed using the ITS region, and its secondary structureinformation, to determine the phylogeographic position of thistropical Southeast Asian population in relation to the otherlineages. Our findings demonstrate that the global dispersal routeof P. pungens may not be attributed exclusively to anthropogenicfactors (e.g. ships’ ballast water), but rather to global oceaniccirculation patterns and an ancient geological event.

2. Materials and methods

2.1. Sampling and culture collections

Plankton samples were collected from six coastal locations inMalaysia (South China Sea) between 2009 and 2012, by verticalhauls of a 20-mm mesh plankton net from 3 to 5 m to the surface.The locations were Santubong and Muara Tebas, in Kuching(Sarawak); Kudat, in Sabah; Teluk Batik, in Perak; Klang, in

Selangor; and Port Dickson, in Negeri Sembilan (Fig. 1, Supple-mentary Table S1). Clonal cultures of P. pungens were isolated andestablished as in Lim et al. (2012a,b). In brief, single chains ofPseudo-nitzschia cells were isolated using a finely drawn Pasteurpipette under an Olympus IX51 inverted light microscope(Olympus, Tokyo, Japan) and inoculated into a micro-well plate.Clonal cultures were established in SWII medium (Iwasaki, 1961)with a salinity of 30 psu, pH of 7.8–7.9, and maintained at25 � 0.5 8C under a light intensity of 100 mmol photons m�2 s�1,12:12-h L:D photoperiod, provided by cold-white fluorescent bulbsin a growth chamber (SHEL Lab, Cornelius, OR, USA). Forty sevenstrains were established into clonal culture (Supplementary TableS1).

2.2. DNA extraction, PCR and taxon sampling

Cultures of P. pungens at mid-exponential phase were harvestedby centrifugation (9300 � g for 10 min). Genomic DNA of P.

pungens was extracted with cetyl-trimetyl ammonium bromide(CTAB) lysis buffer and precipitated with ethanol as in Lim et al.(2012a). The nuclear encoded internal transcribed spacer (ITS)region was amplified using the primer set ITS1 and ITS4 (Whiteet al., 1990). Gene amplification of 25 ml PCR mixtures was as inLim et al. (2012a). Thirty five cycles (30 s at 94 8C, 30 s at 51 8C,1 min at 72 8C) were performed with an initial denaturation step of4 min at 94 8C and a final elongation step of 7 min at 72 8C. Theamplicons were subsequently purified using a QIAquick PCRpurification kit (QIAGEN, Hilden, Germany). Sequencing wasperformed on both strands using the same primers. AvailableITS sequences of P. pungens were retrieved from NCBI GenBank. Atotal of 137 sequences were retrieved, and sequences withambiguous bases were excluded; only 117 sequences were usedin the analyses. Sequences obtained in this study were deposited inNCBI GenBank, with the accessions provided (SupplementaryTable S1).

2.3. ITS folding

Using the ITS2 database (Koetschan et al., 2012; Merget et al.,2012), ITS secondary structures were predicted by homologymodeling (Wolf et al., 2005a). ITS secondary structures of P.

pungens, manually predicted according to Casteleyn et al. (2008),served as a template. The ITS2 proximal stem was taken intoconsideration (Cote and Peculis, 2001; Keller et al., 2009). 5.8Ssecondary structures were manually predicted according toGottschling and Plotner (2004). Consensus structures weredisplayed using RNALogo (Chang et al., 2008).

2.4. Genetic diversity

In accordance with Keller et al. (2010), phylogenetic analysesfollowed the procedures outlined in Schultz and Wolf (2009). Aglobal multiple sequence-structure alignment was automaticallygenerated in 4SALE v1.7 (Seibel et al., 2006, 2008), wherebysequences and their secondary structures were simultaneouslyaligned using an ITS2 sequence-structure specific scoring matrix(Seibel et al., 2006). Based on the sequence-structure alignment,phylogenetic relationships were reconstructed by Neighbor-joining (NJ) through the use of an ITS2 sequence-structure specificgeneral time reversible (GTR) substitution model as implementedin ProfDistS v0.9.9 (Friedrich et al., 2005; Wolf et al., 2008).Compensatory base changes (CBCs) as well as hemi-CBCs wereidentified using the CBCAnalyzer option implemented in 4SALEv1.7 (Wolf et al., 2005b; Seibel et al., 2006, 2008). Additionally,using MEGA5 (Tamura et al., 2011), the sequence alignment wasused to calculate uncorrected pairwise p-distances.

Fig. 1. Maps showing sampling stations. (Upper panel) General location of the stations in Malaysia. Detailed locations of stations (A) Santubong and Muara Tebas, in Kuching

(Sarawak); (B) Kudat, in Sabah; (C) Teluk Batik, in Perak; (D) Klang, in Selangor; and (E) Port Dickson, in Negeri Sembilan.

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152144

2.5. Population structure inference

To obtain an estimate of population structure, KST (the analog ofFST that incorporates interallelic distances) was calculated from ananalysis of molecular variance (AMOVA), using Arlequin 3.5.1.2(Excoffier and Lischer, 2010). Population pairwise F-statistics werealso calculated to estimate the gene flow, according to Weir andCockerham (1984). A median-joining network was reconstructedusing Network 4.6.1.0 to infer haplotype relationships (Bandelt et al.,1999). The input file for population structure analysis was convertedusing GenAIEx 6.5b3 (Peakall and Smouse, 2012). Populationstructure and admixture were investigated using Bayesian clusteringanalyses implemented in STRUCTURE v. 2.3.4 (Pritchard et al., 2000).The most likely number of K clusters was identified through 10iterations performed at each K value, from 1 to 10; each run wasbased on 10,000 burn-in periods and 10,000 Markov Chain MonteCarlo repetitions. Simulations were performed using the admixturemodel with the LOCPRIOR option, assuming correlated allele

frequencies among populations. The log likelihood of these datawas used to estimate the best K following the recommendation ofPritchard et al. (2000). Multiple runs generated by STRUCTURE wereanalyzed using Structure Harvester Web v0.6.93 (Earl and Vonholdt,2012). The individual membership coefficient (qind) was obtainedfrom Structure Harvester for the selected K value. Cluster Matchingand Permutation Program, CLUMPP v. 1.1.2 (Jakobsson and Rosen-berg, 2007) was used to determine the optimal assignation of clustersfor the analyzed individuals, maximizing similarity between thedifferent structure replications. The results from the Bayesiananalysis were visualized using DISTRUCT v. 1.1 (Rosenberg, 2004).

3. Results

3.1. ITS folding and structural information

Nucleotide sequences of P. pungens obtained in this study weredeposited in NCBI GenBank and accession numbers are given in

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152 145

Supplementary Table S1. The P. pungens var. aveirensis isolatesfrom Malaysia are identical genetically to those described byChurro et al. (2009) from the Aveiro coastal lagoon, Portugal, andthe P. pungens var. pungens Malaysian isolates are identical to P.pungens var. pungens in Casteleyn et al. (2009a).

The consensus structures of P. pungens ITS1 and ITS2 computedfrom all the sequences analyzed are illustrated in Fig. 2, andsequence divergences were determined (Table 1). The secondarystructures of ITS1 and ITS2 are highly conserved, given fivecommon helices (I–V) in ITS1 (Fig. 2A), and four helices (I–IV) witha pseudo-helix, IIa, in ITS2 (Fig. 2B). We observed a U–U mismatchat the base of helix II and a AAA motif between helix II and pseudo-helix IIa, which are not shown in Casteleyn et al. (2008), butdescribed for eukaryotes (cf. Schultz et al., 2005). The folding ofITS2 was determined based on the 5.8S–28S interaction (Fig. 2B,Keller et al., 2009). This interaction formed a proximal stem, whichhas been demonstrated in yeast to be a processing site in pre-rRNAprocessing (Cote and Peculis, 2001). Hence, this confirmed thetermini of the ITS2 transcript, i.e. the exact nucleotide positionwhere the ITS2 starts (50-CCCAC-30) and ends (50-AUACA-30) (seeFig. 2B for details).

Analysis of the ITS1 consensus secondary structure of P. pungens

showed two hemi-compensatory base changes (HCBCs) in helices I,

Fig. 2. The (A) ITS1 and (B) ITS2 consensus secondary structure of P. pungens inferred with

helix IIa in ITS2).

II and III, three HCBCs in helix V, and 26 single nucleotidepolymorphisms (SNPs) for the whole structure. In ITS2, one HCBCwas identified in helices II and IV, one CBC and nine HCBCs in helixIII, and 26 SNPs for the whole structure.

3.2. Phylogeography of P. pungens

The ITS tree revealed three major clades (I, II, III), with strongsupport of high posterior probability in the NJ analysis (Fig. 3).Clade I represents strains from the North Sea (Belgium, Denmark,France, The Netherlands), NE Atlantic (Spain, Portugal), NWAtlantic (Canada), NE Pacific (USA), NW Pacific (Japan) and SPacific (New Zealand); strains belong to var. pungens. Clade IIcontains strains from the NE Pacific, belonging to both var. pungens

and var. cingulata; var. cingulata was found only in this clade.Strains co-occurring in both clade I and clade II represent thehybrids (33 clones) of var. pungens and var. cingulata in the naturalenvironment. Both clades I and II were strongly supported by ahigh posterior probability value of 0.97 (Fig. 3A). Clade III wasseparated into two sub-clades, i.e. sub-clade IIIa and sub-clade IIIb,with strong posterior probability support (0.95, Fig. 3B). This cladecomprised both var. pungens and var. aveirensis. Sub-clade IIIa wasmainly from the W Pacific (Straits of Malacca, Malaysian Borneo,

RNA Logo. Major helices are labeled (helix I–V in ITS1) and (helix I–IV and pseudo-

Table 1Sequence divergence (uncorrected p-values) within and among the three Pseudo-nitzschia pungens clades, shown as minimum-maximum (average).

Clade I Clade II Clade III

ITS* ITS1 ITS2 ITS* ITS1 ITS2 ITS* ITS1 ITS2

Clade I (n = 75) 0–0.007 (0.001) 0–0.020

(0.001)

0–0.010 (0.001)

Clade II (n = 35) 0.006–0.022

(0.014)

0–0.032

(0.013)

0.013–0.029

(0.021)

0–0.011

(0.004)

0–0.020

(0.005)

0–0.013

(0.004)

Clade III (n = 54) 0.022–0.038

(0.026)

0.016–0.051

(0.024)

0.032–0.052

(0.041)

0.025–0.043

(0.032)

0.016–0.051

(0.029)

0.042–0.061

(0.049)

0–0.018

(0.005)

0–0.028

(0.005)

0–0.026

(0.007)

* ITS region = ITS1, 5.8S rDNA and ITS2.

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152146

China and Vietnam). Sub-clade IIIb comprised four NE Atlanticstrains (from Portugal), one from W Atlantic (Mexico) and one fromW Pacific (Malaysia). The splits among clades are substantial, withaverage sequence divergences ranging from 1.4 to 3.2% in the ITSregion (Table 1). Within clades, sequence divergences arerelatively low, averaging 0.1% in clade I; 0.4% in clade II; and0.5% in clade III (Table 1).

3.3. Haplotype diversity

Sequence analysis revealed 60 unique haplotypes in 164sequences from the eight geographical regions. The whole ITSregion was highly variable, with 105 differing nucleotide sites; thedetails of each haplotype are listed in Supplementary Table S2.Fig. 4 shows the median-joining network derived from the wholeITS sequences of P. pungens.

A haplogroup of P. pungens from Malaysia comprised 14haplotypes (h2, h4, h5, h42, h44, h46–h48, h51, h52, h56–h59). Thepopulation of the Strait of Malacca formed a haplogroup, withseven haplotypes, followed by the strains from Sulu Sea (Kudat),with five haplotypes. Kuching, which is situated at the southernSouth China Sea (Fig. 1), has only one haplotype (Fig. 4).

Fig. 3. Neighbor-joining tree of P. pungens derived from P. pungens ITS sequence-struct

Rectangles indicate var. cingulata; dark circles indicate var. aveirensis. Posterior probabilit

Only one haplotype (h1) was observed in more than onepopulation; strains belong to this haplotype were from the NorthSea, NE Atlantic, NW Atlantic, NE Pacific, NW Pacific and S Pacific.Haplotype h43 occurred in the NE and W Atlantic populations.Among the rest of the haplotypes, none were found in more thantwo populations. Haplotype h2, and adjacent haplotypes, wereconfined to the W Pacific, whereas h16 and h50, with adjacenthaplotypes, were restricted to the NE Pacific. Among thehaplogroups, the NE Pacific showed the highest haplotypediversity, with 30 haplotypes (Fig. 4).

Pseudo-nitzschia pungens grouped not only according to thehaplotypes but also according to their geographical locales. In thisnetwork, G3 had long branch-lengths, with G1 and G2 separated by15 mutational steps (Fig. 4). From the network, haplotype h1 wasthe haplotype with the highest frequency in G1, followed byhaplotype h2 in G3. The grouping in this network is in agreementwith the clustering in the NJ phylogenetic tree (Fig. 3).

According to AMOVA, 75% of the total ITS rDNA variation wasfound among the populations, and only 25% within populations(KST = 0.75). Standardized multilocus (FST) values were relativelyhigh between strains from the W Pacific compared to the othersfrom the NW, NE and S Pacific; NE and NW Atlantic (0.69–0.88),

ure information. The tree was rooted with P. multiseries (DQ062664) (not shown).

ies of more than 0.50 are shown. Figure is split into (A) and (B) for better illustration.

Fig. 4. Median-joining network derived from P. pungens ITS nucleotide sequences. Haplotypes are designated by numbers. Each tick on the line connecting haplotypes

represents a mutational step. The haplogroups are assigned to G1, G2 and G3, where G1 corresponds to Clade I, G2 to clade II and G3 to clade III in Fig. 3. The dashed ovals

correspond to sub-clades IIIa and IIIb.

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152 147

except for those from the W Atlantic (0.54) (Table II). Among thestrains of P. pungens from the W Pacific, pairwise geneticdifferences ranged from 1.04 to 3.17%.

3.4. Population structure

The most likely number of K is inferred using the methoddescribed in Evanno et al. (2005). The second order statisticdevised showed a peak at K = 2, the first value with P value >0.05 inthe simulations (Supplementary Fig. S1). The global distributionmap for P. pungens is plotted according to the locations of thestrains listed in Supplementary Table S1 and the literature (Lelonget al., 2012; Trainer et al., 2012) (Fig. 5A). The STRUCTURE analysisrevealed two genetic clusters (Fig. 5B). The phylogeographicdistribution of P. pungens is congruent with the STRUCTUREanalysis, where the W Pacific represents one distinct phylogroupand the rest of the populations (North Sea, NE/NW/W Atlantic,NW/SW Pacific and NE Pacific) are panmictic.

4. Discussion

4.1. Global patterns of P. pungens genetic variation

Across eukaryotes, the ITS region of the nuclear-encodedribosomal RNA gene has been evaluated as useful for inferringphylogenies and/or for distinguishing species (e.g. Coleman, 2003;

Muller et al., 2007). Our data show three distinct ITS entities(clades I, II and III; Fig. 3), in agreement with Casteleyn et al. (2008).Several strains in both clade I and II were reported as hybrids of P.

pungens var. pungens and P. pungens var. cingulata in the naturalenvironment of the NE Pacific (USA) (Casteleyn et al., 2009a). Thiswas further supported in our analyses by the moderately low FST

values, indicative of a genetic exchange through mating. The highdegree of gene flow observed in clade I between the North Sea andNE Atlantic (0.30) is comparable to that of the marine diatomSkeletonema marinoi in the North Sea (0.011–0.267; Godhe andHarnstrom, 2010).

Pseudo-nitzschia pungens has always been described as cosmo-politan, with a global distribution (Hasle, 2002; Casteleyn et al.,2008; Lelong et al., 2012). However, our study shows that geneflow is restricted between tropical/subtropical P. pungens (growingat >20 8C; NOAA, 2013) and temperate P. pungens (10–20 8C;NOAA, 2013). This is evidenced by the high FST value of W Pacificpopulations when compared to those of the NE Pacific, North Seaand NW Atlantic populations (0.81, 0.88 and 0.86, respectively).Similar trends were also observed in other plankton, e.g. thecalanoid copepod Subeucalanus pileatus, where gene flow was lowbetween populations of the W Pacific and North Atlantic(FST = 0.979–0.985; Goetze and Ohman, 2010). The limit of geneflow was discussed in Casteleyn et al. (2010) using species-specificmicrosatellites markers. It is still premature to compare popula-tions from NW Pacific and S Pacific, even though they showed high

Fig. 5. The geographical distribution of P. pungens and the population structure. (A) Colored lines represent global thermohaline circulations (ocean ‘‘conveyor belt’’); blue line

indicates cold, salty, dense, deep water and red line indicates warm, shallow water. Colored circles indicate the strains as verified in the present study based on ITS sequences:

gray = clade I; blue = clade II; red = clade III. Black dots indicate morphology-based records of P. pungens; modified from Lelong et al. (2012). Water surface temperature based

on NOAA/NESDIS GEO-Polar Blended 5-km SST analysis (NOAA, 2013). (B) Population structure based on STRUCTURE analysis estimated from 164 individuals for K = 2. Each

individual is represented by a vertical bar that is partitioned into K color fragments of the individual’s estimated membership fractions in K cluster.

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152148

FST values, because the population size is small (n = 2). Furtherattention is warranted before a strong conclusion can be drawn;indeed, available sequence data of P. pungens from the maincontinents of the Southern Hemisphere (South America, SouthAfrica, Australia) are still limited.

To date, clade III is confined to tropical (W Pacific, W Atlantic) andsubtropical (NE Atlantic) waters. The relatively high FST values(0.81–0.88; Table 2) also explain the restriction between popula-tions of clade III and those of the other two clades from temperateregions, suggesting that they are reproductively isolated. Clade III isclustered into two sub-clades (IIIa and IIIb) with phylogeographicstructuring. Sub-clade IIIa comprises tropical P. pungens populationsfrom the W Pacific (South China Sea), whereas sub-clade IIIb ismainly from the subtropical W and NE Atlantic. In spite of thepopulations from South China Sea (W Pacific) and W/NE Atlanticbelonging to two different oceans, the gene flow (FST = 0.54–0.69) iscomparable to that between the North Sea and NE Atlantic comparedwith NE Pacific (FST = 0.60 and 0.46, respectively). Thus, the distancebetween populations from the South China Sea (W Pacific) and theNorth Atlantic Ocean (W/NE Atlantic) did not prevent geneticexchange among the populations in clade III.

Table 2Genetic divergence of the ITS region in Pseudo-nitzschia pungens populations. Upper diag

average number of pairwise genetic differences within population (PiX, %); lower diagon

W Pacific NW Pacific NE Pacific S Pa

W Pacific (n = 49) 0.48 0.83 0.81 0.83

NW Pacific (n = 2) 2.69 0 0.42 0

NE Pacific (n = 49) 3.17 1.11 0.76 0.42

S Pacific (n = 2) 2.69 0 1.11 0

North Sea (n = 37) 2.76 0.06 1.17 0.06

NE Atlantic (n = 15) 2.27 0.77 1.71 0.77

NW Atlantic (n = 9) 2.69 0 1.11 0

W Atlantic (n = 1) 1.04 2.77 3.25 2.77

The geographic structuring in P. pungens clade III is supported inthe haplotype analysis. However, this pattern is tentative, as thenumber of available ITS sequences from China and Vietnam islimited (n = 1) (Fig. 4). Genetic differentiation within P. pungens canbe broadly linked to the geographical origin of the strains. Clades I,II and III can be considered as three distinct phylogroups that couldhave evolved separately over a long period of time at differentgeographic regions with different ecological conditions. The resultsof NJ and median-joining network analyses showed clear group-ings that are congruent with geographical locales. This was furthersupported by STRUCTURE analysis with detection of admixing, asshown in Fig. 5B. The populations of the North Sea, the NE Atlanticand the NW Atlantic occupy a huge geographic area connected bythe North Atlantic Ocean, and together with strains from the NEand S Pacific, form one panmictic population. On the other hand,the South East Asian populations, which originated from threedifferent areas of the Indo-Pacific region, i.e. the Indian Ocean(Strait of Malacca, Malaysia), the South China Sea and the Sulu Sea(W Pacific), represent another panmictic population.

It is noteworthy that P. pungens var. pungens is distributedthroughout the lineages (within clades I and II, as well as clade III in

onal: Population pairwise standardized multilocus (FST) values; diagonal element:

al: average number of pairwise genetic differences between populations (PiXY, %).

cific North Sea NE Atlantic NW Atlantic W Atlantic

0.88 0.69 0.86 0.54

�0.33 �0.12 0 1

0.60 0.46 0.52 0.77

�0.33 �0.12 0 1

0.13 0.30 �0.06 0.96

0.83 1.21 0.15 0.41

0.06 0.77 0 1

2.83 2.06 2.77 0

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152 149

this study), where the temperate ocean populations are anti-tropical, i.e. they can only be found north and south of the tropicalzone but not in the tropics (Casteleyn et al., 2010). A possibleexplanation for this is that an ecological barrier could haveseparated P. pungens var. pungens into two geographic regions,tropical and temperate. This separation could be caused bydifferences in water temperature, salinity and light intensity,which have occurred in both regions, providing clinal differentia-tion (ecotypes without any morphological changes) that allowedsubsequent genetic differentiation of the populations of var.pungens above and below the tropics. Even though P. pungens var.pungens from temperate (Northern and Southern hemispheres) isseparated by the equator, this did not affect the gene flow.

Information on diatom resting stages is important forunderstanding the population dynamics, including gene flow,dispersion and colonization, of diatoms. In the eastern Bering Seashelf of the North Pacific Ocean, the life cycle strategy of restingstages allows cells to survive the dark winter season and formblooms during the spring (Tsukazaki et al., 2013). A similar strategywas encountered in the tropical East China Sea of the WesternPacific Ocean, where Zhang et al. (2010) showed that thegermination of diatom resting stages in sediment samples wassignificantly affected by light intensity, rather than by watertemperature and salinity. There is contradictory informationregarding the existence of resting stages in Pseudo-nitzschia spp.(Lelong et al., 2012), and no evidence of them was found in the Gulfof Naples (Montresor et al., 2013). On the other hand, resting stagesof Pseudo-nitzschia sp. were reported in recent marine sediments ofPeter the Great Bay, Russia (Orlova and Morozova, 2009). As well, P.

multiseries vegetative cells germinated from the sediments of theEast China Sea (Zhang et al., 2010). Clearly, Pseudo-nitzschia spp.must possess some kind of resting stage that allows them to retainthe physiological capacity to germinate upon the restoration ofproper environmental conditions.

Our hypothesis is that the thermohaline circulation of theglobal ‘‘conveyor belt’’ (Roach, 2004) may allow P. pungens var.pungens to be drawn to the depths, where it travels in a lowmetabolic state from the N Atlantic (Northern Hemisphere) to the SAtlantic (Southern Hemisphere) (blue line in Fig. 5A), where it thenupwells and grows. One branch of the circulation carries cellsnorthward into the Indian Ocean and the other branch up into theW Pacific. The warm shallow current (red line in Fig. 5A) can thendeliver the tropical P. pungens var. pungens and P. pungens var.aveirensis from the W Pacific, across the Indian Ocean and thennorthward, where the cells are eventually returned to the NAtlantic, and the cycle begins again (Fig. 5A). However, deep oceancurrents have not always existed or circulated in the same way, buthave been shut down or changed direction many times during theEarth’s history (Broecker, 1997; Driscoll and Haug, 1998). Ashutdown in global oceanic circulation may have allowed forprolonged periods of geographical isolation, during which geneticdifferentiation occurred. The evolutionary diversification patternof P. pungens is consistent with this historical ocean circulation.

4.2. Phylogeographic hypothesis and diversification of P. pungens

The raphid pennate diatoms were predicted to diverge from themore recent araphid pennate diatoms �93.8 Mya, and started amajor diversification process �75 Mya (Sorhannus, 2007). InPseudo-nitzschia, particularly, P. pungens diverged from P. multiseries

between 1.9 and 5.9 Mya (Sorhannus, 2007). Based on thedivergence time estimated by Casteleyn et al. (2010), we speculatethat the tropical and subtropical P. pungens (clade III) underwentspeciation and diverged 0.6–1.6 Mya during the Lower Pleistocene(Calabrian). Casteleyn et al. (2010) provided several explanations forthis speciation. Another hypothesis for the global diversification of P.

pungens could be the formation of the Isthmus of Panama, alsoknown as the Central American (Inter-American) Seaway, whichclosed �3 Mya. Before the seaway closed, both oceans had the samesalinity, which was maintained by the flow from the Atlantic and thePacific Oceans through the Central America Seaway. However, whenthe Isthmus of Panama formed, water exchange between theAtlantic and Pacific Oceans was restricted, salinities changed and theconnection of conspecific planktonic populations from both oceanswas disrupted (Haug et al., 2004). This event could have separatedthe P. pungens populations into two ocean basins, thereby allowingthem to become isolated from each other. The closure resulted in amarked reorganization of ocean circulation, starting �4.6 Mya; itenhanced the flow of the Gulf Stream and possibly intensifiedNorthern Hemisphere glaciations due to the introduction of warm,saline water (Haug and Tiedemann, 1998). The tropical Atlantic andCaribbean became saltier via evaporation, and trade windsdeposited the water vapor westward across the isthmus into thePacific via rainfall, resulting in a less saline Pacific Ocean (Haug et al.,2004). Transport via ships’ ballast water could also spread P. pungens

around the world. However, the time frame for this species to evolveis limited, as the use of water for ballast in vessels began only afterthe 1880s, when steel-hulled vessels were introduced (NationalResearch Council (NRC), 1996; Pughuic, 2001).

Changes in ocean circulation driven by temperature and salinitymight have triggered the differentiation of P. pungens intomorphological (differences in valvocopula structure and bandstriae density) and genetic varieties, through adaptation. Inparticular, the structure of the band striae (cingular bands),composed of three bands (the valvocopula, second band and thirdband), is the most pronounced characteristic that distinguishesamong the three varieties of P. pungens. The valvocopulae of P.

pungens var. pungens (clade I) have a single row of poroids, andthose of P. pungens var. cingulata (clade II) have two rows of 2–3poroids. The second band may have one or two poroids per stria;even if one poroid is present, it is much smaller compared to P.

pungens var. pungens (Villac and Fryxell, 1998). On the other hand,P. pungens var. aveirensis (clade III) is characterized by two types ofvalvocopulae, one with one row of oval poroids and the other withsquare poroids split into 2–3 parts (Villac and Fryxell, 1998; Churroet al., 2009). This phenotypic variation may have been triggered byadaptation, in particular to water temperature, as an environmen-tal factor that influenced the speciation in P. pungens.

Laboratory experiments have shown morphological differencesin Pseudo-nitzschia species when grown at different temperatures.For example, P. multiseries had the usual 3–4 rows of poroids in thestriae when grown at 5 8C and 15 8C, but this decreased to 2–3 rowsat 25 8C (Lewis et al., 1993). Similarly, the number of rows ofporoids of the temperate species P. seriata decreased from four atits usual growth temperature (4 8C) to two, at 15 8C (Hansen et al.,2011). Thus, the speciation of P. pungens populations could betriggered when given enough time in cold water environments (cf.Butlin et al., 2012). The highly dispersed P. pungens can be areflection of the degree of adaptation to different environments,resulting in the independent evolution of P. pungens var. aveirensis

(in tropical and subtropical waters) and P. pungens var. cingulata (intemperate waters).

4.3. Toxicity of P. pungens

The distinction among phytoplankton species, especiallyPseudo-nitzschia and their structured populations, remains con-tentious due to difficulties in defining the species’ boundaries andthe lack of available genetic and ecological data. Compensatorybase changes correlating with the biological species concept(Muller et al., 2007; Coleman, 2009; Wolf et al., 2013) have beenwidely adopted, along with a combination of morphology and

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152150

phylogenetic reconstructions, to discriminate among species ofPseudo-nitzschia (Churro et al., 2009; Lim et al., 2012a; Lundholmet al., 2012). This has been particularly important when dealingwith potentially toxic species. Unfortunately, such data are limitedfor the three varieties of P. pungens, with the exception of thosepresented in Churro et al. (2009).

Up until 1994, P. pungens was considered as a non-toxic species(Bates et al., 1998). However, it was then proven to toxifyGreenshell mussels in Marlborough Sounds, New Zealand (SPacific), and to produce low levels of DA (0.47 pg cell�1) in oneisolate (Rhodes et al., 1996). No other toxic strain was reporteduntil a strain of P. pungens from Twin Harbors (Washington State,USA; NE Pacific) again produced low levels of DA (0.07 pg cell�1)(Trainer et al., 1998). Another isolate from offshore WashingtonState (NE Pacific) produced even less DA (0.0018 pg cell�1) (Baughet al., 2006). An isolate from the Bay of Crozon (France, E Atlantic)subsequently showed a level of 0.2 pg DA cell�1 (Calu et al., 2009).All of the above studies either pre-date the acceptance that P.

pungens could be separated into three varieties, or else the varietywas not specified. Churro et al. (2009) did not detect DA in culturesof P. pungens var. aveirensis from Portugal (NE Atlantic).Moschandreou et al. (2012) reported that one strain (out of nine)of P. pungens var. pungens from Greek coastal waters (NEMediterranean basin) produced trace (but not quantifiable) levelsof DA. This was the first report of toxicity in P. pungens var. pungens.More data are required to find patterns of toxicity among the threeP. pungens varieties.

Interestingly, cultured strains of both P. pungens var. pungens

and var. aveirensis examined in Malaysia were found to be non-toxic (Lim et al., 2010). Strains of P. pungens (variety not specified)isolated from the East China Sea (W Pacific) were also reported tobe non-toxic (Li et al., 2005). This is consistent with the finding thatP. brasiliana was not toxic in tropical waters (Malaysia [Lim et al.,2010, 2012b,c]; China [Wang et al., 2012]), but toxic in subtropicalwaters (Tunisia; SW Mediterranean [Sahraoui et al., 2011]),although it should be noted that P. brasiliana from Spain (NWMediterranean) was reported to contain DA below the limit ofdetection (0.53 fg DA cell�1) (Quijano-Scheggia et al., 2010). It ishence hypothesized that low temperatures favor the production ofDA in certain species. This is also supported by Amato et al. (2010),who showed that the ability for DA production in P. multistriata

was suppressed at 27 8C (the optimum temperature was at 18 8C).Interestingly, it should be noted that the toxic P. pungens

documented in Rhodes et al. (1996) and Moschandreou et al.(2012) were maintained at a lower temperature (18 � 1 8C and19 � 2 8C, respectively) compared to the apparently non-toxic strainsfrom Malaysia (25 � 0.5 8C).

5. Conclusion

Our results show that gene flow is limited between P. pungens

from tropical and temperate waters. The morphological varieties ofP. pungens var. pungens, var. cingulata and var. aveirensis areconsistent with genetic groupings. We hypothesize that oceancirculation and changes in physical barriers could explain thecurrent dispersal patterns of P. pungens, as well as other marinespecies, rather than anthropogenic activities alone. These condi-tions might thus play an important role in understanding theevolutionary pattern of marine species diversification. Moleculardata were essential in defining this biogeography.

Acknowledgments

We thank Dr. Matthias Wolf (Department of Bioinformatics,Biocenter, University of Wuerzburg, Germany) for his reviewedcomments on the ITS2 modeling, and Iratxe Montes (Department

of Genetics, Physical Anthropology and Animal Physiology,University of the Basque Country UPV/EHU, Leioa, Spain) for herkind assistance in using STRUCTURE. This work was funded by theMalaysian government through the Ministry of Education underJSPS COMSEA Matching Fund (GA002-2014); the Ministry ofScience, Technology and Innovation under Science Fund (04-01-09-SF0092); and University of Malaya under UMRP (RU006-2014).HC Lim was supported by the Ministry of Higher Education MyPhDscholarship, and this work formed part of his PhD project.

[SS]

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.hal.2014.06.004.

References

Adams, N.G., Trainer, V.L., Rocap, G., Herwig, R.P., Hause, L., 2009. Genetic popula-tion of Pseudo-nitzschia pungens (Bacillariophyceae) from the Pacific Northwestand the North Sea. J. Phycol. 45, 1037–1045.

Amato, A., Ludeking, A., Kooistra, W.H.C.F., 2010. Intracellular domoic acid produc-tion in Pseudo-nitzschia multistriata isolated from the Gulf of Naples (TyrrhenianSea, Italy). Toxicon 55, 157–161.

Amzil, Z., Fresnel, J., Le Gal, D., Billard, C., 2001. Domoic acid accumulation in Frenchshellfish in relation to toxic species of Pseudo-nitzschia multiseries and P.pseudodelicatissima. Toxicon 39, 1245–1251.

Bandelt, H.J., Forster, P., Rohl, A., 1999. Median-joining networks for inferringintraspecific phylogenies. Mol. Biol. Evol. 16, 37–48.

Bates, S.S., Garrison, D.L., Horner, R.A., 1998. Bloom dynamics and physiology ofdomoic-acid-producing Pseudo-nitzschia species. In: Anderson, D.M., Cembella,A.D., Hallegraeff, G.M. (Eds.), Physiological Ecology of Harmful Algal Blooms.Springer-Verlag, Heidelberg, pp. 267–292.

Baugh, K.A., Bush, J.M., Bill, B.D., Lefebvre, K.A., Trainer, V.L., 2006. Estimates ofspecific toxicity in several Pseudo-nitzschia species from the Washington coast,based on culture and field studies. Afr. J. Mar. Sci. 28, 403–407.

Broecker, W.S., 1997. Thermohaline circulation, the Achilles heel of our climate system:will man-made CO2 upset the current balance? Science 278, 1582–1588.

Butlin, R., Debelle, A., Kerth, C., Snook, R.R., Beukeboom, L.W., Cajas, R.F.C., Diao, W.,Maan, M.E., Paolucci, S., Weissing, F.J., Zande, L.V.D., Hoikkala, A., Geuverink, E.,Jennings, J., Kanakare, M., Knott, K.E., Tyukmaeva, V.I., Zoumadakis, C., Ritchie,M.G., Barker, D., Immonen, E., Kirkpatrick, M., Noor, M., Garcia, C.M., Schmitt, T.,Schilthuizen, M., 2012. What do we need to know about speciation? TrendsEcol. Evol. 27, 27–39.

Calu, G., Martin-Jezequel, V., Lefau, E., Sechet, V., Lassus, P., Weigel, P., Amzil, Z.,2009. The influence of nitrogen speciation on growth and toxicity of Pseudo-nitzschia multiseries and P. pungens in batch and continuous cultures. In: Lassus,P. (Ed.), Seventh International Conference on Molluscan Shellfish Safety, Edi-tions Quæ, Nantes, p. 7.

Casteleyn, G., Chepurnov, V.A., Leliaert, F., Mann, D.G., Bates, S.S., Lundholm, N.,Rhodes, L., Sabbe, K., Vyverman, W., 2008. Pseudo-nitzschia pungens (Bacillar-iophyceae): a cosmopolitan diatom species? Harmful Algae 7, 241–257.

Casteleyn, G., Adams, N.G., Vanormelingen, P., Debeer, A.-E., Sabbe, K., Vyverman,W., 2009a. Natural hybrids in the marine diatom Pseudo-nitzschia pungens(Bacillariophyceae): genetic and morphological evidence. Protist 160, 343–354.

Casteleyn, G., Evans, K., Backeljau, T., D’hondt, S., Chepurnov, V., Sabbe, K., Vyver-man, W., 2009b. Lack of population genetic structuring in the marine planktonicdiatom Pseudo-nitzschia pungens (Bacillariophyceae) in a heterogeneous area inthe Southern Bight of the North Sea. Mar. Biol. 156, 1149–1158.

Casteleyn, G., Leliaert, F., Backeljau, T., Debeer, A.-E., Kotaki, Y., Rhodes, L., Lund-holm, N., Sabbe, K., Vyverman, W., 2010. Limits to gene flow in a cosmopolitanmarine planktonic diatom. Proc. Nat. Acad. Sci. 107, 12952–12957.

Chang, T.H., Horng, J.T., Huang, H.D., 2008. RNALogo: a new approach to displaystructural RNA alignment. Nucleic Acids Res. 36, W91–W96.

Chepurnov, V.A., Mann, D.G., Sabbe, K., Vannerum, K., Casteleyn, G., Verleyen, E.,Peperzak, L., Vyverman, W., 2005. Sexual reproduction, mating system, chloro-plast dynamics and abrupt cell size reduction in Pseudo-nitzschia pungens fromthe North Sea (Bacillariophyta). Eur. J. Phycol. 40, 379–395.

Cho, E.S., Park, J.G., Oh, B.C., Cho, Y.C., 2001. The application of species specific DNA-targeted probes and fluorescently tagged lectin to differentiate several species ofPseudo-nitzschia (Bacillariophyceae) in Chinhae Bay. Korean Sci. Mar. 65, 207–214.

Churro, C.I., Carreira, C.C., Rodrigues, F.J., Craveiro, S.C., Calado, A.J., Casteleyn, G.,Lundholm, N., 2009. Diversity and abundance of potentially toxic Pseudo-nitzschia Peragallo in Aveiro coastal lagoon, Portugal and description of anew variety, P. pungens var. aveirensis var. nov. Diatom. Res. 24, 35–62.

Coleman, A.W., 2003. ITS2 is a double-edged tool for eukaryote evolutionarycomparisons. Trends Genet 19, 370–375.

Coleman, A.W., 2009. Is there a molecular key to the level of biological species ineukaryotes? A DNA guide. Mol. Phylogenet. Evol. 50, 197–203.

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152 151

Cote, C.A., Peculis, B.A., 2001. Role of the ITS2-proximal stem and evidence forindirect recognition of processing sites in pre-rRNA processing in yeast. NucleicAcids Res. 29, 2106–2116.

Driscoll, N.W., Haug, G.H., 1998. A short circuit in thermohaline circulation: a causefor Northern Hemisphere glaciation? Science 282, 436–438.

Earl, D., Vonholdt, B., 2012. STRUCTURE HARVESTER: a website and program forvisualizing STRUCTURE output and implementing the Evanno method. Conserv.Genet. Resour. 4, 359–361.

Evanno, G., Regnaut, S., Goudet, J., 2005. Detecting the number of clusters ofindividuals using the software structure: a simulation study. Mol. Ecol. 14,2611–2620.

Evans, K.M., Hayes, P.K., 2004. Microsatellite markers for the cosmopolitan marinediatom Pseudo-nitzschia pungens. Mol. Ecol. Notes 4, 125–126.

Evans, K.M., Kuhn, S.F., Hayes, P.K., 2005. High levels of genetic diversity and lowlevels of genetic differentiation in North Sea Pseudo-nitzschia pungens (Bacillar-iophyceae) populations. J. Phycol. 41, 506–514.

Excoffier, L., Lischer, H.E.L., 2010. Arlequin suite ver 3.5: a new series of programs toperform population genetics analyses under Linux and Windows. Mol. Ecol. Res.10, 564–567.

Falkowski, P., Scholes, R.J., Boyle, E., Canadell, J., Canfield, D., Elser, J., Gruber, N.,Hibbard, K., Hogberg, P., Linder, S., Mackenzie, F.T., Moore Iii, B., Pedersen, T.,Rosenthal, Y., Seitzinger, S., Smetacek, V., Steffen, W., 2000. The global carboncycle: a test of our knowledge of earth as a system. Science 290, 291–296.

Friedrich, J., Dandekar, T., Wolf, M., Muller, T., 2005. ProfDist: a tool for theconstruction of large phylogenetic trees based on profile distances. Bioinfor-matics 21, 2108–2109.

Godhe, A., Harnstrom, K., 2010. Linking the planktonic and benthic habitat: geneticstructure of the marine diatom Skeletonema marinoi. Mol. Ecol. 19, 4478–4490.

Goetze, E., Ohman, M.D., 2010. Integrated molecular and morphological biogeog-raphy of the calanoid copepod family Eucalanidae. Deep-Sea Res., II 57,2110–2129.

Gottschling, M., Plotner, J., 2004. Secondary structure models of the nuclear internaltranscribed spacer regions and 5.8S rRNA in Calciodinelloideae (Peridiniaceae)and other dinoflagellates. Nucleic Acids Res. 32, 307–315.

Hansen, L.R., Soylu, S.I., Kotaki, Y., Moestrup, Ø., Lundholm, N., 2011. Toxin produc-tion and temperature-induced morphological variation of the diatom Pseudo-nitzschia seriata from the Arctic. Harmful Algae 10, 689–696.

Hasle, G.R., 1993. Nomenclatural notes on marine planktonic diatoms. The familyBacillariaceae. Nova Hedwigia Beih. 106, 315–321.

Hasle, G.R., 2002. Are most of the domoic acid-producing species of the diatomgenus Pseudo-nitzschia cosmopolites? Harmful Algae 1, 137–146.

Haug, G.H., Tiedemann, R., 1998. Effect of the formation of the Isthmus of Panama onAtlantic Ocean thermohaline circulation. Nature 393, 673–676.

Haug, G.H., Tiedemann, R., Keigwin, L.D., 2004. How the Isthmus of Panama put icein the Arctic drifting continents open and close gateways between oceans andshift Earth’s climate. Oceanus Mag. 42, 1–4.

Holtermann, K.E., Bates, S.S., Trainer, V.L., Odell, A., Armbrust, E.V., 2010. Masssexual reproduction in the toxigenic diatoms Pseudo-nitzschia australis and P.pungens (Bacillariophyceae) on the Washington Coast, USA. J. Phycol. 46, 41–52.

Iwasaki, H., 1961. The life-cycle of Porphyra tenera in vitro. Biol. Bull. 121, 173–187.Jakobsson, M., Rosenberg, N.A., 2007. CLUMPP: a cluster matching and permutation

program for dealing with label switching and multimodality in analysis ofpopulation structure. Bioinformatics 23, 1801–1806.

Keller, A., Schleicher, T., Schultz, J., Muller, T., Dandekar, T., Wolf, M., 2009. 5.8S–28SrRNA interaction and HMM-based ITS2 annotation. Gene 430, 50–57.

Keller, A., Forster, F., Muller, T., Dandekar, T., Schultz, J., Wolf, M., 2010. IncludingRNA secondary structures improves accuracy and robustness in reconstructionof phylogenetic trees. Biol. Direct 5, 4.

Koetschan, C., Hackl, T., Muller, T., Wolf, M., Forster, F., Schultz, J., 2012. ITS2Database IV: interactive taxon sampling for internal transcribed spacer 2 basedphylogenies. Mol. Phylogenet. Evol 63, 585–588.

Kotaki, Y., Koike, K., Ogata, T., Sato, S., Fukuyo, Y., Kodama, M., 1996. Domoic acidproduction by an isolate of Pseudo-nitzschia multiseries, a possible cause for thetoxin detected in bivalves in Ofunato Bay, Japan. In: Yasumoto, Y., Oshima, Y.,Fukuyo, Y. (Eds.), Harmful and Toxic Algal Blooms. Intergov. Oceanogr. Comm.,UNESCO, Paris, pp. 151–154.

Lelong, A., Hegaret, H., Soudant, P., Bates, S.S., 2012. Pseudo-nitzschia (Bacillario-phyceae) species, domoic acid and amnesic shellfish poisoning: revisitingprevious paradigms. Phycologia 51, 168–216.

Lewis, N.I., Bates, S.S., McLachlan, J.L., Smith, J.C., 1993. Temperature effects ongrowth, domoic acid production, and morphology of the diatom Nitzschiapungens f. multiseries. In: Smayda, T.J., Shimizu, Y. (Eds.), Toxic PhytoplanktonBlooms in the Sea. Elsevier Sci. Publ. B.V, Amsterdam, pp. 601–606.

Li, A., Yu, R., Wang, Y., Yan, T., Zhou, M., 2005. Morphological and toxicity char-acteristics of Pseudo-nitzschia pungens strain PP 0201-01 isolated from the EastChina Sea. Chin. J. Oceanol. Limnol. 23, 418–426.

Lim, H.C., Su, S.N.P., Mohamed-Ali, H., Kotaki, Y., Leaw, C.P., Lim, P.T., 2010. Toxicityof diatom Pseudo-nitzschia (Bacillariophyceae) analyzed using high perfor-mance liquid chromatography (HPLC). J. Sci. Technol. Tropics 6, S116–S119.

Lim, H.C., Leaw, C.P., Su, S.N.P., Teng, S.T., Usup, G., Noor, N.M., Lundholm, N., Kotaki,Y., Lim, P.T., 2012a. Morphology and molecular characterization of Pseudo-nitzschia (Bacillariophyceae) from Malaysian Borneo, including the new speciesPseudo-nitzschia circumpora sp. nov. J. Phycol. 48, 1232–1247.

Lim, H.C., Lim, P.T., Su, S.N.P., Kotaki, Y., Leaw, C.P., 2012b. Morphological observa-tion of two species of Pseudo-nitzschia (Bacillariophyceae). Coast. Mar. Sci. 35,52–57.

Lim, H.C., Lim, P.T., Su, S.N.P., Teng, S.T., Leaw, C.P., 2012c. Genetic diversity ofPseudo-nitzschia brasiliana (Bacillariophyceae) from Malaysia. J. Appl. Phycol.24, 1465–1475.

Lundholm, N., Hasle, G.R., Fryxell, G.A., Hargraves, P.E., 2002. Morphology, phyloge-ny and taxonomy of species within the Pseudo-nitzschia americana complex(Bacillariophyceae) with descriptions of two new species, Pseudo-nitzschiabrasiliana and Pseudo-nitzschia linea. Phycologia 41, 480–497.

Lundholm, N., Bates, S.S., Baugh, K.A., Bill, B.D., Connell, L.B., Leger, C., Trainer, V.L.,2012. Cryptic and pseudo-cryptic diversity in diatoms–with descriptions ofPseudo-nitzschia hasleana sp. nov. and P. fryxelliana sp. nov. J. Phycol. 48, 436–454.

Merget, B., Koetschan, C., Hackl, T., Forster, F., Dandekar, T., Muller, T., Schultz, J.,Wolf, M., 2012. The ITS2 database. J. Vis. Exp. 61, e3806.

Miller, P.E., Scholin, C.A., 1996. Identification of cultured Pseudo-nitzschia (Bacillar-iophyceae) using species-specific LSU rRNA-targeted fluorescent probes. J.Phycol. 32, 646–655.

Montresor, M., Di Prisco, C., Sarno, D., Margiotta, F., Zingone, A., 2013. Diversity andgermination patterns of diatom resting stages at a coastal Mediterranean site.Mar. Ecol. Prog. Ser. 484, 79–95.

Moschandreou, K.K., Baxevanis, A.D., Katikou, P., Papaefthimiou, D., Nikolaidis, G.,Abatzopoulos, T.J., 2012. Inter- and intra-specific diversity of Pseudo-nitzschia(Bacillariophyceae) in the northeastern Mediterranean. Eur. J. Phycol. 47, 321–339.

Muller, T., Philippi, N., Dandekar, T., Schultz, J., Wolf, M., 2007. Distinguishingspecies. RNA 13, 1469–1472.

NOAA, 2013. Sea Surface Temperature (SST) Contour Charts. National Oceanograph-ic and Atmospheric Administration, Office of Satellite and Product Operations,National Environmental Satellite, Data, and Information Service Online hhttp://www.ospo.noaa.gov/Products/ocean/sst/contour/i (29 May 2014).

National Research Council (NRC), 1996. Stemming the Tide: Controlling Introductionsof Nonindigenous Species by Ships’ Ballast Water. Marine Board Committee onShips’ Ballast Operations, National Academy Press, Washington, DC, pp. 22.

Orlova, T.Y., Morozova, T.V., 2009. Resting stages of microalgae in recent marinesediments of Peter the Great Bay, Sea of Japan. Russian J. Mar. Biol. 35, 313–322.

Orsini, L., Sarno, D., Procaccini, G., Poletti, R., Dahlmann, J., Montresor, M., 2002.Toxic Pseudo-nitzschia multistriata (Bacillariophyceae) from the Gulf of Naples:morphology, toxin analysis and phylogenetic relationships with other Pseudo-nitzschia species. Eur. J. Phycol. 37, 247–257.

Peakall, R., Smouse, P.E., 2012. GenAlEx 6.5: genetic analysis in Excel. Populationgenetic software for teaching and research—an update. Bioinformatics 28,1537–1539.

Pritchard, J.K., Stephens, M., Donnelly, P., 2000. Inference of population structureusing multilocus genotype data. Genetics 155, 945–959.

Pughuic, D., 2001. Ballast water management and control: an overview. Trop. Coasts8, 42–49.

Quijano-Scheggia, S., Garces, E., Andree, K.B., de la Iglesia, P., Diogene, J., Fortuno,J.M., Camp, J., 2010. Pseudo-nitzschia species on the Catalan coast: characteri-zation and contribution to the current knowledge of the distribution of thisgenus in the Mediterranean Sea. Sci. Mar. 74, 395–410.

Rhodes, L., White, D., Syhre, M., Atkinson, M., 1996. Pseudo-nitzschia species isolatedfrom New Zealand coastal waters: domoic acid production in vitro and links withshellfish toxicity. In: Yasumoto, T.,Oshima, Y., Fukuyo, Y. (Eds.), Harmful and ToxicAlgal Blooms. Intergov. Oceanogr. Comm., UNESCO, Paris, pp. 155–158.

Roach, J., 2004. Ocean Conveyor Belt sustains sea life, study says. National Geo-graphic News. online. http://news.nationalgeographic.com/news/2004/06/0615_040614_SouthernOcean.html (29 May 2014).

Rosenberg, N.A., 2004. DISTRUCT: a program for the graphical display of populationstructure. Mol. Ecol. Notes 4, 137–138.

Sahraoui, I., Bates, S.S., Bouchouicha, D., Hadj Mabrouk, H., Sakka Hlaili, A., 2011.Toxicity of Pseudo-nitzschia populations from Bizerte Lagoon Tunisia, southwestMediterranean, and first report of domoic acid production by P. brasiliana.Diatom. Res. 26, 293–303.

Schultz, J., Wolf, M., 2009. ITS2 sequence-structure analysis in phylogenetics: ahow-to manual for molecular systematics. Mol. Phylogenet. Evol. 52, 520–523.

Schultz, J., Maisel, S., Gerlach, D., Muller, T., Wolf, M., 2005. A common core ofsecondary structure of the internal transcribed spacer 2 (ITS2) throughout theEukaryota. RNA 11, 361–364.

Seibel, P.N., Muller, T., Dandekar, T., Schultz, J., Wolf, M., 2006. 4SALE–A tool forsynchronous RNA sequence and secondary structure alignment and editing.BMC Bioinf. 7, 498.

Seibel, P.N., Muller, T., Dandekar, T., Matthias, W., 2008. Synchronous visual analysisand editing of RNA sequence and secondary structure alignments using 4SALE.BMC Res. Notes 1, 91.

Sorhannus, U., 2007. A nuclear-encoded small-subunit ribosomal RNA timescale fordiatom evolution. Mar. Micropaleontol. 65, 1–12.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:molecular evolutionary genetics analysis using maximum likelihood, evolu-tionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28,2731–2739.

Teng, S.T., Leaw, C.P., Lim, H.C., Lim, P.T., 2013. The genus Pseudo-nitzschia (Bacillar-iophyceae) in Malaysia, including new records and a key to species inferredfrom morphology-based phylogeny. Bot. Mar. 56, 375–398.

Trainer, V.L., Adams, N.G., Bill, B.D., Anulacion, B.F., Wekell, J.C., 1998. Concentrationand dispersal of a Pseudo-nitzschia bloom in Penn Cove, Washington, USA. Nat.Toxins 6, 113–126.

Trainer, V.L., Cochlan, W.P., Erickson, A., Bill, B.D., Cox, F.H., Borchert, J.A., Lefebvre,K.A., 2007. Recent domoic acid closures of shellfish harvest areas in WashingtonState inland waterways. Harmful Algae 6, 449–459.

H.C. Lim et al. / Harmful Algae 37 (2014) 142–152152

Trainer, V.L., Bates, S.S., Lundholm, N., Thessen, A.E., Cochlan, W.P., Adams, N.G.,Trick, C.G., 2012. Pseudo-nitzschia physiological ecology, phylogeny, toxicity,monitoring and impacts on ecosystem health. Harmful Algae 14, 271–300.

Tsukazaki, C., Ishii, K.I., Saito, R., Matsuno, K., Yamaguchi, A., Imai, I., 2013. Distri-bution of viable diatom resting stage cells in bottom sediments of the easternBering Sea shelf. Deep-Sea Res., II 94, 22–30.

Turrell, E., Bresnan, E., Collins, C., Brown, L., Graham, J., Grieve, M., 2008. Detection ofPseudo-nitzschia (Bacillariophyceae) species and amnesic shellfish toxins inScottish coastal waters using oligonucleotide probes and the Jellett RapidTestTM. Harmful Algae 7, 443–458.

Vale, P., Sampayo, M.A.M., 2001. Domoic acid in Portuguese shellfish and fish.Toxicon 39, 893–904.

Villac, M.C., Fryxell, G.A., 1998. Pseudo-nitzschia pungens var. cingulata var. nov.(Bacillariophyceae) based on field and culture observations. Phycologia 37,269–274.

Vrieling, E.G., Koeman, R.P.T., Scholin, C.A., Scheerman, P., Peperzak, L., Veenhuis, M.,Gieskes, W.W.C., 1996. Identification of a domoic acid-producing Pseudo-nitzschia species (Bacillariophyceae) in the Dutch Wadden Sea with electronmicroscopy and molecular probes. Eur. J. Phycol. 31, 333–340.

Wang, P., Liang, J., Lin, X., Chen, C., Huang, Y., Gao, Y., Gao, Y., 2012. Morphology,phylogeny and ITS-2 secondary structure of Pseudo-nitzschia brasiliana (Bacil-lariophyceae), including Chinese strains. Phycologia 51, 1–10.

Weir, B.S., Cockerham, C.C., 1984. Estimating F-statistics for the analysis of popula-tion structure. Evolution 38, 1358–1370.

White, T.J., Bruns, T., Lee, S., Taylor, J.W., 1990. Amplification and direct sequencingof fungal ribosomal RNA genes for phylogenetics. In: Innis, M.A., Gelfand, D.H.,Sninsky, J.J., White, T.J. (Eds.), PCR Protocols: A Guide to Methods and Applica-tions. Academic Press Inc., New York, NY, pp. 315–322.

Wolf, M., Achtziger, M., Jorg, S., Dandekar, T., Muller, T., 2005a. Homology modelingrevealed more than 20,000 rRNA internal transcribed spacer 2 (ITS2) secondarystructures. RNA 11, 1616–1623.

Wolf, M., Friedrich, J., Dandekar, T., Muller, T., 2005. CBCAnalyzer: inferring phy-logenies based on compensatory base changes in RNA secondary structures. InSilico Biol. 5, 291–294.

Wolf, M., Ruderisch, B., Dandekar, T., Schultz, J., Muller, T., 2008. ProfDistS: (profile-)distance based phylogeny on sequence-structure alignments. Bioinformatics24, 2401–2402.

Wolf, M., Chen, S., Song, J., Ankenbrand, M., Muller, T., 2013. Compensatory basechanges in ITS2 secondary structures correlate with the biological speciesconcept despite intragenomic variability in ITS2 sequences—a proof of concept.PLoS One 8, e66726.

Zhang, Y.Y., Lu, S.H., Zhang, C.S., Gao, Y., 2010. Distribution and germination ofviable diatom resting stage cells in sediments of the East China Sea. ActaOceanol. Sin. 29, 121–128.