16
DNA-barcoding: A case study in the diatom genus Frustulia (Bacillariophyceae) Pavla Urbánková* & Jana Veselá Department of Botany, Faculty of Science, Charles University in Prague, Benátská 2, 12801 Prague, Czech Republic * Corresponding author: [email protected] With 2 figures and 3 tables Abstract: Frustulia was used as a model genus to test four candidate markers (D1-D2 LSU rDNA, V4-SSU rDNA, rbcL-3 P, and COI-5 P) for DNA barcoding of diatoms. The Frustulia strains included in the study were primarily isolated from Europe and New Zealand. In agreement with previous reports, the results suggest that a dual-locus barcode comprising a partial sequence of the large ribosomal subunit (LSU) and a partial sequence of the large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase (rbcL) is the best option for DNA barcoding of diatoms. Unfortunately, this approach is limited by the lack of a distinct bar- code gap for these markers. Overall, the results of this study suggest that Frustulia species cannot currently be resolved without the use of additional non-molecular evidence. Effective use of DNA barcoding will require a better understanding of the mechanisms that generate and maintain genetic diversity in diatoms. Key words: diatom, Frustulia, DNA barcoding, D1-D2 LSU, rbcL-3 P, V4 SSU, COI-5 P Introduction DNA barcoding uses short DNA fragments (400600 bp) to diagnose species (Hebert et al. 2003). Considering the increasing availability of DNA sequencing, DNA barcoding is a promis- ing method of species identication for both basic and applied research. Although the feasibil- ity of DNA barcoding has been demonstrated in well-studied groups of organisms (Barrett & Hebert 2005, Ward et al. 2005, Baker et al. 2009), limited knowledge of the other groups of organisms currently hinders its use for species identication (Meyer & Paulay 2005, Meier et al. 2006, Trewick 2008). This problem is especially evident for many protist groups; despite their undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009, Stern et al. 2010). Only a small fraction of the diversity of diatoms, which may include up to 200,000 species (Mann & Droop 1996), has been described to date. In poorly characterized groups such as the diatoms, DNA barcoding may be used as a tool for dis- covery of yet undescribed diversity (Hebert et al. 2004, Mann et al. 2010). However, opposition to the use of DNA barcoding for species delineation has emerged; in general, its opponents warn against replacement of an integrative taxonomical approach with typological utilization of a Nova Hedwigia, Beiheft 142, 147–162 Article Stuttgart, September 2013 © 2013 J. Cramer in Gebr. Borntraeger Verlagsbuchhandlung, Stuttgart, Germany www.borntraeger-cramer.de 1438-9134/2013/0142-0147 $ 4.00 C

DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

  • Upload
    others

  • View
    14

  • Download
    0

Embed Size (px)

Citation preview

Page 1: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

DNA-barcoding: A case study in the diatom genus Frustulia (Bacillariophyceae)

Pavla Urbánková* & Jana Veselá

Department of Botany, Faculty of Science, Charles University in Prague, Benátská 2, 12801 Prague, Czech Republic

* Corresponding author: [email protected]

With 2 fi gures and 3 tables

Abstract: Frustulia was used as a model genus to test four candidate markers (D1-D2 LSU rDNA, V4-SSU rDNA, rbcL-3 P, and COI-5 P) for DNA barcoding of diatoms. The Frustulia strains included in the study were primarily isolated from Europe and New Zealand. In agreement with previous reports, the results suggest that a dual-locus barcode comprising a partial sequence of the large ribosomal subunit (LSU) and a partial sequence of the large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase (rbcL) is the best option for DNA barcoding of diatoms. Unfortunately, this approach is limited by the lack of a distinct bar-code gap for these markers. Overall, the results of this study suggest that Frustulia species cannot currently be resolved without the use of additional non-molecular evidence. Effective use of DNA barcoding will require a better understanding of the mechanisms that generate and maintain genetic diversity in diatoms.

Key words: diatom, Frustulia, DNA barcoding, D1-D2 LSU, rbcL-3 P, V4 SSU, COI-5 P

Introduction

DNA barcoding uses short DNA fragments (400�600 bp) to diagnose species (Hebert et al. 2003). Considering the increasing availability of DNA sequencing, DNA barcoding is a promis-ing method of species identifi cation for both basic and applied research. Although the feasibil-ity of DNA barcoding has been demonstrated in well-studied groups of organisms (Barrett & Hebert 2005, Ward et al. 2005, Baker et al. 2009), limited knowledge of the other groups of organisms currently hinders its use for species identifi cation (Meyer & Paulay 2005, Meier et al. 2006, Trewick 2008). This problem is especially evident for many protist groups; despite their undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009, Stern et al. 2010). Only a small fraction of the diversity of diatoms, which may include up to 200,000 species (Mann & Droop 1996), has been described to date. In poorly characterized groups such as the diatoms, DNA barcoding may be used as a tool for dis-covery of yet undescribed diversity (Hebert et al. 2004, Mann et al. 2010). However, opposition to the use of DNA barcoding for species delineation has emerged; in general, its opponents warn against replacement of an integrative taxonomical approach with typological utilization of a

Nova Hedwigia, Beiheft 142, 147–162 ArticleStuttgart, September 2013

© 2013 J. Cramer in Gebr. Borntraeger Verlagsbuchhandlung, Stuttgart, Germany www.borntraeger-cramer.de 1438-9134/2013/0142-0147 $ 4.00

C

F-NOVA142_S001-190.indb 147F-NOVA142_S001-190.indb 147 27.08.2013 10:31:3427.08.2013 10:31:34

Page 2: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

148 Pavla Urbánkova & Jana Veselá

single molecular marker (Will & Rubinoff 2004, DeSalle et al. 2005, Prendini 2005). At present, DNA barcoding is envisioned as a method to pinpoint potential new lineages that merit further exploration (Moniz & Kaczmarska 2009, Mann et al. 2010), rather than a panacea to the species concept dispute (Mann 1999, 2010).

Some advances towards effective DNA barcoding in diatoms have been already made. Sev-eral candidate barcode markers have been tested on datasets spanning the taxonomical distance from a single genus (Evans et al. 2007, Trobajo et al. 2011) to a whole class of diatoms (Moniz & Kaczmarska 2009, 2010, Pniewski et al. 2010, Hamsher et al. 2011, Zimmermann et al. 2011, Luddington et al. 2012), up to the distance across the entire range of algal or protist diver-sity (Sherwood & Presting 2007, Wylezich et al. 2010). Selection of a suitable barcode marker generally follows two criteria, sometimes referred to as discriminatory power and universality. Whereas discriminatory power evaluates the ability of the marker to separate closely related spe-cies, universality describes methodological issues such as the availability of universal primers, the quality of obtained sequences, and problems related to sequence alignment (Hamsher et al. 2011). Currently, the best performing barcode markers for diatoms are as follows: (i) the 3’ end of the large subunit of the rbcL (rbcL-3 P), (ii) a 540 bp fragment situated 417 bp downstream of the start codon of the rbcL (540 bp rbcL), (iii) the 5’ end of the mitochondrial cytochrome c oxidase I gene (COI-5 P), (iv) a partial sequence of the large ribosomal subunit (D1-D3 LSU, usually either D1-D2 or D2-D3), and (v) the V4 sub-region of the small ribosomal subunit (V4 SSU) (Evans et al. 2007, Moniz & Kaczmarska 2009, 2010, Hamsher et al. 2011, MacGillivary & Kaczmarska 2011, Trobajo et al. 2011, Zimmermann et al. 2011, Luddington et al. 2012). The 5.8 S gene combined with the second internal transcribed spacer is also a potential barcode marker for diatoms; however, even though this marker displayed suffi cient universality and good discrimination power when assessed using a dataset comprising a wide range of diatom taxa (MacGillivary & Kaczmarska 2011), it was rejected in several studies due to substantial intra-clonal diversity that hampered alignment of even closely related lineages (Behnke et al. 2004, Poulíþková et al. 2010, Trobajo et al. 2011). The universal plastid amplicon, the entire sequence of the small ribosomal subunit, and the entire sequence of the large subunit of rbcL are unsuit-able for DNA barcoding. Sequencing of the latter two markers requires several internal primers; therefore, these markers fail to satisfy the universality criterion. The universal plastid amplicon, which was suggested as a marker for all eukaryotic algae and cyanobacteria (Sherwood & Prest-ing 2007), has low discriminatory power; this marker achieved only 20 % success for diatom species identifi cation in a previous study (Hamsher et al. 2011).

In this study, we attempted to select markers for screening species diversity within the diatom genus Frustulia. The performance of four candidate barcode markers (D1-D2 LSU, V4 SSU, COI-5 P, and rbcL-3 P) was examined using selected European isolates that were previously used by Veselá et al. (2012), as well as new strains from more recently sampled locations in Europe, New Zealand, and sub-Antarctic Marion Island.

Materials and methods

The selected Frustulia strains previously used by Veselá et al. (2012) were included in this study. A number of new strains isolated from samples collected in New Zealand, Greenland, Marion Island, and several European locations were also included (Table 1). Isolation and cultivation of the strains was performed as described by Veselá et al. (2012). The frozen biomass and voucher materials have been stored in the Laboratory of Phycology, Department of Botany, Charles Uni-versity in Prague, Czech Republic.

The D1-D2 LSU sequences (552–687 bp) were used to provide an initial estimate of the ge-netic diversity of the Frustulia strains examined. Veselá et al. (2012) initially used primers D1 R

F-NOVA142_S001-190.indb 148F-NOVA142_S001-190.indb 148 27.08.2013 10:31:3427.08.2013 10:31:34

Page 3: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

149DNA-barcoding: A case study in the diatom genus Frustulia

and D2 C, which were suggested by Trobajo et al. (2009) to amplify D1-D2 LSU. However, due to their low specifi city, these primers often amplifi ed heterotrophic culture contaminants. Con-sequently, the D1 R and D2 C primers were only used by Veselá et al. (2012) to obtain partial sequences that aided the design of specifi c primers, namely LSU-DF1 and LSU-740DR. The wider applicability of these primers was tested in silico by analyzing the collection of diatom sequences available in the GenBank database. DNA amplifi cation and sequencing of the D1-D2 LSU marker was performed using the LSU-DF1 and LSU-740DR primers, as described previ-ously (Veselá et al. 2012). Raw sequences were edited using SeqAssem software (SequentiX, Klein Raden, Germany) and screened for ambiguous positions according to the method described by Beszteri et al. (2005). Proper codes were used in the alignment and ambiguous positions were deleted in individual pairwise sequence comparisons. Unique sequences were manually aligned and then analyzed using MEGA 4 software (Tamura et al. 2007). The fi nal alignment spanned 648 bp and contained two indels. Identifi cation of molecular operational taxonomic units (MO-TUs) was based on the similarity matrix of the uncorrected pairwise distances (p-distances) calculated using MEGA 5 software (Table 2). The p-distances within and between species were visualized using the neighbor-joining clustering algorithm. After examination by both light mi-croscopy and scanning electron microscopy, the MOTUs were identifi ed to the species level according to recent diatom monographs (Lange-Bertalot 2001, Beier & Lange-Bertalot 2007).

Strains used to evaluate the V4 SSU, COI-5 P, and rbcL-3 P molecular markers were selected on the basis of the D1-D2 LSU MOTUs. The intention of the selection method was to cover intra-MOTU sequence variability and include allopatric strains. This approach was designed to minimize the risk of overlooking cryptic lineages that are not diversifi ed in D1-D2 LSU rDNA but are potentially identifi able by the more divergent COI-5 P and rbcL-3 P markers (Amato et al. 2007, Hamsher et al. 2011).

The COI-5 P marker (632–697 bp) was amplifi ed and sequenced using DiamF3 (Hamsher et al. 2011) and KEdtmR (Evans et al. 2007) primers. The PCR profi le was identical to that used by Hamsher et al. (2011), with the exception that the initial denaturation was extended to 10 min, as required for use of AmpliTaq Gold DNA polymerase (Applied Biosystems, California, USA). The rbcL-3 P marker (763–865 bp) was amplifi ed and sequenced using cfD (Hamsher et al. 2011) and DPrbcL7 (Levialdi Ghiron 2006) primers. The PCR profi le was as follows: 95 °C initial denaturation for 10 min; 40 cycles of 94 °C for 1 min, 48 °C for 1 min, and 72° C for 2 min 30 s; and a fi nal extension of 72 °C for 10 min. The V4 SSU marker (397 bp) was selected in silico from the entire SSU sequence (1681–1739 bp); this region was amplifi ed using the 34 F primer (Thüs et al. 2011) and the ITS5-DR primer (CCT TGT TAC GAC TTC ACC TTC C), which was designed based on alignment of diatom ITS1 sequences available in the GenBank database. The 528 F, 536 R, 1055F, and 1055 R internal primers (Montresor et al. 2004) were also used for sequencing. The PCR profi le was: 94 °C initial denaturation for 10 min; 35 cycles of 95 °C for 1 min, 51 °C for 1 min, and 72 °C for 2 min; and a fi nal extension of 72 °C for 10 min. The SSU sequence of Frustulia vulgaris (Twaites) De Toni was obtained from GenBank (AM502038).

The sequences of the rbcL-3 P, COI-5 P, and V4 SSU markers were analyzed in the same way as the D1-D2 LSU marker; raw sequences were screened for ambiguities and aligned, p-distances and number of base-pair differences matrices were constructed (Tables 2 and 3), and then neighbor-joining phylograms were generated.

Results

A total of 14 MOTUs were identifi ed based on the variability of the D1-D2 LSU sequences across the Frustulia isolates studied (Tables 1 and 2, Fig. 1). According to their morphology, six of these MOTUs were recognized as previously described species: (1) F. aotearoa Beier &

F-NOVA142_S001-190.indb 149F-NOVA142_S001-190.indb 149 27.08.2013 10:31:3427.08.2013 10:31:34

Page 4: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

150 Pavla Urbánkova & Jana Veselá

Table 1. List of Frustulia strains, including the date and locality of collection. GenBank accession numbers for all tested markers are shown. Strains are grouped into D1-D2 LSU-based MOTUs that were named ac-cording to the morphotypes of strains. Information regarding ambiguous nucleotide positions in the D1-D2 LSU marker (number of ambiguities in individual sequences / total number of ambiguous positions in a particular MOTU) is shown in parentheses. A dash (–) indicates an unsuccessful amplifi cation; “na” indi-cates amplifi cations that were not attempted.

Strain Locality Date GeneBank accessions D1-D2

LSUV4 SSU rbcL-3P COI-5P

F. crassinervia lineage III (2–5/6) F237 Aquitaine (FR) 2010 HE601712 na HF562235 HF562259F350 Connemara (IE) 2010 HE601712 HF562290 HF562235 HF562258

F. crassinervia-saxonica lineage IV (2–3/3)F28 Abisko (SE) 2008 HE601713 na – – F219 Azores Islands (PT) 2009 HE601713 na – – F259 Aquitaine (FR) 2010 HE601714 HF562292 HF562236 –

F. crassinervia-saxonica lineage V (1–4/4)F45 Aquitaine (FR) 2009 HE601723 HF562291 HF562237 – F277 Kerry (IR) 2010 HE601721 na na – NZ14 South Island (NZ) 2010 HE601716 na – – BB2 Greenland (DK) 2011 HE601721 na HF562238 – 20-2F Bornholm (DK) 2011 HE601721 na HF562238 – 20-2E Bornholm (DK) 2011 HF562272 na – – 21-9B Bohemia (CZ) 2011 HF562272 na HF562239 HF562260

+ HE601715 – HE601724 (Veselá et al.2012)F. crassinervia-saxonica lineage VI (0–5/8)19-2B Bornholm (DK) 2011 HE601733 HF562294 HF562240 HF562255NZ136 South Island (NZ) 2010 HE601733 na HF562241 – NZ176 South Island (NZ) 2010 HF562273 na – – AD9 Aquitaine (FR) 2011 HE601733 na – – 22-4E Bohemia (CZ) 2011 HE601733 na partial HF56225724-2F Bohemia (CZ) 2011 HE601733 na HF562240 HF562257

+ HE601725 – HE601733 (Veselá et al. 2012)F. saxonica lineage VII (0-2/4)NZ17 South Island (NZ) 2010 HF562275 na na naNZ33 South Island (NZ) 2010 HF562276 HF562293 – HF562262NZ34 South Island (NZ) 2010 HF562276 na HF562242 – NZ39 South Island (NZ) 2010 HF562278 na HF562243 – NZ103 South Island (NZ) 2010 HF562278 na na naNZ139 South Island (NZ) 2010 HF562276 na na naNZ144 South Island (NZ) 2010 HF562277 na na naNZ169 South Island (NZ) 2010 HF562279 na na naNZ184 South Island (NZ) 2010 HF562274 na HF562254 –

F. erifuga (0)F197 Bohemia (CZ) 2009 HE601709 HF562295 HF562244 – F380 Beara (IE) 2010 HE601710 na HF562245 –

F-NOVA142_S001-190.indb 150F-NOVA142_S001-190.indb 150 27.08.2013 10:31:3427.08.2013 10:31:34

Page 5: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

151DNA-barcoding: A case study in the diatom genus Frustulia

Lange-Bertalot, (2) F. cassieae Lange-Bertalot & Beier, (3) F. erifuga Lange-Bertalot & Kram-mer, (4) F. gondwana Lange-Bertalot & Beier, (5) F. maoriana Lange-Bertalot & Beier, and (6) F. vulgaris (Twaites) De Toni. One of the MOTUs from New Zealand is very likely conspecifi c with New Zealand populations that were reported previously (Beier & Lange-Bertalot 2007, Kil-roy 2007). However, due to uncertainty about the conspecifi city of these populations with the ho-lotype population, it was treated as F. cf. magaliesmontana Cholnoky (for a detailed discussion see Beier & Lange-Bertalot 2007). Two of the identifi ed MOTUs were morphologically distinct from Frustulia species described so far; these lineages have been reported previously but have

Strain Locality Date GeneBank accessions D1-D2

LSUV4 SSU rbcL-3P COI-5P

F. gondwana (5/5)NZ22 South Island (NZ) 2010 HF562284 na HF562246 HF562263NZ24 South Island (NZ) 2010 HF562284 – – – NZ37 South Island (NZ) 2010 HF562284 HF562288 HF562246 HF562263

F. maoriana (2–4/4)NZ27 South Island (NZ) 2010 HF562268 na HF562247 HF562264NZ32 South Island (NZ) 2010 HF562270 na na naNZ35 South Island (NZ) 2010 HF562271 HF562296 HF562247 HF562264NZ36 South Island (NZ) 2010 HF562269 na na naNZ44 South Island (NZ) 2010 HF562271 na HF562247 HF562264

F. cf. magliesmontana (0)NZ159 South Island (NZ) 2010 HF562280 na HF562248 HF562265NZ162 South Island (NZ) 2010 HF562280 HF562287 HF562248 HF562265

F. aotearoa (0)NZ177 South Island (NZ) 2010 HF562281 HF562286 HF562249 HF562266NZ188 South Island (NZ) 2010 HF562281 na partial –

F. spA (0)NZ13 South Island (NZ) 2010 HF562282 – – –NZ30 South Island (NZ) 2010 HF562282 – HF562250 HF562267

F. spB (1/1)F378 Beara (IE) 2010 HE601711 HF562297 HF562251 HF562261F381 Beara (IE) 2010 HE601711 na HF562251 HF562261

F. cassieae (0)NZ190 South Island (NZ) 2010 HF562283 HF562289 HF562252 –

F. vulgaris (0)MIC10-15 Marion Island (ZA) 2012 HF562285 na HF562253 HF562256MIC10-38 Marion Island (ZA) 2012 HF562285 na HF562253 – AT-108Gel03 GenBank na na AM502038 na na

F-NOVA142_S001-190.indb 151F-NOVA142_S001-190.indb 151 27.08.2013 10:31:3427.08.2013 10:31:34

Page 6: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

152 Pavla Urbánkova & Jana VeseláTa

ble

2. P

airw

ise

com

paris

on o

f D1-

D2

LSU

(low

er le

ft) a

nd V

4-SS

U (u

pper

righ

t) se

quen

ces:

p-d

ista

nces

(num

ber o

f bas

e-pa

ir di

ffere

nces

). In

cong

ru-

ence

s with

D1-

D2

LSU

-bas

ed M

OTU

s are

out

lined

by

a so

lid li

ne.

D1-

D2

LSU

/ V4

SSU

lin

eage

IVlin

eage

Vlin

eage

VI

linea

ge V

IIlin

eage

III

F. m

aor.

F. sp

BF.

gon

d.F.

eri

f.F.

cf.

mag

.F.

aot

e.F.

spA

F. c

ass.

F. v

ulg.

F. c

rass

iner

via-

saxo

nica

lin

eage

VI

<

0.00

30.

000

0.00

00.

005

0.01

20.

012

0.01

70.

015

0.01

50.

029

0.02

2–

0.03

20.

036

(0

–2)

(0)

(0)

(2)

(5)

(5)

(7)

(6)

(6)

(12)

(9)

(13)

(15)

F. c

rass

iner

via-

saxo

nica

lin

eage

V 0

.007

–0.0

080.

000

0.00

00.

005

0.01

20.

012

0.01

70.

015

0.01

50.

029

0.02

2–

0.03

20.

036

(4

–5)

(0)

(0)

(2)

(5)

(5)

(7)

(6)

(6)

(12)

(9)

(13)

(15)

F. c

rass

iner

via-

saxo

nica

lin

eage

IV 0

.016

–0.0

200.

013–

0.01

5<

0.00

20.

005

0.01

20.

012

0.01

70.

015

0.01

50.

029

0.02

2–

0.03

20.

036

(1

0–12

)(8

–9)

(0–1

)(2

)(5

)(5

)(7

)(6

)(6

)(1

2)(9

)(1

3)(1

5)

F. sa

xoni

ca

linea

ge V

II 0

.021

–0.0

270.

015–

0.01

70.

020–

0.02

5<0

.003

0.00

70.

007

0.01

50.

012

0.01

20.

024

0.02

4–

0.02

70.

032

(1

3–16

) (9

–10)

(12–

15)

(0–2

)(3

)(3

)(6

)(5

)(5

)(1

0)(1

0)(1

1)(1

3)

F. c

rass

iner

via

linea

ge II

I0.

046–

0.05

40.

041–

0.04

70.

043–

0.05

20.

039–

0.04

70.

000

0.00

00.

012

0.01

0.01

0.02

20.

027

–0.

029

0.02

4(2

7–32

) (2

5–28

)(2

6–31

)(2

3–28

)(0

)(0

)(5

)(4

)(4

)(9

)(1

1)(1

2)(1

0)

F. m

aori

ana

0.0

31–0

.036

0.02

8–0.

029

0.03

5–0.

040

0.03

1–0.

037

0.03

4–0.

041

0.00

00.

012

0.01

0.01

0.02

20.

027

–0.

029

0.02

4

(18–

21)

(16–

17)

(20–

23)

(18–

21)

(20–

23)

(0)

(5)

(4)

(4)

(9)

(11)

(12)

(10)

F. sp

B0.

072–

0.07

70.

071–

0.07

30.

073–

0.07

70.

077–

0.08

20.

086–

0.09

30.

072–

0.07

60.

000

0.00

20.

002

0.01

90.

022

–0.

029

0.02

7(4

4–47

)(4

3–44

)(4

4–47

)(4

7–49

)(5

2–56

)(4

2–43

)(0

)(1

)(1

)(8

)(9

)(1

2)(1

1)

F. g

ondw

ana

0.0

96–0

.100

0.09

1–0.

094

0.09

3–0.

098

0.08

9–0.

091

0.09

3–0.

098

0.08

80.

067–

0.06

80.

000

0.00

00.

017

0.01

9–

0.02

70.

024

(5

5–58

)(5

3–54

)(5

4–56

)(5

2–53

)(5

4–56

) (5

0–51

)(3

9–40

)(0

)(0

)(7

)(8

)(1

1)(1

0)

F. e

rifu

ga0.

069–

0.07

20.

066–

0.06

80.

071–

0.07

70.

069–

0.07

30.

068–

0.07

30.

060–

0.06

30.

046–

0.04

70.

056–

0.05

80.

002

0.01

70.

019

–0.

027

0.02

4(4

2–44

) (4

1–42

)(4

3–47

)(4

2–44

)(4

1–44

)(3

5–36

)(2

8–29

)(3

3–34

)(1

)(7

)(8

)(1

1)(1

0)

F. c

f. m

agal

iesm

onta

na 0

.073

–0.0

780.

069

0.07

2–0.

075

0.06

3–0.

067

0.06

6–0.

072

0.06

2–0.

064

0.08

4–0.

086

0.09

50.

081

0.00

00.

022

–0.

015

0.01

2

(47–

48)

(42)

(44–

45)

(38–

40)

(40–

43)

(36)

(51–

52)

(55)

(49)

(0)

(9)

(6)

(5)

F. a

otea

roa

0.09

7–0.

100

0.09

40.

101–

0.10

50.

095–

0.09

80.

092–

0.09

80.

089–

0.09

00.

118–

0.11

90.

127

0.11

6–0.

118

0.08

60.

000

–0.

032

0.03

2(5

8–61

)(5

7)(6

2–63

)(5

8–59

)(5

9)(5

1–52

) (7

2–73

)(7

4) (7

1–72

)(5

2)(0

)(1

3)(1

3)

F. sp

A 0

.079

–0.0

820.

079

0.08

1–0.

085

0.07

0–0.

073

0.07

6–0.

082

0.07

1–0.

072

0.09

7–0.

098

0.11

30.

088–

0.09

00.

072

0.05

70.

000

––

(4

8–51

)(4

8)(4

9–51

)(4

3–44

)(4

6–49

) (4

1–42

)(5

9–60

)(6

6)(5

4–55

)(4

4)(3

5)(0

)

F. c

assi

eae

0.10

8–0.

114

0.10

8–0.

110

0.11

9–0.

123

0.10

8–0.

113

0.11

4–0.

117

0.09

7–0.

099

0.10

9–0.

110

0.14

0.11

2–0.

114

0.11

70.

113

0.10

3na

.0.

012

(57–

60)

(57–

58)

(62–

65)

(57–

60)

(60–

62)

(51–

52)

(58–

59)

(74)

60–

61(6

2)(6

0)(5

5)(5

)

F.vu

lgar

is 0

.105

–0.1

100.

102–

0.10

40.

105–

0.11

20.

098–

0.10

30.

105–

0.11

00.

092–

0.09

40.

111–

0.11

20.

129

0.10

7–0.

109

0.09

90.

107

0.09

60.

098

na.

(6

4–67

)(6

2–63

)(6

4–68

)(6

0–62

)(6

4–66

) (5

3–55

)(6

8–69

)(7

6) 6

6–67

(60)

(68)

(59)

(54)

F-NOVA142_S001-190.indb 152F-NOVA142_S001-190.indb 152 27.08.2013 10:31:3427.08.2013 10:31:34

Page 7: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

153DNA-barcoding: A case study in the diatom genus Frustulia

Tabl

e 3.

Pai

rwis

e co

mpa

rison

of C

OI-

5P (l

ower

left)

and

rbcL

-3P

(upp

er ri

ght)

sequ

ence

s: p

-dis

tanc

es (n

umbe

r of b

ase-

pair

diffe

renc

es).

Inco

ngru

ence

s w

ith D

1-D

2 LS

U-b

ased

MO

TUs a

re o

utlin

ed b

y a

solid

line

.

CO

I-5P

/ rbc

L-3P

lin

eage

VI

linea

ge V

linea

ge IV

linea

ge V

IIlin

eage

III

F. m

aor.

F. e

rif.

F. g

ond.

F. sp

BF.

aot

e.F.

cf.

mag

.F.

spA

F. c

ass.

F. v

ulg.

F. c

rass

iner

via-

saxo

nica

lin

eage

VI

< 0.

005

(0–5

)0.

001–

0.00

60.

017

0.02

4–0.

026

0.02

30.

024

0.04

8–0.

052

0.04

9–0.

053

0.04

9–0.

053

0.04

4–0.

050

0.06

8–0.

072

0.05

8–0.

062

0.07

2–0.

074

0.08

1<

0.00

5 (0

–3)

(1–5

)(1

3)(1

9–21

)(1

8)(1

9)(3

8–41

)(3

9–42

)(3

9–42

)(2

2–37

)(5

4–57

)(4

2–45

)(5

7–59

)(6

4)

F. c

rass

iner

via–

saxo

nica

lin

eage

V0.

003–

0.00

5<

0.00

4 (0

–3)

0.01

6–0.

017

0.02

3–0.

028

0.02

1–0.

025

0.02

1–0.

025

0.04

9–0.

053

0.05

2–0.

053

0.05

2–0.

053

0.04

5–0.

050

0.07

0–0.

072

0.06

0–0.

061

0.06

0–0.

061

0.07

9–0.

083

(2–5

)na

.(1

2–14

)(1

8–22

)(1

7–19

)(1

7–20

)(3

9–42

)(4

1–42

)(4

1–42

)(2

3–38

)(5

5–57

)(4

3–44

)(5

8)(6

3–66

)

F. c

rass

iner

via–

saxo

nica

lin

eage

IV–

–na

.0.

020–

0.02

20.

013–

0.01

80.

026

0.05

8–0.

059

0.05

60.

058

0.04

6–0.

053

0.06

6–0.

067

0.06

20.

076

0.08

4

–(1

5–17

)(1

0–14

)(2

0)(4

4–45

)(4

3)(4

4)(2

5–35

)(5

0–51

)(4

3)(5

8)(6

4)

F. sa

xoni

ca

lin

eage

VII

0.08

4–0.

085

0.08

8–

0.00

3 (2

)0.

024–

0.02

70.

025–

0.02

60.

050–

0.05

40.

050–

0.05

30.

050–

0.05

30.

050–

0.06

30.

072–

0.07

60.

062–

0.06

40.

079–

0.81

0.08

9–0.

092

(53–

54)

(56)

na.

(19–

21)

(20–

21)

(40–

43)

(40–

42)

(40–

42)

(28–

42)

(57–

60)

(45–

46)

(63–

64)

(71–

73)

F. c

rass

iner

via

lin

eage

III

0.13

9–0.

141

0.13

7–0.

138

–0.

152–

0.15

30.

000

(0)

0.02

1–0.

022

0.05

0–0.

052

0.05

00.

050–

0.05

10.

037–

0.04

60.

066–

0.06

70.

065

0.07

40.

079–

0.08

1(8

4–95

)(8

3–87

)(9

3–96

)0.

002

(1)

(17)

(39–

41)

(39–

40)

(39–

40)

(20–

31)

(51–

53)

(47)

(57–

59)

(61–

64)

F. m

aori

ana

0.15

0–0.

152

0.14

8–0.

149

–0.

152

0.11

7–0.

123

0.00

0 (0

)0.

052–

0.05

30.

052

0.05

20.

045–

0.05

00.

075–

0.07

60.

065

0.06

50.

077

(95–

96)

(94)

(96)

(71–

78)

0.00

0 (0

)(4

1–42

)(4

1)(4

1)(2

4–36

)(5

9–60

)(4

7)(6

0)(6

1)

F. e

rifu

ga–

––

––

–>0

.004

(3)

0.01

4–0.

015

0.01

5–0.

016

0.05

5–0.

074

0.08

2–0.

087

0.06

70.

073–

0.07

40.

079–

0.08

3

–(1

1–12

)(1

2–13

)(3

4–45

)(6

5–69

)(4

8)(5

8–59

)(6

3–66

)

F. g

ondw

ana

0.16

1–0.

163

0.16

0–

0.15

30.

149–

0.15

20.

137–

0.13

8–

0.00

0 (0

)0.

009

0.05

5–0.

067

0.08

5–0.

086

0.06

70.

074

0.07

9(1

02–1

03)

(101

)(9

7)(9

0–96

)(8

7)0.

000

(0)

(7)

(32–

44)

(67–

68)

(48)

(59)

(63)

F. sp

B0.

150–

0.15

50.

153–

0.15

5–

0.15

4–0.

155

0.16

3–0.

165

0.13

8–0.

144

–0.

087–

0.08

90.

000

(0)

0.05

7–0.

071

0.08

5–0.

086

0.06

50.

072–

0.07

30.

081

(93–

99)

(93–

98)

(94–

98)

(99–

104)

(84–

91)

(54–

55)

0.00

0 (0

)(3

3–46

)(6

7–68

)(4

7)(5

7–58

)(6

4)

F. a

otea

roa

0.19

0–0.

192

0.19

0–

0.17

90.

188–

0.19

10.

179

–0.

152

0.16

40.

000

(0)

0.06

2–0.

078

0.06

1–0.

077

0.05

9–0.

061

0.07

1–0.

076

(108

–109

)(1

16)

(109

)(1

14–1

16)

(109

)(8

7)(1

00)

na.

(37–

49)

(31–

44)

(29–

47)

(34–

60)

F. c

f. m

agal

iesm

onta

na0.

192–

0.19

60.

191–

0.19

3–

0.18

5–0.

188

0.16

9–0.

171

0.17

5–0.

180

–0.

150–

0.15

20.

167–

0.16

80.

162–

0.16

30.

000

(0)

0.07

80.

083

0.09

2–0.

093

(116

–124

)(1

16–1

22)

(112

–119

)(1

02–1

08)

(106

–114

)(9

1–96

)(1

02–1

06)

(98–

99)

0.00

0 (0

)(5

5)(6

6)(7

3–74

)

F. sp

A0.

177–

0.17

80.

182

–0.

169

0.17

9–0.

180

0.17

9–

0.14

40.

157–

0.15

80.

164

0.16

2–0.

163

na.

0.07

10.

089

(108

–109

)(1

11)

(103

)(1

09)

(101

)(8

8)(9

6)(1

00)

(88–

89)

na.

(51)

(64)

F. c

assi

eae

––

––

––

––

––

––

na.

0.07

7–0.

089

–(6

1–64

)

0.21

3–0.

218

0.21

5–

0.21

30.

225–

0.22

60.

198

–0.

180

0.21

00.

178

0.19

60.

210

–0.

000

(0)

(129

–132

)(1

30)

(129

)(1

36–1

37)

(120

)(1

09)

(127

)(1

08)

(119

)(1

27)

na.

F-NOVA142_S001-190.indb 153F-NOVA142_S001-190.indb 153 27.08.2013 10:31:3427.08.2013 10:31:34

Page 8: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

154 Pavla Urbánkova & Jana Veselá

not yet been offi cially described. Therefore, they were named according to the original reports as F. spA (Kilroy 2007) and F. spB (Urbánková 2011). Species F. spB previously appeared in Veselá et al. 2012 – as F. rhomboides var. crassinervia lineage II. The remaining fi ve lineages (herein labeled III�VII) had an uncertain taxonomical ranking within the complex of pseudocryptic lineages that comprises nominal species F. crassinervia (Brebisson) Lange-Bertalot & Krammer and F. saxonica Rabenhorst. Lineages III�VI have recently been examined using a multidisci-plinary approach (Veselá et al. 2012, therein treated as F. rhomboides (Ehrenberg) De Toni sensu lato). Lineage VII from New Zealand fell within this cluster according to both morphology and molecular phylogeny.

The LSU-DF1 and LSU-740DR primers successfully amplifi ed D1-D2 LSU in almost all strains examined; failure of these primers to amplify D1-D2 LSU in a few strains was most likely caused by low DNA concentration. Two problems were encountered during analysis of the D1-D2 LSU sequencing data. First, the presence of intra-clonal polymorphism in the D1-D2 LSU region (up to fi ve ambiguous positions per sequence; Table 1) meant that careful sequence edit-ing was necessary. Second, unambiguous nucleotide homologization within a 50 bp stretch of the alignment was complicated by the presence of an indel and adjacent variable positions in the loop of stem C1 (Poulíþková et al. 2010). However, since the latter problem did not complicate resolution among closely related lineages, these positions were retained in the alignment.

The V4 SSU region was selected in silico from the entire SSU sequence. Amplifi cation of the entire SSU sequence was successful for all strains, except those belonging to the F. spA lineage. Considering that we were able to localize exact sequences of the primers proposed by Zimmer-mann et al. (2011), amplifi cation of V4 SSU in the genus Frustulia should not present any major obstacles.

The COI-5 P sequence was amplifi ed using the Diam-F3 and KEdtmR primers; these prim-ers displayed the highest universality (35 %) in a previous study (Hamsher et al. 2011). In the current study, universality was 49 % (19 of 39 strains) among all taxa and reached 70 % (14 of 20 strains) following elimination of the 19 sequences of strains belonging to the closely related semicryptic lineages IV�VII, for which amplifi cation was mostly unsuccessful (Table 1). Failure of the Diam-F3 and KEdtmR primers to amplify COI-5 P in closely related lineages may have been caused by a common mutation in one of the primer-binding sites. In addition, these primers often amplifi ed culture contaminants, which decreased the quality of the sequences obtained.

The cfD+DPrbcL7 primers were used to amplify rbcL-3 P. Although Hamsher et al. (2011) reported 92 % amplifi cation success with this primer pair, only 77 % (30 of 39) of the strains included in this study were successfully amplifi ed. The lower success rate achieved in this study was caused by poor performance of the primers in lineages IV�VII (similar to COI-5 P). Exclu-sion of these lineages increased the amplifi cation success to 95 % (19 of 20 strains examined) (Table 1).

The smallest distance between two MOTUs in the D1-D2 LSU dataset was p = 0.007 dif-ferences per site (diff./site). This distance was greater than the divergence within MOTUs (p � 0.003 diff./site; Table 2) and exceeded the highest divergence encountered between sexu-ally compatible strains, which was p = 0.005 diff./site in Pseudo-nitzschia pseudodelicatissima (Hasle) Hasle (Amato et al. 2007; comparison of GenBank entries AY550126.1 and AY550127.1 trimmed to fi t our dataset) and p = 0.004 diff./site in Nitzschia palea (Kützing) W. Smith (Tro-bajo et al. 2009). The remaining Frustulia MOTUs were separated by p-distances ranging from p = 0.013 to p = 0.129 diff./site.

Divergence of the entire SSU rDNA region is low (Evans et al. 2007, Moniz & Kaczmarska 2009); therefore, V4 SSU sequences were obtained for only a single representative D1-D2 LSU-based MOTU. The V4 SSU marker failed to separate semicryptic lineages IV�VI. In addition, two other sets of morphologically and geographically distinct MOTUs, namely F. gondwana vs. F. erifuga, and F. maoriana vs. F. crassinervia lineage III, could not be separated (Fig. 1). The

F-NOVA142_S001-190.indb 154F-NOVA142_S001-190.indb 154 27.08.2013 10:31:3527.08.2013 10:31:35

Page 9: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

155DNA-barcoding: A case study in the diatom genus Frustulia

Fig. 1. Unrooted neighbor-joining tree for the D1-D2 LSU (left) and V4-SSU (right) markers, based on p-distances.

F-NOVA142_S001-190.indb 155F-NOVA142_S001-190.indb 155 27.08.2013 10:31:3527.08.2013 10:31:35

Page 10: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

156 Pavla Urbánkova & Jana Veselá

p-differences between the remaining MOTUs in the V4 SSU dataset ranged from p = 0.003 to p = 0.036 diff./site (Table 2).

Divergence of COI-5 P was the highest of all markers examined. With the exception of lin-eages V and VI, the p-distances among amplifi ed MOTUs ranged from p = 0.08 to p = 0.196 diff./site (Table 3). These distances exceeded the minimal inter-specifi c distance recorded for Sellaphora species (p = 0.012 diff./site; Hamsher et al. 2011). The divergence among strains belonging to lineages V and VI (p � 0.008 diff./site) was similar to the highest intra-specifi c divergences recorded for diatoms (Sellaphora: p = 0.008 diff./site, Evans et al. 2007; Nitzschia: p = 0.01 diff./site, Trobajo et al. 2011). In addition, marker COI-5 P did not separate strains from lineages V and VI according to the D1-D2 LSU-based MOTUs (Fig. 2).

Fig. 2. Unrooted neighbor-joining tree for the rbcL-3 P (left) and COI-5 P (right) markers, based on p-distances. An asterisk indicates partial sequences. D1-D2 LSU-based MOTUs that were not resolved by these markers are outlined by a dashed line

F-NOVA142_S001-190.indb 156F-NOVA142_S001-190.indb 156 27.08.2013 10:31:3527.08.2013 10:31:35

Page 11: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

157DNA-barcoding: A case study in the diatom genus Frustulia

Sequence divergence of the rbcL-3 P marker ranged from p = 0.009 to 0.093 diff./site. This divergence was distinctly lower than that of the COI-5 P marker and very similar (or slightly lower) to that of the D1-D2 LSU marker (Table 3). Similar to the results obtained for COI-5 P, the rbcL-3 P marker failed to discriminate lineages V and VI, which were recognized in LSU D1-D2 phylogeny as separate lineages (Fig. 2). There is no barcode gap for rbcL-3 P (Hamsher et al. 2011); therefore, the divergence between strains from lineages V and VI (p = 0.006 diff./site) may represent intra-specifi c variability or the presence of several, potentially cryptic, species. The relatively high divergences within lineage VII (p = 0.005 diff./site) and F. erifuga (p = 0.004 diff./site) are similarly problematic.

Discussion

In this study, Frustulia was used as a model genus to examine the suitability of selected candi-date markers for DNA barcoding of diatoms. Similar to previous reports, none of the potential markers included in this study displayed 100 % universality and discriminatory power. Although the amplifi cation success of the COI-5 P marker is relatively high when examining closely re-lated lineages (Evans et al. 2007, Trobajo et al. 2011), it drops dramatically when the sampling is enhanced to cover the phylogeny of several genera (Moniz & Kaczmarska 2009, Hamsher et al. 2011). In the current study, amplifi cation of this marker from F. crassinervia-saxonica, which is geographically widespread and is the most abundant lineage of the whole genus, was often unsuccessful. Therefore, despite its high discriminatory power, the COI-5 P marker is unsuitable for DNA barcoding, unless we fi nd better primers. The opposite problem was encountered for the V4 SSU marker. Although this study and previous studies (Zimmermann et al. 2011, Luddington et al. 2012) demonstrated successful amplifi cation of this marker from a wide range of diatoms, its discriminatory power was low; specifi cally, the V4 SSU marker failed to separate Frustulia morphospecies, which were otherwise clearly distinguished by the remaining markers. A similar problem has previously been reported for morphologically distinct species from several other genera (Zimmermann et al. 2011, Luddington et al. 2012). The V4 SSU marker seems to be less sensitive than light microscope observations and is therefore unsuitable for exploration of spe-cies diversity within the diatoms.

The remaining two markers included in this study, namely rbcL-3 P and D1-D2 LSU, displayed nearly similar variability across the genus Frustulia. Moreover, the primers used to amplify these markers displayed adequate universality. Primers for the rbcL-3 P region were designed or tested using taxa representing vast diatom diversity (Jones et al. 2005, Levialdi Ghiron 2006, Hamsher et al. 2011). The universality of the primers used for amplifi cation and sequencing of D1-D2 LSU were tested in silico on a dataset that also represented wide diatom diversity. Furthermore, additional primers for this region, or the overlapping D2-D3 region, have been reported in sev-eral previous studies (Trobajo et al. 2009, Wylezich et al. 2010, Hamsher et al. 2011). Use of D1-D2 LSU as a marker had only minor drawbacks. The fi rst drawback was the presence of an indel and adjacent highly variable nucleotides in the D2 domain; these features hampered the alignment of a 50 bp stretch of the sequence from distantly related Frustulia species. While this problem did not affect resolution between closely related taxa and is not worsened by extended sampling (Hamsher et al. 2011), it should not complicate the use of partial D1-D3 LSU rDNA sequences such as the D1-D2 (this study and Trobajo et al. 2011) or D2-D3 (Hamsher et al. 2011) regions. The second drawback related to D1-D2 LSU is the relatively high intra-clonal variabil-ity of some Frustulia strains (1�5 ambiguous positions; Table 1), which forced us to screen all sequences visually. Unfortunately, omission of this time-demanding step is problematic because it may impede the resolution of closely related lineages and detection of hybrids (Sonnenberg et al. 2007). The presence of intra-clonal variability and the differing levels of variability among

F-NOVA142_S001-190.indb 157F-NOVA142_S001-190.indb 157 27.08.2013 10:31:3527.08.2013 10:31:35

Page 12: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

158 Pavla Urbánkova & Jana Veselá

lineages/species are noteworthy. Processes responsible for concerted evolution of numerous rDNA cistrons (e.g. intra-genomic homogenization, mutation rate, genetic drift, and natural se-lection) depend on species-specifi c characteristics such as the frequency of sexual reproduction, population size and structure (Alverson 2008, VanKuren et al. 2012). Further investigation of the variable intra-clonal diversity between the closely related lineages in this study may enable a better understanding of the attributes of sister species and the speciation of diatoms.

The choice between the use of the rbcL-3 P and D1-D3 LSU markers is complicated by the fact that neither has a distinct barcode gap. A single base-pair divergence in rbcL-3 P may signify either intra- or inter-specifi c divergence (Hamsher et al. 2011). Moreover, introgression of rbcL and failure of D1-D3 LSU to separate two sexually incompatible species has been reported (Am-ato et al. 2007). These problems may be partly overcome by the joint use of rbcL-3 P and D1-D3 LSU in a dual-locus barcode (Trobajo et al. 2009, Mann et al. 2010, Hamsher et al. 2011). The use of this or an alternative dual-locus barcode (MacGillivary & Kaczmarska 2011) will increase the robustness of conclusions drawn from DNA barcoding studies by combining information from different parts of the organismal genome. However, evidence from the selected markers may be confl icting and assignment of the taxonomical status may therefore be problematic, es-pecially for closely related lineages (Alverson 2008, Degnan & Rosenberg 2009, McManus & Katz 2009, Bittner et al. 2010).

One example of confl icting evidence from multiple barcode markers occurred in the current study; specifi cally, separation of lineages V and VI by D1-D2 LSU was not supported by either the rbcL-3 P or COI-5 P markers. However, biological relevance of this separation has been shown previously (Veselá et al. 2012). The varying abundances of lineage V and VI at numerous localities across Europe suggest that they have different ecological preferences with respect to pH, conductivity, and/or habitat type (Veselá et al. 2012). In addition, even though considerable overlap in morphology complicates their unambiguous identifi cation, geometric morphometric analyses proved that lineages V and VI are signifi cantly different. The remaining pseudocryptic F. crassinervia-saxonica lineages were separated by at least one additional individual marker. Lineage III was separated by V4 SSU, COI-5 P, and rbcL-3 P; lineage IV was separated by rbcL-3 P; and lineage VII was separated by V4 SSU, COI-5 P, and rbcL-3 P. Unfortunately, the rare occurrence, scattered distribution, and/or limited sampling of lineages III, IV, and VII did not enable inference of their ecological preferences. Additional evidence to support diversifi ca-tion currently exists for lineage VII only. Despite extensive sampling in Europe, this lineage has only been isolated from New Zealand. As it regularly occurred at high abundance together with lineage VI, the absence of lineage VII in Europe cannot be explained by general rarity or lack of suitable habitats (Vyverman et al. 1998, Sabbe et al. 2001). Moreover, considering the exceptionally high rate of diatom endemicity in Australasia (Vyverman et al. 1998, Sabbe et al. 2001, Beier & Lange-Bertalot 2007), it is likely that the distribution of lineage VII is geographi-cally restricted to either this region or New Zealand alone. In spite of this evidence, we decided not to elevate any of the pseudocryptic lineages to a specifi c status because their morphologies overlapped with each other and/or with lineages V and VI (Veselá et al. 2012, unpublished mor-phometric data for lineage VII) and also spanned at least two nominal species (F. crassinervia and F. saxonica). The observed incongruence among tested markers could have resulted from hy-bridization of lineages that are different ecotypes or geographically separated populations of one species (potentially an incipient species if the gene fl ow is suffi ciently reduced by incomplete sexual barriers, ecological selection, or geographic distance). Alternatively, the phylogenetic pat-terns could have resulted from sexually incompatible species that retain ancestral polymorphism and similar morphology. Further independent evidence, such as microsatellite markers or data from mating or ecophysiological experiments, are required to determine whether they are worthy of individual taxonomic recognition. The remaining Frustulia lineages, which were separated by both D1-D2 LSU and rbcL-3 P, were morphologically distinct and did not display pseudocryptic

F-NOVA142_S001-190.indb 158F-NOVA142_S001-190.indb 158 27.08.2013 10:31:3527.08.2013 10:31:35

Page 13: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

159DNA-barcoding: A case study in the diatom genus Frustulia

diversity or conspecifi city among morphospecies. However, this result may change if geographi-cally or ecologically extended sampling recovers their closer relatives.

Based on the results of this study, we concur with use of a dual-locus barcode comprising partial rDNA sequences, such as the D1-D2 LSU region (Trobajo et al. 2011) or the D2-D3 LSU region (Hamsher et al. 2011), and either rbcL-3 P (Hamsher et al. 2011) or the 540-bp rbcL frag-ment (MacGillivary & Kaczmarska 2012). Notably, the diffi culties associated with assigning taxonomical status to closely related lineages belonging to the species complex F. crassinervia-saxonica that were encountered in this study highlight the dynamic nature of diatom species. This dynamicity causes ongoing controversy regarding the species concept (Mann 1999, 2010). Mann (2010) suggested acquisition of the “Unifi ed Species Concept” sensu de Queiroz (2007). According to this concept, diatom species are viewed as separately evolving lineages that gradu-ally acquire specifi c attributes such as distinctive morphology, ecology, or reciprocal monophyly (de Queiroz 2007). Only reciprocal monophyly is used for identifi cation and delineation of spe-cies by barcoding techniques; therefore, the “species problem” cannot be solved by DNA bar-coding alone. However, better insights into the mechanisms that generate and maintain genetic variability may simplify and improve the decision making process. In particular, research that addresses population genetics, biogeography, or ecological differences among sister lineages may contribute to our understanding of the speciation mechanisms in diatoms and other protists (DeSalle et al. 2005, Trewick 2008, Baker et al. 2009, Monaghan et al. 2009). The knowledge gained from these studies could be incorporated into algorithms for the screening of diatom diversity. These algorithms could replace the commonly used threshold approach to species de-lineation, which often fails because it ignores the fact that the distribution of intra-specifi c and inter-specifi c variability among species is overlapping rather than discrete (Will & Rubinoff 2004, Meyer & Paulay 2005, Meier et al. 2006, Dexter et al. 2010). As a result, threshold ap-proach overlooks a considerable number (up to 30 %) of distinct, but less divergent, taxa (Meyer & Paulay 2005, Meier et al. 2006). Algorithms that enable more natural species delineation are already being devised for macroorganisms (Little & Stevenson 2007, Monaghan et al. 2009, Ca-siraghi et al. 2010); however, effective use of these algorithms in diatoms requires an enhanced appreciation of the differences in speciation between macro- and microorganisms, as well as its peculiarities in this class.

Acknowledgements

We would like to thank Dr. Pieter Vanormelingen for providing us with strains of F. vulgaris, Dr. Paul Broady and members of his lab for enabling sampling in New Zealand, and to our col-leagues from the phycological research group for their help with sampling. Funding for this work was provided by the Czech Science Foundation project no. P506-13-07210 P.

References

Amato, A., Kooistra, W. H. C. F., Ghiron, J. H. L., Mann, D. G., Pröschold, T. & Montresor, M. (2007): Re-productive isolation among sympatric cryptic species in marine diatoms. – Protist 158: 193–207.

Alverson, A. J. (2008): Molecular systematics and the diatom species. – Protist 159: 339–353.Baker, A. J., Tavarese, S. & Elbourne, R.F. (2009): Countering criticisms of single mitochondrial DNA gene

barcoding in birds. – Mol. Ecol. Resour. 9: 257–268.Barrett, R. D. H. & Hebert, P. D. N (2005): Identifying spiders through DNA barcodes. – Can. J. Zool. 83:

481–491.Behnke, A., Friedl, T., Chepurnov, V. A. & Mann, D. G. (2004): Reproductive compatibility and rDNA

analyses in the Sellaphora pupula species complex (Bacillariophyta). – J. Phycol. 40: 193–208.

F-NOVA142_S001-190.indb 159F-NOVA142_S001-190.indb 159 27.08.2013 10:31:3527.08.2013 10:31:35

Page 14: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

160 Pavla Urbánkova & Jana Veselá

Beier, T. & Lange-Bertalot, H. (2007): A synopsis of cosmopolitan, rare and new Frustulia species (Bacil-lariophyceae) from ombotrophic peat bogs and minerotrophic swamps in New Zealand. – Nova Hed-wigia 85: 73–91.

Beszteri, B., Ács, E. & Medlin, L. K. (2005): Ribosomal DNA sequence variation among sympatric strains of the Cyclotella meneghiniana complex (Bacillariophyceae) reveals cryptic diversity. – Protist 156: 317–333.

Bittner, L., Halary, S., Payri, C., Cruaud, C., de Reviers, B., Lopez, P. & Bapteste, E. (2010): Some con-siderations for analyzing biodiversity using integrative metagenomics and gene networks. – Biol. Direct 5: 47.

Casiraghi, M., Labra, M., Ferri, E., Galimberti, A. & De Mattia, F. (2010): DNA barcoding: a six-question tour to improve users’ awareness about the method. – Briefi. Bioinform. 11: 440–453.

Degnan, J. H. & Rosenberg, N. A. (2009): Gene tree discordance, phylogenetic inference and the multispe-cies coalescent. – Trends Ecol. Evol. 24: 332–340.

de Salle, R., Egan, M. G. & Siddall, M. (2005): The unholy trinity: taxonomy, species delimitation and DNA barcoding. – Phil. Trans. R. Soc. London B Biol. Sci. 360: 1905–1916.

de Queiroz, K. (2007): Species concepts and species delimitation. – Syst. Biol. 56: 879–886.Dexter, K. G., Pennington, T. D. & Cunningham, C. W. (2010): Using DNA to assess errors in tropical tree

identifications: How often are ecologists wrong and when does it matter? – Ecol. Monogr. 80: 267–286.Evans, K. M., Wortley, A. H. & Mann, D. G. (2007): An assessment of potential diatom „barcode“ genes

(cox1, rbcL, 18 S and ITS rDNA) and their effectiveness in determining relationships in Sellaphora (Bacillariophyta). – Protist 158: 349–364.

Hamsher, S. E., Evans, K. M., Mann, D. G., Poulíþková, A. & Saunders, G. W. (2011): Barcoding diatoms: exploring alternatives to COI-5 P. – Protist 162: 405–422.

Hebert, P. D. N., Cywinska, A., Ball, S. L. & Dewaard, J. R. (2003): Biological identifications through DNA barcodes. – Proc. R. Soc. London B Biol. Sci. 270: 313–321.

Hebert, P. D. N., Stoeckle, M. Y., Zemlak, T. S. & Francis, C. M. (2004): Identifi cation of birds through DNA barcodes. – PLoS Biol. 2: 1657–1663.

Jones, H. M., Simpson, G. E., Stickle, A. J. & Mann, D. G. (2005): Life history and systematics of Petroneis (Bacillariophyta), with special reference to British waters. – Eur. J. Phycol. 40: 61–87.

Kilroy, C. B. (2007): Diatom communities in New Zealand subalpine mire pools: distribution, ecology and taxonomy of endemic and cosmopolitan taxa. PhD-thesis.

Luddington, I. A., Kaczmarska, I. & Lovejoy, C. (2012): Distance and Character-Based Evaluation of the V4 Region of the 18 S rRNA Gene for the Identifi cation of Diatoms (Bacillariophyceae). – PLoS ONE 7: e45664.

Lange-Bertalot, H. (2001): Diatoms of Europe. Vol. 2: Navicula sensu stricto, 10 genera separated from Navicula sensu lato, Frustulia. 526 pp. Gantner Verlag.

Levialdi-Ghiron, J. H. (2006): Plastid phylogeny and chloroplast inheritance in the planktonic pennate dia-tom Pseudo-nitzschia (Bacillariophyceae). Doctoral thesis, Universita Degli Studi Di Messina.

Little, D. P. & Stevenson, D. W. m. (2007): A comparison of algorithms for the identifi cation of specimens using DNA barcodes: examples from gymnosperms. – Cladistics 23:1–21.

Mann, D. G & Droop, S. J. M. (1996): Biodiversity, biogeography and conservation of diatoms. – Hydro-biologia 336: 19–32.

Mann, D. G. (1999): The species concept in diatoms (Phycological Reviews 18). – Phycologia 38: 437–95.Mann, D. G. (2010): Discovering diatom species: is a long history of disagreements about species-level

taxonomy now at an end? – Plant Ecol. Evol. 143: 251–264.Mann, D. G., Sato, S., Trobajo, R., Vanormelingen, P. & Souffreau, C. (2010): DNA barcoding for species

identifi cation and discovery in diatoms. – Cryptogamie Algol. 31: 557–577.MacGillivary, M. L. & Kaczmarska, I. (2011): Survey of the effi cacy of a short fragment of the rbcL gene

as a supplemental DNA barcode for diatoms. – J. Eukaryot. Microbiol. 58: 529–536.McManus, G. B. & Katz, L. A. (2009): Molecular and morphological methods for identifying plankton:

what makes a successful marriage? – J. Plankton Res. 31: 1119–1129.Meier, R., Shiyang, K., Vaidya, G. & Peter, K. L. N. (2006): DNA barcoding and taxonomy in Diptera: a tale

of high intraspecifi c variability and low identifi cation success. – Syst. Biol. 55: 715–728.Meyer, C. P. & Paulay, G. (2005): DNA barcoding: error rates based on comprehensive sampling. – PLoS

Biol. 3: 2229–2238.

F-NOVA142_S001-190.indb 160F-NOVA142_S001-190.indb 160 27.08.2013 10:31:3527.08.2013 10:31:35

Page 15: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

161DNA-barcoding: A case study in the diatom genus Frustulia

Monaghan, M. T., Wild, R., Elliot, M., Fujisawa, T., Balke, M., Inward, D. J., Lees, D. C., Ranaivosolo, R., Eggleton, P., Barraclough, T. G. & Vogler, A. P. (2009): Accelerated species inventory on Madagascar using coalescent-based models of species delineation. – Syst Biol. 58: 298–311.

Moniz, M. B. J. & Kaczmarska, I. (2009): Barcoding diatoms: Is there a good marker? – Mol. Ecol. Resour. 9: 65–74.

Moniz, M. B. J. & Kaczmarska, I. (2010): Barcoding of diatoms: Nuclear encoded ITS revisited. – Protist 161: 7–34.

Montresor, M., John, U., Beran, A. & Medlin, L. K. (2004): Alexandrium tamutum sp. nov. (Dinophyceae): a new nontoxic species in the genus Alexandrium. – J. Phycol. 40: 398–411.

Pniewski, F. F., Friedl, T. & Lataáa, A. (2010): Identifi cation of diatom isolates from the Gulf of GdaĔsk: testing of species identifi cations using morphology, 18 S rDNA sequencing and DNA barcodes of strains from the Culture Collection of Baltic Algae (CCBA). – Oceanol. Hydrobiol. St. 34: 3–20.

Poulíþková, A., Veselá, J., Neustupa, J. & Škaloud, P. (2010): Pseudocryptic diversity versus cosmopolitan-ism in diatoms: a case study on Naviculla crytpocephala Kütz. (Bacillariophyceae) and morphologically similar taxa. – Protist 161: 353–368.

Prendini, L. (2005): Comment on “Identifying spiders through DNA barcodes.” – Can. J. Zool. 83: 498–504.Sabbe, K., Vanhoutte, K., Lowe, R. L., Bergey, E. A., Biggs, B. J. F., Francoeur, S., Hodgson, D. & Vyver-

man, W. (2001): Six new Actinella (Bacillariophyta) species from Papua New Guinea, Australia and New Zealand: further evidence for widespread diatom endemism in the Australasian region. – Eur. J. Phycol. 36: 321–340.

Sherwood, A. R. & Presting, G. G. (2007): Universal primers amplify a 23 S rDNA plastid marker in eu-karyotic algae and Cyanobacteria. – J. Phycol. 43: 605–608.

Sonnenberg, R., Nolte, A. W. & Tautz, D. (2007): An evaluation of LSU rDNA D1–D2 sequences for their use in species identifi cation. – Front. Zool. 4: 6.

Stern, R. F., Horak, A., Andrew, R. L., Coffroth, M.-A., Andersen, R. A., Küpper, F. C., Jameson, I., Hop-penrath, M., Véron, B., Kasai, F., Brand, J., James, E. R. & Keeling, P. (2010): Environmental Barcoding Reveals Massive Dinofl agellate Diversity in Marine Environments. – PLoS ONE 5: e131991.

Tamura, K., Dudley, J., Nei, M. & Kumara, S. (2007): MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. – Mol. Biol. Evol. 24: 1596–1599.

Thüs, H., Muggia, L., Pérez-Ortega, S., Favero-Longo, S. E., Joneson, S., O’Brien, H., Nelsen, M. P., Duque-Thüs, R., Grube, M., Friedl, T., Brodie, J., Andrew, C. J., Lücking, R., Lutzoni, F. & Gueidan, C. (2011): Revisiting photobiont diversity in the lichen family Verrucariaceae (Ascomycota). – Eur. J. Phycol. 46: 399–415.

Trewick, S. A. (2008): DNA Barcoding is not enough: mismatch of taxonomy and genealogy in New Zea-land grasshoppers (Orthoptera: Acrididae). – Cladistics 24: 240–254.

Trobajo, R., Clavero, E., Chepurnov, V. A., Sabbe, K., Mann, D. G., Ishihara, S. & Cox, E. J. (2009): Mor-phological, genetic and mating diversity within the widespread bioindicator Nitzschia palea (Bacillari-ophyceae). – Phycologia 48: 443–459.

Trobajo, R., Clavero, E., Evans, K. M., Vanormelingen, P., McGregor, R. C. & Mann, D. G. (2011): The use of partial cox1 rbcL and LSU rDNA sequences for phylogenetics and species identification within the Nitzschia palea complex (Bacillariophyceae). – Eur. J. Phycol. 45: 413–425.

Urbánková, P. (2011): Molecular diversity and distribution of the species complex Frustulia rhomboides (Bacillariophyceae). – Master thesis, depon. in the Library of Department of Botany, Faculty of Science, Charles University in Prague.

Vankuren, N. W., Den Bakker, H. C., Morton, J. B. & Pawlowska, T. E. (2012): Ribosomal RNA gene di-versity, effective population size, and evolutionary longevity in asexual Glomeromycota. – Evolution 67: 207–224.

Veselá, J., Urbánková, P., ýerná, K. & Neustupa, J. (2012): Ecological variation within traditional diatom morphospecies: diversity of Frustulia rhomboides sensu lato (Bacillariophyceae) in European freshwater habitats. – Phycologia 51: 552–561.

Vyverman, W., Sabbe, K., Mann, D. G., Kilroy, C., Vyveman, R., Vanhoutte, K. & Hodgson, D. (1998): Eunophora gen. nov. (Bacillariophyta) from Tasmania and New Zealand: description and comparison with Eunotia and amphoroid diatoms. – J. Phycol. 3: 95–111.

Ward, R. D., Zemlak, T. S., Innes, B. H., Last, P. R. & Hebert, P. D. N. (2005): DNA barcoding Australia’s fi sh species. – Phil. Trans. R. Soc. London 360: 1847–1857.

F-NOVA142_S001-190.indb 161F-NOVA142_S001-190.indb 161 27.08.2013 10:31:3627.08.2013 10:31:36

Page 16: DNA-barcoding: A case study in the diatom genus Frustulia ... · undisputed ecological and economic importance, these organisms are still poorly characterized (McManus & Katz 2009,

162 Pavla Urbánkova & Jana Veselá

Will, K. W. & Rubinoff, d. (2004): Myth of the molecule: DNA barcodes for species cannot replace mor-phology for identifi cation and classifi cation. – Cladistics 20: 47–55.

Wylezich, C., Nies, G., Mylnikov, A. P., Tautz, D. & Arndt, H. (2010): An Evaluation of the Use of the LSU rRNA D1-D5 Domain for DNA-based Taxonomy of Eukaryotic Protists. – Protist 161: 342–352.

Zimmermann, J., Jahn, R. & Gemeinholzer, B. (2011): Barcoding diatoms: evaluation of the V4 subregion on the 18 S rRNA gene including new primers and protocols. – Org. Divers. Evol. 11: 173–192.

F-NOVA142_S001-190.indb 162F-NOVA142_S001-190.indb 162 27.08.2013 10:31:3627.08.2013 10:31:36