Identification of north-western Atlantic Porphyra (Bangiaceae, Bangiales) based on sequence variation in nuclear SSU and plastid rbc L genes

Phycologia (2003) Volume 42 (2), 109-122 Published 29 April 2003

Identification of north-western Atlantic Porphyra (Bangiaceae, Bangiales) based on sequence variation in nuclear SSU and plastid rbcL genes

ANITA S. KLEINl,2* , ARTHUR C. MATHIESON3.4, CHRISTOPHER D. NEEFUS3, DANIELLE F. CAIN2, HEATHER A. TAYLORl,

BRIAN W. TEASDALE3, ANDREW L WEST2, EDWARD J. HEHRE4, JULIET BRODIE5, CHARLES YARlSH6 AND AARON L WALLACEl

lDepartment of Biochemistry and Molecular Biology, University of New Hampshire, Durham, NH 03824, USA

2Graduate Program in Genetics, University of New Hampshire, Durham, NH 03824, USA

3Department of Plant Biology, University of New Hampshire, Durham, NH 03824, USA

4Jackson Estuarine Laboratory, University of New Hampshire, Durham, NH 03824, USA

5Bath Spa University College, Bath, BA2 9BN, England, UK

6Department of Ecology and Evolutionary Biology, University of Connecticut Stamford, CT 06901, USA

A.S. KLEIN, A.C MATHIESON, CD. NEEFUS, D.F CAIN, H.A. TAYLOR, B .W. TEASDALE, A.L WEST, E.J. H EHRE, J. BRODIE,

C Y ARISH AND A.L WALLACE. 2003. Identification of north-western Atlantic Porphyra (Bangiaceae, Bangiales) based on

sequence variation in nuclear SSU and plastid rbcL genes. Phycologia 42: 1 09- 1 22.

Six species of Porphyra have commonly been recognized in the north-western Atlantic from Long Island Sound to the

Canadian Maritimes: P. amplissima, P. leucosticra, P. linearis, P. minima, P. purpurea, and P. umbilicalis. Distinguishing

them with certainty has been problematic. A DNA-based system of molecular identification was developed using partial sequences of the nuclear small subunit ribosomal RNA gene (SSU) or the plastid ribulose-I ,S-bisphosphate carboxylase­

oxygenase large subunit gene (rbcL). Multiple samples of each taxon were surveyed for intraspecific variation. Intraspecific

SSU divergences for Porphyra 'Ieucosficta', P. 'miniata', P. 'umbilicalis', and P. 'purpurea' ranged from 0% to 1 %. There

was more variation for P. 'all1plissima' (0-2. 1 %) and P. 'linearis' (0-3.5%); however, each taxon was monophyletic. No

intraspecific differences were observed for these taxa in rbcL (one to eight samples per taxon). These sequences were

compared with P. yezoensis US I, introduced to Maine, and with P. 'dioica', a north-east Atlantic Porphyra easily confused

with P. 'purpurea'. To discriminate between P. 'purpurea', P. 'umbilicalis', and P. 'Ieucosticla' , SSU variation was used

to design primers for the Allele-Specific Polymerase Chain Reaction®>. With molecular tools, we could classify over 80%

of the monostromatic specimens surveyed, but the residue of unidentifiable specimens may indicate the existence of further

monostromatic species in the north-west Atlantic. Porphyra 'purpurea' was found to occur further south than previously

recorded. A morphologically cryptic Porphyra was discovered at Herring Cove, Nova Scotia, Canada.t Phylogenetic anal­

yses using SSU or rbcL sequences showed 'soft incongruence' between gene trees, i.e. the topologies of the phylograms

were similar but not identical, with only weak to moderate bootstrap support for the nodes that differed. Both trees strongly

supported a clade including P. 'purpurea', P. 'umbilicalis', P. 'linearis', and P. 'dioica'. Porphyra sp. Herring Cove was

allied with the remaining Porphyra taxa in the SSU tree. The rbcL phylogeny was less well resolved, consisting of a

polytomy of a P. 'purpurea'-P. 'umbilicalis'-P. 'linearis'-P. 'dioica' clade, Porphyra sp. Herring Cove, a clade comprising

P. 'amplissima' and P. 'minima', and a P. 'suborbiculata'-P. 'Ieucosticla'-P. yezoensis clade.

INTRODUCTION

The blade-forming red algal genus Porphyra C. Agardh grows

on rocky, cold- to warm-temperate shorelines throughout the

world (Brodie et ai. 1 996; Yoshida 1997) and is the basis of

a multibillion-dollar aquaculture industry in Asia (Hanisak

1 998). In the early 1990s, Coastal Plantations Inc. established

a small commercial nori aquaCUlture operation in northern

Maine, USA, which was licensed to grow culti vars U5 l and

H25 of the north-west Pacific taxon P. yezoensis Ueda (Levine

1998). Currently, there is considerable interest in fostering

Porphyra aquaCUlture in New England (USA), based on in­

digenous Porphyra species (Yarish et al. 1 998, 1 999), which

immediately prompts the question of which species occur in

this region. Prior to this study, s ix species of Porphyra had

been recorded from New England and the Canadian Maritime

Provinces: P. amplissima (Kjellman) Setchell & Hus in Hus,

P. leucosticta Thuret in Le Jolis, P. linearis Grevil le, P. min­

iata (c. Agardh) C. Agardh, P. purpurea (Roth) c. Agardh,

and P. umbilical is Kiitzing (Taylor 1 957; Schneider et al.

1 979; Mathieson & Hehre 1986; Bird & McLachlan 1 992;

Hehre & Mathieson 1 993; Silva 1 999). All but P. miniata

were originally described from locations in the north-east At­

lantic (http://www.algaebase.com).

* Corresponding author ([email protected]).

t Note added in proof: Porphyra sp. Herring Cove has been desig­nated P. birdiae CD. Neefus & A.C Mathieson. [Neefus CD., Ma­thieson A.C, Klein A.S. , Teasdale B . , Bray T & Yarish C 2002. Porphyra birdiae sp. nov. (Bangiales, Rhodophyta): a new species from the northwest Atlantic. Algae 1 7: 203-2 1 6. ] .

The present taxonomy of Porphyra is based largely on mor­

phological characters, including the colour, size, shape and

thickness of the thallus, cell dimensions, reproductive cell di­

vision sequences, and the distribution of fertile thallus tissues.

The cUiTently recognized north-west Atlantic species of Por­

phyra can usually be distinguished using these characters, ex­

cept for vegetative or juvenile specimens. For Porphyra, blade

morphology is extremely simple and there are few morpho-

1 09

1 10 Phycologia, Vol. 42 (2), 2003

logical characters on which to base species identification

(Lindstrom & Cole 1 993; Brodie et at. 1 996). In some instanc­

es, both monoecious and dioecious fronds have been attributed

to the same species (Taylor 1957 ; B ird & McLachlan 1992)

and different karyotypes have been reported for what is cur­

rently a single taxon (Kapraun & Freshwater 1987; Kapraun

et at. 1 99 1 ; Lindstrom & Cole 1992; Mitman 1 992; Mitman

& van der Meer 1994; Wilkes el at. 1999). These contradic­

tory patterns l ikely result from cryptic variabil ity, sexual dif­

ferences, polyploidy, or misidentification. Accordingly, Lind­

strom & Cole ( 1993) recommend that at least one nonmor­

phological diagnostic character should be used to verify the

identification of individual specimens; DNA-based molecular

markers provide such characters.

Because of the considerable taxonomic confusion and mis­

identification of many older and recent herbarium accessions

of Porphyra, and because gene sequences have not been ob­

tained from type specimens, we have chosen to qualify our

identifications by putting the species epithet in single quotes.

A similarly cautious approach to authenticating source mate­

rial (i .e. requiring DNA to be obtained from type material

before identification is accepted as certain) was previously

advocated by Brodie et at. ( 1998). The same conservative

approach has also been used when refening to Porphyra se­

quences obtained from GenBank, when the source material

was not from type specimens (Freshwater et at. 1994; Ragan

et at. 1994).

Bird el al. ( 1992) noted that, compared to higher plants and

green algae, several genera of red algae have unusual ly high

levels of sequence variation (� 1 5 %) for the nuclear-encoded

small subunit ribosomal RNA (SSU). Regions of the gene that appear to be phylogenetically informative i ncl ude the S' end

of SSU (approximately from helix 6 to helix 19) and the mid­

dle of the gene (helices 19-27). Both regions have been

shown to vary between species (C.J. Bird, personal commu­

nication), and do not include a group I intron present in hel ix

5 1 at the 3' end of the gene (Sti l ler & Waaland 1 993; Oliveira

& Ragan 1 994; Kunimoto et at. 1999b). The intron exhibits

intrapopulation variabil ity (Stil ler & Waaland 1993; Oli veira

& Ragan 1994), but its phylogenetic propelties and uti l ity as

a species-specific marker are as yet unclear (Muller et at.

1998, 200 I ; Kunimoto et at. 1 999b). Sequences of SSU have

been previously determined for several North Atlantic Por­

phyra taxa, including specimens identified as P. 'amptissima',

P. 'leucosticta', P. 'miniata', and P. 'purpurea' (Ragan et at.

1994; Oliveira et at. 1 995). The SSU sequences (GenBank

L36049, L26202) for two samples of P. 'umbilicalis' collected

from Nova Scotia (Ragan et at. 1 994) differed by 0.3%, hint­

ing that there was a low level of intraspecific sequence vari­

ation for this gene.

The plastid-encoded, ribulose-I,5-bisphosphate carboxyl­

ase-oxygenase large subunit (rbcL) gene has been shown to

be phylogenetically informative at the family, genus, and spe­

cies levels in the Rhodophyta (Freshwater et at. 1994). At the

beginning of our studies in 1 995, only three sequences of

Porphyra rbcL were available in GenBank, including a ful l ­

length sequence for P . 'purpurea' (Reith & Munholland 1 993)

and partial sequences from P. rosengurtii J. Coli & J. Cox

and a topotype specimen of P. 'carolinensis' J. Coli & J . Cox

(Freshwater et at. 1 994). Porphyra suborbiculata Kjellman

has recently been shown to include P. carolinensis and the

New Zealand taxon P. lilliputiana W. Nelson, G. Knight & M. Hawkes (Broom et at. 2002).

Phylogenetic analyses of the Bangiophyceae have previ ­

ously been conducted based o n the sequences o f a single or

at most two examples of each species and this limited sam­

pl ing is inadequate to assess intraspecific variation or to reveal

cryptic variation in morphologically similar taxa (Braverstock

& Mortiz 1 996). The number of samples or populations per

taxon needed to rel iably resolve relationships among groups

depends on several factors, including the amount of sequence

polymorphism evident for each species, the evolutionary dis­

tance between taxa, and the relative rate of evolution of the

gene(s) being examined. Using isozyme analysis, Lindstrom

& Cole ( 1 992, 1 993) noted that several Porphyra species have

such high levels of intraspecific variation that the current tax­

onomy is called into question; some of these taxa may rep­

resent species complexes.

The primary objective of thi s study was to develop molec­

ular screens that can reliably and objectively sort specimens

of north-west Atlantic Porphyra into taxa. The screens are

based on interspecific sequence differences in the SSU and

rbcL genes. An additional goal was to re-examine the phy­

logenetic relationships among north-west Atlantic Porphyra.

We assessed intraspecific sequence variation from SSU and

rbcL of north-west Atlantic Porphyra taxa, plus the morpho­

logically similar north-east Atlantic species, P. dioica Brodie

& Irvine, and cultured strains of P. yezoensis (U5 1 and H25)

from northern Maine (Levine 1 998; Yarish et at. 1 998, 1999).

Using new sequence data, allele-specific polymerase chain re­

actions (AS-PCR: Okayama et at. 1 989; Wu et at. 1 989) were

designed that were capable of providing accurate separations

of three monostromatic species, P. 'umbilicalis', P. 'leucos­

tieta' and P. 'purpurea', among hundreds of samples collected

in population surveys. Using AS-PCR and sequence analysis,

we have begun to document the distributional patterns of

north-west Atlantic Porphyra species.

MATERIAL AND METHODS

Sampling strategy

Collections were made throughout the year at ecologically dif­

ferent sites throughout New England and the Canadian Mar­

itime Provinces (New Brunswick, Nova Scotia and Prince Ed­

ward Island), using protocols simil ar to those outlined by Ma­

thieson & Hehre (1986) and Mathieson et at. ( 1 998). Samples

of all conspicuous Porphyra taxa were obtained from diverse

intertidal (on foot) and shallow subtidal habitats (by SCUBA).

Specimens were returned to the Jackson Estuarine Laboratory

for processing; in warm weather or when travel time from the

collecting site exceeded one hour, samples were kept on ice.

Tentative identifications to species were made based on mor­

phology, using a variety of taxonomic references, viz. Taylor

( 1 957), Coll & Cox ( 1 977), Kornmann ( 1 986, 1 994), Korn­

mann & Sahling ( 1 99 1 ), Schneider & Searles ( 1 99 1), B ird & McLachlan ( 1 992), Villalard-Bohnsack ( 1995), Brodie & Ir­

vine ( 1 997), and Sears (2002). Assumptions about the season­

al occunence and ecology of different taxa were supplemental

factors that helped in the initial sorting of field samples. For

example, all monostromatic Porphyra collected in New

Klein et al . : Molecular identification of north-west Atlantic Porphyra J 1 1

Hampshire during the summer months were assumed to be

either P. 'umbilicalis' or P. 'leucosticta', based on observa­

tions from earlier ecological studies (Taylor 1 957; Kingsbury

1 969; Hehre & Mathieson 1 970; Mathieson & Hehre 1 986).

The nomenclature employed primarily fol lows South & Tit­

tley ( 1 986) and Bird & McLachlan ( 1 992), except for recent

changes made by B rodie et al. ( 1 996), Brodie & Irvine ( 1 997),

Silva ( 1 999), Broom et al. (2002) and Sears (2002). Herbar­

ium voucher specimens documenting our studies are deposited

in the Albion R. Hodgdon Herbarium (NHA) of the University

of New Hampshire, USA (Table 1 ) . Subsets of each collection

were either frozen at -20°C or -80°C, pressed as herbarium

specimens, or dried in silica gel for later extraction of DNA.

DNA extraction and gene amplification

Small amounts of tissue (50-200 mg) were pulverized in liq­

uid nitrogen in a chilled mortar. Grinding was fol lowed by

DNA extraction in high-salt cetyldi methylethylammonium

bromide (CTAB) buffer (Doyle & Doyle 1 990), fol lowed by

extraction with phenol-chloroform, as described by Stiller & Waal and ( 1 993). DNA was precipitated with isopropanol and

then redissolved in 1 0 mM Tris-HCI, I mM EDTA buffer (pH

8.0; TrisEDTA). DNA was stored at 4°C or -80°C.

Because Porphyra thall i are frequently contaminated by

both macro- and microscopic epiphytes (Taylor 1 957; Bird et

at. 1 992), PCR primers were designed that were selective ei­

ther for the genus Porphyra in general or for particular species

of Porphyra. Highly conserved regions of SSU and rbcL in

Porphyra were distinguished, based on those accessions avail­

able in GenBank when this project was initiated ( 1 995 for

SSU and 1 997 for rbcL). The primers developed for Porphyra

in general would probably also anneal to template DNAs from

the sister genus Bangia Lyngbye, because Porphyra has been

demonstrated to be paraphyletic with respect to Bangia (Oliv­

eira et al. 1 995; Muller et al. 1 998).

Primers for SSU were designed based on the GenBank ac­

cessions of P. 'purpurea' (L2620 1 ), P. 'amplissima'

(L36048), P. 'miniata' ( L26200) and P. 'leucosticta'

(L26 1 99) (Ragan et al. 1 994). The Lasergene application

Primer Select (version 3 .72, DNASTAR, Madison, WI) was

used to select optimal primer pairs. The quality of each primer

pair was confirmed using Amplify 1 . 1 (Engels 1 993). The

approximate positions of these primers relative to SSU and

their sequences are l isted in Table 2, and corresponding DNA

amplification profiles are l isted in Table 3 .

Allele-specific S S U primers for individual taxa were de­

signed, based on sequence analyses of 8- 1 0 samples for each

species, these being collected at different seasons or geograph­

ical locations. Each primer pair ampli fied SSU fragments from

P. 'leucosticta', P. 'purpurea', or P. 'umbilicalis'. The rela­

tive positions of these primers, their sequences, and corre­

sponding PCR amplification profiles are shown in Tables 2

and 3. Standard PCR protocol was used for SSU amplification

(Mull i s & Faloona 1 987). Final concentrations of reagents

were 200 f.lM deoxyribonucleoside triphosphates (dNTPs), 0.4

f.lM for each primel� 2.25 mM MgClz, and 1 X Taq buffer B

(PIN M 1 66 1 ; Promega, Madison, WI). One to four microliters

(25-75 ng) of template and two units of Taq DNA polymerase

were added to each 50 f.ll reaction. Amplifications were car­

ried out in thin wall, 200 f.ll tubes or in 96-well plates in an

Ml thermocycler (Ml Research, Watertown, MA) equipped

with a Hot Bonner.@>.

For rbcL analyses, one set of primers, RBCL I and RBCL2,

was designed based on three previously determined Porphyra

sequences: P. 'rosengurtii' (U04042), P. 'suborbiculata' ( =

P. 'carolinensis'; U0404 1 ) (Freshwater et at. 1 994), and P.

'purpurea' (NC000925 ; Reith & Munholland 1 995). A second

set of primers, RBCL3 and RBCL4, was designed after pre­

liminary sequence analysis of several amplified fragments of

New England Porphyra taxa (Table 2). Amplification reac­

tions for rbcL contained 1 .5 mM MgCI2, J f.lM of each primer

pair, 200 f.lM dNTPs, 2 units of Taq polymerase, and 1 -4 f.lJ

(25-75 ng) of the template. To enhance specificity of ampli­

fication, a Hot Start procedure (D' Aquila et at. 1 99 1 ) was

used to initiate PCR with primers RBCL I and RBCL2.

DNA sequencing and phylogenetic analysis

All DNA sequences unique to this study were submitted to

GenBank, either as individual sequences or as population data

sets (Table 1 ) . The same specimens, or samples from the same

collection, were used to amplify and sequence both SSU ri­

bosomal DNA (rDNA) and the rbcL gene (Table I). To pre­

pare template DNAs for sequencing, amplified DNAs were

purified by gel electrophoresis on low meJting point agarose

(Life Technologies, Gaithersburg, MD), with the agar plugs

being digested with agarase (Sigma, St Louis, MO). Cycle

sequencing was carried out according to the manufacturer's

i nstructions, either with a Taq DyeDeoxy Terminator Cycle

Sequencing Kit with AmpliTaq (PIN 40 1 384; Perkin-Elmer,

Foster City, CA) or with an ABI Prism Dye Terminator Cycle

Sequencing Ready Reaction Kit with AmpliTaq DNA Poly­

merase FS (PIN 402080). Additional primers for sequencing

are listed in TabJe 2. Fluorescent-labelled DNAs were sepa­

rated on a 6% polyacrylamide gel using an Applied Biosys­

tems (ABI) Model 373 Automated DNA Sequencer at the

University of New Hampshire Sequencing Facility.

Sequences were generated with ABI DNA Sequencing Soft­

ware, version 2 . l . I , Base Caller ABT50, and were edited using

SegEd (Version 1 .0 .3 ; ABI, Foster City, CA). Edited sequenc­

es were assembled using the Lasergene Seq man application

(DNASTAR). Initially, a - 1 1 00 bp fragment of the rbcL was

sequenced. However, only those 8 1 5 bp that aligned with

available sequences for the outgroups were used in the phy­

logenetic analysis . Multiple sequence alignments for rbcL

were carried out with the MegAlign application of Lasergene.

For SSU, the two partial gene sequences for each taxon were

concatenated (total length - 990 bp) and then rDNA segments

were aligned, based on predicted secondary structures, using

the Dedicated Comparative Sequence Editor (DCSE: De Rijk

& De Wachter 1 993, http://rrna.uia.ac.be/dcse/). Al l stem do­

main alignments were verified manually. Corresponding se­

quences for two other bangiophycean red algae (Rintoul et at.

1 999), Erythrocladia Rosenvinge sp. (L26 1 88, AF087 1 1 7)

and Erythrotrichia carnea (Dillwyn) 1. Agardh (L26 1 88,

AF087 1 1 8), were used as outgroups in both sets of al ign­

ments. Only those regions of the rDNAs that could be readily

aligned were used for phylogenetic analysis (950 bp).

Sequence alignments were imported into PAUP *4.06b

(Swofford 1 998; http://paup.csit.fsu.edu/) for phylogenetic

Table 1. Algal samples with sequence data from this study. ...... tv

Col1ections Accession numbers GenBank accession numbers Albion Hodgdon Sample � ;:,-

Species Locations Dates H erbarium (NHA)' number SSU2 rbcL '< (") Cl Porphyra 'amplissima' Gove Point, Cobs cook Bay, North 8Jul. 1 995 58 1 86 1 AF358288, AF3583 1 2 AY028522 c;-

Lubec, Maine 2 AF358289, AF358313 AF02 1 034 O'Q 5'

3 AF358290, AF3583 1 4 . 4 AF35829 1 , AF3583 1 5 -< 0 5 AF358292, AF358316 :-' 6 AF358293, AF3583 1 7 .,. tv 7 AF358294, AF358318

� 8 AF358275, AF358274

Christmas Cove, Darmariscotta Riv- 1 8 May 1 996 63333, 63485-63487 1 AF358298, AF358322 tv 0 er, South Bristol, Maine 2 AF358299, AF358323 0 VJ Five Island, Sheepscot River, 20 Jun. 1 996 none available I AF358296, AF358320

Georgetown, Maine 2 AF358297, AF358321

P. 'dioica' Cruden Bay, Aberdeenshire, Scot- Jan. 1995 none available I AF3 1 9773, AF3 1 9774 AF095859 land, UK3 2 AY028524

North of Aberystwyth, Wales, UK 4 Apr. 1 998 AF081291

P. 'leucosticta' Gove Point Cobscook Bay, North 8 Jul. 1 995 58185, 699 1 5, 699 1 6, 6 1 945, 6 1 946 1 AF358278, AF358279 AF078744 Lubec, Maine 2 AF358345, AF358352

3 AF358346, AF358353 4 AF358347, AF358354

Pemaquid Point, Bristol, Maine 8 Aug. 1 996 64968 1 AF358348, AF358355 2 AF358349, AF358356

Rachel Carson Salt Pond Preserve, 9 Aug. 1 996 65006 AF358406, AF358407 AY028526 Chamberlain, Maine

Two Lights State Park, Cape Eliza- 9 Aug. 1 996 65284-65286 1 AF358350, AF358357 AY028525 beth, Maine 2 AF358351, AF358358

Fort Stark, New Hampshire 25 Aug. 1998 7 1 773 AF27 10784

P. 'linearis' S outh Bristol (near Turnip 1.), 1 8 Feb. 1996 60868, 6 1 075 1 AF358327, AF358334 Maine 2 AF358328, AF358335

3 AF358329, AF358336 4 AF358330, AF358337

Seapoint, Kittery, Maine 14 Feb. 1996 6 1 160 1 AF038587, AF038588 AF078745 2 AF358324, AF35833 1 3 AF358325, AF358332 4 AF358326, AF358333

P. 'miniata' Fink Cove, Nova Scotia, Canada 23 Jun. 1996 65301 1 AF358359, AF358385 AF02 1 033 2 AF358360, AF358386 AY028529 3 AF3583 6 1 , AF358387 4 AF358362, AF358388 5 AF358363, AF358389 6 AF358364, AF358390 7 AF358365, AF35839 1 8 AF358366, AF358392 9 AF358282, AF358283

Table 1. Continued.

Species

P. 'purpurea'

P. 'suborbiculata'

Porphyra sp. Herring Cove

P. 'umbilicalis'

P. yezoensis U5 1

Collections

Locations Dates

Cape Elizabeth Light, Cape Eliza­ 28 Apr. 1 997 beth, Maine

Yarmouth Harbor, Nova Scotia, 28 Sep. 1 996 Canada

Avonport, Nova Scotia, Canada 30 Sep. 1 996

Herring Cove, Nova Scotia, Canada 28 Sep. 1 996

Ross Island, Grand Harbor, New I Nov. 1 996 Brunswick, Canada

Lighthouse Cove, Dipper Harbor, I Nov. 1 996 New Brunswick, Canada

Leighton Cove, Whiting, Maine 24 Mar. 1 996

Camp Ellis, Saco, Maine 6 Feb. 1 996

Waterford, Connecticut Dec. 1 995

Masonboro, North Carolina 20 May 1 998

Herring Cove, Nova Scotia, Canada 28 Sep. 1 996

Sand Beach, Bar Harbor, Mount 22 Feb. 1 996 Desert Island, M aine

Red Point, Swans Island, Maine 1 2 Mar. 1 996

Bagaduce Falls, Brooksville, Maine 23 Mar. 1 996

Reids State Park, Georgetown, 8 Aug. 1 996 Maine

Fort Williams, Portland Head, 27 Jan. 1 996 Maine

Two Lights State Park, Cape Eliza­beth, Maine

Dover Point, New Hampshire

Fort Stark, Newcastle, New Hamp­shire

Eastport, Maine

28 Jan. 1 996

25 Aug. 1 998

20 Jun. 1 998

7 Oct. 1995

, Morphologically similar material was col lected at the same time for DNA extraction.

Accession numbers Albion Hodgdon

Herbarium (NHA)'

64304, 67135

65050

65045-65048

65 1 86-65 1 96

652 1 5-65224

62857

73866, 73867

none avai lable

65044

60822

6 1 1 68

638 1 1

638 1 6-638 1 8, 64962

6 1 996

not avai lable

7 1 783

7 1 780

598 1 9

Sample number

1 2

1 2 3

2 3

1 2

I 2

I 2

1 2

1 2

1 2 3

GenBank accession numbers

SSU2 rbcL

AF358284, AF358285

AF358374, AF358383 AF358375, AF358384

AF358373, AF358382 AY028533

AF358370, AF358379 AY028532 AF35837 1 , AF358380 AF358372, AF35838 1

AF358367, AF358376 AF358368, AF35837 AF358369, AF358378

AF358405, AF358367 AY028535 AY0285 3 1

AF358276, AF358277 AF078746 AF362362, AF362363 AY028534

AY028536 AY028537

AFI33792 AF078743

AY028523

AY l O0474, AY l O0473 AF319460

AF358393, AF358399 AY028540

AF358395, AF358401 AY02854I AF358286, AF358287

AF358394, AF358400 AF358280, AF35828 1

AF358396, AF358402 AY028539 AF358397, AF358403 AF358398, AF358404

AF078747

AY028538

AF27 1 0774

AF27 I 076 (2)

AF364074, AF364075 AF02 1 032

2 Each template was amplified and sequenced with two sets of primers: Por3 and Por4, and either Pori and Por2 or Por l 5 and Porl 6, two entries correspond to the respective sequences from the same template; P. 'suborbiculata' was amplified and sequenced with Pori I and Por l 9, and Porl7 and Porl8. 3 From conchocelis cultures. 4 Represents identical sequences for two isolates.

� � 5' � �

� o � () c � 0.: (lJ g. :!l () � o· ::l o ...., ::l o S­� (lJ ;!;. > ;-::l g. ;0 � � i3

...... VJ

1 1 4 Phyc% gia, Vol. 42 (2), 2003

Table 2. Primers for DNA amp lification and sequencing.

Gene fragment'

SSU

SSU

SSU sequencing Porphyra '!eucosticra' SSU

P. 'umbilicalis' SSU

SSU

SSU sequencing P. '!eucosricra' SSU

SSU amplification

P. 'suborbiculala' SSU amplification

P. 'suborbicLl!ala' SSU amplification with Pori 1 P. 'purpurea' SSU

Porphyra rbcL

Porphyra rbcL

Sequencing rbcL Sequencing rbcL Sequencing rbcL Sequencing rbcL

Primer

Pori Por2 Por3 Por4 Por5 Por6 Por7 Por8 Por9 PorlO PorI I Porl2 Porl3 Porl4 Por l 5 Porl6 Por l 7 Por l 8 Por l 9 Purl Pur2 RBCLI RBCL2 RBCL3 RBCL4 RBCLseq l PORRBCL5 PORRBCL6 PORRBCL7

, Fragment amp li fied or specificity for sequencing (or both) .

Sequence (5' -1 3')

GGGGGCCCTTTACGGTCCTCT CAATCCCCAGTCGGCATCGTT GCCAGTAGTCATATGCTTGTCTCA GGGCCCCCTATTGTTATTTT TAGCGTATATTAAAGTTGTTG CCTCAACGCCCGACTCACGAA AGACAAACTCCCACTGGCAACAAG CTTTTGTGCTTTGTGATGATTC AGAAGACACCCTCACAGAAACA TGCCTACACGCCCGACTC CCTTGTTACGACTTCTCCTTCCTC CCTGAGAAACGGCTACCA CAGAGCGCTTTGAGATGATTC CCCGACAAGAGGAGACAAACT ACGGGGAGGTAGTGACAAGAA CCCCCAGAACCCAAAGAC CCTTCGGGAACGCTTTGAG CGTCTTCGATCCCCTAACTTTC CCATTCAAATTTCTGCCCTATCA GGGGCGCGGACACAGAAGC TCATACAACAAATCCCCGCCAATC AACTGCTTCTATTATTGGTA CTTCCGAGTAATCTGTAAA ATGCGCTGGAGAGAAAGAT ACGTTGGCTGTTGGAGTCT TCATGTCATTTTTACGAGCCC TTTGACCAGCATGAATACCAC GAGTTAATATTTTCATCATCTT TGTAAATGGATGCGTATGGC

Relative position and orientation2

490-5 1 0 (F) 1 070- 1 050 (R)

1 7-40 (F) 497-478 (R) 598-6 1 8 (F) 1 33- 1 5 3 (F) 664-64 1 (R) 1 85-206 (F) 645-624 (R) 1 75- 1 92 (F)

1 787-1 764 (R) 356-373 (F) 1 93-2 1 3 (F)

676-656 (R) 468-488 (F) I 1 30-1 I 13 (R)

1 84-202 (F) 658-637 (R) 30 1 -324 (F) 654-672 (F)

1 084- 1 06 1 (R) 360-379 (F) 957-939 (R) 646-664 (F)

1 463-1 445 (R) 877-857 (R)

1 1 68-1 1 48 (R) 635-6 1 4 (R) 952-97 1 (F)

2 Relative to the coding sequence of the gene; (F) = forward, (R) = reverse.

analysis. The neighbour-joining (NJ) algorithm was used to

construct phylogenetic hypotheses for rbcL. Several character

weightings of the first, second and third codon positions were

explored. For SSU, character sets corresponding to secondary

structures (stems, bulges, and loops identified in DCSE) were

defined and various weightings of these features were ex­

plored. Uncorrected distances with among-site rate variation

were used with an NJ algorithm to construct the SSU phylog­

eny under the criterion of minimum evolution (ME). The dis­

tance model outlined by Tamura & Nei ( 1 993) was applied

with the NJ algorithm for the rbcL phylogeny. Various char­

acter weightings for first, second and third codon positions of

rbcL were explored. One thousand bootstrap replicates were

performed in order to assess the robustness of phylogenetic

reconstructions.

Gene trees were also examined using maximum parsimony

(MP) and maximum likelihood (ML) criteria, as implemented

in PAUP The topologies of the MP, ML and ME phylogenies

were compared using the parsimony and likelihood implemen­

tations of the Kishino-Hasegawa test. In order to bypass a bug

in PAUP* 4.0 1 Ob (http://paup.csit.fsu.edu/problems.html, 5 Au­

gust 2002), the distribution for likelihood score was set to

RELL.

RESULTS

SSU sequence analysis and reclassification of Porphyra samples

Partial SSU sequences were obtained from Porphyra samples

collected in the north-west Atlantic (Table 1) and correspond-

ing SSU sequences were also obtained for the north-east At­

lantic taxon P. 'dioica' (AF3 l 9773 and AF3 1 9774), the north­

west Pacific P. 'yezoensis' cultivar US I (Table I ), and P. 'su­

borbiculata' (AF I 33792), which is an apparently cosmopoli­

tan or perhaps introduced taxon ( Humm 1 979; Broom et al.

2002). The partial SSU sequence for P. 'yezoensis' US 1 was

identical to several other Japanese P. 'yezoensis' cultivars

(Kunimoto et al. 1999a).

There was good correspondence between existing SSU se­

quence data for north-west Atlantic Porphyra (Ragan et al.

1 994; Oliveira et al. J 995) and populations sampled in this

study. At least one sequence of each taxon of north-west At­

lantic Porphyra sampled (Table 4) was 2: 99% identical to

those previously reported. Intraspecific variation ranged from

0-0.34% in P. 'umbilicalis' (eight samples), to 0-2. 1 % in P.

'amplissima' ( 1 2 samples) and 0-3.5% for P. 'linearis' (eight

samples). Some intraspecific variants represented samples

from different locations and seasons; for example, P. 'am­

plissima' AF358293 and AF3583 1 7 from Cobscook Bay, Lu­

bec, Maine, differed by 1 .7% from AF358299 and AF358323

from the Darmariscotta River in South Bristol, Maine. In other

samples, sequence differences were observed for two different

samples of one taxon from a single collection. For example,

in P. 'amplissima' from Lubec, Maine, AF358288 and

AF3583 1 2 di ffered by 2. 1 % from AF358293 and AF3583 1 7

in partial SSU sequences; i n P. 'linearis' from Seapoint,

Maine, AF358327 and AF358334 differed by 3.5% from

AF358329 and AF358336.

Several samples collected during the summer of 1 995 from

Gove Point in Lubec, Maine, were initially identified by mor­

phological examination as P. 'umbilicalis'; subsequent se-

Klein et al.: Molecular identification of north-west Atlantic Porphyra 1 15

Table 3. Amplification profiles.

Primers Hot start' Anneal Extension Denature2 Cycles

POI'] and Por2 no 56°C, J min noc, 45 s 93°C, 30 s 30 Por3 and Por4 no 4SoC, ] min noc, 45 s 93°C, 30 s 30 Por6 and Por7 no 56°C, 1 mm noc, 45 s 93°C, 30 s 30 PorS and Por9 no 45°C, ] min noc, 45 s 93°C, 30 s 30 Por 1 0 and Por l ] no 50°C, 1 mm noc, 2 min 93°C, 30 s 30 Por 1 3 and Por l 4 no 54°C, 1 min noc, 45 s 93°C, 30 s 30 Por l S and Por l 6 no 47°C, ] min noc, 45 s 93°C, 30 s 30 Pur l and Por2 no 56°C, I min noc, 45 s 93°C, 30 s 30 RBCL I and RBCL2 yes 46°C, 1 min 72°C, I min 94°C, 30 s 30 RBCL3 and RBCL4 no 44°C, I min 72°C, I min 94°C, 30 s 30

, The reaction mixture was heated to 94°C for 3 minutes and then the temperature was reduced to SO°C while Taq polymerase was added. 2 All reactions heated initially to 93°C or 94°C for 3 minutes to denature templates prior to the first round of amplification. After 30 cycles of

amplification, a final cycle was carried out with an extension time of 1 0 minutes.

quence comparisons of SSU fragments showed that these sam­

ples were very different from gene sequences previously re­

ported for this species (L36049, L26202; Ragan et al. 1 994).

By contrast, the SSU sequences from Gove Point Porphyra

templates were less than � 0.5% divergent from P. 'leucos­

ticta' (L26 1 99) and so they were reclassified as P. 'leucostic­

ta' (Table 1 ) . Phylogenetic analysis confirmed that these Gove

Point Porphyra formed a monophyletic group with the P. 'leu­

costicta' accessions (data not shown). The SSU sequences ob­

tained for samples from Leighton Cove, Whiting, Maine

(AF358276 and AF358277), were > 99% identical to P. 'pur­

purea' (L2620 J ) , although they had been tentatively identified

by vegetative morphology as P. 'umbilicalis'. Finally, com­

parisons of SSU sequences with GenBank accession L26200

led to several Cape Elizabeth, Maine, vegetative samples be­

ing reclassified as P. 'amplissima'; they had initial ly been

identified as P. 'miniata' (Table 1 ) .

Evaluation of rbcL, new distributional records, and

cryptic taxa

A � 1 1 00 bp fragment of rbcL was amplified and sequenced

from 3 1 Porphyra specimens, which generaJly represented the

same or similar samples (i.e. from the same collection) as

those for which SSU sequences were obtained (Table 1 ) . The

rbcL fragments sequenced ranged from one (P. 'linearis') to

eight samples (P. 'purpurea'). In contrast to the intraspecific

sequence variation observed for SSU, little or no intraspecific

variation was observed in the rbcL gene. The rbcL sequences

of two P. 'umbilicalis' individuals sampled from an estuary

in New Hampshire were identical to each other and differed

by I bp from the rbcL sequence of two individuals of P.

'umbilicalis' from a nearby coastal site (Table 1) . The rbcL

sequences for the U5 1 and H25 cultivars of P. yezoensis were

identical to each other (data not shown).

The sequence of the rbcL fragment (AF078743, Table 1 )

amplified from a Waterford, Connecticut, collection (late au­

tumn 1 995), which was morphologically identified as P. 'leu­

costicta', was compared by Blastn Search (Altschul et al.

1990, 1997) with the GenBank database (August 1 997). It was

99% identical to a partial sequence of P. 'suborbiculata' ( =

P. 'carolinensis' U0404 1 ; Freshwater et al. 1994). The dif­

ference was a track of N in U0404 1 , this ambiguous sequence

being resolved in AF078743. Freshwater et al. 's (1994) rbcL

sequence (U04041) was obtained from a specimen collected

at Fort Fisher, North Carolina. The type location of P. caro­

linens is is not far away, at Fort Macon, Bogue Bank, North

Carolina, and material from Fort Fisher was also examined by

Coli & Cox (1977) in their original description of these spe­

cies. Further comparisons of P. 'suborbiculata' were made

with samples collected at Masonboro Island, North Carolina

(supplied by D.W. Freshwater). The partial rbcL sequence for

this accession (AY028523) was identical to the Waterford

sample (Table 1). Based on these data, further morphological

characterization, and recent results from Broom et al. (2002),

Table 4. Comparisons of intraspecific sequence divergence of Porphyra species for SSU.

Reference GenBank Percent identity Number of Percent divergence sequence(s) from with reference samples within taxon sequence

Species prior studies sequence sequenced in this study (%)'

P. 'amplissima' L3604S 99.5 1 3 0-2. 1 P. ' leucosticta' L26 1 99 99.9 9 0-0.46 P. 'linearis' none NA S 0-3.5 P. 'l1'liniata' L26200 99.6 1 2 0-0.97 P. 'purpurea' L2620 1 99.S 1 0 0-0.90 P. 'umbilicalis' L36049 99.4 S 0-0.34

L26202 99.5 AB01 379 99.5

P. yezoensis US I D79976 1 00 2

, Between Sand 1 3 sequences representing an individual taxon were aligned with the Clustal method as implemented in MegaJign, Lasergene. Sequences were trimmed to the same relative beginning and ends. The alignments were converted to Nexus format, and then imported into PAUP* 4.0a and the 5' and middle segments of each SSU were assembled by concatenation. Uncorrected pairwise distances for each sequence were calculated using the algorithm employed by PAUP* NA, none available.

1 1 6 Phycologia, Vol. 42 (2), 2003

A bp 185-213 located on sense strand

Species p. 'leucosticta' P. 'umbilicalis'

Complement of bp 643-677

Species P. 'leucosticta' P. 'umbilicalis'

B

<i>X Haelll

;0 .;:; � � �-3 "'"

t

." <:> .;:; � � """ � �

I

Por8

'''� Porl3 I

CTCTTTCCAGAGCGCTTTGAGATGATTC CT----TTTG--TGCTTTGTGATGATTC

r F"lp1 TGTTGCCAG-TGGGAGTTTGTCTCCTCTTGTCGGG TGTTTCT--GTGAGGG--TGTCTTCTTTTTCTGGG

483bp

C

<i>X Haelll

;0 .;:; � � �-3 "'"

t

." <:> .;:; � � ;;;' � �

460 bp

Fig. 1. A. Positions of allele-specific primers, corresponding to regions of high sequence divergence between Porphyra 'leucosricra' and P. 'umbilicalis'. B. Amplification products with PorS and Por9 (specific for P. 'umbilicalis'), using positi ve and negative control templates. C. Amplification with Por l 3 and Por l 4 (specific for P. 'leucosricta') us­ing positive and negative control templates. Samples were separated on 1 .4% agarose in 0.5 Tris Borate EDTA (TBE) buffer, with 0.005% ethidium bromide at 5 V cm '. Molecular weight markers correspond to bacteriophage q,XJ74 replicative form, cut with restriction enzyme Haell!. DNA fragments were visual ized under UV light.

the Atlantic distributional range for P. 'suborbiculata' can be

extended northwards, from North Carolina to Long Island

Sound.

A cryptic taxon was discovered among samples collected

at Herring Cove, Nova Scotia, Canada, in late September

1996. The sample was morphologically identified as P. 'pur­

purea' , but rbcL sequence data demonstrated that it is a pre­

viously unknown taxon from the north-west Atlantic. A Blastn

search of GenBank (August 2000) showed that the cryptic

taxon from HetTing Cove is no more than 9 1 % similar to any

other known Porphyra rbcL sequence. Hereafter, it wil l be

referred to as 'Porphyra sp. Herring Cove'. Similar material

was found later (autumn 1998) within Cobscook Bay, Maine.

Given the novel features of the Herring Cove rbcL sequences,

the SSU sequences for this material were also determined

(AY lO0474 and AY I00473).

CANADA QUEBEC

USA

" r-"I' l __ \

\ \ \ \ \ \

,! \

S I

MAINE

GULF OF MAINE 43

N

1 40

67 63

Fig. 2. Geographical distribution of Porphyra taxa confirmed by AS­peR. Between 2 and 20 samples were screened per collection. 0, P. 'Ieucosricta';., P. 'umbilicalis'; x, P. 'purpurea'. Map adapted from Sears (2002).

Development of AS-PCR molecular screens to

discriminate between three Porphyra taxa

AS-PCR molecular screens were developed to differentiate the

three most common monostromatic taxa, P. 'umbilicalis' , P.

'/eucosticta' , and P. 'purpurea' . The allele-specific primers

(Table 2) were designed to anneal to positions in the SSU that

differed maximally among the three rDNA genes but were

invariant within each taxon. For example, the sequences com­

plementary to primers Por8 and Por 1 3 differed by two inser­

tion-deletions and five base substitutions over 28 bases (Fig.

l A) . Primer pair Por l 3 and Por 1 4 was specific for P. 'leu­

costicta' ; in PCR, these primers supported amplification of a

483 bp fragment (Fig. IB). Primers Por8 and Por9 were spe­

cific for P. 'umbilicalis' , supporting the ampl ification of a 460

bp fragment (Fig. l C) . Primers Pur l and Pur2 were specific

for P. 'purpurea'; a standard amplification reaction using this

primer pair produced a 430 bp fragment (data not shown).

Positive and negative control templates were chosen for the

AS-PCR screens from which SSU was previously character­

ized by sequencing. Each allele-specific primer pair amplified

a fragment of predicted size from the positive control but did

not support DNA amplification from the negative control (Fig.

l B , C). Sequences of fragments amplified with allele-specific

primers confirmed that the primer pairs amplified SSU DNA

(data not shown).

Application of AS-PCR and DNA sequencing to identify

field collections of Porphyra

Table 5 summarizes the number of monostromatic samples

examined by AS-PCR or DNA sequencing. The efficacy of

the AS-PCR screens was limited to the three common species

for which the primers were designed to be targeted. Approx­

imately 1 8% of the > 450 samples screened by AS-PCR were

ambiguous: either the SSU fragments did not amplify, or they

amplified weakly with one of the three sets of allele-specific

primers, or they were amplified by two sets of the primers.

For example, DNAs from eight samples of Porphyra sp. Her-

Klein et a1. : Molecular identification of north-west Atlantic Porphyra 1 1 7

Table 5. Monostromatic samples identified by AS-peR.

Identi fication

P. ' leucoslicta' P. 'purpurea' P. 'umbilicalis' Ambiguous Total

Number of samples (%)

1 10 (24 . 1 ) I I I (24.2) J54 (33.7)

82 (1 7.9) 457

ring Cove were not amplified with any of the three sets of

AS-PCR primers. Other ambiguous results would be expected

if collections incl uded additional cryptic species or any other

Porphyra species whose rDNA sequence at the two primer

annealing sites might be similar, but not identical, comple­

ments to the corresponding AS-PCR primers.

Figure 2 summarizes the geographical distributions of P.

'umbilicalis' , P. 'leucosricta' , and P. 'purpurea' identified by

molecular methods. The number of Porphyra specimens from

each collection whose identity was confirmed by molecular

assays was 2-20. Geographical records for P. 'purpurea' ex­

tended as far south as Long Island Sound. Mixed populations

of P. 'purpurea' and P. 'umbilicalis' were found in collections

from Camp Ellis, Maine (February 1 996). The relative abun­

dance of Porphyra species may vary seasonally, e.g. 14 sam­

ples of P. 'umbilicalis' were collected in January 1996 at Two

Lights State Park on Camp Elizabeth, Maine, but the four

samples collected at this site in August 1996 were P. 'leu­

cosricta' . Over the course of this study, more than 700 sam­

ples were analysed by a DNA-based technique (sequencing or

AS-PCR). The combined DNA analyses indicated that ap­

proximately 25% of the field collections were misidentified

during the initial morphological inspection (data not shown).

Phylogenetic relationships between north-west Atlantic

Porphyra

Gene trees for the partial SSU and rbcL gene sequences were

constructed for native Porphyra taxa from New England and

the Canadian Maritime Provinces, together with P. 'dioica'

from Europe and the introduced Pacific Asiatic P. yezoensis

U5 1 . Single SSU sequences, representative of each taxon

(Cain 2000), were included in the phylogenetic analysis. Fol­

lowing reclassification of some accessions, all SSU sequences

within a designated taxon were shown to obey the principle

of monophyly (analyses not shown). Eryrhrorrichia carnea

and Erythrocladia sp. were used as outgroups. Secondary

structural alignments were determined using DCSE and only

those SSU segments that could be readi ly aligned (950 bp)

were used in phylogenetic analyses. The sequences for Por­

phyra rbcL began and ended at different positions relative to

those available for Eryrhrorrichia carnea and Erythrocladia

sp.; hence, only 8 15 bp were used in these phylogenetic anal­

yses.

Between species di vergences were 4 - 14% for SSU and 3-

1 1 % for rbcL. To assess whether the individual codon posi­

tions for rbcL were saturated for base substitutions (Griffiths

1 997), pairwise distances were computed for each codon po­

sition and plotted against the total uncorrected distances. Plots

of first- and third-position changes against total distances were

linear, with high correlation coefficients (R) of - 0.9 (data not

shown). With respect to the outgroups, Erythrotrichia carnea

and Erythrocladia sp., our analysis suggests that Porphyra

rbcL is not saturated for third-position substitutions.

Distance phylograms from SSU and rbcL are shown in Fig.

3. Phylogenies estimated using MP and ML produced similar

topologies, with bootstrap values at the same level of support

as in ME analysis (data not shown). The parsimony version

of the Kishino-Hasegawa test was used, as implemented in

PAUP* 4.0b I 0, to compare the phylogenetic topologies in the

three types of analysis for each gene; no significant differenc­

es were detected between ME, MP, and ML. Several weight­

ings applied for different codon positions of rbcL did not

change the overall topology of phylogenies, although signifi­

cant down-weighting of the third codon position decreased

resolution of the distance phylogram (data not shown).

The overall topologies of the distance phylograms are sim­

ilar but not identical for the partial SSU and partial rbcL trees

(Fig. 3). The Porphyra taxa in the SSU gene tree separate into

two strongly supported clades. One consists of Porphyra sp.

Herring Cove, P. 'miniata' , P. 'amp/issima' , P. 'suborbicu­

lata' , P. 'leucosticta' and P. yezoensis US I ; the other includes

P. 'dioica', P. 'purpurea' , P. 'linearis' and P. 'umbilicalis'.

A clade consisting of P. 'dioica' , P. 'purpurea', P. 'linearis'

and P. 'umbilicalis' is also strongly supported in the rbcL tree

(Fig. 3b). In the rbcL tree, P. 'amplissima' and P. 'miniata'

become si ster taxa with moderate bootstrap support, whereas

the SSU phylogeny does not support sister taxon relationships

for these two distromatic species. In the rbcL gene tree, P.

'suborbiculara' , P. 'leucosticta' and P. yezoensis form a clade

with weak bootstrap support. The rbcL tree has an unresolved

polytomy for these three clades and Porphyra sp. Hen·i ng

Cove. In both the SSU and rbcL gene trees, P. 'dioica' , which

has frequently been confused with P. 'purpurea' (Brodie & Irvine 1 997), is less closely affi l iated to this species than to

P. 'umbilicalis' .

DISCUSSION

Species identification

In this study, gene sequences of Porphyra species from New

England and the Canadian Maritime Provinces were compared

with each other and with sequences from previous studies of

Porphyra and other Rhodophycean algae (Freshwater et al.

1 994; Ragan et al. 1994; Mizukami et al. 1 998; Kunimoto et

al. 1 999a, b). The data were used to reclassify a significant

proportion of our samples. Identifications based on gene se­

quences are only as good as the initial morphological deter­

minations, and ideally identifications of samples used in mo­

lecular studies should be confirmed by inclusion of corre­

sponding molecular genetic data markers for type specimens.

Unfortunately, DNA extractions from 100-year-old type ma­

terials are still (and may remain) unreliable (Savolainen et al.

1995); therefore, it is premature to attempt molecular analyses

with precious type herbarium material . Eventually, however,

molecular analyses of type specimens may help to resolve

confusion regarding the taxonomy, phylogeny, ecology, and

biogeography of complex genera such as Porphyra.

Gene sequence analysis improved the discri mination be­

tween species in field collections. Comparison of the Gove

1 1 8 Phycologia, Vol. 42 (2), 2003

a SSU

100

74

99

P. 'amplissima ' 73

1,...--P. 'leucosticta '

P. yezoensis

P. 'suborbiculata '

'--------P. 'miniata '

P. sp Herring Cove

P. 'umbilicalis '

P. 'purpurea '

Erythrocladia

--0.0 1 substitutions/site Erythrotrichia carnea

b rbcL

1 00

,.---- P. 'amplissima '

1.... ____ P. 'miniata '

,.-----P. 'suborbiculata ' 55

1 00 P. 'leucosticta '

93 P. 'dioica '

1 00 1 00 P. 'linear is '

P. 'umbilicalis '

P. 'purpurea '

'-------P. sp. Herring Cove

,.--------Erythrocladia sp.

- 0.0 1 substitutions/site Erythrotrichia 'carnea '

Fig. 3. Minimum evolution distance phylograms from partial SSU (a) and rbcL gene sequences (b). The two SSU fragments (- 440 bp + -540 bp) that were sequenced for each taxon were concatenated and then aligned, based on RNA secondary structure predictions. The rbcL alignments correspond to 8 1 5 bp of the full-length gene (positions 380- 1 1 95). Multiple sequence alignments were analysed with the NJ algorithm in PAUP 4.01 0b, with 1 000 bootstrap repl icates. Tamura-Nei corrections were applied in the rbcL analysis. Only bootstrap values > 50 are shown.

Point herbarium samples (which were reclassified from P.

'umbilicalis' to P. 'leucosticta' based on SSU sequences) with

other P. 'leucosticta' is helping to refine our concept of the

range of vegetative morphologies of this taxon (C.D. Neefus

& A.C. Mathieson, in preparation). Porphyra 'amplissima'

was not c learly identified from the north-west Atlantic until

the early J 990s (Bird & McLachlan 1 992; Hehre & Mathieson

1 993); SSU sequence comparisons help to distinguish this tax­

on from the other distromatic taxon (P. 'miniata') in this re­

gion. Prior to this study, no published gene sequences had

been reported for P. 'linearis' (Table 1 ) . HistoricalJy, P. 'ii­

nearis' has been identified as a form or subspecies of P. 'um­

bilicalis' (see references in Taylor 1 957; Silva 1 999) and dis­

tinguishing P. '/inearis' from P. 'umbiiicalis' i s difficult, be­

cause young blades of the l atter can be of the size and stature

expected for P. 'linearis' (Bird & McLachlan 1 992). The SSU

sequence data helped us to evaluate herbarium samples and

subsequently to improve our ability to distinguish these two

taxa morphologicalJy (C.D. Neefus & A.c. Mathieson, in

preparation). Simi lar approaches have been taken to identify

cryptic invasions of Neosiphonia harveyi (1. Bailey) Choi,

Guiry & G.W. Saunders (Mcivor et ai. 200 1 , as Polysiphonia

harveyi; Choi et at. 200 1 ) and to differentiate between Gra­

teloupia C. Agardh species (Marston & Villalard-Bohnsack

2002).

The difficulty of making accurate species identification in

Porphyra is reinforced by a recent paper on phylogenetic anal­

ysis of Group I introns in Bangiales (Muller et ai. 200 1 ) . Over

50 accessions of Porphyra were examined for two SSU in­

trons and their presence or absence was mapped onto an SSU-

based phylogeny. Porphyra 'umbilicalis' sequences appear i n

three different clades and P . 'purpurea' appears in two sister

clades. For P. 'umbilicalis', 1 -2% variation occurred among

samples in the same clade, but substantially more sequence

divergence was found between samples in different clades,

with P. 'umbilicalis' HG (AF I 75549) differing from P. 'um­

bilicalis' HF (L25202) by - 4-6%. Muller et al. state that

the identification of Porphyra species in their study is tenta­

tive, because of the lack of a definitive global key. Our results,

however, indicate that it i s possible to correct species identi­

fications of north-west Atlantic Porphyra, so that each taxon

is monophyletic for SSU. Kunimoto et ai. ( 1 999a) have made

similar observations for accessions of P. 'yezoensis' and P.

'tenera'.

Sequence variation in SSU and rbcL genes

To evaluate intra- and interspecific polymorphism of Porphyra

from New England and the Canadian Maritime Provinces, se­

quence divergences have been determined for partial SSU and

rbcL gene fragments. As shown here, as well as in previous

studies (Ragan et at. 1 994; Oliveira et ai. 1 995; Kunimoto et

ai. 1 999a), the interspecific SSU divergence is very high. In

our study, SSU sequence divergence (- 990 bp) was 4-1 4%

and rbcL (8 1 5 bp) divergence was 3-1 1 %. Interspecific se­

quence variation for partial SSU genes in north-west Atlantic

Porphyra was thus comparable with that observed by Kuni­

moto et ai. ( 1 999b) for Japanese species (0.4- 1 1 %).

Kunimoto et ai. ( 1 999a) were concerned that the low level

of variation in SSU (0.4%) between P. 'tenera' and P. ye-

Klein et al . : Molecular identi fication of north-west Atlantic Porphyra 1 1 9

zoensis was insufficient to distinguish these as individual spe­

cies. Wild and cultivated varieties of P. yezoensis were iden­

tical over the whole length of SSU rDNA, but differed over

the more variable internal transcribed spacer I sequence

(ITS 1 ) . The P. 'tenera' ITS I was less similar to the P. ye­

zoensis ITS sequences than these were to each other. There­

fore, Kunimoto et al. ( 1 999a) concluded that the small amount

of sequence variation in the SSU between P. tenera and P.

yezoensis was significant in a phylogenetic context.

The SSU sequence of P. 'suborbiculata' from this study is

99% identical to that of P. lilliputiana, a species initially con­

sidered to be endemic to New Zealand (Nelson et af. 1 998),

and also 99% identical to that of P. 'carolinensis' (Broom et

af. 2002). Porphyra 'suborbiculata' is found i n Japan, Korea

and the Indian Ocean (Silva et af. 1 996) . Based on these and

other molecular data, Broom et af. (2002) recommended these

species be synonymized and we have i mplemented their rec­

ommendation. Broom et al. (2002) recommend that biogeo­

graphical studies of Porphyra, supplemented with careful

morphological and multilocus molecular analyses, are needed

to resolve apparent vicariant patterns.

The > 9% rbcL divergence between the cryptic Porphyra

sp. Herring Cove (Nova Scotia) and other Porphyra taxa ex­

amined here is greater than the pairwise differences between

most of the currently recognized taxa of Porphyra in the

north-west Atlantic. In a wider phylogenetic comparison, in­

cluding north-east Pacific Porphyra taxa, Porphyra sp. Her­

ring Cove was sister to, but distinct from , a potential new

taxon, to be called P. aestivalis Lindstrom (S.c. Li ndstrom,

personal communication). A detailed morphological descrip­

tion of Porphyra sp. Herring Cove is in preparation by C.D.

Neefus & A.c. Mathieson (personal communication). The

Herring Cove alga certainly represents a species hitherto un­

recorded in the north-west Atlantic, but whether it is a new

species or simply a new distributional record of a previously

recognized taxon cannot be answered until comparable mo­

lecular data are available for the other (> 1 33) recognized

Porphyra species (Yoshida et al. 1 997).

There is no generally recognized threshold value of se­

quence divergence that, by itself, can be used as a criterion

for subdividing a taxon into an additional species or infraspe­

cific taxa. Prior to our studies, SSU and rbcL intraspecific

sequence divergence for north-west Atlantic Porphyra species

had not been evaluated. We found that intraspecific SSU var­

iation was 0- 1 % (Table 4), except in P. 'amplissima' and P.

'linearis' . In a recent paper by Muller et af. (200 1 ) , samples

of P. 'miniata' and P. 'umbilicalis' varied by 1 -2% vs other

individuals within the same taxon. By contrast, we observed

very l ittle intraspecific variation in rbcL sequence variation

« 0. 1 %), albeit for a much smaller number of samples ( 1 -

6) per taxon.

The high level of sequence divergence within P. 'linearis'

and P. 'amplissima' samples suggests that further work is

needed to resolve whether these species need to be split and

whether subtle morphological or ecological characters corre­

late with specific sequence variants. However, careful mor­

phological assessments of P. 'leucosticta' , which has lower

intraspecific sequence variation « 0.5% for SSU, a relatively

conserved gene) than that in P. 'linearis' and P. 'amplissima' ,

suggest that this taxon could be further divided (Neefus el af.

2000). This is likely also to be true for several other wide-

spread north-west Atlantic taxa, such as P. 'purpurea' and P.

'umbilicalis' . Other, more sensitive molecular markers, such

as isozymes, the nuclear ribosomal internal transcribed spacer

(ITS), or microsatellites (Avise 1 994), can provide higher res­

olution in investigations of population structure, biogeograph­

ical variation, and ecophysiology, which should ultimately

lead to a re-evaluation of some of the north-west Atlantic

species (Neefus et af. 2000).

Molecular screens for species identification

Sequence data, variable PCR products and restriction frag­

ment length polymorphisms (RFLPs) have been used to dif­

ferentiate taxa in several algal groups (Stiller & Waaland

1 993; Broom et al. 1 999; Kuni moto et af. 1 999b; Teasdale et

al. 2002). The use of AS-PCR, which was designed to differ­

entiate the three monostromatic taxa P. 'leucosticta' , P. 'pur­

purea' and P. 'umbilicalis,' provides a rapid method to screen

l arge numbers of vegetative monostromatic samples from di­

verse ecological sites along the north-west Atlantic coast. The

three primer pairs were selected to anneal to regions of the

SSU that varied maximally between P. 'umbilicalis' , P. 'pur­

purea' , and P. 'leucosticta' . A portion of the samples ( � 1 8%

of 457 samples; Table 5) produced ambiguous results, which

could be explained if the DNA template in these cases rep­

resented a different, otherwise unrecorded monostromatic spe­

cies, i .e. so that none of the three sets of primers was a perfect

match in primer-template annealing. Furthermore, the three

sets of primers would potentially fail to discriminate between

additional taxa whose SSU sequence was similar to that of

one or more of the target species. Once all the species prev­

alent in the north-west Atlantic have been identified, appro­

priate primer pairs could be designed to discriminate addi­

tional taxa, based on SSU sequences.

Bird & McLachlan ( 1 992) and Li ndstrom & Cole ( 1 992)

suggest that the species richness of Porphyra for the north­

west Atlantic has previously been underestimated; our results

confirm their speculation. At least eight species of Porphyra

were recognized in our study: P. 'amplissima' , P. 'suborbi­

culata' , P. 'linearis' , P. 'leucosticta' , P. 'miniata' , P. 'pur­

purea' , P. 'umbilicalis' , and Porphyra sp. Herring Cove. The

SSU and rbcL sequence data from this study provide an im­

portant reference and also make it feasible to compare north­

western species with the north-east Atlantic Porphyra (e.g. P.

'dioica' ) or Porphyra species from the other oceans (Broom

et af. 1 999, 2002; Kunimoto et af. 1 999b) .

Phylogenetic analyses

A minor refinement of our phylogenetic analyses of partial

SSU sequences vs earlier studies (Ragan et af. 1 994; Oliveira

et al. 1 995) is that our alignments are based on the predicted

secondary structure (De Rijk & De Wachter 1 993). Our tree

topologies for partial SSU and rbcL sequences were generally

similar but not identical to those of previous studies (Fig. 3) .

Both the SSU and rbcL partial gene trees strongly support a

sister-taxon relationship between P. 'linearis' and P. 'umbil­

icalis' . Such results contrast with other treatments, based on

i sozymes, which ally P. 'linearis' with P. 'purpurea' (Lind­

strom & Cole 1 992, 1 993) . Some of the confusion about P.

'linearis' may result from the existence of two morphologi­

cally similar species, with different chromosome numbers

1 20 Phycologia, Vol. 42 (2), 2003

(S.c. Lindstrom, personal corrununication). Our molecular re­

sults support earlier taxonomic studies (e.g. Taylor 1 957), in

which P. ' linearis' was thought to be closely related to P.

'umbilicalis' [P. 'umbilicalis' forma 'linearis' (Greville) Ro­

senvinge] .

Comparison o f the S S U phylogeny i n this study with the

recent study by MUller et at. (200 1 ) is hampered by apparent

differences in species identification. Several taxa (i .e. P. 'pur­

purea' , P. 'umbilicalis' and P. 'miniata' ) were not monophy­

letic in MUller et at. 's analysis, whereas in ours each species was clearly homogeneous with respect to any other and mono­

phyletic.

Both the SSU and rbcL gene phylogenies in this study con­

firm the findings of Yamazaki et at. ( 1 996), which suggest an

affinity between the Pacific Asiatic P. 'yezoensis' and the

North Atlantic P. 'leucosticta' . Although bootstrap support

was weak in the partial SSU tree, the sister-taxon relationship

between these species was strongly supported in the rbcL phy­

logeny.

The demonstration of synonymy between P. 'carolinensis'

and P. 'suborbiculata' (Broom e/ at. 2002) contradicts the

proposition that there are no species in common between the

North Atlantic and the North Pacific (Yoshida et al. 1 997).

Alternative hypotheses to explain the presence of species in

both the North Atlantic and the North Pacific are transarctic

migration during interglacial periods (S.C. Lindstrom, person­

al communication) or that Pacific species were introduced into

the northern Atlantic via shipping trade in the 1 9th century

(Brodie et al. 1 998; see also Villalard-Bohnsack 1 998). Humm

( 1 979) also speculated that the occurrence of P. 'sub­

orbiculata' ( = P. 'carolinensis' ) in North Carolina (Coli & Cox 1 977) may be because of recent introduction.

The major clades in both of our gene trees (Fig. 3) are

consistent with those observed in previous phylogenetic stud­

ies using SSU (Oliveira et at. 1 995 ; MUller et at. 1 998), al­

though earlier investigators examined different subsets of Por­

phyra taxa. The SSU gene tree (Fig. 3a) does not support any

correlation between number of cell layers and phylogenetic

relationship, agreeing with the conclusions drawn by Oliveira

et at. ( 1 995). By contrast, the rbcL gene tree defined the dis­

tromatic P. 'miniata' and P. 'amplissima' (which can be mono­

stromatic or di stromatic on the same blade: J. Brodie, unpub­

l ished observation) as sister taxa. Our rbcL tree (Fig. 3b)

loosely grouped the monostromatic taxa P. 'leucosticta', P.

'yezoensis', and P. 'suborbiculata' into a distinct clade.

The gene trees presented in this study are based on partial

gene sequences. Some of the clades had weak bootstrap sup­

port and topologies may shift for these clades if ful l-length

gene sequences are analysed. Such ' soft' incongruities (See­

lanan et at. 1 997) between the two gene trees do not require

explanation at this point. However, well-supported incongru­

ities between the nuclear and plastid gene trees are present in

our analyses, for example in the position of P. 'miniata' , and

incongruities have also been observed at a higher taxonomic

level in the Bangiophycideae (Oliveira & Bhattacharya 2000).

Several hypotheses may be put forward to account for in­

congruities between the rbcL and SSU gene trees. One or both

genes may be saturated for base substitutions, leading to an

erroneous tree (Swofford et at. 1 997). In our analyses, mu­

tational saturation of rbcL was not apparent, but it has been

found to be problematic in phylogenetic reconstruction in the

related bangiophycean alga Bangia atropurpurea (Roth) C.

Agardh (K.M. MUller, personal communication). Secondly,

the gene tree for SSU may be erroneous because of concerted

evolution in the high-copy number rDNA (Hil l i s et al. 1 99 I ) .

Thirdly, differences in gene trees may reflect different evo­

lutionary histories between chloroplast and nuclear gene lin­

eages, perhaps as a consequence of chloroplast capture. One

way to test these various hypotheses would be to compare the

rbcL and SSU gene trees against those for other nuclear and

plastid genes (Olmstead & Palmer 1 994). D. Mathews and S.

Minocha (personal communication) have found that the to­

pology of a gene tree for intron 1 of the nuclear-encoded ac­

tin} gene is congruent with the rbcL gene tree in this study.

Such evidence suggests that concerted evolution of the rDNA

may obscure the evolutionary history of Porphyra species and

that an accurate phylogeny for the genus wil l require the ex­

amination of multiple genes.

Several other research groups are generating gene sequence

data for many of the recognized species of Porphyra world­

wide (Kunimoto et at. I 999a, b; Nelson et al. 200 1 ; Broom

et at. 2002; S.c. Lindstrom, K.M. MUller and R. Sheath, per­

sonal communication). Thus far, only a l i mited amount of data

has been released. A synthesis of more gene sequence data,

with appropriate phylogenetic and morphological compari­

sons, is needed to resolve the evolutionary history and prob­

lems of synonymy in the genus Porphyra and its paraphyletic

sister taxon Bangia (see Oliviera et al. 1 995; MUller et at.

1 998).

ACKNOWLEDGEMENTS

The present study was supported by NOAA's New Hamp­

shire-Maine Sea Grant College Program as project RlFMD-

1 40. We would also like to acknowledge the help of several

individuals who made available sequence data, prior to pub­

lication, or who provided samples for comparison, viz. Dr

Sandra Lindstrom (University of British Columbia, Vancou­

ver, British Columbia, Canada); Drs Kirsten MUller (Univer­

sity of Waterloo, Ontario, Canada) and Robert Sheath (Cali­

fornia State University San Marcos, CA); and Drs Yuzuru

Mizumaka and Masahiko Kunimoto (Japanese National Fish­

eries Institute). Drs Sandra Lindstrom and Carolyn Bird (In­

stitute for Marine B ioscience, National Research Council of

Canada, Halifax, Nova Scotia, Canada) also provided invalu­

able discussions and advice during our studies. In particular,

we want to thank a third anonymous reviewer for insightful

comments and suggestions for further analysis. This article is

published as Contribution Number 356 from the Jackson Es­

tuarine Laboratory and the Center for Marine Biology at the

University of New Hampshire, USA.

REFERENCES

ALTSCHUL S.P., GISH W., MILLER W. , MYERS E.W. & LIPMAN D.l. 1 990. Basic local alignment search tool. Journal of Molecular Biology

2 1 5 : 403-4 10. ALTSCHUL S.P., MADDEN T.L., SCHAFFER A.A., ZHANG l. , ZHANG Z.,

MILLER W. & LIPMAN D.l. 1 997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Ac­

ids Research 25: 3389-3402.

Klein et al . : Molecular identification of north-west Atlantic Porphyra 1 2 1

AVISE J.c. 1 994. Molecular markers, natural history, and evolution.

Chapman and Hall, New York. 499 pp. BIRD c.J. & McLACHLAN J. 1 992. Seaweed flora of the Maritimes. 1.

Rhodophyta - the red algae. Biopress, Bristol, UK. 1 77 pp. B llm c.J., RICE E.L., MURPHY C.A. & RAGAN M .A. 1 992. Phyloge­

netic relationships in the Gracilariales (Rhodophyta) as determined by 18S rDNA. Phycologia 3 1 : 5 1 0-522.

B RAVERSTOCK P.R. & MORITZ C. 1 996. Project design. In: Molecular

systematics (Ed. by D.M. Hill is, C. Moritz & B .K . Mable), pp. 1 7-28. Sinauer Associates, Sunderland, Massachusetts.

BRODIE J. & IRVINE L.M. 1 997. A comparison of Porphyra dioica sp. nov. and P. purpurea (Roth) c. Ag. (Rhodophyta: Bangiophycidae) in Europe. Cryptogamie, A lgologie 1 8 : 283-297.

BRODIE J., HAYES P. K., BARKER G.L. & IRVINE L.M. 1 996. Molecular and morphological characters distinguishing two Porphyra species (Rhodophyta: Bangiophycidae). European Journal of Phycology 3 1 : 303-308.

BRODlE 1., HAYES P.K., BARKER G.L., IRVINE L.M. & BARTSCH I . 1 998. A reappraisal of Porphyra and Bangia (Bangophycidae, Rhodophy­ta) in the northeast Atlantic based on the rbcL-rbcS intergenic spac­er. Journal of Phycology 34: 1 069- 1 074.

BROOM J.E., JONES W.A., HILL D.F., KNIGHT G.A. & NELSON W.A. 1 999. Species recognition in New Zealand Porphyra using 1 8S rDNA sequencing. Journal of A pplied Phycology 1 I : 42 1-428.

B ROOM J.E., NELSON W.A., Y ARISH C., JONES W.A., AGUILAR ROSAS

R. & AGUILAR ROSAS L.E. 2002. A reassessment of the taxonomic status of Porphyra suborbiculata, Porphyra carolinensis and Por­

phyra liLLiputiana (Bangiales, Rhodophyta) based on molecular and morphological data. European Journal of Phycology 37: 227-236.

CAIN D.F. 2000. Genetic characterization and phylogenetic analysis of native western North Atlantic Porphyra, using the 1 8S ribosomal gene. MS thesis. University of New Hamphsire, Durham, New Hampshire. 76 pp.

CHOI H.G., KIM M.S., GUIRY M.D. & SAUNDERS G.W. 200 1 . Phylo­genetic relationships of Polysiphonia (Rhodomelaceae, Rhodophy­ta) and its relatives based on anatomical and nuclear small-subunit rDNA sequence data. Canadian Journal of Botany 799: 1 465-1 476.

COLL J. & Cox J. 1 977. The genus Porphyra C. Ag. (Rhodophyta, Bangiales) in the American North Atlantic. 1. New species from North Carolina. Botanica Marina 20: 1 55-1 59.

D' AQUILA R.T., BECHTEL L.J., VITELER J.A., ERON 1.1., GORCZYCA P. & KAPLIN J.C. 1 99 1 . Maximizing sensitivity and specificity of PCR by preamplification heating. Nucleic A cids Research 1 9 : 37-49.

DE RUK P. & DE WACHTER R. 1 993. DCSE, an interactive tool for sequence alignment and secondary structure research. Computer

A pplied Biosciences 9: 735-740. DOYLE I.J. & DOYLE J.L. 1 990. Isolation of plant DNA from fresh

tissue. Focus 1 2: 1 3- 1 5. ENGELS W.R. 1 993. Contributing software to the Internet: the Amplify

program. Trends in Biochemical Sciences 1 8 : 448-450. FRESHWATER D.W., FREDERICQ S. , B UTLER B.S . , HOMMERSAND M.H. &

CHASE M.W. 1 994. A gene phylogeny of the red algae (Rhodophyta) based on plastid rbcL. Proceedings of the National Academy of

Science of the United States of A merica 9 1 : 728 1-7285. GRIFFITHS C.S. 1 997. Correlation of functional domains and rates of

nucleotide substitution in cytochrome b. Molecular Phylogenetics

and Evolution 7: 352-365. HANISAK M.D. 1 998. Seaweed cultivation: global trends. World A qua­

culture 29: 1 8-2 1 . HEHRE E.J . & MATHIESON A.C. 1 970. Investigations of New England

marine algae n. Composition, seasonal occurrence and reproductive periodicity of the marine Rhodophyceae in New Hampshire. Rho­

dora 72: 1 94-239. HEHRE E.J. & MATHIESON A.C. 1 993. Porphyra amplissima ( Kjellman)

Setchell et Hus: new records of an 'arctic' seaweed in southern Maine, New Hampshire, and northern Massachusetts. Rhodora 95: 1 84- 1 87.

HILLIS D.M., MORITZ c., PORTER C.A. & BAKER R.J. 1 99 I . Evidence for biased gene conversion in concerted evolution of ribosomal DNA. Science 25 1 : 308-3 1 0.

HUMM H.J. 1 979. The marine algae of Virginia. Special Papers i n Marine Science, number 3 . University Press o f Virginia, Charlottes­vi l le. 263 pp.

KAPRAUN D.F. & FRESHWATER D.W. 1 987. Karyological studies of five species of Porphyra (Bangiales, Rhodophyta) from the North At­lantic and Mediterranean. Phycologia 26: 82-87.

KAPRAUN D.F., HINSON T.K. & LEMUS A.J. 1 99 1 . Karyology and cy­tophotometric estimation of interspecific and intraspecific nuclear DNA variation in four species of Porphyra (Rhodophyta). Phycol­

ogia 30: 458-466. KINGSBURY J .M. 1 969. The seaweeds of Cape Cod and the islands.

Chatham Press, Chatham, Massachusetts. 2 1 2 pp. KORNMANN P. 1 986. Porphyra yezoensis bei Helgoland - eine entwick­

lungsgeschichtliche Studie. Helgoltinder wissenschaftliche Meere­

suntersuchungen 40: 327-342. KORNMANN P. 1 994. Life histories of monostromatic Porphyra species

as a basis for taxonomy and classification. European Journal of

Phycology 29: 69-7 1 . KORNMANN P. & SAHLING P.-H. 1 99 1 . The Porphyra species of Hel­

goland (Bangiales, Rhodophyta). Helgoltinder Meeresuntersuchun­

gen 45: 1-38. KUNIMOTO M., KITO H., KAMIN1SHI Y, MIZUKAMI Y. & MURASE N.

I 999a. Molecular divergence of the SSU rRNA gene and internal transcribed spacer 1 in Porphyra yezoensis (Rhodophyta). Journal

of A pplied Phycology 1 1 : 2 1 1 -2 1 6. KUNIMOTO M., KITO H. , YAMAMOTO Y., CHENEY D.P., KAMLNISHI Y. &

MIZUKAMI Y. 1 999b. Discrimination of Porphyra species based on small subunit ribosomal RNA gene sequence. Journal of Applied

Phycology 1 1 : 203-209. LEVINE I.R. 1 998. Commercial cultivation of Porphyra (nori) in the

United States. World A quaculture 29: 37-38, 45-47. LINDSTROM S.c. & COLE K.M. 1 992. Relationships between some

North Atlantic and North Pacific species of Porphyra. Canadian

Journal of Botany 70: 1 355-1 363. LINDSTROM S.c. & COLE K.M. 1 993. The systematics of Porphyra:

character evolution in closely related species. Hydrobiologia 260/ 26 1 : 1 5 1 - 1 57.

MARSTON M.F. & VILLALARD-BoHNSACK M . 2002. Genetic variabil ity and potential sources of Grateloupia doryphora (Halymeniaceae, Rhodophyta) an invasive species in Rhode Island waters (USA). Journal of Phycology 38: 639-658.

MATHIESON A.C. & HEHRE E.J. 1 986. A synopsis of New Hampshire seaweeds. Rhodora 88: 1- 1 39.

MATHIESON A.C., DAWES c.J. & HEHRE E.J. 1 998. Floristic and zo­nation studies of seaweeds from Mount Desert Island, Maine: an historical comparison. Rhodora 1 00: 333-379.

McIVOR L., MAGGS C.A., PROVAN J. & STANHOPE M.J. 200 1 . rbcL

sequences reveal mUltiple cryptic introductions of the Japanese red alga Polysiphonia harveyi. Molecular Ecology 1 0: 9 1 1-9 1 9.

MITMAN G.G. 1 992. Meiosis, blade development and sex determina­

tion in Porphyra umbilicalis (L.) 1. Agardh from A vonport, Nova

Scotia, Canada. PhD thesis. Dalhousie University, Nova Scotia, Canada. 1 42 pp.

MITMAN G.G. & VAN DER MEER J.P. 1 994. Meiosis, blade develop­ment, and sex determination in Porphyra purpurea (Rhodophyta). Journal of Phycology 30: 1 47-1 59.

MLZUKAMI Y, KAMINISHI Y., KUNIMOTO M., KOBAYASHI M., MURASE

N. & KITO H. 1 998. Comparison of partial nucleotide sequences in the exonic region of a small subunit ribosomal RNA gene for dis­crimination of laver (Porphyra) species and cultivars. Fisheries Sci­

ence, Tokyo 64: 886-89 1 . MOLLER K.M., SHEATH R.G., VIS M.L., CREASE T.J. & COLE K.M.

1 998. Biogeography and systematics of Bangia (Bangiales, Rho­dophyta) based on the Rubisco spacer, rbcL gene and 1 8S rRNA gene sequences and morphometric analyses. I. North America. Phy­

cologia 37: 1 95-207. MOLLER K.M., CANNONE J.J., GUTELL R.R. & SHEATH R.G. 200 1 . A

structural and phylogenetic analysis of the group IC 1 introns in the order Bangiales (Rhodophyta). Molecular Biology and Evolution

1 8 : 1 654-1 667.

1 22 Phycoiogia, Vol. 42 (2), 2003

MULLlS KB. & FA LOON A F.A. 1 987. Specific synthesis of DNA i n vitro via a polymerase-catalyzed chain reaction. I n : Recombinant

DNA Part F (Ed. by R. Wu), pp. 335-350. Academic Press, San Diego.

NEEFUS e.D., MATHIESON A.C., Y ARISH e., KLEIN A.S., WEST A., TEASDALE B . & HEHRE E.1. 2000. Five cryptic species of Porphyra

from the Northwest Atlantic. Journal of Phycology 36, supplement: 73.

NELSON W.A., KNIGHT G.A. & HAWKES M.W. 1 998. Porphyra lilLi­

putiana sp. nov. (Bangiales, Rhodophyta): a diminutive New Zea­land endemic with novel reproductive biology. Phycological Re­

search 46: 57-6 1 . NELSON W.A., BROOM J.E. & FARR TJ. 200 1 . Four new species of

Porphyra (Bangiales, Rhodophyta) from the New Zealand region described using traditional characters and 1 8S rDNA sequence data. Cryptogam ie, A lgologie 22: 263-284.

OKAYAMA H., CURIEL D.T, BRANTLY M .L., HOLMES M.D. & CRYSTAL

RG. 1 989. Rapid, nonradioactive detection of mutations in the hu­man genome by allele-specific amplification. Journal of Laboratory

and Clinical Medicine 1 1 4 : L OS-I 1 3 . OLlVE1RA M.C. & BHATTACHARYA D. 2000. Phylogeny of the Bangio­

phycidae ( Rhodophyta) and the secondary endosymbiotic origin of algal plastids. A merican Journal of Botany 87: 482-492.

OLIVEIRA M.C. & RAGAN M.A. 1 994. Variant forms of a group I intron in nuclear small-subunit rRNA genes of the marine red alga. Mo­

lecular Biology and Evolution I I : 1 95-207. OLIVEIRA M.e., KURNIAWAN J. , BIRD e.J., RICE E.L., MURPHY e.A.,

SINGH R.K, GUTELL R.R. & RAGAN M.A. 1 995. A prel iminary in­vestigation of the order Bangiales (Bangiophycidae, Rhodophyta) based on the sequences of nuclear small-subunit ribosomal RNA genes. Phycological Research 43: 7 1 -79.

OLMSTEAD R.G. & PALMER J.D. 1 994. Chloroplast DNA systematics: a review of methods and data analysis. A merican Journal of Botany

8 1 : 1 205-1 224. RAGAN M.A., BIRD e.l. , RlCE E.L., GUTELL R.R., MURPHY C.A. &

SINGH R.K. 1 994. A molecular phylogeny of the marine red algae ( Rhodophyta) based on the nuclear small-subunit rRNA gene. Pro­

ceedings of the National A cademy of Science of the United States

of A merica 9 1 : 7276-7280. REITH M. & MUNHOLLAND J. 1 993. A high-resolution gene map of the

chloroplast genome of the red alga Porphyra purpurea. Plant Cell

5: 465-475. REITH M. & MUNHOLLAND J. 1 995. Complete nucleotide sequence of

the Porphyra purpurea chloroplast genome. Plant Molecular Biol­

ogy Reporter 1 3 : 333-335. RINTOUL TL., SHEATH R.G. & V IS M. 1 999. Systematics and bioge­

ography of the Compsopogonales (Rhodophyta) with emphasis on the freshwater famil ies in North America. Phycologia 38: 5 1 7-525.

SAVOLAINEN V.W., CENOUD P., SPICHIGER R., MARTINEZ M.D., CREVE­COEUR M. & MANEN J.F. 1 995. The use of herbarium specimens in DNA phylogenetics: evaluation and improvement. Plant Systemat­

ics and Evolution 1 97 : 87-98. SCHNEIDER e.W. & SEARLES R.B. 1 99 1 . Seaweeds of the southeastern

United States. Duke University Press, Durham, North Carolina. 553 pp.

SCHNEIDER C.W., SUYEMOTO M.M. & YARlSH e. 1 979. An annotated

checklist of Connecticut seaweeds. State Geological and Natural H istory Survey, Connecticut Department of Environmental Protec­tion, Hartford, Connecticut. 20 pp.

SEARS J .R. [Ed.] 2002. NEAS keys to benthic marine algae of the

northeastern coast of North A merica from Long Island Sound to

the Strait of Belle Isle, ed. 2. Contribution No. 2 . Northeast Algal Society, Dartmouth, Massachusetts. 1 6 1 pp.

SEELANAN T, SCHNABEL A. & WENDEL J.F. 1 997. Congruence and consensus in the cotton tribe (Mal vaceae). Systematic Botany 22: 259-290.

SILVA P.e. 1 999. Porphyra. Index Nominum Algarum. Web page:

http://ucjeps.berkeley.edu/rlmoe/porphyra.html (revised 20 Decem­ber 1 999). University Herbarium, University of California, Berke­ley.

SILVA P.C., BASSON P.w. & MOE R .L. 1 996. Catalogue of the benthic marine algae of the Indian Ocean. University of California Publi­

cations in Botany 79: 1 - 1 259. SOUTH G.R. & TITTLEY LA. 1 986. Checklist and distributional index

of the benthic algae of the North A tlantic Ocean. Huntsman Marine Laboratory, St Andrews, New Brunswick, Canada, and British Mu­seum (Natural History), London. 76 pp.

STILLER l .W. & WAALAND I .R. 1 993. Molecular analysis reveals cryp­tic diversity in Porphyra (Rhodophyta). Journal of Phycology 29: 506-5 1 7.

SWOFFORD D.L. 1 998. PA UP*. Phylogenetic analysiS using parsimony

(* and other methods), version 4. Sinauer Associates, Sunderland, Massachusetts.

SWOFFORD D.L., OLSEN G.J., WADDELL P.J. & HILLIS D.M. 1 997. Phy­logenetic inference. In: Molecular systematics (Ed. by D.M. Hil l is, e. Moritz & B . K. Mable), pp. 407-5 1 5. S inauer Associates, Sun­derland, Massachusetts.

TAMURA K . & NEI M. 1 993. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Molecular Biology and Evolution 1 0: 5 1 2-526.

TAYLOR W.R. 1 957. Marine algae of the northeastern coast of North

A merica. University of Michigan Press, Ann Arbor. 509 pp. TEASDALE B.W., WEST A.L., TAYLOR H. & KLEIN A.S. 2002. A simple

restriction fragment length polymorphism (RFLP) assay to discrim­inate common Porphyra (Bangiophyceae, Rhodophyta) taxa from the northwest Atlantic. Joumal of Applied Phycology 1 4 : 293-298.

VILLA LARD-BoHNSACK M. 1 995. Illustrated key fO the seaweeds of

New England. Rhode Island Natural History Survey, Kingston, Rhode Island. 1 46 pp.

V ILLALARD-BoHNSACK M. 1 998. Non-indigenous benthic algal species introduced to the northeastern coast of North America. In: NEAS

keys to the benthic marine algae of the northeastern coast of North

A merica from Long Island Sound to the Strait of Belle Isle (Ed. by J. Sears), pp. 1 30-1 3 1 . Contribution No. I . Northeast Algal Society, Dartmouth, Massachusetts.

WILKES R.I. , Y ARISH C. & MITMAN G.G. 1 999. Observations on the chromosome numbers of Porphyra (Bangiales, Rhodophyta) popu­lations from Long Island Sound to the Canadian Maritimes. A lgae

14 : 2 1 9-222. Wu D.Y., UGOZZLI L., PAL B .K & WALLACE R.B. 1 989. Allele-spe­

cific enzymatic amplification of beta globin genomic DNA for di­agnosis of sickle cell anemja. Proceedings of the National Academy

of Science of the United States of A merica 86: 2757-2760. YAMAZAKl S. , KlTADE Y., MARUYAMA T. & SAGA N. 1 996. Phyloge­

netic position of Porphyra yezoensis based on the 1 8S rDNA se­quence. Journal of Marine Biotechnology 4: 230-232.

YARISH e., WILKES R., CHOPIN T, FEI X.G., MATHIESON A.e., KLEIN

A.S., NEEFUS e.D., MITMAN G.G. & LEVINE 1. 1 998. Domestication of indigenous Porphyra (nori) species for commercial cultivation in northeast America. World Aquaculture 29: 26-30.

YARlSH C. , CHOPIN T., WILKES R., MATHIESON A.e., FEl X.G. & Lu S. 1 999. Domestication of nori for northeast America: the Asian experience. Bulletin of the Aquaculture A ssociation of Canada 99: 1 1 - 1 7 .

YOSHIDA T . 1 997. The history and future prospects o f systematics of Bangiaceae, Rhodophyta. Natural History Research, Special Issue 3: 1 -4.

YOSHIDA T, NOTOYA M., KIKUCHI N. & M IYATA M. 1 997. Catalogue of species of Porphyra in the world, with special reference to the type locality and bibl iography. Natural History Research, Special Issue 3: 5- 1 8.

Accepted 10 October 2002