Gracilaria species (Gracilariaceae, Rhodophyta) from southeastern

Phycological Research

2002;

50:

295–311

Gracilaria

species (Gracilariaceae, Rhodophyta) from southeastern Australia, including a new species,

Gracilaria perplexa

sp. nov.: Morphology, molecular relationships and agar content

Kellie Byrne,

1

Giuseppe C. Zuccarello,

2

John West,

1

* Ming-Long Liao

1

and Gerald T. Kraft

1

1

School of Botany, University of Melbourne, Victoria 3010, Australia and

2

National Herbarium Nederlands, University of Leiden, PO Box 9514, 2300 RA Leiden, The Netherlands

SUMMARY

Select species of the agarophyte

Gracilaria

werestudied from southeastern Australia. The morphologyand anatomy of species is described and molecularrelations are inferred based on plastid and mitochon-drial DNA sequence data. Agar yields and qualities aredetermined for each species.

Gracilaria chilensis

,found in Tasmania and Victoria, is morphologically andmolecularly similar to

G. chilensis

from New Zealandand Chile and has low agar yields of 11–16%.

Gracilariacliftonii

from Victoria, has high crude agar yield (52%)and is molecularly uniform.

Gracilaria perplexa

sp.nov., known only from Botany Bay, New South Wales,has an agar yield of 39%. The agar of

G. perplexa

isunusual in requiring the addition of 0.1 mol L

–1

NaClfor alcohol precipitation and is cold-water (25

°

C)soluble because of the very high sulfate ester content.Molecular phylogeny shows that

G. perplexa

is closelyrelated to

Gracilaria preissiana

from western Australia,but differs from the latter in its reduced branching andnarrower more terete axes

.

Key words: agar content,

cox

2–3 spacer,

Gracilaria

,

Gracilaria perplexa

, Gracilariaceae, RuBisCo spacer.

INTRODUCTION

Gracilaria

is the third largest genus in the Rhodophyta,with over 150 described species. It is cosmopolitan,with the greatest standing stocks and numbers ofspecies occurring in temperate seas.

Gracilaria

speciesare commercially important in the production of thephycocolloid agar (Armisen 1995). Many recent studieshave thus been directed toward resolving their taxon-omy and suitability for commercial exploitation. Thefull potential for utilization of

Gracilaria

is unlikely tobe realized without the establishment of sound taxo-nomic criteria for identifying species (Bird 1995).Progress in this activity over the past 40 years hasbeen slow and has mostly come from comparative

morphological studies of Chang and Xia (1963),Yamamoto (1978), Edelstein

et al

. (1978), Bird andMcLachlan (1982), Fredericq and Hommersand(1989a,b), Abbott

et al

. (1991) and Withell

et al

.(1994). These investigations have established theimportance of spermatangial structures, gonimoblastplacentation, traversing-filament orientation and carpo-sporophyte anatomy in delineating species and related-ness. At the same time, major monographic surveys ofthe

Gracilaria

floras of Japan (Ohmi 1958; Yamamoto1978), China (Zhang and Xia 1992), Asian countries(Trono

et al

. 1983; Dinh 1992), New Zealand (Nelson1987), Australia (Withell

et al

. 1994; Womersley1996), and Atlantic North America (Fredericq andNorris 1992) were conducted, which further clarifiedspecies concepts and biogeographic distribution.

The first extensive anatomical and taxonomic studyof the genus

Gracilaria

in Australia was made by May(1948), who dealt mostly with terete species anddescribed nine species, 24 forms, one variety, and five

incertae sedis

taxa. She separated the species into twomain groups (

Macrocystideae

and

Microcystideae

) basedon thallus transverse sections.

Macrocystideae

have anabrupt transition between peripheral and medullarycells and generally have dense granular cell contents,whereas

microcystideae

have no sharp boundarybetween the two tissue types, the cells becominggradually smaller toward the periphery. May (1948)disregarded reproductive structures entirely, relyingexclusively on vegetative and habit features to distin-guish taxa. She regarded cystocarp shape and size,thallus color and size and cell-wall thicknesses as notparticularly valuable for the delineation of the speciesshe studied. It was not until 46 years later thatcomparable broad-scale studies on Australian

Gracilaria

were again attempted (Withell

et al

. 1994; Womersley1996).

*To whom correspondence should be addressed.Email: [email protected] editor: T. Motomura.Received 11 May 2002; accepted 23 August 2002.

296

K. Byrne

et al.

Withell

et al

. (1994) emphasized spermatangialmorphology, carposporophyte anatomy, and tetras-porangial configurations, as well as vegetative andhabit features, in a detailed treatment of nine of the21 Australian species that they recognized. Theydescribed two new species, re-named another, andnewly recorded a further four species for Australia. LikeMay (1948), Withell

et al

. (1994) found the type oftransition in cell size from cortex to medulla (whetherabrupt or gradual) to be consistent within species.Characteristics such as habit and cystocarpic featuresvaried considerably in some species, but were highlyuniform in others.

The identification of slender, terete and irregularlybranched

Gracilaria

species has been a constantproblem for taxonomists, particularly as many popula-tions are only rarely reproductive in the field. Labora-tory culture to obtain missing life history stages,particularly the male gametophytes that so often eludedetection even in well-known species, is one approachto complete anatomical characterization. Another prom-ising approach is the comparative analysis of nucle-otide sequences of selected genomic DNA fragmentssuch as the small subunit of ribosomal RNS (18SrDNA) (Bird

et al

. 1992, 1994), the ribosomal internaltranscribed spacer (Goff

et al

. 1994) and the rbcL-rbcS intergenic spacer (RuBisCo spacer) (Destombeand Douglas 1991; Goff

et al

. 1994). Another promis-ing genetic region is the mitochondria encoded inter-genic spacer, and parts of flanking genic regions,between the cytochrome oxidase 2 and cytochromeoxidase 3 genes (

cox

2–3 spacer) (Zuccarello

et al

.1999a). These molecular tools, when combined withanatomical data, are capable of resolving taxonomicproblems at virtually every taxonomic level from strainsand populations to classes and phyla. The presentstudy has aimed to analyze samples of

Gracilaria

fromsoutheastern Australia taxonomically and molecularly,and to characterize their agars, in order to increase ourunderstanding of this potentially commercially impor-tant genus in Australia.

MATERIALS AND METHODS

Anatomical methods

Anatomical observations were made on a previouslyunnamed species from the Sydney region (designated

Gracilaria perplexa

sp. nov.),


Bird,McLachlan et Oliveira (Marion Bay, Glenelg River,Phillip Island, Hobson’s Bay) and

Gracilaria cliftonii

Withell, Millar et Kraft (Portland, Portsea). Theanatomy of

G. cliftonii

is detailed extensively by Withell

et al

. (1994) and Womersley (1996; as

Gracilariaramulosa

J. Agardh), and as the material of the presentstudy matches these descriptions very precisely, itsmorphology is only briefly considered.

Fresh, dried and formalin-fixed tissues were hand-sectioned, stained in 1% acidified aniline-blue solutionand mounted on glass slides in 20% Karo syrup.Anatomical features were photographed using a ZeissC35 camera and a Zeiss MC 100 camera on a ZeissGFL bright field microscope. Low-power habit photo-graphs were taken using a Nikon F801 Camera on aWild dissecting microscope.

Culture methods

Unialgal culture methods are described in West andZuccarello (1999). Culture isolates were all maintainedat 20–26

°

C, 12 : 12 LD daily photoperiods, 10–15

µ

molphotons m

–2

s

–1

coolwhite fluorescent lighting, MPM/2culture medium (30% salinity; McBride and West1999). For faster growth and reproduction, cultureswere placed under 20–25

µ

mol photons m

–2

s

–1

on arotary shaker (70 r.p.m).

Molecular phylogeny methods

Samples used for molecular analysis are listed inTable 1. Field samples were cleaned of surface epi-phytes or debris, and portions of each specimenanalyzed were pressed as a voucher.

DNA extractions followed the Chelex protocol of Goffand Moon (1993) as outlined in Zuccarello

et al

.(1999b). A fresh or dried thallus tip of approximately3–5 mm was used.

RuBisCo spacer amplification followed the proce-dures in Zuccarello

et al

. (1999b) using the rbcF1forward and rbcR

2

M

2

reverse primers.

Cox

2–3 spaceramplification followed the procedures in Zuccarello

et al

. (1999a). The internal transcribed spacer 1 regionwas amplified as outlined in Zuccarello

et al

. (1999c).Polymerase chain reaction (PCR) products werechecked by electrophoresing in a 1–2% agarose gel,stained in ethidium bromide, visualized under UV lightand photographed.

The PCR products were prepared for sequencing bypolyethylene glycol (PEG) precipitation. A total of45

µ

L of crude PCR product was added to 45

µ

L of 2XPEG precipitation mix (26.7% PEG 8000, 0.6 mol L

–1

Na-acetate (pH 5.2) and 6.5 mmol L

–1

MgCl

2

), brieflyvortexed and left at room temperature for 10 min. Theprecipitated DNA was pelleted by centrifugation for15 min at 9447 g. The supernatant was gentlyremoved by pipette, leaving the DNA pellet in the tube.The DNA was washed twice with 1 mL of 100%ethanol and centrifuged for 15 min after each wash.The residual ethanol was then removed and the DNApellet dried at room temperature. The pellet was thenresuspended in 20–40

µ

L of H

2

O.Automated sequencing was performed on an ABI Prism

377 DNA Sequencer (Perkin-Elmer) after cycle sequenc-ing of the purified PCR product with dye-labeled

Gracilaria perplexa

sp. nov.

297

dideoxynucleotides. Cycle sequencing was performedon a P-E 2400 Thermal Cycler (Perkin Elmer), andsamples cleaned up by the ethanol precipitationmethod following the manufacturer’s recommenda-tions. Sequences were assembled using the computersoftware supplied with the sequencer. All sequenceswere aligned visually in

SEQPUP

(free software distrib-uted by D. G. Gilbert, Biology Department, IndianaUniversity, Bloomington, IN, USA).

Phylogenetic relationships were inferred with

PAUP

*4.0b6-8 (Swofford 2001).

Ptilophora prolifera

was used as an outgroup. Maximum parsimony (MP)trees were constructed using the heuristic searchoption, 500 random sequence additions, tree bisection-reconnection (TBR) branch swapping, unorderedunweighted characters, gaps treated as missing data.The program

MODELTEST

version 3.06 (Posada andCrandall 1998) was used to find the model ofsequence evolution that best fitted each data set by ahierarchical likelihood ratio test (

α

= 0.05) (Posadaand Crandall 2001). When the best sequence evolutionmodel had been determined, maximum likelihood (ML)and distance searches were performed in

PAUP

* usingthe estimated parameters (substitution model, gammadistribution, proportion of invariable sites, and transition/

transversion ratio). Distance trees were constructedusing neighbor-joining reconstruction (NJ). Maximumlikelihood was also used to construct the most likelytree from the data set (5 random additions). Supportfor individual internal branches was determined bybootstrap analysis (Felsenstein 1985), as implementedin

PAUP

*, and a decay index (Bremer 1988). Forbootstrap analysis, 1000 bootstrap data sets weregenerated from resampled data (5 random sequenceadditions), for both the MP and NJ analysis. Decayindices, on a strict consensus of the most parsimonioustrees, were calculated with

AUTODECAY

version 4.0.2(Eriksson 1998).

Polysaccharide methods

Both field and cultured materials were used for poly-saccharide analyses. The field material was cleaned byhand to remove epiphytes or debris. The cleanedmaterial was then dried

in vacuo

for 24 h at 60

°

C. Thedried material was then ground into powder usingliquid nitrogen and a mortar and pestle.

The powdered specimen was hydrated overnight indistilled water (

ca

1 g in 100 mL). The seaweedsuspension was then heated in a water bath for 1 h at

Table 1.

Samples used for phylogenetic study

Species Location Culture/sample no.

RuBisCoGenBank no.

cox

2–3GenBank no.

ITS GenBank no.

Gracilaria sp.

1 (‘

verrucosa

’)

1

Dumaguete, Philippines 3448 AY131298 AY131322


Marion Bay, Tasmania 3948 AY131299 AY131316 AY131297


Glenelg River, Victoria 3960 AY131300 AY131315 AY131296


Phillip Island, Victoria 3967 AY131301 AY131314 AY131295Gracilaria chilensis 2 Auckland, New Zealand NA U21356 AF034265Gracilaria cliftonii Portland, Victoria 3957 AY131302 AY131321Gracilaria cliftonii Portsea, Victoria 3984 AY131303 AY131319Gracilaria cliftonii Phillip Island, Victoria E197 AY131304 AY131320Gracilaria eucheumatoides 3 Lombok, Indonesia field AY131305 AY131318Gracilaria pacifica 4 San Juan Island, Washington NA U21353Gracilaria perplexa Bare Island, Botany Bay, NSW 3903 AY131306 AY131325Gracilaria perplexa Bare Island, Botany Bay, NSW E252 AY131307 AY131326Gracilaria perplexa Kurnell, Botany Bay, NSW E199 AY131308 AY131327Gracilaria preissiana 5 Carnac Island, Western Australia field AY131309 AY131324Gracilaria robusta 6 Monterey, California NA U21355Gracilaria secundata Kurnell, Botany Bay, NSW field AY131311 AY131323Gracilaria tenuistipitata 7 Hainan Island, China 3404 AY131310 AY131317Gracilaria tikvahiae 8 Nova Scotia, Canada NA U21357Gracilaria ‘verrucosa’ 9 Norway NA U21358Gracilariopsis sp.10 Qingdao, China NA U21350Gracilariopsis tenuifrons 11 Venezuela NA U21351Ptilophora prolifera 12 Garden Island, Western Australia NA AY131312 AY131313

ITS, internal transcribed spacer. NA, not applicable, sequences retrieved from GenBank. The 4-digit numbers refer to J. A. Westculture numbers. Sample numbers beginning with E indicate DNA extractions of field-collected material.

Authorities for species listed above. Gracilaria chilensis Bird, McLachlan et Oliveira, G. cliftonii Withell, Millar et Kraft, G. eucheuma-toides Harvey, G. robusta Setchell, G. secundata Harvey, G. tenuistipitata Chang et Xia, G. tikvahiae McLachlan, G. verrucosa (Hudson)Papenfuss, Gracilariopsis tenuifrons (Bird et Oliveira) Fredericq et Hommersand, Ptilophora prolifera (Harvey) J. Agardh.

Details and references for species listed above. 1Collected by H. P. Calumpong, Silliman University, Dumaguete, Negros Oriental, ThePhilippines, 28.v.1994. 2Culture from M. Pederson, see Goff et al. (1994). 3Collected by J. Huisman, iv.1999. 4See Goff et al. (1994).5Collected by J. Huisman, iv.1999. 6See Goff et al. (1994). 7Culture from M. Pederson, see Goff et al. (1994). 8See Goff et al. (1994).9See Goff et al. (1994). 10Culture from M. Pederson, see Goff et al. (1994). 11Culture from D. Kapraun, see Goff et al. (1994). 12Collectedby J. Huisman, 1999.

298 K. Byrne et al.

95–100°C, with occasional stirring. The extractionmixture was centrifuged to separate the insoluble algalmaterial from the extracted water-soluble polysaccha-ride. The pellet was re-suspended and the extractionwas repeated twice. The successive polysaccharidesolutions (supernatants) from each sample were com-bined, heated to ca 90°C, and filtered successivelythrough a Whatman no. 541 filter paper, Whatman no.542 filter paper and Whatman GF/A glass microfibrefilter. The filter, funnel and solution were kept hot(>60°C) throughout the filtration process. The refinedpolysaccharide solution was then heated to >60°C,placed on a magnetic stirrer with rapid stirring, andprecipitated by adding 3 vols of hot (>60°C) 2-propa-nol. The precipitate was collected by centrifugation,washed with 60% 2-propanol and left to dehydrate in99.5% 2-propanol. After drying overnight at 60°C in vacuoover P2O5, the yield of crude polysaccharide wasdetermined as a percentage of the dry weight of thealgal sample.

Films for Fourier transform infrared (FTIR) spectro-scopy analysis were prepared by evaporating aqueoussolutions of the polysaccharide (2.5 mL of 4 mg mL–1

solution) at 60°C in plastic dishes (3.2 cm diameter)(Liao et al. 1993). The spectra were recorded on a

Perkin-Elmer Series 2000 FTIR spectrophotometer intransmittance mode (16 scans, collected at a resolu-tion of 4 cm–1).

RESULTS

MorphologyThe following species have been identified from theeight sampling sites studied; unialgal isolates weremade from a single individual from most sampling sites(culture numbers are shown in Table 1):

Gracilaria chilensis Bird, McLachlan et Oliveira (Bird et al. 1986: 2928, figs 2–13; Womersley 1996: 23, fig. 5; Bird et al. 1990: 375, Figs 1–5)Synonym: Gracilaria sordida Nelson (Nelson 1987:90–98, figs 6–9).

Type Locality: Penco, Bahía Conceptíon, Chile(Romo, 1.ix.1982); holotype and isotypes in NRCC,9131 (a) and (b).

Distribution: Chile, from Bahía Herradura to RíoMaullin; New Zealand, North, South, Stewart and

Fig. 1. Map showing distribution

of the southeastern Australian

Gracilaria populations studied.

a. Glenelg River. b. Hobson’s Bay.

c. Bare Island. d. Portland. e. Port-

sea. f. Phillip Island. g. Marion

Bay.

Gracilaria perplexa sp. nov. 299

Chatham Islands; southeast Australia, from Cowell andEyre Peninsula in South Australia to Hobson’s BayVictoria; eastern Tasmania.

Glenelg River, Victoria (38°23′S, 141°35′E) (Fig. 1)Habitat: Plants on rocks at 50-cm depths along thewest bank of the river mouth, north of the blockingsand bar and when measured on two occasions (highand low tide), the salinity was 18‰ and 16‰, respec-tively (G. and R. Kraft, K. Byrne, B. McDonald, 7.iv.99,MELU A042868).

Morphology: Thalli dark brown to black, 20–21 cmhigh, irregularly branched, and generally stringy inappearance, fronds terete and slender throughout,laterals of varying length, acutely pointed and basallyunconstricted. Primary axes are up to 850 µm indiameter, branching to three orders, with some of theultimate laterals hooked (Fig. 2a), numerous primaryaxes arising from a discoid holdfast.

Vegetative Anatomy: Cortex-to-medulla transitiongradual, with cells increasing to 230 µm in diametertowards the center of the frond (Fig. 2d), cortex is twoto three cells thick, outer cells anticlinally elongatedand the inner cells subisodiametric. Surface corticalcells 7–10 µm broad by 12–15 µm in length. A sub-cortex of one or two cell layers overlies a medulla ofseven to nine cells.

Reproduction: Sessile cystocarps scattered radiallyon all branch orders except the lower main axes(Fig. 2b). Cystocarps beaked, shaped like depressedspheres, basally constricted and reaching 850–900 µmin width, traversing cells or filaments descending intocystocarp floor and growing outward into lower pericarp(Fig. 2c,e), pericarp of six to nine cells reaching70–100 µm thick, carposporangia slightly pyriform tosubspherical, reaching 20–25 µm in diameter; malesand tetrasporophytes not observed.

In laboratory culture, isolate 3960 did not reproduce.

Marion Bay, Tasmania (42°49′S, 147°54′E) (Fig. 1)Habitat: Free-floating in estuarine area with salinity of28–35‰ and subject to freshwater flooding. Plantsforming a mat several centimeters thick at a depth of30–50 cm (B. Stevens, 20.ii.1999, MELU A042866,042865).

Morphology: Plants yellow to deep green/black incolor, terete throughout, irregularly branched to threeorders, to 13–14 cm in length. They are exclusivelyfree-floating and thus have no anchoring parts,although plants grown from tetraspores in cultureproduce discoid holdfasts (Fig. 3a,b). Primary branchesoccasionally have long flagelliform extensions.Branches on more sparingly branched thalli are gener-ally stouter, with higher-order laterals occasionallyslightly constricted at the base. Branching becomesdenser distally, with the upper long main branchesproducing numerous laterals of varying lengths with

acute apices. In laboratory culture, plants reach10–12 cm in length with primary axes to 1.4 mm indiameter.

Vegetative Anatomy: Cortex-to-medulla transition isgradual, with cell diameters increasing to 290 µmtowards center of the thallus (Fig. 3c). Cortex is two tothree cells thick, outermost cells being anticlinal andelongated. Cortical cells are 7.5 µm in breadth and15 µm in length, with the innermost cells isodiametric.Subcortical layer is one or two cells thick and sur-rounds a medulla consisting of 9–10 cells.

Reproduction: Field plants only known to reproducevegetatively. Culture isolate 3948 did not reproduce,whereas isolate 3947 formed all reproductive phases.Plants in culture became tetrasporic after about30 days at 25°C, 10–20 µmol photons m–2 s–1, 30‰salinity, the tetraspores germinating into dioeciousgametophytes that bore gametes after about 28 days.Sessile cystocarps depressed spheres with pointedbeaks, basally constricted, reaching 800–1000 µm inwidth (Fig. 3d,f), scattered radially on all branch ordersexcept for lower main axes, traversing cells or filamentsdescending into cystocarp floor and growing outwardlyinto the lower pericarp. Carpospores pyriform to sub-spherical, reaching 20–30 µm in diameter. Sperma-tangial mother cells line bases of relatively shallow pits(Textorii-type), surrounded by clavate cells 18–25 µmhigh and 4–5 µm in diameter (Fig. 3c,e). Tetraspor-angia are decussate-cruciate and scattered in anunmodified outer cortex.

Phillip Island, Victoria (38°28′S, 145°14′E) (Fig. 1)Habitat: Plants attached to shells scattered across amuddy/sandy substratum on a sheltered intertidal flatat Woody Point, Phillip Island (G. Kraft and B. Stevens,19.iv.1999, MELU A042867).

Morphology: Thalli light green and terete through-out, 16–20 cm in length, and stringy in appearance,branching irregular to three orders, mature primaryaxes to 900 µm in diameter, first-order branchesoccasionally having long slender extensions, all lateralswith acute apices and lacking basal constrictions.

Vegetative Anatomy: Cortex-to-medulla transitiongradual with cell diameters increasing to 210 µmtowards the thallus center, cortex two to three cellsthick, having anticlinally elongated outermost cells,cortical cells 7–9 µm in breadth by 12–15 µm inlength, with the innermost cells isodiametric; subcorti-cal layer one to two cells thick surrounding a medullaof 9–10 cells.

Isolate 3967 did not reproduce in culture.

Hobson’s Bay, Victoria (37°50′S, 144°53′E) (Fig. 1)Habitat: Attached to muddy substrata near mangrovesin intertidal zone, Stoney Creek backwash, Hobson’sBay, Spotswood, Melbourne (J. West, 25.iii.1999,MELU A042869).

300 K. Byrne et al.

Morphology: Mature field plants, not reproductive,similar in morphology and length to those from Marion

Bay, Glenelg River and Phillip Island. Isolates of tetra-sporophyte in culture (3954) reproduced as described

Fig. 2. a–e. Gracilaria chilensis from the Glenelg River, Victoria (MELU A042868). a. Frond grown for 6 months in culture from a

3.2-mm excised branch tip. b. Sessile, basally constricted cystocarps releasing carpospores (arrows) from a culture-grown female

gametophyte. c. Lower periphery of cystocarp showing traversing filaments (arrows). d. Cross-section of lower axis showing gradual tran-

sition from cortex to medulla. e. Cross-section of cystocarp showing narrow basal placentation with downwardly penetrating peripheral

traversing filaments (small arrow) and irregularly contoured central sterile carposporophyte tissue. Linear outer gonimoblast filaments

(larger arrows), and terminal chains of ovoid to subspherical carposporangia (arrow heads).


Fig. 3. a–f. Gracilaria chilensis from Marion Bay, Tasmania (MELU A042865, A042866). a. Non-reproductive plant grown for 6 months

in culture from a 3.5-mm long excised branch tip. b. Male gametophyte grown for 6 months in culture from a tetraspore. c. Branch cross-

section of male gametophyte showing Textorii-type spermatangial conceptacles (arrows) and a gradual cortex-to-medulla transition. d. Pyriform

cystocarps on a female gametophyte grown from a tetraspore in culture. e. Detail of Textorii-type male conceptacles showing clavate

paraphyses (arrows) between adjacent clusters of spermatangial mother cells. f. Cross-section of cystocarp showing broad basal placentation

with downwardly penetrating peripheral traversing filaments (arrow head) irregularly contoured small-celled central sterile carposporophyte

tissue, and terminal ovoid carposporangia (arrows).

302 K. Byrne et al.

for Marion Bay isolate. Culture isolate 3954 formed allreproductive phases.

Gracilaria perplexa Byrne et Zuccarello sp. nov., Bare Island, New South Wales (34°00′S, 151°14′E) (Fig. 1)Frondes teretes, ad 15 cm [altae], varie ramosae,usque ad quadruplo subdichotomae, apicibus acutis,interdum arcuatae; axes primarii 1.5–1.7 mm indiam., cortice 30–40 µm lato, e stratis cellularum 2–4constanti, cellulis hyalinis dispersis (basibus trichoma-tum) instructo, abrupte in medullam transienti, cellulismedullosis polygonis, ad 250–5500 µm in diam. Gen-eratio typi Polysiphoniae. Gametophyta dioica; cysto-carpia globosa, rostrata, ad basin leviter constricta, ad1.3 mm in diam., solitaria vel interdum aggregata,filamentis nutriciis ad fundum loculi cystocarpii cres-centibus; carposporangia subsphaerica, ad 25 µm indiam., pericarpio 220–270 µm crasso, e stratis cellu-larum 7–10 constanti; spermatangia in conceptaculiscompositis (typi Polycavernosae) 50–85 µm latis insubcortice vel medulla exteriori disposita; tetraspor-angia cruciatim decussata, 17–26 × 26–29 µm, incortice exteriori dispersa.

Locus Typicus: Bare Island (lat. aust. 33°59′, long.orient. 151°13′, Fig. 1), La Perouse, Botany Bay, NewSouth Wales, Australia.

Holotypus: MELU 042862, K. Byrne and J. Bartlett,22.v.1999. Isotypus: MELU A042863 and A042864, N.Paul, xii.1998.

Distributio: E loco typico et Kurnell, NSW (lat. aust.34°00′, long. orient. 151°12′) tantum cognita.

Terete shoots to 15 cm, branching variable, sub-dichotomous to four orders, acute tips, sometimesarcuate, primary axes 1.5–1.7 mm diam., cortex30–40 µm wide, two to four cell layers, with scatteredhyaline cells (hair cell bases). Transition to medullaabrupt. Medullary cells, polygonal, to 250–320 µmdiam. Polysiphonia-type life history, gametophytes dio-ecious, cystocarps globose, rostrate, slightly con-stricted at base, to 1.3 mm diameter, solitary orsometimes aggregated; traversing filaments directedtoward cystocarp cavity floor, carposporangia subspheri-cal to 25 µm diameter, pericarp, 220–270 µm thickwith 7–10 cell layers; spermatangia borne in com-pound conceptacles 50–85 µm across (Polycavernosa-type) in subcortex/outer medulla; tetrasporangia cruci-ate-decussate, 17–26 × 26–29 µm scattered in outercortex.

Type Locality: Bare Island (33°59′S, 151°13′E, Fig. 1),La Perouse, Botany Bay, New South Wales, Australia.

Holotype: MELU 042862, K. Byrne and J. Bartlett,22.v.1999. Isotype: MELU A042863 and A042864, N.Paul, xii.1998.

Distribution: Known only from type locality andKurnell, New South Wales (34°00′S, 151°12′E).

Etymology: The species designation ‘perplexa’ indi-cates the frequent difficulties in resolving the charac-teristics of many Gracilaria species. Molecular analysishas proven essential as evidence of the differencebetween many species.

Morphology: Mature thalli burgundy to light green,cartilaginous, cylindrical to subcylindrical, reaching15 cm in length (Fig. 4a), fronds extremely variablewith subdichotomous branches to four orders, basallyconstricted. Primary axes to 1.5–1.7 mm in diameter.On some plants, branching arcuate and apically atten-uated, becoming more dense distally. Attachment by alacerate holdfast.

Vegetative Anatomy: Cortex-to-medulla transition abruptwith cell diameters reaching 250-(400)–550 µmtowards thallus center (Fig. 4f). Cortex surface withscattered hyaline cells of undetermined function (pos-sibly basal remnants of hair cells), two to four cellsthick and 30–40 µm wide. Cortical cells 10–20 µm indiameter, quadrate to spherical in outline. Medullareaches 1500–1600 µm in diameter.

Reproduction: Gametophytes dioecious, cystocarpssolitary or, less commonly, aggregated radially on allbut the lower main axes. Cystocarps to 1.3 mm indiameter, globoid, rostrate, and slightly constricted atthe base (Fig. 4b). Traversing cells or filaments aredirected exclusively towards cystocarp cavity floor(Fig. 4d,e), sterile gonimoblast cells subisodiametricwith relatively thick walls, carposporangia subspheri-cal, to 25 µm in diameter, pericarp 220–270 µm wideand 7–10 cell layers thick; spermatangia produced byspermatangial mother cells, which form a branchsystem covering the entire inner surface of the com-pound (>4 cavities) conceptacles (Polycavernosa-type),50–85 µm across within the subcortex and outermedulla (Fig. 4c); Tetrasporangia cruciate-decussate,17–26 × 26–29 µm, scattered in outer cortex (Fig. 4g).

Remarks: Entity 3903 was isolated from Bare Island,New South Wales on 6.xii.1998. The Polysiphonia-typesexual life-history was completed in culture. A secondcollection from Bare Island was made on 29.iii.2000.It constitutes a distinct new species of Gracilaria in theAustralia flora. The large subunit of ribulose bisphos-phate carboxylase was sequenced for this species(culture isolate 3903, GenBank Accession no.AY131306) and, when added to large rbcL data, it isshown to be closely related to Gracilaria preissiana(Sonder) Womersley from Western Australia (Hommer-sand, M., 2002 personal communication). We com-pared the morphological characters of G. preissianawith G. perplexa (Table 2). Gracilaria preissiana haswider and often flattened branches and is more fre-quently branched than G. perplexa.


Fig. 4. a–g. Gracilaria perplexa Byrne et Zuccarello, sp. nov. from Bare Island, Sydney (MELU 042862). a. Field-collected female gametophyte

showing lacerate holdfast (arrow). b. Basally constricted cystocarps on a female gametophyte grown in culture. c. Cortex of a male gametophyte

showing the many-chambered Polycavernosa-type spermatangial conceptacles. d. Sterile gonimoblast tissue and descending traversing filaments

(arrowheads). e. Cross-section of cystocarp showing descending traversing filaments (arrowheads) and anticlinally elongated basal sterile tissue

(star). f. Cross-section of thallus showing gradual cortex to medulla transition. g. Cruciate tetrasporangia (arrow) embedded in the cortex.

304 K. Byrne et al.

Gracilaria cliftonii Withell, Millar et KraftGracilaria ramulosa J Agardh (Agardh 1876: 471;Womersley 1996: 18, pl. 1, figs 1,3; De Toni 1900:439; Sonder 1881: 22).

NON G. ramulosa (Martius) Greville in Saint-Hilaire(1833): 457.

Type Locality: Swan River at Fremantle.Lectotype: TCD, Herb. Harvey Alg. Aust Exsicc. no.

323-a.Distribution: Cottesloe, Western Australia, to Walk-

erville, Victoria, and around Tasmania (Womersley1996).

Portland (38°20′S, 141°36′E) (Fig. 1)Material Examined: Plants at 2-m depths attached tostones and shells debris in sandy substrata in a densepopulation in the intertidal area of Portland (K. Byrneand B. McDonald, 18.iv.1999).

Morphology: Thalli burgundy to dark red, 20–25 cmhigh, cartilaginous and terete throughout (Fig. 5a).Branching subdichotomously to five orders, withprimary axes (to ca 2 mm in diameter), laterals withattenuate to acute apices. Attachment is by an encrust-ing holdfast.

Vegetative Anatomy: Cortex-to-medulla transitiongradual, with cells increasing to 250 µm in diametertowards frond center (Fig. 5b). Cortex two to four cellsthick, outer cells anticlinally elongated and inner cellssubisodiametric. Surface cortical cells are up to12.5 µm broad and 12.5 µm in length.

Reproduction: Cystocarps are sessile and scatteredradially on all branch orders except for the lower mainaxes (Fig. 5e). Cystocarps are strongly beaked, globose,basally constricted, and reach 1700 µm in width(Fig. 5c). Traversing cells or filaments both descendinto the cystocarp floor and radiate into the mid tolower pericarp (Fig. 5d). Pericarp is 12–14 cells and

Table 2. Morphological comparison of Gracilaria preissiana and Gracilaria perplexa

Gracilaria preissiana† Gracilaria perplexa

Thallus Erect to 24 cm. Lower axes and major branches are subterete or compressed. Upper branches are terete or subterete with apices acute. Branching irregular, dichotomously divide, many short branchlets and short, undivided lateral arising from blade margins. Obtuse apices, all blades linear to slightly flagellate

Erect to 13 cm, subcylindrical to terete throughout. Branching irregular with subdichotomous acute apices, widely spaced to three or four orders throughout. Branches occasionally basally constricted. Thallus is cartilaginous. Both discoid and lacerate holdfasts have been observed

Stipe width and length 2 cm long (0.5) 0.1–2.2 (2.7) mm diam. 1 cm longLower branches 2.5–4.0 (5.8) mm wide and 0.3–0.5 (0.6) mm thick (1.0) 1.2–1.5 (2.0) mm wideLower axes 1.7–2.3 (2.6) mm wide and 1.3–1.9 mm thick 1.5–1.7 mm diam.Major branches 1.4–1.8 (2.1) mm wide and 1.3–1.6 mm thickCortex–medulla transition Abrupt AbruptCortex 1–2 cells thick (1) 2–4 cells thick and (25) 30–37 (40) µm wideCortex cell shape Hemispherical to ovoid IsodiametricCortex cell size (7) 9–17 µm long by (6) 8–14 (16) µm diam. 10–20 µm diam.Medulla 6–9 cells broad (900) 1500–1600 µm in widthMedulla cell shape Polygonal-subspherical Polygonal-subsphericalMedulla cell size (65) 90–180 (240) µm diam. (95) 250–320 (550) µm diam.Hair and basal cells Abundant except in pressed specimens PresentGametophytes Dioecious, occasionally bisexual DioeciousMale Spermatangia globose to obovoid, 2–4 µm,

Polycavernosa-type conceptacles in cortex and medulla, 100–250 µm deep by 115–200 µm wide

Spermatangia globose 4–5 µm, Polycavernosa-typeconceptacles in subcortex to outer medullary regions, 50–85 µm across, consisting of >4 cavities

Female cystocarps Cystocarps, solitary ostiolate Cystocarps solitary, globoid and slightly rostrateTraversing filaments Common to abundant, from gonimoblast cells

penetrating deep into gametophyte tissue basal to carposporophyte, and lower third of pericarp

Exclusively directed towards floor of cystocarp cavity

Cystocarp size (0.5) 1.0–1.6 mm high by (0.7) 1.6–2.0 mm wide To 1.3 mm wideCystocarp base Sometimes constricted at base Slightly constricted at baseCystocarp position Occurring on upper third of thallus Occurring on secondary branches and abovePericarp tissue 240–360 µm wide and 11–17 cells thick 220–270 µm wide and 7–10 cell layers thickCarpospore shape and size Obovoid, globose or oval, 26–38 µm by 20–27 µm Subspherical, up to 25 µm diam.Carpospore origin Borne terminally on gonimoblast filaments Borne terminally on gonimoblast filamentsTetrasporangia Cruciately divided, 28–40 µm long by 25–30 µm wide,

in outer cortex, mainly in laterals and smaller branchesCruciately decussately divided, 26–29 µm long by 17–26 µm wide, scattered in outer cortex, borne on all branches

†From Withell et al. (1994).


250–300 µm thick. Carposporangia are ovoid, reach-ing 37–40 µm in diameter. Males and tetrasporophyteswere not observed.

Portsea (38°19′S,144°43′E) (Fig. 1)Material Examined: Plants at ca 2-m depth attached toshells and rocks, scattered and arising individually

under the Portsea pier. (K. Byrne and N. Watt,25.iv.1999, MELU A042861, A042860, A042872).

Morphology, Vegetative Anatomy and Reproduction:As described for the Portland specimens.

Remarks: The material studied conforms to recentdescriptions of G. cliftonii by Withell et al. (1994) andWomersley (1996). Neither isolate (3957, 3984) grewsatisfactorily in laboratory culture.

Fig. 5. a–e. Gracilaria cliftonii from Portland, Victoria (MELU A042860, A042861, A042872). a. Field-collected sterile

thallus. b. Cross-section of thallus showing extremely gradual cortex-to-medulla transition. c. Rostrate cystocarp on a field-collected

thallus. d. Traversing filaments (arrows) entering the pericarp from peripheral gonimoblast tissue. e. Cross-section of mature cystocarp.

306 K. Byrne et al.

Phylogenetic analysisAll sequence data has been deposited in GenBank(Table 1).

The plastid-encoded RuBisCo spacer produced anunambiguously aligned data set of 317 characters with83 potentially parsimony-informative characters. Allphylogenetic reconstruction methods (MP, NJ and ML)produced similar tree topologies. Maximum parsimonyproduced seven equally most parsimonious trees of285 steps (Consistency index (CI) = 0.6807). Thetopology of the ML tree (–ln likelihood score =1650.736) is shown in Fig. 6.

The three samples of G. chilensis from Marion Bay,Phillip Island and Glenelg River, and the GenBanksequence of G. chilensis from New Zealand, formed awell-supported group with Gracilaria tenuistipitata. Therelationship between these three G. chilensis samplesand the G. chilensis from New Zealand is not resolvedin the most parsimonious tree (only 2 bp differencebetween all samples). Gracilaria tenuistipitata differsby 7–9 bp from the G. chilensis samples. The first

ribosomal intergenic spacer reveals some base pairvariation between these G. chilensis samples (Table 3),although no pattern is obvious.

The three samples of G. cliftonii form a well-supported group, with a possible sister relationship,although without bootstrap support, to Gracilariaeucheumatoides and Gracilaria secundata.

Gracilaria perplexa samples formed a well-supportedgroup, distinct from the other Australian species, withG. preissiana as its sister species.

Cox2–3 spacer regionThe mitochondrial-encoded cox2–3 spacer producedan unambiguously aligned data set of 362 characterswith 107 potentially parsimony-informative characters.All phylogenetic reconstruction methods (MP, NJ andML) produced similar tree topologies. Maximum parsi-mony produced 24 equally most parsimonious trees of307 steps (CI = 0.7655). The topology of the ML tree(–ln likelihood score = 1682.763) is shown in Fig. 7.

The relationships inferred for the cox2–3 spacerregion shows similar tree topology to the RuBisCo tree,although there are fewer taxa upon which to basephylogenetic inferences since there are no cox2–3spacer regions for Gracilaria species in GenBank. Someof the species relationships at the termini of thebranches are more resolved or better supported due tothe known higher mutation rate of this spacer com-pared with the RuBisCo spacer (Zuccarello and West2002). Although the cox2–3 sequence has not beendetermined for the New Zealand G. chilensis, the treesstrongly support the clade of the three Australiansamples, as previously shown for the RuBisCo tree.Similarly, G. tenuistipitata forms a close relationship tothe G. chilensis samples, in this case a well-supportedsister relationship.

The three samples of G. cliftonii form a stronglysupported clade. Gracilaria secundata Harvey andGracilaria sp.1. (‘verrucosa’-like) form a sister groupwith these samples. The sister group relationship withthe Philippine species might be expected from the‘verrucosa’-type males of both species (Withell et al.1994; fig. 5h).

Fig. 6. Maximum likelihood topology of inferred relationships

between Gracilaria, Gracilariopsis samples using RuBisCo spacer

DNA sequence data (–ln likelihood = 1640.7363). Ptilophora

prolifera was used as an outgroup. Numbers above branch-line left

of diagonal, parsimony based bootstrap values, in percentages;

above branch-line right of diagonal, distance based bootstrap

values; below branch-line, decay values.

Table 3. Distance matrix of 412 aligned bases of the internal

transcribed spacer 1 for samples of Gracilaria chilensis

Sample 1 2 3 4

1. 3948 – 0.015 0.005 0.0152. 3967 6 – 0.020 0.0273. 3960 2 8 – 0.0204. NZ GenBank 6 11 8 –

Absolute distances (in base pairs) below diagonal; Kimura-2-parameter distances above diagonal. See Table 1 for sampledetails.


Again, G. perplexa forms a separate well-supportedclade containing G. preissiana as a well-supportedsister species. Samples of G. perplexa were collectedon three occasions at Bare Island (10.xii.1998,22.vi.1999 and 10.iii.2000) and at Kurnell (29.iii.2000);all samples had identical cox2–3 spacer sequences.

Polysaccharides

Physical propertiesThe polysaccharide of G. perplexa differed from that ofthe other taxa analyzed in requiring the addition of0.1 mol L–1 NaCl to the extract for alcohol precipitationand being soluble in H2O at 25°C. The polysaccharidepreparations of the other samples showed the charac-teristic agar property of being insoluble in H2O at 40°C.In 0.5% (w/v) solution, their gelation was ion-independent,of good clarity and thermoreversible. The crude poly-saccharide yields of samples from five Gracilariapopulations differ substantially. Gracilaria cliftonii pro-duced an exceptionally high yield (52%) of crude

polysaccharide, exceeding that of G. perplexa (39%).Samples of G. chilensis from Marion Bay, Phillip Islandand Glenelg River all produced low crude polysaccha-ride yields of between 11 and 16%.

Fourier transform infrared spectroscopyThe FTIR spectra of polysaccharide preparations fromall Gracilaria populations were recorded. The diagnos-tic regions (1400–700 cm–1) for three preparations areshown in Fig. 8 for comparison. For all preparations,except that of G. perplexa, the FTIR patterns obtainedwere characteristic of agars, showing essentially noabsorption peaks over the range 850–800 cm–1 that areindicative of various equatorial and axial sulfate esters atsignificant levels. The spectra also consistently show avery weak band of absorption at ca 1250 cm–1, reflect-ing very low levels of total sulfate ester (Stancioff andStanley 1969). Strong bands at 932.75 and 892.05 cm–1,characteristic of agar biopolymers, demonstrate thepresence of 3,6-anhydrogalctosyl and unsulfated,3-linked galactopyranosyl residues, respectively (Whyteet al. 1985).

The preparation from G. perplexa exhibits an intenseabsorption band at 850 cm–1 (Fig. 8), denoting that thepolysaccharide is heavily sulfated at the C(O)4 positionof the 3-linked galactosyl residues (Stancioff andStanley 1969). In addition, the much higher sulfatecontent of the preparation from G. perplexa is reflectedin the broad band at ca 1251.91 cm–1.

Fig. 7. Maximum likelihood topology of inferred relationships

between Gracilaria samples using cox2–3 spacer DNA sequence

data (–ln likelihood = 1682.7630). Ptilophora prolifera was used

as an outgroup. Numbers above branch-line left of diagonal, par-

simony based bootstrap values, in percentages; above branch-line

right of diagonal, distance based bootstrap values; below branch-

line, decay values.

Fig. 8. Fourier transform infrared spectra of polysaccharide

preparations from Gracilaria chilensis, Gracilaria cliftonii and

Gracilaria perplexa. Bands at 933 and 892 cm–1 demonstrate 3,6-

anhydrogalctosyl and unsulfated 3-linked galactopyranosyl residues,

respectively. For G. perplexa the broad band at ca 1250 cm–1

indicates a high sulfate content. An intense absorption band at

850 cm–1 denotes heavy sulfation at the C(O)4 position of the

3-linked galactosyl residues.

308 K. Byrne et al.

DISCUSSIONThe characterization of Gracilaria species is largelybased on gross plant morphology and the developmen-tal morphologies of vegetative and reproductive struc-tures (Withell et al. 1994; Womersley 1996). In agenus as large and morphologically plastic as Gracilariaobviously is, it would seem inevitable that the relativelyfew consistent features that classically distinguishspecies are going to show much overlap and ambiguity,particularly in cases where field populations areseldom if ever represented by all life history stages.This is particularly true of the filiform types that makeup the bulk of dominant Gracilaria species in virtuallyevery world region. Withell et al. (1994) focusedmainly on eastern and western Australian members ofthe genus, on a number of particularly morphologicallydistinctive taxa, and paid relatively little attention tothe much more widespread terete forms, of which some15 are named in their appendix of currently acceptedAustralian Gracilaria species. The research reportedhere has concentrated on these ambiguous types thatmake up the Gracilaria populations indigenous toVictoria, Tasmania and parts of New South Wales.

To break morphological ambiguity, plasticity andincomplete expression of life history phases that sooften characterize field populations of Gracilaria,molecular, polysaccharide and laboratory culture tech-niques were jointly applied. The RuBisCo spacer andcox2–3 spacer have been applied by other workers toestablish phylogenetic relatedness within several redalgal genera (Zuccarello and West 1997; Zuccarelloet al. 1999c), including Gracilaria (Goff et al. 1994).

Gracilaria chilensisWomersley (1996), in a monograph of southern Aus-tralian Gracilaria species, attributes several estuarinepopulations from the southeastern part of the continentto G. chilensis, a species recently described from Chileby Bird et al. (1986). The original definition of thisspecies relied heavily on the type of spermatangialconfiguration, the shallowly sunken Textorii-type. Justafter the Bird et al. (1986) paper appeared, Nelson(1987) described the new species, G. sordida, basedon New Zealand populations of what had previouslyalso been identified as G. verrucosa. Comparison ofmorphology, reproductive anatomy, chromosome num-bers, organellar DNA restriction profiles and inter-fertility studies led Bird et al. (1990) to regardG. sordida and G. chilensis as conspecific, the lattername having nomenclatural priority. Goff et al. (1994)also found no sequence variation in the nuclear inter-nal transcribed spacer region between Chilean andNew Zealand samples of this species. Our data indicatethat the Australian samples, from Hobsons Bay, PhillipIsland, Glenelg River and Marion Bay, belong to the

same species as the New Zealand sample, thus sup-porting Womersley’s identification of Australian popu-lations as this species (Womersley 1996). Additionalsupport is provided by the laboratory culture of tetra-sporophytes from Hobsons Bay (3954) and Marion Bay(3947). Tetraspores developed into gametophytes andthe males had the typical Textorii-type spermatangialstructures (Fig. 3c,f) distinctive of G. chilensis andwhich distinguish it from a number of similar-appearingtaxa, including several Australian representativesdesignated by May (1948) as various forms ofG. confervoides (= G. verrucosa). We conclude fromthese results that G. chilensis in Australia is character-istic of habitats in which large salinity fluctuations arethe norm, such as the estuary of the Glenelg River, thesloughs of Marion Bay, and the intertidal mudflats ofwave-sheltered northern Phillip Island and HobsonsBay. As pointed out by Womersley (1996), theseenvironments are typical of G. chilensis as describedpreviously from both Chile (Bird et al. 1986) and NewZealand (Nelson 1987). The Marion Bay population, inparticular, shows great promise as a potential feedstock for abalone mariculture, growing in dense aggre-gates of unattached, vegetatively fragmenting frondsin slack water subject to wide salinity fluctuations. Itsassignment to G. chilensis (plants from Marion Baypreviously had been regarded as likely forms ofG. secundata and were filed under that name in MELU)has a number of important implications, not least ofwhich is that, far from being unique to Tasmania, thisspecies is native to Victoria and thus would notconstitute an exotic import subject to special Environ-mental Protection Agency and Australian Quarantineand Inspection Service regulations if its cultivationwere to be undertaken by the abalone industry in Victoria.

Molecular evidence indicates that G. perplexa, fromcentral New South Wales, is a sister species toG. preissiana from Western Australia. Comparison ofthe morphology and anatomy of these two species(Table 2) show that these two species can be readilydistinguished by their gross morphology, with G. preissianabeing more compressed (flattened) on the main andlower axes and having many short undivided lateralsarising from blade margins. The restricted range ofG. perplexa to Botany Bay probably has more to do withthe lack of effort in collecting in the lower intertidal/shallow subtidal in this region than a restricted rangeof this species.

Polysaccharides and Fourier transform infrared spectroscopyThe agar yields of most Gracilaria species have beenreported to be over the range of 10–30%. The excep-tionally high yield of crude agar preparation obtainedfrom G. cliftonii was thus surprising, although itsquality remains to be accurately analyzed. In contrast,


the three samples of G. chilensis produced agars at thelow end of the yield range. It should be noted, however,that different agarophytes vary significantly in theoptimum conditions for their agar extraction. Theextraction protocol adopted in this study representsonly one approach, by no means optimized for eachspecies. Furthermore, G. chilensis is established as anagarophyte of commercial importance, based on itsagar yield, quality and growth performance in fieldculture (e.g. Pickering et al. 1993).

All the polysaccharide preparations obtained, exceptthat from G. perplexa, showed physicochemical proper-ties characteristic of agars. Interestingly, no polysac-charide could be obtained from the G. perplexaprecipitation process without the addition of NaCl. Thisnecessary addition of NaCl, together with the FTIRspectrum and cold-water solubility of the preparationfrom G. perplexa, initially suggested that it containedκ-/β-carrageenan when compared with infrared dataestablished by Liao et al. (1993), or a highly sulfatedagar (Whyte et al. 1985). Anatomical features con-firmed that the species could only be a Gracilaria, agenus not known to produce any cell-wall hydrocolloidother than agar. Confirmation of the agar identity of thepolysaccharide from G. perplexa can readily beachieved by using 13C nuclear magnetic resonancespectroscopy (Usov 1984). Agars with high levels ofsulfation have been recorded from several species ofGracilaria in recent literature, such as Gracilaria edulisby Villanueva and Montano (1999). Like that ofG. perplexa, these agars have heavy sulfation at theC(O)4 position of the 3-linked galactosyl residues.Fourier transform infrared spectroscopy, a very usefultechnique for qualitative and semiquantitative compar-isons of the 3,6-anhydrogalactosyl residues and sulfateesters (Stancioff and Stanley 1969; Rochas et al.1986), was employed in this study for rapid screeningof the cell-wall polysaccharides, although testing ofother quality parameters is needed for overall assess-ment of the agar samples.

In conclusion, our study integrates several biological/molecular approaches in this study of the commerciallyimportant genus, Gracilaria, in southeastern Australianwaters. We have shown that G. chilensis is common inVictorian waters, G. cliftonii produces the highest agaryield of the investigated species, and we have describeda new species G. perplexa Byrne et Zuccarello fromNew South Wales.

ACKNOWLEDGMENTSWe thank Jason Bartlett for technical help, Nick Paul forcollections of Gracilaria perplexa, John Huisman forcollections of Gracilaria preissiana and Ptilophora pro-lifera, and H. P. Calumpong for collections of Gracilaria‘verrucosa’. We also appreciate the financial supportand technical assistance from the Kilkunda Abalone

Farm, Phillip Island, Victoria, particularly Brett Stevens.The Australian Research Council (A199170561999–2001; SG0935526 (1994); S198122824(1998; S005005 (2000)) supported part of this work.

REFERENCESAbbott, I. A., Zhang, J. and Xia, B. 1991. Gracilaria mixta,

sp. nov. and other western Pacific species of the genus(Rhodophyta: Gracilariaceae). Pac. Sci. 45: 12–27.

Agardh, J. 1876. Species genera et ordines algaraum.Volumen tertium: de Florideis curae posteriores, Pt 1.Publisher unknown, Leipzig, Vii + 724 pp.

Armisen, R. 1995. World-wide use and importance ofGracilaria. J. Appl. Phycol. 7: 231–43.

Bird, C. J. 1995. A review of recent taxonomic concepts anddevelopments in the Gracilariaceae (Rhodophyta). J. Appl.Phycol. 7: 255–67.

Bird, C. J. and McLachlan, J. 1982. Some underutilizedtaxonomic criteria in Gracilaria (Rhodophyta, Gigartinales).Bot. Mar. 15: 557–62.

Bird, C. J., McLachlan, J. and de Oliveira, E. C. 1986.Gracilaria chilensis sp. nov. (Rhodophyta, Gigartinales),from Pacific South America. Can. J. Bot. 64: 2928–34.

Bird, C. J., Nelson, W. A., Rice, E. L., Ryan, K. G. andVillemur, R. 1990. A critical comparison of Gracilariachilensis and G. sordida (Rhodophyta, Gracilariales).J. Appl. Phycol. 2: 375–82.

Bird, C. J., Rice, E. L., Murphy, C. A. and Ragan, M. A. 1992.Phylogenetic relationships in the Gracilariales (Rhodophyta)as determined by 18S rDNA sequences. Phycologia. 31:510–22.

Bird, C. J., Ragan, M. A., Critchley, A. T., Rice, E. L. andGutell, R. R. 1994. Molecular relationships among theGracilariaceae (Rhodophyta): Further observations on someundetermined species. Eur. J. Phycol. 29: 195–202.

Bremer, K. 1988. The limits of amino acid sequence datain angiosperm phylogenetic reconstruction. Evolution 42:795–803.

Chang, C. F. and Xia, B. M. 1963. Polycavernosa, a newgenus of the Gracilariaceae. Studia Mar. Sin. 3: 119–26,pls 1,2.

De Toni, G. B. 1900. Sylloge Algarum. Vol. IV. Florideae.Sectio II. Publisher unknown, Patavii (Padova),pp. 387–774 + 775–6 (Index).

Destombe, C. and Douglas, S. E. 1991. Rubisco spacersequence divergence in the rhodophyte alga Gracilaria ver-rucosa and closely related species. Curr. Genet. 19: 395–8.

Dinh, N. H. 1992. Vietnamese species of Gracilaria andGracilariopsis. In Abbott, I. A. (Ed.) Taxonomy of Eco-nomic Seaweeds, with Reference to Some Pacific andWestern Atlantic Species, Vol. III. Report no. T-CSGCP-023, California Sea Grant College Program, La Jolla, CA,pp. 207–10.

Edelstein, T., Chen, L. C. -M. and McLachlan, J. 1978.Studies on Gracilaria (Gigartinales, Rhodophyta): repro-ductive structures. J. Phycol. 14: 92–100.

310 K. Byrne et al.

Eriksson, T. 1998. AutoDecay, Version 4.0. Program distrib-uted by the author at http://www.bergianska.se/personal/TorstenE/. Department of Botany, Stockholm University,Stockholm.

Felsenstein, J. 1985. Confidence intervals on phylogenies: anapproach using the bootstrap. Evolution 39: 783–91.

Fredericq, S. and Hommersand, M. H. 1989a. Proposal of theGracilariales ord. nov. (Rhodophyta) based on an analysisof the reproductive development of Gracilaria verrucosa.J. Phycol. 25: 213–27.

Fredericq, S. and Hommersand, M. H. 1989b. Comparativemorphology and taxonomic status of Gracilariopsis(Gracilariales, Rhodophyta). J. Phycol. 25: 228–41.

Fredericq, S. and Norris, H. N. 1992. Studies on cylindricalspecies of western Atlantic Gracilaria (Gracilariales,Rhodophyta): G. cylindrica Boergesen and G. boldgettiiHarvey. In Abbott, I. A. (Ed.) Taxonomy of EconomicSeaweeds, with Reference to Some Specific Pacific andWestern Atlantic Species, Vol. III. Report no. TCSGCP-023, California Sea Grant College Program, La Jolla, CA,pp. 211–31

Goff, L. J. and Moon, D. A. 1993. PCR amplification ofnuclear and plastid genes from algal herbarium specimensand algal spores. J. Phycol. 29: 381–4.

Goff, L. J., Moon, G. A. and Coleman, A. W. 1994. Moleculardelineation of species and species relationships in the redalgal agarophytes Gracilariopsis and Gracilaria (Gracilari-ales). J. Phycol. 30: 521–37.

Liao, M. -L., Kraft, G., Munro, S., Craik, D. J. and Bacic, A.1993. Beta/kappa-carrageenans as evidence for continuedseparation of the families Dicranemataceae and Sarco-diaceae (Gigartinales, Rhodophyta). J. Phycol. 29:833–44.

McBride, D. L. and West, J. A. 1999. Patterns of tetrasporedischarge in Caloglossa and Murrayella (Ceramiales,Rhodophyta). Phycol. Res. 47: 21–6.

May, V. 1948. The algal genus Gracilaria in Australia. Com-monw. Aust. Counc. Sci. Int. Res. Bull. 235: 1–64, pls1–15.

Nelson, W. A. 1987. The New Zealand species of GracilariaGreville (Rhodophyta, Gigartinales). N. Z. J. Bot. 25: 87–98.

Ohmi, H. 1958. The species of Gracilaria and Gracilariopsisfrom Japan and adjacent waters. Mem. Fac. Fish.Hokkaido Univ. 6: 1–66, pls 1–10.

Pickering, T. D., Sladden, V. H., Furneaux, R. H., Hemming-son, J. A. and Redfearn, P. 1993. Comparison of growthrate in culture, dry matter content, agar content and agarquality of two New Zealand red seaweeds, Gracilariachilensis Bird, McLachlan et Oliveira and Gracilaria trun-cata Kraft. J. Appl. Phycol. 5: 85–91.

Posada, D. and Crandall, K. A. 1998. Modeltest: testing themodel of DNA substitution. Bioinformatics 14: 817–18.

Posada, D. and Crandall, K. A. 2001. Selecting the best-fitmodel of nucleotide substitution. Syst. Biol. 50: 580–601.

Rochas, C., Lahaya, M. and Yaphe, W. 1986. Sulfate contentof carrageenan and agar determined by infrared spectros-copy. Bot. Mar. 29: 335–40.

Saint-Hilaire, A. 1833. Voyage Dans le District Des Diamanset Sur le Litteral Du Brésil, Vol. 2. Publisher unknown,Paris, 457 pp.

Sonder, O. G. 1881. Additamenta pro Supplentum volumnisundecimi. Algae. In von Mueller, F. (Ed.) Fragmenta Phy-tographiae Australiae, Vol. 11. (Suppl.) GovernmentPrinter, Melbourne, pp. 105–6.

Stancioff, D. J. and Stanley, N. F. 1969. Infrared andchemical studies on algal polysaccharides. Proc. Int.Seaweed Symp. 6: 595–609.

Swofford, D. 2001. PAUP*. Phylogenetic Analysis Using Par-simony (*and Other Methods), Version 4. Sinauer Associ-ates, Sunderland, MA.

Trono, G. C. Jr, Azanza-Corrales, R. and Manuel, D. 1983.The genus Gracilaria (Gigartinales, Rhodophyta) in thePhillipines. Kaliksan Phil. J. Biol. 12: 15–41.

Usov, A. I. 1984. NMR spectroscopy of red seaweed poly-saccharides: agars, carrageenans, and xylans. Bot. Mar.27: 189–202.

Villanueva, R. and Montano, N. 1999. Highly methylated agarfrom Gracilaria edulis (Gracilariales, Rhodophyta). J. Appl.Phycol. 11: 225–7.

West, J. A. and Zuccarello, G. C. 1999. Biogeography ofsexual and asexual reproduction in Bostrychia moritziana(Rhodomelacaeae, Rhodophyta). Phycol. Res. 47: 115–23.

Whyte, J. C., Hosford, S. P. C. and Englar, J. R. 1985.Assignment of agar or carrageenan structures to red algalpolysaccharides. Carbohydr. Res. 140: 336–41.

Withell, A. F., Millar, A. J. K. and Kraft, G. T. 1994.Taxonomic studies of the genus Gracilaria (Gracilariales,Rhodophyta) from Australia. Aust. Syst. Bot. 7: 281–352.

Womersley, H. B. S. 1996. The Marine Benthic Flora ofSouthern Australia. Part IIIB. Gracilariales, Rhodymeni-ales, Corallinales and Bonnemaisoniales. Australian Bio-logical Resources Study, Canberra, 329 pp.

Yamamoto, H. 1978. Systematic and anatomical study of thegenus Gracilaria in Japan. Mem. Fac. Fish. Hokkaido Univ.25: 97–152.

Zhang, J. and Xia, B. 1992. Studies on two new Gracilariafrom south China and a summary of Gracilaria species inChina. In Abbott, I. A. (Ed.) Taxonomy of EconomicSeaweeds with Reference to Some Pacific and WesternAtlantic Species, Vol. III. Report no. T-CSGCP-023, Cali-fornia Sea Grant College Program, La Jolla, CA,pp. 195–206.

Zuccarello, G. C. and West, J. A. 1997. Hybridization studiesin Bostrychia: 2. Correlation of crossing data and plastidDNA sequence data within B. radicans and B. moritziana(Rhodophyta, Ceramiales). Phycologia 36: 293–304.

Zuccarello, G. C. and West, J. A. 2002. Phylogeography of theBostrychia calliptera/B. pinnata complex (Rhodomelaceae,Rhodophyta) and divergence rates based on nuclear, mito-chondrial and plastid DNA markers. Phycologia 41: 49–60.

Zuccarello, G. C., Burger, G., West, J. A. and King, R. J.1999a. A mitochondrial marker for red algal intraspecificrelationships: variable populations and maternally inher-ited. Mol. Ecol. 8: 1443–8.


Zuccarello, G. C., West, J. A., Kamiya, M. and King, R. J.1999b. A rapid method to score plastid haplotypes in redseaweeds and its use in determining parental inheritanceof plastids in the red alga Bostrychia (Ceramiales). Hydro-biologia 401: 207–14.

Zuccarello, G. C., West, J. A. and King, R. J. 1999c. Evolu-tionary divergence in the Bostrychia moritziana/B. radi-cans complex (Rhodomelaceae, Rhodophyta): molecularand hybridisation data. Phycologia 38: 234–44.

Documents

Gracilaria species (Gracilariaceae, Rhodophyta) from southeastern