2

Systematics,phylogenyandbiogeography
ofJuncaginaceae
Dissertation
zurErlangungdesGrades
DoktorderNaturwissenschaften
amFachbereichBiologie
derJohannesGutenbergUniversitt
Mainz
SabinevonMering
geb.inErfurt
Mainz,Juni2013
Dekan:
1.Berichterstatter:
2.Berichterstatter:
TagdermndlichenPrfung:

TriglochinmaritimaL.SaltmarshinDenmark(Photo:SvM).
Fortherearesomeplantswhichcannotliveexceptinwet;
andagainthesearedistinguishedfromoneanotherbytheirfondnessfordifferentkindsofwetness;
sothatsomegrowinmarshes,othersinlakes,othersinrivers,otherseveninthesea[].
Somearewaterplantstotheextentofbeingsubmerged,whilesomeprojectalittlefromthewater;
ofsomeagaintherootsandasmallpartofthestemareunderthewater,buttherestofthebodyis
altogetheraboveit.
Theophrastus(370c.285B.C.)onaquaticplantsinEnquiryintoPlants(HistoriaPlantarum)

TABLEOFCONTENTS
INTRODUCTION 1
CHAPTER1: Phylogeny,systematics,andrecircumscriptionofJuncaginaceae
acosmopolitanwetlandfamily 7
CHAPTER2: Phylogeny,biogeographyandevolutionofTriglochinL.(Juncaginaceae)
morphologicaldiversificationislinkedtohabitatshiftsratherthanto
geneticdiversification 25
CHAPTER3: RevisionoftheMediterraneanandsouthernAfrican
Triglochinbulbosacomplex(Juncaginaceae) 51
CHAPTER4: TetronciumanditsonlyspeciesT.magellanicum(Juncaginaceae):
distribution,ecologyandlectotypification 91
CHAPTER5: MorphologyofMaundiasupportsitsisolatedphylogeneticposition
intheearlydivergentmonocotorderAlismatales 103
CONCLUSIONSANDOUTLOOK 141
REFERENCES 143
APPENDICES 169
Appendix1.Listofaccession(Chapter1)
Appendix2.Voucherinformation(Chapter2)
Appendix3.PCRinformation(Chapter2)
Appendix4.ListofstudiedspecimensofTetronciummagellanciumWilld.(Chapter4)
Appendix5.ListofstudiedspecimensofMaundiatriglochinoidesF.Muell.(Chapter5)
Appendix6.IdentificationkeysforTriglochininEurope
SUMMARY 189
ZUSAMMENFASSUNG 191
LEBENSLAUF
ACKNOWLEDGEMENTS

Publicationsarisingfromthisthesis
Chapter 1 of this thesis has been published in the peerreview proceedings of the Monocots IV
conference(Copenhagen,2008):
von Mering S. & Kadereit J. W. 2010: Phylogeny, systematics and recircumscription of
Juncaginaceae a cosmopolitan wetland family. Pp. 5579 in: Seberg O., Petersen G.,
Barfod A. S. & Davis J. I. (ed.), Diversity, phylogeny, and evolution in the monocotyledons
Proceedings of the Fourth International Conference on the Comparative Biology of the
MonocotyledonsandtheFifthInternationalSymposiumonGrassSystematicsandEvolution.
Aarhus:AarhusUniversityPress.
Chapter2ofthisthesiswillbesubmittedtoMolecularPhylogeneticsandEvolution:
von Mering S. & Kadereit J. W. Phylogeny, biogeography and evolution of Triglochin L.
(Juncaginaceae) morphological diversification is linked to habitat shifts rather than to
geneticdiversification.
Chapter3ofthisthesishasbeenpublishedintheEdinburghJournalofBotany:
KckeA.V.*,vonMeringS.*,MucinaL.&KadereitJ.W.2010:RevisionoftheMediterranean
and southern African Triglochin bulbosa complex (Juncaginaceae). Edinburgh Journal of
Botany67:353398.DOI:10.1017/S0960428610000041.
*Theseauthorscontributedequallytothispublication.
Chapter4ofthisthesishasbeenpublishedinWilldenowia:
von Mering S. 2013: Tetroncium and its only species T. magellanicum (Juncaginaceae):
distribution, ecology and lectotypification. Willdenowia 43: 1324. DOI: 10.3372/wi.43.
43102.
Chapter5ofthisthesiswillbepublishedintheBotanicalJournaloftheLinneanSociety:
Sokoloff, D. D., von Mering, S.*, Jacobs, S. W. L. & Remizowa, M. V. 2013 (in press): Flower
structure of Maundia supports its isolated phylogenetic position in the earlydivergent
monocotorderAlismatales.Bot.J.Linn.Soc.
*SvMconductedmorphologicalstudiesonherbariummaterialofMaundiaandrelatedtaxa,gathered
charactersandliteratureinformationforMaundiaandrelatedtaxa,andworkedonthemanuscript.
Introduction
1
Introduction
The monocot order Alismatales is a cosmopolitan and highly diverse group comprising 13
(APG III 2009) or 14 families with about 166 genera and ~4500 species (Stevens 2001+; Table 1 in
Chapter1).Itformsawellsupportedmonophyleticclade(APGII2003,APGIII2009,Soltisetal.2005
and references therein) and the current circumscription includes wellknown groups such as the
speciesrichandhighlydiverseArumfamily(Araceae;e.g.,Cabreraetal.2008,Cusimanoetal.2011)
aswellasallmarinefloweringplants(seagrasses;e.g.,denHartog1970,Green&Short2003).
Molecular phylogenetic studies have shown that Araceae are sister to all remaining families
of the order, and within the latter group Tofieldiaceae are sister to the socalled core Alismatales
(e.g.,Lesetal.1997,Chaseetal.2006,Ilesetal.2013,Les&Tippery2013).CoreAlismatales(e.g.,
Iles et al. 2013), approximately corresponding to the traditional order Helobiae (Engler 1909), is a
morphologicallyhighlydiversemonophyleticgroupofaquaticandwetlandplantsfoundinallregions
of the world. Core Alismatales can be divided into two informal groups based on floral character
istics: a petaloid and a tepaloid clade (Posluszny & Charlton 1993). While petaloid alismatids
comprise the three families Hydrocharitaceae, Butomaceae, and Alismataceae (incl. Limnochari
taceae), tepaloid alismatids comprise the following eight families: Scheuchzeriaceae, Apono
getonaceae, Juncaginaceae, Posidoniaceae, Cymodoceaceae, Ruppiaceae, Potamogetonaceae, and
Zosteraceae.
As an earlydiverging monocot lineage, Alismatales are one key to the understanding of

monocot evolution. Several studies of members of Alismatales have concentrated on comparative
flower morphology and development (e.g., Posluszny & Charlton 1993, Posluszny et al. 2000,
Remizowa et al. 2012b). However, some taxa have been overlooked by researchers and very little
information about their biology is available in the literature. Among those poorly known taxais the
small arrowgrass family Juncaginaceae Rich., a family of annual or perennial grasslike herbs
characterised by basal, sheathing leaves and spikelike inflorescences (Fig. 1) and underground
storageorganssuchasbulbs,rhizomes,stolons,ortuberousroots.
The family Juncaginaceae was established at the beginning of the 19

th
century when Louis
Claude Marie Richard used the name Juncagines and gave the following short description plus a
Frenchtranslationofit:Juncagines.Caps2.Spermas.Akenium:Sem.erectum.Embryoperispermi
cus,orthotropus,brachypodus.(Richard1808).Someyearslater,Richardlistedthefollowinggenera
comprising Juncagineae: Lilaea, Cathanthes [a synonym of Tetroncium], Triglochin, Scheuchzeria
(Richard 1815). Juncaginaceae is a conserved name (nomen conservandum, nom. cons.). Circum
scription of the family and delimitation of genera have changed considerably since Richards times.
MoredetailsonthehistoryofclassificationaregiveninChapter1(seealsoTable1inChapter1).
The last comprehensive study of the family dates back to the beginning of the 20
th
century
(Buchenau 1903). Few broader studies focussing on members of the family were published
afterwards; publications were mostly restricted to regional studies (e.g., Markgraf 1981, Haynes
2004) or focussing on one or few taxa (e.g., Davy & Bisphop 1991). Notable exceptions are the
revision of the tuberousrooted Triglochin species (Aston 1993, 1995) and the recent treatment for
theFloraofAustralia(Aston2011)coveringamajorpartofthefamily.
Introduction
2
The number of accepted genera in Juncaginaceae has varied between three and five (e.g.,
Dahlgrenetal.1985,Takhtajan1997,Haynesetal.1998)andspeciesnumbersrangefrom12to15
(e.g.,Haynesetal.1998,Stevens2001+,Mabberley2008).However,thesedataarebasedonlimited
information and considerably underestimate the real number of species. Recent treatments have
shownthatthenumberofrecognisedspeciesexceeds30(e.g.,Kckeetal.2010;Aston1993,1995,
2011).
The family Juncaginaceae has an almost cosmopolitan distribution, with a notable gap in
large parts of the tropics (for a distribution map see Fig. 1 in Chapter 1). Australia is the centre of
species diversity with about 26 species in 3 genera (Aston 2011). Australian taxa include a group of
annual Triglochin species and the Triglochin procera complex [Cycnogeton] (Aston 2011). Further
species are native to the southern hemisphere: several taxa of the Triglochin bulbosa complex are
endemic to South Africa (see Chapter 3; Fig. 1), and the monotypic Tetroncium is restricted to
southernSouthAmerica(seeChapter4;Fig.1).Whilesomespeciesarewidespreadintemperateor
cooler regions of both hemispheres (e.g., T. palustris, T. maritima) others are restricted to small
areas(e.g.,T.gaspensis,T.buchenaui).
Despite its small size, the family shows considerable ecological diversity. Members of
Juncaginaceae are windpollinated herbs which can be found in freshwater wetlands (slowflowing
rivers,bogs,fens),inbrackishwater (e.g.,estuaries),andin coastalsaltmarshes.Sometaxainhabit
only seasonally wet terrestrial sites such as vernal pools or inland salt pans (e.g., annual species of
Triglochin, T. bulbosa L. subspp.). Species of the T. procera complex (Cycnogeton) are true aquatics,
livinginfreshwaterhabitatsinAustralia(Aston1993,1995,2011).Triglochinisprobablybestknown
for its coastal salt marsh species found in many parts of the world (e.g., T. maritima, T. striata).
However, even those taxa are not confined to coastal habitats, but also occur in inland habitats.
Nevertheless, occurrence is usually linked to (seasonal) wetlands and often to habitats with higher
salinity.Triglochinspeciesoccurinhabitatsfromsealevelupto4,000metresintheRockyMountains
(Haynes & Hellquist 2000) and even more than 5,000 metres altitude in the Himalayas (Guo et al.
2010).
At the beginning of this project little was known about phylogenetic relationships within
Juncaginaceae. Almost no sequence data were available for members of the family and inter and
intrageneric relationships remained largely unresolved. Despite the interesting distribution pattern,
biogeographic aspects had not been studied in detail. In addition, the urgent need for taxonomic
revisions of several groups within Juncaginaceae had been emphasised by different authors (e.g.,
Thieret1988,H.I.Aston,pers.comm.).
In this thesis I have investigated the systematics, phylogeny and biogeographical history of
the family Juncaginaceae based on morphological and molecular data. Molecular markers from all
three plant genomes (nuclear, chloroplast and mitochondrial) were used in this first study of
Juncaginaceae on a global scale and an extensive specieslevel phylogeny was generated, with a
focus on the extraAustralian taxa (Chapter 1 and 2). In the course of this project, more than 7,500
specimens from about 40 herbaria have been revised to clarify species delimitation and to gain
insight into the specific diversity of Triglochin and relatives (Chapter 3 and 4). The main questions
included: Are Juncaginaceae in the traditional circumscription a monophyletic group? How are the
generarelatedtoeachother?Howoldisthegroupandwherediditoriginate?
Introduction
3
Fig.1.Juncaginaceae.Representativespeciesofthefamilyshowinggeneralhabitandhabitatdiversity.A.
Triglochinmaritima(coast,Galicia,Spain),B.T.bulbosassp.quarcicola(quartzfield,Knersvlakte,SouthAfrica),
C.Triglochinlaxiflora(rockpool,Malta),D.Triglochinscilloides(syn.Lilaeascilloides),(poolinBotanicGarden
Mainz),E.Tetronciummagellanicum(Sphagnumbog,Argentina),F.Cycnogetonsp.(freshwater,Australia).
Photos:S.v.Mering,U.Schmiedel,S.Mifsud,R.Greissl,R.Douzet,SAJF,A.Schnhofer.

Introduction
4
In the five chapters of this thesis I have used a wide range of methods to address these
questions and to extend our knowledge of Juncaginaceae, from classical taxonomic revisions to
molecularsystematicandbiogeographicapproaches.
In Chapter 1 a phylogenetic analysis of the family and members of Alismatales was

conducted to clarify the circumscription of Juncaginaceae and to understand intrafamilial
relationships. For the first time, all accepted genera or those associated with the family in the past
were analysed together. Phylogenetic analysis of three molecular markers (rbcL, matK, and atpA)
showedthatJuncaginaceaewerenotmonophyletic.Asaconsequencethefamilyisrecircumscribed
to exclude Maundia which is proposed to belong to a separate family Maundiaceae. In the new
classification the number of accepted genera in Juncaginaceae is reduced to three: Tetroncium,
Cycnogeton, and Triglochin. Tetroncium is weakly supported as sister to the rest of the family. The
reinstated Cycnogeton (formerly included in Triglochin) is highly supported as sister to Triglochin
s.str. The enigmatic Lilaea is nested within Triglochin s. str. and highly supported as sister to the T.
bulbosa complex. The results of the molecular analysis are discussed in combination with
morphological characters. Additionally, a key to the genera of the family is given and several new
combinationsaremade.
In Chapter 2 the phylogenetic relationships within the largest genus Triglochin were
investigated. A specieslevel phylogeny was constructed based on molecular data obtained from
nuclear (ITS, internal transcribed spacer) and chloroplast sequence data (psbAtrnH spacer, matK
gene). Based on the phylogeny of the group divergence times were estimated and ancestral
distribution areas reconstructed. The monophyly of Triglochin is confirmed and relationships
between the major lineages of the genus are resolved. A clade comprising the Mediterranean/
African T. bulbosa complex and the American T. scilloides (formerly Lilaea s.) is sister to the rest of
the genus which contains two main clades. In the first, the widespread T. striata is sister to a clade
comprisingannualTriglochinspeciesfromAustralia.ThesecondcladecomprisesT.palustrisassister
to the T. maritima complex, of which the latter is further divided into a Eurasian and an American
subclade. Diversification in Triglochin began in the Miocene or Oligocene, and most disjunctions in
Triglochin were dated to the Miocene. Taxonomic diversity in some clades is strongly linked to
habitat shifts and can not be observed in old but ecologically invariable lineages such as the non
monophyleticT.maritima.
Chapter 3 is an taxonomical treatment carried out in collaboration with colleagues. The

Triglochin bulbosa complex (Juncaginaceae), a monophyletic group from the Mediterranean region
and Africa was revised based on the study of a large number of herbarium specimens. One new
species, Triglochin buchenaui, and two new subspecies, T. bulbosa subsp. calcicola and subsp.
quarcicola,aredescribedfromSouthAfrica.Furthermore,twotaxaareelevatedtospeciesrankand
two are reinstated. Altogether, seven species and four subspecies are recognised, thus, more than
doubling the number of accepted Triglochin species for this region. An identification key, detailed
descriptions and accounts of the ecology and distribution of the taxa are provided. An IUCN
conservationstatusisproposedforeachtaxon.
Chapter 4 deals with the monotypic genus Tetroncium from southern South America. The
remarkableT.magellanicumistheonlydioeciousspeciesofthefamily.Thetaxonomichistoryofthe
speciesisdescribed,typematerialistraced,andalectotypeforthenameisdesignated.Basedonan
Introduction
5
extensive study of herbarium specimens and literature a detailed description of the species and
notesonitsecologyandconservationstatusareprovided.Forthefirsttime,adetailedmapshowing
thecompleteknowndistributionareaofT.magellanicumispresented.
In the final chapter, Chapter 5, colleagues and I studied the flower structure of the rare
Australian endemic Maundia triglochinoides which has recently been shown not to be amember of
Juncaginaceae based on molecular data. In the light of this finding reassessment of views on the
morphologicalevolutionandclassificationofAlismatalesisrequired.AsthemorphologyofMaundia
ispoorlyknownandsomekeyfeaturesarecontroversiallydescribedintheliterature,inflorescences,
flowers and fruits were studied using serial mictrotome sections and scanning electron microscopy.
Theisolatedphylogeneticplacement,affinitiestorelatedtaxa,andtheevolutionofcertainfeatures
are discussed. As Maundia exhibits a mosaic of features characteristic of other families of tepaloid
coreAlismatalesitssegregationasaseparatefamilyisplausible.

Introduction
6
Chapter1
Phylogeny,systematicsandrecircumscriptionofJuncaginaceae
acosmopolitanwetlandfamily
SabinevonMering&JoachimW.Kadereit
Published in: Seberg O., Petersen G., Barfod A. S. & Davis J. I. 2010 (ed.): Diversity, phylogeny, and
evolution in the monocotyledons Proceedings of the Fourth International Conference on the
Comparative Biology of the Monocotyledons and the Fifth International Symposium on Grass
SystematicsandEvolution.Pp.5579.Aarhus:AarhusUniversityPress.
ABSTRACT
Juncaginaceae are a small monocot family of mostly coastal and wetland herbs of almost
cosmopolitan distribution. A phylogenetic analysis of the family and members of Alismatales was
conducted to clarify the circumscription of Juncaginaceae and to understand intrafamilial
relationships. For the first time, all genera associated with the family in the past were analysed
together. Two plastid (rbcL and matK) and one mitochondrial gene (atpA) were sequenced. The
separate and combined analysis of the three markers showed that Juncaginaceae are not
monophyletic in their current circumscription. The family is recircumscribed to exclude Maundia
which is proposed to belong to a separate family Maundiaceae. In the new classification
Juncaginaceaecomprisethreegenera:Tetroncium,Cycnogeton,andTriglochin.Tetronciumisweakly
supported as sister to the rest of the family. The reinstated Cycnogeton (formerly included in
Triglochin) is highly supported as sister to Triglochin s.s. The enigmatic Lilaea is nested within
Triglochins.s.andhighlysupportedassistertotheT.bulbosacomplex.Theresultsofthemolecular
analysis are discussed in combination with morphological characters. A key to the genera of the
family is presented, and six new combinations are proposed: Cycnogeton alcockiae, Cycnogeton
dubium, Cycnogeton microtuberosum, Cycnogeton multifructum, Cycnogeton rheophilum, and
Triglochinscilloides.
1
Keywords.Alismatales,Cycnogeton,Lilaea,Maundia,Tetroncium,Triglochin.
1
ThisarticleisdedicatedtothememoryofDr.SurreyW.L.Jacobs(19462009)inappreciationofhis
outstandingcontributionstotheknowledgeofAustralianwaterplantsandingratefulacknowledge
mentofhissupportoftheprojectdescribedhere.

Chapter1:PhylogenyandsystematicsofJuncaginaceae
8
9
INTRODUCTION
TheorderAlismatales(14familieswithca.4490species;Stevens2001+),oneoftheearliest
diverging lineages of monocotyledons (Janssen and Bremer 2004), comprises mainly aquatic and
wetland plants. While several groups of Alismatales have received considerable attention (e.g.,
seagrasses: e.g., den Hartog 1970; Les et al. 1997; Araceae: e.g., Mayo et al. 1997; Cabrera et al.
2008),somesmallerfamiliesoftheorderhavenotbeenstudiedindetail.OneoftheseareJuncagi
naceae(Arrowgrassfamily),partofthesocalledaquaticclade(Juddetal.2007)orcoreAlismatales
(Stevens2001+).
Ascurrentlycircumscribed,Juncaginaceaecomprisefourgenera,Triglochin,Lilaea,Maundia,
and Tetroncium (Haynes et al. 1998; Stevens 2001+) with together approximately 2535 annual or
perennialspecies.Despiteitssmallsize,thefamilyshowsconsiderableecologicaldiversity.Members
of Juncaginaceae are windpollinated, grasslike herbs which can be found in freshwater (slow
flowing rivers, bogs, fens), in brackish water (e.g., estuaries), and in salt marshes, but also in only
seasonally wet terrestrial sites (e.g., annual species of Triglochin, T. bulbosa L. subspp.). The family
hasanalmostcosmopolitandistributionwithAustraliaascentreofspecificdiversity(Fig.1).
Triglochin is the largest genus of the family and is distributed almost worldwide. Widely
circumscribed(inthefollowingcalledTriglochins.l.)itcomprisesthemostlyhalophyticarrowgrasses
(Triglochins.s.)andthewaterribbons(T.proceraR.Br.andrelatedspeciesoftheT.proceracomplex)
which are important components of Australian freshwater communities. The latter complex is
sometimes segregated from Triglochin as Cycnogeton. The number of recognized species in
Triglochin varies greatly in the literature, ranging from 12 to 24 (e.g., Haynes et al. 1998; Stevens
2001+;Govaerts2008;Mabberley2008).However,recentrevisions(Aston1995;Kckeetal.,2010;
H. I. Aston, pers. comm.) have revealed considerably higher numbers of probably more than 30
species.
Fig.1.GeographicaldistributionofthecurrentlyrecognizedgeneraofJuncaginaceae.Mappreparedusing
OnlineMapCreation(Weinelt1996onwards).
2/6
1/2
1/4
1/8
3/5
2/c.20
10
While Triglochin is distributed almost worldwide, the monotypic genera Tetroncium and
Maundia are restricted to relatively small areas in southern South America and eastern Australia,
respectively (Fig. 1). Tetroncium magellanicum is a typical component of Sphagnum L. bogs in
Patagonia and Tierra del Fuego, and on some subantarctic islands (Falkland Islands, Gough Island).
Maundiatriglochinoidesisfoundinswamps,creeks,orshallowfreshwaterincoastalQueenslandand
NewSouthWales.Lilaea,thethirdmonotypicgenus,isdistributedfromsouthernSouthAmericato
southernCanada(Fig.1).Lilaeascilloidesgrowsemergentorsubmergedinshallowwaterofseasonal
pools or neighbouring mud flats. This species is also naturalised in Australia (Aston 1967, 1977;
Australian Plant Census 2009; H. I. Aston, pers. comm.) and the Iberian Peninsula (Gonzlez 1968;
Navaetal.2000;RomeroBujn2007).
The circumscription of Juncaginaceae has changed throughout history. The most important
historicalclassificationsaresummarizedinTable1.InolderclassificationsthegenusScheuchzeriaL.
wasincludedinthefamily.Morphological(e.g.,Tomlinson1982;Posluszny1983)andmorerecently
molecular data (Les et al. 1997) have shown the distinctness of Scheuchzeria, which is now
commonlyplacedinitsownmonotypicfamilyScheuchzeriaceae.ThecorrectplacementofLilaeahas
also been problematic. This genus has often been treated as the only member of Lilaeaceae
Dumortierbecauseofitsdivergentfloralmorphology(e.g.,Taylor1909;Tomlinson1982).Resultsof
karyological (Larsen 1966), embryological (Agrawal 1952; Yamashita 1970), and molecular studies
(Les et al. 1997) have, however, shown the close relationship between Lilaea and Juncaginaceae,
resultingintheinclusionofLilaeainJuncaginaceaebymostmodernauthors(e.g.,Haynesetal.1998;
Stevens2001+).Someauthors(e.g.,NoveloandLot2001;Novelo2003),however,retainLilaeaceae
as a separate family. Finally, the monotypic Maundia has been treated also as a separate family,
MaundiaceaeNakai(1943).Subsequently,thisfamilywasacceptedonlybyTakhtajan(1997).
GenericlimitswithinJuncaginaceaealsohavebeenassesseddifferentlybydifferentauthors
(Table 1). The status of Cycnogeton and Maundia has changed several times through history. Both
taxa were originally described as monotypic genera, but later treated either as sections (Micheli
1881) or subgenera of a broadly circumscribed Triglochin (subg. Cycnogeton (Endl.) Buchenau and
subg. Pseudotriglochin (Micheli) Buchenau, respectively; Buchenau and Hieronymus 1889). While
Cycnogeton in subsequent treatments usually was included in Triglochin without recognition as
subgenus(e.g.,Aston1977,1995;Haynesetal.1998),Maundiawasgenerallyacceptedasaseparate
genus(Aston1977;Haynesetal.1998).
NomolecularstudyofallgeneraofJuncaginaceaehasyetbeenpublished.Molecularstudies
ofmonocotsorAlismatalesincludedonlyveryfewrepresentativesofthefamilyandsometimesnone
atall.AnanalysisofAlismatalesbasedonrbcLsequencedata(Lesetal.1997)containedonlythree
species (Triglochin maritima L., Cycnogeton procerum, and Lilaea scilloides which formed a strongly
supported clade with Cycnogeton as sister to a strongly supported Lilaea/Triglochin clade. In a
biogeographicalstudyofhydrophytes(Lesetal.2003),ITSsequencesofTetronciumandCycnogeton
were used to estimate divergence times, but not to reconstruct the phylogeny of these taxa. Thus,
although the position of Juncaginaceae within Alismatales is relatively unambiguous, phylogenetic
relationships among genera of Juncaginaceae are still unclear mainly because Maundia and
Tetronciumhaveneverbeenincludedinmolecularphylogeneticanalysesandbecausethesampleof
Triglochins.l.hasbeentoosmall.EventhoughsomeanatomicalinformationisavailableforMaundia
(Tomlinson 1982) the knowledge of this latter genus and Tetroncium is fragmentary and mainly
limitedtoFloratreatments(e.g.,Thompson1961;Correa1969;Aston1977;BroughtonandMcAdam
2005).
TABLE1.ClassificationofgeneraassociatedwithJuncaginaceaefrom1881topresent.
1
"Genusomninoanomalum:Juncagineispropriisvixaffine.",
2
only
Maundiatriglochinoides,
3
onlyTriglochinprocera,
4
incl.Lilaea,n.s.notspecified.
Micheli1881 Buchenau&
Hieronymus1889
Buchenau1903 Hutchinson1934,
1959
Dahlgrenetal.
1985
Takhtajan1997 Haynesetal.
1998
thisstudy
Juncagineae Juncaginaceae Scheuchzeriaceae Juncaginaceae Juncaginaceae Juncaginaceae Juncaginaceae Juncaginaceae
(1112spp.) (15spp.) (17spp.) (n.s.) (c.20spp.) (15spp.) (c.15spp.) (c.30spp.)
Lilaea
1
Lilaea Lilaea Cycnogeton Cycnogeton Cycnogeton Lilaea Triglochin
4
Scheuchzeria Scheuchzeria Maundia Maundia Lilaea Tetroncium Maundia Cycnogeton

Tetroncium Tetroncium Scheuchzeria Tetroncium Maundia Triglochin Tetroncium Tetroncium
Triglochin Triglochin Tetroncium Triglochin Tetroncium Triglochin
sect.Eutriglochin subg.Eutriglochin Triglochin Triglochin

sect.Cycnogeton
3
subg.Cycnogeton
3
subg.Eutriglochin
sect.
Pseudotriglochin
2
subg.
Pseudotriglochin
2
subg.
Cycnogeton
3

Lilaeaceae Lilaeaceae
(1sp.) (1sp.)
Lilaea Lilaea
Maundiaceae Maundiaceae
(1sp.) (1sp.)
Maundia Maundia
Scheuchzeriaceae Scheuchzeriaceae Scheuchzeriaceae Scheuchzeriaceae Scheuchzeriaceae
(1sp.) (1sp.) (1sp.) (1sp.) (1sp.)
Scheuchzeria Scheuchzeria Scheuchzeria Scheuchzeria Scheuchzeria
C
h
a
p
t
e
r
1
:
P
h
y
l
o
g
e
n
y
a
n
d
s
y
s
t
e
m
a
t
i
c
s
o
f
J
u
n
c
a
g
i
n
a
c
e
a
e
1
1
12
The objective of this study is to provide a comprehensive phylogenetic analysis of

Juncaginaceaebasedonmoleculardata.Thisphylogenywillbeinterpretedonthebackgroundofthe
morphologyofthefamilyasfarasknown.Toobtainabetterunderstandingofrelationshipsamong
generapresentlyassignedtoJuncaginaceae,andtoevaluatethedelimitationofthefamily,weused
the chloroplast (cp) genes rbcL and matK as well as the mitochondrial (mt) gene atpA. In this study
we will (1) investigate the monophyly of the family Juncaginaceae in its current circumscription, (2)
clarify relationships among the genera of the family, and (3) propose a revised classification of
Juncaginaceaewhereclearevidencefrommolecularandotherdataisavailable.
MATERIALSANDMETHODS
Taxon SamplingAltogether, nine species currently recognized as members of Juncagi

naceae, the monotypic Lilaea, Maundia, Tetroncium, and six species considered representative of
Triglochin, plus Scheuchzeria palustris L. (Scheuchzeriaceae) were included in our analysis. Our
molecularstudies(S.vonMeringandJ.W.Kadereit,unpubl.data)haveshownthatseveralmembers
of the morphologically welldefined Australian Triglochin procera complex (waterribbons) form a
monophyletic group. Difficulties in obtaining high quality DNA prevented the inclusion of an annual
species of Triglochin in this study. However, species of this group formed a monophyletic clade
within Triglochin s.s. (S. von Mering and J. W. Kadereit, unpubl. data). Sequences of members of all
other families of Alismatales were downloaded from GenBank. Voucher information and GenBank
accessionnumbersaregiveninAppendix1.
Molecular MarkerThe plastid rbcL gene was chosen in this study because rbcL sequences
areavailableformembersofallotherfamiliesofAlismatales,providingarichsourceforoutgroups.
Additionally,matKwasusedtoimproveresolutionand/orsupport,andtheatpAgenehasbeenused
intheanalysisofmonocotandAlismatalesrelationships(e.g.,Davisetal.1998,2004,2006;Petersen
etal.2006).
DNAExtraction,Amplification,andSequencingTotalgenomicDNAswereextractedeither
fromsilicadriedleaves,frommaterialpreservedinsaturatedNaClCTABsolutionsupplementedwith
200 mM sodium ascorbate (Thomson 2002), or from herbarium material using NucleoSpin plant
DNA extraction kits (MachereyNagel, Dren, Germany) following the manufacturers protocol. The
standard25lPCRreactionmixconsistedofthefollowing:2.5mMMgCl2,200MdNTPs,1pMper
primer, 0.025 U/l Taq polymerase, 12 l of DNA extract in the reaction buffer provided by the
manufacturerofthepolymerase,and1%BSA.PCRreactionswerecarriedoutinaBiometraT3ora
PTC 100
TM
thermocycler (MJ Research, Inc., MA, U.S.A.) using the programme: 60 sec at 94C,
followedby35cyclesof18secat94C,30secat52C,60secat72Candaposttreatmentof8min
at72CforrbcL.TheprogrammeusedfortheamplificationofthematKandatpAsequencesdiffered
inthehigherannealingtemperature(55C)andalongerelongationtime(90sec).
The primers used for amplification of the three markers are summarized in Table 2. RbcL
sequences were amplified in three overlapping fragments with slightly modified standard rbcL
primers: 1F and 579R, 507F and 994R, and 955F and 1460R. To amplify and sequence the matK
region one primer was modified after Mller and Borsch (2005): JUmatK 480F, and the standard
primer trnK 2R was used as reverse primer (Johnson and Soltis 1994). The following primers were
used for amplification of atpA: atpA FA1 and atpA BA1 (Davis et al. 2004). PCR products were
checked on 0.8% agarose gels and purified directly using a PCR purification kit (QIAGEN GmbH,
Hilden, Germany). Purified PCR products were cyclesequenced with the ABI Prism Dye Terminator
Cycle Sequencing Ready Reaction Kit (BD 3.0 in 10 l reactions) by Perkin Elmer using the PCR
13
primerslistedaboveandfollowingthemanufacturersprotocol.Productswerepurifiedandanalysed
by GENterprise (Mainz, Germany). Forward and reverse sequences were manually edited and
mergedintoconsensussequencesusingSequencher4.1.2(GeneCodesCorp.,AnnArbor,Michigan),
and aligned manually in MacClade 4.1 (Maddison and Maddison 2000). Alignment of rbcL and atpA
was straightforward. The matK sequences downloaded from GenBank were prealigned
automatically using the programme ClustalX (Thompson et al. 1997) and afterwards adjusted
manually.Duetodifficultieswithamplificationandsequencingsomesequencesarenotavailablefor
alltaxaandonlypartialsequencescouldbegeneratedformatKandatpA.
Morphological DataMorphological characters of Juncaginaceae and related families of
Alismatales were compiled from numerous sources, mainly from Buchenau (1903), Aston (1977,
1993, 1995), Tomlinson (1982), Dahlgren et al. (1985), Kubitzki (1998), Igersheim et al. (2001),
Stevens (2001+), and Mabberley (2008). These were supplemented by own observations of living
and/orpreservedmaterialofsomegeneraofJuncaginaceae.Wehavenotcodedthemorphological
data for phylogenetic analysis because the available information is fragmentary for some taxa and
codingwouldhaveresultedinanincompletedatamatrix.
Phylogenetic AnalysisAll datasets were separately analysed using Maximum Parsimony
(MP)andMaximumLikelihood(ML)implementedinPAUP*4.10b(Swofford2003).TheatpAdataset
showedlittlesequencevariationandresultedinapoorlyresolvedtree.ThesameappliestothematK
data set with slightly higher resolution. Therefore, a Partition Homogeneity Test (Farris et al. 1994;
implemented in PAUP*) with 100 homogeneity replicates, 10 random addition sequences, tree
bisectionreconnection (TBR) branch swapping on, best only and MULTREES on was performed to
testwhetherthethreedatasets(rbcL,matK,atpA)couldbecombined.Nosignificantincongruence
wasdetectedbetweenthedatasets(pairwiseILDtest:p=0.85forrbcLandmatK,p=0.13forrbcL
andatpA,p=0.04formatKandatpA).Inconsequence,wealsoanalysedacombineddatamatrixof
the cpDNA data and all three genes. All phylogenetic data sets were deposited in TreeBASE (study
accessionnumberS2667).
Maximum Parsimony (MP) analyses were performed using PAUP* with 1000 replicated
heuristic searches using the same heuristic search settings as described above for the Partition
Homogeneity Test. Gaps were treated as missing and gaps were not coded. Branch support was
assessed with 100 bootstrap (BS) replicates with 10 random taxon additions each and TBR and
MULTREESon.
TABLE2.PrimersusedinPCRandsequencing.*Slightlymodifiedstandardprimer.
Primername Sequence
rbcL1F 5'ATGTCACCACAAACAGAAACTAAAGCA3'
rbcL579R 5'AAATCAAGTCCACCRCG3'
rbcL507F* 5'TATTGGGATGTACTATTAAAC3'
rbcL994R* 5'CCTTCYAGTTTACCTAC3'
rbcL955F* 5'CGYATGTCTGGTGGAGATC3'
rbcL1460R 5'CCTTTAGTAAAAGATTGGGCCGAG3'
JUmatK480F 5'CATCTYGAAATHTTGGTTC3'
trnK2R 5'AACTAGTCGGATGGAGTAG3'
atpAFA1 5'CAGTTGGAGATGGGATTGCACG3'
atpABA1 5'GGCAGTGGTTCATATTGTGGTTG3'
14
For Maximum Likelihood (ML), the appropriate model of DNA substitution for the inference
of phylogenetic relationships under ML was estimated using Modeltest 3.06 (Posada and Crandall
1998).BestfitmodelswereselectedbytheAkaikeInformationCriterion(PosadaandBuckley2004)
and implemented in the corresponding data matrices (see Results). ML heuristic searches and
bootstrapbranchsupport(BS)wereperformedinPAUP*with100replicatedheuristicsearchesand
thesamesettingsasintheMPanalysis.
Following Chase et al. (2000) in presenting and discussing the results, bootstrap support of
50%74%isconsideredlow,75%84%moderate,and>85%high.
RESULTS
Molecular Data Phylogenetic AnalysisRelationships within Juncaginaceae and the

delimitation of the family within Alismatales were reconstructed based on variation in the plastid
genes rbcL and matK, the mitochondrial gene atpA, and a combination of these. Statistics for all
analysesaresummarisedinTable3.AnoverviewofallresultsisgiveninFig.2,wheresimplifiedtrees
includingMPandMLbootstrapvaluesareprovided.
CPDNADATATherbcLdatasetcomprised38speciesofallfamiliesofAlismatalesplusAcorus
(Acoraceae) as outgroup. Of the 1177 nucleotide positions included in the alignment, 358 were
variable and 235 parsimony informative. The MP analysis resulted in six shortest trees (consistency
index,CI=0.476;retentionindex,RI=0.649)withalengthof1033steps(Table3).TheMLanalysis
(GTR+I+G, shape parameter = 0.5070, base frequencies 0.2850 0.1876 0.2184, rate matrix 1.0354
3.9222 0.5658 1.0758 5.3542) yielded one best tree. No major incongruencies were found when
comparing the topologies of the MP strict consensus tree and the ML tree when only clades with
goodsupportwereconsidered.TheMLtreeisillustratedinFig.3anddescribedhere.
Inthistree(Fig.3),membersofthefamiliesAraceaeandTofieldiaceaearesistertoallother
Alismatales (core Alismatales). These strongly supported core Alismatales (ML BS 93%, MP BS 95%)
comprise two subclades: 1) members of the families Alismataceae (incl. Limnocharitaceae),
Butomaceae,andHydrocharitaceaeformahighlysupportedclade(MLBS92%,MPBS96%),and2)a
largecladeinwhichScheuchzeriaandAponogetonL.f.aremoderately(toweakly)supported(MLBS
77%, MP BS 57%) as sister groups to a clade comprising Juncaginaceae and several aquatic families
(Potamogetonaceae, Zosteraceae, Cymodoceaceae, Posidoniaceae, Ruppiaceae). The latter clade
plus Maundia is sister to all other Juncaginaceae. This clade, in the following called Maundia/
Potamogeton clade, is weakly supported in this data set (ML BS 68%, MP BS 71%) and not well
resolved. Tetroncium is weakly supported (ML BS 66%, MP BS 64%) as sister to the remaining
Juncaginaceae. The latter clade, here called Triglochin s.l., is highly supported (ML BS 98%, MP BS
98%)andcanbedividedintoT.rheophila(T.proceracomplex)andacladecomprisingseveralother
Triglochinspp.(Triglochins.s.).LilaeaisnestedwithinTriglochins.s.(MLBS100%,MPBS100%),and
ishighlysupportedassistertomembersoftheT.bulbosacomplex(MLBS94%,MPBS97%).
ThematKdatasetcomprised31taxaand911charactersofwhich478werevariableand309
parsimony informative (Table 3). The MP analysis resulted in two shortest trees (CI = 0.576, RI =
0.701)of1190steps.TheMLreconstruction(TVM+G,shapeparameter=0.6700,basefrequencies
0.3160 0.1483 0.1578, rate matrix 1.4885 2.9422 0.1615 1.2022 2.9422) resulted in one tree. This
and the MP strict consensus tree are congruent, but differ in resolution. The ML tree resolves a
trichotomyofTriglochins.l.,Tetroncium,andtheMaundia/Potamogetonclade(MLBS73%),whereas
TABLE3.Summaryofsequencecharacteristicsandtreestatisticsfortheindividualgenesandforthecombinedanalyses.bpbasepairs.
Dataset rbcL matK atpA rbcL+matK rbcL+matK+atpA

No.taxa 38 31 23 31 19
No.characters/alignmentlength(bp) 1177 911 903 2088 2985
No.(%)variablecharacters 358(30.4) 478(52.5) 209(23.1) 804(38.5) 900(30.2)
No.(%)informativecharacters[MP] 235(20.0) 309(33.9) 141(15.6) 527(25.2) 514(17.2)
No.trees[MP] 6 2 10 1 1
treelength[MP] 1033 1190 346 2075 1783
CI/RI[MP] 0.476/0.649 0.576/0.701 0.760/0.869 0.543/0.672 0.651/0.685
No.trees[ML] 1 1 1 1 1
Modelofsequenceevolution[ML] GTR+I+G TVM+G TVM+G GTR+I+G GTR+I+G
C
h
a
p
t
e
r
1
:
P
h
y
l
o
g
e
n
y
a
n
d
s
y
s
t
e
m
a
t
i
c
s
o
f
J
u
n
c
a
g
i
n
a
c
e
a
e
1
5
16
Fig.2.PhylogeneticanalysisofrbcL,matK,atpA,andthecombineddatasets.MPandMLbootstrapvalues
givenaboveandbelowtheline,respectively.Onlybootstrapvalueshigherthan70%included.Dash():clade
notrecoveredinMPanalysis.
1
Triglochinbarrelieri,T.elongata,Lilaeascilloides,T.maritima,T.palustris,T.striata.
2
Potamogeton
distinctus,Heterozosteratasmanica,Zosteranoltii,Z.marina,Phyllospadixtorreyi.
3
Amphibolisantarctica,
Cymodoceaserrulata,Ruppiamaritima,Haloduleuninervis,Posidoniaoceanica.
4
Haloduleuninervis,
Potamogetondistinctus,Heterozosteratasmanica,Zosteranoltii,Phyllospadixtorreyi/iwatensis.
5
as
1
but
incl.T.laxifloraandexcl.T.palustris.
6
Scheuchzeriapalustris,Butomusumbellatus,Otteliaacuminata,
Caldesiaoligococca,Sagittarialatifolia.
7
as
1
butexcl.T.palustris.
Missingpartsofthetree(dashedline):Aponogetonfenestralis
a,d
,Scheuchzeriapalustris
a,d,e
,Hydrocharis
dubia
a,b,d
,Limnobiumlaevigata
a,b,d
,Najasmarina
a
,Vallisneriaamericana
a
,Otteliaacuminata
a,b,d,e
,Stratiotes
aloides
a,b,d
,Butomusumbellatus
a,b,d,e
,Hydrocleysnymphoides
a,b,d
,Limoncharisflava
a,b,d
,Alismaplantago
aquatica
a,b,d
,Caldesiaoligococca
a,b,d,e
,Arisaematriphyllum/tortuosum
a,b,c,d,e
,Gymnostachysanceps
a,b,c,d,e
,
Orontiumaquaticum
a,b,c,d,e
,Pleeatenuifolia
a,b,c,d,e
,Tofieldiacalyculata
a,b,c,d,e
,Acoruscalamus
a,b,c,d,e
.
17
Fig.3.MLphylogramofrbcLdatasetofAlismatales.MPandMLbootstrapvaluesgivenaboveandbelowthe
line,respectively.CurrentlyrecognizedtaxaofJuncaginaceaeinbold.Dash():cladenotrecoveredinMP
analysis.
TABLE4.Speciesno.,habitatandmorphologyofJuncaginaceaeandrelatedfamilies.Compiledfromvarioussources(seetextfordetails).n/anotapplicable,?
unknown.
Taxon Genera/
Species
Habitat Underground
organs
Leaves Flowers No.carpels Carpel
fusion
Placentation Ovules Endosperm
formation
Scheuchzeriaceae 1/1 Sphagnumbogs,
marshes
rhizomes,
stolons,turions
ligulate 3merous 3(6) free basalaxile anatropous,erect helobial
Aponogetonaceae 1/43 aquatic,
freshwater
rhizomes,
corms
eligulate variable 36 free basalor
marginal
anatropous helobial
Juncaginaceae
(Tetroncium)
1/1 Sphagnumbogs rhizomes eligulate 2merous 4 partly
fused
basal anatropous,erect ?
Juncaginaceae
(Triglochins.s.)
1/ca.20 freshtobrackish
water,marshes
rhizomes,
stolons,bulbs
ligulate,
auriculate
3merous 36 fused basal anatropous,erect nuclear
Juncaginaceae
(Lilaea)
1/1 aquatic,seasonal
pools
rootstocks ligulate 1merous 1 n/a basal anatropous,erect nuclear
Juncaginaceae
(Cycnogeton)
1/ca.8 aquatic,
freshwater
rhizomes,root
tubers
eligulate 3merous 36 free basal anatropous,erect ?
Juncaginaceae
(Maundia)
1/1 aquatic,
freshwater
rhizomes eligulate perianthless
or2merous
(2)4 partly
fused
apical orthotropous,
pendulous
?
Posidoniaceae 1/9 marine rhizomes ligulate perianthless 1 n/a apical orthotropous,
pendulous
helobial
Ruppiaceae 1/110 freshtobrackish
water,notmarine
rhizomes eligulate,
auriculate
perianthless (2)4(16) free apical campylotropous,
pendulous
?(helobialand
nuclear
reported)
Cymodoceaceae 5/16 marine rhizomes ligulate perianthless 2 free apical orthotropous,
pendulous
nuclear
Zosteraceae 2/14 marine,rarely
brackishwater
rhizomes ligulate perianthless 1 n/a apical orthotropous,
pendulous
?(helobialand
nuclear
reported)
Potamogetonaceae 4/102 aquatic,
freshwater,rarely
brackishwater
rhizomes,
stolons,turions
mostly
ligulate
generally4
merous
(1)4(7) free apical orthotropous(later
campylotropous),
pendulous
helobial
C
h
a
p
t
e
r
1
:
P
h
y
l
o
g
e
n
y
a
n
d
s
y
s
t
e
m
a
t
i
c
s
o
f
J
u
n
c
a
g
i
n
a
c
e
a
e
1
8
19
theMPtreeessentiallyresultedinapolytomyofTriglochins.l.,Tetroncium,theMaundia/Potamogeton
clade,Aponogeton,andScheuchzeria(Fig.2).
MTDNA DATAThe atpA data set comprised 23 taxa and 903 characters of which 209 were
variable and 141 parsimony informative (Table 3). The MP analysis resulted in ten shortest trees (CI =
0.760, RI = 0.869) of 346 steps. ML reconstruction (TVM+G, shape parameter = 0.7648, base fre
quencies 0.2934 0.2118 0.2387, rate matrix 1.7514 2.0460 0.7166 0.4594 2.0460) resulted in 1 tree.
ResolutionintheatpAdatasetwaspoor.Tetronciumisweaklysupportedassistertoahighlysupported
Triglochin s.l. (ML BS 100%, MP BS 100%). This group is part ofa large polytomy including members of
most families of core Alismatales (Fig. 2). The ML tree and the MP strict consensus tree are similar
exceptforthepositionofPotamogetonL.IntheMLtreethisgenusissistertoallothercoreAlismatales,
whileitispartofthepolytomyintheMPanalysis.
COMBINED MOLECULAR DATAThe combined cp data set (rbcL and matK) comprised 31 taxa and
2088 characters of which 804 were variable and 527 parsimony informative (Table 3). The MP analysis
resultedinoneshortesttree(CI=0.543,RI=0.672)of2075steps.TheMLreconstruction(GTR+I+G,
shape parameter = 0.8140, base frequencies 0.2953 0.1713 0.1908, rate matrix 1.4691 3.3168 0.3026
1.04913.7757)resultedinonetree.
Thethreegenedataset(rbcL,matK,andatpA)comprised19taxaand2985charactersofwhich
900 were variable and 514 parsimony informative (Table 3). The MP analysis resulted in one shortest
tree (CI = 0.651, RI = 0.685) of 1783 steps. The ML reconstruction (GTR+I+G, shape parameter =
0.7257, base frequencies 0.2890 0.1898 0.2121, rate matrix 1.6790 3.2857 0.3953 0.8581 3.8972)
resultedinonetree.
Inthetwocombineddatasets,thetopologiesoftheMPandMLtreesareidentical(Fig.2)and
differ only in branch support. As in the rbcL data set, Scheuchzeria and Aponogeton are the first
diverginglineagesinthelargersubcladeofthecoreAlismatales.Reconstructionofthetwogeneandthe
threegenedatasetdiffermainlyinthepositionofTetroncium.WhereasinthelessresolvedcpDNAdata
set Tetroncium forms a trichotomy with Triglochin s.l. and the Maundia/Potamogeton clade, the three
gene data set moderately supports Tetroncium as sister to Triglochin s.l. (ML BS 73%, MP BS 76%). In
bothdatasets,Triglochins.l.canbedividedintoT.rheophilaandahighlysupportedcladecomprising
TABLE5.ComparisonofLilaea,Triglochins.s.andtheT.proceracomplex(=Cycnogeton).?unknown.
Character Lilaea Triglochins.s. Triglochinproceracomplex

(=Cycnogeton)
Habitat seasonal
pools
marshes,oftensaline,brackishwater,
seasonalpools
freshwater
Cyanogenic + +
Chromosomeno. 2n=12 2n=12,18,24,36,48,etc.(upto144) 2n=16,32,64
Habit annual annualorperennial perennial
Leaves ligulate ligulate,auriculate Eligulate
Carpels 1 36,fused 36,free(tofused)
Carpophore n/a mostlypresent Absent
Endosperm
formation
nuclear nuclear ?
20
severalspeciesofTriglochins.s.andLilaea(MLBS100%,MPBS100%).Thissistergrouprelationshipis
highly supported (ML BS 100%, MP BS 100%). The position of Maundia as sister to a clade containing
membersoftheaquaticfamiliesisbettersupportedinthecpDNA.
MorphologicalDataSelectedmorphologicalcharactersforJuncaginaceaeandrelatedfamilies
arecompiledinTable4,andforLilaea,Triglochins.s.,andtheTriglochinproceracomplex(Cycnogeton)
inTable5.
DISCUSSION
Circumscription of Juncaginaceae and Intrafamilial RelationshipsThe circumscription of

Juncaginaceae has changed through time particularly with respect to the inclusion or exclusion of
Scheuchzeria,Lilaea,andMaundia(Table1).
SCHEUCHZERIAScheuchzeriapalustrisisararespeciesthatisnativetocooltemperateregionsof
theNorthernhemisphere,whereitgrowsinwetSphagnumbogs.ItsharesthishabitatwithTetroncium,
which has a similar ecology, but is confined to the Southern hemisphere. Even though the genus has
been included in Juncaginaceae in earlier classifications, it is now generally acknowledged (and also
supportedbyourdata)thatScheuchzeriabelongstoaseparatefamily.
Recently the family Juncaginaceae has been considered to consist of four genera, Triglochin,
Tetroncium, Maundia, and Lilaea (e.g. Haynes et al. 1998; Stevens 2001+). In this circumscription, it is
difficulttodetectanyconvincingmorphologicalsynapomorphiesforthefamily.
In this study, for the first time, all genera at some point affiliated with Juncaginaceae were
analysed together. Our results clearly show that Juncaginaceae as currently circumscribed are not
monophyletic(Figs.2,3).
MAUNDIAMaundiadoesnotgroupwiththeremainingtaxaofthefamilyinanyofouranalyses.
The genus is moderately supported as part of a clade comprising members of Posidoniaceae,
Ruppiaceae, Cymodoceaceae, Zosteraceae, and Potamogetonaceae (here called Potamogeton clade).
Some morphological characters support the close relationship between Maundia and this clade of
aquaticfamilies(Maundia/Potamogetonclade).MaundiaaswellasmostmembersofthePotamogeton
cladeshowapicalplacentationandonependulous,orthotropousovulepercarpel(e.g.Buchenau1903;
Aston1977;Dahlgrenetal.1985;Kubitzki1998;Table4).Thesecharactersandtheaquatichabitatare
listedbyStevens(2001+)aspotentialsynapomorphiesofthePotamogetonclade.Incontrasttothis,all
other Juncaginaceae have one basal anatropous ovule per carpel (e.g., Buchenau 1903; Tomlinson
1982). The flower structure of Maundia is peculiar and has led to different interpretations. Flowers of
Maundiahaveeitherbeeninterpretedtohavetwo(tofour)perianthsegments(Mueller1858;Bentham
1878;Buchenau1903;Nakai1943)ortheseorganshavebeenregardedastwobracts(Jacobs2009;H.I.
Aston, pers. comm.). When interpreted as bracts, the then perianthless flowers would constitute
another similarity to several members of the Potamogeton clade. The stamens of Maundia have been
interpretedaseither(fourto)sixsessile,bilocular(tetrasporangiate)anthers(Bentham1878;Buchenau
1903; H. I. Aston, pers. comm.) or as (eight to) 12 unilocular (bisporangiate) anthers (Mueller 1858).
Thecaearealmostseparate(probablythereasonforMuellersinterpretationasunilocularanthers),but
adnateinpairstoacommonconnective(Bentham1878;H.I.Aston,pers.comm.).Thischaracterisalso
found in members of the Potamogeton clade (e.g., Posidoniaceae and Zosteraceae; Tomlinson 1982;
Dahlgren et al. 1985; Stevens 2001+). We favour the interpretation of Maundia flowers as lacking a
perianthand possessingsixbilocularanthers,butdevelopmental studiesareneeded tofullyclarifythe
21
floralstructure.UnlikemostotherJuncaginaceae(exceptTetroncium,seebelow)whichpossesscarpels
that are free or fused but separate at maturity, the carpels of Maundia are fused (almost to the apex)
andremainunitedatmaturity.
The combined molecular and morphological evidence thus indicates that Maundia cannot be
regardedascloselyrelatedtotheremaininggenera ofJuncaginaceae.Consequently,thisgenusshould
beexcludedfromthefamily.SeveralpotentialsynapomorphiesunitingMaundiawiththefamiliesofthe
Potamogeton clade can be identified (apical placentation, one orthotropous, pendulous ovule, and
perianthless flowers). However, the exact relationships of the genus in the Potamogeton clade cannot
be determined with our data. Taxon sampling in the Maundia/Potamogeton clade is low in our
combinedanalyses.Inthethreegenedatasetonlyonememberofthedifferentfamiliesofthiscladeis
included (Potamogeton). Therefore, better sampling within this clade is necessary to resolve the
relationships of Maundia. The lack of more detailed information about Maundia (e.g., karyological,
palynological, and embryological data) and the uncertainty regarding the interpretation of the flower
structuredoesnotallowamorespecificplacementinthePotamogetoncladeeither.Noclearaffinities
to one of the other families were found and Maundia might also form a separate lineage within this
orderwhichincludesseveralmonotypicormonogenericlineages(e.g.,Scheuchzeriaceae,Butomaceae,
andAponogetonaceae).ThiswouldsupportatreatmentofMaundiaastheonlygenusofMaundiaceae
as proposed by Nakai (1943) and accepted by Takhtajan (1997). Based on the currently available
knowledge,thisclassificationisadoptedhere.
TETRONCIUMThe monotypic Tetroncium is weakly supported as sister to Triglochin s.l. in the
rbcLandthethreegenedataset.However,thissistergrouprelationshipisnotrecoveredinallanalyses.
AlthoughTetronciumandTriglochins.l.haveseveralcharactersincommon(Table4),noneofthesecan
beinterpretedassynapomorphic.Thetwogeneraareclearlydifferentinflowermorphologyandother
characters.Thus,incontrasttoTriglochinwithbisexual,trimerousflowers,Tetronciumisdioeciousand
has dimerous flowers. Carpels of Tetroncium are fused (basally to lower half) and do not separate at
maturity,whereascarpelsofTriglochins.l.areeitherfused(Triglochins.s.)orfree(T.proceracomplex)
and mostly separate at maturity. Fruits of Tetroncium are reflexed (similar to Carex pulicaris L.), a
characternotknownfromTriglochins.l.Thetwogeneraalsodifferintheirseeds.WhileinTriglochinthe
endosperm is lacking (used up) in the mature seeds (as in most core Alismatales, e.g., Dahlgren et al.
1985), seeds of Tetroncium are endospermic (Hooker 1844; Buchenau 1903). Pollen data do not
contradict a close relationship between Tetroncium and Triglochin s.l. (Grayum 1992). However, the
genera[Lilaea,Tetroncium,Triglochin]arequiteuniformpalynologically,andhardlytobedistinguished
on this basis from Potamogeton (Grayum 1992). The stiff swordshaped leaves of Tetroncium lack a
ligule (Buchenau 1903; own obs.), while leaves of Triglochin s.s. (incl. Lilaea) are ligulate or auriculate,
buteligulateintheT.proceracomplex.Thislattercharacterthusdoesnotcontradicttheplacementof
Tetroncium as sister to Triglochin s.l. Provided such relationship would be correct, it would imply that
ligulateleavesoriginatedwithinTriglochins.l.
Tomlinson(1982)correctlystatesthat littleisknownabout Tetroncium. Althoughthe currently
available data provide no unambiguous support for the relationships of Tetroncium and several
characters seem autapomorphic, we retain it as a member of Juncaginaceae. The finding of nuclear
endosperm formation would provide good support for the continued inclusion of Tetroncium in
Juncaginaceae.
TRIGLOCHINS.L.AlldatasetsrevealedahighlysupportedTriglochins.l.comprisingTriglochins.s.
with Lilaea nested inside and T. rheophila of the T. procera complex (BS 100%, Fig. 2). This clade
(Triglochins.l.)wasevenrecoveredwithmoderatesupport (MP BS83%)ina phylogenyobtainedfrom
22
theconserved5.8SrRNAgeneoftheITSregion(only163bp;S.vonMeringandJ.W.Kadereit,unpubl.
data).
The species of the T. procera complex are morphologically (Aston 1993, 1995) and molecularly
(this study; S. von Mering and J.W. Kadereit, unpubl. data) clearly differentiated from the remaining
Triglochinspecies(Table5).Potentialsynapomorphiesofthismonophyleticgroupincludethepresence
ofroottubers,thelackofaligule(unlesstheliguleoriginatedwithinTriglochins.l.asdiscussedabove),
the absence of a carpophor (carpels are free), and a chromosome base number of x = 8 (Robb and
Ladiges 1981). Also, in contrast to several species of Triglochin s.s. and Lilaea, T. procera is not
cyanogenic(Gibbs1974).
TheT.proceracomplexattimeshasbeentreatedatgenericrankasCycnogeton,whichwasfirst
described in 1838 as a monotypic genus, comprising only C. huegelii. Later, two Triglochin species (T.
linearis and T. procera) were included in Cycnogeton (Sonder 1856; Buchenau 1867). However,
CycnogetonlaterwastreatedassectionorsubgenusCycnogetonofTriglochin(Micheli1881;Buchenau
andHieronymus1889;Buchenau1903,Table1).InherrevisionoftheT.proceracomplexAston(1995,
p. 332) wrote that the tuberousrooted species form a natural grouping based on their thickened,
woody, fibrecovered rhizomes and their conspicuous storage tubers terminal on the roots. These
subterranean features are quite unlike those of other species currently placed in Triglochin and could
possibly be used as a distinguishing character applicable at generic rank. If further studies within the
familysupportedsuchagenericdistinctionthenthenameCycnogetonEndl.shouldbereinstated..
Ourmoleculardataincombinationwiththemorphologicaldistinctnessofthegroup(Table5)in
our opinion warrant the segregation of the T. procera complex (waterribbons) as Cycnogeton as
proposedbyAston(1995).
LILAEASurprisingly, Lilaea scilloides is nested within Triglochin s.s. This position is highly
supported in all data sets, with high support for a sister group relationship between Lilaea and the
Triglochinbulbosa complex(highlysupportedinall analysesexcepttheatpA dataset).Thus,Triglochin
s.s.wouldbeparaphyleticifLilaeawerenotincluded.
Lilaea has often been placed in its own family Lilaeaceae (e.g., Schumann 1894; Taylor 1909)
based on its divergent floral morphology. This enigmatic species has unisexual and bisexual flowers of
five different types (see Posluszny et al. 1986 for details). All flowers are monomerous, i.e., have only
onecarpelinfemaleflowers,onestamenandoneperianthsegmentinmaleflowers,andacombination
of both in bisexual flowers. In contrast to this, flowers of Triglochin s.s. are always bisexual and
trimerous. In spite of these striking differences in floral morphology, the two taxa share a number of
characters. For example, the vegetative habit of Triglochin s.s. and Lilaea is similar and both taxa have
semitereteleaveswithsheathandligule.Furthermore,nuclearendospermformationwasdescribedfor
both taxa (Agrawal 1952), and both have the same chromosome base number of x = 6 (Larsen 1966).
Table5summarizescharactersfoundinLilaeaandTriglochins.s.
Although several studies had recognized the close relationship between Lilaea and
Juncaginaceae (e.g., Markgraf 1936; Larsen 1966; Tomlinson 1982), the recognition of the position of
LilaeawithinTriglochins.s.probablywashamperedbytheautapomorphicdivergenceofLilaea.
The morphological characters discussed above and compiled in Tables 4 and 5 largely support
thecladesrecoveredinourmolecularphylogeneticanalyses.Eventhoughlargemorphologicaldatasets
areavailableforthemonocotyledons(e.g.,Chaseetal.1995;StevensonandLoconte1995)knowledge
of some taxa of Juncaginaceae is incomplete. Especially embryological, karyological, and palynological
dataarelacking,andmoreworkisneededtoallowacombinedanalysisofmorphologicalandmolecular
data.
23
With the removal of Maundia from Juncaginaceae, a recircumscription of the family is

necessary. In its new circumscription Juncaginaceae are characterised by having flowers in spikelike
inflorescences, nuclear endosperm formation (unknown for Tetroncium and Cycnogeton), basal
placentation, and one anatropous ovule per carpel. None of these characters can be regarded as
synapomorphic. Our results necessitate several new combinations relating to the inclusion of Lilaea in
Triglochins.s.andthereinstatementofCycnogeton.
TAXONOMICTREATMENT
ProposedNewClassificationforJuncaginaceae
JuncaginaceaeRich.,Dmonstr.Bot.9.Mai1808[asJuncagines],nom.cons.
Annualorperennialherbswithrhizomesorbulbs,sometimeswithtuberousroots(Cycnogeton),
mostly scapose. Leaves terete or flattened, sheathing, ligulate or eligulate. Inflorescence spikelike.
Flowers inconspicuous, trimerous, dimerous or monomerous (Triglochin scilloides), bisexual or
unisexual, then plants monoecious or dioecious (Tetroncium), or with bisexual and some unisexual
flowers(Triglochinscilloides).Fruitsorpartialfruitsindehiscent.
Threegenerawith30species,subcosmopolitan,mostlytemperate.Centreofspecificdiversity
inAustralia.
KEYTOTHEGENERAOFJUNCAGINACEAE
1.Plantsmonoeciousorwithbisexualflowers,leavessemitereteorflattened,ligulateor
eligulate,flowersusually3merous(rarely1merous),ofalmostcosmopolitandistribution
..............................................................................................................................................2
1.Plantsdioecious,leavesstiff,swordshaped,eligulate,flowers2merous,fromSphagnum
bogsinsouthernSouthAmerica(PatagoniaandTierradelFuego)andonsomesubantarctic
islands(FalklandIslands,GoughIsland)
.........................................................................................................................1.Tetroncium
2.Plantswithrhizomesandtuberousroots,leavesflattened,eligulate,fruitswithout
carpophore,freshwateraquaticsfromAustralasia
...................................................................................................................2.Cycnogeton
2.Plantswithbulbsorrhizomes,leavessemiterete,ligulateorauriculate,fruitsmostlywith
carpophore,plantsfrommosttemperateregionsoftheworld
.......................................................................................................................3.Triglochin

24
1.TETRONCIUMWilld.,Mag.NeuestenEntdeck.GesammtenNaturk.Ges.Naturf.FreundeBerlin2:17.
1808.
TETRONCIUMMAGELLANICUMWilld.,Mag.NeuestenEntdeck.GesammtenNaturk.Ges.Naturf.Freunde
Berlin2:17.1808.
2.CYCNOGETONEndl.,Stirp.Herb.Hgel.:22.1838(Ann.Wien.Mus.2:210.1838).
CYCNOGETONHUEGELIIEndl.,Stirp.Herb.Hgel.:23.1838(Ann.Wien.Mus.2:211.1838).Triglochin
huegelii(Endl.)Aston,Muelleria8:3:346.
CYCNOGETONLINEARE(Endl.)Sond.,Linnaea28:225.1851.TriglochinlinearisEndl.,Pl.Preiss.2:54.1848.
CYCNOGETONPROCERUM(R.Br.)Buchenau,Abh.Naturwiss.VereineBremen1:224.1867.Triglochin
proceraR.Br.,Prodr.Fl.Nov.Holland.:343.1810.
Cycnogetonalcockiae(Aston)Mering&Kadereit,comb.nov.TriglochinalcockiaeAston,Muelleria8:85.
1993.
Cycnogetondubium(R.Br.)Mering&Kadereit,comb.nov.TriglochindubiaR.Br.,Prodr.Fl.Nov.
Holland.:343.1810.
Cycnogetonmicrotuberosum(Aston)Mering&Kadereit,comb.nov.TriglochinmicrotuberosaAston,
Muelleria8:88.1993[asT.microtuberosum].
Cycnogetonmultifructum(Aston)Mering&Kadereit,comb.nov.TriglochinmultifructaAston,Muelleria
8:90.1993[asT.multifructum].
Cycnogetonrheophilum(Aston)Mering&Kadereit,comb.nov.TriglochinrheophilaAston,Muelleria8:
94.1993[asT.rheophilum].
3.TRIGLOCHINL.,Sp.Pl.:338(1753).
Triglochinscilloides(Poir.)Mering&Kadereit,comb.nov.PhalangiumscilloidesPoir.,Encycl.(Lamarck)
5:251.1804.Lilaeascilloides(Poir.)Hauman,Publ.Inst.Invest.Geogr.Fac.Filos.LetrasUniv.Buenos
Aires,A10:26.1925.
Excludedtaxa
MAUNDIAF.Muell.,Fragm.1:22.1858.
M.TRIGLOCHINOIDESF.Muell.,Fragm.1:23.1858.Triglochintriglochinoides(F.Muell.)Druce,Bot.Soc.
Exch.ClubBrit.Isles4:651.1916(publ.1917).
TriglochinmaundiiF.Muell.,Fragm.6:83.1867,nom.inval.,nom.prov.
MaundiaceaeNakai,ChosakuronbunMokuroku[Ord.Fam.Trib.Gen.Sect....nov.ed.]:213.1943
Chapter2
Phylogeny,biogeographyandevolutionofTriglochinL.(Juncaginaceae)
morphologicaldiversificationislinkedtohabitatshifts
ratherthantogeneticdiversification
SabinevonMering&JoachimW.Kadereit
ManuscriptforMolecularPhylogeneticsandEvolution
ABSTRACT
A specieslevel phylogeny is presented for Triglochin, the largest genus of Juncaginaceae

(Alismatales) comprising about 30 species of annual and perennial herbs. Triglochin has an almost
cosmopolitan distribution with Australia as centre of species diversity. TransAtlantic and trans
African disjunctions exist in the genus. Phylogenetic analyses were conducted based on molecular
data obtained from nuclear (ITS, internal transcribed spacer) and chloroplast sequence data (psbA
trnHspacer,matKgene).Basedonthephylogenyofthegroupdivergencetimeswereestimatedand
ancestraldistributionareasreconstructed.OurdataconfirmthemonophylyofTriglochinandresolve
relationships between the major lineages of the genus. The sister group relationship between the
Mediterranean/AfricanT.bulbosacomplexandtheAmericanT.scilloides(formerlyLilaeas.)ishighly
supported. This clade is sister to the rest of the genus which contains two main clades. In the first,
thewidespreadT.striataissistertoacladecomprisingannualTriglochinspeciesfromAustralia.The
secondcladecomprisesT.palustrisassistertotheT.maritimacomplex,ofwhichthelatterisfurther
dividedintoaEurasianandanAmericansubclade.DiversificationinTriglochinbeganintheMiocene
orOligocene,andmostdisjunctionsinTriglochinweredatedtotheMiocene.Taxonomicdiversityin
some clades is strongly linked to habitat shifts and can not be observed in old but ecologically
invariablelineagessuchasthenonmonophyleticT.maritima.
Keywords.Alismatales,Cycnogeton,disjunctions,dispersal/LDD,Lilaea,moleculardating

Chapter2:Phylogeny,biogeographyandevolutionofTriglochin
26

27
Introduction
Triglochin L. is an almost cosmopolitan genus with between 25 and 30 currently recognized

species (e.g., Aston, 2011; Kcke et al., 2010; Table 1, 2), making it the largest genus in the small
monocot family Juncaginaceae (Alismatales) which besides Triglochin comprises Cycnogeton Endl.
andTetronciumWilld.(vonMeringandKadereit,2010).InthestrictsenseTriglochinincludesannual
or perennial grasslike herbs characterised by sheathing leaves with ligules or auricles and spike
likeinflorescences.
Triglochinprobablyisbestknownforitscoastalsaltmarshspeciesfoundinmosttemperate
regions of the world. However, species of Triglochin occur in a wide variety of wet to dry habitats
suchasestuaries,seasonalwetlands(vernalpools,rockpools),andsemidesertinlandhabitatsfrom
sealeveltouptoca.5,000maltitude(Guoetal.,2010).
The generic circumscription of Triglochin has long been unstable (see Table 1 for details).
Followingarecentmolecularphylogeneticstudy,theTriglochinproceracomplex(thesocalledwater
ribbons; Aston, 1995, 1993) was segregated from Triglochin s. l. and reinstated as Cycnogeton Endl.
basedonaphylogeneticstudyandmorphological,karyologicalandecologicalcharacters(vonMering
and Kadereit, 2010). Furthermore, the monospecific Lilaea Bonpl. was found to be nested within
Triglochin and transferred into the genus as T. scilloides (Poir.) Mering & Kadereit. The newly
circumscribed Triglochin is highly supported as a monophyletic group (Les and Tippery, 2013; von
MeringandKadereit,2010).
DuetoitsfocusonthedelimitationofJuncaginaceaeandgenericrelationshipsinthefamily,
our previous study (von Mering and Kadereit, 2010) included only a relatively small sample of
Triglochin. Species of Triglochin are found on all continents except Antarctica, and the genus has a
distributionintemperatetoarcticandsubantarcticregions(Fig.1,Table2).Thefewspeciesgrowing
in tropical regions are mostly restricted to high elevations (e.g., T. scilloides in the Andes, T. milnei
Horn in subSaharan Africa). While several Triglochin species such as T. maritima L., T. palustris L.,
andT.striataRuiz&Pav.areverywidespread,othersareendemicsconfinedtorelativelysmallareas
(e.g., T. gaspensis Lieth & D.Lve, T. compacta Adamson, T. buchenaui Kcke, Mering & Kadereit).
Australia is the centre of species diversity with about 15 annual species (Aston, 2011; Keighery,
2003). These small, sometimes tiny annual plants are ephemerals inhabiting temporarily wet sites
such as clay pans, rock pools or soilfilled depressions and sometimes edges of lakes or swamps
(Aston,2011;Keighery,1979;Table2).
Despite its interesting distribution pattern, the biogeographic history of Triglochin has not
beenanalysedinanydetail.InastudyofdistributionpatternsCamp(1947)onlypresentedamapof
thegeneraofJuncaginaceaeandbrieflydiscussedthedistributionofTriglochinandrelatives.Hornaf
Rantzien(1961)notedthatTriglochinisofconsiderableinterestfromtheviewofphytogeography
(p. 82). He suggested an Antarctic origin of the family and its spread from this primary centre of
distributiontootherregionsinthesouthernhemisphere(Australiantemperateregion,extratropical
SouthAmerica,SouthAfrica).Hefurtherwrote:Somespeciesapparentlymigratednorthwardsfrom
these austral centres to the northern hemisphere, probably along the mountain ranges. (Horn af
Rantzien, 1961, p. 83). Based on the southern hemisphere distribution of several taxa related to
Juncaginaceae,aGondwanaoriginhasbeenproposedforthefamilybyRavenandAxelrod(1974).
Hereweinvestigatethephylogeny,biogeographyandevolutionofTriglochinusingsequence
dataobtainedfromonenuclear(nrITS,internaltranscribedspacer)andtwochloroplastDNAregions
(psbAtrnHintergenicspacer,matKgene).Themainobjectivesofthisstudyare:(1)toestablisha
28
robustphylogeneticframeworkforTriglochin,(2)toestimatedivergencetimesofitsmajorlineages,
(3) toreconstructitsbiogeographic history,and(4) toinvestigaterelationshipsbetweentaxonomic,
geneticandecologicaldiversitiesintwocladesofthegenus.
Fig. 1. Distribution of Triglochin based on selected specimen data from GBIF (http://data.gbif.org/), comple
mentedbygeoreferencedspecimendatafromarevision(Kckeetal.,2010).Duetothelackofspecimendata
fromsomecountries(e.g.,Russia)intheGBIFportal,partsofthedistributionareaineasternEuropeandAsia
aremissing.
MaterialandMethods
TaxonsamplingToresolveintragenericphylogeneticrelationships,threeDNAregions(nrDNA:ITS;
cpDNA: psbAtrnH, matK) were sequenced for this study. We sampled multiple accessions of
Triglochin representing almost all currently recognised extraAustralian species and subspecies and
almost the entire distribution range of the genus (Appendix 2). The recent Flora of Australia
treatment of Triglochin (Aston, 2011) recognises 14 annual species. Based on the assumption that
these species form a monophyletic group (see also 4.1), only four species of this group were
included.WithintheT.maritimacomplex,wesampledaccessionsfromseveralpartsofEurope,Asia
(includingT.maritimasubsp.asiaticaKitag.ofuncleartaxonomicstatus)andAmerica.Inadditionto
the two species accepted in Flora of North America (T. maritima and T. gaspensis; Haynes and
Hellquist, 2000) we also included two taxa recognised only in regional treatments from western
North America (T. concinna Burrt Davy vars. concinna and debilis (M.E.Jones) J.T.Howell, e.g., Keil,
2012). South American material of the T. maritima complex is limited to one sample of T. concinna
fromChile(occurringalsoinArgentinaandBolivia).AlthoughT.mexicanaKunthisacceptedbysome
authors(e.g.,EspejoSerenaandLpezFerrari,1996;NoveloandLot,2001),itsstatusisunresolved.
No material of this taxon was available for our study. For the widespread T. palustris and T. striata
accessions covering the entire distribution area were sampled where possible. The synonymy of
severalpublishednameswitheitherT.palustrisorT.striataissupportedbythestudyofherbarium
material and data from the literature. Our sampling of the T. bulbosa complex was exhaustive. We
included all taxa accepted in a recent revision (Kcke et al., 2010) except for T. bulbosa L. ssp.
tenuifolia(Adamson)Horn,arareendemicfromtheCaperegionofSouthAfrica.
29
For the ITS region, a dataset comprising 78 accessions was generated. This dataset was
complemented by reduced datasets for the two plastid markers used. The newly generated psbA
trnH dataset comprised 55 sequences. The matK dataset with eight sequences combined data from
our previous study (von Mering and Kadereit, 2010) with newly generated sequences of Australian
annualsandCycnogeton.Duetodifficultieswithamplificationandsequencing,somesequenceswere
notavailableforalltaxa.SinceCycnogetonishighlysupportedassistergrouptoTriglochin(Lesand
Tippery,2013;vonMeringandKadereit,2010),representativespeciesofthisgenuswereincludedas
outgroup in the cpDNA datasets. ITS sequences of Cycnogeton could not be aligned with those of
Triglochin.
DNA was extracted from herbarium specimens or from silicadried leaf samples. Voucher
informationandGenBankaccessionnumbersarecompiledinAppendix2.
Table1.CircumscriptionofTriglochinfrom1881topresent.
1
onlyTriglochinprocera,
2
only
Maundiatriglochinoides,
3
excl.Maundia,
4
excl.Cycnogeton,
5
incl.Cycnogeton,
6
incl.Lilaea,
n.s.notspecified.
Author/Source Numberofspecies
Micheli1881 Triglochin 9
sect.Eutriglochin
sect.Cycnogeton
1
sect.Pseudotriglochin
2
Buchenau&Hieronymus1889 Triglochin 12
subg.Eutriglochin
subg.Cycnogeton
1
subg.Pseudotriglochin
2
Buchenau1903
Triglochin
3
13
subg.Eutriglochin
subg.Cycnogeton
Hutchinson1934,1959
Triglochin
3,4
n.s.
Dahlgrenetal.1985
Triglochin
3,4
c.16
Takhtajan1997
Triglochin
3,4
13
Haynesetal.1998
Triglochin
3,5
12
Aston2011 Triglochin
3,5
24(Australia)
vonMering&Kadereit2010,
thisstudy
Triglochin
3,4,6
c.2530
30
DNA extraction, amplification, and sequencing The molecular work followed standard protocols.
Total genomic DNA was extracted using commercial plant DNA extraction kits (NucleoSpin,
MachereyNagel, Dren, Germany or DNeasy plant extraction kit, Qiagen, Hilden, Germany)
according to the manufacturers protocols with minor modifications. Details of PCR amplification of
allDNAregionsaresummarisedinAppendix3.PCRproductswerepurifiedwithastandardcleanup
kit(Qiagen,Hilden,Germany)followingthemanufacturersprotocolwithminormodifications.Cycle
sequencing was carried out with BigDye Terminator 3.1 (Applied Biosystems, Foster City, CA, USA)
using the same primers as for PCR amplification (Appendix 3). The fluorescently labelled samples
were sequenced at GENterpriseGenomics GmbH or StarSEQ GmbH (both Mainz, Germany). Some
sequences were run on an ABI 3130xl Genetic Analyzer or ABI 3730 automated sequencer at the
BotanicalMuseumCopenhagen,Denmark.
Sequence alignment and phylogenetic analyses Forward and reverse strands were manually
edited and assembled using Sequencer 4.1.2 or 4.10.1 (Gene Codes Corp., Ann Arbor, MI, USA).
Consensus sequences were then aligned manually using Maclade 4.1 (Maddison and Maddison,
2000).
Sequence divergence in the ITS region was too high for the unambiguous alignment of ITS
sequences of Triglochin and the outgroup Cycnogeton, and therefore midpoint rooting had to be
chosenforthisdataset.TherootchosenthiswaywasidenticaltotherootsuggestedbythecpDNA
analysisrootedwithCycnogeton. In thecombinedITS/psbAtrnHdatasetwe added the5.8Sregion
of the ITS sequence of Cycnogeton rheophilum to the alignment (in addition to the psbAtrnH
sequence)inordertoreducetheamountofmissingdataforthistaxon.
Sequence analysis revealed that the following accessions had identical ITS sequences: (1) T.
maritima from Europe incl. Russia (11 accessions), (2) T. maritima from the Caucasus region (5
accessions), (3) T. buchenaui and T. elongata from South Africa. These were represented in the
phylogenetic analyses by a single terminal taxon and are marked on the trees by origin/numbers
behind the taxon name. The same applies to identical psbAtrnH sequences generated for eight
accessions.
Phylogenetic analyses were conducted (1) separately for each dataset, i.e. ITS with a global
sample of Triglochin species, psbAtrnH with a reduced and matK with a small sample, (2) for the
combineddatamatrixofITSandpsbAtrnH,and(3)foracombineddatasetofallthreemarkers.
Separate and combined datasets were analysed using both Maximum Likelihood (ML) and Bayesian
Inference(BI).Thebestfitmodelsofnucleotidesubstitutionforeachseparateandcombineddataset
were determined using jModelTest v2.1.1 (Posada, 2008) and applying the Akaike Information
Criterion(AIC).
Maximum Likelihood (ML) analyses were performed with RAxML (Stamatakis et al., 2008)
usingtheGTR+Gmodel.Statisticalsupportwasassessedwith100bootstrapreplicates.Inthispaper,
bootstrapsupportof7584%isconsideredmoderateand85100%strong.
TheBayesianInferencetopology(BI)ofthedatasetswasfoundusingBEASTv1.7.4(Drumm
ond et al., 2012). The Bayesian analysis was carried out with a Markov Chain Monte Carlo (MCMC)
algorithm of 10 million generations, sampling every 1000 generations. A Yule tree prior was
employedinallruns.
For all BEAST analyses, convergence of resulting posterior distributions for parameter
estimates was examined using Tracer v1.5 (Rambaut and Drummond, 2007) to ensure that all
parameters had an effective sampling size (ESS) over 200. After discarding the first 10% as burnin,
maximum clade credibility trees were calculated using TreeAnnotator v1.7.4 (Rambaut and
31
Drummond, 2009). Trees were visualised and edited using FigTree v1.3.1 (Rambaut, 2006). For
conversionoffileformatsforuseindifferentanalysestheprogrammeALTER(GlezPeaetal.,2010)
wasused.
In order to compare levels of genetic differentiation between two clades (the T. bulbosa
complex and the T. maritima complex, see 3.1), pairwise genetic distances (uncorrected pdistance
and Kimura 2parameter, K2P) were calculated in PAUP* (Swofford, 2002) using alignments of
completeITSsequences.
Molecular dating Divergence times were estimated using BEAST v1.7.4 (Drummond et al., 2012)
which employs a Bayesian MCMC approach to simultaneously estimate tree topology, substitution
rates and node ages. No fossils of Juncaginaceae are known. We therefore used a secondary
calibration with an age for the split between Araceae and the rest of Alismatales obtained in an
angiospermwidedatinganalysisusingfossilcalibration(Belletal.,2010).Thissplitwasfixedat122
millionyearsago(Ma)withastandarddeviation(SD)of6Ma.AgeestimatesforJuncaginaceaefrom
a monocotwide, fossil calibrated analysis (Janssen and Bremer, 2004; 82 and 52 Ma for stem and
crown group age, respectively) could not be used as the sampling in that study did not include all
genera of the family. A dataset used earlier, including members of all genera of Juncaginaceae and
further representatives of Alismatales (based on the plastid genes rbcL and matK, see von Mering
andKadereit,2010fordetails),wasusedasinputdata.DatinginBEASTreliedonaYuleprocesstree
prior, and either a relaxed lognormal clock with a normal distribution and the GTR+G+I model or a
strictclockwasselected.MCMCchainswererunfor60milliongenerationsintherelaxedclockand
for 10 million generations in the strict clock models, with sampling every 10000th and 1000th
generation,respectively.
BecauserbcLandmatKsequenceswereavailableonlyforalimitednumberoftaxa,theage
estimateobtainedforthesplitbetweenCycnogetonandTriglochin(40Ma,SD=5)wasthenusedfor
anotherdatinganalysisbasedonthecombinedITS/psbAtrnHdataset.
Biogeographic analysis To reconstruct ancestral distribution areas of Triglochin (and Juncagina

ceae), we conducted a statistical dispersalvicariance analysis (SDIVA, based on DIVA, Ronquist,
1997)usingtheprogrammeRASPv2.1b(Yuetal.,2011;Yuetal.,2010).Becauseavailablemolecular
data either could not be aligned for all three genera of Juncaginaceae (ITS) or the sampling of
relevanttaxawaslimited(rbcL,matK,psbAtrnH),weusedoneconsensustreeofallthreegeneraof
Juncaginaceae based on different datasets as input. Only groups which were fully resolved in our
phylogeny were included in the topology (and reduced to one representative terminal) to provide
the required completely bifurcating tree. Therefore, the SDIVA analysis could not calculate
optimisedareasoverasetoftreesandthusdidnottakeintoaccounttopologicaluncertainty.
The following six geographic areas were defined: (A) Australia and New Zealand, (B) Eurasia
(excl. Mediterranean region), (C) Mediterranean region, (D) subSaharan Africa, (E) North America
and(F)SouthAmerica.TheAustralianCycnogeton(withC.multifructumandC.rheophilum)andthe
SouthAmericanTetronciumwereincludedinouranalysesandcodedaccordingly.
The maximum number of possible areas at each node (maxareas) was set to two. This
reflects that most species of Triglochin occur in only one or two areas with only three species (T.
maritima, T. palustris and T. striata) being more widespread. Additional runs were performed to
exploretheeffectsofareaconstraints,onethatallowedmaximallyfourareasfortheancestralrange
ofacladeandonethatdidnotconstrainthenumberofareas.
32
Results
Phylogenetic analyses The final ITS matrix (after removal of identical sequences) comprised 58
terminals, and the alignment was 668 bp long. Monophyly of most species groups and widespread
species is highly supported in our analysis (Fig. 2). A clade comprising Triglochin scilloides plus the
monophyleticT.bulbosacomplexishighlysupportedassistertotheremainderofthegenus.Within
the latter, two main clades are found: (1) Triglochin striata is moderately supported as sister to a
clade comprising Australian annual Triglochin, and (2) Triglochin palustris plus the T. maritima
complex, the latter consisting of two highly supported clades comprising Eurasian and American
accessions,respectively.
The psbAtrnH dataset (after removal of identical sequences) comprised 45 accessions (plus two
outgroup taxa), and the alignment was 488 bp long. Phylogenetic analyses resulted in several well
supported clades corresponding to species or species complexes. Relationships among these were
essentiallyunresolvedduetolowsequencevariation(resultsnotshown).
Figure 3 compares trees obtained from a reduced ITS dataset and from a small combined
(matK and psbAtrnH) cp dataset (eight accessions incl. Cycnogeton rheophilum as outgroup). The
inferredtopologieswereidenticalandfullyresolvedinbothdatasets,withhighsupportfoundforall
clades.
Equally,resultsfromtheITSandpsbAtrnHanalysesdidnotshowanysupportedtopological
conflictandwerecombined.ThecombinedmatrixofITSandpsbAtrnHcomprised35accessionsand
had 1216 nucleotide positions. Phylogenetic analyses resulted in the following topology (Fig. 4): a
highlysupportedcladecomprisingT.scilloidesandtheT.bulbosacomplexissistertotherestofthe
genuswhichcomprisestwomainclades,(1)amoderatelysupportedcladecomprisingT.striataand
AustralianannualTriglochin,and(2)amoderatelysupportedcladeoftheT.maritimacomplexandT.
palustris.
Pairwisegenetic distancesfortheT. bulbosa complexareslightly higher comparedtotheT.
maritima complex. Values for the Kimura 2parameter and uncorrected pdistances (mean and
maximumdistances)aresummarisedinTable3.
Molecular dating analyses Divergence time estimates (mean ages with confidence intervals,
highestposteriordensity=HPD)fordifferentbiogeographicallyimportantnodesaresummarisedin
Table 4 based on relaxed and strict clock models and two different datasets. The relaxed and the
strict clock models resulted in considerably different age estimates, with the strict clock model
yielding older ages. Age estimates using the ITS/psbAtrnH dataset resulted in even older ages.
Differences in age estimates were so large that inferred ages sometimes lay in different geological
timeperiods.Theresultsofthestrictclockarepreferredherebecauseonlyoneconstraintwasused
toobtainthem.However,otherageestimateswillbeconsideredinthediscussionwherenecessary.
Using the strict clock model, the split between Triglochin and Cycnogeton occurred around
53.8 Ma (95% HPD: 45.262.4 Ma; Fig. 5), i.e. in the Eocene or Palaeocene. Diversification of
TriglochinbeganintheOligocene(orlateEocene),withmostoftheextantspeciesdiversitydatingto
theOligoceneorMiocene.Achronogram(Fig.4)showsthemeanageswith95%HPD.
33
Fig.2.PhylogenyofTriglochinbasedonITSsequences.MaximumlikelihoodbootstrapandBayesianposterior
probabilityvaluesareshownalongthebranches.

34
ITS cpDNA
(matK+psbAtrnH)
Fig.3.PhylogenyofTriglochin.TopologiesoftheITSdataset(left;midpointrooting)andacombinedcpdataset
(right)ofmatKandpsbAtrnHsequences(Cycnogetonrheophilumasoutgroup).Maximumlikelihoodbootstrap
andBayesianposteriorprobabilityvaluesareshownaboveandbelowthebranches,respectively.

35
Fig.4.PhylogenyofTriglochin.ChronogramofTriglochinobtainedfromBEASTusingthecombinedITSand
psbAtrnHdataset.Bootstrapandposteriorprobabilityvaluesareshownalongthebranches.Bluebarsshow
the95%highestposteriordensity(HPD)credibilityintervalsofmeanageestimates.Thelettersindicatenodes
asusedinTable3.
Fig.5.ChronogramforJuncaginaceaeinferredfromaBayesiandatinganalysisusingcpDNAdata(rbcLand
matK)andBEASTmodeledunderastrictclockmodel.Meanagesaregiveninmillionyears(Ma)forselected
nodes.Bluebarsrepresentthe95%highestposteriordensity(HPD)credibilityintervalsofmeanageestimates.
OligoOligocene,PalPalaeocene,PlPliocene.
A
B
C
F
D
G
36
BiogeographicanalysesResultsoftheDIVAanalysesareshowninFig.6.DIVAanalysiswith
maxarea=2inferredthattheancestraldistributionareaofTriglochinisacombinedareaofAustralia
andAfricaorAustraliaand(NorthorSouth)America.ApartfromtheinclusionofNorthAmerica,this
suggestsanancestralareaintheSouthernhemisphere(Fig.6).TheancestorofTetroncium,
CycnogetonandTriglochin,i.e.oftheentirefamily,isinferredtohavebeendistributedinAustraliaor
SouthAmerica.Whenfourareasareallowedateachnode(maxareas=4),Australia,Africaor
AmericawerereconstructedastheancestralareaofJuncaginaceae.
Fig.6.DispersalvicariancescenariosforTriglochinreconstructedwithRASP.Optimizationwiththemaximum
numberofareaunits(maxareas)settotwo.AAustralia,BEurasia,CMediterraneanregion,DAfrica,E
NorthAmerica,FSouthAmerica.
37
Discussion
PhylogeneticrelationshipsinTriglochin
Our study provides a robust phylogeny of Triglochin (Figs. 24) by resolving relationships
betweenallspecies groupsofthegenusandmostextraAustraliantaxa. With regardtothegeneral
topology,resultsfromearlierstudiesbasedonrbcL(LesandTippery,2013;vonMeringandKadereit,
2010)wereconfirmed,andmuchbettersupportwasfoundforrelationshipsamongclades.
T. scilloides/T. bulbosacomplex The inclusion of the formerly separate genus Lilaea in Triglochin
(von Mering and Kadereit, 2010) and its sister group relationship to the T. bulbosa complex was
confirmed by analyses of all datasets. For a detailed discussion of the inclusion of this species in
TriglochinseevonMeringandKadereit(2010).Noobvioussynapomorphiesforthecladecomprising
T. scilloides and the T. bulbosa complex could be detected. Within the T. bulbosa complex, which is
characterisedby(narrowly)ovoidfruitsconsistingofthreefertilecarpelsthatseparateatmaturity,a
highlysupportedcladeofT.milneiandT.compactaissistertotherestofthisgroup(Fig.2,4).These
twospeciesaremorphologicallysimilartoeachother(e.g.,outlineofseedshowsthroughmericarp
surface) and share the presence of outgrowths at the base of the carpophores (Kcke et al., 2010).
However,theyaregeographically,ecologicallyandphenologicallyclearlydifferentiated(Kckeetal.,
2010). Relationships among the remaining taxa are not fully resolved, and our results suggest that
some further adjustments to taxon circumscription in the T. bulbosa complex might be necessary
(Kckeetal.,2010),especiallywithregardtoT.bulbosawhichappearstobeparaphyletic(Fig.2,4).
The three sampled subspecies of T. bulbosa (material of subsp. tenuifolia was not available) do not
formamonophyleticgroupbecausethetwoMediterraneanspeciesT. barrelieriandT.laxifloraare
nested in this otherwise South African clade. A sample of naturalised T. bulbosa from Western
Australia collected close to an Air Force base (Aston, 2011) is highly supported as sister to an
accession from South Africa, indicating its African origin. South Africa has been identified as an
importantsourceofAustraliasalienspecies(ScottandPanetta,1993).
T. striata/Australian annual Triglochin Triglochin striata is recovered as sister to the lineage of

Australian annual species in all datasets, with high support in the combined datasets (Fig. 2, 4).
However, support for this sister group relationship is only moderate in the ITS dataset. No obvious
synapomorphies for this clade could be detected. An analysis of the relevant literature (e.g., Aston,
2011;Romanowski,1998;Thieret,1988)andextensiveherbariummaterialshowsthatT.striataisa
widespreadandmorphologicallyvariabletaxon.Numerousformshavebeendescribedfromdifferent
parts of the world which differ in plant size and form or length of leaves (e.g., Gandoger, 1919).
Various authors have treated these forms at different taxonomic levels as forms, varieties,
subspeciesorevenspecies.However,currentlyallthesetaxaareregardedassynonymsofT.striata
(Buchenau, 1882; Goverts, 2013). As long as no global revision of T. striata is available, we treat T.
striataasonepolymorphicspecies.AnnualTriglochininAustraliaisarelativelyspeciesrichgroup(14
spp. recognized by Aston, 2011, several more expected, e.g., Keighery, 2003). Apart from their
annuallifeformandsmallsize,thesespeciessharecharacterssuchasfiliform,auriculateleavesand
fruitswiththreefertileandthreesterilecarpels(Aston,2011;S.vonMering,pers.obs.).Preliminary
molecular results for a larger number of species (S. von Mering et al., unpubl. data) suggest that
AustralianannualTriglochinisamonophyleticgroup.
Table2.Distribution,habitat,lifeform,floweringtimeandchromosomenumbersofTriglochin,CycnogetonandTetroncium(Juncaginaceae).Compiledusing
thefollowingmainreferences:Aston,1993,1995,2011;Guoetal.,2010;HaynesandHellquist,2000;Goldblatt&Johnson,1979+;Keil,2012;Kckeetal.,
2010;vonMeringandKadereit,2010;vonMering,2013;Tzvelev,2013.
Taxon Distribution Habitat Lifeform Floweringand

fruitingtime
Chromosome
number
Triglochinbulbosa complex
TriglochinbarrelieriLoisel. Mediterraneanregion,reaching
theAtlanticcoastofNWFrance
andMorocco
coastalsaltmarshesandalonglagoons, less
frequentlyoncoastalrocksexposedtosaltspray,in
depressionsincoastaldunesorincorkoakforests
perennial spring(Feb.
May)
2n=30,32,
36
TriglochinbuchenauiKcke,
Mering&Kadereit
SouthAfrica restrictedtolower(periodicallyflooded)tidalzoneof
saltmarshestuaries
perennial throughoutthe
year
?
Triglochinbulbosa L. SouthAfrica variable(seesubspecies),rangingfromtemporary
poolstodry(shaded)inlandhabitats,butneverin
habitatsunderdirectinfluenceoftidalflooding
perennial (early)spring
(Jul.Nov.)
?
Triglochinbulbosassp.bulbosa SouthAfrica temporarywetlands ofcoastallowlands,e.g.,edges
ofvernalpools,ditchesandotherdepressions;on
heavyclayeytolighterclayeysandysoils;oftenin
clayfilleddepressionsamongstabilisedsandydunes;
alt.0100(500)m
spring(Aug.
Nov.)
?
Triglochinbulbosassp.calcicola
Mering,Kcke&Kadereit
SouthAfrica onlimestoneoraeoliniteincoastalregions spring(Aug.
Nov.)
?
Triglochinbulbosassp.quarcicola
SouthAfrica restrictedtosalinequartzfieldsoftheKnersvlakte,
especiallydryslopesofkoppies,coveredwithalayer
ofwhiteangularquartzgravel
earlyspring(Jul.
Aug.)
?
Triglochinbulbosassp.tenuifolia
(Adamson)Horn
SouthAfrica sheltered,shadyinlandhabitatsonmountainslopes
oftheCapePeninsulaandLangebaanPeninsula;
restrictedtocoarse,sandysoilsderivedfromgranites;
alt.below400m
spring(Jul.
Sep.)
?
Triglochincompacta Adamson SouthAfrica deep,mainlynutrientpoorsands;inlandsanddunes
(alwaysoutsidecurrentdirectinfluenceofthesea)
perennial autumn(mostly
Mar.May)
?
C
h
a
p
t
e
r
2
:
P
h
y
l
o
g
e
n
y
,
b
i
o
g
e
o
g
r
a
p
h
y
a
n
d
e
v
o
l
u
t
i
o
n
o
f
T
r
i
g
l
o
c
h
i
n
3
8
Triglochinelongata Buchenau SouthAfrica upperandsupratidalzonesofestuarinesaltmarshes

andestuarinerivers;coastalhabitats(elevatedbanks
ofsandybeaches,rocksexposedtosaltspray);some
inlandlocalities
year
?
TriglochinlaxifloraGuss. Mediterraneanregion (France,
Italy,Portugal,Spain,Cyprus,
Greece,Malta,Algeria,Morocco)
saltmarshesandalong
lagoons,butalsoincorkoakorpineforestsandin
Mediterraneantemporalpools,sometimesin
limestoneareas;alt.0500(1100)m
perennial autumn(mainly
Sep.Nov.)
2n=18
Triglochinmilnei Horn subSaharanAfrica(Angola,DR
Congo,Malawi?,SouthAfrica,
Tanzania,Zambia,Zimbabwe)
seasonal(summerwet)wetlandsandseasonallywet,
sometimesburnt,grasslands;alt.above(500)1000
m,inTanzaniaupto1900m
perennial spring summer
(Oct.Feb.)
?
Triglochinmaritima complex
1,2

Triglochinmaritima L.
3
widespreadinEurasiaandthe
Americas
coastalsaltmarshes,inlandsaline,brackish,alkaline
marshes,moistalkalinemeadows,seeps,mudflats,
streamandlakemargins;alt.05200m
perennial spring autumn 2n=12,24,
36,48,96,
120,144,156
Triglochingaspensis Lieth&
D.Lve
NENorthAmerica tidalsaltwatermarshes,regularlysubmerged perennial summer(Jul.
Aug.)
2n=96
Triglochinpalustris L. widespreadinEurasiaandthe
Americas(includingGreenland);
alsoinNewZealand
coastalandinlandmarshes,moistalkalinemeadows,
wetflats,streamandlakemargins;alt.04500m
perennial summer
autumn
2n=24
Triglochinscilloides(Poir.)Mering
&Kadereit
Canada,USA;Mexico;Argentina,
Chile,Ecuador,Peru;introduced
inSEAustralia,IberianPeninsula
shallowwaterandadjacentmudflats,seasonalpools,
stream,pondandlakemargins;alt.01700m
annual spring summer 2n=12
Triglochinstriata Ruiz&Pav. Australia,NewZealand,French
Polynesia;Africa;USA,Mexico,
Caribbean;SouthAmerica;
introduced:IberianPeninsula
coastalalkaline,brackishtofreshwatermarshes,
interduneswamps,saltmarshes,tidalestuaries,
streamandlakeedges,inlandsprings
year(but
seasonalin
individual
regions)
2n=24
C
h
a
p
t
e
r
2
:
P
h
y
l
o
g
e
n
y
,
b
i
o
g
e
o
g
r
a
p
h
y
a
n
d
e
v
o
l
u
t
i
o
n
o
f
T
r
i
g
l
o
c
h
i
n
3
9
AustralianannualTriglochinspp.
Triglochinsp.A(FloraofAustralia) Australia dampsandloamandclaysoilsofintermittently
floodeddepressions,swamps,boreoverflowsand
watercourses,includinggibberplainsandgranite
rockholes;waterfreshtosometimesslightlysaline
annual Jul.Oct.;Apr.
Jun.(Feb.)[N.T.]
?
Triglochincalcitrapa Hook. Australia ondamp,winterwetsandyclay,salinesoilsofflats,
claypansandlakeedges
annual Aug.Oct. ?
Triglochincentrocarpa Hook. Australia salinesandsorclayedgingsaltlakes annual lateJul.Oct. ?
Triglochinhexagona J.M.Black Australia indamptosaturated,semisaline,sandtoclaysoilsof
ephemerallakesandfloodplains(rarelyedging
freshwaterswamps)
annual (Feb.)Apr.Sep.
[N.T.,W.A.],
(Sep.)Oct.Apr.
[otherStates]
?
Triglochininsingiana (J.M.Black)
Aston
Australia indampsand tosandyclaysoilsorshallowwaterof
lakeandwatercourseedgesorephemerallyflooded
depressions;alsolimestonecoastalsandsanddamp
shallowsoilsofrockoutcrops;chieflyfreshwater
situations,sometimessemisaline
annual Mar.,Jun.Oct. ?
Triglochinlongicarpa (Ostenf.)
Aston
Australia indampsand,loamandclaysalineorsubsalinesoils,
usuallyedgingsaltlakesandseasonallyflooded
claypans;alsoinshallowsoilsongraniteoutcrops
annual (Jul.)Aug.Oct.
(Nov.)
?
Triglochinminutissima F.Muell. Australia usuallyinsalinetosubsalinesandsorsandyclays
borderingephemeralwatersofsaltpansorlakes,or
insubsalinepoolsingraniterocks,rarelyinheathland
swamps;alsocoastalcliffs
annual Aug.Oct.,Dec. ?
Triglochinmucronata R.Br. Australia ondampsalinesandorsandyclaysoils,occ.loam;
usuallycoastaloraroundthemarginsofsaltlakesor
inshallowseasonallyfloodeddepressions
annual Jul.Nov.[W.A.],
(Aug.)Sep.
Nov.(Dec.)
[easternStates]
?
C
h
a
p
t
e
r
2
:
P
h
y
l
o
g
e
n
y
,
b
i
o
g
e
o
g
r
a
p
h
y
a
n
d
e
v
o
l
u
t
i
o
n
o
f
T
r
i
g
l
o
c
h
i
n
4
0
Triglochinmuelleri Buchenau Australia coastalplaininsand,loamorclayofwinterwetflats

andswamps,oredgingdrainagelinesandcreeks;
salinetofresh
annual Sep.Nov. ?
TriglochinnanaF.Muell. Australia mesophyticorephemerallyfloodedfreshwatersites
(claypans,creekedges,swamps,roadsideseepages,
shallowdepressions),sometimesedgingsaltlakesor
subsalineswamps
annual Aug.Dec.[se
areas],Apr.
Dec.elsewhere
?
Triglochinprotuberans Aston Australia recordedfromacreekedge,anAcacia woodlandflat,
andfrequentlyfromsoilsinephemeralrockpoolsof
graniteoutcrops,oncefromsalineclay;soilsofred
loamandgrittyredbrownclayrecorded
annual Aug.Oct. ?
Triglochinstowardii N.E.Br. Australia saltaffectedclayflatadjacenttosaltlake,salt
impactedMelaleucalateriflorashrubland,aclayflat
inEucalyptuswandoowoodland,flatsinanopenwet
reedyarea,aroadsideditch,andfromthebaseofa
quartzitehill
annual Sep.Oct. ?
Triglochintrichophora Neesex
Endl.
Australia chieflyinsandsandcalcareoussoilsoflimestone
bedrockandcoastaldunes,inmoistsitesincl.mossy
seepagesandsprings
annual Jul.Dec. ?
Triglochinturrifera Ewart Australia insoakagesordamptowetsoilsnearrecedingfresh
waterofswamps,pools,smalldepressionsor
temporarystreams
annual Aug.Nov. ?
CycnogetonEndl.
Cycnogetonalcockiae (Aston)
Mering&Kadereit
Australia usuallyinfresh,still,clearwaterto30cmdeepin
ephemeralswamps;occasionalinsmallstreams,
stagnantwaterorcoastalbrackishhabitats
perennial Sept. Mar. ?
C
h
a
p
t
e
r
2
:
P
h
y
l
o
g
e
n
y
,
b
i
o
g
e
o
g
r
a
p
h
y
a
n
d
e
v
o
l
u
t
i
o
n
o
f
T
r
i
g
l
o
c
h
i
n
4
1
Cycnogetondubium(Aston)
Mering&Kadereit
Australia,NewGuinea usuallyinsaturatedsoilsorstillephemeralfresh
waterto50cmdeep,inswamps,creeksor
floodplains;alsoinwaterto2mdeeporstrong
flowingstreamsinthetropics
perennial (Sep.)Nov.
Mar.[N.S.W.,
Vic.];
throughoutthe
yearelsewhere
?
Cycnogetonhuegelii Endl. Australia fresh,stilltoflowingwaterto1mdeep,edging
permanentlakesandstreamsorsometimesin
ephemerallyfloodedsites
perennial (Jul.)Aug.
Dec.(Jan.)
?
Cycnogetonlineare (Endl.)Sond. Australia fresh,usuallystillandephemeralwaterto40cm
deep,inswamps,ditchesandlowlyingfloodland;
occasionallyinsmallstreamsandcreekpools
perennial (May)Jul. Oct.
(Dec.)
?
Cycnogetonmicrotuberosum
(Aston)Mering&Kadereit
Australia nearcoastalregions,infresh,stillorslowflowing
waterto50(120)cmdeep,insmallcreeks,swamps
andfarmdams,onsilt,peat,loamorsandsoils;
stagnantwateroftenhighlyeutrophicandhumic
fromfarmlandrunoff;alt.c.3100(400)m
perennial (Aug.)Nov.
Apr.(May)
2n=32
Cycnogetonmultifructum(Aston)
Mering&Kadereit
Australia fresh,stilltoslowflowingwaterto70cmdeepin
seasonaltopermanentswamps,irrigationchannels,
ditches,lagoonsandcreeks,mostlyonsandytoheavy
greyclays;alt.usually<200m,rarelyto1050m
perennial Aug. May(
Jun.)[Vic.,s.
N.S.W.];Apr.
Sep.[n.tropics]
2n=16
Cycnogetonprocerum (R.Br.)
Buchenau
Australia stilltoslowflowingfreshwaterto2mdeepmostlyin
permanentswamps,lagoonsandstreams,but
withstandsstrandingforextensiveperiods
year,mainly
Sep.Mar.
?
Cycnogetonrheophilum (Aston)
Mering&Kadereit
Australia clearflowingfreshwaterto1mdeepinpermanent,
oftenrockystreamsandriversusuallysubjectto
severeflooding;recordedtoc.970malt.
perennial Aug. Feb. 2n=64

C
h
a
p
t
e
r
2
:
P
h
y
l
o
g
e
n
y
,
b
i
o
g
e
o
g
r
a
p
h
y
a
n
d
e
v
o
l
u
t
i
o
n
o
f
T
r
i
g
l
o
c
h
i
n
4
2
TetronciumWilld.
TetronciummagellanicumWilld. southernSouthAmerica
(Argentina,Chile,FalklandIslands);
GoughIsland
differentpeatlandtypes,especiallySphagnum and
cushionbogs
perennial summer(mainly
Dec.Jan.)
?
1
unresolvedcomplexrequiringaglobalrevision;
2
taxonomicstatusofT.mexicanaKunthunclear;
3
includingssp.asiaticaKitag.andT.concinnaBurrtDavyvars.concinnaand
debilis(M.E.Jones)J.T.Howell;statusofthesetaxaneedsreevaluation.N.S.W.NewSouthWales,N.T.NorthernTerritory,Vic.Victoria,W.A.WesternAustralia.
C
h
a
p
t
e
r
2
:
P
h
y
l
o
g
e
n
y
,
b
i
o
g
e
o
g
r
a
p
h
y
a
n
d
e
v
o
l
u
t
i
o
n
o
f
T
r
i
g
l
o
c
h
i
n
4
3
44
T.palustris/T.maritimacomplexThelineage comprisingT.palustrisandtwocladescomposedof
membersoftheT.maritimacomplex(EurasianandAmericanmaterial)ishighlysupportedinallour
analyses. Relationships among the three highly supported clades are basically unresolved. A sister
group relationship of T. palustris to the T. maritima complex is only weakly supported in the
combined ITS/psbAtrnH dataset. There are good morphological characters clearly distinguishing T.
palustris from the T. maritima complex. All members of the T. maritima complex have ovoid fruits
composed of six fertile carpels that fully separate at maturity (e.g., Correa, 1969; Haynes and
Hellquist, 2000). Presence of six fertile carpels is a diagnostic character not found in any other
Triglochin species and a synapomorphy for the complex. In contrast, T. palustris is characterised by
narrowlyovoid,clubshapedfruitswhicharecomposedofthreefertileandthreesterilecarpels.The
fertile carpels split from the base of the fruit and spread widely, thus resembling the eponymous
arrow head (Juncaginaceae = arrowgrass family). In accordance with several other authors (e.g.,
Jafri,1973;Keil,2012;Markgraf,1981),ourobservationsclearlyshowthatthecarpelsusuallyremain
distally united with the carpophore. However, carpels may separate completely in the second year
(M. Remizowa, pers. comm.). Furthermore, T. palustris differs from members of the T. maritima
complex by having slender, fugacious stolons with terminal bulbs which secure reproduction and
hibernation(e.g.,Buchenau,1903;Markgraf,1981).
Wide distribution, morphological variation and polyploidy in the T. maritima complex have
resultedintherecognitionofseveraltaxaatdifferenttaxonomiclevelsandinchangingclassifications
(e.g.,FordandBall,1988;LveandLve,1958).Severalauthorshavestressedtheneedforaglobal
study of the T. maritima complex (Lve and Lve, 1958; Reveal, 1977; Thieret, 1988). A numerical
studyofmorphologicalvariationinthecomplex(FordandBall,1988)concludedthattwospecies,T.
maritimaandT.gaspensis,shouldberecognised.However,thisstudywasbasedonspecimensfrom
eastern and central North America and Europe and thus did not include western North American
material sometimes recognized as a distinct taxon (e.g., Keil, 2012). In the Flora of North America
(Haynes and Hellquist, 2000) only T. maritima and T. gaspensis are accepted, while regional
treatmentsfromwesternNorthAmericarecogniseT.concinnawithtwosubspecies(e.g.,Keil,2012).
In the Panarctic Flora (Tzvelev, 2013) it is noted that Triglochin maritima s. lat. is an unresolved
aggregate of probably several taxa. Our molecular results indicate that the T. maritima complex is
divided into two highly supported geographical clades, a Eurasian and an American clade,
respectively. There are two possible options to deal with the taxonomic implications of this result.
Either all taxa of the T. maritima complex are included in a polymorphic T. maritima with several
intraspecifictaxa.Alternatively,severaldistinctspeciescouldberecognised.

45
Table3.PairwisegeneticdistancesintheT.maritimaandT.bulbosacomplexes.K2PKimura2
parameterdistance,pdistanceuncorrectedpdistanceordissimilarity.
Table4.EstimateddivergencetimesforselectednodesasusedinFigs.4and5.Meanagesarein
millionyears(Ma)andvaluesgiveninbracketsarethe95%HPD(Ma).SDstandarddeviation.
Node Description rbcL/matK rbcL/matK ITS/psbAtrnH

Relaxedclock Strictclock Relaxedclock
A Juncaginaceae(Tetronciumvs.
rest)
44.1(67.427.4) 70.2(80.460.4) n.a.
B Cycnogetonvs.Triglochin 39.6(54.019.6) 53.8(62.445.2) fixedat40.0
(SD=5/6)
C T.scilloides/T.bulbosa
complexvs.rest
21.6(36.311.3) 30.6(36.125.0) 38.8(46.730.8)
D T.scilloidesvs.T.bulbosa
complex
16.3(25.45.9) 24.9(30.019.7) 26.1(36.516.7)
E T.bulbosacomplex(T.
milnei/T.compactavs.rest)
n.a. n.a. 13.6(20.27.4)
F T.striata/Australianannuals
vs.T.palustris/T.maritima
complex
18.8(22.53.2) 16.5(20.811.7) 30.3(39.721.0)
G T.palustrisvs.T.maritima
complex
4.1(13.71.2) 9.8(13.36.4) 15.5(24.97.9)
H T.maritimacomplex(Eurasian
vs.Americanclade)
n.a. n.a. 10.0(16.44.3)
Biogeography
The ancestral distribution of Triglochin was inferred to be Australia and (North or South)
America or Australia and Africa. Apart from the inclusion of North America this supports the
hypothesisthatTriglochinoriginatedinthesouthernhemisphereandspreadfromtheretothenorth
as suggested by earlier authors (e.g., Camp, 1947; Horn af Rantzien, 1961). Ancestral area
reconstruction suggested an origin of Juncaginaceae in Australia or South America. The split of
Tetroncium from Cycnogeton plus Triglochin in the Late Cretaceous/Paleocene (mean: 70 Ma, 95%
Pairwisegenetic
distance
T.maritima
complex
T.bulbosa
complex
MeanK2Pdistance 0.045 0.055
Max.K2Pdistance 0.079 0.087
Meanpdistance 0.043 0.053
Max.pdistance 0.075 0.082
46
HPD: 8060 Ma under the strict clock model; mean 45 Ma, 95% HPD: 6727 Ma under the relaxed
clock model; Fig. 5, Table 4), which essentially confirms an estimated divergence time between
Tetroncium and Cycnogeton of 51.7 Ma (based on ITS sequences) by Les et al. (2003), is consistent
with a distribution of the ancestor of this lineage from South America to Australia via Antarctica.
ConnectionsbetweenthoseareasexisteduntiltheopeningoftheDrakePassageandtheTasmanian
Gateway which separate Antarctica from South America and Australia, respectively. Despite some
controversyabouttheexacttimingoftheopeningoftheDrakePassage(estimatedtobetween42to
17 Ma; e.g., Lyle et al., 2008; Scher and Martin, 2006; Upchurch, 2008, and references therein), it
probably opened near the Eocene/Oligocene boundary. The opening of the Tasmanian Gateway is
largely considered to have occurred around 34 Ma (e.g., McLoughlin, 2001; Stickley et al., 2004;
Upchurch, 2008) resulting in the isolation of the Australian continent from other landmasses. Thus,
landconnectionsbetweenAustralia,AntarcticaandSouthAmericaallowingexchangeofbiotacould
have existed in the Late Eocene and Early Oligocene and might explain the disjunction between
Tetroncium(SouthAmerica)andCycnogeton(Australia).
Assumingasouthhemisphericaloriginofthefamily,thenorthernhemispherewascolonised
at least four times in the following clades: (1) T. maritima complex/T. palustris, (2) T. striata (apart
fromitswidedistributioninthesouthernhemispherethisspeciesisalsonativetotheCaribbeanand
southern and western North America), (3) T. barrelieri and T. laxiflora, both distributed in the
Mediterranean region (these two species may have reached the Mediterranean region
independently, see below), and (4) T. scilloides, which is distributed in the Americas from Chile to
Canada.
Diversification in Triglochin apparently started in the Late Eocene/Early Oligocene and
intensifiedintheMiocene(Table4;Fig.4,5).SeveraldisjunctionsexistinTriglochin,someofwhich
areTransAtlantic(T.scilloidesvs.T.bulbosacomplex,T.maritimaEurasianvs.Americanclade)and
sometransAfrican(T.bulbosa complex).Possibleexplanationsincludeancientisolation(vicariance)
andmorerecenteventsoflongdistancedispersal(LLD)acrossseabarriers.
T.scilloides/T.bulbosacomplexTriglochinscilloides,anativeofwesternNorthandSouthAmerica,
diverged from its sister clade, the T. bulbosa complex from the Mediterranean region and sub
Saharan Africa, in the Oligocene or Miocene (Fig. 5, Table 4). The split between the two lineages
occurredwellafterthepresumedseparationofAfricaandSouthAmericawhichhasbeensuggested
tohavetakenplacebetween11080MaintheCretaceous(e.g.,McLoughlin,2001;Upchurch,2008).
Therefore, Gondwanan vicariance is a highly unlikely explanation for this transAtlantic disjunction
which probably is the result of LLD, possibly by sea water (e.g., Renner, 2004; Won and Renner,
2006).RelationshipsbetweenAmericanandAfricancladeshavebeenreviewedbyChristenhuszand
Chase (2013), emphasizing the importance of dispersal for this type of relationship. TransAtlantic
sistercladeshavealsobeenstudiedbyLavinetal.(2004)andRenner(2004).
Diversification of the T. bulbosa complex is likely to have started in the Miocene. The sister
grouprelationshipof T.milneiand T. compactato therestofthecomplexsuggestsanoriginof the
group in subSaharan Africa, possibly in South Africa. This is consistent with the results of our DIVA
analysis.OurdatapartlysuggesttwocolonisationsoftheMediterraneanregionfromSouthAfrica,by
theautumnfloweringT.barrelieriandthespringfloweringT.laxiflora,butrelationshipsarenotfully
resolved. The disjunction between subSaharan Africa and the Mediterranean region could either
have been established by longdistance dispersal (e.g., by migratory birds resting in wetlands) or by
migrationthroughthesocalledaridcorridorconnectingsouthwesternandnortheasternAfrica(e.g.,
Jrgens,1997andreferencestherein)andvicariancefollowingitsdisruption(e.g.,Galleyetal.,2007;
47
Thiv et al., 2011). Different age estimates exist for the disruption of the corridor, some dating it to
the Pleistocene, others to the Miocene (e.g., Bobe, 2006; Schuster et al., 2006). Triglochin is absent
from some regions of the corridor such as northeastern Africa and occurs only further north in the
Mediterranean region (T. barrelieri, T. laxiflora) and further south (T. milnei in parts of Central and
EastAfrica).Thispatterncouldbetheresultofextinction,andT.milneiastheonlymorewidespread
African species (South Africa to Angola and Tanzania; Kcke et al., 2010) might be a relic of a
formerly wider distribution of the group. Our data do not allow us to decide whether this trans
Africandisjunctionistheresultoflongdistancedispersalorvicariance.
Diversification of the T. bulbosa complex in the Miocene coincides with age estimates for
other plant groups of (partly) similar distribution, e.g. Androcymbium Willd. (CaujapCastells et al.,
2001;delHoyoetal.,2009)andThamnosmaTorr.&Frm.(Thivetal.,2011).
T. striata/Australian annual Triglochin The split between the Australian annual lineage and the
widespread T. striata took place in the Miocene. Apart from its distribution on the southern
continents (Australia, Africa, South America), T. striata has colonised the Caribbean and North
America (Corell and Corell, 1982; Haynes and Hellquist, 2000; Leon, 1946) and has been introduced
to the Iberian Peninsula (Talavera, 2010). It also seems to have the ability to establish on isolated
oceanic islands, e.g., French Polynesia and some subAntarctic islands (S. von Mering, study of
herbarium material). Typical (but not exclusive) habitats of T. striata are coastal marshes and tidal
estuaries(seeTable2).Therefore,andalsoconsideringfruitmorphology,dispersalbyseacurrentsor
birdsseemsalikelyexplanationforthewidedistributionofT.striata.
T. palustris/T. maritima complex Divergence of T. palustris from the T. maritima complex was
datedtothe Miocene(Fig.4,Table4). Anunexpectedfindingofouranalysesisthesplitbetweena
Eurasian and an American clade in the T. maritima complex. This disjunction could be the result of
LDDormigrationvialandbridgesthatlinkedthecontinentsonthenorthernhemisphereinthepast.
The relevant land bridges are (1) an Eocene North Atlantic Land Bridge, and (2) an Early Miocene
BeringLandBridge(TiffneyandManchester,2001andreferencestherein).Duetotheageofthesplit
betweentheAmericanandEurasianT.maritimaclade(mean:30.3Ma,95%HPD:21.039.7Ma),the
North Atlantic Land Bridge is a rather unlikely explanation for this disjunction. The Bering Land
Bridge, however, could have served as a migratory route for the group. Considering the high
dispersability of T. maritima by sea currents and probably also by birds (Davy and Bishop, 1991;
Koutstaletal.,1987),thecurrentdistributionoftheT.maritimacomplexcouldalsobetheresultof
longdistancedispersal.
Dispersal of fruits or mericarps For some species of Triglochin it is known that fruits have the
ability to float for long times, e.g. T. maritima (Markgraf, 1981), for others this has been strongly
suspected, e.g. T. striata (see above). Mericarps of T. maritima can float for more than six months
and retain viability for up to five months (Koutstal et al., 1987). Flotation of Triglochin fruits is
facilitatedbythepossessionofairfilledcavities(aerenchyma;e.g.,Markgraf,1981).Theimportance
of water for dispersal of Cycnogeton species has also been reported (e.g., Aston, 2011; Keighery,
1975). Besides dispersal by flotation, birds have been suggested as a dispersal vector for species of
Juncaginaceae, at least over short distances (Haynes et al., 1998). Endozoochorous dispersal by
waterfowlisverylikely,butfeedingexperimentsareneededtoverifythatfruitsofTriglochinsurvive
gut passage and germinate afterwards (pers. comm., J. Figuerola). Epizoochory in mud on bird feet
has been suggested repeatedly (e.g., Pojar, 1974). Fruits or mericarps with mucronate, spurred or
48
evenhookedapicesand/orbasesarefoundinseveralspeciesofTriglochin(e.g.,T.palustris,several
Australian annuals such as T. calcitrapa Hook., T. mucronata R.Br.), making epizoochory by
adherence to feathers plausible. However, this possibility has been questioned for the Australian
annualspecies(Keighery,1975).
Taxonomic,geneticandecologicaldiversity
AmongperennialTriglochin,ouranalysisidentifiedtwocladesofroughlycomparablecrown
groupageandgeneticdiversity,i.e.,theT.maritimacomplex(mean:10Ma,95%HPD:4.2516.4Ma;
meanK2Pdistance:0.045,max.pdistance:0.075)andtheT.bulbosacomplex(mean:13.6Ma,95%
HPD: 7.420.2 Ma; mean K2P distance: 0.055, max. pdistance: 0.082). These two groups, however,
differ strongly in taxonomic diversity with only two or three species (with two varieties) in the T.
maritima complex but seven species (with three subspecies) in the T. bulbosa complex. This
differenceappearstobecorrelatedwithecologicaldiversification.
When a broad concept of T. maritima is applied as, e.g., in Flora of North America (Haynes
and Hellquist, 2000), the T. maritima complex comprises the very widespread T. maritima and the
narrowendemicT.gaspensis.WhileT.maritimaisdistributedalongthecoastsoftemperateregions
of the world but also widely in saline inland localities, T. gaspensis is ecologically more specialised
and restricted to the lower tidal zone of salt marshes where plants are subject to daily flooding
(Haynes and Hellquist, 2000; Lve and Lieth, 1961). Triglochin gaspensis is ecologically,
morphologicallyandkaryologicallydifferentiatedfromT.maritima(LveandLieth,1961)buttheITS
sequences of the two species are almost identical. In addition to these two species, some authors
recogniseT.concinnawithtwovarietiesasdistinctfromT.maritima(e.g.,Keil,2012;MunzandKeck,
1968;Reveal,1977).WhereasthetypevarietyofT.concinnaistypicallyfoundincoastalsaltmarshes
(e.g.,Brayshaw,2000;Keil,2012),T.concinnavar.debilisisfoundininlandhabitatssuchasonsinter
and travertine deposits, at the edges of hot spring pools as well as alkali or salt lakes at higher
elevations in western North America (ca. 4002500 m, Great Basin region; Keil, 2012; S. v. Mering,
pers. obs. on herbarium material). Triglochin mexicana has been reported to occur in brackish or
salinesitesalonglakesorinswamps(EspejoSerenaandLpezFerrari,1996;NoveloandLot,2001).
However,thistaxonneedstobereevaluatedtoclarifyitsstatus.
Triglochin maritima occupies a very similar niche in the Americas and Eurasia, i.e., mainly
coastalsaltmarshesandinlandsalinehabitats(saltoralkalilakes,saltpans,saltmarshes,etc.).The
speciesisrelativelytoleranttophysicalstressandchangingconditions(toleratingsalineandalkaline
conditions,waterlogging etc.).Itisafacultative halophyteandapparentlylessstrictlyhalophyticat
high latitudes and altitudes (Davy and Bishop, 1991). In summary, the T. maritima complex essen
tially comprises one species (T. maritima) which is ancient, very widespread and ecologically rather
invariable, and one (T. gaspensis) or two (T. concinna) more narrowly distributed species of
specializedecology.
In comparison to the T. maritima complex, ecological diversity, including flowering time
diversification,isconsiderablyhigherinthetaxonomicallymorediverseT.bulbosacomplex(Table2).
Whilesomespeciesoccurincoastalhabitatssuchassaltmarshesortidalestuaries(e.g.,T.barrelieri,
T. elongata, T. buchenaui), others are never found in habitats under the direct influence of tidal
flooding (e.g., T. bulbosa; see Table 2, Kcke et al., 2010). Edaphic differentiation is especially
pronouncedintheSouthAfricanT.bulbosa(Kckeetal.,2010).Whilethewidelydistributedsubsp.
bulbosa is waterbound and occurs in temporarily wet coastal and inland habitats, the other three
subspecies grow in fully terrestrial inland habitats characterised by granite (subsp. tenuifolia),
49
limestone(subsp.calcicolaMering,Kcke&Kadereit)andquartz(subsp.quarcicolaMering,Kcke&
Kadereit).
Interesting similarities can be found when comparing T. elongata/T. buchenaui (T. bulbosa
complex) and T. maritima/T. gaspensis (T. maritima complex). These two species pairs comprise a
widespreadandaregionallyendemicspecies,ofwhichtheregionalendemicinbothcasesgrowsin
regularly flooded habitats (T. buchenaui, Kcke et al., 2010; T. gaspensis, Lieth and Lve, 1961). ITS
sequenceswithinthesetwospeciespairsareidenticaloralmostidentical.
Considering that crown group age and genetic diversity (as measured with the DNA regions
used in our phylogenetic analyses) are similar in the T. maritima and T. bulbosa complexes, but
taxonomic diversity (reflecting morphological diversity) is clearly higher in the latter, we conclude
that taxonomic/morphological diversity in these two clades of Triglochin is linked to ecological
diversity,whichwouldillustratetheprominentroleofecologyinspeciation(Schluter,2009;Sobelet
al.,2010).Itseemsplausiblethathabitatshiftstriggeredtaxonomic/morphologicaldiversification.
Probablyourargumentcanbeextendedtomostothercladesofthegenus.Thus,habitatdiversityis
lowinT.scilloides(annual),T.striataandT.palustris.Ofthese,T.scilloidesisusuallyfoundinshallow
water and adjacent mud flats of seasonal pools or other temporary wetlands (e.g., Haynes and
Hellquist,2000;Keil,2012),T.striataiscommonincoastalsaltmarshesandestuariesbutalsofound
insalineinlandhabitats(e.g.,HaynesandHellquist,2000;Lisowskietal.,1982;Moore,1970),andT.
palustrisinhabitssalineas wellasfreshwatermarshesorwetmeadows(Table2).Australianannual
Triglochin mostly inhabit ephemeral wetlands or winterwet habitats (Aston, 2011; Keighery, 1979;
Pignatti and Pignatti, 2005). On the basis of the ecological knowledge available for these species
(Table2),wecannotdecidewhetherthespeciesrichnessofthiscladealsoistheresultofecological
diversificationorofothercauses.

50

Chapter3
RevisionoftheMediterraneanandsouthernAfrican
Triglochinbulbosacomplex(Juncaginaceae)
A.ValerieKcke
*
,SabinevonMering
*
,LadislavMucina&JoachimW.Kadereit
*
Theseauthorscontributedequallytothispublication.
Publishedin:EdinburghJournalofBotany67(3):353398(2010).
ABSTRACT
TheTriglochinbulbosacomplex(Juncaginaceae)fromtheMediterraneanregionandAfricais
revised.Onenewspecies,Triglochinbuchenaui,andtwonewsubspecies,T.bulbosasubsp.calcicola
and subsp. quarcicola, are described from South Africa. The only two Mediterranean taxa in the
complex(Triglochinbarrelieri,T.laxiflora)areelevatedtospeciesrank.Altogethersevenspeciesand
four subspecies are recognized: Triglochin barrelieri, T. buchenaui, T. bulbosa subsp. bulbosa, T.
bulbosa subsp. calcicola, T. bulbosa subsp. quarcicola, T. bulbosa subsp. tenuifolia, T. compacta, T.
elongata, T. laxiflora and T. milnei. An identification key, detailed descriptions and accounts of the
ecologyanddistributionofthetaxaareprovided.AnIUCNconservationstatusisproposedforeach
taxon.
Keywords.Triglochin,Juncaginaceae,revision,taxonomy,morphology,typification,ecology,
MediterraneanFloristicRegion,SouthAfrica,CapePeninsula,CapeFloristicRegion,IUCN
conservationassessment.

Chapter3:RevisionoftheTriglochinbulbosacomplex
52

53
INTRODUCTION
Triglochin L. is the largest of the currently recognized four (Haynes et al., 1998; Stevens,
2001+)orfive(Dahlgrenetal.,1985)generaofthesmallmonocotyledonfamilyJuncaginaceae,and
contains2535annualandperennialspecies.Thegenushasanalmostcosmopolitandistributionwith
Australiaascentreofspecificdiversity.CommonhabitatsofTriglochinarecoastalsaltmarshes,tidal
mudflats,freshwaterwetlands,seasonalpoolsandinlandsaltpans.
In the Mediterranean region and southern Africa Triglochin is represented by a group of
widelydistributedtaxawithovoidtoverynarrowlyovoidfruitsconsistingofthreefertilecarpelsthat
separateatmaturity.Acarpophoremaybepresentorabsent.Sincealltaxabelongingtothisgroup
at some point have been included in Triglochin bulbosa L., the group is here called the T. bulbosa
complex.InamolecularphylogeneticstudyofTriglochin(vonMering&Kadereit,unpubl.data),this
complexforms a wellsupported monophyletic clade. In addition to the Triglochin bulbosa complex,
T.maritimaL.,T.palustrisL.andT.striataRuiz&Pav.canbefoundintheMediterraneanregion,and
T.striatainsouthernAfrica.ThesethreelatterspeciescaneasilybedistinguishedfromtheTriglochin
bulbosa complex by fruit characters. The fruits of Triglochin maritima are composed of six fertile
carpels,T.palustrishasfruitswithonlythreecarpelsthatdonotseparateatmaturity,andT.striata
iseasilyrecognizablebyroundishfruitsconsistingofthreefertileandthreesterilecarpels.
TheTriglochinbulbosacomplexisasystematicallydifficultgroupasevidentfromitscomplex
taxonomic history. In South Africa, five species of the Triglochin bulbosa complex have been
recognized in the past, but currently only a broadly circumscribed T. bulbosa is accepted within the
complex (e.g., Obermeyer, 1966; Goldblatt & Manning, 2000; Glen & Cook, 2003; TrinderSmith,
2003).Inthisbroadcircumscription,Triglochinbulbosashowssubstantialmorphological(e.g.,inleaf
and fruit characters and underground parts) and ecological variation (Fig. 1). In the Mediterranean
area, Triglochin barrelieri and T. laxiflora have been variously treated as separate species (e.g.,
Talavera, 1987; S. Talavera, ms. Flora Iberica) or as subspecies of the South African T. bulbosa (e.g.,
Bols&Vigo,2001).
No revision of the entire complex has ever been attempted. This paper reports the findings
of an analysis of extensively researched herbarium material and our own collections from the
Mediterranean area and South Africa. Based on morphology, phenology, ecology and geographical
distribution,sevenspeciesandfoursubspeciesarerecognized.TheseareTriglochinbarrelieriLoisel.,
T. buchenaui Kcke, Mering & Kadereit, sp. nov., T. bulbosa L. (with subspp. bulbosa, tenuifolia
(Adamson) Horn, calcicola Mering, Kcke & Kadereit, subsp. nov., quarcicola Mering, Kcke &
Kadereit, subsp. nov.), T. compacta Adamson, T. elongata Buchenau, T. laxiflora Guss and T. milnei
Horn.
TAXONOMICHISTORY
Triglochin bulbosa L. was described by Linnaeus in 1771 based on collections by Thunberg

fromtheCapeofGoodHope(Buchenau,1896).However,Barrelier'saccountof"Juncagomaritima"
(1714)fromthewesternMediterraneanisthefirstdescriptionofamemberoftheTriglochinbulbosa
complex. Despite its morphological distinctness stressed by Barrelier (1714), Linnaeus did not
recognize Barrelier's taxon as a separate species but included it as var. in Triglochin palustris L.
(Linnaeus,1762).LoiseleurDeslongchamps(1806)atfirstincludedthistaxonasTriglochinpalustrisL.
var. in his Flora Gallica, but in the amendments to this Flora (LoiseleurDeslongchamps, 1807)
recognizeditasadistinctnewspecies,T.barrelieriLoisel.Anaffinitytoorapossiblesynonymywith
54
Triglochin bulbosa was suggested ("An Triglochin bulbosum. Lin. Mant. 226 ?", Loiseleur
Deslongchamps,1807).ThesecondMediterraneantaxon,TriglochinlaxifloraGuss.,wasdescribedby
Gussone(1825)fromSicilyandCorsica.
Subsequently, Triglochin barrelieri and T. laxiflora were both recognized as distinct species
(e.g., LoiseleurDeslongchamps, 1828; Battandier & Trabut, 1884; Bonnet, 1896; Jahandiez & Maire,
1931).OtherauthorsacceptedonlyTriglochinlaxifloraasaspecieswhileT.barrelieriwastreatedas
a synonym (e.g., Micheli, 1881; Buchenau, 1903) or as a subspecies of the South African T. bulbosa
(e.g.,Hayek,1933;Maire,1952).Rouy(1912)wasthefirsttoreducethetwoMediterraneantaxato
subspecies of the South African Triglochin bulbosa (T. bulbosa subsp. barrelieri (Loisel.) Rouy, and
subsp.laxiflora(Guss.)Rouy).SomerecentFlorastreatthemasdistinctspecies,Triglochinbarrelieri
andT.laxiflora(Talavera,1987;S.Talavera,ms.FloraIberica),andothersassubspeciesofT.bulbosa
L.(e.g.,deBols&Vigo,2001).
Based mainly on differences in fruit size and pedicel length, Pau (1914) described Triglochin
barrelieri var. maura Pau, a variety endemic to Morocco, which was later also treated as T. bulbosa
var. maura (Pau) Maire (Maire, 1952) or as a synonym of T. bulbosa subsp. barrelieri (Govaerts,
2008).
Buchenau (1903) newly described Triglochin elongata Buchenau from South Africa and
recognizedT.bulbosa(syn.T.barrelieri)andT.laxiflora.TheoccurrenceofthelatterinSouthAfrica
wasrecordedwithreservationsbecauseonlyonespecimencouldbeassignedtoTriglochinlaxiflora
byBuchenau(1903).Bennett(1897)alsobelievedTriglochinlaxifloratooccurinSouthAfrica,while
HornafRantzien(1961)excludedthisspeciesfromtheSouthAfricanflora.
Adamson(1939),whocontributedconsiderablytothecompilationofthe"FloraoftheCape
Peninsula",studiedtheSouthAfricanspeciesofTriglochininmoredetailandalsopublishednoteson
habitat preferences. He provided detailed descriptions and diagnostic characters for the South
African taxa of the Triglochin bulbosa complex known to him (Adamson, 1939). Adamson described
two new Triglochin species, T. tenuifolia Adamson (1939) and T. compacta Adamson (1943), and
accepted T. elongata. Thus, Adamson (1939, 1943) recognized the following four species of the
TriglochinbulbosacomplexinSouthAfrica:T.bulbosa,T.elongata,T.tenuifoliaandT.compacta.We
followAdamsonintherecognitionofhisnewlydescribedT.compactaandT.tenuifolia,buttreatthe
latterasasubspeciesofT.bulbosa.
The last treatment of the African species of Triglochin was carried out by Horn af Rantzien
(1961). This author did not regard his publication as a full revision since it was based on herbarium
material from Kew and Stockholm only. He provided an identification key to the species and
subspecies of Triglochin in Africa which was mainly based on differences in fruit characters and
underground organs. In his study only taxa with bulbs were investigated in great detail, especially
withrespecttoflowerandfruitmorphology.Asaresultofthishedescribedanewspecies,Triglochin
milnei Horn, from southern Central Africa (Tanzania, Angola, Zambia, eastern South Africa),
recognizable by its conspicuously larger flowers and fruits. Apparently Horn af Rantzien (1961) was
notawareoftheexistenceofTriglochincompactaanddidnotincludethistaxoninhisstudy.Hedid,
however, include Triglochin tenuifolia in T. bulbosa as subsp. tenuifolia (Adamson) Horn, and also
treated the other taxa at intraspecific rank. This resulted in the recognition of five subspecies of
Triglochin bulbosa: subsp. bulbosa (South Africa), subsp. tenuifolia (South Africa), subsp. barrelieri
(Mediterranean), subsp. laxiflora (Mediterranean) and subsp. maura (Pau) Horn (Morocco). Such
treatment seemed justified by great similarities within the complex and the occurrence of
transitional forms. Like Adamson (1939), Horn af Rantzien (1961) also accepted Triglochin elongata
asaseparatespecies.However,thistaxonwasoverlookedbylaterauthors.
55
Fig.1.TriglochinbulbosaL.subsp.calcicolaMering,Kcke&Kadereitsubsp.nov.A,habit,B,habitat(De
Hoop);T.bulbosaL.subsp.quarcicolaMering,Kcke&Kadereitsubsp.nov.C,habitat(Moedverloren),D,
habit;T.compactaAdamson.E,bulbs,F,habitat(Goukamma),G,infructescence;T.elongataBuchenau.H,
habit(withrhizomes),L,habitat(Overberg);T.buchenauiKcke,Mering&Kadereitsp.nov.I,habitat(Knysna),
K,habit.PhotosA,B,F,HL:L.Mucina;CD:U.Schmiedel;E,G:J.Naud.
56
MostrecentFlorasandchecklistsofCentralandsouthernAfricaacceptTriglochinbulbosaas
the only species (plus the distinct T. striata, which does not fall into the T. bulbosa complex; e.g.,
Obermeyer,1966;Goldblatt&Manning,2000;Glen&Cook,2003;TrinderSmith,2003)andtreatT.
milnei as a synonym (e.g., Obermeyer, 1966; Lisowski et al., 1982). Only Cook (2004) recognized T.
milnei as a distinct species. The World Checklist of Monocotyledons (Govaerts, 2008) currently
recognizes only Triglochin bulbosa with a broadly circumscribed subsp. bulbosa (Tanzania to South
Africa),andsubspp.barrelieriandlaxiflora(bothintheMediterranean).
MATERIALANDMETHODS
Thisrevisionisbasedmainlyonthestudyofca.800specimensfromAV,B,BC,BM,BOL,EA,
FB,HBG,JE,K,M,MSB,MJG,MO,NAP,NBG,NBGSAM,NBGSTE,NY,P,POZG,RNG,RO,S,STEU,
WHB,ZandZT(abbreviationsaccordingtoHolmgren&Holmgren,1998).Namesweretypifiedasfar
aspossible,andmosttypematerialcitedhasbeenseen,insomecasesasdigitalimages(indicatedby
!andphoto!,respectively).
The distribution maps are based on information from herbarium labels and unambiguous
records in the literature. Coordinates of older collections were taken mainly from the NGA GEOnet
Names Server (GNS) (National GeospatialIntelligence Agency, http://earth
info.nga.mil/gns/html/index.html) and the Gazetteer of the Cape Region (Institute of Systematic
Botany,UniversityZurich,http://www.systbot.uzh.ch/Bestimmungsschluessel/GazeteerRSA.html).In
addition,theGoogleEarthwebbasedsatelliteimagerydatabase(<earth.google.com>)wasused.All
mapsweregeneratedusingtheprogrammeArcGIS9.0(ESRI,2006).Thelocationsofnewcollections
made between 2000 and 2008 were ascertained by using a GPS device. Some information on
collector names and/or localities in Africa was verified using Aluka (Ithaka Harbors, Inc.,
www.aluka.org). Measurements were made on herbarium material and may deviate from fresh
material. Longer field trips were made in 2006 (MarchMay) in South Africa, in October 2006 in
Sardinia(Italy),andin2007(MarchApril)inFranceandSpain.Theconservationstatusofeachtaxon
wasassessedusingtheIUCNRedListcategoriesandcriteria(IUCN,2001).
TAXONOMICTREATMENT
TriglochinL.,Sp.Pl.:338(1753).Typespecies:T.palustrisL.,"HabitatinEuropaeinundatis
uliginosis"(lectoLINN466.1photo!).
JuncagoSg.,Pl.Veron.3:90(1754).
JuncagoTourn.exMoench,Methodus:644(1794).
TristemonRaf.,Amer.MonthlyMag.&Crit.Rev.4(3):192(1819).
AbbotiaRaf.,NewFl.(Rafinesque)1:36(1838).
HexaglochinNieuwl.,Amer.Midl.Naturalist3:10(1913).
Note. None of the above synonyms has been used in the recent literature. We have not seen any
typematerialforthesenames.

57
DescriptionoftheTriglochinbulbosacomplex
Erectperennialherbs,590cmhigh,withscaposeinflorescencesandrhizomesor(oftenaggregated)
bulbs. Base of plants covered with fibrous remains of old leaves. Leaves basal, distichous,
succulent,linear,terete(semiterete),ligulate,sheathingatthebase.Inflorescenceaspike.Flowers
hermaphrodite, 3merous, ebracteate, proterogynous. Perianth of six green or sometimes reddish
tepals;tepalsconchiform,deciduous.Stamenssixorthree,sessileorsubsessile;anthersdorsifixed,
dithecous, dehiscing longitudinally, extrorse. Infructescence a raceme. Fruits ovoid to very narrowly
ovoid (Figs 2, 6). Fertile carpels three, connate, sessile, with one ovule and sessile stigma each,
usually separating into oneseeded mericarps at maturity. Sometimes an outer whorl of three
vestigial carpels present (T. milnei and T. compacta). Carpophore mostly present, sometimes
carpophore very thin or lacking and then mericarps not always separating. Seeds erect, embryo
straight.
Notes. Flowers of Triglochin usually have six stamens. However, this number may be reduced to
threeasreportedforTriglochinmilnei(HornafRantzien,1961).Suchreductionwasalsonotedinthis
study in Triglochin bulbosa subsp. calcicola. The gynoecium of Triglochin is variable. In principle,
carpels are arranged in two whorls of three carpels each (e.g., Tomlinson, 1982). Whereas in some
species (e.g., Triglochin maritima) all six carpels are fertile, in others (e.g., Triglochin striata) the
three carpels of the outer whorl are sterile (e.g., Tomlinson, 1982; pers. obs.). In the Triglochin
bulbosa complex only three fertile carpels are present. However, in some species (especially
Triglochincompacta,T.milnei)membranousoutgrowthsarepresentatthebaseofthecarpophore,
which have been interpreted as remains of the second whorl of carpels. Size and form of these
outgrowths are variable. Sometimes they are extremely small, but in Triglochin milnei these
structures are often very distinctive. Horn af Rantzien (1961) considered them sterile carpels and
accordinglydescribedTriglochinmilneiashavingsixcarpels,ofwhichthreearesterile.However,not
allspecimensstudiedhaveoutgrowthsthatcanbeidentifiedassterilecarpels.Fruitsinthecomplex
mostlyseparatefrombottomtotop.Insomespeciesthischaracterseemstobevariable,i.e.,carpels
separate from bottom to top or from top to bottom. Currently we are not able to give accurate
informationonthemodeofcarpelseparationforalltaxaandaccordinglydonotknowwhetherthis
character can be used for the distinction of taxa. This character should be further studied on fresh
material.
Distribution. TheTriglochinbulbosacomplexisnativetothe Mediterraneanregion,extendingalong

the Atlantic coast to NW France and Morocco (Fig. 3), and to southern Central Africa extending to
SouthAfrica(Figs7,11,12).AsubspeciesofTriglochinbulbosahasbeenreportedtobenaturalized
in Western Australia (Western Australian Herbarium, 1998 and onwards; Australian Plant Census,
2008;seealsounderT.bulbosasubsp.bulbosa).
Habitat and Ecology. Plants of this group are typically found in hygrohalophytic habitats such as
coastalsaltmarshes,tidalmudflatsandalonglagoons.Theyalsooccuroncoastalrocksexposedto
saltsprayorintemporarypools,ininlandsaltpans,indry,nonsalinehabitatsonmountainslopesor
in seasonally wet grassland and along river banks. Some collections were made at altitudes above
1000m.

58
Fig.2.FruitsofA,TriglochinbarrelieriLoisel.;B,T.laxifloraGuss.;C,T.compactaAdamson;D,T.milneiHorn;E,
T.elongataBuchenau;F,T.buchenauiKcke,Mering&Kadereit.DrawingsbyL.Klckner(Mainz).
59

Fig.3.GeographicaldistributionofTriglochinbarrelieriLoisel.andT.laxifloraGuss.
Chromosomes. In Triglochin chromosome numbers of up to 2n = 144 are known (e.g., Lve & Lve,
1958;Robb&Ladiges,1981).NochromosomecountsareavailableforthesouthernAfricanspecies
ofthecomplex,andonlyfewfortheMediterraneanspecies.ForTriglochinlaxiflora2n=18(Gard&
MalheirosGard,1953;Bartoloetal.,1977;Talaveraetal.,1995;Castroetal.,2007),andforT.
barrelieri 2n = 30 (Castro & Carvalho Fontes, 1946), 32 (Dahlgren et al., 1971; Bartolo et al., 1977)
and36(Castroetal.,2007)havebeenreported.
Nomenclatural notes. Although Linneaus (1753) treated Triglochin as neuter in his original
publication of the name, the feminine gender has to be used (ICBN Vienna Code, 2006, Art. 62.2b
andEx.5).
KeytothespeciesoftheTriglochinbulbosacomplex
Thefollowingkeyisbasedonherbariummaterial.Characterssuchassizeofflowers,fruitsor
pedicels often are not sufficient to identify species because of the existence of (in this respect)
intermediate forms. Underground parts are sometimes lacking in herbarium specimens but are
important for the identification of species. When the geographical origin of the material and/or its
phenologyareunknown,specialattentionmustbepaidtothepresenceofeitherbulbsorrhizomes.
Notesonhabitatanddistributionshouldalsofacilitatetheidentificationofspecies.Asalreadystated
by Horn af Rantzien (1961, p. 102): "In identifications statements on locality, habitat, and time of
60
anthesisareofmoreusethanthoseonmorphologicalcharacters."Thisimpliesthatthecombination
ofmorphologicalattributes,ecologyandgeographicaldistributiondefinesspeciesbest.
1a.PlantsfromtheMediterraneanregionortheAtlanticcoastofwesternEurope
andNorthAfrica..............................................2
1b.PlantsfromCentralorsouthernAfrica........................................3
2a.Leavespresentatfloweringtime,mostlyaslongasorlongerthanplantheight;
pedicelsspreading(Fig.2A);plantsfloweringinspring(iiv)............................1.T.barrelieri
2b.Leavesformedafterflowering,usuallymuchshorterthanplantheight;
pedicelsascending,appressedtoinfructescenceaxis(Fig.2B);
plantsfloweringinautumn(ixxii)...................................2.T.laxiflora
3a.Plantswithrhizomes,ofsaltmarshestuariesorestuarinerivers...................................................4
3b.Plantswithbulbs,ofinlandhabitats,neverunderdirectinfluenceoftidalflooding................5
4a.Inflorescencesdense,with10100flowers;infructescences(6)1025cmlong;fruitswith
carpophore;plantsgrowingintheupperzoneofsaltmarshestuaries
andestuarinerivers..............................................................................................6.T.elongata
4b.Inflorescenceslax,with317flowers;infructescences27cmlong,fruitswithout
carpophore;plantsrestrictedtothelowertidalzoneofsaltmarshestuaries
...................................................................................................7.T.buchenaui
5a.Flowersusually>3mmlong;fruits(8)1014mmlongand3mmwide(Fig.2D);
plantsfloweringandfruitinginsummer(xii/iii);plantsfromCentral
andsouthernAfrica(incl.easternSouthAfrica)............................5.T.milnei
5b.Flowersusually<3mmlong;fruits410mmlongand12mmwide;
plantsfloweringandfruitinginspring(viixi)orautumn(ii/iiiv);
plantsfromSouthAfrica...6
6a.Leavespresentatfloweringtime,mostlyaslongasorlongerthanplantheight;
plantsfloweringandfruitinginspring(viixi)........................................................3.T.bulbosa
6b.Leavesformedafterflowering,mostlyonlyhalfaslongasplantheight,
plantsfloweringandfruitinginautumn(ii/iiiv).................................................4.T.compacta
1.TriglochinbarrelieriLoisel.,Fl.Gall.ed.1.725(1807),ed.2.1:264(1828).Triglochin
bulbosasubsp.barrelieri(Loisel.)Rouy,Fl.France(Rouy)13:271(1912).Type:
France,Arles(Arelatem),"habitatArelatemadRhodanu",Artauds.n.(holoAV
photo!;Fig.4).Figs2A,3.
TriglochinpalustreL.var.,Sp.Pl.ed.2.1:483(1762).
TriglochinatlanticaWilld.herb.,testeKunth,Enum.Pl.(Kunth)3:143(1841).Type:notinBW.
JuncagochabraeiBubani,Fl.Pyren.(Bubani)4:8(1901).
Triglochinbarrelierivar.mauraPau,Bol.RealSoc.Esp.Hist.Nat.14:425(1914).Type:
Morocco,adripasfl.Lukos,juxtaElAraix,1iii1930,FontQuers.n.(synMPUMairephoto!),
interCeutaetTetauen,l.d.RincndeMedik,13iii1930,FontQuers.n.(synMPUMaire
photo!).Triglochinbulbosavar.maura(Pau)Maire,Fl.AfriqueN.1:213(1952).
Triglochinbulbosavar.genuinaMaire,Jahandiez&Maire,Cat.Maroc859(1934).
Plants1045cmhigh,withfusiformbulbs;bulbscoveredbydry,brown,stiff,sometimesspikyfibres,
these up to 4 cm long. Plants without rhizomes. Leaves usually many, present at flowering time,
61
mostly as long as or longer than plant height, uniform. Inflorescences rather dense, with 1025(30)
flowers. Flowers (2.5)3.04.5 mm long. Pedicels elongated at fruit maturity, up to 27 mm long,
rarely longer, spreading at angles of 45 to 90 from the inflorescence axis, slightly curved inwards
towardstheapex.Infructescences816cmlong.Fruitsnarrowlyovoid,510mmlong,rarelylonger,
12mmwide(Fig.2A),sometimesturningdarkorblackatfruitmaturity.Mericarpslinear,connateat
thecarpophore;tipsonlyslightlycurvedoutwards.
Distribution.ThespeciesiswidelydistributedintheMediterraneanregion(withanoteworthygapin
Egypt) reaching the Atlantic coast of NW France and Morocco (Fig. 3). It is known from all large
Mediterraneanislands.
Habitat and ecology. Triglochin barrerlieri is commonly found in coastal wetlands, e.g., in salt
marshesandalonglagoons,andlessfrequentlyoncoastalrocksexposedtosaltspray,indepressions
incoastaldunesorincorkoakforests(e.g.,inAndalusia,Spain).InNorthAfricaittypicallyoccursin
sebkhas (sabkhas: rarely flooded salt marshes of desert coasts). The species usually grows in very
openvegetationandisoftenassociatedwithSarcocorniafruticosaandJuncusacutus.
Phenology.Floweringinspring(iiv),leavespresentatfloweringtime.
Chromosomenumber.2n=30(Castro&CarvalhoFontes,1946),32(Dahlgrenetal.,1971,Bartoloet
al.,1977),36(Castroetal.,2007).
Notes. This species is very variable in overall size and fruit and pedicel length. Triglochin barrelieri
var. maura has been described from Morocco as a taxon of larger size and with longer pedicels.
However,suchlargeformsarealsofoundinotherpartsofthespeciesdistributionrangeandarenot
limitedtoMorocco.PresenceofTriglochinbarrelieriinIsraelseemspossiblebecausethedescription
of T. palustris for this area by Post (1933) partly matches T. barrelieri. We have, however, not seen
anymaterialofthespeciesfromIsraelwhereitmaybeextinct.
Taxonomicremarks.TriglochinbarrelieriisclearlydistinctfromtheonlyotherMediterraneanspecies
of the complex, T. laxiflora, by flowering in spring (vs. autumn), the presence of leaves at flowering
time (vs. leaf formation after flowering), and by having spreading pedicels (vs. appressed to
infructescenceaxis).
ThedistinctionbetweenTriglochinbarrelieriandT.bulbosasubsp.bulbosaisdifficult,andthesetwo
species are morphologically very similar to each other. They can be distinguished by geographical
originandecology,andresultsofourmolecularstudies(vonMering,Kcke&Kadereit,unpubl.data)
supporttheirrecognitionasseparatespeciesbecausetheyarenotclosestrelativetoeachother.
Nomenclatural notes. In the protologue of Triglochin barrelieri the following localities are given:
HabitatinmaritimisNeutstriae,Occitaniae,Provinciae;adripasRhodanipropeArelatemcollegitD.
ARTAUD. (LoiseleurDeslongchamps, 1807). The herbarium of LoiseleurDeslongchamps in Avignon
(AV) contains several specimens bearing the name Triglochin barrelieri. Only the specimen cited
above corresponds to the protologue (location and collector), was annotated by Loiseleur
Deslongchampshimself(P.Moulet,pers.comm.)andthusprobablyrepresentstheholotype(Fig.4).
62
ProposedIUCNconservationstatus.LeastConcern(LC).Thespeciesusedtobewidelydistributedin
the Mediterranean region and is probably not under immediate threat. However, coastal wetlands
(wherethespeciesistypicallyfound)areamongthemostvulnerablehabitatsintheMediterranean
region.Largeareasarealreadylost,degradedorunderseverethreatduetodrainage,construction,
urbanisationandtourism(e.g.,Finlayson&Moser,1991;MedWet,1996).Onlyfewdataareavailable
onthecurrentareaofoccupancyduetotheverylownumberofrecentcollections.Severalcollecting
localities of older specimens have been destroyed by settlements. Other populations might also be
destroyedorthreatenedbyhabitatdestruction.Therefore,theareaofoccupancymightbedeclining,
and the global conservation status of this species may be prone to change. Further conservation
assessments are also necessary at the national level. Triglochin barrelieri is already protected in
Aquitaine, France (Musum National dHistoire Naturelle, 20032006) and in the Italian regions of
Molise,Basilicata,CalabriaandSicily(Contietal.,1997).
Selected specimens examined. Europe. FRANCE. Aquitaine. Dp. Gironde: Bordeaux, Bassin d'Arcachon, vi
1831, Endress s.n. (B, BM, HBG, JE, M, P). Bretagne. Dp. Finistre: Loctudy and Combrit, without date,
Bonnemaisons.n.(AV).Dp.Morbihan:Gvres,15vii1847,Debrezs.n.(ZT);Gvres,1847,Pontarliers.n.(P);
Gvres,1847,Tasls.n.(P);Gvres,3vi1851/4vii1847,Toussaintss.n.(P);PortLouis,vi1847,Pontarliers.n.
(ZT),1849,Greniers.n.(ZT);PortLouis,30vi1877,Lloyds.n.(BM);PortLouis,1891,Renous.n.(P);PortLouis,
27v1912,Charriers.n.(BM).Corsica.Dp.HauteCorse:Ghisonaccia,30v1900,Rotgs339(P);SaintFlorent,
river l'Aliso, 20 v 1973, Auquier 3037, Baguette & Lambinon (M). Dp. CorseduSud: Bonifacio, 29 iii 1849,
Kralik787(P);PortoVecchio,19iv1868,Mabilles.n.(BM,P,K);Bonifacio,SantaManza,12iv1876,Revelire
s.n.(BM);Bonifacio,12v1880,Reverchons.n.(BM);Bonifacio,SantaManza,12v1880,Reverchon246(JE,P);
Bonifacio,SantaManza,2vi1894,Stefanis.n.(P);Bonifacio,Piantarella,5iv1895,Stefanis.n.(BCSENNEN,P),
9 iv 1895, Stefani s.n. (M, ZT); PortoVecchio, iv 1911, Stefani s.n. (P); iv 1912, Roux s.n. (P); PortoVecchio, E
sideofStabiaccoestuary,13iv1939,BrenanC.408(K).LanguedocRoussillon.Dp.Aude:Leucate,30v1887,
Chevallier s.n. (M); La Nouvelle, iv 1899, Sennen s.n. (BCSENNEN), 10 vi 1904, Sennen s.n. (P); Dp. Hrault:
Maguelonne,1vi1821,Cufinos.n.(HBG,RO);Ste(formerlyCette):1834,Gaths.n.(BM),3v1882,Neyras.n.
(JE),4v1882,Neyras.n.(B,JE,RO),9v1891,Dirmdes.n.(K),v1905,Bees.n.(P);Roquehaute,nearBziers,v
1855, Personnat s.n. (P); Montpellier, v 1857, Cramer s.n. (HBG); Maguelonne, 13 vi 1857, Cosson s.n. (P);
Palavas, 27 v1899, Morot s.n. (P); Palavas, 20 v1908, Vichet s.n. (P); Lattes, Montpellier, 23 iv 1914,Sudi s.n
(ZT); E of Palavas, 10 v 1914, Noaik s.n. (Z); Prols, 1 vi 1915, Braun s.n. (ZT); St. Marcel, Prols, 1 vi 1915,
BraunChur s.n. (Z); Montpellier, v 1921, Jenny s.n. (ZT); Etang de Prols, Palavas, 12 v 1934, Zobrist s.n (ZT);
Dp. PyrnesOrientales: Vendres, 28 iv 1891, Sennen s.n. (M), 7 vii 1891, Sennen s.n. (BCSENNEN).
ProvenceAlpesCte d'Azur. Dp. BouchesduRhne: tang de Marignane, 7 vi 1894, Rouy & Blaise s.n. (P);
SaintLouisduRhne, v 1906, Thibaut s.n. (P, ZT); without locality, v 1909, Thiebaut s.n. (M); Les Saintes
MariesdelaMer, 12 iv 1957, H. & E. Walter 68 (B); Ile des Riges, 24 iv 1967, Lpnitz F 233, F 234, F 235
(MJG); Etang de Vaccars, 11 iv 1969, Hecker F 320 (MJG). Dp. Var: Hyres, v 1839, Kugel s.n. (BM, M);
Toulon,LesSablettes,16v1848,Bourgeau361(M,P);Hyres,1853,Jordans.n.(HBG);Hyres,v1863,Huet&
Jacquin s.n (B); Hyres, 29 iv 1869, Shuttleworth s.n. (P); Toulon, Les Sablettes, 9 v 1870, Antheman s.n (B);
Hyres,iv1873,Allards.n(P,Z);Toulon,23vi1887,Chevaliers.n.(P);Hyres,14iv1898,Mouillefarines.n.(P);
Hyres,30iv1903,Raines.n.(P);SaintTropez,LesSalins,iii1910,A.&E.G.Camuss.n.(P),iv1910,A.&E.G.
Camuss.n.(P),22iii1911,A.&E.G.Camuss.n.(P);Hyres,25iii1913,Begers.n.(B);Hyres,25iii1913,Noaik
s.n. (Z); Hyres, 3 iv 1913, Thompson s.n. (BM); Saint Tropez, Les Salins, 31 iii 1919, Hibon s.n. (P); Hyres,
Ceinturon, 14 iv 1925, Ruppert s.n. (M); Hyres, La Londe, river Pansard, 30 iv 1929, Ruppert s.n. (M); Saint
Tropez,LesSalins,5iv1931,Pauls.n.(M);Hyres,30iii1932,Oberneder&Oberneder457(BM);Hyres,5iv
1955,Schumachers.n.(HBG).ITALY.Apulia.Prov.Brindisi:SalinediPuntadellaContessa,23v2007,diPietro
& Wagensommer s.n. (MJG), 09 iv 2008, di Pietro s.n. (MJG). Prov. Foggia: Monte Gargano, Manfredonia, iv
1876,Comes,Freda,Pedicions.n.(RO);Mte.Gargano,mouthofriverCervaro,14kmSofManfredonia,roadto
Bari,20iv1964,Dietrichs.n.(M),Hertel3575(M),Podlech9554(MSB);betweenScialeBorgiaandthemouth
63
of river Cervaro, 1214 km S of Manfredonia, 22 v 1972, Hein s.n. (B). Prov. Taranto: TarantoGinosa, 18 iv
1879, Pasquale s.n. (RO). Lazio. Prov. Rome: Ostia, near Rome, v 1925, Meebold s.n. (K). Sardinia. Prov.
Cagliari: Cagliari, vi 1854, Huet du Pavillon s.n. (P); Cagliari, La Madeleine, 3 iv 1958, Titden 128 (K); Cagliari,
near Maddalena, 6 iv 1858, Schweinfurth s.n. (HBG); Quartu Sant Elena, Stagno di Simbirizzi, ca. 20 m, 18 iv
1975,Urmis.n.(Z);DomusdeMaria,SWofStagnionidesuSali,4iv1992,Stud.biol.Acad.Turic.initinere92
12(Z).Prov.OlbiaTempio:Terranova,11iii1885,ForsythMajors.n.(HBG);GolfoAranci,1iv1909,Poeverlein
25009 (B); near Olbia, 8 iv 1962, Podlech 7872 (MSB); Porto di Cugnana, iv 1966, Merxmller 20931 &
Oberwinkler,Merxmller20932&Oberwinkler(M);3kmSofPalau,19iv1973,Humphries&Richardson378
(RNG).Prov.Oristano:Oristano,propelocumdictumTorreGrande,19iii1912,Fioris.n.(BC,BM,RO,Z).Prov.
Sassari:StintinoPeninsula,CapoFalcone,3iv1968,HeckerI27,I28(MJG);CapoFalcone,ca.10m,13iv1971,
Hecker I 500 (Hec.691) (MJG); Capo Falcone, 015 m, 9 iv 1973, Humphries & Richardson 167 (BM); Capo
Falcone,4x2006,vonMerings.n.(MJG).Sicily.Prov.Agrigent:Licata,iv1902,Ross376(B,HBG,M,ZT);Isola
di Lampedusa, 10 iii 1907, Martarana s.n. (RO). Prov. Caltanisetta: San Cataldo, 1847, Rabenhorst s.n. (BM);
SanCataldo,26v1875,Porta&Rigo299(JE,P).Prov.Catania:Catania,26iii1856,E.&A.HuetduPavillons.n.
(BM,M,P).Prov.Siracusa:Siracusa,26iv1898,Rigo161(HBG,M,P);nearSiracusa,iv1898,Rigos.n.(B,BM,
M,P,Z,ZT);Pachino,31iii1952,Harries&Doppelbaur343(M);nearSiracusa,iii2005,Uhinks.n.(MJG).Prov.
Trapani: Marsala, v (no year given), Todaro s.n. (BM, RO); Trapani, 9 iv 1966, H. & H. Doppelbaur 14565 (M).
Tuscany. Prov. Grosseto: Follonica, 27 v 1911, Fiori s.n. (BCSENNEN). Prov. Livorno: Livorno, Paduletta, 19 v
1861,Savis.n.(B,P).Prov.Pisa:nearPisa,1868,vanHeurcks.n.(BC).SPAIN.Andalusia.Prov.Cdiz:Bonanza,
17iv1951,Alston10267(BM);entreSanFernandoyChiclana,28v1951,deBoloss.n.(BC);betweenTarifaand
PuertoReal,8iv1953,Merxmller&Wiedmann648/53(M);nearLosBarrios,ca.150m,14v1967,Brinton
Lee1306(BM);Tarifa,RiodelaJara,iii1973,Deils.n.(FB);Tarifa,19iv1974,Carr1636(RNG);PuertodeSanta
Mara, 13 iv 1978, Luque & Valds s.n. (B, BC, HBG, M, RNG); entre San Fernando y Chiclana, 4 v 1979,
Cabezudo,Gallego&Talaveras.n.(RNG);Tarifa,mouthofRioJara,20vi1990,Deil5967(FB);Tarifa,Zaharade
losAtunes,16v1992,Deil6329(FB).Prov.Sevilla:IslaMenor,RioGuadalquivir,17iv1972,Heywood,Moore,
Bramwell, et al. 218 (RNG). Balearic Islands. Mallorca: near Porrata, 5 iv 1921, WeltiHug s.n. (Z); SE of
Albufera, v 1985, Beckett s.n. (RNG). Catalonia. Prov. Girona: Castell d'Empries, 27 iv 1908, Sennen 613
(BM); Castell d'Empries, vii 1916, Pascual s.n. (BCSENNEN); Rosas, v 1943, Bolos s.n. (BC); Aiguamolls de
Castell d'Empries, 14 v 1983, Blanch & Valls s.n. (M). Prov. Tarragona: Delta de l'Ebre, Sant Carles de La
Rpita,2iv1934,FontQuers.n.(BC);Alfaquespeninsula,nearSantCarlesdelaRpita,2iv1934,Koch34/50
(ZT);Deltadel'Ebre,Sales,16ii1975,Baladas.n.(BC).Galicia.Prov.ACorua:Olveira,24iv1981,Amich,Rico,
Snchez&Giraldezs.n.(BC);Olveira,SantaUxadeRibeira,4v1985,ValdsBermejo10148&GrupoBotanico
Gallego (BC, M, Z). Valenciana. Prov. Valencia: Valencia, iv 1892, Reverchon s.n. (M); Sagunt(o) (French:
Sagonte), iv 1893, Reverchon 834 (B, BM, HBG, M, P, Z). PORTUGAL. Dist. Aveiro: Aveiro, Mina, 20 v 1954, J.
Matos,A.Matos&Marques4752(P);lhavo,GafanhadaBoavista,20v1954,J.Matos,A.Matos&Marques
4768 (P); lhavo, between Forte da Barra and Ponte Nova (Costa Nova), 30 v 1979, Marques 1260 (B). Dist.
Coimbra: Figueira da Foz, Gala, 11 v 1950, J. Matos & A. Matos s.n. (BM). Dist. Faro (Algarve): between Vila
Real de San Antnio and Castro Marim, 22 iv 1956, MalatoBeliz et al. 2890 (RO); Martinhal near Sagres, 1 iv
1971,Davis50919(RNG);Alvor,iii1990,Univ.ofOxfordPlantSciencesundergraduatefiledcourse(RNG).Dist.
Leiria:SofNazar,7iv1961,C.&J.Poelt207,227(M).Dist.Lissabon:Sacavm,31iii1943,CarvalhoFonts
8909 (BC). Dist. Porto: Porto, Boucas, Mattosinhos, v 1895, Sampaio s.n. (M, P). Dist. Stubal: Bank of river
Tejo ("ad Tagum"), Coina, iv 1840, Welwitsch 237 (P); "ad Tagum", 18481850, Welwitsch 993 (BM);
Transtagana (SE Portugal, later Prov. Alemtejo), near Fornos dEl Rei, vi 1851, Welwitsch 333 (P); Trafaria, iii
1888,Daveau348(P).CYPRUS.Dist.Famagusta:nearParalimnivillage,10iv1953,Kennedy1782(D3755)(K).
Dist.Limassol:Akrotiri,WsideofLimassolsaltlake.15iv1992,Lambinon92/Cy/359&vandenSande(MSB).
UK.Gibraltar.Palmones,7ii1913,WolleyDod1573(K),11iii1922,WolleyDod349(BM).CROATIA.Dalmatia.
Pag(island),nearVlasici,13v2005,Bernhardts.n.(WHB);nearSibenik,8vi2006,Bernhardt&Hermanowski
s.n.(WHB).GREECE.nearAthens,1820,Sartoris.n.(M);Faliro(alsoFaliron/Phaliron,latinPhalerum),iv1849,
Orphanidess.n.(P);Eleusine,19iii1876,Heldreichs.n.(HBG);v1876,Pichlers.n.(P);Faliro,iv1886,Heldreich
s.n. (B), 8 iv 1897, Heldreich s.n. (Z); Athens, Loutsa, 15 ii 1931, Guiol 1714 (BM). Peloponnese. Ahaia, near
Kalogria,33kmWofPatra,2v1995,Turland849(BM);betweenBriniaandManolada(Ilia),17v1996,Raabe
64
s.n.(B).Messeni:Ep.Pilia:VoidhokoiliaBaybetweenPetrohoriandPaleokastroNavarinou,05m,17iv1979,
Greuter & Merxmller 17215 (B, M); Methoni, iv 1981, Bowen 2051 (RNG). Lakonia: Ep. Epidhavros Limiras,
oppositeislandElafonisos,10iv1979,Greuter&Merxmller17076(B,M,P);mouthofriverEvrotas,nearElos,
iv 1979, Bowen 1307 (RNG); Vinglafia, opposite Elafonisos, 21 ii 1991, Jagel s.n. (B); Argolis, Ep. Ermionidhos,
3.3kmENEPortoheli,05m,14iv1996,Eisenbltter&Willing43.814,43.815(B);nearKalogriaSWofPatras,
25 iv 1991, Raabe s.n. (B); Evvia: ep. Karistias, W of Karistos, 010 m, 4 v 1996, S. & B. Snogerup 12651 (B);
Loutra Kounoupelli, road to Kalogria/Metohi, 7 v 1996, Raabe s.n. (B). Chios. c. 1930, Guiol s.n. (BM). Corfu.
Kastrades, am Hyllischen Hafen, 3 m, 22 iv 1896, Baenitz s.n. (B, HBG, M, P, ZT); Potami, 21 v 1887, name
illegible(ZT);Alikes,iv1980,Bowen1897(RNG).Crete.NearSuda(alsoSouda),iii1846,Heldreichs.n.(BM,P);
locationillegible,iii1846,Heldreichs.n.(BM,K,P).Kos.Ncoast,nearMarmari,03m,12iv1974,Buttler18237
(M).Naxos.Chora,15v1992,Bhling1491(B);Alluvai,17iv1993,Bhling2312(B).Samos.Tigani,2iv1934,
K.H.&F.Rechinger3642(BM,K).TURKEY.Prov.Izmir.Smyrna(=Izmir),29iii1854,Balansas.n.(P,ZT);Smyrna,
nearHalcarBunar(HalkarBunar),12v1906,Bornmller10018(B,BM,HBG,JE,P,ZT);CesmePeninsula,endof
GlbahceBay,25ii1966,Bocquet1133(ZT);betweenKarsyakaandIzmirnearBairakli,5iv1966,Bocquet1887
(ZT).
North Africa.ALGERIA. LaCalle (=El Kala),11 v1841, Durieu de Maisonneure s.n. (P);Algiers,Marais ou Baba
Alg., iv 1850, Jamin s.n. (P); Algiers, MaisonCarre, iii 1856, Wolfe s.n. (K); Algiers, MaisonCarre, 8 iv 1894,
Chevalier s.n. (P); Marais de la Meteija, iv 1851, Jamin s.n. (P); Salines dArgess, 6 v 1875, Cosson s.n. (P); La
Calle,18v1918,Clavs.n.(P);Algiers,maraisdelaRaasauraprsFortdel'Eau,28ii10iv1960,d'Alleizette
s.n.(M);nearGaraetzelMkhada,23iii1968,Leippert7107(B);LaCalle,510m,11v1971,Davis52174(RNG,
BM). LIBYA. Benghazi, 9 iv 1883, Ruhmer 315 (JE, HBG, P); Benghazi, 1884, Petrovich s.n. (Z); Tadjoura, near
Tripoli,8iv1886,Letourneuxs.n.(P);Tripoli,Mlacha,22iii1887,Taubert87(JE);Cyrenaica,withoutlocality,13
v1887,Taubert87(P);roadbetweenBenghaziandTocra,19iii1929,Riklis.n.(ZT);ElHania,SebkaChascem
elChelb, 8 Mag.(v) 1934, Pampanini & PichiSermolli 165 (K); near Tripoli, 29 iii 1938, Kracht 38/001, 38/002
(B);ElMagazin,nearBezenicem,20iv1938,Maire&Weiller1461(P);Barsis(alsoBersis),11iii1958,Guichard
CYR/96/58(BM);45kmNofBenghazi,3iv1970,Davis50488(K,RNG).MOROCCO.Tanger,CapSpartel,20iv
1911,Pitards.n.(P);Bouznika,23iv1979,60m,Lewalle9020(P);OuladAmara,NdeBouznika,40kmSWof
Rabat,23iv1979,Damblon79/63(B,RNG);Tanger,CapSpartel,10m,30iv1986,Deil1321(FB);OuedLaou,
ca.8kmEofTetouanonroadtoMartil,25m,25,ii1994,Jury13327&Taleb,Upson,Walters(RNG);Larache,
bridgeoverriverLoukkos,17iii1995,3m,Mateos,Reina,Sangalli,Sardn&Valds4824/95(RNG);Tahaddart,
Sansouire, 17 v 2003, Deil & Hamman 34 (FB); Larache, 15 v 2003, Deil & Hamman 5 (FB). TUNISIA. Bne
(=Annaba), 1834, Steinheil s.n. (P); Tunis, 4 iii 1874, DometAdanson s.n. (P); Tunis, Goletta, 9 iii 1880,
Armitage s.n. (RO); Tunis, La Goulette, 2 v 1883, Cosson, Letourneux, Reboud, Barratte, Bonnel s.n. (P);
HammamelLif,8v1883,Cosson,Letourneux,Reboud,Barratte&Bonnels.n.(P);ZaouietelMgaiz,18v1883,
Cosson,Letourneux,Reboud,Barratte&Bonnels.n.(P);CapBon,betweenKurbaandMenzel,Temim,v1883,
Cosson,Letourneux,Reboud,Barratte&Bonnels.n.(P);Annaba,nearBoudjima,28iii1884,Letourneuxs.n.(P);
RuinsUtique,21vi1887,Letourneuxs.n.(P);betweenDindjaandlakeIchkeul,23vi1887,Letourneuxs.n.(P);
betweenDjebelIchkeulandMateur,26vi1887,Letourneuxs.n.(P);Utique,4v1888,Barrattes.n.(P);Menzel,
Djemil,11v1888,Barrattes.n.(P);Tunis,LaGoulette,19v1888,Barrattes.n.(P);HammamelLif,20v1888,
Barratte s.n. (P); Menzel, Djemil, 2 vi 1888, Cosson, Barratte & Duval s.n. (P); HammamelLif, 6 v 1893,
Chevaliers.n.(P);DjebelDjeloud,iv1909,Pitards.n.(ZT);nearProtville,betweenTunisandBizerta,16iv1938,
Simpson38298(BM);CarthageTunisroad,28iv1939,Sandwith2845&Simpson39779(BM);ca.7kmSWof
Hammamet,14iv1968,Wagenitz1177(B);Tunis,12iv1971,Bolay&Kramers.n.(B);Kelibia,EsideofCapBon
peninsula,28iv1975,Davis56931&Lamond(BM,RNG).
2.TriglochinlaxifloraGuss.,IndexSeminumHortoBoccad.(PAL)1825:s.p.(1825);Fl.
Sic.Prod.1:451(1827).T.bulbosasubsp.laxiflora(Guss.)Rouy,Fl.France(Rouy)13:272
(1912).Type:Italy,Sicily,Palermo,PalermoaMontePeregrino[=MontePellegrino],
Gussones.n.(lectoNAPphoto!,designatedhere;Fig.5).Figs2B,3.
TriglochinpalustreDesf.,Fl.Atlant.1:322(1798),nonL.(nom.illeg.).
65
Plants1025(35)cmhigh,withfusiformbulbs;bulbscoveredbydry,brownfibreswhichareusually
stiff,sometimessoftorspiky;fibresupto5cmlong.Plantswithoutrhizomes.Leavesfew(mostly24
perbulb),formedafterflowering,usuallymuchshorterthanplantheight(mostlyreachingonlyhalf
of the height of plants). Outer leaves shorter and wider than inner leaves, up to 6 cm long,
mucronate, inner leaves abruptly narrowing above the ligule, less than 1.5 mm wide. Inflorescences
lax, with 415(25) flowers. Flowers 2.03.0 mm long. Pedicels 23 mm long, rarely longer, usually
muchshorterthanfruit,ascending,appressedtoinfructescenceaxis,divergingfromitatanglesof
up to 45, slightly curved inwards towards the apex. Infructescences 310 cm long. Fruits narrowly
ovoid, 510 mm long, rarely longer, 12 mm wide (Fig. 2B). Mericarps linear, connate at the
carpophore;tipscurvedoutwards.
Distribution. Widely distributed in the Mediterranean region, largely sympatric with Triglochin
barrelieributabsentfromtheAtlanticcoastofFranceandfrompartsoftheeasternMediterranean
(Fig.3).
Habitatandecology.Thisspeciesistypicallyfoundinsaltmarshesandalonglagoons,butalsoincork
oak or pine forests (e.g., Morocco, Andalusia) and in Mediterranean temporal pools (e.g., Corsica,
Malta). Triglochin laxiflora is sometimes found in limestone areas (e.g., Malta, Sicily). Growing from
0500m(e.g.,MontePellegrino,Sicily),inAlgeriareachingca.1100m(e.g.,DjebelOuach).
Phenology.Thespeciesflowersandfruitsinautumn(ixxii)andleavesareformedafterflowering.
Chromosomenumber.2n=18(Gard&MalheirosGard,1953;Bartoloetal.,1977;Talaveraetal.,
1995;Castroetal.,2007).
Notes. The distributional gaps in many parts of the eastern Mediterranean might be a collecting
artefact resulting from the autumnal flowering time of the species. Dandy (1980) indicated Albania,
Crete, Greece and exJugoslavia as part of the distribution area in the eastern Mediterranean.
TriglochinlaxifloraisnotknownfromTurkey(Uotila,1984).
Although reported from South Africa by some authors (Bennett, 1897; Buchenau, 1903), these
recordsareduetomisidentificationandrefer,atleastinpart,toTriglochinelongataorT.compacta.
Triglochin laxiflora does not occur there and has to be excluded from the South African flora
(Adamson,1939;HornafRantzien,1961).
Taxonomic remarks. See Triglochin barrelieri for a short comparison with the only other Medi
terraneanspeciesofthecomplex.
Nomenclatural notes. In the protologue of Triglochin laxiflora Gussone (1825) does not cite any
specimens or localities, but in his Florae Siculae Prodromus (Gussone, 1827) the following localities
are given: Palermo a Monte Peregrino a Monte Castellana sotto Baida Mondello Sferracavallo,
Catania,Agosta,Siracusa.SeveralTriglochinspecimensfromGussonescollectionsintheHerbarium
Neapolitanum (NAP) were studied. Four specimens of Triglochin laxiflora are labelled with localities
that were cited in Gussone (1827). The specimen from the Monte Pellegrino in Palermo was here
chosen as lectotype (Fig. 5) because it contains several fruiting and some flowering plants in good
condition.
66
Variation. One herbarium sheet (BC 76587) is a plant grown from bulbs in the Botanic Garden of
Barcelona.Thisspecimenhasexceptionallylargefruitswhichare1012mmlong.
ProposedIUCNconservationstatus.LeastConcern(LC).Thespeciesusedtobewidelydistributedin
the Mediterranean region and probably is not under immediate threat. However, the habitat
degradation problems described for Triglochin barrelieri also apply here. Mediterranean temporary
pools have also been destroyed in considerable numbers by human activities (e.g., Grillas et al.,
2004; Ruiz, 2008). Only few data are available on the current area of occupancy. In contrast to the
many collections of the species from the 19
th
century only a small number of recently collected
specimens has been seen. Several collecting localities of older specimens have been destroyed by
settlements. Other populations might also be destroyed or threatened by habitat destruction. This
mightimplythatthespecieshasbecomerare,andtheglobalconservationstatusofthisspeciesmay
be prone to change. Further conservation assessments are also necessary at the national level.
Triglochin laxiflora is a protected species in France (Musum National dHistoire Naturelle, 2003
2006)andinLazioandSicily,Italy(Contietal.,1997).
Specimensexamined.Europe.FRANCE.Corsica.Bonifacio,11x1856,Revelire494(BM,P);PortoVecchio,16x
1858,Revelires.n.(P);Ajaccio,20ix1868,Mabilles.n.(JE,K,P);Bonifacio,26x1880,Reverchon393(BM,JE,
P);Ajaccio,1881,Reverchons.n.(ZT);Bonifacio,x1885,Reverchon393(B);Ajaccio,Parata,16x1888,LeGrand
s.n.(P,RO);Bonifacio,9x1901,Stefanis.n.(BM,WTU),6x1903,Stefanis.n.(ZT),14x1910,Stefanis.n.(P);
Ajaccio, Parata, 1 x 1916, ForsythMajor 292 (K); Cargese, x 1923, Wyatt 63 (K); Calvi, golfe de la Revellata,
plagedel'Alga,14x1979,Lambinon79/1043,Bellotte,Dellens&Monfort(BC,MSB,RNG);Bonifacio,Tonnara
plage,3x1993,Dutartres.n.(B,M,MSB,RNG).ITALY.Sardinia.Prov.CarboniaIglesias:Portoscuso,9x1912,
Bonomis.n.(BM).Prov.OlbiaTempio:SantaTeresaGallura,Tempio,Bancamino,20ix1881,Reverchon194(B,
HBG, JE, K, M, RO, ZT); Terrnanova (=Olbia), 2 x 1922, Rikli s.n. (ZT); Terranova, x 1922, Schinz s.n. (ZT); road
betweenOlbiaandSanTeodoro,km310,LeVecchieSaline,closetolagoons(stagno),6x2006,vonMerings.n.
(MJG).Prov.Sassari:Sassari,ix1895,Nicotras.n.(RO),ix1899,Nicotras.n.(B,M);IsoladiReulino,23x1988,
Bocchieris.n.(B).Sicily.Prov.Agrigent:Licata,x1911,Rosss.n.(ZT).Prov.Palermo:Palermo,Mte.Pellegrino,
ix1855,E.&A.HuetduPavillons.n.(BM,P,ZT);Mondello,x1879,Lojaconos.n.(K,P);Castelbuono,x1885,
Lojaconos.n.(P);Palermo,x1885,Todaros.n.(ZT),1886,Lojaconos.n.(K),x1892,Todaros.n.(RO);Palermo,
Aquasanta, 19 x 1897, Fisch s.n. (ZT); Palermo, x 1901, Ross 377 (B, BC, HBG, JE, M, ZT); Palermo, Mte.
Pellegrino, 300 m, 25 x 1903, Spencer s.n. (M); Palermo, x 1903, Ross s.n. (HBG, M), x 1906, Ross s.n. (BC);
Palermo,Mte.Pellegrino,430m,30x1935,Lusirnes.n.(RO);Palermo,Mte.Pellegrino,400m,7ix1964,Davis
40243 (RNG). Puglia. Prov. Taranto: Taranto, Leucaspide, 23 x 1905, Lacaita 4808 (BM), 2 xi 1909, Lacaita
11461 (BM, P). Tuscany. Prov. Livorno: Isola di Capraia, 7 i 1898, Bguinot s.n. (RO). MALTA. Mtahleb, W of
Rabat,nearcoastalcliffs(limestone),ca.600m,28x1974,Westra&Rooden252(Z);Mosta,WiedilGhasel,22
xi2007,Mifsuds.n.(MJG).SPAIN.Prov.Cdiz:Cdiz,LagunadelaPaja,Chiclana,20m,29x1925,Gross.n.(M,
RNG), ix 1929, Font Quer s.n. (RNG); entre Los Barrios y Casas del Castao, Puerto de la Cebada, 1 xi 1967,
Galiano&Valds128/67(RNG);Cdiz,LagunadelaJanda,3xi1978,Galiano,Rivera&Valds7097/78(RNG).
Prov.IslasBaleares:Menorca.MarinadeBenisarmeaySanAntonio,ix/xsinanno,Tremolss.n.(BC);without
locality, 13 x 1865, Rodriguez s.n. (P); near Mahn, 21 x 1872, Rodriguez s.n. (P); near Mahn, 21 x 1872,
Rodriguezs.n.(P);S.Antonio,x1872,Rodriguezs.n.(P).
NorthAfrica.ALGERIA.Algiers,PointePescade,ix(withoutyear),Jamin208(P);LaCalle(=ElKala),17xi1840,
DurieudeMaisonneures.n.(P);Mostaganem,14ix1848,Barlanez(partlyillegible)s.n.(P);Algiers,CapCaxine,
x 1850, Jamin s.n. (P); Algiers, Cap Caxine, xii 1850, Jamin s.n. (P); Algiers, x 1851, Jamin s.n. (P); Bne
(=Annaba),1860,Letourneuxs.n.(B,K,P);Annaba,OuedFourcha,25x1865,Tribouts.n.(BM,P);Constantina,
mountain Djebel Ouach, ca. 1100m, 10 xi 1868, Paris s.n. (BM, P); Algiers, 1869, Durando s.n. (P); Algiers, La
Reghaia, xi 1880,Battandier & Trabut 986(P, RO); Berrouaghia, road to Mda, 1100 m, 6 x1886, Luizet s.n.
(P); MaisonCarre, x 1886, Luizet s.n. (P); Annaba, SaintClous, 6 xi 1890, Luitfroy 337 (P); Bne, 29 x 1891,
67
Luitfroy 370 (P); Bne, 1906/1907, Gandoger s.n. (BM). MOROCCO. Lukos, El Araix, 4 xii 1929, Font Quer s.n.
(BM); Chaouia, Bouskoura, 190 m, x 1932, Gattefoss s.n. (M, P); 20 xii 1936, Gattefoss s.n. (B, JE, K, P);
Tanger,betweenCapSpartelandtheAirport,30x1993,75m,GarcaMurillo,Gibb&TalaveraST40/93(RNG).
3.TriglochinbulbosaL.,Mant.Pl.:226(1771).Type:SouthAfrica,CapeofGoodHope,
(neoLINN466.3photo!,BOLphoto!);MalmesburyDist.,atturnouttoGansekraal,15ix
1940,Compton9451(epiNBG!,designatedhere,Fig.8).Figs6,7.
TriglochinbulbosarobustiorRoem.&Schult.,Syst.Veg.,ed.15bis[Roemer&Schultes],
7(2):1585(1830)(testeKunth,Enum.Pl.(Kunth)3:144(1841).
TriglochinpatensSteud.,Nomencl.Bot.,ed.2,2:709(1841)nomilleg.=T.bulbosaL.
robustior(testeKunth,Enum.Pl.(Kunth)3:144,1841).
Plants550cmhigh,withfusiformbulbsandwithoutrhizomes,bulbscoveredbydry,brownorblack,
soft or stiff fibres, these 0.53.0 cm long. Leaves present when flowering, uniform, as long as or
longer than plant height. Inflorescences lax or dense, with 450 flowers. Flowers 1.53.0 mm long.
Pedicels elongated at fruiting time, 210 mm long, diverging at angles of 45 or 4590 from the
infructescence axis, not or only slightly curved inwards towards the apex. Infructescences 220 cm
long. Fruits narrowly to very narrowly ovoid, 410(12) mm long, 0.52.0 mm wide (Fig. 6), turning
dark or black at fruit maturity. Mericarps linear, connate at the carpophore; tips not or only slightly
curvedoutwards.
Distribution.EndemictoSouthAfrica,whereitisrestrictedtotheWesternCapeProvince(mainlyon
CapePeninsula,WestCoastandinOverberg)andtheHantamregionoftheNorthernCapeProvince
(Fig.7).
Habitatandecology.Variable(seesubspecies),rangingfromtemporarypoolstodry(shaded)inland
habitats,butneverinhabitatsunderdirectinfluenceoftidalflooding.
Phenology.Floweringinspring(viixi).
Chromosomenumber.Unknown.
Nomenclatural notes. The neotype (LINN 466.3, photo!) designated by Obermeyer (1966) is
ambiguous. Most importantly, its fruits are not mature which makes identification difficult.
Additionally no data on locality or collecting date are given. Thus, an additional epitype (Compton
9451,NBG;Fig.8)wasdesignatedherewhichshouldbeusedinconjunctionwiththeneotype.
Taxonomic remarks. We characterize Triglochin bulbosa as a bulbous species with narrowly to very
narrowlyovoidfruits,theoutlineofseedsnotvisiblethroughthepericarp,andpedicelsspreadingat
an angle of 4590 from the inflorescence axis. The fruits of T. bulbosa often turn dark or black at
maturity unlike those of the other South African species of the complex. The species differs from T.
compacta, which has a similar distribution area, in flowering time (spring vs. late summer/autumn)
and the presence of leaves at flowering time (vs. leaf formation after flowering). Additionally, the
fruits of T. bulbosa have no basal outgrowths and the outline of the seeds is not visible on the
mericarpsurface(asinT.compacta).
68
Fig.4.FruitsandinfructescencestructureofTriglochinbulbosaL.A,T.bulbosasubsp.bulbosa;B,T.bulbosa
subsp.tenuifolia(Adamson)Horn;C,T.bulbosasubsp.CalcicolaMering,Kcke&Kadereitsubsp.nov.;D,T.
bulbosasubsp.quarcicolaMering,Kcke&Kadereitsubsp.nov.DrawingsbyL.Klckner(Mainz).
Adamson (1939) newly described Triglochin tenuifolia on the basis of narrow leaves, fewflowered
spikes, and its characteristic habitat. However, this species was later reduced to a subspecies of
Triglochin bulbosa by Horn af Rantzien (1961). The newly circumscribed Triglochin bulbosa shows
substantialvariationinwidthofleavesaswellassizeandnumberofflowersandfruits.Severalforms
are geographically and ecologically welldifferentiated, but transitions in morphological characters
are found. Thus, we recognize four subspecies in Triglochin bulbosa: subsp. bulbosa, subsp.
tenuifolia,subsp.calcicolaandsubsp.quarcicola.
69
Keytothesubspecies
1a.Plants(8)2050cmhigh,robust;leaves13mmwide;fruits12mmwide;
widelydistributedintheWesternCapeandsouthernregionsoftheNorthernCape
........................................................................3A.subsp.bulbosa
1b.Plants(5)825(35)cmhigh,delicate;leaveslessthan1mmwide;fruitsless
than1mmwide;distributedintheWesternCape..........................................2
2a.Inflorescenceslaxwith415(25)flowers,pedicelsdivergingat4590or90
fromtheinflorescenceaxis,plantsgrowingongranitesoilsorinquartzfields
............................................................3
2b.Inflorescencesdensewith(4)1030flowers,pedicelsdivergingat45
fromtheinflorescenceaxis,plantsgrowingonlimestoneoraeolinitein
coastalregionsoftheWesternCape
........................................................................................3C.subsp.calcicola
3a.Pedicelsdivergingatanglesof4590fromtheinflorescenceaxis,plants
growingongranitesoilsonmountainslopesoftheCapePeninsulaand
LangebaanPeninsula
..............................................................................3B.subsp.tenuifolia
3b.Pedicelsdivergingatanglesofmostly90fromtheinflorescenceaxis,
plantsgrowinginquartzfieldsoftheKnersvlakte
...................................................................................................................3D.subsp.quarcicola
3A.subsp.bulbosaFigs6A,7.
Plants (8)2050 cm high, with fusiform bulbs, bulbs covered by dry, rather stiff, brown or black
fibres, these up to 3 cm long, usually separate, sometimes fused into layers. Leaves as long as or
longer than plant height, 13 mm wide. Inflorescences dense to lax, with (8)1550 flowers. Flowers
1.53.0 mm long, rarely longer. Pedicels 210 mm long, diverging at angles of 45 to 90 from the
inflorescenceaxis,notoronlyslightlycurvedinwardstowardstheapex,butoftenfruitsbentfurther
upwards.Infructescences620cmlong.Fruitsnarrowlyovoid,510(12)mmlong,12mmwide(Fig.
6A).Mericarpconnateatthecarpophore;tipsslightlycurvedoutwards.
Distribution.EndemictoSouthAfrica,restrictedtotheWesternCapeProvinceandsouthernregions
(Hantam)oftheNorthernCapeProvince(Fig.7).
Habitat and ecology. Found in temporary wetlands, such as edges of vernal pools, in ditches and in
other depressions, mainly in vegetation types classified as Cape Vernal Pools and Cape Inland
Saltpans,andrarelyalsoSwartlandGraniteRenosterveld(Mucina&Rutherford,2006).Thehabitats
arecharacterisedbyheavyclayeytolighterclayeysandysoils,derivedfromnutrientrichsubstrates
such as shale, granite or geologically young sandy sediments of marine origin covering coastal
lowlands; here it is often found in clayfilled depressions among stabilised sandy dunes. Water
loggingseemstobeamajorhabitatstructuringfactor.Thetaxonwasfoundataltitudesofupto500
m(CederbergMountains),butmostlocalitiesarelimitedtoaltitudesbelow100ma.s.l.
Phenology.Floweringandfruitinginspring(viiixi).
70

Fig.5.GeographicaldistributionofTriglochinbulbosaL.subsp.bulbosa,T.bulbosasubsp.tenuifolia(Adamson)
Horn,T.bulbosasubsp.calcicolaMering,Kcke&Kadereitsubsp.nov.,andT.bulbosasubsp.quarcicola
Mering,Kcke&Kadereitsubsp.nov.
Notes. One specimen (Wrz 04.10.06.01, PERTH n.v., STU!) which is morphologically very similar to
Triglochin bulbosa ssp. bulbosa was collected in 1994 in Australia (Western Australia, SwanValley
near Perth), where the taxon has become established as a recent introduction (H. Aston, pers.
comm.;seenotesforT.bulbosacomplex).
Taxonomicremarks.Subspeciesbulbosadiffersfromtheothersubspeciesinhavingwiderleaves(13
vs.<1mmwide)andwiderfruits(12vs.<1mmwide).Theplantsaremorerobustandoftenlarger
thanthoseoftheothersubspecies.
Proposed IUCN conservation status. Least concern (LC). The subspecies is widely distributed in the
WesternCapeProvince,withsomelocalitiesfoundinsouthernregionsoftheneighbouringNorthern
CapeProvince.Itdoesnotappeartobeimmediatelythreatened.However,asseasonalpools(asall
lowland wetlands in general) are becoming rare due to cultivation of land and construction work
(Mucina&Rutherford,2006),theconservationstatusshouldbereassessedinthenearfuture.
71
Specimensexamined.SOUTHAFRICA.NorthernCape.Nieuwoudtville,CloudskraalFarm,7xi1962,Barker9799
(NBG); Nieuwoudtville, 3 miles S of Nieuwoudtville, 7 ix 1963, Nordenstam 3032 (S). Western Cape.
Stellenbosch, Hottentotts Holland, sine dato, Zeyher & Ecklon s.n. (P); Paarl, Klein Draakensteen, 10 ix 1827,
Drge 8798 (P); Cape Town, Simons Bay, Constantia, 18531856, Wright s.n. (NY); Cape Town, Raapenburg
Vlei,viii1882,Guthrie1242(BOL);Caledon,nearvillage,vii1892,Guthrie2526(NBG);CapeTown,Kenilworth,
anno 1892, Bolus 7926 (BOL); Ceres, Mostertsberg, 1200 feet, x 1894, MacOwan 1991 (BM, P, NBGSAM, Z);
CapeTown,KenilworthRaceCourse,5ix1897,WolleyDod2858(NBG);CapeTown,nearMaitlandCemetery,2
x 1897, Wolley Dod 3221 (BOL); Cape Town, Vaarsche Vley, 17 x 1897, Wolley Dod 3139 (BOL); Cape Town,
Simonstown, Constantia, Bergvliet Farm, 11 ix 1916, Purcell 66 (NBGSAM); Cape Town, Simonstown,
Constantia,BergvlietFarm,Eoflake,11ix1917,Purcell3(NBGSAM);Stellenbosch,behindBoysHighSchool,8
xi1918,Garside1139(K);Worcester,GoudiniRoad,x1921,Michell2833(BOL);Clanwilliam,ZeekoeVlei,1500
feet,ix1925,Levyns1192(BOL);CapeTown,SomersetWest,ix1931,Schmidt437(M);Bredasdorp,between
Bredasdorp and Elim, damp places near salt vlei, ix 1933, Levyns 4542 (BOL); Bredasdorp, The Poort, sides of
temporary pool, 100 feet, ix 1933, Levyns 4460 (NBG); Stellenbosch, Faure, 9 ix 1934, Garside 4649 (K); Cape
Town, Rondebosch Common, 15 vi 1936, Adamson 940 (BOL); Cape Town, Kenilworth Race Course, 6 x 1936,
Adamson1200(BOL);Tulbagh,SaronFlats,viii1937,Compton&party1992/36(BOL);Clanwilliam,Alpha,28vi
1938,MartinNBG1258/37(NBG);OlifantsrivierValley,11milesNofClanwilliam,2ix1938,Salter7529(BOL);
Napier,24viii1940,Bond470(NBG);Darling,15ix1940,Esterhysen3864(BOL);[Yzerfontein],atturnoutto
Gansekraal, 15 ix 1940, Compton 9451 (epi NBG); Piketberg, Berg River, 21 ix 1940, Compton 9474 (NBG);
Clanwilliam, Pakhuis, 29 ix 1940, Compton 9553 (NBG); Clanwilliam, Pakhuis, x 1940, Esterhuysen 3170 (BOL);
Cape Town, Hout Bay, 4 ix 1941, Compton 11294 (NBG); Malmesbury Dist., Mamre Hills, 7 ix 1941, Compton
11608(NBG);Strand,8viii1942,Parker3708(BOL,NBG);CapeTown,Milnerton,31viii1942,Compton13424
(NBG); Cape Town, Firgrove, 3 ix 1942, Compton 13458 (BOL); Malmesbury Dist., Mamre Hills, 12 ix 1942,
Barker1596(NBG);Stellenbosch,StellenboschVlakte,8iv1943,Jordaan18(NBG);MalmesburyDist.,footof
Mamre Hills, 22 ix 1943, Henderson 1858 (NBG); Malmesbury Dist., Kalabas Kraal, 28 ix 1943, Barker 3546
(NBG);Piketberg,PapkuilsVlei,30ix1943,Leighton119(BOL);Piketberg,SoutKloof,betweenSauerandBerg
River, 1 x 1943, Compton 15122 (NBG); Piketberg, Zoutkloof, 1 x 1943, Leighton 397 (BOL); Bredasdorp, The
Poort,2xi1943,Barker2491(NBG);CapeTown,MelkboschRoad,swamp,31viii1944,Compton15865(NBG);
Stellenbosch, Faure, swamp, 16 xi 1944, Compton 15984 (NBG); Citrusdal, 1 ix 1945, Leighton 1389 (BOL);
Ceres,ColdBokkeveld,nearElandskloof,9ix1945,Leighton1270(BOL);Ysterfontein[Yzerfontein],12ix1945,
Compton 17379 (NBG); Darling, ix 1945, Stokoe 59918 (NBGSAM); Cape Town, Kirstenbosch, ix 1945,
Esterhuysen11856(BOL,NBG);Citrusdal,CitrusdalVlei,1xi1945,Compton17106(NBG);Paarl,flatsNofPaarl,
30 viii 1946, Leighton 1961 (BOL, NY); Cape Town, Kenilworth Race Course, ix 1946, Lewis 1900 (NBGSAM);
Piketberg, Berg Valley, ix 1947, Lewis 2380 (NBGSAM); Clanwilliam, Nardouw Kloof, ix 1947, Stokoe 61194
(NBGSAM); Cape Town, Kraaifontein, 4 x 1947, Compton 20091 (NBG); Stellenbosch, Golfplatz, 6 viii 1948,
Rehm1950(M);Stellenbosch,StellenboschVlakte,ix1948,Malherbes.n.(NBG);CapeTown,Claremont,19ix
1951, Salter 9302 (BM); Cape Town, between Fernwood and Liesbeek River, 24 ix 1951, Salter 9045 (BOL);
Darling,DarlingFloraReserve,25vii1956,Rycroft1973(NBG);Darling,DarlingFloraReserve,4x1956,Lewis
5087 (NBG); Cape Town, Rondebosch Common, 15 x 1962, Nordenstam 1564 (S); Cape Town, Durbanville,
Peaslake,15viii1963,Taylor5009(NBG);CapeTown,RondeboschCommon,7ix1963,Stauffer5136(NBG,P,
Z); Darling, 16 ix 1963, H. & E. Walter s.n. (B); Stellenbosch, 3 miles from Simondium to Stellenbosch, 16 ix
1968,Marsh653(NBGSTE);Caledon,HappyValleyFarm,HighNoonEstates,onroadfromCaledontoFrensch
HoekPass,belowEzeljacht,30ix1971,Barker10834(NBG);Worcester,Bokkekraal,NWsideofBrandvleiDam,
18 ix 1974, Mauve & Oliver 254 (K); Cape Town, Rondebosch Common, 21 ix 1974, Nordenstam & Lundgren
1973 (S); Bettys Bay, Groot Vleie, moist peaty area to south of vleis, 13 xi 1974, Boucher 2657 (NBGSTE);
Darling, Platteklip, ix 1976, Liebenberg 8301 (K); Clanwilliam, outside Sandberg on road to Witels Kloof, 13 x
1976,Hugo671(NBGSTE);CapeTown,IsoetesVlei,anno1977,Gubb50(NBG);Malgas,DeHoopPotbergNat.
Res.,Potberg,flatsnearMelkbosheuwel,moistsandyflats,16ix1979,Burgers2252(STE);Tulbagh,foothillsof
WitzenbergMountains,7ix1980,Schonken314(NBGSTE);Malmesbury,Klipfonein,16ix1982,vanZyl3247
(NBG);CapeTown,RondeboschCommon,15ix1983,Koutnik1262(BOL);Worcester,WorcesterCommonage,
26viii1985,Bayer4837(NBG);CapeTown,SomersetWest,publicopenspaceatWestridge,5ix1993,Runnalls
72
574(NBGSTE);CapeTown,Kraaifontein,Scottsdene,betweenScottsdeneLibraryandWolwefonteinRd.,29ix
1997,Cupido48(NBG);Malgas,DeHoopNat.Res.,Potbergsection,valleyofPotbergsrivier,3iv2006,Kcke&
Mucina030406/22(MJG);Darling,TienieVersveldFlowerReserve,attheentrancetothereserve,14iv2006,
Kcke & Mucina 140406/01 (MJG), Yzerfontein, E of Salt Pan, 14 iv 2006, Kcke & Mucina 140406/33 (MJG),
140406/34(MJG);Stellenbosch,Klipheuwel,Sofroadcrossing,19iv2006,Kcke&Steffen190406/02(MJG);
Velddrif, road Velddrif to Piketberg, Tasaars Kuil Farm 253, at turnoff to Aurora, 19 iv 2006, Kcke & Steffen
190406/27 (MJG); Wellington, road between Perderberg and Wellington, Boland Agricultural College, 30 vii
2006,Mucina300706/21(MJG);Darling,roadbetweenMalmesburyandDarling,LangekloofFarm,4viii2006,
Mucina&Jakubowsky040806/02(MJG);Darling,roadbetweenMalmesburyandDarling,nearWaylands,4viii
2006, Mucina & Jakubowsky 040806/05 (MJG); Velddrif, road Velddrif to Piketberg, Tasaars Kuil Farm 253, at
turnofftoAurora,4viii2006,Mucina&Jakubowsky040806/28(MJG);Elim,roadBosheuwelandWiesdrif,14x
2006, Mucina 141006/20 (MJG); Paternoster, Tietiesbaai, coastal rocks at the entrance to Cape Columbine
Nature Reserve, 25 x 2008, Mucina 251008/05 (MJG), 251008/06 (MJG); Darling, Tienie Versveld Flower
Reserve,attheentrancetothereserve,26x2008,Mucina261008/04(NBG).
3B.subsp.tenuifolia(Adamson)Horn,SvenskBot.Tidskr.55:85(1961).Triglochin
tenuifoliaAdamson,J.S.Afr.Bot.5:3031(1939).Type:SouthAfrica,TableMountainWest
side,PipeTrack,shelteredslopes,5ix1936,Adamson986(holoT.tenuifoliaBOL!).Figs6B,
7.
Plants 825(35) cm high, delicate, slender, with fusiform bulbs, bulbs covered by dry, rather stiff or
soft, brown fibres, these < 1 cm long. Leaves as long as or much longer than plant height, very
narrow, about 0.51 mm wide. Inflorescences lax, with 415 flowers. Flowers 1.52.5 mm long.
Pedicelsaslongasorlongerthanfruitatfruitingtime,510mmlong,divergingatanglesof45to90
from the inflorescence axis, slightly curved inwards towards the apex. Infructescences 25 cm long.
Fruits very narrowly ovoid, 58 mm long, 0.51.0 mm wide (Fig. 6B). Mericarp connate at the
carpophore;tipsslightlycurvedoutwards.
Distribution.Thissubspeciesis(asfarasknown)restrictedtothelowerslopesofTableMountainon
theCapePeninsulaandtothePostberg(nearLangebaan,WestCoastNationalPark)(Fig.7).
Habitatsandecology.Thesubspeciesisfoundinsheltered,shadyinlandhabitatsonmountainslopes
ataltitudesbelow400m.Itisrestrictedtocoarse,sandysoilsderivedfromgranites(ArchaeanCape
Granite Suite). The vegetation types supporting this taxon are classified as Cape Peninsula Granite
RenosterveldandSaldanhaGraniteStrandveld(Mucina&Rutherford,2006).
Phenology.Floweringandfruitingviiix,onespecimen(withflowersandfruits)wascollectedinv.
Notes.ThepreferenceofthistaxonforgranitesoilswasalreadynotedbyAdamson(1939).
Taxonomic remarks. Subspecies tenuifolia differs from the other subspecies in having lax inflores
cenceswith415flowersandpedicelsdivergingatanglesof4590fromtheinflorescenceaxis.This
subspecies is restricted to granite soils on mountain slopes of the Cape Peninsula and Langebaan
Peninsula.
73
Nomenclatural notes. At least one paratype of Triglochin tenuifolia (Dmmer 1050, SAM) cited by
Adamson (1939) represents a misidentified specimen of T. compacta. Not all material cited by
Adamsonhasbeenseenbyus.
Proposed IUCN conservation status. Near Threatened (NT): The taxon meets the area requirements
undercriterionBforthreatened(EOO<20,000km
2
and/orAOO<2,000km
2
)andisdeclining,butthe
populationisnotseverelyfragmented,andthetaxonoccursat12locations.
Judgingfromtheavailableherbariummaterial,thesubspecieshasaveryrestricteddistributionarea.
The extent of occurrence (EOO) is estimated to be only several square kilometres. It is known from
parts of the Cape Peninsula and one locality at the Postberg (Langebaan Peninsula). All of the
historical localities are today protected within the Table Mountain National Park and the private
PostbergNatureReserve(contractualportionoftheWestCoastNationalPark).Nodataareavailable
on the current area of occupancy (AOO) of Triglochin bulbosa subsp. tenuifolia as all herbarium
specimensseenareatleast60yearsold,apartfromonecollectionfromPostbergdatingfrom1966.
SeveralcollectingsitesarelocatedwithinthebordersofCapeTownandmighthavebeendestroyed
longago.Fieldworkisneededtoassessthecurrentrangeandtheconservationstatusofthistaxon.
Specimensexamined.SOUTHAFRICA.WesternCape.CapeTown,Stinkwater,anno18751880,Rehmann1194
(BM, Z); Cape Town, locality illegible, 14 vii 1883, Wilms 36451 (BM, Z); Cape Town, top ridge over
Klassenbosch, 15 viii 1897, Wolley Dod 2918 (BOL); Cape Town, Lion's Head, lower west slopes, 29 viii 1897,
WolleyDod 2915 (para BOL); Cape Town, Signal Hill, ix 1897, Thode 6089 (NBGSTE); Cape Town, Table
Mountain,PipeTrack,aboveCampBay,viii1920,Michells.n.(BOL);CapeTown,Simonstown,BaviaansKloof,4
v1927,Salter344/10(paraBM);CapeTown,HoutBayNek,23viii1928,Hutchinson&Pillans104(BOL);Cape
Town,KloofNek,27vii1929,Scotts.n.(paraBM);CapeTown,TableMountainWside,PipeTrack,5ix1936,
Adamson986(holoT.tenuifolia,BOL);CapeTown,TableMountain,Wsideat1000feet,11viii1940,Adamson
2899 (BM); Cape Town, Table Mountain, west slopes, 5 ix 1943, Esterhuysen 27104 (BOL); Cape Town, Table
Mountain, Spring Buttress, ix 1945, Stokoe 59916 (NBGSAM); Cape Town, Table Mountain, Llandudno end,
1000 feet, 19 x 1958, Esterhuysen 27923 (BOL); Langebaan, E slopes of Postberg, 9 ix 1966, Barker 10471
(NBG).
3C.subsp.calcicolaMering,Kcke&Kadereitsubsp.nov.Type:SouthAfrica,Western
Cape,DeHoopPotbergNatureReserve,Dronkvlei,alongtracktoKoppieAlleen,shallow
sandysoilsoverlimestone,lowrestiad[restioid]vegetation,frequentindampplacesand
hollows,11ix1979,Burgers2172(holoNBGSTE!;Fig.9).Figs1AB,6C,7.
Differtasubspecietypicahabituminoreetgraciliore,foliisperangustis,infructescentiadensa;soloin
substratocalcareo.
Plants(5)1025cmhigh,stifflyerect,withsmall,fusiformbulbs,bulbscoveredbythinsoftorthicker
stiff,brownfibres,these1cmlong,fibressometimesnotseparatebutfusedintolayers.Leavesas
longasinfructescence,verynarrow,0.5(1.0)mmwide.Inflorescencesdense,with(4)1030flowers.
Flowers 1.52.5 mm long, number of stamens sometimes reduced to three. Pedicels 24 mm long,
divergingatanglesofapproximately45fromtheinflorescenceaxis,slightlycurvedinwardstowards
theapex.Infructescences28cmlong.Fruitsverynarrowlyovoid,47mmlong,rarelylonger,mostly
< 1 mm wide, very regularly spaced (Fig. 6C). Mericarp connate; tips short and slightly curved
outwards.Carpophoreonlyverythinandshortorlacking.
74
Fig.6.HolotypeofTriglochinbulbosasubsp.calcicolaMering,Kcke&Kadereitsubsp.nov.(NBGSTE).
Distribution.ThistaxonislimitedtopartsoftheWestCoast(LangebaanPeninsula),thecoastofFalse
Bay(SwartklipandMacassar)andtotheSouthCoast(DeHoopandStillbaairegions)(Fig.7).
75
Habitatandecology.SubspeciescalcicolaisastrictendemicofTertiarycalcareousbedrockssuchas
limerich aeolinites (False Bay) and limestones of the Bredasdorp Group (De Hoop and Stillbaai
regions, Fig. 1B) and of the Sandveld Group (Langebaan and Saldanha regions). The sparse
populationsofthissubspecieshereoccurinshallowloamfilleddepressions(filledwithfineloamand
limerichsand)inlimestonepavements;thesedepressionsarewetinwinterandspringanddryout
completely in summer. The vegetation types housing this taxon are De Hoop Limestone Fynbos,
CancaLimestoneFynbos,andSaldanhaLimestoneStrandveld(Mucina&Rutherford,2006).
Phenology.Floweringandfruitinginspringfromviiixi.
Etymology. The epithet calcicola refers to the typical substrate of the species (hard limerich
substratessuchaslimestonesandaeolinites).
Notes.Thissubspeciesisnotwellstudied.Itsdelimitationfromthebroadlysympatric(allcooccuring
intheWestCoastregion)subsp.bulbosaandsubsp.tenuifoliarequiresfurtherresearch.
Taxonomic remarks. Subspecies calcicola differs from the other subspecies in having dense
inflorescenceswith(4)1030flowersandpedicelsdivergingatanglesofapproximately45fromthe
inflorescence axis. This subspecies is restricted to limestone or aeolinite in coastal regions of the
WesternCape.
Proposed IUCN conservation status. Vulnerable (VU B2ab(iii)). This subspecies is ecologically
specialised and occurs only on limestone substrates, which are generally rare in the Western Cape.
Its populations on the South Coast (Overberg and Canca) and part of the False Bay coast do not
experience direct threats because of current effective conservation (De Hoop Nat. Res., Pauline
Bohnen Nat. Res., Wolvengat Nat. Res.). However, the populations on the West Coast (Langebaan
andSaldanha)arefoundoutsideformallyprotectedareas;limestoneminingaswellasdevelopment
of coastal settlements may pose serious threats to these populations. Based on the limited area of
occupancy (AOO, estimated to be less than 2000 km
2
) and severely fragmented distribution area,
plusthethreattothehabitat(atleastinsomeareas)thissubspeciesishereclassifiedasVulnerable.
Specimensexamined.SOUTHAFRICA.WesternCape.Bredasdorp,ThePoort,limestoneridge,indepressionsin
rock on top, 400 feet, ix 1933, Levyns 4491 (BOL); Strand, Macassar Downs, between St. Joseph's Tomb and
sea, 2 ix 1934, Garside 4630 (B, K); Strand, 8 viii 1942, Parker 3707 (BOL, NBG); Bredasdorp, The Poort, 0.5
miles E of the road BredasdorpElim, on rocks below limestone hill, 19 ix 1962, Nordenstam 1503 (S); Cape
Town, Swartklip, 6 ix 1972, Taylor 8169 (NBGSTE); Stillbaai, ridge below reservoir, southerly aspect, gentle
slope, limestone, 60 m, 24 viii 1979, Bohnen 6191 (NBGSTE); Malgas, De HoopPotberg Nat. Res., Dronkvlei,
along track to Koppie Alleen, 11 ix 1979, Burgers 2172 (holo NBGSTE); Malgas, De Hoop, 4 km from
MoerasfonteinturnoffonBredasdorpSkipskoproad,8viii1984,Fellingham732(NBGSTE);Malgas,DeHoop,
Buffelsfontein, 9 viii 1984, van Wyk 1773 (NBGSTE); Vredenburg, lower NW slopes of limestone koppie at
Paternoster,28vii1998,Goldblatt&Manning10928(MO,NBG);Malgas,DeHoopNat.Res.,plainsbelowthe
slopeatroadfromentranceofreservetotheinformationcenter,limestonepavement,28ix2008,Mucina&
Merukov 280908/04 (MJG); Malgas, De Hoop Nat. Res., 28 ix 2008, Merukov KM03/27 (NBG), 14 x 2008
KM27/24 (NBG), KM28/51 (NBG); Langebaan Peninsula, Jacobsbaai, 26 x 2008, Merukov KM44/44 (NBG),
KM45/41 (NBG), 30 x 2008, Merukov KM52/28 (NBG), KM53/33 (NBG); Langebaan Peninsula, Paternoster,
CapeColumbine,29x2008,MerukovKM50/34(NBG),KM51/44(NBG).
76
3D.subsp.quarcicolaMering,Kcke&Kadereitsubsp.nov.Type:SouthAfrica,Western
Cape,VanrhynsdorpDiv.,farmMoedverloren,KnersvlakteConservationArea,onquartz
fields,BIOTABiodiversityObservatoryNo28(Moedverloren),ha.78,3128'3.7"S,01826'
52.8"E,1ix2008,Schmiedel124811(holoMJG;isoNBG;Fig.10).Figs6D,7.
Differtasubspecietypicahabituminoreetgraciliore,foliisperangustis,pedicellisdivergenssub
angulocirca90;soloinsubstratoquarcitico.
Plants 520 cm high, delicate, slender, with small, fusiform bulbs, bulbs covered by dry, soft or stiff
brownfibres,these12cmlong.Plantswithoutrhizomes.Leavesuniform,aslongasorshorterthan
infructescence,verynarrow,about0.51mmwide.Inflorescenceslax,with415(25)flowers.Flowers
(1.5)2.03.0 mm long. Pedicels shorter than fruit at fruiting time, 25 mm long, held at angles of
almost90fromtheinflorescenceaxis,usuallynotcurvedinwardstowardstheapex.Infructescences
lax, axis sometimes flexuose. Infructescences 38 cm long. Fruits very narrowly ovoid (to almost
ellipsoid), 58 mm long, 0.51.0 mm wide (Fig. 6D). Mericarps linear, sometimes with short basal
spurs,connate;tipsnotcurvedoutwards.Carpophoreverythinorlacking.
Distribution. Triglochin bulbosa subsp. quarcicola is only found in the Western Cape Province of
SouthAfricaandappearstobeendemictotheKnersvlakte(amajorcentreofplantendemismofthe
country;vanWyk&Smith2001;Fig.7),aslightlyundulating,lowlyingbasininsouthernNamaqua
land (near the town of Vanrhynsdorp). Subspecies quarcicola is the only subspecies of Triglochin
bulbosa found outside the Fynbos Biome. The farm Moedverloren is known to harbour a high
numberofveryrestrictedlocalendemics(U.Schmiedel,pers.comm.;B.Nordenstam,pers.comm.).
Ecology and habitats. The subspecies is restricted to saline quartz fields (Fig. 1C), especially the dry
slopes of koppies, covered with a layer of white, angular quartz gravel (Schmiedel & Jrgens 1999).
The vegetation housing populations of subsp. quarcicola is a dwarf succulent shrubland, rich in
succulent shrubs and stone plants (Knersvlakte Quartz Vygieveld according to Mucina &
Rutherford,2006).TheclimateoftheKnersvlakteistypicalforwinterrainfallsemidesert,withmean
annual precipitation of around 120 mm and mean annual temperature of about 18 C (Mucina &
Rutherford,2006).
Phenology. Flowering and fruiting in early spring (viiviii). The subspecies seems to grow only
following sufficient rainfall (U. Schmiedel, pers. comm.). According to notes on one specimen (Le
Roux2205,STE),theleavesaredeciduous.
Notes. According to notes on two specimens, flowers are brownishpurple or purple (Nordenstam
587,S;LeRoux2205,STE).
Taxonomic remarks. Subspecies quarcicola differs from the other subspecies in having lax
inflorescences with 415(25) flowers and pedicels diverging at angles of mostly 90 from the
inflorescenceaxis.ThissubspeciesisrestrictedtoquartzfieldsoftheKnersvlakte.
ProposedIUCNconservationstatus.Vulnerable(VUD2).Triglochinbulbosasubsp.quarcicolashowsa
veryrestrictedareaofoccupancy(AOO,estimatedtobelessthan100km
2
)andiscurrentlyonly
77
Fig.7.HolotypeofTriglochinbulbosasubsp.quarcicolaMering,Kcke&Kadereitsubsp.nov.(MJG).
78
known from three locations (Fig. 7), where it was found only in small populations (B. Nordenstam,
pers. comm.; U. Schmiedel, pers. comm.; see alsoa remark on the label of one of the specimen (Le
Roux2205,STE).
The farm Moedverloren (or Moedverloor on some maps) is currently managed by CapeNature, the
provincialnatureconservationauthority,aspartoftheKnersvlakteProtectedArea(U.Schmiedel,
pers. comm.). Some of the populations are found, however, outside the borders of formally
protectedareas,andthesurvivalofthistaxondependsontheownersoftheseneighbouringfarms.
Further field studies are required to estimate more precisely the area of occupancy and extent of
occurrence.
Specimensexamined.SOUTHAFRICA.WesternCape.Knersvlakte,Koekenaap,15vii1962,Nordenstam587(S);
20 vii 1962, Hall 2397 (NBGSAM); Knersvlakte, Vanrhynsdorp, Moedverloor [Moedverloer in Hall 4124], 28
vii1962,Nordenstam790(S);10viii1962,Nordenstam898(S);23viii1971,Hall4124(NBG);5viii1981,Hall
5078 (NBG); Knersvlakte, Lutzville, 7 km N of Hol River railway station on the farm Klipdrift, 10 viii 1977, Le
Roux 2205 (NBGSTE); Knersvlakte, Vanrhynsdorp, farm Moedverloren, Knersvlakte Conservation Area (BIOTA
BiodiversityObservatoryNo28),quartzfields,1ix2008,Schmiedel124811(holoMJG,isoNBG).
4.TriglochincompactaAdamson,J.S.AfricanBot.9:152(1943).Type:SouthAfrica,
Karbonkelberg,1000ft,28ii1943,Compton14436(holoNBG!).Figs1EG,2C,11.
Plants (10)1550 cm high, with fusiform bulbs and without rhizomes; bulbs covered by dry, rather
soft, brown fibres, these up to 5 cm long. Leaves few, formed after flowering, mostly reaching only
halfoftheheightofplants.Outerleavesstrapshaped,shorterandwiderthaninnerleaves,upto4
cm long, mucronate, inner leaves abruptly narrowing above the ligule, less than 0.5 mm wide,
distinctly longer than outer leaves, leaves generally larger and more numerous after fruit maturity.
Inflorescencesdense,with(6)1050flowers.Flowers1.53.0mmlong.Pedicelselongatedatfruiting
time, up to 27 mm long, mostly curved inwards towards the apex. Infructescences 520 cm long.
Fruitsnarrowlyovoid,610mmlong,12mmwide(Fig.1G,2C). Mericarpsdorsallycurvedupwards,
connate at the carpophore; tips curved outwards. Carpophore often with three basal membranous
outgrowths.
Distribution.EndemictoSouthAfrica:restrictedmainlytotheWesternCapeProvincewhereitoccurs
intheCapePeninsulaasfareastasGoukammaontheGardenRouteandinSeweweekspoortinKlein
Swartberg Mountains,theCederbergandinahandfuloflocalitiesinthe NorthernCapeProvincein
the surroundings of Nieuwoudtville and near Kommagas (Namaqualand) (Fig. 11). The Kommagas
collection is an unsual outlier and requires further attention. Unfortunately newly dated material
fromthislocalitywasnotavailable.
Habitat and ecology. Triglochin compacta is invariably found on deep (mainly nutrientpoor) sands.
When of marine orgin, these sands form stabilised (and highly leached) inland sand dunes, always
outside the current direct influence of the sea. In some places at higher altitudes these deep sandy
sediments are derived in situ from arenites such as Table Mountain and Nardouw sandstones. The
vegetation types characteristically supporting this taxon are Hangklip Sand Fynbos, Cape Flats Sand
Fynbos, Albertinia Sand Fynbos, Knysna Sand Fynbos, Cederberg Sandstone Fynbos and Bokkveled
SandstoneFynbos(sensuMucina&Rutherford,2006).
79

Fig.8.GeographicaldistributionofTriglochincompactaAdamson,T.elongataBuchenau,andT.buchenaui
Kcke,Mering&Kadereitsp.nov.
80
Phenology.Floweringmostlyinautumnfrom(ii)iiiv,somefruitingmaterialwascollectedinvi.
Taxonomic remarks. Ever since its description by Adamson (1943), this species has been incorrectly
classified as being synonymous with Triglochin bulbosa. However, the two species can be easily
distinguished by phenology as well as morphology (see T. bulbosa for a comparison between these
twosympatricspecies).TheaffinitiesofTriglochincompactatoT.milneiwillbediscussedunderthe
latterspecies.
ProposedIUCNconservationstatus.DataDeficient(DD).Thelownumberofcollectionsofthistaxon
can be a result of poor sampling (failure to distinguish this species in the field from the much more
common Triglochin bulbosa subsp. bulbosa), habitat loss, natural rarity, or a combination of these
factors. Some of the sandveld areas (system of lowland stabilised dunefields) have been suffering
from pressure of urban sprawl through both the spread of formal housing development and the
explosivespreadofinformalsettlementsontheCapeFlats(nowpartoftheCapeTownmetropolis).
Theconservationstatusofthistaxoncanonlybejudgedonceanintensivesearchforextantlocalities
hasbeenconducted.WethereforesuggestclassifyingTriglochincompactaasDataDeficient.
Specimensexamined.SOUTHAFRICA.NorthernCape.Niewoudtville,sinedato,Leipoldt808(BOL);Niewoudt
ville,19ii1932,MaughanBrowns.n.(BOL);[Namaqualand],Komaggas("Bethel,Kammaggas"),24iii1952,van
Niekerk3833(BOL);Nieuwoudtville,FarmArendskraal,30iii1982,Snijman597(NBG).WesternCape.Capeof
GoodHope,annoca.1770,AugeorOldenburgs.n.(BM);iiioriv1771,Banks&Solanders.n.(BM);CapeTown,
Wynberg,23iii1892,Schlechter607(Z);Hermanus,Vogelgat,11iv1897,Schlechter10424(BM,PRE,Z);Cape
Town, Muizenberg, iii 1908, Dmmer 1050 (NBGSAM); Caledon, 21 ix 1916, Purcell 104 (NBGSAM); Cape
Town,Constantia,BergvlietFarm,Peter'sHill,26iii1917,Purcell12(NBGSAM);CapeTown,LadiesMileHill,12
iv1918,Purcell65(NBGSAM);Knysna,Goukamma,WestHill,iv1928,Fourcade3925(BOL,NBGSTE);Strand,
on sand dunes near the sea, 22 vi 1940, Parker 3515 (BOL, NBG); Cape Town, Slangkop, 9 v 1942, Bond 1498
(NBG);CapeTown,Karbonkelberg,25ii1943,Isaac10(BOL);28ii1943,Compton14436(holoNBG);4iii1943,
Leighton 934 (NBG); 30 iv 1944, Compton 15657 (NBG); 2 iv 1945, Levyns s.n. (Adamson 3615) (BOL); Cape
Peninsula, Buffels Bay, 27 v 1945, Compton 17061 (NBG); Cape Town, slopes above Rheboksdam Bay, 27 v
1945, Leighton 970 (BOL); Cederberg, Sandfontein Peak, 5 iv 1947, Esterhuysen 13870 (BOL); Cape Town,
hillside above Kommetjie, vi 1947, Lewis 2379 (NBGSAM); Clanwilliam, Krom River, 1 iv 1956, Esterhuysen
25489 (BOL); Cape Town, Kalk Bay Mountain, slope below Boyes Drive, 12 v 1974, Goldblatt 1792 (NBG);
Stillbaai, Panorama Circle, 10 iii 1979, Bohnen 5112 (NBG, PRE); Stellenbosch, BoOnderpapagaaiberg, 20 v
1988, Becker 0001 (NBG); Sedgefield, Buffelsbaai, Goukamma River estuary, camping site, 6 v 2006, Kcke &
Mucina 060506/02 (MJG); Stillbaai, Pauline Bohnens exgarden on Panoramasingel, 18 iii 2008, Naud et al.
s.n.(NBG).
5.TriglochinmilneiHorn,SvenskBot.Tidskr.55:85(1961).Type:ZAMBIA,MwinilungaDist.,mile
SofMatonchifarm,ca.1350m,30x1937,MilneRedhead3012(holoKphoto!,BOLphoto!;
isoSphoto!).Figs2D,12.
Plants(15)2060cmhigh,withbulbs,bulbscoveredbydry,rathersoft,brownfibres,theseupto5
cm long. Plants without rhizomes. Leaves few at flowering time, much shorter than, and mostly
reachingonlyhalfoftheheightofplants,becominglargerandnumerousafterfruitmaturity.Outer
leavesstrapshaped,shorterandwiderthaninnerleaves,upto4cmlong,withmucronatetips,inner
leaves abruptly narrowing above the ligule, >12 mm wide, distinctly longer than outer leaves.
81
Inflorescenceslax,with530flowers.Flowers23mmlong,usually>3.0mmlong.Pedicelselongated
atfruitmaturity,then315mmlong,divergingatanglesof45andcurvedinwardstowardstheapex.
Infructescences 614 cm long. Fruits narrowly ovoid to ovoid, (8)1014 mm long, 3 mm wide (Fig.
2D).Mericarpsconnateatthecarpophore.Carpophorewiththreebasalmembranousoutgrowths.
Distribution. Angola, Democratic Republic of Congo (DRC), Tanzania, Zambia, Zimbabwe; in South
Africa in KwaZuluNatal, southern Mpumalanga and Zuurberg, Griqualand East, Eastern Cape (Fig.
12). Usually found at altitudes above (500)1000 m, in Tanzania at up to 1900 m. According to
Govaerts(2008),TriglochinmilneiisalsofoundinMalawi,butnoherbariummaterialfromtherehas
beenseen.
Habitat and ecology. Triglochin milnei is found in seasonal (summerwet) wetlands and seasonally
wet, sometimes burnt, grasslands, often on grey or black clayey soils. At least some of these
grasslands of the mid and highaltitudes in Tanzania and Zimbabwe (and possibly also in Malawi)
would qualify as afromontane. At low altitudes in the DRC and Zambia this taxon is found in so
called dambos seasonally waterlogged, predominantly grasscovered depressions bordering
headwater drainage lines (Mackel, 1985). On the Mpumalanga Highveld and in the KwaZuluNatal
Midlands (South Africa), this taxon appears to be limited to azonal wetlands embedded within the
GrasslandBiome.
Phenology.Floweringandfruitingfromxii(iii).
Nomenclaturalnotes.Theholotype(MilneRedhead3012,K)doesnothavefruitsandshouldbeused
togetherwithfruitingmaterial.
Taxonomicremarks.TriglochinmilneiismorphologicallysimilartoT.compactabyhavingoutgrowths
atthebaseofthecarpophoreandnarrowlyovoidtoovoidfruits,inwhichtheoutlineoftheseedsis
oftenvisibleonthemericarpsurface.However,thefruitsdifferinwidth(3mmvs.12mm)andin
most cases in length ((8)1014 mm vs. 610 mm). Furthermore, the two species are geographically,
ecologicallyandphenologicallyclearlydifferentiated.
MostflorascoveringthedistributionareaofthespeciesdonotrecognizeTriglochinmilnei,buttreat
it as a synonym of T. bulbosa (e.g., Bennett, 1902; Obermeyer, 1966; Lisowski et al., 1982). Napper
(1971) accepted Triglochin milnei (p. 3: "The circumscription of T. bulbosa in F.S.A. 1: 93 (1966),
which includes T. milnei, is less satisfactory.") but doubted the sole record of this species in South
Africa. However, the study of extensive Triglochin material from South Africa and comparison with
material of T. milnei from other parts of Africa revealed that the species occurs in eastern South
Africa.Nevertheless,onlylimitedmaterialisavailableforSouthAfricaandmorecollectingisneeded.
Proposed IUCN conservation status. Data Deficient (DD). Triglochin milnei seems to be widely distri
butedinCentralandsoutheasternAfricaandisprobablynotunderimmediatethreat.However,little
is known about the current area of occupancy. The species is underrepresented in herbarium
collectionsandfewotherdataareavailable.IneconomicallymoredevelopedAfricancountries,such
asSouthAfrica,subtropicalandwarmtemperategrasslandsareunderpressurefromintensivestock
grazingassociatedwithfrequent(oftentoofrequent)burning.Sparseandvulnerablewetlandswithin
82

Fig.9.GeographicaldistributionofTriglochinmilneiHorn.
thesegrasslandsalsosufferfromoverutilizationbyhighgrazingstockconcentrations.Threatsand
currentconservationstatusatthenationallevelarelikelytodifferamongcountries,andtherefore
differentialassessmentsareneeded.InSouthAfricathespeciesisonlyknownfromsix,mostlyvery
old,collections.
Specimens examined. ANGOLA. Dist. Huilla, Alto plana, ad ripas Rivi de Lopollo, prope Ohai (Hy), 5000 feet,
xi/xii 1859, Welwitsch 3017 (para BM, P); Benguela [formerly Benguella], country of the Ganguellas and
Ambuellas, anno 1910, Gossweiler s.n. (para K). DEMOCRATIC REPUBLIC OF CONGO. 1.5 km E of Kabiasha,
Katanga,1020m,7xii1967,Malaisse6189(POZG);HautShaba,closetoLubumbashi(formerlyElisabethville),
8kmWofKasapaUniversity,25xii1968,Lisowski84458(POZG);Kinshasa,HautKatanga,closetoLubumbashi,
on the fringes of Natwebo, 1220 m, 22 x 1969, Lisowski 731 & 732 (POZG), 27 xi 1970, Lisowski 733 (POZG);
PlateaudeKundelungu,Katanga,bankofKalundariver,1590m,9i1971,Lisowski,Malaisse&Symoens12645
&12698(POZG);PlateaudesKundelungu,3kmNofsourceofriverLutshipuka,9i1971,Lisowski,Malaisse&
Symoens12644(BOL).TANZANIA.Uyansi,LakeChaya,ca.1240m,4i1926,Peter45784(B);Sumbawanga,6200
feet,30i1950,Bullock2364(paraK,Sphoto!);12kmEofSongeabyNonganongastream,1050m,28xii1955,
MilneRedhead&Taylor7934(paraB,EA,P,Sphoto!);IringaDist.,4milesNofIringa,GreatNorthRoad,5150
feet, 5 ii 1962, Polhill & Paulo 1362 (B, EA, P); 7 miles from Iringa on Dabaga road, 26 xii 1965, Harris 10256
83
(EA); Sumbawanga, Tatanda Mission, 1700 m, Bidgood, Mbago & Vollesen 2409 (K); Nkansi Distr., 5 km on
NamanyereKarongaroad,1500m,4iii1994,Bidgood,Mbago&Vollesen2603(P).SOUTHAFRICA.KwaZulu
Natal.Inanda,i1800,MedleyWood997(BM);GriqualandEast,ZuurbergMountains,x1883,Tyson1866(BOL);
Alexandra Dist., Station Dumisa, Farm Friedenau, Umgaye, 600 m, 15 x 1908, Rudatis 440 (para BM, P);
Krantzkop ("Kranskop"), 4500 feet, xii 1910, Thode 3870 (NBGSTE); Krantzkloof, Gillitto Kloof, 1800 feet, x
1921, Haygarth s.n. (STE 260, 261). Mpumalanga. Wakkerstroom Dist., N of Dirkiesdorp, SE of main road, on
thefarmRoodekraal,1340m,13xii1995,Balkwill9389(PRE).ZAMBIA.MwinilungaDist.,mileSofMatonchi
farm,16xii1937,MilneRedhead3693(paraSphoto!);5milesEofLusaka,4200feet,10i1958,King403(para
K); Chadiza, 850 m, 25 xi 1958, Robson 686 (BM); Lusaka, 4 xii 1964, Robinson 6266 (M); Kitwe, 9 ii 1969,
Fanshawe10515(K).ZIMBABWE.CharterDist.,Charter,4000feet,27xii1926,Eyles4587(paraK).
6.TriglochinelongataBuchenau,inA.Engler,DasPflanzenreich4.14:10,fig.3(1903).
Type:SouthAfrica,Div.Malmesbury,UmgegendvonHopefield,Bachmann1662,1693(syn
B);SouthAfrica,Pondoland,Bachmann309(synB);SouthAfrica,Div.Malmesbury,
UmgegendvonHopefield,WegnachCoeuratenberg,xi1886,Bachmann1693(lecto.,
designatedhere,Z!;Fig.13);drawingoffruitandhabit(fromBachmann1693,B)in
Buchenau(1903),fig.3,p.11(epi,designatedhere;Fig.14).Figs1H,1L,2E,11.
Plants1590cmhigh,withrhizomes(Fig.1H).Rhizomeswhitishtobeigewithbracts.Leavesuniform,
thickened at the base but not forming a bulb, as long as or longer than plant height, 12 mm wide.
Leaf bases covered by soft, whitish to brown fibres, fibres mostly 36 cm, rarely up to 15 cm long.
Inflorescences dense, with 10100 flowers. Flowers 1.52.5 mm long. Pedicels elongated at fruiting
time, then 26 mm long. Infructescences (6)1025 cm long. Fruits narrowly ovoid to ovoid, often
secund, 59 mm long and 1.02.5 mm wide (Fig. 2E). Mericarps connate at the carpophore; tips
mostlycurvedoutwards.
Distribution. Endemic to South Africa: distributed mainly in coastal regions of the KwaZuluNatal,
Eastern,WesternandNorthernCapeProvinces(Fig.11).InthelatterProvincetherearetwooutlying
localities always located near the mouths of intermittent rivers draining the coastal plains of
Namaqualand. We have failed to detect any morphological uniqueness of these Namaqualand
populations. They are exceptional in terms of their geographical distribution and deserve further
study. Many isolated inland localities are presumably of relictual character as they are found
abundantly in regions frequently flooded by the sea during the altitermal marine transgressions
(Overberg,partsofWestCoast).
Habitatandecology.Triglochinelongatacommonlygrowsintheuppertidalandsupratidalzonesof
estuarine salt marshes (Fig. 1L), usually on heavy clayey (partly also clayeysandy) saline soils. In
coastal habitats it is found on elevated banks of sandy beaches, edges of estuarine rivers as well as
on rocks exposed to salt spray. Along some estuarine rivers it penetrates deeply inland (Bushmans
River in the Eastern Cape), and in some regions (West Coast, Breede River, Overberg) it is found in
(obviously relictual) inland localities, on banks of intermitted rivers (in Namaqualand) and on the
edges of saline and brackish temporary pans, here often intermingled with dense Juncus stands.
According to Mucina & Rutherford (2006) these habitats belong to the following vegetation units:
Arid Estuarine Salt Marshes, Cape Estuarine Salt Marshes, Cape Inland Saltpans, Cape Lowland
FreshwaterWetlandsandCapeSeashoreVegetation.
Phenology.Floweringthroughouttheyearaslongaswaterisavailable.
84
Nomenclatural notes. In the protologue of Triglochin elongata, Buchenau (1903) cites three
specimens which probably were destroyed in B during World War II. A duplicate of one of these
syntypes was discovered (Bachmann 1693, Z; Fig. 13) and is designated as lectotype here. This
specimen, however, has only flowers but no fruits. The fruit (and habit) of one of the original
syntypes (Bachmann 1693, B) is illustrated in Buchenau (1903, fig. 3, p. 11; Fig. 14). This drawing
shouldbeusedasanepitype(designatedhere)inconjunctionwiththelectotype.
Taxonomic remarks. Triglochin elongata is similar to T. buchenaui in having rhizomes instead of

bulbs,butdiffersinhavingdenseinflorescenceswith10100flowers(vs.laxinflorescenceswith317
flowers), (6)1025 cm long infructescences (vs. 27 cm long), and fruits with a carpophore. Unlike T.
buchenaui, which is restricted to the lower tidal zone of salt marsh estuaries, T. elongata is
ecologicallymorevariableandgrowsintheupperzoneofsaltmarshestuariesandestuarinerivers.
Variation. Several exceptionally large individuals (up to 90 cm) with long infructescences and
numerousfruitshavebeenseen.
Proposed IUCN conservation status. Least Concern (LC). Triglochin elongata is a widespread species
recorded from many riversides near the sea and almost all major estuaries in South Africa (Fig. 11).
However, some of these estuaries are under serious pressure by silt deposition or reduced
freshwaterflow(Turpieetal.,2002;Turpie,2004).
Specimens examined. SOUTH AFRICA. Eastern Cape. Uitenhage Div. [probably Swartkops Estuary near Port
Elizabeth], xii sine anno, Ecklon & Zeyher 620 (BOL); Port Elizabeth, Zwartkopsrivier, sine dato, Zeyher 4326
(NBGSAM); East London, Kwenxura (= Kwenqura) River mouth, 23 xii 1900, Galpin 5815 (NBG); [Transkei],
Kentani, coast, 100 feet, anno 1903, Pegler 335 (BOL); [Transkei], Kentani, 4 xii 1905, Pegler 335 (SAM, Z);
[Transkei],KeiRiverBridge,19x1931,Taylor3669(NBG);Tsitsikamma,NaturesValley,GrootRiverlagoon,ii
1932,J.&B.Rennie545(BOL);[Transkei],nearKeiRivermouth,25xi1945,Compton17675(NBG);[Transkei],
Coffee Bay, iii 1947, Lewis 2374 (NBGSAM); East London, 27 x 1962, Batten 1.Pl.115. (NBG); East London,
NahoonRiver,26vi1963,Tlken402(NBGSTE);PortAlfred,KowieRiverestuary,28i1964,Mauve&Wells7
(NBG); East London, Bridle Drift, 2 iv 1966, Pamphlett 5 (NBG); Humansdorp, on water edge, 21 xi 1972,
Montgomery 182 (NBGSTE); Tsitsikamma, Natures Valley, Groot River estuary, 6 v 2006, Kcke & Mucina
060506/29 (MJG, NBG); Jefferey's Bay, Kabeljous River estuary, 7 v 2006, Kcke & Steffen 070506/01 (MJG);
Port Alfred, Kowie River estuary, 9 v 2006, Kcke & Steffen 090506/22 (MJG). KwaZuluNatal. [Herbarium
Natal, locality burnt, possibly Durban], xi 1882, Medley Wood 396 (BM); Durban, shores at bay, 20 xi 1884,
Medley Wood 396 (BOL); Durban, xi sine anno, Medley Wood 925 (BM); Durban, Congella, sine dato, Medley
Wood11985(BOL);Durban,x1888,Wilms2278(BM);Durban,Levy,iii1894,Kuntzes.n.(NY);Durban,Beach
Terminus,vi1913,Thode4901(NBGSTE);Durban,Isipingo,22iv1921,ForbesSTE12526(NBGSTE);1xi1926,
Schrders.n.(ZT);Durban,Fusel,ix1933,Meebold12994(M);Durban,IsipingoBeach,13viii1949,Ward917
(K);PortEdward,PalmBeach,21iii1992,Weigend2335(M);PortEdward,UmtamvunaRivermouth,9iv2006,
Pienaar, Jakubowsky & Swelankomo 090406/1 (MJG, NBG). Northern Cape. Namaqualand, Hondeklipbaai,
Spoeg River mouth, 5 m, 18 x 1980, Le Roux & Parsons 59 (NBGSTE); Namaqualand, Kleinzee, 5 ix 2005,
Mucina 050905/04 (MJG). Western Cape. Cape Town, in humidasis in littore maris False Bay prope
Muizenberg, iii 1892, Schlechter 7184 (BOL); Bonnievalle, swamp near Breede River, iii sine anno, Marloth
11592(NBGSTE);CapePeninsula,MuizenbergVley,11iii1896,WolleyDod969(BOL);14xi1897,WolleyDod
3658 (BOL); Cape Town, Paarden Island, ii 1908, Dmmer 1171 (NBGSAM); Cape Town, Lakeside, i 1918,
Michell s.n. (BOL); West Coast, Graafwater, ix 1923, Adamson 39042 (NBGSAM); Cape Peninsula, E of
85
Paulsberg, 3 v 1929, Salter 344/9 (K); Robertson, Breede River, 25 ix 1935, Lewis s.n. (BOL); Cape Town,
PaardenIsland,24iii1936,Adamson843(BOL);CapeTown,LakesideVlei,17x1936,Adamson1235(BOL);17
xi 1936, Adamson 1370 (BOL); 17 i 1937, Adamson 1614 (BOL); Cape Town, Sand Vlei, 17 xi 1936, Adamson
1373(K);CapeTown,SandVlei,Eside,17xi1936,Adamson1370(NBGSAM);CapeTown,Muizenberg,Sand
Vlei,20iii1938,Adamson1689(BM);WestCoast,betweenBokbaaiandDarling,15ix1940,Esterhuysen3866
(BOL); Caledon, at streamside, xi 1940, Esterhuysen 3865 (BOL); Hermanus, Onrust, beside the lagoon, 22 iii
1944, Leighton 413a (BOL), Leighton 413 (NBG); Gansbaai, Frikkiesbaai [Uilkraalsmond], 21 iv 1946, Leighton
1669 (BOL); Gansbaai, Franskraal, 21 iv 1946, Leighton 1669 (NY); George, Kaaimansgat, Kaaimans River, xi
1947, Wilman s.n. (BOL, K); Plettenberg Bay, 24 ix 1967, Thompson 588 (NBGSTE); Cape Town, Cape Flats,
Isoetesvlei,9iii1970,Strauss10(NBG);PearlyBeach,KleinHagelkraal,50feet,10iii1979,Thompson3898(M,
NBG);Stillbaai,KransfonteinFarm,humusrichriverbankinforest,7x1980,Bohnen7731(NBGSTE);Gansbaai,
Franskraal, N bank of Uilkraal River, 2 m, 3 xi 1987, OCallaghan 3/11/2 (NBGSTE); Struisbaai, at entrance to
CoronaFarm,28xii2003,Mucina281203/05(MJG);CapeTown,Zandvlei,nearstationandboatclubarea,20
viii 2004, Walton 341 (MJG); Gansbaai, Franskraal, Uilkraalsmond Estuary, N of the road bridge, 8 v 2005,
Mucina 080505/05 (MJG); Malgas, De Hoop Nat. Res., Potberg section, Cupidos Kraal, 3 iv 2006, Kcke &
Mucina030406/14(MJG);Struisbaai,VogelzangFarm,4iv2006,Kcke&Mucina040406/22,040406/24(MJG,
NBG);Gansbaai,Franskraal,UilkraalsmondEstuary,Noftheroadbridge,5iv2006,Kcke&Mucina050406/03
(MJG), 050406/07 (MJG), 050406/09 (MJG); Hermanus, Fisherhaven, near jetty at the Botrivier Lagoon, 5 iv
2006, Kcke & Mucina 050406/11 (MJG, NBG); Bredasdorp, Patryskraal on road between Oupos and
Bredasdorp,5iv2006,Kcke&Mucina050406/18(MJG,NBG);Gansbaai,Franskraal,Uilkraalsmondestuary,5
iv2006,Kcke&Mucina050406/19(MJG,NBG);Struisbaai,WofZoetendalsvlei,atroadStruisbaaitoElim,km
59.9, 5 iv 2006, Kcke & Mucina 050406/24 (MJG, NBG); Struisbaai, at entrance to Corona Farm, 5 iv 2006,
Kcke&Mucina050406/27(MJG);CapeTown,Milnerton,RietvleiNat.Res.,closetoparkingarea,20iv2006,
Kcke&Steffen200406/01,200406/03(MJG);onthewaytobirdhide,20iv2006,Kcke&Steffen200406/10
(MJG); at bird hide, 20 iv 2006, Kcke & Steffen 200406/17 (MJG); Hermanus, Voklip, 23 iv 2006, Kcke &
Mucina 230406/18 (MJG); Mosselbaai, Hartenbos, near the bridge over river, 4 v 2006, Kcke & Mucina
040506/24 (MJG, NBG); [Cape Town], Faure, Vergenoeg Farm, 4 x 2006, Boucher 7412 (NBG); Elim, near
homestead of Heuningrug Farm, 14 x 2006, Mucina 141006/23 (MJG); Malgas, De Hoop Nature Reserve,
Potberg section, near Bultfontein, 21 x 2006, Mucina 211006/17 (MJG); Struisbaai, Vogelzang Farm, N of
HeuningnesRiverestuary,17iii2007,Mucina170307/03(MJG),170307/04(MJG);Struisbaai,DeMondNature
Reserve,HeuningnesRiverestuary,leftbank,17iii2007,Mucina170307/13A(MJG).
7.TriglochinbuchenauiKcke,Mering&Kadereit,sp.nov.Type:SouthAfrica,
WesternCape,WestCoastNationalPark,EofSeebergBirdHide,intertidalsaltmarshflats,
14iv2006,Kcke&Mucina140406/06(holoMJG;isoNBG;Fig.15).Figs1IK,2F,11.
Herbascaposa,rhizomatosa,nonbulbosa,ad30cmalta.DiffertaTriglochinelongatainflorescentia
laxacum317flores,fructibusangustusovoideususquefereglobosus,carpophorisdeficiens.
Plants930cmhigh,withrhizomes.Rhizomeswhitishtobeigewithbracts.Leavesuniform,thickened
at base but not forming a bulb, as long as or longer than infructescence, leaf bases covered by
usually(very)few,soft,whitishtobeigefibreswhichcanbemorethanhalfaslongasheightofplant.
Inflorescenceslax,with317flowers.Flowers2mmlong.Pedicelsupto2mmlongatfruitingtime,
usually shorter. Infructescences 27 cm long, fruits distant. Fruits narrowly ovoid to ovoid, 48 mm
longand1to>2mmwide(Fig.2F).Mericarpsdorsallycurvedupwards,filledwithair,tipsnotcurved
orweaklycurvedoutwards,carpophoreabsent.
86
Fig.10.HolotypeofTriglochinbuchenauiKcke,Mering&Kadereitsp.nov.(MJG).
87
Distribution. Endemic to South Africa, restricted to major estuaries of the West and South Coast,
spanning Groot Berg River estuary near Velddrif (Western Cape Province) and Kariega estuary near
KentononSea(EasternCapeProvince)(Fig.11).
Habitatandecology.Triglochinbuchenauiisatypicalspeciesofperiodicallyfloodedlowersaltmarsh
estuaries(Fig.1I).Thesoilsinthesehabitatsaresandytosandyclayey.Triglochinbuchenauiusually
occurs together with Sarcocornia tegetaria, T. striata, Cotula coronopifolia, Poecilolepis ficoidea,
Chenolea diffusa and various Limonium species. The vegetation in these habitats is classified
exclusivelyasCapeEstuarineSaltMarshes(Mucina&Rutherford,2006).
Phenology.Floweringthroughouttheyear.
Taxonomicremarks.SeeTriglochinelongataforacomparisonofthesetworhizomatousspecies.
Etymology.ThespeciesisnamedafterFranzGeorgPhilippBuchenau(18311906),aGermanbotanist
whocontributedsignificantlytotheknowledgeofTriglochinandJuncaginaceae.
ProposedIUCNconservationstatus.LeastConcern(LC).Triglochinbuchenauiisawidespreadspecies
recordedfromanumberoflargelagoonsandestuariesinSouthAfrica(Fig.11)andthereforeisnot
immediately threatened. The Berg River and Swartkops River estuaries support large populations of
Triglochin buchenaui and are also among the highest ranked South African estuaries in terms of
botanical and conservation importance (Turpie et al., 2002). The estuaries of the Langebaan and
KnysnaLagoonsaswellasestuariesofKrommeandSwartkopsenjoyformalprotectioneitheraspart
ofnationalparksorlocalauthoritynaturereserves.
Specimens examined. SOUTH AFRICA. Eastern Cape. Port Elizabeth, Redhouse, i 1915, Paterson 288 (BOL);
Boesmansriviermond, Bushmans River mouth, 14 i 1936, Dyer 3365 (PRE); Port Elizabeth, Swartkops River
mouth, 4 ii 1964, Mauve & Wells 38 (NBGSTE); Tsitsikamma, Blaubaai by Robbehoek, 8 iii 1979, Taylor 9954
(NBGSTE); KentononSea, Ghio Wetland, below the bridge over Bushmans River, 26 xi 2005, Mucina
261105/05 (MJG); St. Francis Bay, KrommeRiver estuary,at the bridge, 7 v 2006, Kcke & Steffen 070506/07
(MJG); Port Elizabeth, Swartkops River estuary, 8 v 2006, Kcke & Steffen 080506/06 (MJG); KentononSea,
GhioWetland,atbridgeoverBushmansRiver,9v2006,Kcke&Steffen090506/04(MJG);9v2006,Kcke&
Steffen 090506/08 (MJG); KentononSea, Kariega River estuary, 9 v 2006, Kcke & Steffen 090506/13 (MJG).
WesternCape.Knysna,Woodbourne,edgeoflagoon,31i1924,Duthie876(NBGSTE);Knysna,Belvidere,edge
oflagoon,22i1925,Duthie876(BOL);Langebaan,wetplaces,ix1925,Leipoldt27103(BOL);Knysna,Knysna
Lagoon,6ii1964,Mauve36(PRE),Wells37(PRE);LangebaanLagoon,nearOesterwal,4v1967,Simons1991
(BOL); Knysna, Natures Valley, Groot (Wes) River mouth, 8 iv 1981, Parsons 161 (NBGSTE); Velddrif, Port
Owen,BergRiver,13ii1981,LeRoux2855(NBGSTE);16ii1986,Boucher5123(NBGSTE);22ii1988,Becker
0002(NBGSTE);Velddrif,saltmarshoppositeBergRiverMouth,14x1986,OCallaghan157(NBG);Velddrif,N
bank of Berg River, Wreck, 4 ii 1987, OCallaghan 1468 (NBGSTE); Velddrif, Berg River, W of Port Owen, 8 ix
1987, OCallaghan 8/13 (NBGSTE); Langebaan, West Coast Nat. Park, Churchhaven, 21 viii 1997, Mucina
6435/11 (PRE); Langebaan, West Coast National Park, Geelbek at Bird Hide, 14 iv 2006, Kcke & Mucina
140406/27(MJG);Velddrif,BergRiverestuary,atCarinusBridge,19iv2006,Kcke&Steffen190406/15(MJG,
NBG), 190406/19 (MJG); Stillbaai, Goekoe River estuary, E bank of the river, 4 v 2006, Kcke & Mucina
040506/10 (MJG); Knysna, Knysna Lagoon, turnoff to Leisure Island, 5 v 2006, Kcke & Mucina 050506/05
(MJG);PlettenbergBay,BitouRiverestuary,6v2006,Kcke&Steffen060506/25(MJG),060506/27(MJG).
88
DISCUSSION
The seven species of the Triglochin bulbosa complex recognized here are not only
morphologically distinct but also differentiated in terms of geographical distribution, ecology,
phenologyand,whereknown,chromosomenumber.
IntheMediterraneanregion,TriglochinlaxifloraisautumnfloweringwhereasT.barrelieriis
springflowering. The two Mediterranean species also have different ploidy levels, with Triglochin
laxiflorahaving2n=18andT.barrelierihaving2n=30,32or36chromosomes.InAfrica,Triglochin
milnei is the only species of the complex found in several countries of Central and southern Africa.
The distribution of Triglochin milnei does not overlap with the other four species in South Africa. In
easternSouthAfricaitistheonlyspeciestobefoundininlandlocalitiesandataltitudesabove500
metres. Both rhizomatous species, Triglochin elongata and T. buchenaui, grow sympatrically (in a
broad sense) along the South and West coasts of South Africa. However, they differ in ecology by
Triglochinbuchenauibeinglimitedtolowertidalhabitats,whileT.elongataisfoundinuppertidalor
semiterrestrial inland habitats such as edges of water courses and pans. Where the two bulbous
species,TriglochinbulbosaandT.compactaoccurinthesamearea(e.g.,CapePeninsula),theydiffer
both in phenology and ecology. Triglochin bulbosa flowers between July and November, while the
flowering time of T. compacta is between February/March and May. Triglochin bulbosa subsp.
bulbosa occurs on Cape Peninsula on Archaean granites, while T. compacta is found on quartzite
sandsderivedfromOrdovicianTableMountainsandstone.
The importance of edaphic factors (geology, soils and associated pedohydrological
conditions) in diversification of the Cape flora has been widely suspected (Marloth, 1908) and in
manyinstancesalsowelldocumented(Rourke,1972;Goldblatt,1979;Williams,1982;Linder&Ellis,
1990; Kurzweil et al., 1991; see also Linder, 2003 and van der Niet et al., 2006). Triglochin bulbosa
may be a good candidate for the study of adaptive radiation of a noninsect pollinated taxon. The
four subspecies now distinguished within Triglochin bulbosa are clearly differentiated along pedo
hydrological habitat axes. The only waterbound subspecies of Triglochin bulbosa is the widely
distributed subsp. bulbosa occurring in temporarily wet (flooded) both coastal and inland habitats,
while the other three subspecies radiated into fully terrestrial inland habitats characterised by
contrasting geologies, such as granite (subsp. tenuifolia), limestone (subsp. calcicola) and quartz
fields(subsp.quarcicola).
Morphologyofundergroundpartsappearstobecorrelatedwithwateravailability.Whereas
thespecieswithrhizomes(Triglochinelongata,T.buchenaui)arelimitedtohabitatswithconstantly
available water, the bulbous species (Triglochin bulbosa, T. compacta, T. milnei, T. barrelieri, T.
laxiflora) grow in places where at least abundant water is available only for a short period of the
year.
ThesimilaritiesindifferentiationbetweenTriglochinspeciesfromtheMediterraneanFloristic
RegionandtheCapeFloristicRegion,respectively,arestriking.Thus,morphologicallysimilarspecies
from the two regions are flowering either in spring (Mediterranean: Triglochin barrelieri, South
Africa: T. bulbosa) or in autumn (Mediterranean: T. laxiflora, South Africa: T. compacta). This
divergence of flowering time was already noted by Buchenau (1896) and Rainha (1944) for the
MediterraneanregionandbyAdamson(1943)forSouthAfrica.Thespeciesofthecomplexthusarea
nice example for parallel ecological differentiation under similar climatic conditions in
mediterraneantypeecosystems.
SeveralspeciesoftheTriglochinbulbosacomplexareofconservationconcern.Eventhough
only Triglochin bulbosa subspp. calcicola and quarcicola are threatened according to the IUCN Red
89
List categories (IUCN, 2001), the habitats of many taxa (e.g., seasonal pools, estuaries) are under
manifoldpressurefromhumanactivities.

90

Chapter4
TetronciumanditsonlyspeciesT.magellanicum(Juncaginaceae):
distribution,ecologyandlectotypification
SabinevonMering
Publishedin:Willdenowia43(1):1324.
ABSTRACT
Tetronciummagellanicum(Juncaginaceae)wasdescribedbyWilldenowin1808,basedonmaterial
collectedbyCommersonattheStraitofMagellanduringBougainvillesvoyagearoundtheworld.
Typematerialofthisspecieswastracedandalectotypeforthenameisdesignated.Adescriptionof
thespeciesandnotesonitsecologyandconservationstatusareprovided.Forthefirsttime,a
detailedmapshowingtheknowndistributionareaofT.magellanicumispresented.
Additionalkeywords:GoughIsland,HerbariumWilldenow,lectotype,peatlands,southernSouth
America,typification

Chapter4:TetronciumanditsonlyspeciesT.magellanicum
92
93
Introduction
ThemonotypicgenusTetronciumWilld.belongstothesmallmonocotfamilyJuncaginaceae
(Alismatales) and is sister to the other two genera of the family, Cycnogeton R.Br. and Triglochin L.
(von Mering & Kadereit 2010). Tetroncium magellanicum Willd. is the only dioecious species in the
family(Fig.1B,6).ItisaperennialherbgrowingmainlyinthepeatlandsofsouthernSouthAmerica
andsomeneighbouringislands(Fig.1A,4).
Tetroncium magellanicum was described by Carl Ludwig Willdenow in 1808. In the
protologue, Willdenow stated that the new genus and species are based on material collected by
Commerson at the Strait of Magellan (Willdenow 1808). Philibert Commerson (sometimes spelled
Commeron) was a French naturalist who accompanied Louis Antoine de Bougainville during his
voyage around the world from 1766 to 1769 (Bougainville 1772). The botanical collections of this
voyagereachedParisin1774,andduplicatesweresubsequentlydistributedtoseveralbotanistsand
botanicalinstitutionsinEurope(Stafleu&Cowan19761988).
Tetroncium magellanicum was mentioned in several publications on the early botanical
explorationofsouthernSouthAmerica(e.g.,Hooker1847,Gay1849,Alboff1896,Reiche1907)and
also in later works covering the flora of the region (e.g., Correa 1969, Moore 1968, 1983,
Marticorena&Quezada1985,Zuloagaetal.2008).FranzBuchenau,whocontributedsignificantlyto
ourknowledgeofJuncaginaceae,alsoaddedtoourknowledgeofTetroncium(e.g.,Buchenau1868,
1882, 1903). However, the species has never been studied in full detail and Tomlinson (1982)
correctly stated that little is known about it. For example, embryological and karyological data are
lacking, and details of flower and fruit development as well as the mode of fruit dispersal are only
incompletelyknown.
In the course of taxonomic studies in Juncaginaceae I noted that the name Tetroncium
magellanicum has not been typified. Apart from typifying the species, I will also provide a
description, information on its ecology, an assessment of its conservation status (according to the
IUCN threat categories, IUCN 2001), and a distribution map based on data obtained from revised
herbariumspecimensandreliableliteraturesources.
Fig.1.Tetronciummagellanicum.A.Typicalhabitat,continentalbogdominatedbySphagnummagellanicum,
nearUshuaia,Argentina.B.Female(centre)andtwomaleinflorescences(photographs:R.Douzet,SAJF,
UniversityGrenoble).
94
Materialandmethods
Collections from Herbarium Willdenow (housed at the Botanic Garden and Botanical
Museum BerlinDahlem, Freie Universitt Berlin as separate collection BW; e.g., Hiepko 1972) and
fromseveralothermajorherbaria(seeAcknowledgementsforafulllist)wererevisedtolocatetype
materialofthespecies.Additionally,onlinedatabasesofherbariaorvirtualherbariawerereviewed,
as well as the GBIF Data Portal (Global Biodiversity Information Facility 2007+), JSTOR Plant Science
(2010+), Flora del Conosur (2009+), Flora Argentina (2012+), British Antarctic Survey (BAS) Higher
PlantsHerbarium(2009+),andUKOverseasTerritories(UKOTs)OnlineHerbarium(2011).
Altogether,morethan150specimensofTetronciummagellanicumwerestudied.Specimens
thatwereonlyseenasascanorphotographaremarkedwithimage!inthelistofspecimensseen
below. A lectotype for the name of this species is designated here according to the International
CodeofBotanicalNomenclature(McNeill&al.2006;Fig.3).
Information given on labels of historic specimens is usually limited, important specifications
such as (exact) collection dates or localities are often missing. Primary literature such as original
traveloguesofexpeditionsandsecondaryliteraturewasconsultedtodecipher,verifyandwherever
possiblecomplementavailablelabelinformation.Godleysaccountsofbotanicalexplorationofthe
Southernhemisphere(Godley1965,1970)wereespeciallyvaluableinthisrespect.Furthermore,the
followingonlineresourceswereused:TaxonomicLiteratureII(TL2,Stafleu&Cowan19761988and
supplements),JSTORPlantScienceCollectionofPlantCollectors(JSTORPlantScience2010+)andthe
HUHIndexofBotanists(HarvardUniversityHerbaria2011+).
The distribution map was created using georeferenced localities obtained from revised
herbarium specimens and reliable literature sources. Georeferencing of localities was facilitated by
the use of electronic gazetteers such as GEOnet Names Server (GNS 1994+), GeoNames (2012) or
GoogleEarth(Google2012),andthemapwasgeneratedusingtheEDITmapViewer(2012).
Nameandtypification
TetronciumWilld.,Mag.NeuestenEntdeck.GesammtenNaturk.Ges.Naturf.FreundeBerlin2:17.
1808.
=CathanthesRich.,Mm.Mus.Hist.Nat.1:365.1815.[Type:nondesignatus]
Notes. The material of Tetroncium that Willdenow had on hand at the time of writing the
descriptionwasnamedTriglochinreflexumandsenttohimbythelateProfessorMartinVahlfrom
Copenhagen (Willdenow 1808; Fig. 2). Willdenow noticed its distinctness from Triglochin and
describedthenewgenus(Willdenow1808).
Little is known about the identity of Cathanthes Rich., a name established only seven years
after Tetroncium by Richard (1815). Richards brief description of Cathanthes in his listing of genera
of Juncaginaeae reads as follows: [Lilaea.] Cathanthes, floribus dioici, reclinatis, etc., a sequente
diversa.[Triglochin.Scheuchzeria.](Richard1815).Kunth(1841)andHooker(1843)citeCathanthes
as a synonym of Tetroncium. This statement seems well supported by the few characters noted in
theprotologueasTetronciumistheonlydioecioustaxoninthegroup,andpossesesreflexedfemale
flowersandfruits.NoCathanthesspeciesweredescribedlaterandnospecimensbearingthisname
were found. Buchenau (1868, 1903) and Micheli (1881) mention Catanthes as a synonym of
Tetroncium.Thisismostprobablytheresultofaspellingerror.
95
Fig.2.ProtologueofTetronciummagellanicumWilld.(Willdenow1808;frommicrofiche,libraryoftheBGBM).
TetronciummagellanicumWilld.,Mag.NeuestenEntdeck.GesammtenNaturk.Ges.Naturf.Freunde
Berlin2:17.1808.
Lectotype(designatedhere):HabitatadFretumMagellanicum[StraitofMagellan],[December
1767],Commersons.n.(BW17531020!;Fig.3C);isolectotypes(designatedhere):BW1753101
0!,BM000820840!,C!,G00098732[image!],P01759004!,P01759005!).
TriglochinreflexaVahlexKunth,Enum.Pl.3:142(1841),magellanicum,nom.inval.(ICNArt.
36.1(c)).
TriglochinmagellanicaVahlexKunth,Enum.Pl.3:142(1841),reflexum,nom.inval.(ICNArt.
36.1(c)).
Notes. Material of T. magellanicum located in the Herbarium Willdenow comprises two sheets
placedinoneofthetypicalbluefoldersofthishistoriccollection.TheyarelabelledT.magellanicum
1 and T. magellanicum 2 (B W 1753101 0 and B W 1753102 0). This material clearly matches
thedetailsoftheprotologue,thelabelattachedtothefoldergivesthelocalityascitedaboveandthe
designation Triglochin reflexum written in Willdenows hand (Fig. 3A). The name Tetroncium
magellanicumandtheliteraturereferencewrittenonthislabelonthefolderwaslateraddedbyD.
F. K. Schlechtendal, as explained by the note Schlechtendal p., i.e. Schlechtendal pater (Latin for
father),writtenbyhissonD.F.L.Schlechtendal.Anothersmallerlabelonthefolder,readingVahl.
W., was also written by D. F. L. Schlechtendal and is indicating that the material originated from
VahlandWilldenow(Fig.3B).
ItseemscertainthatbothspecimenswereinthehandsofWilldenow.Therefore,alectotype
hastobedesignated.ThefirstofthetwosheetsinHerbariumWilldenowconsistsofafemaleanda
maleplant.Incontrast,onlyamaleplantismountedonthesecondsheet.However,thelatterbears
anadditionallabelreadingTriglochinmagellanicumwrittenbyWilldenowhimself(P.Hiepko,pers.
comm.;Fig.3C)anditisthereforechosenhereaslectotype(BW17531020).Theotherspecimenis
designatedasisolectotype.
ThelectotypeandisolectotypefromBareavailableonlineashighresolutionimagesthrough
the Digital Herbarium (Rpert 2000+). Duplicates of Commersons collection of T. magellanicum
(representingisolectotypes)werefoundinBM,C,GandP.Mostoftheseisolectotypes(andfurther
96
Fig.3.TetronciummagellanicuminHerbariumWilldenowA:folderwithlabel;B:additionallabel;C:specimen
designatedaslectotype(barcodeBW17531020/ImageId:267191),seetextfordetails.Photographs:
Herbarium,BotanicGardenandBotanicalMuseumBerlinDahlem(Rpert2000+).[Publishedat
http://data.bgbm.org/herbarium/BW17531020].
historic material mentioned below) are accessible online via JSTOR Plant Science or the collection
databasesoftheherbaria,respectively.
Label information of the type material in B does not include a collection date or a locality
apart from Fretum Magellanicum. However, labels attached to one of the isolectotypes present in
Paris (P01759004) give further information on the date and locality where it was collected. The
inscription Xbri 1767 is indicating that the material was collected in December 1767. This is
confirmed by the fact that the passage of the Strait of Magellan lasted from December 1767 to
January 1768 (Bougainville 1772). Several labels list localities where the species was found in the
Strait of Magellan: Baye francaise, Baye Bougainville, Port Galant & in Sylvis Commersonianis, In
97
alpibus Commersonianis, Port Bougainville. However, it remains unclear in which of these precise
localitiesthematerialwascollected.
The label on the isolectotype of T. magellanicum in C was written by M. Vahl (information
from P. Wagner via O. Seberg/G. Petersen, pers. comm.), the botanist who sent the material to
Willdenow. A specimen located in the Herbier Genve (G00098732) and another in the Herbarium
HAL (HAL0109729), both annotated as type material (by R. R. Haynes & L. B. HolmNielsen in 1985
andbyN.Tkachin2010,respectively)couldrepresentfurtherisolectotypes.However,inbothcases
the source could not be identified with certainty and the specimens are therefore not listed as
isolectotypes until further evidence becomes available. Further sets of material collected during
Bougainvilles circumnavigation were distributed to several other herbaria (Stafleu & Cowan 1976
1988)whichaccordinglymighthousefurthertypematerialofTetroncium.
Triglochin reflexum was used as a provisional name by Vahl and not validated by later
publication.Willdenow(1808)referredtothisnameintheprotologueofTetronciummagellanicum.
ThenameTriglochinmagellanicumwasusedbyWilldenowontheherbariumsheet(seesmalllabel
on lectotype, Fig. 3C). Both names were cited by Kunth (1841) as synonyms under Tetroncium
magellanicum,butthisdidnotresultinthevalidpublicationofthesenames(seeArt.34.1c,McNeill
&al.2006).
Ic.:Hooker1843,t.534;Hooker1847,t.128(Fig.5);Correa1969,Fig.15,p.36.
Description(basedonownobservationsandmeasurementsofherbariummaterialandpartlyon
Hooker1847,Buchenau1903,Moore1968,1983;seealsoFig.6)
Perennial,rhizomatous,glabrousherb,525(35)cmhigh.Rhizomeascending,upto20cmlong,24(
6) mm in diameter, producing aboveground stems towards the apex. Stems ascending or erect,
branchingnearbase,denselycoveredwithbrownleafremainsatbaseandwithleavestowardsapex.
Leavesdistichous,simple,coriaceous,rigid,linearensiform(swordshaped),acute,(15)20100(120)
1.53.0 mm, equitant, with basal sheath but without ligule or auricles. Plant dioecious, scapose;
scapes erect, 25200(250) mm long. Inflorescences terminal, dense, c. 1050 mm long, ebracteate
spikes of up to 30 flowers; pedicels very short (c. 0.5 mm) or absent. Male flowers: tepals 4,
yellowishwithreddishbrownspots,concaveorconchiform,broadlyovate,acuteorsubobtuse,1.2
2.0 0.51.5 mm. Stamens 4, opposite tepals and inserted at their base, (sub)sessile; anthers
yellowish, extrorse. Female flowers: tepals 4, as in male flowers but narrower, ovate to ovate
lanceolate. Carpels 4, subulate, fused from base to about half of their length, with one basal
anatropousovulepercarpel;stylesdivergent.Fruits4locular(orpossiblyunilocularwithincomplete
septa),dry,indehiscent,reflexed,narrowlyconical,withlongpersistentbeaklikestyles,48(10)c.
1 mm, smooth, reddish brown, usually oneseeded through abortion of three ovules. Seeds linear
oblong,endospermic.Polleninaperturate,subspheroidal,mediumsized,dispersedinmonads,exine
reticulate(Kupriyanova1948,Grayum1992).Chromosomenumberunknown.
Tetroncium differs from the other genera of Juncaginaceae mainly in its dioecy and flower
merosity (dimerous vs. trimerous in Triglochin and Cycnogeton). Further differences include seeds
withendosperm,thegeneralhabit,aswellasleafformandstructure(ensiform,rigidandcoriaceous
in Tetroncium vs. semiterete, succulent in Triglochin and flattened, strapshaped, spongy in
Cycnogeton).
98
Note: Willdenow (1808, Fig. 2) as well as Hooker (1843) erroneously described the flowers apart
from the gynoecium as trimerous instead of dimerous. This was clarified by Hooker (1847, see also
Fig.5)andlaterdiscussedbyBuchenau(1882).Furthermorphologicalanddevelopmentalstudiesare
neededtoexamineflowerandfruitstructureinmoredetail.
Distribution
The distribution area of T. magellanicum comprises mainly southern South America

(PatagoniaandTierradelFuego)northwardstoc.40SinwesternArgentinaandtoc.37SinChile
(Fig. 4 & 5). Collections from Parque Nacional Nahuel Huapi in the Andes (Argentina) as well as
ParqueNacionalNahuelbutaandParqueNacionalChilointheChileanCoastalRangerepresentthe
northernlimitsofthedistributionofthespecies.ItisalsodistributedintheFalklandIslands(Eastand
WestFalkland)butabsentfromtheLafoniaregionofEastFalklandprobablyduetosummerdrought
there(Broughton&McAdams2005).Furthermore,TetronciumisfoundonGoughIslandintheSouth
Atlantic Ocean (Wace 1961, Groves 1981) but not known from the other islands of the Tristan da
Cunhagroup(N.Gremmen,pers.comm.).
Fig.4and5givethefirstdetailedmapoflocalitiesrepresentingtheknowndistributionarea
of the species. Previously, only a map showing a rough outline of the distribution area (without
GoughIsland,Camp1947)andamaplimitedtoTierradelFuego(Moore1983)hadbeenpublished.
Habitat,ecologyandconservationaspects
Tetronciummagellanicumisfoundindifferentpeatlandtypes,especiallySphagnumbogsand
cushion bogs. These peatlands are part of characteristic vegetation sometimes termed Magellanic
moorland (e.g., Moore 1983, Arroyo & al. 2005 and references within). Tetroncium is common in
continental bogs dominated by Sphagnum magellanicum Brid., where the moss forms orange or
reddish carpets or hummocks (Fig. 1A). Tetroncium magellanicum predominantly occurs in wetter
parts of the peatlands and is frequently found accompanied by e.g. Carex magellanica Lam. (e.g.,
Kleinebecker&al.2007).InPacificbogsdominatedbycushionformingvascularplantsTetronciumis
associated with cushion plants such as Astelia pumila (J.R.Forst.) Gaudich. and Donatia fascicularis
J.R.Forst. & G.Forst. or Drosera uniflora Willd. (e.g., Dusn 1900, Moore 1983, Kleinebecker & al.
2007; A. Vogel, pers. comm.; B. Ruthsatz, pers. comm.). In a transition zone both peatland types
intermingleoroccurmosaiclikesidebyside(Kleinebecker&al.2007).Tetronciumisalsofrequentin
this Sphagnum magellanicum cushion plant mixed type and along a moisture gradient in all three
types (Kleinebecker & al. 2007). Highest frequencies were, however, recorded from Sphagnum
cuspidatumhollows(exclusivelyfoundincontinentalSphagnumbogsandSphagnumcushionmixed
bogs)andwetS.magellanicumcarpetsoftensurroundinghollowsofcontinentalbogs(Kleinebecker
&al.2007).
Habitats belonging to the Magellanic moorland are continuously distributed from the
extremesouthofthecontinenttoaround43SinChileandfromsealeveltoabovetreeline(Pisano
1983). Tetroncium magellanicum is not only known from most parts of this region but also from
outlying areas of moorland vegetation occurring further north, e.g. in the Cordillera de Piuchu
(Chilo Island) and the Cordillera de Nahuelbuta, both in the Chilean coastal range (Ruthsatz &
Villagrn1991,Arroyo&al.2005,seeFig.4&5).
99
Fig.4.DistributionofTetronciummagellanicumsouthernSouthAmericaincludingFalklandIslandsand
GoughIslandintheSouthAtlantic(ca.2,800kmfromCapeTown,SouthAfricaand3,300kmfromtheSouth
Americanmainland);mapcreatedusingtheonlinetoolEDITmapViewer(2012).
Fig.5.DistributionofTetronciummagellanicumChileandArgentinamapcreatedusingtheonlinetool
EDITmapViewer(2012).
100
In southern Chile, Magellanic moorland occurs together with subantarctic evergreen forest
dominated by Nothofagus betuloides (Mirb.) Oerst. (e.g., Dusn 1903, Ruthsatz & Villagrn 1991).
Furthereastsuchpeatlandscanalsobefoundinthetransitionzonetodeciduousforestdominated
by N. pumilio (Poepp. & Endl.) Krasser and N. antarctica (G. Forst.) Oerst. Tetroncium grows in
clearingsoropenpartsoftheseNothofaguswoodlands.
Habitats on the Falkland Islands are bogs, dwarf shrub heath and acid grassland (Moore
1968,1983).OnGoughIslandT.magellanicumisoneofthefewabundantfloweringplantsfoundin
high altitude bogs, i.e. Sphagnumdominated peat bogs above 600 m (Wace 1961; N. Gremmen,
pers.comm.).
The species is usually found from sea level to about 500 metres. Higher altitudes of up to
1300 metres are only recorded from the northernmost part of the distribution area (e.g., Parque
NacionaldeNahuelbuta,Ricardi5697&Marticorena,CONC;Zuloagaetal.2008).
Tetroncium is predominantly found in wetter sites and tolerates frequent and prolonged
periodsofinundation.Sometimesonlyupperleavesandinflorescencesareemergingfromthewater.
Patagonian bog ecosystems are characterised by extreme nutrientpoor conditions and T.
magellanicumappearstobehighlyefficientinnutrientacquisition(Schmidtetal.2010).
Phenology: Flowering and fruiting in summer, mainly from December to January (sometimes to
April). The species is probably windpollinated as most other species of the family. Fruits of T.
magellanicum are indehiscent (pers. obs.). Their long, rigid appendages formed by the persistent
styles might serve dispersal by animals such as birds. However, pollination as well dispersal
mechanisms have not been directly observed in Tetroncium. Male and female plants can grow
intermingled (see, e.g., specimen James 1500, BM, SI) but unisexual populations can be also
separatedbysomedistance(R.W.Woods,pers.comm.).ThisimpliesthatT.magellanicumisableto
reproducevegetatively,probablythroughregenerationfromrhizomefragments.
Proposed IUCN conservation status: Least Concern (LC). Tetroncium magellanicum is relatively
widespreadinsouthernSouthAmerica,locallyabundant,andnotunderimmediatethreat.However,
asintheNorthernhemisphere,peatlandsintheSouthernhemisphereareincreasinglythreatenedby
drainage(e.g.,forurbandevelopmentorroadconstruction)andotherlandusechanges,inparticular
peatmining.Inrecentdecades,thenumberofpeatextractionsiteshasrisencontinuouslyespecially
inTierradelFuego,andfurtherextractionpermitshavebeenissued(e.g.,Blanco&delaBalze2004,
Iturraspe&Urciuolo2004,Grootjansetal.2010;A.Vogel,pers.comm.).
Tetroncium magellanicum is not confined to continental Sphagnum bogs which are most
affected by this growing trend in peat extraction but also occurs in oceanic cushion bogs which are
protected to a higher degree. Nevertheless, conservation assessments might become more
importantinthefuture,especiallyonaregionalscale.
Fortunately,arelativelyhighproportionofsouthernSouthAmericanwetlandsareprotected
as part of National Parks or National Reserves (e.g., Parque Nacional Tierra del Fuego in Argentina,
BernardoO'HigginsNationalParkandCabodeHornosNationalParkinChile).However,protectionof
different peatland types is unbalanced, and Nothofagus forests further north in the Coastal range
containing disjunct areas of peatlands have been reported to be under threat (Arroyo & al. 2005).
Further threats to fragile wetland ecosystems include the introduction of nonnative species and
tourism.
101
Fig.6.TetronciummagellanicumPlatefromFloraAntarcticaillustratingmaleandfemaleplants,
flowersandfruits(Hooker1847).[Publishedathttp://biodiversitylibrary.org/page/13448638].
Gough Island is recognized as a nature reserve and an ordinance entitled 'Conservation of

Native Organisms and Natural Habitats (Tristan da Cunha) Ordinance 2006' which replaced the
previous conservation ordinance from 1976 gives protection to all all native plants including T.
magellanicum.TheislandwasalsoinscribedtotheUNESCOWorldHeritageListin1995(modifiedin
2004)aspartoftheGoughandInaccessibleIslandsWorldHeritageSite.Nevertheless,threatstothe
islands ecosystems are not to be underestimated and include (further) introduction of nonnative
speciesandclimatechange(e.g.,Jones&al.2003,Middleton&Kleinebecker2012).

102

Chapter5
MorphologyofMaundiasupportsitsisolatedphylogeneticposition
intheearlydivergentmonocotorderAlismatales
DmitryD.Sokoloff,SabinevonMering,SurreyW.L.Jacobs
*
andMargaritaV.Remizowa
ManuscriptwillbepublishedintheBotanicalJournaloftheLinneanSociety(inpress)
ABSTRACT
According to recent molecular phylogenetic data, the rare Australian endemic Maundia
triglochinoidesdoesnotformacladewithtaxatraditionallyclassifiedasmembersofJuncaginaceae.
Therefore, views on the morphological evolution and taxonomy of Alismatales require re
assessment. As the morphology of Maundia is poorly known and some key features have been
controversiallydescribedintheliterature,theflowers,fruits,inflorescenceaxesandpeduncleswere
studiedusinglightandscanningelectronmicroscopy.Inflorescencesarebractlessspikeswithflowers
arranged in trimerous whorls. Except in the inflorescence tip (where the flower groundplan is
variable), flowers possess two tepals in transversalabaxial positions, six stamens in two trimerous
whorls and four carpels in median and transversal positions. Fruits are indehiscent. The shared
possession of orthotropous ovules supports the molecular phylogenetic placement of Maundia as
sister to a large clade including Potamogetonaceae and related families. Maundia and Aponogeton
spp.sharethesamehighlyunusualfloralgroundplan,ahomoplasticsimilaritythatcanbeexplained
byspatialconstraintsindevelopinginflorescences.ThenucellarcoenocyteofMaundiaappearstobe
uniqueamongmonocots.AsMaundiaexhibitsamosaicoffeaturescharacteristicofotherfamiliesof
tepaloidcoreAlismatales,itssegregationasaseparatefamilyisplausible.
Keywords:anatomy,Aponogetonaceae,bract,flower,fruit,Juncaginaceae,Maundiaceae,nucellus,
ovule,Potamogetonaceae,tepal,vasculature
*MaterialcrucialforthisstudywasprovidedbythelateSurreyJacobs,whosadlypassedaway
beforethisstudycouldbecompleted.HecontributedtoearlierdiscussionsonMaundia,provided
importantadditionalinformationandalloweduseofphotographstakenbyhiminthefield.

Chapter5:MaundiaandevolutionofAlismatales
104
105
Introduction
ThecoreAlismatales(e.g.,Iles,Smith&Graham,2013),aspeciespoorbutmorphologically
highly diverse monophyletic group of aquatic and wetland plants traditionally known as the order
Helobiae (Engler, 1909; Eckardt, 1964), superorder Alismatanae (Takhtajan, 2009) or Alismatiflorae
(Dahlgren,Clifford&Yeo,1985)orsubclassAlismatidae(Takhtajan,1987,1997;Les&Tippery,2013),
havelongbeenafocalpointofdiscussionregardingtheevolutionaryhistoryofmonocots.Thegroup
wasoftenviewedasabasalmonocotlineage,probablysistertotherestofthemonocotyledons,an
opinion supported by the highly unstable floral groundplan, frequent occurrence of apocarpy and
some other features considered as potentially primitive (Wettstein, 1924; Takhtajan, 1966, 1987;
Cronquist, 1981). Molecular phylogenetic data do not support the hypothesis of the basalmost
placementofHelobiaeamongmonocots,butstillshowthattheorderAlismatalesbelongstoagroup
ofearlydivergentmonocots(Chaseetal.,2000,2006;Davisetal.,2004;Grahametal.,2006;APGIII,
2009; Iles, Smith & Graham, 2013). Thus, Helobiae are significant for the understanding of early
monocot evolution. Molecular phylogenetic trees suggest that apocarpy and an unstable flower
groundplan could be derived rather than ancestral features in core Alismatales (Doyle & Endress,
2000; Chen et al., 2004; Endress & Doyle, 2009; Remizowa, Sokoloff & Rudall, 2010; Sokoloff,
Remizowa & Rudall, 2013). The high interest in members of core Alismatales has made this group
one of the most extensively studied with respect to comparative flower morphology and
development (reviewed by Posluszny & Charlton, 1993; Posluszny, Charlton & Les, 2000; Remizowa
etal.,2012b).However,afewkeytaxaremainpoorlyknown,particularlyduetotechnicalproblems
with obtaining appropriate plant material. In this paper, we present the first detailed data on
peduncle, flower and fruit anatomy in Maundia triglochinoides F.Muell., a member of core
Alismatales whose phylogenetic placement has been reassessed using molecular phylogenetic data
(vonMering&Kadereit,2010;Iles,Smith&Graham,2013;Les&Tippery,2013).Thenewphylogeny
requiresanupdatedcomparativeanalysisofmorphologicalcharactersinMaundiaF.Muell.,manyof
whicharecurrentlyeitherpoorlyknownorforwhichexistinginterpretationsarecontroversial.
The Australian endemic M. triglochinoides is an erect rhizomatous perennial herb restricted
tofreshwaterswampsandstreamsincoastalNewSouthWales,extendingintosouthernQueensland
(Aston, 2011). Due to habitat loss and fragmentation, the range of the species has been much
reduced (Sainty & Jacobs, 2003) and it is listed as Vulnerable in New South Wales (Schedule 2,
ThreatenedSpeciesConservationAct)andinQueensland(Schedule3,NatureConservation(Wildlife)
Regulation 2006 SL No. 206 [Nature Conservation Act 1992]). Traditionally, the monotypic genus
MaundiawasclassifiedasamemberofthefamilyJuncaginaceae(Mueller,1858;Hutchinson,1959;
Eckardt, 1964; Cronquist, 1981; Dahlgren, Clifford & Yeo, 1985; Haynes, Les & HolmNielsen, 1998;
Takhtajan, 1966, 2009; Seberg, 2007). Earlier classifications treated the family in a wide sense to
include Scheuchzeria L. using the names Juncaginaceae s.l. (Buchenau & Hieronymus, 1889;
Shipunov, 2003) or Scheuchzeriaceae (Buchenau, 1903), respectively. Nakai (1943) proposed a
monogeneric family Maundiaceae. Subsequently, this family was accepted by Takhtajan (1987,
1997).MoleculardatasupportedtheideathatMaundiashouldbeexcludedfromJuncaginaceae(von
Mering & Kadereit, 2010). According to Iles, Smith & Graham (2013), Aponogeton L.f. (Apono
getonaceae), Scheuchzeria (Scheuchzeriaceae), Triglochin L. (Juncaginaceae) and Maundia form
successive branches in a grade leading to a group of more specialized aquatic Alismatales such as
Zosteraceae,Potamogetonaceae,Posidoniaceae,CymodoceaceaeandRuppiaceae.
106
Todate,thereisnoconsensusregardingthefamilyplacementofMaundia.Aston(2011)and
Reveal & Chase (2011) continued to use the traditional concept of Juncaginaceae, while Reveal
(2011), Stevens (2001+) and Les & Tippery (2013) accepted the monogeneric Maundiaceae.
Furthermore,APGIII(2009)suggestedthatmorestudyisneededbeforeMaundiaceaeisrecognized
as another monogeneric family in Alismatales. According to APG III (2009) and Stevens (2001+), it
might be better in this case to create a larger single family for the larger clade. As reviewed by von
Mering & Kadereit (2010), several morphological characters of Maundia flowers are controversially
interpreted in the literature. These include the presence or absence of a perianth and bracts and
interpretation of stamens as bisporangiate and monothecal or tetrasporangiate and dithecal,
respectively. We use our new anatomical evidence to discuss these issues. In addition, we improve
existing descriptions of carpel arrangement in Maundia. Ovule type (orthotropous vs. anatropous)
was used as the main morphological character distinguishing Maundiaceae from Juncaginaceae
(Nakai, 1943; Takhtajan, 1987, 1997). Although the orthotropous ovule is nearly always indicated in
descriptions of Maundia, detailed descriptions of ovule anatomy are not available. Existing
descriptionsoffruitmorphologyinMaundiaarecontroversial(e.g.,Bentham,1878;Cronquist,1981;
Aston, 2011). As pointed out by Thieret (1988), gynoecia of Maundia and Tetroncium Willd.
apparently differ from those of other Juncaginaceae, and a developmental study of fruits of both
genera is called for. The present study improves knowledge on fruits and seeds/ovules ofMaundia.
Finally,weprovideobservationsonthevegetativeanatomyandpollenmorphologyofMaundia.
Materialandmethods
The following collection was studied: Maundia triglochinoides: Australia, New South Wales,
PortersCreekWetland,Wyong,entrypointintoswamp,331536.7S,1512611.4E,elev.14m,3
Dec2008,L.Stanberg&G.SaintyLS80(NSW810429,duplicatesinC,K,MJG).Theplantoccurredin
a depression (20% shaded) in Melaleuca linearifolia (Link) Craven woodland with occasional
scattered Eucalyptus robusta Sm., with associated species such as Typha orientalis C.Presl., Villarsia
exaltata (Sims) G.Don, Persicaria sp., Baumea rubiginosa Boeckeler, B. articulata (R.Br.) S.T.Blake,
Alternanthera denticulata R.Br., Juncus polyanthemus Buchenau, Lachnagrostis filiformis Trin.,
Sagittaria platyphylla (Engelm.) J.G.Sm., Carex polyantha F.Muell., C. appressa R.Br., Cyperus
eragrostisLam.,RanunculusinundatusR.Br.exDC.,Hypolepissp.Inthislocality,Maundiawaslocally
abundant,forminganalmostpurecontinuousstandinadepressioncoveringc.1acre,inwater20
30cmdeep.
Inflorescences and fruits were fixed in FAA and stored in 70% ethanol. For light microscope
observations, material was sectioned using standard methods of paraplast embedding and serial
sectioning at 15 mm thickness (e.g. Barykina et al., 2004). Sections were stained with
picroindigocarmineandcarbolicfuchsine(Axenov,1967)oralcianblueandsafraninandmountedin
Biomount. Both cross sections and longitudinal serial sections were made, of flowers, fruits and
inflorescence axes. In addition, freehand sections of peduncles and fruits were treated with
phloroglucinolandhydrochloricacid(toreveallignificationofcellwalls)orI/KI(torevealstarch)and
subsequentlyobservedin glycerol.Sectionswereexaminedandimagesofthemwere takenusinga
Zeiss Axioplan microscope. Threedimensional models of floral vasculature were constructed using
3DDoctor. For scanning electron microscopy (SEM), the material was dissected in 96% ethanol and
dehydrated through absolute acetone and criticalpoint dried using a Hitachi HCP2 critical point
dryer,thencoatedwithgoldandpalladiumusinganEikoIB3ioncoater(Tokyo,Japan)andobserved
using CamScan 4 DV (CamScan, UK) at Moscow University. In addition to fixed material, herbarium
107
specimensfromseveralcollectionswerestudied.ThesearelistedintheAppendix.Terminologyused
in the Results section reflects our preferred morphological interpretation; other interpretations are
reviewedandcriticallyevaluatedintheDiscussion.
Results
GENERALMORPHOLOGY,PEDUNCLEANATOMYANDINFLORESCENCESTRUCTURE
Maundiatriglochinoides(Fig.1AE)isaperennialaquaticherbwithallleavesrestrictedtoa
creeping rhizome. Erect flat linear, eligulate foliage leaves emerge from the water. Inflorescence
pedunclesareterete,long,exposingflowersabovethewaterlevel,andlackinganyfoliageleavesor
scales.
The peduncle (Figs 1FI, 2) is covered by a onelayered epidermis. Epidermal cells are
elongated along the peduncle. Stomata are present at least in the distal part of the peduncle (Fig.
2D).Guardcellsareelongatedalongthelengthofthepeduncle.Cortexiseitherabsent(inthiscase,
theoutermostvascularbundlesareadjacenttotheepidermis)orrepresentedbyonetoeightlayers
ofthinwalledcells(Figs1GI,2A).Thestelecontainsnumerousvascularbundlesarrangedwithouta
clearpatternincrosssectionsofpeduncle.Centralbundlesarelargerthanperipheralones(Fig.1A).
Fig.1.Maundiatriglochinoides.AE,livingplants(photographstakeninnaturebyS.Jacobs).A,inflorescence
(bractlessspike).B,postantheticinflorescence.C,D,detailsofimmaturefruits.E,inflorescenceaxiswithfruits.
FI, details of freehand transverse section of peduncle, treated with phloroglucinol and hydrochloric acid
(lignified cell walls orange). F, one of central vascular bundles. GI, peripheral vascular bundles, epidermis of
peduncle bottom. G,H, inverted peripheral bundles (typical condition). I, obliquely oriented bundle (rare
condition). Scale bars,100m (FI). ac, air canals; f, fibres associatedwith vascularbundle;blackarrowhead,
phloem;whitearrowhead,xylemtracheid;asterisk,protoxylemlacuna.
108
Fig. 2. Maundia triglochinoides. Peduncle anatomy and stoma on peduncle (scanning electron microscopy,
SEM). A, Peduncle in transverse section. B, Detail of vascular bundle. C, Septum on air canal. D, Stoma on
peduncle. Scale bars: 1 mm (A); 200 m (B); 100 m (C); 10 m (D). ac, air canal; cvb, large central vascular
bundle; ph, phloem; pvb, small peripheral vascular bundle; pxl, protoxylem lacuna; se, septum in air canal; t,
tracheidwithspiralthickenings.
Thevascularbundlesarecollateral(Figs1FI;1B).Inlargebundles,theprotoxylemisrepresentedby
a lacuna (Figs 1F, 2B). The tracheids of the metaxylem form a horseshoeshaped row in cross
sections, adjacent to the phloem (Fig. 1B). Tracheids possess spiral thickenings (up to five parallel
spirals per cell). Vessels were not observed. Most peripheral bundles are inverted, i.e., with xylem
orientedtowardstheepidermisandphloemtowardsthecentreofthepeduncle(Fig.1G,H).Oneof
theobservedperipheralbundleswasobliquelyoriented(Fig.1I).Eachlargebundleissurroundedby
analmostcompletesheathofthinwalledlignifiedfibres(Fig.1F).Smallestbundlespossessfibres
109
Fig. 3. Maundia triglochinoides. Flower groundplan, normal flowers (scanning electron microscopy, SEM). A,
Flower from the central part of the inflorescence, top view. B, Postanthetic flower from the abaxial side. C,
Flowerfromtheinflorescencebase,topview.D,Flowerwithgynoeciumremoved.Scalebars:500m(AD).c,
carpel;gs,stalkofremovedgynoecium;ia,inflorescenceaxis;is,innerwhorlstamen;os,outerwhorlstamen;
te,tepal.
alongthephloemsideonly(Fig.1G).Mediumsizedbundlespossesstwogroupsoffibres,i.e.,along
the xylem and the phloem side (Fig. 1I). All space between the bundles is filled by an aerenchyma
with large air canals separated by uniseriate files of thinwalled cells (Fig. 1A); these cells contain
starchgrains(therearefewergrainsinthecellsattheperipheryofthepeduncle).Alongthelength
110
Fig.4.Maundiatriglochinoides.Stamens,tepals,pollengrains,stomaontepal(scanningelectronmicroscopy,
SEM). A, Median abaxial outerwhorl stamen not associated with a tepal and tepals on radii of transversal
abaxialinnerwhorlstamens.B,Flowerwithgynoeciumremovedshowingmedianabaxialouterwhorlstamen,
twotepalsandinnerwhorlstamenssituatedonradiiofthetepals.C,D,Tepalstamenpairsindifferentviews.
E, Removed stamen, adaxial view. F, Removed tepal, adaxial view. G, Basal part of young fruit to show non
abscisedstamens.H,Stomaontepal.I,Pollengrain.J,Exinesculpture.Scalebars:1mm(A,B);500m(C,D,F,
G);300m(E);20m(H);5m(I);1m(J).c,carpel;gs,stalkofremovedgynoecium;ia,inflorescenceaxis;is,
innerwhorlstamen;os,outerwhorlstamen;te,tepal;th,theca.
111
of the peduncle, the air canals are divided into chambers by transverse septa (Fig. 1A, C). Narrow
perforations connecting adjacent air lacunae are present between cells forming a septum (Fig. 1C).
Cells forming septa lack starch grains. No specialized mechanical elements are present in peduncles
exceptthefibresassociatedwithvascularbundles.
Fig. 5. Maundia triglochinoides. Transverse sections of anthetic flowers (light microscopy, LM). AE, A
descending series of sections of a flower. A, Level of oblique carpel mouths. B, Level of free ascidiate carpels
with carpel canals (arrowheads) displaced towards ventral sides of carpels. C, Carpels are united via floral
centre, just below the level of ovule attachment; note four ventral carpel bundles in the central part of the
gynoecium.D,Aboutthelevelofthemicropylesofovules(belowthemicropyleintheuppercarpel,abovethe
micropyleinthelefthandcarpel);atthislevel,thecaeoftheinnerwhorlstamensareunited,whereasthecae
oftheouterwhorlstamensarefreefromeachother.E,Belowtheovarylocules,eachcarpelwithadorsaland
a ventral bundle; thecae of outerwhorl stamens are united. F, Another flower with asymmetric carpel
arrangement and an unvascularized bulge (large arrowhead) that could be interpreted as an incipient fifth
carpel.Smallarrowheadindicatesanenlargedcellflankingthedorsalbundle;thesecellswillbeconspicuousin
fruits.G,Stamenthecaatthelevelaboveseparationoffreethecae.H,Antheratthelevelofunited
thecae; note the occurrence of two vascular bundles. Scale bars: 300 m (AF, H); 100 m (G). c, carpel; db,
dorsalcarpelbundle;is,innerwhorlstamen;os,outerwhorlstamen;ov,ovule;te,tepal;vb,ventralbundle.
112
Inflorescencesareunbranchedspikes(Fig.1A).Flowersarearrangedalongtheinflorescence
axis in regularly alternating trimerous whorls. In preanthetic inflorescences, the internodes of the
inflorescenceaxisareshort,theflowersaredenselyspaced,andtheinflorescenceaxisisusuallynot
visiblewithoutremovingflowers.Themostproximalinternodecanbeslightlylongerthantheother
internodes (this is also typical for several species of Triglochin S.v.M., unpubl. data). Six
orthostichies of flowers can be easily recognized. In postanthetic inflorescences, internodes of the
inflorescence axis are longer and visible between the flowers or young fruits. At these stages,
different flowers of the same whorl may be inserted at slightly different levels of the inflorescence
axis. No flowersubtending bracts (or any rudiments) were observed (Figs 3B, D, 4A, B). Flowers are
completelysessile,sometimeswiththeexceptionoftheuppermostflowersinaspike.Asnoflower
stalksaredevelopedafteranthesis,fruitsremainsessileandperpendiculartotheinflorescenceaxis
(Fig.1CE).
MORPHOLOGYANDVASCULARANATOMYOFFLOWERS
Almost all flowers, except the uppermost ones, exhibit a stable groundplan. There are two
tepals in transversalabaxial positions, six stamens in two alternating trimerous whorls (an outer
whorlwithamedianabaxialandtwotransversaladaxialstamensandaninnerwhorlwithamedian
adaxialandtwotransversalabaxialstamens)andfourcarpels,ofwhichtwoareinmedianandtwoin
transversal positions (Figs 3, 4A, B, 5AE). In one flower, a small, unvascularised outgrowth was
foundinthetransversaladaxialposition;thiscouldbeinterpretedasanincipientfifthcarpel(Fig.5F,
arrowhead). In this flower, two carpels situated on the opposite radius were more closely spaced
thanintypicalflowers(Fig.5F).
Tepals are green and ca. 1.5 times as long as the stamens. They have narrow bases and are
attached to the receptacle at the radii of the transversalabaxial inner whorl stamens (Figs 3, 4A, B,
5AF).Stamenandtepalbasescanuniteforaveryshortdistance(Fig.4C,D).Thetepalsareinserted
atapproximatelythesamedistancefromtheflowercentreastheouterwhorlstamens(Figs3B,4A,
B).Theyhaveashortclawgraduallyextendedintoanalmostorbicularblade,whichiscurvedinwards
(Fig.4C,D,F).Thetepalbladeisconspicuouslythick,consistingofseveralcelllayersincrosssections
ofthemiddlepart(Fig.5A).Abundantstomataarepresentontheabaxialsurfaceofthetepalblade.
Theguardcellsarebeanshaped(Fig.4H).Distinctcuticularridgesarepresentalongtheouterorifice
oftheaperture.Epidermalcellsintheimmediatevicinityoftheguardcellsaresmallerthantherest
oftheepidermalcells;theirnumberandarrangementrelativetotheguardcellsdonotappeartobe
preciselyfixed.
Stamens are yellow, tetrasporangiate and dithecal (Fig. 5H). Stamen filaments are absent,
andantherconnectivesarewide(widerthanlong)andveryshort,ca.onethirdaslongasthethecae
(Fig.4CE).Freethecaearestronglyextendedabovetheconnectiveandslightlyextendedbelowthe
connective (Figs 4E, 5). Anther dehiscence is extrorse (Figs 3B, C, 4A, B). The line of dehiscence
terminates a short distance from the acute distal end of a theca. At the obtuse proximal end of a
theca,thedehiscencelinecurvesfromtheoutertotheinnersideofthetheca(Fig.4E).Thestamen
epidermislacksstomata.Theendotheciumcellsshowfibrousthickenings.Pollengrainsarespherical,
inaperturate,andthepollensurfaceisreticulate(Fig.4I,J).
Carpels are pronouncedly ascidiate. In mature preanthetic flowers (the youngest stage
available in this study), the carpels are congenitally united at the base via the floral centre, and the
unitedpartisaboutaslongasthefreepartsofthecarpels(Fig.6B).Thegrowthoftheventralsides
offreepartsofthecarpelsisapparentlydelayedwithrespecttotheirdorsalsidesonearlystagesof
113
Fig.6.Maundiatriglochinoides.Gynoeciumatanthesisandfruits(scanningelectronmicroscopy,SEM).A,Free
part of carpel at anthesis, ventral side. B, Longitudinally dissected anthetic flower showing relative length of
unitedandfreepartsofcarpels.C,Detailof(B)showingovule.D,Crosssectionofyoungfruit.E, Topviewof
youngfruit.F,Sideviewofyoungfruit.G,Longitudinalsectionofyoungfruit.H,Detailof(G).Scalebars:100
m(A,C,H);500m(B);1mm(D,E,F,G).c,carpel;ii,innerintegument;mc,micropyle;nc,nucellus;oi,outer
integument;s,developingseed;th,stamentheca.
114
Fig. 7. Maundia triglochinoides. Threedimensional reconstructions of floral vasculature. Fusion bundles =

bundles formed by fusion of traces from different organ types. AD, Different views of the entire floral
vasculature. A, Top view. B, Oblique top view. C, Bottom view. D, Side view. E, Top view of vasculature with
gynoeciumbundlesremoved.F,Topviewofvasculaturewithgynoeciumandfusionbundlesremoved.G,The
sameas(F),bottomview.H,I,Top(H)andside(I)viewofgynoeciumvasculature.db,dorsalbundle;vb,ventral
bundle.
development. In the earliest available stage, the carpel mouth (which is strongly oblique) is located
onitsinnersidebutdoesnotextendalongtheentirelengthofthefreepartofacarpel(Fig.6A,B).
Ontheinnersideofthefreepartofacarpel,thereisadistinctcongenitallyclosedportionbelowthe
mouth, which is the morphologically ventral area, and a portion above the mouth, which is part of
the morphologically dorsal surface (Fig. 6A). Short stigmatic papillae are present around the carpel
mouth(Fig.6E).Incrosssectionsbelowthecarpelmouth,acanalelongatedinaradialplaneis
115
Fig.8.Maundia triglochinoides. Sketches of cleared tepals viewed from the adaxial side showing variations in
vasculature.Scalebar(commontoallimages):500m.
present(Fig.5B).Thecanalisverynarrowcomparedtothewidthoftheentirecarpel.Itislocatedon
theventralsideofthecarpelveryclosetoitssurface(Fig.5B).
Each carpel has a single ovule inserted ventrally just below the level of carpel separation
fromthefloralcentre(Fig.6B,C).Theovuleispendent,bitegmicandorthotropous.Themicropyleis
formed by the inner integument. The ovary locule is narrow compared to the carpel width, and
circularincrosssection(Fig.5C,F).Theloculeisclosertothefloralcentrethantothedorsalsideofa
carpel(Fig.5C,F).Theovulefillsthelocule(Figs5C,F,6B,C).Itisinclosecontactwiththeloculewall
exceptinthemicropylarregion(Fig.5D).
Each flower is supplied by a single strand of conductive tissues, which could also be viewed
as a group of closely spaced bundles. Conductive tissues soon form a complete or incomplete ring
(with xylem on the inner and phloem on the outer side) that is subdivided upwards into distinct
individual bundles. Each tepal usually receives a single vascular bundle from the receptacle, which
furtherbranchesinthetepalblade(Figs5A,B,F,7AG,8).Bundlesarelocatedclosertotheadaxial
thantotheabaxialtepalsurface.Examinationoftotalremovedtepalsallowedassessingthevariation
in tepal vascularisation (Fig. 8). Some of the examined tepals were apparently threetraced.
Anastomoses between tepal bundles are usually but not always present. Vein endings remain free.
The endings in the tepal blade are usually thicker than the bundle(s) in the tepal claw. Some free
endingsaredirectedtowardsthetepalbase(Fig.8).Eachstamenthecahasasinglevascularbundle
extending into the very proximal portion of its free part (Fig. 5D); the rest of the theca is non
vascularised (Fig. 5G). Usually, these two bundles remain distinct in the anther connective (Fig. 5H)
andfusetoformacommonstamentraceintheflowerreceptacle(Fig.7AG).Innerwhorlstamens
arevariablewithrespecttothepresenceorabsenceofthiscommonstamentrace.Inthelattercase,
thebundlefromeachthecadownwardsseparatelyreachestheringofconductivetissuesatthebase
ofthereceptacle(Fig.7EG).Thetepaltracejoinsthecommontraceofaninnerwhorlstamenofthe
same radius or enters the stele of the receptacle in between the free traces of the two thecae of
suchaninnerwhorlstamen(Fig.7AG).Eachcarpelhasadorsalandaventralbundle(Figs5BF,H,
7H,I,9DF).Theventralbundlesuppliestheovule,thencontinuesforacertaindistancealongthe
116
Fig.9.Maundiatriglochinoides.Descendingseriesoftransversesectionsofyoungfruit(lightmicroscopy,LM).
AC,Upperportionofthefruitwithfreecarpeltips.A,Carpeltipwithobliquecarpelmouth.B,Carpelclearly
ascidiate, with an open canal displaced towards its ventral side. C, Detail of section below (B) showing post
genitallyclosedcarpelcanalandtwoventralbundlesoneithersideofit.DI,Levelsatwhichcarpelsareunited
via floral centre. D, Just below the level of carpel separation. At this level, the upper lefthand and the lower
righthandcarpelshavepostgenitallyclosedcarpelcanals,whereastwoothercarpelsstillhaveopencanals.E,
Upperpartofovarylocules;twocarpelswithlargeovules(sectionedatthelevelofthefuniculus)completely
fillingtheloculesandtwocarpelswithsmaller,apparently(still?)unfertilizedovules.F,Detailof(E).G,Section
at the middle part of the young fruit, below the level of micropyles of the two smaller ovules. H, Detail of
pericarptissueindorsalpartofacarpelwithlargeaircanals.I,Belowovarylocules,eachcarpelsuppliedbya
singlevascularbundle.Scalebars:500m(A,B,DG,I);100m(C);50m(H).ac,aircanal;cb,commissural
vascularbundleconnectingdorsalandventralbudles;cc,carpelcanal;cm,carpelmouth;db,dorsalbundle;fc,
floralcentre;lc,largecellsinpericarp;ol,ovarylocule;ov,ovule;vb,ventralbundle;arrowheads,postgenitally
closedcarpelcanal.
postgenitally closed carpel canal (Fig. 9D) and ultimately splits tangentially into two bundles above
the level of its insertion (Fig. 9C). These two branches curve along either side of the carpel canal,
approacheachotherandformajointcommissuralbundlethatuniteswiththedorsalbundle(Fig.
117
Fig. 10. Maundia triglochinoides. Pericarp histology on longitudinal sections of young fruits (light microscopy,
LM).A,OutlineofatransversesectionbasedonFigure9Gshowingorientationofsectionsillustratedin(BG).
BG,Longitudinalsections.B,Upperpartofafruit.C,D,Lowerpartofafruit.EG,Middlepartofafruit.Scale
bars:500m(AE);40m(F);200m(G).en,endosperm;ex,exocarp;fn,funiculus;if,intercarpellaryfurrow;
ii,innerintegument;lc,largecellsinpericarp;ol,ovarylocule;ov,ovule;asterisks,aircanals.
9B).Thedorsalandtheventralbundleofeachcarpelproximallyuniteintoasinglecarpeltrace(Figs
7H, I, 9I). Proximally in the receptacle, the vascular traces of the four carpels are connected by
anastomosesforminganalmostcompleteringofconductivetissues(Fig.7H,I).Moreproximallythis
ring again splits into separate strands that irregularly join stamen and tepal traces to form the
proximalringofconductivetissueintheflowerreceptacle(Fig.7D).
118
Fig.11.Maundiatriglochinoides.Transversesectionsofovulesinpostantheticflowers(lightmicroscopy,LM).
A,Funiculuswitharingofvasculartissue.B,Sectionaboveendosperm.C,Sectionofanapparentlyunfertilized
ovule. DF, Sections at the level of endosperm above embryo. G, Section through globular proembryo
surrounded by endosperm. H, Section through gigantic basal cell of suspensor. I, Section through micropyle.
Scale bars: 50 m (AI). db, dorsal bundle; en, endosperm; gc, gigantic basal cell of suspensor; ii, inner
integument (tegmen); lc, large cells in pericarp; nc, nucellus; oi, outer integument (testa); pc, pericarp; pe,
globularproembryo;vb,ventralbundle;asterisks,aircanalsinouterintegument.
119
Fig.12.Maundiatriglochinoides.Chalazalpartsofovulesofpostantheticflowersonlongitudinalsections(light
microscopy,LM).A,Funiculus,chalazaandadjacentpartofnucellus.B,Chalazalpartofnucelluswithnucellar
coenocytic structure. CE, Successive serial sections through a nucellar coenocyte. F, Funiculus, chalaza and
adjacentpartofnucellusofanapparentlyunfertilizedovule.G,Detailof(F).Scalebars:200m(A);100m(B,
F); 50 m (CE, G). ce, nucellar coenocyte; en, endosperm, fn, funiculus; ii, inner integument (tegmen); nc,
nucellus;oi,outerintegument(testa);pc,pericarp;arrowhead,freeendofdegeneratingwallbetweennucellar
cellsformingcoenocyte.
120
Fig.13.Maundiatriglochinoides.Outerintegument(testa)anatomyonlongitudinalsectionsoffertilizedovules
(lightmicroscopy,LM).A,Tangentialsectionofanovulethroughmiddlelayerofouterintegument.B,Detailof
(A).CE,Radialsectionsofovules.C,Levelofupperpartofendosperm.D,Levelofmiddlepartofendosperm.
E, Detail of outer epidermis of outer integument showing wide pores in anticlinal walls. F, Tangential section
throughouterepidermisofouterintegumentshowinganticlinalwallsincrosssection.Scalebars:100m(A);
20m(BF).en,endosperm;fn,funiculus;ii,innerintegument;lc,largecellsinpericarp;nc,nucellus;oi,outer
integument(testa);pc,pericarp;asterisks,aircanalsinouterintegument;arrowheads,poresinanticlinalwalls
ofouterepidermisofouterintegument(exotesta).
POSTANTHETICCHANGESINGYNOECIAANDOVULESOFTYPICALFLOWERS
Most, but not all flowers in an inflorescence develop fruits. In addition, a fruit may develop fewer
thanfourseeds.Thismaybeduetofailureofcrosspollinationiftheplantsareselfincompatible.
After pollination, the part of the gynoecium, where the carpels are united via the floral
centre (Figs 6D, 9EG), elongates considerably and becomes much longer than the free parts of the
carpels (Fig. 6F, G). In addition, growth is more extensive in the floral centre and in ventral parts of
thecarpels.Asaresult,thecarpelmouthsurroundedbyremainsofstigmaticpapillaeisonlyslightly
oblique at these late developmental stages (Fig. 6E) and carpels appear united via the floral centre
alongtheirentirelength.Inaveryshortbasalmostportionofthefruits,thecarpelflanksareunited,
andthefurrowsbetweenthecarpelsareabsent(Figs4G,6F).
Fruits are sessile on the inflorescence axis (Figs 1CE, 4G). Stamens and often also tepals
remain attached at the fruiting stage (Figs 1CE, 4G). They are much shorter than the fruits, which
aregreen,atleastatimmaturestages.Wehavenoevidenceoffruitdehiscence,mericarpseparation
121
Fig. 14. Maundia triglochinoides. Micropylar parts of ovules of postanthetic flowers in longitudinal sections
(lightmicroscopy,LM).A,Sectionthroughmicropyle.B,Sectionthroughmicropyleofanapparentlyunfertilized
ovule.C,D,Nonmedianlongitudinalsectionsshowingproembryoatglobularstage.Scalebars:100m(A);50
m (BD). en, endosperm; gc, gigantic basal cell of suspensor; ii, inner integument (tegmen); nc, nucellus; oi,
outerintegument(testa);pc,pericarp;pe,globularproembryo.
or specialized fruit abscission in our fixed material. Observations made on herbarium collections
stronglysupporttheindehiscentnatureofthefruitsofMaundia.
Incrosssection(Figs6D,9EG),thecarpellocules,whichareroundedinoutline,arecloseto
the flower centre. Seeds are almost as wide as the locules, thus filling them completely. Only
immature fruits were available, but further widening of locules is highly unlikely because of
histological differentiation of endocarp and mesocarp cells in the latest available stages. Free
peripheral (dorsal) parts of carpels are thick in a radial and especially in a tangential plane (Figs 6D,
9EG). Lateral peripheral sides of adjacent carpels are in close proximity to each other, whereas
adjacentcarpelsareseparatedbyaconsiderablespacenearthefloralcentre(Figs9F,10A).
The exocarp is onelayered (Fig. 10F) and composed of short thinwalled epidermal cells.
Stomataarepresentonthefruitsurface,atleastinthedistalpart.Guardcellsarebeanshaped.The
number and arrangement of epidermal cells surrounding the guard cells do not appear to be
precisely fixed. Most of the multilayered mesocarp is composed of mediumsized thinwalled cells
122
(Figs6D,9G,10).Inthemiddleandouterpartofthemesocarp,largesphericalorradiallyelongated
intercellular spaces are present (Figs 9G, H, 10CE). The dorsal vascular bundles are massive and
located close to the carpel locules (Figs 9F, 11I). On the left and right sides of the dorsal bundle,
groupsofverylarge(almostisodiametric)thinwalledcellsarepresent(Figs9G,10G,11I).Thesecells
arerecognizablealreadyinantheticflowers(Fig.5F).Theyhavealargenucleusandalargevacuole.
The large cells do not degenerate at least until the stages illustrated in Figs 6D, 9 and 10. In the
middle part of the fruit, each carpel locule is surrounded by several layers of fibres with relatively
thinlignifiedcellwalls.Fibresclosesttotheloculesareelongatedalongthefruitlength(Figs6H,10D,
11D).
We believe that only the innermost cell layer can be identified as an endocarp (i.e., a
derivative of the inner epidermis of carpels), because the deeper cell layers are not aligned to the
innermostcelllayerinawaythatcouldbeinterpretedasaresultofcelldivisionsinpericlinalplanes.
Therestofthecellselongatedalongthefruitlengththereforebelongtothemesocarp.Someother
mesocarp fibres are oriented transversally (Figs 6H, 10D, G, 11D, I). These are situated (1) between
the longitudinal fibres and dorsal bundles, extending along the inner margins of the groups of large
cellsand(2)intheperipheralpartofthemesocarpadjacenttothefurrowsdividingthecarpels.The
ventralbundles,whicharemuchsmallerthanthedorsalbundles,aresituatedattheperipheryofthe
floral centre adjacent to the sheath of fibres surrounding the locules. The floral centre is
parenchymatous(Fig.9E,F).Thesheathoffibresdoesnotcompletelyencircletheloculesalongtheir
fulllength.Gapsarepresentintheapicalpartofthelocules(Fig.10B)andontheleftandrightside
intheproximalpartofthelocules(Fig.11I).
Crystals, tanniferous cells, oil cells or other cell types with conspicuous content are absent
from the ovary wall or pericarp. Starch grains are abundant in parenchyma cells of the floral centre
(these cells occupy a crossshaped area on transverse sections of young fruits, with edges of the
cross alternating with ovary locules, Fig. 9G). Scattered starch grains are present in mesocarp cells
situatedintheperipheralpartsofthecarpels.
The ovules elongate considerably after fertilization, along with the elongation of the entire
portionofthegynoecium consistingof unitedcarpels(Fig.6G). Thefuniculus remainsmuchshorter
thanandaswideastheovule(Figs10B,12A,13A).Ithasacircularringofvascularbundlesthatdoes
not extend into integuments or the nucellus (Fig. 11A). The integuments are free from each other
andfromthenucellusthroughouttheirentirelength(Fig.12),circularincrosssection(Fig.11).The
outer integument consists of three to four cell layers and possesses a continuous outer and inner
epidermisandatissuewithabundantlacunaeinbetween(Figs11,13).Incrosssection,largelacunae
are separated by smaller cells (or radial cell pairs) linking the outer and the inner epidermis. The
lacunae(aswellasthecellsoftheintegument)areelongatedalongthelengthoftheovule.Cellsof
the outer epidermis of the outer integument possess characteristic thickenings on their radial walls
(Fig. 13C, E, F). These can be interpreted as secondary thickenings with numerous very large and
denselyspacedpores,eachextendingalongthewholedepthoftheradialwall.Theinnerintegument
istwolayeredwithalmostequalisodiametriccells.Lacunaeareabsentintheinnerintegument.The
two integuments are tightly appressed to each other (with cuticle in between), and the inner
integumentisappressedtothenucellus,alsowithcuticleinbetween.Themicropyleisformedbythe
innerintegument,whichismorethantwolayeredinthisregion(Fig.14A).Inthemicropylarregion,
cellsoftheinnerepidermisoftheinnerintegumentareelongatedinaradialdirectionandtheirwalls
arethickened(Fig.14A).
In cross sections through central parts of fertilized ovules with developing endosperm and
embryo,thenucellusformstwotofourcelllayerssurroundingtheendosperm(Fig.11DG).Nucellar
123
Fig.15.Maundiatriglochinoides.Flowerswiththreetepalsatinflorescencetip(scanningelectronmicroscopy,
SEM).A,Topviewofthreeflowersformingawhorlatinflorescencetip.Upperrightflower(alsoillustratedinB)
has three carpels, lower right flower has four carpels (adaxial one is smaller), left flower had damaged
gynoecium,whichwasthereforeremoved.B,Anotherviewoftheflowerfrom(A)withtricarpellategynoecium.
Scalebars:1mm(A);500m(B).c,carpel;gs,stalkofremovedgynoecium;is,innerwhorlstamen;os,outer
whorlstamen;te,transversalabaxialtepal;
te*,adaxialtepal.
cells are thinwalled and elongated along the length of the ovule. Cells of the outer epidermis are
larger than the rest of the nucellar cells. The innermost cell layers appear to degenerate during the
course of endosperm development (hence the difference in the number of cell layers observed in
different ovules). In the micropylar part of the ovule, the number of cell layers in the nucellus
increases, and the cells are isodiametric. A narrow conical nucellar beak is present, extending
towards the micropyle (Fig. 14A). In most ovules observed, the nucellar beak was short. In one
unfertilizedovule(inadevelopingfruitwithotherovulesfertilized),thenucellarbeakwasverylong
and protruding through the micropyle and widening at its distal side (Fig. 14B). Cells of the chalazal
part of the nucellus are almost isodiametric, except for nucellar epidermis cells. These cells are
uniform,thinwalledinyoungovulesaswellasinunfertilizedovulesindevelopingfruits(Fig.12F,G).
However, in fertilized ovules, the chalazal part of the nucellus undergoes a radical transformation
(Fig. 12AE). The cell walls separating the nucellar cells closest to the chalaza disappear, thus
resulting in the formation of a very large multinucleate coenocyte. The coenocyte contains a large
vacuole with numerous cytoplasmic strands extending in chalazalmicropylar direction (Fig. 12BE).
Nucleiofthecoenocyteappeartobefunctional.Atleasttheirnucleoliarestainedinthesamewayas
innucleiofnormalnucellarcells.Stagesofdegenerationofcellwallsbetweenadjacentnucellarcells
were documented (Fig. 12E, arrowhead). Several layers of unfused uninucleate nucellar cells are
presentbetweenthenucellarcoenocyteandtheendosperm.Thesecellsaremuchlargerthanthose
incontactwiththechalazalsideofthecoenocyte.
124
Fig.16.Floraldiagrams.AC,Maundiatriglochinoides.A,Normalflower.B,C,Flowersfrominflorescencetips.
D, E, Aponogeton subconjugatus. F, Scheuchzeria palustris L. G, Triglochin maritima L. H, Potamogeton sp. I,
Ruppia sp. (fourcarpellate flower). Black arches in (F, H), flowersubtending bracts; black dots, main
inflorescence axis. Organ arrangement (but not stamen and carpel shape) in (D, E) is based on Tomlinson
(1982).Gynoeciumoutlinein(G)isbasedonIgersheimetal.(2001).Stamenandtepalshapein(H)isinspired
bySingh(1965,Potamogetonindicus).
The globular stage of embryo development has a long and narrow suspensor ending in a
giganticbasalcellwithalargenucleus(Fig.14C,D).
125
FLOWERVARIATIONATTHEINFLORESCENCETIP
Twoinflorescencetipswereavailablefordetailedinvestigations.Inoneofthem(Fig.15),all
flowersuptotheinflorescencetipwerearrangedincleartrimerouswhorls.Inthisspecimen,flowers
of the uppermost whorl differed from typical flowers of Maundia (Fig. 16A) in the occurrence of
threetepals.Eachofthesethreetepalswasinsertedonaradiusofoneoftheinnerwhorlstamens,
so that two were transversalabaxial and the third was medianadaxial. Furthermore, one of the
three flowers of the uppermost whorl possessed three carpels (inserted on the radii of the outer
whorl stamens, Figs 15B, 16B), another had four carpels, and the gynoecium of the other was
damaged,makingitunsuitableforinvestigation.Inthefourcarpellategynoecium,theadaxialcarpel
was smaller than other carpels (Fig. 15A). Note that the adaxial carpel was missing from the
tricarpellategynoeciumofanotherfloweronthesameinflorescence.
In the other inflorescence that was investigated in detail, two distal lateral flowers were
attached at different levels of the inflorescence axis (below them, flowers were arranged in
trimerouswhorlstypicalofMaundia).Theinflorescenceaxiswasnotextendedabovethelevelofthe
uppermost flower. In the absence of flowersubtending bracts in Maundia, a morphologically
terminalpositionfortheuppermostflowercannotbecompletelyruledout.However,astheflower
wasturnedtowardsonesideoftheinflorescenceandnotdevelopmentallyaccelerated,wepreferto
interpretitasmorphologicallylateral.Inthisinflorescence,thedistalmostflower(Fig.16C)hasfour
carpels in positions typical of flowers of Maundia, four stamens and four tepals. Two stamens are
tetrasporangiate and dithecal (i.e., of a normal structure); these are in the medianadaxial and
medianabaxialpositions.Twootherstamensarebisporangiatemonothecal;thesearelocatedinthe
leftandrighttransversaladaxialpositions.Thefourtepalsareindiagonalpositions.Twoofthemare
asymmetricandassociatedwitheachofthemonothecalstamens(vasculartracesofastamenanda
tepal in both cases unite in the flower receptacle). Two other tepals are symmetrical and not
associatedwithanystamens.
Discussion
PEDUNCLEANATOMYANDOCCURRENCEOFSTOMATA
Maundia shares many features of peduncle anatomy with other aquatic and semiaquatic
membersofAlismatales,includingthepresenceofaerenchymawithaircanalsdividedintochambers
by transverse, minutely perforate septa, and tracheids rather than vessels in the metaxylem
(Tomlinson,1982).AprotoxylemlacunaispresentinthepedunclebundlesofmanyAlismatales,but
according to Tomlinson (1982) it is not conspicuous in most Juncaginaceae (except Cycnogeton
Endl.). Sclerenchyma forming a peripheral ring is present in peduncles of Juncaginaceae and
ScheuchzeriaceaebutabsentfromPotamogetonaceaeandapparentlyAponogetonaceae(Tomlinson,
1982). According to our data, the ring of sclerenchyma is absent from peduncles of Maundia. The
occurrence of peripheral bundles is recorded in stems and peduncles of many members of the
tepaloid clade of core Alismatales. An apparently unusual feature of Maundia is the inverted
orientationofperipheralbundles,withxylemorientedtowardstheperipheryofpeduncle.Tomlinson
(1982) did not record inverted peripheral bundles in peduncles of close phylogenetic relatives of
Maundia. We were able to reinvestigate the anatomical sections of Cycnogeton procerum (R. Br.)
Buchenau(Juncaginaceae)usedbyTomlinsonanddepositedintheKewmicroscopeslidecollection.
126
AlthoughTomlinsondidnotdescribeexplicitlytheorientationofperipheralbundlesinpedunclesof
Cycnogeton, we confirm that they are not inverted. We also reinvestigated the material used by
Tomlinson to describe the vegetative anatomy of Maundia. This is a set of sections of different
organs, including a root tuber, which has a peculiar type of clustered root hair arrangement. Since
Maundialacksroottubers,itispossiblethatthematerialismisidentified.Duetothelackofvoucher
specimens, it was impossible to verify the correct identity of the material. Therefore, the present
study probably provides the first detailed information on the vegetative anatomy of Maundia.
Buchenau(1903)presentedonlyaschematicillustrationofaleafcrosssection.Schneider&Carlquist
(1997)describedthevesselstructureinrootsofMaundia,butnootherdetailsofrootanatomywere
given.
StomataaretetracyticinScheuchzeria,paracyticortetracyticinJuncaginaceaeandparacytic
(when present) in Potamogetonaceae (Tomlinson, 1982). In Aponogeton, stomata are described as
each having a pair of indistinct lateral subsidiary cells; one or both subsidiary cells are commonly
segmentedbyamediananticlinaldivision,thesubsidiarycellsbecomingobscure(Tomlinson,1982).
We found no data on the stomata of Maundia in the literature. The stomata documented in this
study on tepals and fruits of Maundia possess several small cells surrounding the guard cells.
Althoughdevelopmentaldataareclearlyneededforbothtaxa,thearrangementofsurroundingcells
is similar in Maundia and Aponogeton and cannot be readily described as either paracytic or
tetracytic. Stomata on peduncles of Maundia are more similar to those described for leaves of
Juncaginaceae(Tomlinson,1982).
In Alismatales, stomata are present on carpels of Araceae, some Alismataceae,
Aponogetonaceae,Butomaceae,Juncaginaceae,LimnocharitaceaeandScheuchzeriaceae(Igersheim,
Buzgo & Endress, 2001). The present study revealed stomata on carpels of Maundia. In contrast,
stomata are absent from carpels in a clade that is sister to Maundia (Potamogetonaceae, Zostera
ceae,Ruppiaceae,Posidoniaceae,Cymodoceaceae).Thisdifferenceisperhapsnotsurprising,asmost
ofthesetaxaflowerunderwater.However,flowersofsomespeciesofPotamogetonL.areexposed
abovethewater,whilesubmergedflowersarefoundinotherspeciesofthegenus(Philbrick1988).
MORPHOLOGICALINTERPRETATIONOFFLORALPARTS
Our data confirm earlier observations that the flowers of Maundia, except the uppermost
ones, uniformly possess two scalelike structures interpreted here as tepals. These phyllomes are
inserted in the transversalabaxial position. Three different interpretations of these scalelike
structureshavebeenproposedintheliterature,namelyasbracts(Uhl,1947;Aston,2011),perianth
members(tepals:Bentham,1878;Buchenau,1903;Nakai,1943;Haynes,Les&HolmNielsen,1998;
Buzgoetal.,2006;Takhtajan,2009orsepals:Mueller,1858)andconnectiveappendages(Markgraf,
1936; Eckardt, 1964; Dahlgren, Clifford & Yeo, 1985). The same range of interpretations has been
proposed for scalelike structures in reproductive structures of other members of core Alismatales
(Helobiae),suchasTriglochins.l.andPotamogeton(seeKunth,1841;Ascherson,1889;Buchenau&
Hieronymus,1889;Markgraf,1936;Miki,1937;Uhl,1947;Eames,1961;Eckardt,1964;Sattler,1965;
Singh, 1965; Posluszny & Sattler, 1973, 1974; Burger, 1977; Lieu, 1979; Dahlgren, Clifford & Yeo,
1985;Posluszny&Charlton,1993;Endress,1995;Mavrodiev&Sokoloff,1998;Posluszny,Charlton&
Les, 2000; Rudall, 2003; Buzgo et al., 2006). The bract interpretation is typically used in the
frameworkofthehypothesisthatreproductivestructurescommonlytermedflowersinthisgroupin
fact represent compact inflorescences (pseudanthia) composed of naked unisexual flowers. This
hypothesis implies that each scalelike structure represents a flowersubtending bract of a male
127
flower, and what is traditionally termed a stamen actually represents an entire male flower (e.g.,
Kunth,1841;Miki,1937;Uhl,1947;Eames,1961).Apseudanthialinterpretationoftheconventional
flowerofcoreAlismataleshasbeencriticisedonthegroundsthatthefeaturesusedinsupportofthis
theory can also be found in some monocots belonging to other monocot orders (summarized in
Endress, 1995; see also Lieu, 1979 and Buzgo et al., 2006, for case studies in Juncaginaceae). We
acceptthiscriticismandfollowtheeuanthialinterpretationforflowersofmostAlismatales(probably
excluding Zannichelliaceae and Cymodoceaceae: Sokoloff, Rudall & Remizowa, 2006; Remizowa et
al.,2012b).
The euanthial interpretation implies that structures traditionally called flowers in taxa such
asPotamogeton(Fig.16H),Triglochin(Fig.16G),andScheuchzeria(Fig.16F)arehomologouswiththe
uniaxial flowers of other angiosperms. In the framework of this view, it is clear that a euanthial
interpretation can also be adopted for Maundia. If the flower of Maundia does not represent a
pseudanthium,itishighlyunlikelythatthetwoscalelikestructuresarebracts.Indeed,theirposition
excludesthepossibilitythattheseareflowersubtendingbracts.Forthereasonsoutlinedbelow,we
reject the interpretation of the two scalelike structures as prophylls (bracteoles) or connective
appendages,andinsteadinterpretthemastepals.
One could argue that the two scalelike structures represent prophylls (bracteoles) on the
pedicel.However,bracteolesareunknowninanyothermemberofthelargecladetowhichMaundia
belongs (Aponogetonaceae plus its sister clade, a group called tepaloid alismatids by Posluszny &
Charlton, 1993). Moreover, the hypothesis implies that bracteoles are present when a flower
subtending bract is absent. As pointed out by Remizowa et al. (2013), such a combination of
charactersisnotobservedinotherearlydivergentmonocots.
Like the pseudanthial concept, a connective appendage interpretation for the scalelike
structuresofflowersofJuncaginaceaes.l.and/orPotamogetonaceaes.l.,isbasedonvariouskindsof
associationbetweenascaleandastamenoccurringonthesameradiusinaflower(Ascherson,1889;
Markgraf, 1936; Eckardt, 1964). This association can be manifested in a common vascular supply,
and/orbasalfusionofastamenandascale.Furthermore,asinTriglochin,thescalesassociatedwith
innerwhorlstamenscanbeinsertedabovetheouterwhorlstamensinthetwowhorledandroecium
(e.g., Goebel, 1928; Uhl, 1947; Rudall, 2003; Remizowa, Sokoloff & Rudall, 2010). The connective
appendage concept has been much criticised based on evidence from flower development and
comparative morphology. In particular, the scalelike structures and stamens appear separately
during flower development in both Potamogeton (Hegelmaier, 1870; Sattler, 1965; Posluszny &
Sattler, 1973, 1974; Posluszny, 1981; Sun, Zhang & Chen, 2000; Nunes et al., 2012) and Triglochin
(Lieu, 1979; Bugzo et al., 2006; Remizowa, Sokoloff & Rudall, 2013). The significance of
developmental data should not be overestimated, because different thecae of the same anther
occasionallyappearseparatelyonthefloralapex(Posluszny&Sattler,1973).NotethatEichler(1875)
did not abandon the connective appendage interpretation for Potamogeton despite the excellent
developmentalstudyofHegelmaier(1870).Itismuchmoreimportantthattheintimaterelationships
betweentepalsandstamensoccurringonthesameradiicanbefoundinawiderangeofmonocots
belongingtodifferentordersandthesecouldmerelyreflectthepronouncedsectorialdifferentiation
in the flowers (Endress, 1995; Remizowa, Sokoloff & Rudall, 2010; Remizowa et al., 2012a). As
pointedoutbyEndress(1995),thepronouncedassociationbetweentepalsandstamensinsertedon
thesameradiiismorelikelytoappearintrimerousthaninpentamerousflowers.
Once we accepted the presence of a perianth in Triglochin and Potamogeton, we see no
argumentagainstacceptingthescalelikestructuresinMaundiaastepals.Liketepalsofrelatedtaxa,
thoseofMaundiaareinsertedontheradiiofstamens.Ourdataonfloralvasculaturedonotshowa
128
substantialdifferencebetweenvascularsupplyoftepalsinMaundiaandsomePotamogetonspecies
(Uhl,1947).
Forthepurposesofevolutionarymorphologyandtaxonomy,themostimportantconclusion
is that scalelike structures of Triglochin, Scheuchzeria, Potamogeton and Maundia are homologous
to each other. When flowers of Triglochin are interpreted as lacking bracts and having tepals but
flowers of Maundia as having bracts but lacking tepals (Aston, 2011), such an interpretation
artificially increases the degree of morphological difference between the two taxa. These problems
shouldbeconsideredincharacterscoringformorphologicalcladisticanalyses.
There are two contrasting interpretations of stamen morphology in Maundia. In one
interpretation,thetypicalflowerhasupto12monothecalbisporangiatestamens(Mueller,1858).In
another interpretation, the flower has tetrasporangiate dithecal stamens, with each stamen being
split to its very base (most authors: e.g., Bentham, 1878; Buchenau 1903; Markgraf, 1936; Aston,
2011). We follow the second interpretation, because it creates a much smaller gap between
Maundia and related taxa. With the exception of obviously reduced and highly transformed taxa
suchasZannichelliaceaeandCymodoceaceae(whereweacceptapseudanthialinterpretationofthe
reproductivestructures,seeabove),allthecloserelativesofMaundiapossesstetrasporangiateand
dithecalstamens.AlthoughthetwothecaeofMaundiastamensarelargelyfree,theyremainbasally
united.TheoccurrenceofindividualvascularbundlessupplyingeachthecainMaundiabynomeans
supports the assignment of each theca to an individual monothecal stamen. Although onetraced
stamenswithasingleunbranchedvascularbundlearemostcommonamongangiosperms,taxawith
multibundled and even multitraced stamens are known from all major phylogenetic lineages,
includingmonocots(e.g.,Remizowaetal.,2011),eudicots(e.g.,Nuraliev,Sokoloff&Oskolski,2011)
and magnoliids (e.g., Canright, 1952). Patterns of variation in stamen supply described here for
MaundiaareidenticaltothosedescribedbyUhl(1947)inPotamogeton.
RELATIVEPOSITIONSOFFLORALPARTSINMAUNDIA:ANINTRIGUINGSIMILARITYTOAPONOGETON
Detailed observations on the flower groundplan in Maundia were published by Markgraf

(1936).Hefoundthatmostflowers(exceptattheverytipoftheinflorescence)possesssixstamens
andfourcarpels.Heinterpreted theflowerastetramerous,withtwoinnerwhorlstamenslost.The
outerwhorlstamensareinmedianandtransversalpositions.Thesestamensarenotassociatedwith
scalelike structures. The two stamens of the inner whorl are in positions intermediate between
abaxialandtransversal.Markgrafimpliedthattheinnerwhorloftheandroeciumalternateswiththe
outerwhorl, andtheinnerwhorlstamensarethereforeindiagonalpositions,andthe twostamens
of the adaxial part of the flower are absent. The two innerwhorl stamens that are present are
associated with scalelike structures, which Markgraf interpreted as connective appendages.
However,accordingtoourinterpretation,thesescalelikestructuresaretepals,andthistermwillbe
used below. In the uppermost part of the inflorescence, Markgraf found flowers with four tepals.
Twoadditionaltepalswereinserted in positionsintermediate betweenadaxialandtransversal.This
observationwasusedbyMarkgrafinsupportofthetetramerousinterpretationoftheentireflower.
The four carpels, according to Markgraf, are inserted in one whorl in diagonal positions. This
orientationofcarpelscreatedaproblemwithalternationofwhorlsintheflower.Indeed,thecarpels
andtheinnerwhorlstamensareinsertedonthesameradii.Markgrafpostulatedthatthegynoecium
of Maundia was originally twowhorled (as in other Juncaginaceae), and the outerwhorl carpels
were lost during evolution, but their positional information is retained. This interpretation was
129
supportedbythesterilityoftheouterwhorlcarpelsinmostspeciesofTriglochin.Insomespeciesof
Triglochin,theouterwhorlcarpelsaremuchreduced.
Diagrams of Markgraf (1936) were reproduced (with reference) by Uhl (1947) and Eckardt
(1964), who apparently did not study material of Maundia in detail themselves. However, both
reproduced diagrams differ from the original, and in both cases no comments regarding these
differences are provided. Uhl (1947) changed the carpel position from diagonal to median and
transversal(i.e.,ontheradiioftheouterwhorlstamens).Eckardt(1964)illustratedeightstamensin
twowhorlsandfourtepalsassociatedwithinnerwhorlstamens,inwhathecalledanormalflower.
Our data show an arrangement of organs that differs from all diagrams discussed above. In
our material, carpel orientation was never diagonal (Fig. 16A, C). The four carpels were always
insertedinmedianandtransversepositions,asillustratedbyUhl(1947).Thisarrangementwasalso
the case of the uppermost flowers in the inflorescence, if these are fourcarpellate. In our
interpretation, the androecium is twowhorled and trimerous. The tepals are associated with the
anterior innerwhorl stamens. We lacked sufficient material to investigate the variation in
inflorescence tip morphology in Maundia, because of the rare nature of the plant and different
preservationofinflorescencetipsinourmaterial.However,noneoftheflowersfromtheupperpart
of inflorescence showed clear evidence of tetramery, except in the gynoecium. Moreover, in the
inflorescence with three distalmost flowers forming a whorl, each flower possessed a trimerous
whorl of tepals and two trimerous whorls of stamens. At least in one case, a completely trimerous
flowerwasobservedattheinflorescencetip,wherethegynoeciumwasalsotrimerous(Fig.16B).
Severalauthors,startingwithMueller(1858),indicatetheoccurrenceoftwotofourperianth
membersinMaundia(Thompson,1961;Haynes,Les&HolmNielsen,1998).However,noneofthem
discuss the position of flowers with different perianth morphology within inflorescences. According
to our data, variation in tepal number occurs only in the final whorl of flowers. Except at the
inflorescence tip, all observed flowers consistently possessed two tepals (see also Aston, 2011).
Apparently,therecordsofvariationbetweentwoandfourtepals,alongwiththecommonpresence
offourcarpels,werethesourceofinterpretationofMaundiaflowersasdimerous(Dahlgren,Clifford
&Yeo,1985;Haynes,Les&HolmNielsen,1998).Ourdatadonotsupportadimerousinterpretation.
Data on carpel number in Maundia also differ in various publications. While Mueller (1858)
indicated three to four carpels and Bentham (1878) noted carpels usually 2 or 3, sometimes 4,
Aston (2011) stated that there are usually four, rarely two or three carpels. Unfortunately, these
authors did not consider flower position in the inflorescence. We did not find a bicarpellate
gynoecium.Ouranalysisoffixedmaterialandherbariumcollectionsshowedthatthefourcarpellate
conditionisthetypicaloneinMaundia(seealsoThompson,1961).
ThefloraldiagramofMaundiaisalmostidenticaltothatofmostspeciesofAponogeton(see
Buzgoetal.,2006).InAponogeton,asinMaundia,flowersarearrangedinspikesandlackanysigns
of flowersubtending bracts. With some exceptions (including A. distachyus L.f. and an early
divergent species A. hexatepalus H.Bruggen Les, Moody & Jacobs, 2005), flowers of Aponogeton
typicallypossesstwotepalsintransversalabaxialpositions,sixstamensintwowhorls(outermedian
stamen abaxial) and three, sometimes four carpels (Singh & Sattler, 1977; Tomlinson, 1982;
Remizowa, Sokoloff & Rudall, 2010). When four carpels are present in Aponogeton, their position is
the same as in Maundia (Tomlinson, 1982). When three carpels were present in our material of
Maundia, their position isthe same as in Aponogeton. In both Maundia and Aponogeton, there is a
tendencytodevelopthethird(adaxial)tepalintheuppermostflowersofaninflorescence.
ThesimilarityinflowergroundplanbetweenMaundiaandAponogetonisintriguingbecause
(1) this is a highly unusual flower organization apparently not found in other monocots and (2) the
130
two genera are not sister taxa (Iles, Smith & Graham, 2013; Les & Tippery, 2013). There are other
examplesinbiologicalevolutionwhensimilarandunusualnoveltiesevolvedindependentlyinclosely
related but not sister lineages. For example, the order Alismatales is the only group of seed plants
thatincludesmarinetaxa,thesocalledseagrasses.Inthreefamiliesofseagrasses,thepollengrains
arefiliform,aconditionuniqueamongangiosperms.Analysisofthischaracterbasedonamolecular
phylogenetic treetopology(Iles,Smith&Graham,2013)andadditionalstructuralevidencestrongly
suggest three independent origins of filiform pollen in Alismatales and more than one shift from
continental aquatic to marine habitats (Remizowa et al., 2012). In the case of parallel evolution of
filiformpollen,asinsomeotherexamples,independentappearanceofanunusualcharacter(which
could be described as a homoplastic tendency: Sanderson, 1991) has an obvious adaptive signify
cance.
In contrast, there is no obvious adaptive significance for the occurrence of two abaxial
transversal tepals in Maundia and Aponogeton. The two genera differ in pollination ecology.
Aponogeton is insectpollinated, with tepals being white, pink(ish)/purple, yellow to green and
attractive in many species, while Maundia is apparently windpollinated (though detailed
observations are needed), with tepals being green and nonattractive. We suggest that
developmental and spatial constraints rather than functional significance are responsible for the
homoplasticappearanceofsimilarfloraltypesinMaundiaandAponogeton.Manymembersofcore
Alismatales have spikes or racemes with whorled arrangement of flowers. Whorled flower
arrangement is otherwise extremely rare in monocots (e.g., Sokoloff et al., 2009; Remizowa et al.,
2013). Flowers are closely spaced in developing racemose inflorescences of Alismatales. In the
absence of flowersubtending bracts (another feature common in Alismatales but otherwise rare in
monocots Remizowa, Sokoloff & Rudall, 2013; Remizowa et al., 2013), physiological interactions
betweenadjacentsitesofflowerinitiationmustbeimportantforprepatterningoffloralorgans.We
arguethatonlyalimited numberofflowerorganizationsareavailablethatallowthemostcompact
spacing of developing flowers and the most complete use of space on the surface of the
inflorescenceaxis.TheconditionfoundinMaundiaandAponogetonisoneofthem.Inthisrespect,it
is tempting to consider the wellknown similarity between flowers of taxonomically unrelated
Potamogeton and some Cyclanthaceae and Pandanaceae (e.g., Miki, 1937) as reflecting similar
spatial constraints. Apart from Alismatales, Cyclanthaceae is another aberrant monocot group with
whorledflowerarrangement.
POLLEN
Our data confirm information on Maundia pollen morphology provided by the Australasian
PollenandSporeAtlas(APSAMembers,2007).PollenmorphologyofMaundiaisverysimilartothat
of related members of Alismatales (Grayum, 1992; Furness & Banks, 2010). As stated by Grayum
(1992)thegenera[Triglochin,Lilaea,Tetroncium]arequiteuniformpalynologically,andhardlytobe
distinguished on this basis from Potamogeton. The same applies to Maundia pollen and further
studies,e.g.onpollenwallultrastructuremightrevealinformativecharacters.Monosulcatepollenof
Aponogeton is different from that of taxa listed above (Grayum, 1992; Furness & Banks, 2010).
Aponogeton also differs from other core Alismatales (no data on Maundia) in possessing
simultaneousratherthansuccessivetypeofmicrosporogenesis(Furness&Banks,2010).
131
OVULESANDSEEDS
Ourstudyfullysupportstheoccurrenceofpendent,orthotropousovulesinMaundia.Toour
knowledge,thisfeaturewasquestionedonlybyBentham(1878)whothoughtthatMuellers(1858)
original description was incorrect, which is not the case. We provide the first detailed observations
on the histology of the integuments and the nucellus in Maundia. According to our data, the outer
integument consists of three to four cell layers and contains conspicuous intercellular canals
(apparently air canals) aligned along the ovule length. These canals are conspicuous in fertilized
ovules with developing embryo. ShafferFehre (1987) studied the mature seeds of Maundia and
found no evidence of a mesotesta and no air canals. She also noticed an obliteration of exotesta
cells. We hypothesize that the mesotesta obliterates along with the exotesta, making the air canals
inconspicuousinmatureseeds.
We assume that the presence of air canals in the testa is of phylogenetic significance in
Alismatales. According to literature, conspicuous air spaces are present in the testa of Butomus L.,
Enhalus acoroides (L.f.) Royle (Hydrocharitaceae) and Aponogeton (Melikian, 1985; Plisko, 1985;
Teryokhin,1985).Wefoundnopublishedevidencefortheoccurrenceofsimilarintercellularspaces
in other Alismatales, though more detailed observations in wide range of taxa in various
developmentalstagesarecertainlynecessary.
Accordingtoavailablepublications,theformationofacoenocyticstructureinthenucellusof
fertilized ovules of Maundia has no exact parallels among other angiosperms. In particular, it has
nothing in common with various kinds of specialized structures described in the chalazal part of
ovulesofvariousangiosperms,suchasahypostase,podium,postamentetc.(reviewedbyShamrov,
2008,seealsoRudall,1997).Inmanyangiosperms,nucellarcellsdegenerateduringendospermand
embryodevelopment,butcelldegenerationtypicallytakesplaceinareasofthenucellusthatareina
directcontactwithendospermandembryo.InMaundia,reorganizationofthenucellustakesplacein
its chalazalmost part, and normal tissue of cells with cell walls remain between the nucellar
coenocyteandtheendosperm.Inaddition,formationofthecoenocytedoesnotcausecelldeathin
Maundia,atleastduringthedevelopmentalstagesavailableforthepresentstudy.Thenucleiofthe
coenocytedonotappeartobedegenerating.
Amongangiosperms,lossofcellwallsbetweencellsintissueofthenucellusresultinginthe
formationofacoenocyticstructureisknownintheeudicotfamilyPodostemaceae(Went,1908;Razi,
1949; JgerZrn, 1967; 1997; Nagendran, Arekal & Subramanyam, 1977; Nagendran, Anand &
Arekal, 1980). Here, the nucellar coenocyte (also known as nucellar plasmodium) serves as a
structure that substitutes an endosperm, which is missing in Podostemaceae. So the functional
significanceofthisfeatureinPodostemaceaeisobvious.Wehavenoplausibleinterpretationofthe
functionalroleofthenucellarcoenocyteinMaundia.Itmayplayaroleintransferringnutrientsfrom
the funiculus to the nucellus. Maundia appears to have an unusually welldeveloped vasculature in
the funiculus. Instead of a single bundle, a ring of vascular bundles is present. Rudall (1997)
highlightedthesimilaritybetweentheformationofthenucellarplasmodiuminPodostemaceaeand
the formation of the coenocytic structure in the monocot Pandanus (Pandanales), in which diploid
nuclei of nucellar cells penetrate into the embryo sac where they further divide (Cheah & Stone,
1975).TheformationofmultinucleatestructuresinbothPandanus(Chubirko,1990;Kamelina,2011)
andMaundiarequiresfurtherinvestigation.Kamelina(2011)suggestedthatthecoenocyticstructure
in Pandanus containing up to 200 diploid nuclei could form due to apomictic endosperm
development.
132
FRUITS
Cronquist(1981)statedthat,asinTriglochinandTetroncium,thematurecarpelsofMaundia
separate from the persistent central axis and open ventrally. Our data do not support this
observation. Although we did not observe fully mature fixed fruits, their anatomical structure
excludes the possibility of regular carpel separation from a persistent central column (as in many
Juncaginaceae)andventralcarpeldehiscence.ThefloralcentreofMaundiahasnomechanicaltissue
that should be expected in a persistent column. The sclerenchymatous layer of each carpel is
continuous on the ventral side, making ventral dehiscence impossible. Our observations of fully
maturefruitsinherbariummaterialalsodidnotrevealcarpelseparationanddehiscence.According
to Bobrov et al. (2009), fruits of Maundia are schizocarps with indehiscent mericarps (regma
syncarpia), but we did not record separation of mericarps in studied fixed and herbarium material.
Our data are congruent with the earlier description of Bentham (1878), noting: the carpels almost
drupaceous, each with a tiny cartilaginous endocarp with an acute dorsal rib. One of specimens of
Maundia (Briggs 10003) includes the note "Individual carpels tending to separate in fallen fruits".
However,inalmostallfruitsthatdroppedoffduringthedryingprocessallcarpelsremainedunited.
Those carpels that have separated each possess a segment of the floral centre adhering to the
ventral surface (B.G. Briggs, Royal Botanic Gardens, Sydney, pers. comm.). This observation agrees
with Aston's (2011) description of fruiting carpels remaining united and falling together or tardily
separatingatmaturity,andindehiscent.
Tosummarize,intheabsenceofregularcarpelseparationbeforefruitdetachmentfromthe
maternal plant, fruits of Maundia do not belong to the carpological type that is characteristic of
Triglochin.WewouldliketonoteinthiscontextthatfruitdiversityinJuncaginaceaerequiresfurther
investigation. In Triglochin maritima L., as in many other species, oneseeded parts (fruitlets or
mericarps, depending on the interpretation of the gynoecium) separate from a persistent stalklike
structure (called carpophore or column) along their entire length. In T. plaustris L., the oneseeded
parts remain united with the stalklike structure only at their distalmost parts. According to
observations in NW Russia (M.V. Remizowa, unpubl. data), the fruits of T. palustris do not
disintegrate completely in the year of their formation. They remain attached to persistent upright
inflorescence axes until the next season when individual oneseeded parts ultimately separate.
Finally,intherecentlydescribedspeciesT.buchenauiKcke,Mering&Kadereit(Kckeetal.,2010),
a carpophore (column) is absent, and oneseeded parts separate each possessing a segment of the
floralcentre,asinoccasionalinstancesofcarpelseparationinMaundia.
The absence of a detailed survey of pericarp histology in Juncaginaceae does not allow
conductingacomprehensivecomparisonwithMaundia.Anillustrateddescriptionoffruitanatomyis
available for Triglochin palustris (Petrova, 1985). In addition, preliminary data on some African
TriglochinspeciesarereportedbyLocketal.(2011).AsharedfeatureofMaundiaandT.palustrisis
the occurrence of mechanical tissue in the inner part of the pericarp, surrounding the fruit locules
(however,mechanicaltissueisalsopresentintheinnerpartofthepericarpinmanyrelatedgroups
suchasPotamogetonaceaeandRuppiaceaeTeryokhin,1985).Itisnotclearfromthedescriptionin
Petrova (1985) whether these cells are isodiametric or elongated along the length of the fruit and
whether all layers of the mechanical tissue belong to the endocarp (as interpreted in Bobrov et al.,
2009).AccordingtoLocketal.(2011),asinMaundia,mesocarpcontributestothemechanicaltissue
surroundingtheloculesinAfricanspecies.Differencesinclude(1)thepresenceofairspacesbetween
theseedandthepericarpinTriglochin,(2)thepresenceoftransversallyelongatedfibresinMaundia,
133
(3)thepresenceoflargecellsflankingthedorsalbundleinMaundia,and(4)muchgreaterwidthof
thepericarpinMaundia.
Ofspecialinterestisthepresenceofintercarpellaryfusioninaveryshortproximalportionof
the fruits in Maundia. This fusion is not detectable at anthesis, apparently because the basalmost
part of the gynoecium forms after fertilization. This phenomenon is significant for understanding
gynoecium evolution in Alismatales, if the fusion of carpel flanks is accepted (other interpretations
wouldbeacceptingtheoccurrenceofawidenedfruitstalkorunitedcarpelstipes,asinHarperocallis
McDaniel in Tofieldiaceae Remizowa et al., 2011). Most members of the tepaloid clade of
Alismatales possess either free carpels or carpels united via the floral centre (Igersheim, Buzgo &
Endress, 2001; Remizowa, Sokoloff & Rudall, 2010). The only wellknown exception is Scheuchzeria,
where carpels form a conspicuous unilocular symplicate zone (see Eber, 1934; Igersheim, Buzgo &
Endress,2001).Carpelfusionviathefloralcentreisinmanyrespectsrelatedtoapocarpy,andboth
conditions are currently considered derived character states in monocots (Endress & Doyle, 2009;
Remizowa,Sokoloff&Rudall,2010).TheconditionfoundinScheuchzeriaislikelyplesiomorphic.The
shortfusionbetweencarpelflanksinthebasalmostfruitregionofMaundiacouldbeconsideredas
arudimentofanancestrallysyncarpousgynoeciumconstruction.Inthisrespect,weemphasizethat
a reinvestigation of gynoecium and fruit structure in Tetroncium, an earlydivergent member of
Juncaginaceae, is urgently needed (see also Thieret, 1988). As long as Tetroncium is placed in
Juncaginaceae and as its flowers (though dioecious and dimerous) are generally similar to those of
TriglochinandCycnogeton,itmightbelogicaltosupposethatcarpelsareunitedviathefloralcentre
inTetronciumasinotherJuncaginaceae.Surprisingly,asillustratedbyHooker(1847),Tetronciumhas
carpelsunitedtoformaunilocularovarywithincompletesepta.IfHookersdataarecorrect,thenin
terms of the gynoecium morphology Tetroncium is closer to Scheuchzeria than to Triglochin,
Cycnogeton(andMaundia).
TAXONOMICANDEVOLUTIONARYIMPLICATIONS
Ourresultsshowthatmorphologydoesnotcontradictthemolecularphylogeneticplacement
of Maundia as sister to a group comprising members of Potamogetonaceae, Zosteraceae,
Ruppiaceae,CymodoceaceaeandPosidoniaceae.AspointedoutbyvonMeringandKadereit(2010),
thepresenceofpendent orthotropousovulesisan obvioussynapomorphyofthislineage, including
Maundia. Most members of the large clade that is sister to Maundia have underwater pollination
and thus their flowers are quite reduced compared with those in Maundia and other Alismatales
with emergent flowers. Less reduced flowers are characteristic of Potamogeton, where, at least in
most species, pollination takes place above the water level. Therefore, it is most appropriate to
compare the floral morphology of Maundia and Potamogeton. Tepal morphology is similar; in both
genera,tepalsareclawed,green,withabundantstomata,singletraced(probablyrarelythreetraced
in Maundia), but the tepal bundle branches in the tepal blade (Hegelmaier, 1870; Uhl, 1947; Singh,
1965; Sattler, 1965). The latter character is of particular importance, because branching tepal
bundlesarenotuniversallypresentinAlismatales(e.g.,unbranchedinTriglochinUhl,1947).Some
patternsofinfraspecificvariationintepalvasculatureinMaundia(thisstudy)resemblethosepresent
inPotamogeton(Hegelmaier,1870;Sattler,1965),suchasthepresencevs.absenceofanastomoses
between bundles. Details of stamen vascular supply (and relationships between stamen and tepal
traces)aresimilarinMaundiaandatleastsomespeciesofPotamogeton(Uhl,1947).Theoccurrence
of four carpels, however, cannot be viewed as a similarity between Maundia and Potamogeton,
because those of Maundia are in median and transversal positions while those of Potamogeton are
134
indiagonalpositions.Intermsofcarpelposition, Maundia couldbe comparedwithTetronciumand
thefourcarpellateflowersofRuppiaL.InRuppia(Fig.16I)andapparentlyinTetroncium,carpelsare
arranged in dimerous whorls. Developmental data are required to understand whether the
gynoeciumofMaundiaisformedbyonetetramerousortwodimerouswhorls.StamensofMaundia
withalmostfreethecaearesimilartothoseofRuppia.
AlthoughthereareimportantsimilaritiesbetweenMaundiaandmembersofitssistergroup,
some morphological characters highlight resemblances with other tepaloid Alismatales. In general
habitaswellasintheoccurrenceofstrongcarpelfusionviathefloralcentre,Maundiastillresembles
Juncaginaceae, the family where the genus has been traditionally placed. In flower groundplan,
MaundiaissimilartoAponogeton,asimilaritythatshouldbetakenintoaccountbecauseoftherarity
and unusual nature of this flower organisation. The occurrence of air canals in the testa is another
sharedfeatureofMaundiaandAponogeton(thoughitisalsopresentinsomemoredistantlyrelated
Alismatales,seeabove).SimilaritiesbetweenMaundiaandAponogetonhavebeennotedasfarback
asbyMueller(1858).Inahandwrittenannotationtooneofhisherbariumspecimens(F.Muellers.n.,
K000098531), Mueller stated: Stigma and sepals like Aponogeton Fruit and anthers like
Triglochin.
Furthermore, Maundia possesses several features that appear to be unique or rare in
Alismatales,includingtheoverallfruitstructure(carpelsunitedviaflowercentre,butspecializedfruit
fragmentationabsent;spongyouterandstonyinnerpericarplayers,largecellsinpericarp),inverted
peripheral bundles in peduncle and the enigmatic formation of the nucellar coenocyte during
embryodevelopment.
As Maundia exhibits a mosaic of features characteristic of other families of tepaloid core
Alismatales (Table 1), and taking into account its isolated phylogenetic placement, we prefer to
segregate it in a family of its own, Maundiaceae. As pointed out by APG III (2009) and discussed by
Iles, Smith & Graham (2013), an alternative to the recognition of another monogeneric family in
Alismataleswouldbetocreateanexpanded(buthighlyheterogeneous)familyforthelargerclade.If
this alternative is followed, we suggest accepting only four families in the order Alismatales: (1)
Tofieldiaceae,(2)Araceae,(3)afamilycomprisingmembersofAlismataceae(incl.Limnocharitaceae),
Butomaceae and Hydrocharitaceae, i.e., the petaloid clade of core Alismatales (Fig. 17A) and (4) a
family comprising members of Aponogetonaceae, Scheuchzeriaceae, Juncaginaceae, Maundiaceae,
Potamogetonaceae (incl. Zannichelliaceae), Zosteraceae, Cymodoceaceae, Ruppiaceae and
Posidoniaceae, i.e., the tepaloid clade of core Alismatales (Fig. 17A). This broad concept would
accommodate similarities between Aponogetonaceae and Maundiaceae as well as those between
Scheuchzeriaceae and Juncaginaceae. Figure 17A shows some potential morphological
synapomorphies (and autapomorphies) of various clades in a molecular phylogenetic tree of core
Alismatales. Analyzing data in Fig. 17A reveals one of the most significant problems of the broad
family concept. Namely, the petaloid clade is marked by several potential synapomorphies, while
assessing clear and unambiguous synapomorphies for the tepaloid clade is more problematic (but
seeStevens,2001+).
Table1.OverviewofselectedcharactersinMaundiaandsomeothermembersofthetepaloidcladeofcoreAlismatales.Thefollowingmajorsourceswere
used:Uhl(1947);Gibbs(1974);Goldblatt&Johnson(1979+);Tomlinson(1982);Plisko(1985);Petrova(1985);Teryokhin(1985);ShafferFehre(1987);
Takhtajan(2009);Igersheimetal.(2001);Lesetal.(2005);vonMering&Kadereit(2010);Furness&Banks(2010);Hellquist&Jacobs(2011);Kamelina(2011).
Character
Aponogeton
(Aponogetonaceae)
Scheuchzeria
(Scheuchzeriaceae)
Cycnogeton
(Juncaginaceae)
Triglochin
(Juncaginaceae)
Tetroncium
(Juncaginaceae)
Maundia
(Maundiaceae)
Potamogeton
(Potamogetonaceae)
Ruppia
(Ruppiaceae)
Foliageleavesonelongateshoots
emergingabovesubstratepresent(P)
orabsent(A)
A P A A A A P P
Leafintravaginalsquamules
transformedintohairs(Yes/No)
N Y ? N ? ? N N
Foliageleavesligulate(L)oreligulate
(E)
E L E L E E L E
Leavesdifferentiatedintopetioleand
lamina(Yes/No)
YorN N N N N N YorN N
Leafapicalporeabsent(A)orpresent
(P)
AorP P ? A ? ? P A
Sclerenchymatouscylinderin
inflorescencepeduncles(flowering
stems)absent(A)orpresent(P)
?A P P P ? A A A
Invertedbundlesininflorescence
peduncles(floweringstems)present
(P)orabsent(A)
A A A A ? P A A
*Crystalsinvegetativeorganspresent
(P)orabsent(A)
P P A A A A A A
**Laticifers(P)orabsent(A) P A ? P ? A(?) A A
Spatheenclosingyounginflorescence
present(P)orabsent(A)
P A A A A A A A
Flowersubtendingbractsabsent(A),
presentandvascularised(V),present
andunvascularized(U)
A V A A A A AorU A
C
h
a
p
t
e
r
5
:
M
a
u
n
d
i
a
a
n
d
e
v
o
l
u
t
i
o
n
o
f
A
l
i
s
m
a
t
a
l
e
s
1
3
5
Tepalspresent(P)orabsent(A) P P P P P P P A
Perianthnormallyconsistingoftwo
transversalabaxialtepals(Yes/No)
Y(N) N N N N Y N
Vascularbundlebranchingintepal
bladepresent(P)orabsent(A)
A(P) A A A ? P P
Plantsmonoecious,flowersusually
bisexual(M)orplantsdioecious(D)
M(D) M M M D M M M
Typicalstamennumber (4)6(18) 6 6 (1)6 4 6 4 2
Stamenfilamentspresent(Yes/No) Y Y N N N N N N
Microsporogenesissuccessive(SU)or
simultaneous(SI)
SI SU ? SU ? ? SU SU
Pollen:monads(M),dyads(D) M D M M M M M M
Pollen:monosulcate(M),inaperturate
(I),triaperturate(T)
M I I I I I I T
Pollination:biotic(B),abiotic(A) B A A A A A A A
Carpelsintwotrimerouswhorls
(Yes/No)
N(Y) N Y Y N N N N
Whenfourcarpelsarepresent,these
areindiagonal(D)ormedianand
transversal(MT)positions
MT ? *** MT MT D MT
Carpelsfree(F)orunited(U) U(F) U U U U U F(U) F
Symplicatezoneofgynoeciumpresent
(P)orabsent(A)
A P A A ? A A A
Stomataoncarpelsurfacepresent(P)
orabsent(A)
P P ? P ? P A A
Ovulespercarpel >1 >1 1 1 1 1 1 1
Ovulesanatropous(A)ororthotropous
(O)
A A A A A O O O
Nucellarcoenocyteabsent(A)or
present(P)
A A ? A ? P A A
C
h
a
p
t
e
r
5
:
M
a
u
n
d
i
a
a
n
d
e
v
o
l
u
t
i
o
n
o
f
A
l
i
s
m
a
t
a
l
e
s
1
3
6
Carpophore/columninfruitsabsent(A)
orpresent(P)
A A A P(A) A A A A
Postantheticelongationofcarpelstalks
absent(A)orpresent(P)
A A A A A A A P
Stamensandtepalsabsciseafter
anthesis(A)ornormallyremainun
shed,atleaststamenconnectives(U)
AorU U A A U U U U
Seedswithpersistent,manylayered
mesotestaofthickwalledcells
(Yes/No)
N Y N N N N N N
Cyanogeniccompoundspresent(P)or
absent(A)
A P A P ? ? A A
Chromosomenumber,2n 16,24,32,40,
56
22 16,32,64 12,18,24,36,
,156
? ? 14,,52,104 20,40
Floraldiagrams,Fig.16 D,E F =G G A,B H I
*DatafromTomlinson(1982)thatneedfurtherverification(seePrychid&Rudall,1999).
**Thischaracterclearlyneedsfurtherinvestigationaspublisheddataareinconsistent.
***FourcarpellategynoeciaonlyrarelyoccurinTriglochin,butbothkindsofcarpelarrangement(DandMT)arerecorded(Goebel,1928).
C
h
a
p
t
e
r
5
:
M
a
u
n
d
i
a
a
n
d
e
v
o
l
u
t
i
o
n
o
f
A
l
i
s
m
a
t
a
l
e
s
1
3
7
138
Fig. 17. A, Molecularbased phylogenetic tree of Alismatales (based on Iles et al., 2013) with potential
morphologicalsynapomorphiesofdifferentclades.BD,Hypotheticalseriesofmorphologicaltransformations
of tepal/stamen complexes in the tepaloid clade of Alismatales. Schematic longitudinal sections, tepal grey,
stamen black. B, Scheuchzeriaceae, Aponogetonaceae. C, Juncaginaceae, Maundiaceae. D, Potamogeton.
Morphological synapomorphies: 1, lateral flowers with floral prophylls; 2, perianth differentiated into sepals
andpetals(reducedinspecializedaquaticformsofHydrocharitaceae);3,stamenpairsinandroeciumpresent
(reduced in specialized aquatic forms of Hydrocharitaceae); 4, placentation diffuselaminar (except in forms
with single ovule per carpel); 5, pollen pantoporate; 6, pollen inaperturate (in specialized submerged aquatic
forms with reduced exine or, in Ruppia, triaperturate); 7, ovary inferior; 8, pollination abiotic (also in
specialized Hydrocharitaceae); 9, leaf intravaginal squamules transformed into hairs; 10, flowersubtending
bractslarge,vascularized,basalonessimilartofoliageleaves;11,pollenindyads;12,stamenfilamentloss;13,
carpelspronouncedlyascidiate;14,carpelswithsingleovule(alsoinsomemembersofthepetaloidclade);15,
ovule pendent and orthotropous; 16, leaves on elongate shoots emerging above substrate (also in some
Hydrocharitaceae);17,stomataoncarpelsurfaceabsent(alsoinspecializedsubmergedHydrocharitaceae);18,
carpels free (as an exception, weakly united via flower centre: Potamogeton crispus); 19, inflorescence axis
139
flattened; 20, flowers monomerous (also in Triglochin scilloides); 21, loss of perianth (also in some other
specialized aquatics); 22, ovule campylotropous, with integumental outgrowth; 23, stamens flattened, tepal
likeafterabscissionofthecae.
Analysing morphological data on tepaloid core Alismatales in a phylogenetic context allows
discussion of stamen evolution in this group (Fig. 17BD). The possession of stamens with sessile
anthers is a synapomorphy of the clade that includes Juncaginaceae and its sister lineage. The
absence of stamen filaments is characteristic of all members of this clade. At first glance, most
species of Potamogeton (Potamogetonaceae) represent an exception because the anthers are
stalked. However, this stalk is common to the anther and the perianth member and appears late in
flowerdevelopmentduetointerprimordialandintercalarygrowth(Hegelmaier,1870;Sattler,1965;
Posluszny and Sattler, 1974). In a phylogenetic context, the common stalk should be better
interpreted as a novel structure rather than a product of congenital fusion of between the stamen
andthetepal.AspointedoutbyPoslusznyandSattler(1974:216),itisnotquitecorrecttosaythat
thetepalisinsertedatthestamenconnectiveinPotamogeton.Withrespecttotherelativeposition
oftheantherandtheadjacenttepal,PotamogetondoesnotdifferfromtaxasuchasJuncaginaceae
s.str.andMaundiaceae.

140

Conclusionsandoutlook
141
Phylogenetic relationships within Juncaginaceae as proposed by this study (Chapter 1) have

beenconfirmedbymorerecentstudies.TheplacementofMaundiaoutsidethefamilyJuncaginaceae
was recovered by two studies using other molecular markers (Iles et al. 2013, Les & Tippery 2013).
However, the exact position of Maundia in relation to the members of the Potamogeton clade
(Chapter1)couldnotbeclarifiedyet(Ilesetal.2013,Les&Tippery2013).Theproposedplacement
in a separate family Maundiaceae (Chapter 1) was discussed by APG III (2009) and Iles et al. (2013)
and Maundiaceae has been accepted by some authors until further evidence becomes available
(Stevens2001+,Les&Tippery2013).ThestudyontheflowerstructureofMaundia(Chapter5)adds
newevidencetothisdiscussion.
Important achievements presented in this thesis comprise the recircumscription of the

family (Chapter 1), investigations of the major lineages within Triglochin (Chapter 1 and 2), and the
revision of the Triglochin bulbosa complex (Chapter 3). It could be demonstrated that the
combination of molecular, morphological, ecological and distributional data (i.e. a total evidence
approach)givesbestresultsforasoundphylogeneticanalysisandclassification.However,thisstudy
hasalsoshownthenecessityforfurthertaxonomicandrevisionaryworkinthefamily.
RecircumscriptionofJuncaginaceaeis animportantbasisforfuturestudiesandallowsnew
insightintotheevolutionaryhistoryofthefamilyandrelatedAlismatales.Forthefirsttime,aspecies
level phylogeny is available for Triglochin (Chapter 2) enabling further research on the evolutionary
historyofthegenus.TherevisionoftheT.bulbosacomplex(Chapter3)andthenewlydescribedtaxa
fromSouthAfricaexpandourknowledgeofthefloraintheCapeFloristicregion.
Duringthecourseoftheproject,furtherresearchquestionsweredetectedthatcouldnotbe
addressedduetodifficultiesinobtaininglivingplantmaterialforstudy.Examplesofsuchknowledge
gaps include chromosome numbers as well as embryological data for Tetroncium and Maundia.
Karyological studies of Triglochin could yield valuable information as chromosome numbers are
unknown, for example, in the annual Australian Triglochin species and some taxa of the T. bulbosa
complex.
Ongoing investigations include taxonomic studies, for example in the Triglochin maritima
complex, including the search for type material and lectotypifications which are essential for future
revisions.CollaborativeresearchonthephylogenyandsystematicsofannualTriglochinspeciesfrom
Australia has recently been initiated. Furthermore, investigations of fruit morphology and
development in the group are continuing in collaboration with colleagues (Chapter 5; Lock et al.
2011).
Online platforms or virtual research environments such as the EDIT Platform for
Cybertaxonomy (Berendsohn 2010) or the Scratchpads (Smith et al. 2011) are increasingly used to
build dynamic online information systems in order to facilitate access to and exchange of scientific
data (Berendsohn et al. 2011, Smith et al. 2011). Information on Juncaginaceae has been very
scattered. To improve access to varied information about the family an online platform has been
established:JuncaginceaeoftheWorld(vonMering2012+),aScratchpadwhichisdevelopedinclose
collaboration with the eMonocot project (http://emonocot.org/). In addition to a continuously
142
updatedclassificationforJuncaginaceae(taxonomicbackbone),itprovidesaccesstoabibliography
andanincreasingnumberoftaxonpageswithdescriptions,images,andspecimeninformation.New
information, e.g. on type specimens or newly described taxa are continuously added to this online
tool. Such an online resource is a basis for studies in the context of etaxonomy, sometimes also
called onlinetaxonomy or webtaxonomy (e.g., Godfray et al. 2007, Mayo et al. 2008) and might
eventuallyresultinanonlinemonographyofthefamily.
Identificationkeyshavebeenpreparedforthegeneraofthefamily(Chapter1)andfortheTriglochin
bulbosa complex (Chapter 3). Identification tools including traditional dichotomous keys (single
access keys) and modern multiaccess keys (generated with the programme Xper2, Ung et al. 2010)
foralltaxaofthefamilywillbepublishedandpresentedinonlineinformationsystems.Twoexample
keys for Triglochin in Europe and Central Europe have already been published on the open access
platform Offene Naturfhrer (Hagedorn et al. 2010) and are available online (see Appendix 6, von
Mering2011).

References
143
REFERENCES
AdamsonR.S.1939:TheCapespeciesofTriglochin.J.S.Afr.Bot.5:2931.
AdamsonR.S.1943:Triglochincompactum.In:BarkerW.F.,GerstnerF.J.,AdamsonR.S.&
ComptonR.H.(ed.),PlantaeNovaeAfricanae.J.S.Afr.Bot.9:152153.
AgrawalJ.S.1952:TheembryologyofLilaeasubulataH.B.K.withadiscussiononitssystematic
position.Phytomorphology2:1529.
AlboffN.1896:ContributionsalaFloredelaTerredeFeu.IIEnumrationdesplantesducanalde
Beagle.RevistaMus.LaPlata7:355392.
APGII(AngiospermPhylogenyGroupII)2003:AnupdateoftheAngiospermPhylogenyGroup
classificationfortheordersandfamiliesoffloweringplants:APGII.Bot.J.Linn.Soc.141:399
436[onlineathttp://dx.doi.org/10.1046/j.10958339.2003.t01100158.x].
APGIII(AngiospermPhylogenyGroupIII)2009:AnupdateoftheAngiospermPhylogenyGroup
classificationfortheordersandfamiliesoffloweringplants:APGIII.Bot.J.Linn.Soc.161:105
121[onlineathttp://dx.doi.org/10.1111/j.10958339.2009.00996.x].
APSAMembers2007:TheAustralasianpollenandsporeatlasV1.0.Canberra:AustralianNational
University.Publishedathttp://apsa.anu.edu.au/[accessed15Dec2012].
ArroyoM.T.K.,MihocM.,PliscoffP.&ArroyoKalinM.2005:TheMagellanicmoorland.Pp.424
445in:FraserL.H.&KeddyP.A.(ed.),Theworldslargestwetlandsecologyandconservation.
Cambridge:UniversityPress.
AschersonP.1889:Potamogetonaceae.Pp.194218in:EnglerA.&PrantlK.(ed.),Dienatrlichen
Pflanzenfamilien2(1).Leipzig:W.Engelmann.
AstonH.I.1967:Aquaticangiosperms.RecordsoffourintroducedspeciesnewtoVictoria.
Muelleria1:169174.
AstonH.I.1977:AquaticplantsofAustralia.Carlton:MelbourneUniversityPress.
AstonH.I.1993:NewAustralianspeciesofTriglochinL.(Juncaginaceae)formerlyincludedinT.
procerumR.Br.Muelleria8:8597.
AstonH.I.1995:ArevisionofthetuberousrootedspeciesofTriglochinL.(Juncaginaceae)in
Australia.Muelleria8:331364.
AstonH.I.2011:Juncaginaceae.Pp.5384in:WilsonA.(ed.),FloraofAustralia39.Alismatalesto
Arales.Melbourne:ABRS/CSIROAustralia.
References
144
AustralianPlantCensus2008:IBISdatabase.CentreforPlantBiodiversityResearch,Councilof
HeadsofAustralianHerbaria.Publishedathttp://www.chah.gov.au/apc/index.html[accessed
21Feb2008].
AxenovE.S.1967:Newmethodofplanttissuesstainingforpreparationofpermanentanatomical
crosssections.ScientificReportsofHighSchool.BiologicalSciences11:125126.
BarrelierJ.1714:PlantaeperGalliam,HispaniametItaliamobservatae.Paris:Ganeau.
BartoloG.,BrulloS.,MajoranaG.&PavoneP.1977:NumericromosomiciperlaFloraItaliana:315
328.Inform.Bot.Ital.9:7187.
BarykinaR.P.,VeselovaT.D.,DeviatovA.G.,DjalilovaH.H.,IljinaG.M.&ChubatovaN.V.2004:
Handbookofthebotanicalmicrotechniques.Moscow:MoscowUniversityPress.
BattandierJ.A.&TrabutL.C.1884:Juncagines.In:Flored'Algeretcataloguedesplantes
d'Algrie.Monocotyldones.Alger:Jourdan.
BeaulieuJ.M.,TankD.C.,DonoghueM.J.2013:ASouthernHemisphereoriginforcampanulid
angiosperms,withtracesofthebreakupofGondwana.BMCEvol.Biol.13:80[onlineat
http://dx.doi.org/10.1186/147121481380].
BellC.D.,SoltisD.E.&SoltisP.S.2010:Theageanddiversificationoftheangiospermsrevisited.
Am.J.Bot.97:12961303[onlineathttp://dx.doi.org/10.3732/ajb.0900346].
BennettA.1897:Naiadaceae.Pp.4151in:ThiseltonDyerW.T.(ed.),Fl.Cap.[Harvey]7.
London:L.Reeve&Co.[onlineathttp://biodiversitylibrary.org/page/30235637].
BennettA.1902:Naiadaceae.Pp.215230in:ThiseltonDyerW.T.(ed.),Fl.Trop.Afr.[Oliveretal.]
8.London:L.Reeve&Co.[onlineathttp://biodiversitylibrary.org/page/378868].
BenthamG.1878:Naiadeae.Pp.164183in:FloraAustraliensis:adescriptionoftheplantsofthe
Australianterritory7.London:L.Reeve&Co.[onlineat
http://biodiversitylibrary.org/page/26125210].
BerendsohnW.G.2010:DevisingtheEDITPlatformforCybertaxonomy.Pp.16in:NimisP.L.&
VignesLebbeR.(ed.),Toolsforidentifyingbiodiversity:progressandproblems.Trieste:EUT
EdizioniUniversitdiTrieste[onlineathttp://hdl.handle.net/10077/3737].
BerendsohnW.G.,GntschA.,HoffmannN.,KohlbeckerA.,LutherK.&MllerA.2011:Biodiversity
informationplatforms:Fromstandardstointeroperability.In:SmithV.&PenevL.(ed.),e
Infrastructuresfordatapublishinginbiodiversityscience.ZooKeys150:7187[onlineat
http://dx.doi.org/10.3897/zookeys.150.2166].
References
145
BlancoD.E.&delaBalzeV.(ed.)2004:LosturbalesdelaPatagonia.Basesparasuinventarioyla
conservacindesubiodiversidad.BuenosAires:WetlandsInternational[onlineat
http://lac.wetlands.org/LinkClick.aspx?fileticket=wkPXlsJhKVE%3D&tabid=1227&mid=4854].
BlattnerF.R.1999:DirectamplificationoftheentireITSregionfrompoorlypreservedplantmaterial
usingrecombinantPCR.BioTechniques27:11801186.
BobeR.2006:TheevolutionofaridecosystemsineasternAfrica.J.AridEnviron.66:564584
[onlineathttp://dx.doi.org/10.1016/j.jaridenv.2006.01.010].
BobrovA.V.F.C.,MelikianA.P.&RomanovM.S.2009:MorphogenesisoffruitsofMagnoliophyta.
Moscow:Librokom.
BoelckeO.,MooreD.M.&RoigF.A.(ed.)1985:TransectabotnicadelaPatagoniaaustral.Buenos
Aires:ConsejoNacionaldeInvestigacionesCientficasyTcnicas(CONICET,Argentina),Instituto
delaPatagonia(Chile),RoyalSociety(GreatBritain).
BolsO.de&VigoJ.2001:Juncagincies.Pp.2628in:FloradelsPasosCatalans4.Barcelona:
Ed.Barcino.
BonnetE.1896:Naiadaceae.Pp.428432in:BonnetE.&BarratteG.(ed.),Expl.Sci.Tunisie,Cat.Pl.
BougainvilleL.A.de1772:Avoyagearoundtheworld.TranslatedfromtheFrenchbyJ.R.Forster.
London:Ridgewood(reprintGreggPress1967).
BrayshawC.T.2000:Juncaginaceae.Pp.7587in:Pondweeds,burreedsandtheirrelativesof
BritishColumbia.Victoria:RoyalBritishColumbiaMuseum.
BritishAntarcticSurvey(BAS)HigherPlantsHerbarium2009+:Publishedat
http://herbaria.plants.ox.ac.uk/bol/bas/[accessed26Oct2012].
BroughtonD.A.&McAdamJ.H.2005:AchecklistofthenativevascularfloraoftheFalklandIslands
(IslasMalvinas):newinformationonthespeciespresent,theirecology,statusanddistribution.
J.TorreyBot.Soc.132:115148.
BuchenauF.G.P.1868[1867]:IndexcriticusJuncaginacearumhucusquedescriptarum.Abh.
Naturwiss.Ver.Bremen1:213224.
BuchenauF.G.P.1882:BeitrgezurKenntnisderButomaceen,AlismaceenundJuncaginaceen.
Bot.Jahrb.Syst.2:465510[onlineathttp://biodiversitylibrary.org/page/5183911].
BuchenauF.G.P.1896:EinFallvonSaisonDimorphismusinderGattungTriglochin.Abh.
Naturwiss.VereineBremen13:408412.
BuchenauF.G.P.1903:Scheuchzeriaceae.Pp.120in:EnglerA.(ed.),Pflanzenr.(Engler)4(14):16.
Leipzig:W.Engelmann.
References
146
BuchenauF.G.P.&HieronymusG.1889:Juncaginaceae.Pp.222227in:EnglerA.&PrantlK.(ed.),
DienatrlichenPflanzenfamilien2(1).Leipzig:W.Engelmann.
BurgerW.C.1977:ThePiperalesandthemonocots:alternativehypothesesfortheoriginof
monocotyledonousflowers.Bot.Rev.43:345393.
BuzgoM.,SoltisD.E.,SoltisP.S.,MaH.,HauserB.A.,LeebensMackJ.&JohansenB.2006:Perianth
developmentinthebasalmonocotTriglochinmaritima.Aliso22:107127.
CabreraL.I.,SalazarJ.H.,ChaseM.W.,MayoS.J.,BognerJ.&DvilaP.2008:Phylogenetic
relationshipsofaroidsandduckweeds(Araceae)inferredfromcodingandnoncodingplastidDNA.
Am.J.Bot.95:11531165[onlineathttp://dx.doi.org/10.3732/ajb.0800073].
CampW.H.1947:Distributionpatternsinmodernplantsandtheproblemsofancientdispersals.
Ecol.Monogr.17:159183.
CanrightJ.E.1952:ThecomparativemorphologyandrelationshipsoftheMagnoliaceae.I.Trendsof
specializationinthestamens.Am.J.Bot.39:484497.
CastroD.de&CarvalhoFontesF.1946:Promeirocontactocitolgicocomaflorahalfilados
salgadosdeSacavm.Brotr.Cinc.Nat.15:3846.
CastroM.,FragaP.,TorresN.&RossellJ.A.2007:Cytotaxonomicalobservationsonflowering
plantsfromtheBalearicIslands.Ann.Bot.Fenn.44:409415.
CaujapCastellsJ.,JansenR.K.,MembrivesN.,PedrolaMonfortJ.,MontserratJ.M.,ArdanuyA.
2001:HistoricalbiogeographyofAndrocymbiumWilld.(Colchicaceae)inAfrica:evidencefrom
cpDNARFLPs.Bot.J.Linn.Soc.136:379392[onlineathttp://dx.doi.org/10.1111/j.1095
8339.2001.tb00581.x].
ChaseM.W.,FayM.F.,DeveyD.S.,MaurinO.,RnstedN.,DaviesJ.,PillonY.,PetersenG.,Seberg
O.,TamuraM.N.,AsmussenC.B.,HiluK.,BorschT.,DavisJ.I.,StevensonD.W.,PiresJ.C.,Givnish
T.J.,SytsmaK.J.,McPhersonM.M.,GrahamS.W.&RaiH.S.2006:Multigeneanalysesof
monocotrelationships:asummary.Pp.6375in:ColumbusJ.T.,FriarE.A.,PorterJ.M.,Prince
L.M.&SimpsonM.G.(ed.),Monocots:ComparativeBiologyandEvolution.ExcludingPoales.
Claremont:RanchoSantaAnaBotanicalGarden.[Aliso22:6375.]
ChaseM.W.,SoltisD.E.,SoltisP.S.,RudallP.J.,FayM.F.,HahnW.H.,SullivanS.,JosephJ.,Molvray
M.,KoresP.J.,GivnishT.J.,SytsmaK.J.&PiresJ.C.2000:Higherlevelsystematicsofthe
monocotyledons:anassessmentofcurrentknowledgeandanewclassification.Pp.316in:
WilsonK.K.L.&MorrisonD.(ed.),Monocots:Systematicsandevolution.Proceedingsof
monocotsII,Thesecondinternationalsymposiumonthecomparativebiologyofthe
monocotyledons.Melbourne:CSIRO.
References
147
ChaseM.W.,StevensonD.W.,WilkinP.&RudallP.J.1995:Monocotsystematics:acombined
analysis.Pp.685730in:RudallP.J.,CribbP.J.,CutlerD.F.&HumphriesC.J.(ed.),
Monocotyledons:systematicsandevolution.Kew:RoyalBotanicGardens.
CheahC.H.&StoneB.C.1975:EmbryosacandmicrosporangiumdevelopmentinPandanus
(Pandanaceae).Phytomorphology25:228238.
ChenJ.M.,ChenD.,GituruW.R.,WangQ.F.&GuoY.H.2004:EvolutionofapocarpyinAlismatidae
usingphylogeneticevidencefromchloroplastrbcLgenesequencedata.Bot.Bull.Acad.Sinica
45:3340.
ChristenhuszM.J.M.&ChaseM.W.2013:BiogeographicalpatternsofplantsintheNeotropics
dispersalratherthanplatetectonicsismostexplanatory.Bot.J.Linn.Soc.171:277286[online
athttp://dx.doi.org/10.1111/j.10958339.2012.01301.x].
ChubirkoM.M.1990:Pandanaceae.Pp.268270in:BatyginaT.B.&YakovlevM.S.(ed.),
Comparativeembryologyoffloweringplants.Monocotyledones.ButomaceaeLemnaceae.
Leningrad:Nauka.
ContiF.,ManziA.&PedrottiF.1997:ListeRosseRegionalidellePiantedItalia.Camerino:WWF
Italia,SocietBotanicaItaliana,UniversitdiCamerino.
CookC.D.K.2004:AquaticandwetlandplantsofsouthernAfrica.Leiden:Backuys.
CorellD.S.&CorellH.B.1982:Juncaginaceae.Pp.7981in:FloraoftheBahamasArchipelago
(includingtheTurksandCaicosIslands).Vaduz:Gantner(ReprintEdition:1996).
CorreaM.N.1969:Juncaginaceae.Pp.3337in:CorreaM.N.(ed.),FloraPatagnica2.
Monocotyledoneae(exceptoGramineae).BuenosAires:INTA.
CrispM.D.&CookL.G.2012:Phylogeneticnicheconservatism:whataretheunderlying
evolutionaryandecologicalcauses?NewPhytol.196:681694[onlineat
http://dx.doi.org/10.1111/j.14698137.2012.04298.x].
CronquistA.1981:Anintegratedsystemofclassificationoffloweringplants.NewYork:Columbia
UniversityPress.
CusimanoN.,BognerJ.,MayoS.J.,BoyceP.C.,WongS.Y.,HesseM.,HetterscheidW.L.A.,Keating
R.C.&FrenchJ.C.2011:RelationshipswithintheAraceae:Comparisonofmorphologicalpatterns
withmolecularphylogenies.Am.J.Bot.98:654668[onlineat
http://dx.doi.org/10.3732/ajb.1000158].
DahlgrenR.M.T.,CliffordH.T.&YeoP.F.1985:Thefamiliesofmonocotyledons.Structure,
evolution,andtaxonomy.Berlin:Springer.
References
148
DahlgrenR.,KarlssonT.&LassenP.1971:StudiesonthefloraoftheBalearicIslandsI.Chromosome
numbersinBalearicangiosperms.Bot.Notis.124:249269.
DandyJ.E.1980:Triglochin.Pp.67in:TutinT.G.,HeywoodV.H.,BurgesN.A.,MooreD.M.,
ValentineD.H.,WaltersS.M.&WebbD.A.(ed.),FloraEuropaea5.Cambridge:Cambridge
UniversityPress.
DavisJ.I.,PetersenG.,SebergO.,StevensonD.W.,HardyC.R.,SimmonsM.P.,MichelangeliF.A.,
GoldmanD.H.,CampbellL.M.,SpechtC.D.&CohenJ.I.2006:Aremitochondrialgenesusefulfor
theanalysisofmonocotrelationships?Taxon55:857870.
DavisJ.I.,SimmonsM.P.,StevensonD.W.&WendelJ.F.1998:Datadecisiveness,dataquality,and
incongruenceinphylogeneticanalysis:anexamplefromthemonocotyledonsusingmitochondrial
atpAsequences.Syst.Biol.47:282310.
DavisJ.I.,StevensonD.W.,PetersenG.,SebergO.,CampbellL.M.,FreudensteinJ.V.,GoldmanD.
H.,HardyC.R.,MichelangeliF.A.,SimmonsM.P.,SpechtC.D.,VergaraSilvaF.&GandolfoM.
2004:Aphylogenyofthemonocots,asinferredfromrbcLandatpAsequencevariation,anda
comparisonofmethodsforcalculatingJackknifeandBootstrapvalues.Syst.Bot.29:467510.
DavyA.J.&BishopG.F.1991:BiologicalfloraoftheBritishIsles:TriglochinmaritimaL.J.Ecol.79:
531555.
denHartogC.1970:Theseagrassesoftheworld.Amsterdam:NorthHollandPubl.
DoyleJ.A.&EndressP.K.2000:Morphologicalphylogeneticanalysisofbasalangiosperms:
comparisonandcombinationwithmoleculardata.Int.J.PlantSci.161:S121S153.
DrummondA.J.,SuchardM.A.,XieD.&RambautA.2012:BayesianphylogeneticswithBEAUtiand
theBEAST1.7.Mol.Biol.Evol.29:19691973[onlineat
http://dx.doi.org/10.1093/molbev/mss075].
DusnP.1900:Juncaginaceae.P.234in:DieGefpflanzenderMagellanslnder.Svenska
expeditionentillMagellanslnderna3(5).Stockholm:P.A.Norstedt[onlineat
DusnP.1903:ThevegetationofWesternPatagonia.ReportsofthePrincetonUniversity
ExpeditionstoPatagonia(18961899)(Botany)8:133.
EamesA.J.1961:Morphologyoftheangiosperms.NewYork:McGrawHill.
EberE.1934:KarpellbauundPlazentationsverhltnisseinderReihederHelobiae.Flora127:273
330.
EckardtT.1964:ReiheHelobiae.Pp.499512in:MelchiorH.(ed.),A.EnglersSyllabusder
Pflanzenfamilien2.Berlin:Borntraeger.
References
149
EDITmapViewer2012:EDITmapViewerserviceoftheEuropeanDistributedInstituteofTaxonomy
(EDIT)project,hostedbytheRoyalMuseumforCentralAfrica.Publishedat
http://edit.africamuseum.be[accessed19Nov2012].
EichlerA.1875:Blthendiagramme1.Leipzig:Engelmann.
EndressP.K.1995:Majortraitsofmonocotflowers.Pp.4379in:RudallP.J.,CribbP.J.,CutlerD.
F.&HumphriesC.J.(ed.),Monocotyledons:systematicsandevolution.Kew:RoyalBotanic
Gardens.
EndressP.K.&DoyleJ.A.2009:Reconstructingtheancestralangiospermfloweranditsinitial
specializations.Am.J.Bot.96:2266[onlineathttp://dx.doi.org/10.3732/ajb.0800047].
EnglerA.1909:SyllabusderPflanzenfamilien.6Aufl.Berlin:Borntraeger.
EspejoSerenaA.&LpezFerrariA.R.1996:Juncaginaceae.Pp.6970in:LasMonocotiledneas
MexicanasunaSinopsisFlorstica1(6).MxicoD.F.:ConsejoNacionaldelaFloradeMxico.
ESRI2006:ArcGIS9.0.UsingArcGISDesktop.Redlands,CA,USA.
FarrisJ.S.,KllersjM.,KlugeA.G.&BultC.1994:Testingsignificanceofincongruence.Cladistics
10:315319.
FinlaysonM.&MoserM.(ed.)1991:Wetlands.Oxford:InternationalWaterfowlandWetlands
ResearchBureau(IWRB).
FloraArgentina2012+:PlantasVascularesdelaRepblicaArgentina.Publishedat
http://www.floraargentina.edu.ar/[accessed26Oct2012].
FloradelConosur2009+:Catlogodeplantasvasculares.Publishedat
http://www2.darwin.edu.ar/Proyectos/FloraArgentina/FA.asp[accessed26Oct2012].
FordB.A.&BallP.W.1988:AreevaluationoftheTriglochinmaritimumcomplex(Juncaginaceae)in
easternandcentralNorthAmericaandEurope.Rhodora90:313337.
FurnessC.A.&BanksH.2010:PollenevolutionintheearlydivergentmonocotorderAlismatales.
Int.J.PlantSci.171:713739[onlineathttp://www.jstor.org/stable/10.1086/654848].
GandogerM.1919:Sertumplantarumnovarum.Parssecunda.Bull.Soc.Bot.France66:286307.
GardA.&MalheirosGardN.1953:Contribuioparaoestudocariologicodealgumasspeciesde
AngiospermicasI.GeneticaIberica5(34):115124.
GayC.1849:HistoriafisicaypoliticadeChile5.Paris(bytheauthor)&Santiago:Museodehistoria
natural[onlineathttp://biodiversitylibrary.org/page/16435603].
References
150
GeoNames2012:GeoNamesgeographicaldatabase.Publishedathttp://www.geonames.org/
[accessed26Oct2012].
GEOnetNamesServer(GNS)1994+:ToponymicinformationisbasedontheGeographicNames
Database,containingofficialstandardnamesapprovedbytheUnitedStatesBoardonGeographic
NamesandmaintainedbytheNationalGeospatialIntelligenceAgency.Publishedat
http://earthinfo.nga.mil/gns/html[accessed26Oct2012].
GibbsR.D.1974:Chemotaxonomyoffloweringplants3.Montreal:McGillQueensUniversity
Press.
GlenR.P.&CookC.D.K.2003:Juncaginaceae.Pp.11181119in:GermishuizenG.&MeyerN.L.
(ed.),PlantsofsouthernAfrica:anannotatedchecklist.Pretoria:NationalBotanicalInstitute.
GlezPeaD.,GmezBlancoD.,ReboiroJatoM.,FdezRiverolaF.&PosadaD.2010:ALTER:program
orientedformatconversionofDNAandproteinalignments.NucleicAcidsRes.38(WebServer
issue):W14W18[onlineathttp://dx.doi.org/10.1093/nar/gkq321].
GlobalBiodiversityInformationFacility2010:GBIFDataPortal.Version1.3.1Publishedat
http://data.gbif.org[accessed26Oct2012].
GodfrayH.C.J.,ClarkB.R.,KitchingI.J.,MayoS.J.&ScobleM.J.2007:Thewebandthestructureof
taxonomy.Syst.Biol.56:943955[onlineathttp://dx.doi.org/10.1080/10635150701777521].
GodleyE.J.1965:Botanyofthesouthernzone.Explorationto1843.Tuatara13(3):140181.
[Onlineathttp://nzetc.victoria.ac.nz/tm/scholarly/teiBio13Tuat03t1bodyd1.html].
GodleyE.J.1970:Botanyofthesouthernzone:Exploration,18471891.Tuatara18(2):5093
[onlineathttp://nzetc.victoria.ac.nz/tm/scholarly/teiBio18Tuat02t1bodyd1.html].
GoebelK.1928:OrganographiederPflanzen.T.1.Jena:G.Fischer.
GoldblattP.1979:BiologyandsystematicsofGalaxia(Iridaceae).J.S.Afr.Bot.45:385423.
GoldblattP.&JohnsonD.E.(ed.)1979+:Indextoplantchromosomenumbers(IPCN).St.Louis:
MissouriBotanicalGarden.Publishedathttp://www.tropicos.org/Project/IPCN[accessed9
March2013].
GoldblattP.&ManningJ.2000:Capeplants.AconspectusoftheCapefloraofSouthAfrica.
Pretoria:NationalBotanicalInstitute&St.Louis:MissouriBotanicalGarden.
GonzlezJ.D.1968:EstudiofitoecolgicodelalagunadeValdovio,enLaCorua.Botanica
Complutensis1:1550.
Google2012:GoogleEarthversion6.2.Publishedathttp://www.google.com/earth/download/ge/
[accessed25Aug2012].
References
151
GovaertsR.2008:WorldChecklistofJuncaginaceae.TheBoardofTrusteesoftheRoyalBotanic
Gardens,Kew.Publishedathttp://www.kew.org/wcsp/monocots[accessed21Feb2008,then
stillbeingUR=underreview(partiallyreviewed)].
GovaertsR.2013:WorldchecklistofJuncaginaceae.FacilitatedbytheRoyalBotanicGardens,Kew.
Publishedathttp://apps.kew.org/wcsp/[accessed20Apr2013].
GrahamS.W.,ZgurskiJ.M.,McPhersonM.A.,CherniawskyD.M.,SaarelaJ.M.,HorneE.F.C.,Smith
S.Y.,WongW.A.,O'BrienH.E.,BironV.L.,PiresJ.C.,OlmsteadR.G.,ChaseM.W.&RaiH.S.
2006:Robustinferenceofmonocotdeepphylogenyusinganexpandedmultigeneplastiddata
set.Aliso22:321.
GrayumM.H.1992:ComparativeexternalpollenultrastructureoftheAraceaeandputativelyrelated
taxa.Monogr.Syst.Bot.MissouriBot.Gard.43:1167.
GreenE.P.&ShortF.T.(ed.)2003:WorldAtlasofSeagrasses.PreparedbytheUNEPWorld
ConservationMonitoringCentre.Berkeley:UniversityofCaliforniaPress.
GrillasP.,GauthierP.,YavercovskiN.&PerennouC.(ed.)2004:Mediterraneantemporarypools1.
Arles:StationbiologiquedelaTourduValat.
GrootjansA.,IturraspeR.,LantingA.,FritzC.&JoostenH.2010:Ecohydrologicalfeaturesofsome
contrastingmiresinTierradelFuego,Argentina.MiresandPeat6:115.
GrovesE.W.1981:VascularplantcollectionsfromtheTristandaCunhagroupofislands.Bull.Brit.
Mus.(Nat.Hist.),Bot.8(4):333420.
GuoY.,HaynesR.R.&HellquistB.C.2010:Juncaginaceae.P.105in:FloraofChina23.Beijing:
SciencePress&St.Louis:MissouriBotanicalGarden[onlineat
http://flora.huh.harvard.edu/china/mss/volume23/Flora_of_China_Vol_23_Juncaginaceae.pdf].
GussoneG.1825:Ind.Sem.Hort.Boccad.1825:12.
GussoneG.1827:Fl.Sic.Prod.1.Neapoli:ExRegiaTypographia.
HagedornG.,PressB.,HetznerS.,PlankA.,WeberG.,vonMeringS.,MartellosS.&NimisP.L.2010:
AMediaWikiimplementationofsingleaccesskeys.Pp.7782in:NimisP.L.&VignesLebbeR.
(ed.),Toolsforidentifyingbiodiversity:progressandproblems.Trieste:EUTEdizioniUniversit
diTrieste[onlineathttp://hdl.handle.net/10077/3753].
HarvardUniversityHerbaria2011+:HUHIndexofBotanists.Publishedat
http://kiki.huh.harvard.edu/databases/botanist_index.html[accessed26Oct2012].
HayekA.1933:ProdromusFloraepeninsulaeBalcanicae3.BerlinDahlem:Verlagdes
Repertoriums.
References
152
HaynesR.R.2004:Juncaginaceae.Pp.452453in:SmithN.,MoriS.A.,HendersonA.,StevensonD.
W.&HealdS.V.(ed.),Floweringplantsoftheneotropics.NewYorkBotanicalGarden.
Princeton:PrincetonUniversityPress.
HaynesR.R.&HellquistC.B.2000:Juncaginaceae.Pp.4346in:FloraofNorthAmericaEditorial
Committee(ed.)1993+,FloraofNorthAmericaNorthofMexico22.NewYork:Oxford
UniversityPress[onlineat
http://www.efloras.org/florataxon.aspx?flora_id=1&taxon_id=133624].
HaynesR.R.,LesD.H.&HolmNielsenL.B.1998:Juncaginaceae.Pp.260263in:KubitzkiK.(ed.),
Fam.Gen.Vasc.Pl.[Kubitzki]4.Berlin:Springer.
HegelmaierF.1870:UeberdieEntwicklungderBlthentheilevonPotamogeton.Bot.Zeitung
(Berlin)28:281289,297305,313319.
HellquistC.B.&JacobsS.W.L.2011:Aponogetonaceae.Pp.4452in:WilsonA.(ed.),Floraof
Australia39.AlismatalestoArales.Melbourne:ABRS/CSIROAustralia.
HiepkoP.(ed.)1972:HerbariumWilldenow.AlphabeticalIndex.Zug:IDC.
HolmgrenP.K.&HolmgrenN.H.1998+:IndexHerbariorum:Aglobaldirectoryofpublicherbariaand
associatedstaff.NewYorkBotanicalGarden'sVirtualHerbarium:http://sweetgum.nybg.org/ih.
HookerJ.D.1844:FloraAntarctica:thebotanyoftheAntarcticvoyage.London:ReeveBrothers.
HookerJ.D.1847:Alismaceae.Pp.359360in:ThebotanyoftheAntarcticvoyageofH.M.
discoveryshipsErebusandTerrorintheYears18391843underthecommandofCaptainSir
JamesClarkRoss.FloraAntarctica1(2).London:ReeveBrothers[onlineat
http://www.biodiversitylibrary.org/item/22023]
HookerW.J.1843:Tetronciummagellanicum.Icon.Pl.6:t.534.London:HippolyteBaillire
[onlineathttp://biodiversitylibrary.org/page/16048908].
HornafRantzienH.1961:NotesontheAfricanspeciesofTriglochin.SvenskBot.Tidskr.55:81
117.
HoyoA.del,GarcaMarnJ.L.&PedrolaMonfortJ.2009:Temporalandspatialdiversificationofthe
AfricandisjunctgenusAndrocymbium(Colchicaceae).Mol.Phylogenet.Evol.53:848861
[onlineathttp://dx.doi.org/10.1016/j.ympev.2009.08.005].
HutchinsonJ.1934:Thefamiliesoffloweringplants2.London:Macmillan.
HutchinsonJ.1959:Thefamiliesoffloweringplants2.Ed.2.Oxford:ClarendonPress.
IgersheimA.,BuzgoM.&EndressP.K.2001:Gynoeciumdiversityandsystematicsinbasal
monocots.Bot.J.Linn.Soc.136:165.
References
153
IlesW.J.D.,SmithS.Y.&GrahamS.W.2013:Awellsupportedphylogeneticframeworkforthe
monocotorderAlismatalesrevealsmultiplelossesoftheplastidNADHdehydrogenasecomplex
andastronglongbrancheffect.Pp.128in:WilkinP.&MayoS.J.(ed.),Earlyeventsin
monocotevolution.Cambridge:CambridgeUniversityPress.
InstitutodeBotnicaDarwinion1997+:FloradelConosur.CatlogodelasPlantasVasculares.
Publishedathttp://www2.darwin.edu.ar/Proyectos/FloraArgentina/FA.asp[accessed26Oct
2012].
IturraspeR.J.&UrciuoloA.B.2004:LestourbiresdelaTerredeFeuenArgentine:unpatrimoine
natureltrsmenace.Gocarrefour79(4):325330.[Publishedat
http://geocarrefour.revues.org/2842].
IUCN2001:IUCNRedListcategoriesandcriteria,Version3.1.IUCNSpeciesSurvivalCommission.
GlandandCambridge:IUCN.
JacobsS.W.L.2009:NewSouthWalesFloraOnline.GenusMaundia.In:PlantNETThePlant
InformationNetworkSystemofBotanicGardenTrust,Sydney,Australia,v.2.0.Publishedat
http://plantnet.rbgsyd.nsw.gov.au/.
JafriS.M.H.1973:Juncaginaceae.Pp.15in:FloraofPakistan48.Karachi:Departmentof
Botany,UniversityofKarachi.
JgerZrnI.1967:EmbryologischeUntersuchungenanvierPodostemaceen.sterreichische
botanischeZeitschrift114:2045.
JgerZrnI.1997:EmbryologicalandfloralstudiesinWeddellinasquamulosaTul.(Podostemaceae,
Tristichoideae).Aquat.Bot.57:151182.
JahandiezE.&MaireR.1931:CataloguedesplantesduMaroc1.Alger:Minerva.
JanssenT.&BremerK.2004:Theageofmajormonocotgroupsinferredfrom800+rbcLsequences.
Bot.J.Linn.Soc.146:385398[onlineathttp://dx.doi.org/10.1111/j.10958339.2004.00345.x].
JohnsonL.A.&SoltisD.E.1994:matKDNAsequencesandphylogeneticreconstructionin
Saxifragaceaes.str.Syst.Bot.19:143156.
JonesA.G.,ChownS.L.,RyanP.G.,GremmenN.J.M.&GastonK.J.2003:Areviewofconservation
threatsonGoughIsland:acasestudyforterrestrialconservationintheSouthernOceans.Biol.
Cons.113:7587[onlineathttp://dx.doi.org/10.1016/S00063207(02)003518].
JSTORPlantScience2010+:Publishedathttp://plants.jstor.org/[accessed26Oct2012].
JuddW.S.,CampbellC.S.,KellogE.A.,StevensP.F.&DonoghueM.J.2007:Plantsystematics:a
phylogeneticapproach.Ed.3.Sunderland:SinauerAssociates.
References
154
JrgensN.1997:FloristicbiodiversityandhistoryofAfricanaridregions.Biodiv.Conserv.6:495
514.
KamelinaO.P.2011:Systematicembryologyoffloweringplants.Monocotyledones.Barnaul:
Artika.
KeigheryG.J.1979:NotesonthebiologyandphytogeographyofWesternAustralianplants.Part3:
Juncaginaceae.Unpubl.report.Perth:KingsPark.
KeigheryG.J.&KeigheryB.J.2003:JuncaginaceaeinWesternAustralia:centreofspeciesdiversityof
anotherfamilyinthismegadiverseregion.P.45in:MonocotsIII:Abstracts:theThird
InternationalConferenceontheComparativeBiologyoftheMonocotyledons&theFourth
InternationalSymposiumonGrassSystematicsandEvolution:31March4April,2003,Ontario,
California.Claremont:RanchoSantaAnaBotanicGarden.
KeilD.J.2012:Triglochin.Pp.13751376in:Baldwin,B.G.,Goldman,D.H.,Keil,D.J.,Patterson,R.,
Rosatti,T.J.,Wilken,D.H.(Ed.),TheJepsonmanual:vascularplantsofCalifornia.2nded.
Berkeley:UniversityofCaliforniaPress.
KleinebeckerT.,HlzelN.,VogelA.2007:GradientsofcontinentalityandmoistureinSouth
Patagonianobrotrophicpeatlandvegetation.FoliaGeobot.42:363382.
KckeA.V.2007:PhylogenieundSystematikvonTriglochin(Juncaginaceae)inSdafrika.
UnpublishedDiplomathesis.Mainz:JohannesGutenbergUniversity.
KckeA.V.,vonMeringS.,MucinaL.,KadereitJ.W.2010:RevisionoftheMediterraneanand
southernAfricanTriglochinbulbosacomplex(Juncaginaceae).EdinburghJ.Bot.67:353398
[onlineathttp://dx.doi.org/10.1017/S0960428610000041].
KoutstalB.P.,MarkusseM.M.&deMunckW.W.1987:Aspectsofseeddispersalbytidal
movements.Pp.226233in:HuiskesA.H.L.,BlomW.P.M.&RozemaJ.(ed.),Vegetation
betweenlandandsea.Dordrecht:JunkPubl.
KubitzkiK.1998:Thefamiliesandgeneraofvascularplants4.Berlin:Springer.
KunthK.S.1841:Juncagineae.Enum.Pl.[Kunth]3:141146.StuttgartandTbingen:J.G.Cotta
[onlineathttp://biodiversitylibrary.org/page/7414507].
KupriyanovaL.A.1948:.[Pollen
morphologyandphylogenyofthemonocotyledons].TrudyBot.Inst.Akad.NaukS.S.S.R.,Ser.1,
Fl.Sist.Vyssh.Rast.7:163262.
KurzweilH.,LinderH.P.&ChesseletP.1991:ThephylogenyandevolutionofthePterygodium
Coryciumcomplex(Coryciinae,Orchidaceae).PlantSyst.Evol.175:161223.
LarsenK.1966:CytotaxonomicalnoteonLilaea.Bot.Notiser119:496497.
References
155
LavinM.,SchrireB.P.,LewisG.,PenningtonR.T.,DelgadoSalinasA.,ThulinM.,HughesC.E.,Beyra
MatosA.&WojciechowskiM.F.2004:Metacommunityprocessratherthancontinentaltectonic
historybetterexplainsgeographicallystructuredphylogeniesinlegumes.Phil.Trans.R.Soc.B
359:15091522[onlineathttp://dx.doi.org/10.1098/rstb.2004.1536].
LeonH.1946:Juncaginaceas.Pp.7879in:FloradeCuba1.Havana:Cultural,S.A.
[ContribucionesocasionalesdelMuseodeHistoriaNaturaldelColegiodelaSalle8(1946)].
LesD.H.,ClelandM.A.&WaycottM.1997:PhylogeneticstudiesinAlismatidae,II:evolutionof
marineangiosperms(seagrasses)andhydrophily.Syst.Bot.22:443463.
LesD.H.,CrawfordD.J.,KimballR.T.,MoodyM.L.&LandoltE.2003:Biogeographyof
discontinuouslydistributedhydrophytes:amolecularappraisalofintercontinentaldisjunctions.
Int.J.PlantSci.164:917932.
LesD.H.,MoodyM.L.&JacobsS.W.L.2005:PhylogenyandsystematicsofAponogeton
(Aponogetonaceae):theAustralianspecies.Syst.Bot.30:503519.
LesD.H.&TipperyN.P.2013:Intimeandwithwater...thesystematicsofalismatid
monocotyledons.Pp.118164in:WilkinP.&MayoS.J.(ed.),Earlyeventsinmonocot
evolution.Cambridge:CambridgeUniversityPress.
LieuS.M.1979:OrganogenesisinTriglochinstriata.Can.J.Bot.57:14181438.
LinderH.P.2003:TheradiationoftheCapeflora,southernAfrica.Biol.Rev.(Cambridge)78:597
638.
LinderH.P.&EllisR.P.1990:ArevisionofPentaschistis(Arundineae:Poaceae).Contrib.Bolus
Herb.12:1124.
LinnaeusC.1753:Sp.Pl.Stockholm:Holmiae.
LinnaeusC.1762:Sp.Pl.,ed.2.Stockholm:Holmiae.
LinnaeusC.1771:Mant.Pl.Altera.Stockholm:Holmiae.
LisowskiS.,MalaisseF.&SymoensJ.J.1982:Juncaginaceae.Pp.16in:Flored'AfriqueCentrale
(ZareRwandaBurundi).Meise:JardinBotaniqueNationaldeBelgique.
LoPrestiR.M.&OberprielerCh.2009:Evolutionaryhistory,biogeographyandecoclimatological
differentiationofthegenusAnthemisL.(Compositae,Anthemideae)intheCircumMediterranean
area.J.Biogeogr.36:13131332.
LockI.E.,RemizowaM.V.,vonMeringS.,KckeA.V.&SokoloffD.D.2011:Flowerandfruit
anatomyinAfricanTriglochin(Juncaginaceae:Alismatales).Pp.144145in:DemidovA.S.(ed.),
Carpologyandreproductivebiologyofhigherplants:ProceedingsoftheRussianconferencewith
References
156
internationalparticipationdedicatedtothememoryofProfessorA.P.Melikian.Moscow:Astra
Polygraphia.
LoiseleurDeslongchampsJ.L.A.1806:Fl.Gall.Part1.Paris:Migneret.
LoiseleurDeslongchampsJ.L.A.1807:Fl.Gall.Part2.Paris:Migneret.
LoiseleurDeslongchampsJ.L.A.1828:Fl.Gall.,ed.2.Part1.Paris:Baillire.
LveA.&LveD.1958:BiosystematicsofTriglochinmaritimumagg.NaturalisteCanad.85:156
165.
LveD.&LiethH.1961:Triglochingaspense,anewspeciesofarrowgrass.Canad.J.Bot.39:1261
1272.
LyleM.,BarronJ.,BralowerT.,HuberM.,OlivarezLyleA.,RaveloA.C.,ReaD.K.&WilsonP.A.2008:
PacificOceanandCenozoicevolutionofclimate.Rev.Geophys.46:147[onlineat
http://dx.doi.org/10.1029/2005RG000190].
MabberleyD.J.2008:MabberleysPlantBook.Aportabledictionaryofplants,theirclassificationand
uses.Ed.3.Cambridge:CambridgeUniversityPress.
MackelR.1985:DambosandrelatedlandformsinAfricaanexamplefortheecologicalapproachto
tropicalgeomorphology.Z.Geomorphol.N.F.Suppl.52:123.
MaddisonW.P.&MaddisonD.R.2000:MacClade,v.4.01.Sunderland:SinauerAssociates.
MaireR.1952:Juncaginaceae.Pp.210214in:Fl.AfriqueN.1.Paris:Lechevalier[onlineat
www.telabotanica.org/projets/66/telechargement/17740].
MarkgrafF.1936:BltenbauundVerwandtschaftbeideneinfachstenHelobiae.Ber.Deutsch.Bot.
Ges.54:191228.
MarkgrafF.1981:Juncaginaceae.Pp.198203in:GustavHegiIllustrierteFloravonMitteleuropa
I.1,2(ed.3).Berlin:Parey.
MarlothR.1908:DasKapland,insonderheitdasReichderKapflora,dasWaldgebietunddieKarroo,
pflanzengeographischdargestellt.Pp.1427in:ChunC.(ed.),WissenschafftlicheErgebnisseder
deutschenTiefseeExpeditionaufdemDampferValdivia189818992(3).Jena:GustavFischer.
MarticorenaC.&QuezadaM.1985:CatlogodelaFloraVasculardeChile.Gayana,Bot.42:1157.
MavrodievE.V.&SokoloffD.D.1998:OnmorphologyofEuropeanspeciesoffamilies
Zannichelliaceae,Ruppiaceae,PotamogetonaceaeandZosteraceae.Bull.MoscowSoc.Natur.,
Biol.Ser.103(5):4960.
References
157
MayoS.J.,AllkinR.,BakerW.,BlagoderovV.,BrakeI.,ClarkB.,GovaertsR.,GodfrayC.,HaighA.,
HandR.,HarmanK.,JacksonM.,KilianN.,KirkupD.W.,KitchingI.,KnappS.,LewisG.P.,Malcolm
P.,vonRaabStraubeE.,RobertsD.M.,ScobleM.,SimpsonD.A.,SmithC.,SmithV.,VillalbaS.,
WalleyL.&WilkinP.2008:Alphaetaxonomy:responsesfromthesystematicscommunitytothe
biodiversitycrisis.KewBull.63:116[onlineathttp://dx.doi.org/10.1007/s1222500890141].
MayoS.J.,BognerJ.&BoyceP.C.1997:ThegeneraofAraceae.Kew:RoyalBotanicGardens.
McLoughlinS.2001:ThebreakuphistoryofGondwanaanditsimpactonpreCenozoicfloristic
provincialism.Austral.J.Bot.49:271300.
McNeillJ.,BarrieF.R.,BuckW.R.,DemoulinV.,GreuterW.,HawksworthD.L.,HerendeenP.S.,
KnappS.,MarholdK.,PradoJ.,PrudhommevanReineW.F.,SmithG.F.,WiersemaJ.H.&
TurlandN.J.(ed.)2012:InternationalCodeofNomenclatureforalgae,fungi,andplants
(MelbourneCode):AdoptedbytheEighteenthInternationalBotanicalCongressMelbourne,
Australia,July2011.RegnumVeg.154.
MedWet1996:MediterraneanWetlandStrategy19962006.EndorsedbytheMediterranean
WetlandsConference,Venice,June1996[onlineathttp://agadir2012.medwet.org/wp
content/uploads/2011/07/MediterraneanWetlandStrategy1996.pdf].
MelikianA.P.1985:Butomaceae.Pp.3334in:TakhtajanA.(ed.),Anatomiaseminum
comparativa.T.1.Leningrad:Nauka.
MicheliM.1881:Juncagineae.Pp.94112in:deCandolleA&deCandolleC.(ed.),Monogr.Phan.
[A.DC.&C.DC.]3.Paris:G.Masson.
MiddletonB.A.&KleinebeckerT.2012:Theeffectsofclimatechangeinduceddroughtand
freshwaterwetlands.Pp.117147in:MiddletonB.A.(ed.),Globalchangeandthefunctionand
distributionofwetlands,globalchangeecologyandwetlands1.Dordrecht:Springer.
MikiS.1937:TheoriginofNajasandPotamogeton.Bot.Mag.(Tokyo)51:472480.
MooreD.M.1968:ThevascularfloraoftheFalklandIslands.Sci.Rep.Brit.Antarc.Surv.60:1202.
MooreD.M.1983:FloraofTierradelFuego.London:AnthonyNelson.
MooreL.B.1970:Juncaginaceae.Pp.78in:MooreL.B.&EdgarE.(ed.),FloraofNewZealand2.
Wellington,N.Z.:A.R.Shearer[onlineathttp://FloraSeries.LandcareResearch.co.nz].
MucinaL.&RutherfordM.C.(ed.)2006:ThevegetationofSouthAfrica,LesothoandSwaziland.
Pretoria:SANBI.
MuellerF.1858:Juncagineae.Pp.2223in:MuellerF.,FragmentaphytographiaeAustraliae1.
Melbourne:AuctoritateGubern.
References
158
MllerK.&BorschT.2005:PhylogeneticsofAmaranthaceaebasedonmatK/trnKsequencedata:
evidencefromParsimony,Likelihood,andBayesianAnalyses.Ann.MissouriBot.Gard.92:66
102.
MuirG.&SchlttererC.1999:LimitationstothephylogeneticuseofITSsequencesincloselyrelated
speciesandpopulationsacasestudyinQuercuspetraea(Matt.)Liebl.Chapter11in:GilletE.
M.(ed.),WhichDNAMarkerforWhichPurpose?FinalCompendiumoftheResearchProject:
Development,optimizationandvalidationofmoleculartoolsforassessmentofbiodiversityin
foresttreesintheEuropeanUnionDGXIIBiotechnologyFWIVResearchProgramMolecularTools
forBiodiversity[onlineat
http://webdoc.gwdg.de/ebook/y/1999/whichmarker/m11/Chap11.htm].
MunzP.A.&KeckD.D.1968:Juncaginaceae.Lilaeaceae.Pp.13201322in:ACaliforniaFlora.
Berkeley:UniversityofCaliforniaPress.
MusumNationald'HistoireNaturelle(ed.)20032006:InventairenationalduPatrimoinenaturel.
Publishedathttp://inpn.mnhn.fr[accessed17Jul2008].
NagendranC.R.,AnandV.V.&ArekalG.D.1980:TheembryosacofPodostemumsubulatus
(Podostemaceae)areinvestigation.PlantSyst.Evol.134:121125.
NagendranC.R.,ArekalG.D.&SubramanyamK.1977:EmbryosacstudiesinthreeIndianspeciesof
Polypleurum(Podostemaceae).PlantSyst.Evol.128:215226.
NakaiT.1943:Maundiaceae.Pp.213in:NakaiT.,Ordines,Familiae,Tribi,Genera,Sectiones
novisedita.Appendix.Tokyo:ImperialUniversity.
NapperD.M.1971:Juncaginaceae.Pp.13in:MilneRedheadE.&PolhillR.M.(ed.),Fl.Trop.E.
Africa.London:CrownAgentsforOverseasGovernmentsandAdministrations.
NavaH.S,FernndezCasadoM..&SurezPrezF.J.2000:Lilaeascilloides(Poir.)Hauman
(Lilaeaceae),enAsturias.AnalesJard.Bot.Madrid58:191.
NoveloA.2003:Lilaeaceae.Pp.15in:RzedowskiJ.&RzedowskiG.(ed.),FloradelBajoyde
regionesadyascentes118.Ptzcuaro,Mich.:InstitutodeEcologa,A.C.
NoveloA.&LotA.2001:Juncaginaceae,Lilaeaceae.Pp.986993in:RzedowskiG.C.de,Rzedowski
J.etal.(ed.),FlorafanerogmicadelValledeMxico.Ed.2.Ptzcuaro,Mich.:Institutode
Ecologa,A.C.(INECOL)andComisinNacionalparaelConocimientoyUsodelaBiodiversidad
(CONABIO)(Reprint2005).[Onlineat
www.biodiversidad.gob.mx/publicaciones/librosDig/pdf/Flora_del_Valle_de_Mx2.pdf].
NunesE.L.P.,deLimaM.C.,deChiaraMooM.C.&CoanA.I.2012:Floraldevelopmentin
Potamogeton(Potamogetonaceae,Alismatales)withemphasisongynoecialfeatures.Aquat.
Bot.100:5661.
References
159
NuralievM.S.,SokoloffD.D.&OskolskiA.A.2011:FloralanatomyofAsianSchefflera(Araliaceae,
Apiales):comparingvariationofflowergroundplanandvascularpatterns.Int.J.PlantSci.172:
735762.
ObermeyerA.A.1966:Juncaginaceae.Pp.9295in:CoddL.E.&deWinterB.(ed.),Floraof
SouthernAfrica1.Pretoria:BotanicalResearchInstitute.
PauC.1914:Triglochinbarrelierivar.maurum.In:FontQuerP.,PlantasdeLarache.Bol.RealSoc.
Esp.Hist.Nat.14:424429.
PetersenG.,SebergO.,DavisJ.I.&StevensonD.W.2006:RNAeditingandphylogenetic
reconstructionintwomonocotmitochondrialgenes.Taxon55:871886.
PetrovaL.R.1985:Juncaginaceae.Pp.4951in:TakhtajanA.(ed.),Anatomiaseminum
comparativa1.Leningrad:Nauka.
PhilbrickC.T.1988:Evolutionofunderwateroutcrossing.Fromaerialpollinationsystems:a
hypothesis.Ann.MissouriBot.Gard.75:836841.
PignattiE.&PignattiS.2005:EphemeralwetlandvegetationofWesternAustralia.Phytocoenologia
35:201218.
PisanoE.1983:TheMagellanictundracomplex.Pp.295329in:GoreA.P.J.(ed.),Ecosystemsof
theWorld4B,Mires:Swamp,Bog,FenandMoor.Amsterdam:Elsevier.
PliskoM.A.1985:Aponogetonaceae,Scheuchzeriaceae.Pp.4448in:TakhtajanA.(ed.),Anatomia
seminumcomparative1.Leningrad:Nauka
PojarJ.1974:ReproductivedynamicsoffourplantcommunitiesofsouthwesternBritishColumbia.
Canad.J.Bot.52:18191834.
PosadaD.2008:jModelTest:phylogeneticmodelaveraging.Mol.Biol.Evol.25:12531256[online
athttp://dx.doi.org/10.1093/molbev/msn083].
PosadaD.&BuckleyT.R.2004:Modelselectionandmodelaveraginginphylogenetics:advantages
oftheAICandBayesianapproachesoverlikelihoodratiotests.Syst.Biol.53:793808.
PosadaD.&CrandallK.A.1998:Modeltest:testingthemodelofDNAsubstitution.Bioinformatics
14:817818.
PoslusznyU.1981:UnicarpellatefloraldevelopmentinPotamogetonzosteriformis.Can.J.Bot.59:
495504.
PoslusznyU.1983:ReevaluationofcertainkeyrelationshipsintheAlismatidae:Floralorganogenesis
ofScheuchzeriapalustris(Scheuchzeriaceae).Am.J.Bot.70:925933.
References
160
PoslusznyU.&CharltonW.A.1993:Evolutionofthehelobialflower.Aquat.Bot.44:303324.
PoslusznyU.,CharltonW.A.&JainD.K.1986:Morphologyanddevelopmentofthereproductive
shootsofLilaeascillodes(Poir.)Hauman(Alismatidae).Bot.J.Linn.Soc.92:323342.
PoslusznyU.,CharltonW.A.&LesD.H.2000:Modularityinhelobialflowers.Pp.6374in:Wilson
K.L.&MorrisonD.A.(ed.),Monocots:systematicsandevolution.Collingswood:CSIRO.
PoslusznyU.&SattlerR.1973:FloraldevelopmentofPotamogetondensus.Can.J.Bot.51:647
656.
PoslusznyU.&SattlerR.1974:FloraldevelopmentofPotamogetonrichardsonii.Am.J.Bot.61:
209216.
PostG.E.1933:Juncaginaceae.P.539in:Fl.Syria,ed.2.Beirut:AmericanPress.
PrychidC.J.&RudallP.J.1999:Calciumoxalatecrystalsinmonocotyledons:areviewoftheir
structureandsystematics.Ann.Bot.84:725739.
RainhaB.V.1944:ATriglochinlaxifloraGuss.emPortugal.Bol.Soc.Broter.19:289294.
RambautA.2006:FigTree,version1.3.1.Computerprogramanddocumentationdistributedbythe
author,website:http://beast.bio.ed.ac.uk/FigTree.
RambautA.&DrummondA.J.2007:Tracer,version1.5.MCMCTraceFileAnalyser.Computer
programanddocumentationdistributedbytheauthors,website:
http://beast.bio.ed.ac.uk/Tracer.
RambautA.&DrummondA.J.2009:TreeAnnotator,version1.7.4.Computerprogramand
documentationdistributedbytheauthor,website:http://beast.bio.ed.ac.uk/TreeAnnotator.
RavenP.H.&AxelrodD.I.1974:Angiospermbiogeographyandpastcontinentalmovements.Ann.
MissouriBot.Gard.61:539673[onlineathttp://www.jstor.org/stable/2395021].
RaziB.A.1949:EmbryologicalstudiesoftwomembersofthePodostemaceae.Bot.Gaz.111:211
218.
ReicheK.1907:GrundzgederPflanzenverbreitunginChile.In:EnglerA.&DrudeO.(ed.),Die
VegetationderErde8.Leipzig:W.Engelmann[onlineat
RemizowaM.V.,KuznetsovA.N.,KuznetsovaS.P.,RudallP.J.,NuralievM.S.&SokoloffD.D.2012a:
FlowerdevelopmentandvasculatureinXyrisgrandis(Xyridaceae,Poales):acasestudyfor
examiningpetaldiversityinmonocotflowerswithadoubleperianth.Bot.J.Linn.Soc.170:93
References
161
RemizowaM.V.,RudallP.J.,ChoobV.V.&SokoloffD.D.2013a(inpress):Racemoseinflorescences
ofmonocots:structuralandmorphogeneticinteractionattheflower/inflorescencelevel.Ann.
Bot.[onlineathttp://dx.doi.org/10.1093/aob/mcs246].
RemizowaM.V.,SokoloffD.D.,CalvoS.,TomaselloA.&RudallP.J.2012b:Flowersand
inflorescencesoftheseagrassPosidonia(Posidoniaceae,Alismatales).Am.J.Bot.99:15921608
[onlineathttp://dx.doi.org/10.3732/ajb.1200227].
RemizowaM.V.,SokoloffD.D.,CampbellL.M.,StevensonD.W.&RudallP.J.2011:Harperocallisis
congenericwithIsidrogalvia(Tofieldiaceae,Alismatales):Evidencefromcomparativefloral
morphology.Taxon60:10761094.
RemizowaM.V.,SokoloffD.D.,RudallP.J.2010:Evolutionaryhistoryofthemonocotflower.Ann.
MissouriBot.Gard.97:617645.
RemizowaM.V.,SokoloffD.D.&RudallP.J.2013b:Differentpatternsofbractreductionin
racemoseinflorescencesofbasalmonocots.Pp.185207in:WilkinP.&MayoS.J.(ed.),Early
eventsinmonocotevolution.Cambridge:CambridgeUniversityPress.
RennerS.S.2004:PlantdispersalacrossthetropicalAtlanticbywindandseacurrents.Int.J.Plant
Sci.165:S23S33.
RevealJ.L.1977:Juncaginaceae.Pp.1822in:HolmgrenA.H.,HolmgrenN.H.,RevealJ.L.&
HolmgrenP.K.(ed.),IntermountainFloraVascularPlantsoftheIntermountainWest,U.S.A.6.
NewYork:ColumbiaUniversityPress.
RevealJ.L.2011:Summaryofrecentsystemsofangiospermclassification.KewBull.66:548.
RevealJ.L.&ChaseM.W.2011:APGIII:BibliographicalinformationandsynonymyofMagnoliidae.
Phytotaxa19:71134[onlineat
http://www.mapress.com/phytotaxa/content/2011/f/pt00019p134.pdf].
RichardL.C.M.1808:DmonstrationsBotaniquesouanalysedufruit[Onlineat
http://biodiversitylibrary.org/page/34447568]
RichardL.C.M.1815:Propositiond'unenouvellefamilledesplantes,IesButomes.Mm.Mus.
Hist.Nat.1:365[onlineathttp://www.biodiversitylibrary.org/page/33882754].
RobbL.&LadigesP.Y.1981:MorphologicalformsandpolyploidyinTriglochinproceraagg.in
Victoria.Austr.J.Bot.29:639651.
RpertD.(ed.)2000+:DigitalspecimenimagesattheHerbariumBerolinense.Publishedat
http://ww2.bgbm.org/herbarium/[lastaccessed26Oct2012].
RomanowskiN.1998:AquaticandWetlandPlants.AFieldGuidefornontropicalAustralia.Sydney:
UniversityofNewSouthWalesPress.
References
162
RomeroBujnM.I.2007:FloraexoticadeGalicia(noroesteibrico).BotnicaComplutensis31:
113125.
RonquistF.1997:Dispersalvicarianceanalysis:Anewapproachtothequantificationofhistorical
biogeography.Syst.Biol.46:195203.
RourkeJ.P.1972:TaxonomicstudiesonLeucospermumR.Br.J.S.Afr.Bot.8(Suppl.):1194.
RouyG.1912:Juncaginaces.Pp.268272in:RouyG.&FoucaudJ.(ed.),Fl.France13.Paris:
Deyrolle[onlineathttp://biodiversitylibrary.org/page/12313405].
RudallP.J.1997:Thenucellusandchalazainmonocotyledons:structureandsystematic.Bot.Rev.
63:140181.
RudallP.J.2003:Monocotpseudanthiarevisited:floralstructureofthemycoheterotrophicfamily
Triuridaceae.Int.J.PlantSci.164:S307S320.
RuizE.2008:ManagementofNatura2000habitats.3170*Mediterraneantemporaryponds.
TechnicalReport.EuropeanCommission.
RuthsatzB.&VillagrnC.1991:VegetationpatternandnutrientecologyofaMagellanicMoorland
ontheCordilleraPiuchuinChilo/Chile.RevistaChilenaHist.Nat.64:461478[onlineat
http://rchn.biologiachile.cl/pdfs/1991/3/Ruthsatz_&_Villagran_1991.pdf].
SaintyG.R.&JacobsS.W.L.2003:WaterplantsofAustralia.PottsPoint,Australia:Saintyand
Associates.
SandersonM.J.1991:Insearchofhomoplastictendencies:statisticalinferenceoftopological
patternsinhomoplasy.Evolution45:351358.
SattlerR.1965:PerianthdevelopmentofPotamogetonrichardsonii.Am.J.Bot.52:3541.
ScherH.D.&MartinE.E.2006:TimingandclimaticconsequencesoftheopeningofDrakePassage.
Science312:428430[onlineathttp://dx.doi.org/10.1126/science.1120044].
Schluter,D.,2009.Evidenceforecologicalspeciationanditsalternative.Science323:737741
[onlineathttp://dx.doi.org/10.1126/science.1160006].
SchmidtS.R.,KleinebeckerT.,VogelA.&HlzelN.2010:Interspecificandgeographicaldifferences
ofplanttissuenutrientconcentrationsalonganenvironmentalgradientinSouthernPatagonia,
Chile.Aquat.Bot.92:149156[onlineathttp://dx.doi.org/10.1016/j.aquabot.2009.11.002].
SchmiedelU.&JrgensN.1999:Communitystructureonunusualhabitatislands:quartzfieldsinthe
SucculentKaroo,SouthAfrica.PlantEcol.147:5769.
References
163
SchneiderE.L.&CarlquistS.1997:Originsandnatureofvesselsinmonocotyledons.2.
JuncaginaceaeandScheuchzeriaceae.Nord.J.Bot.17:397401.
SchumannC.1894:Lilaeaceae.Pp.669676in:MartiusC.F.P.,EichlerA.W.&UrbanI.(ed.),Flora
Brasiliensis3(3).Mnchen&Leipzig:F.Fleischer.
SchusterM.,DuringerP.,GhienneJ.F.,VignaudP.,MackayeH.T.,LikiusA.&BrunetM.2006:The
ageoftheSaharadesert.Science311:821[onlineat
http://dx.doi.org/10.1126/science.1120161].
ScottJ.K.&PanettaF.D.1993:PredictingtheAustralianweedstatusofsouthernAfricanplants.J.
Biogeogr.20:8793.
SebergO.2007:Juncaginaceae.Pp.377378in:HeywoodV.N.,BrummittR.K.,CulhamA.&
SebergO.(ed.),Floweringplantfamiliesoftheworld.Ontario,Canada:FireflyBooks.
ShafferFehreM.1987:SeedandtestastructureinrelationtothetaxonomyoftheAlismatidae.
PhDthesis.London:DepartmentofBiology,KingsCollege.
ShamrovI.I.2008:Ovuleoffloweringplants:structure,functions,origin.Moscow:KMKScientific
Press.
ShipunovA.B.2003:Thesystemoffloweringplants:synthesisofclassicalandmolecularapproaches.
J.GeneralBiol.64:499507.
SinghV.1965:MorphologicalandanatomicalstudiesinHelobiae.II.Vascularanatomyoftheflower
ofPotamogetonaceae.Bot.Gaz.126:137144.
SinghV.&SattlerR.1977:FloraldevelopmentofAponogetonnatansandA.undulatus.Can.J.Bot.
55:11061120.
SmithV.,RycroftS.,BrakeI.,ScottB.,BakerE.,LivermoreL.,BlagoderovV.&RobertsD.2011:
Scratchpads2.0:aVirtualResearchEnvironmentsupportingscholarlycollaboration,
communicationanddatapublicationinbiodiversityscience.ZooKeys150:5370[onlineat
http://dx.doi.org/10.3897/zookeys.150.2193].
SobelJ.M.,ChenG.F.,WattL.R.&SchemskeD.W.2010:Thebiologyofspeciation.Evolution64:
SokoloffD.D.,RemizowaM.V.,LinderH.P.&RudallP.J.2009:Morphologyanddevelopmentofthe
gynoeciuminCentrolepidaceae:themostremarkablerangeofvariationinPoales.Am.J.Bot.
96:19251940.
SokoloffD.D.,RemizowaM.V.&RudallP.J.2013:Issyncarpyanancestralconditioninmonocots
andcoreeudicots?Pp.6081in:WilkinP.&MayoS.J.(ed.),Earlyeventsinmonocotevolution.
Cambridge:CambridgeUniversityPress.
References
164
SokoloffD.D.,RudallP.J.&RemizowaM.V.2006:Flowerliketerminalstructuresinracemose
inflorescences:atoolinmorphogeneticandevolutionaryresearch.J.Exp.Bot.57:35173530.
SonderO.W.1856:PlantaeMuellerianae,Juncagineae.Linnaea28:223225.
SoltisD.E.,SoltisP.S.,EndressP.K.&ChaseM.W.2005:Phylogenyandevolutionofangiosperms.
Sunderland:Sinauer.
StafleuF.A.&CowanR.S.19761988:TaxonomicLiterature:Aselectiveguidetobotanical
publicationsandcollectionswithdates,commentariesandtypes(TL2).Utrecht:Bohn,
Scheltema&Holkema.7vols.,2ded.[Publishedat:http://www.sil.si.edu/digitalcollections/tl2].
StamatakisA.,HooverP.&RougemontJ.2008:ARapidBootstrapAlgorithmfortheRAxMLWeb
Servers.Syst.Biol.75:758771.
StevensP.F.2001+:AngiospermPhylogenyWebsite.Publishedat
http://www.mobot.org/MOBOT/research/APweb[accessed26June2012].
StevensonD.W.&LoconteH.1995:Cladisticanalysisofmonocotfamilies.Pp.543578in:Rudall
P.J.,CribbP.J.,CutlerD.F.&HumphriesC.J.(ed.),Monocotyledons:systematicsandevolution.
Kew:RoyalBotanicGardens.
StickleyC.E.,BrinkhuisH.,SchellenbergS.A.,SluijsA.,RhlU.,FullerM.,GrauertM.,HuberM.,
WarnaarJ.&WilliamsG.L.2004:TimingandnatureofthedeepeningoftheTasmanianGateway.
Paleoceanography19:PA4027[onlineathttp://dx.doi.org/10.1029/2004PA001022].
SunK.,ZhangZ.Y.&ChenJ.K.2000:FloralorganogenesisofPotamogetondistinctusA.Benn.
(Potamogetonaceae).ActaPhytotax.Sin.38:528531.
SwoffordD.L.2002:PAUP*.PhylogeneticAnalysesUsingParsimony(*andOtherMethods).Version
4.Sunderland:SinauerAssociates.
SwoffordD.L.2003:PAUP*.Phylogeneticanalysisusingparsimony(*andothermethods),v.4.0beta
10.Sunderland:SinauerAssociates.
TakhtajanA.1966:Asystemandphylogenyofthefloweringplants.Moscow,Leningrad:Nauka.
TakhtajanA.1987:SystemaMagnoliophytorum.Leningrad:Nauka.
TakhtajanA.1997:Diversityandclassificationoffloweringplants.NewYork:ColumbiaUniversity
Press.
TakhtajanA.2009:Floweringplants.Ed.2.NewYork:Springer.
TalaveraS.1987:Juncaginaceae.Pp.186187in:ValdsB.,TalaveraS.&FernndezGalianoE.
(ed.),FloraVasculardeAndalucaOccidental3.Barcelona:Ketres.
References
165
TalaveraS.2010:Triglochin.Pp.4451in:TalaveraS.(ed.),FloraIberica17.Madrid:RealJardn
Botnico,CSIC.
TalaveraS.,OrtizP.L.,AristaM.&BastidaF.1995:Estudiocariosistematicodealgunas
monocotiledoneasbulbosasdeMarruecos.Lagascalia18:83104.
TaylorN.1909:Lilaeaceae.P.37in:NorthAmericanFlora17.NewYork:NewYorkBotanic
Garden.
TeryokhinE.S.1985:Hydrocharitaceae,Potamogetonaceae,Ruppiaceae.Pp.3843,5155in:
TakhtajanA.,ed.Anatomiaseminumcomparativa.T.1.Leningrad:Nauka.
ThieretJ.W.1988:TheJuncaginaceaeofthesoutheasternUnitedStates.J.ArnoldArbor.69:123.
ThivM.,vanderNietT.,RutschmannF.,ThulinM.,BruneT.&LinderH.P.2011:OldNewWorldand
transAfricandisjunctionsofThamnosma(Rutaceae):intercontinentallongdistancedispersaland
localdifferentiationinthesucculentbiome.Am.J.Bot.98:7687[onlineat
http://dx.doi.org/10.3732/ajb.1000339].
ThompsonJ.1961:Juncaginaceae.Pp.7780in:FloraofNewSouthWales16.Contributionsfrom
theNewSouthWalesNationalHerbarium.FloraSeries118.Sydney:NewSouthWalesNational
Herbarium.
ThompsonJ.D.,GibsonT.J.,PlewniakF.,JeanmouginF.&HigginsD.G.1997:TheClustalXwindows
interface:flexiblestrategiesformultiplesequencealignmentaidedbyqualityanalysistools.
NucleicAcidsRes.24:48764882.
ThomsonJ.A.2002:AnimprovednoncryogenictransportandstoragepreservativefacilitatingDNA
extractionfromdifficultplantscollectedatremotesites.Telopea9:755760.
TiffneyB.H.&ManchesterS.R.2001:Theuseofgeologicalandpaleontologicalevidencein
evaluatingplantphylogeographichypothesesintheNorthernHemisphereTertiary.Int.J.Pl.Sci.
162(6Suppl.):S3S17[onlineathttp://www.jstor.org/stable/10.1086/323880].
TomlinsonP.B.1982:Helobiae(Alismatidae),includingtheseagrasses.Pp.1559in:MetcalfeC.R.
(ed.),Anatomyofmonocotyledons7.Oxford:ClarendonPress.
TrinderSmithT.H.2003:TheLevynsguidetotheplantgeneraofthesouthwesternCape.Contrib.
Bol.Herb.21:1355.
TurpieJ.K.2004:SouthAfricanNationalBiodiversityAssessment2004:TechnicalReport3:Estuary
Component.Report.Pretoria:SANBI.
TurpieJ.K.,AdamsJ.B.,JoubertA.,HarrisonT.D.,CollotyB.M.,MareeR.C.,WhitfieldA.K.,
WooldridgeT.H.,LamberthS.J.,TaljaardS.&vanNiekerkL.2002:Assessmentofthe
References
166
conservationprioritystatusofSouthAfricanestuariesforuseinmanagementandwater
allocation.WaterSA28:191206.
TzvelevN.N.2013:Triglochin.In:ElvenR.(ed.),PanarcticFlora.AnnotatedChecklistofthe
PanarcticFlora(PAF).Publishedathttp://nhm2.uio.no/paf/2101[accessed20Apr2013].
UhlN.1947:StudiesinthefloralanatomyandmorphologyofcertainmembersoftheHelobiae.
Ph.D.Thesis.Ithaca:CornellUniversity.
UKOTsOnlineHerbarium2011:Publishedathttp://herbaria.plants.ox.ac.uk/bol/UKOT[accessed26
Oct2012].
UngV.,DubusG.,ZaragetaBagilsR.&VignesLebbeR.2010:Xper2:introducingetaxonomy.
Bioinformatics26:703704[onlineathttp://dx.doi.org/10.1093/bioinformatics/btp715].
UotilaP.1984:Juncaginaceae.Pp.1315in:DavisP.H.(ed.),FloraofTurkeyandtheEastAegaen
Islands8.Edinburgh:EdinburghUniversityPress.
UpchurchP.2008:Gondwananbreakup:legaciesofalostworld?TrendsEcol.Evol.23:229236
[onlineathttp://dx.doi.org/10.1016/j.tree.2007.11.006].
vanderNietT.S.,JohnsonS.D.&LinderH.P.2006:Macroevolutionarydatasuggestarolefor
reinforcementinpollinationsystemshifts.Evolution60:15961601.
vanSteenisC.G.G.J.1949:Juncaginaceae.P.57in:vanSteenisC.G.G.J.(ed.),FloraMalesiana4.
Djakarta:NoordhoffKolffN.V.
vanWykA.E.&SmithG.F.2001:RegionsoffloristicendemisminSouthernAfrica.Areviewwith
emphasisonsucculents.Pretoria:UmdausPress.
vonMeringS.2011:DieGattungTriglochin(Dreizack,Juncaginaceae)inEuropa.Publishedathttp:/
/www.offenenaturfuehrer.de/wiki/Die_Gattung_Triglochin_(Dreizack,_Juncaginaceae)_in_
Europa_(Sabine_von_Mering).
vonMeringS.2012+:JuncaginaceaeoftheWorld.Publishedat
http://juncaginaceae.myspecies.info/.
vonMeringS.2013:TetronciumanditsonlyspeciesT.magellanicum(Juncaginaceae):distribution,
ecologyandlectotypification.Willdenowia43:1324[onlineat
http://dx.doi.org/10.3372/wi.43.43102].
vonMeringS.&KadereitJ.W.2010:Systematics,phylogeny,andrecircumscriptionofJuncaginaceae
acosmopolitanwetlandfamily.Pp.5579in:SebergO.,PetersenG.,BarfodA.S.&DavisJ.I.
(ed.),Diversity,phylogeny,andevolutioninthemonocotyledons.Proceedingsofthefourth
internationalconferenceonthecomparativebiologyofthemonocotyledonsandthefifth
internationalsymposiumongrasssystematicsandevolution.Aarhus:AarhusUniversityPress.
References
167
WaceN.W.1961:ThevegetationofGoughIsland.Ecol.Monogr.31:337367[onlineat
http://www.jstor.org/stable/1950757].
WeineltM.1996+:OnlineMapCreation,v.4.1.Onlineathttp://www.aquarius.ifmgeomar.de.
WentF.A.F.C.1908:Thedevelopmentoftheovule,embryosacandegginPodostemaceae.Pp.
824832in:KoninklijkeNederlandscheAkademievanWetenschappenProceedings10(2):1907
1908.Amsterdam.
WesternAustralianHerbarium1998+:FloraBase:TheWesternAustralianFlora.Perth:Department
ofEnvironmentandConservation.Publishedathttp://florabase.dec.wa.gov.au/.
WettsteinR.1924:HandbuchderSystematischenBotanik.3Aufl.Leipzig,Wien:F.Deuticke.
WilldenowC.G.1808:NhereBestimmungeinigerLiliengewchse.Mag.NeuestenEntdeck.
GesammtenNaturk.Ges.Naturf.FreundeBerlin2:1430.
WilliamsI.J.M.1972:ArevisionofthegenusLeucadendron(Proteaceae).Contrib.BolusHerb.3:
1425.
WonH.&RennerS.S.2006:DatingdispersalandradiationinthegymnospermGnetum.Syst.Biol.
55:610622.
YamashitaT.1970:EigenartigeWurzelanlagedesEmbryosbeiLilaeasubulataHumb.etBonpl.und
TriglochinmaritimumL.J.Fac.Sci.Univ.Tokyo,sect.III,Bot.10:181205.
YuY.,HarrisA.J.&HeX.J.2010:SDIVA(statisticaldispersalvicarianceanalysis):atoolforinferring
biogeographichistories.Mol.Phylogenet.Evol.56:848850[onlineat
http://dx.doi.org/10.1016/j.ympev.2010.04.011].
YuY.,HarrisA.J.&HeX.J.2012:RASP(ReconstructAncestralStateinPhylogenies),v2.1b.
Availableathttp://mnh.scu.edu.cn/soft/blog/RASP.
ZuloagaF.O.,MorroneO.&BelgranoM.J.(ed.)2008:Juncaginaceae.In:CatlogodelasPlantas
VascularesdelConoSur(Argentina,SurdeBrasil,Chile,ParaguayyUruguay).I.Pteridophyta,
GymnospermaeyMonocotyledoneae.Monogr.Syst.Bot.MissouriBot.Gard.107:465466.

References
168

Appendices
169
APPENDICES
Appendix1:Listofaccessions(Chapter1)
Appendix2:Voucherinformation(Chapter2)
Appendix3:PCRinformation(Chapter2)
Appendix4:ListofstudiedspecimensofTetronciummagellanciumWilld.(Chapter4)
Appendix5:ListofstudiedspecimensofMaundiatriglochinoidesF.Muell.(Chapter5)
Appendix6:IdentificationkeysforTriglochininEurope

Appendices
170
Appendices
171
APPENDIX1.
List of Juncaginaceae and other Alismatales species used in this study (families given in bold).
Accessioninformationislistedasfollows:speciesname;voucherspecimen;andGenBank accession
numbers (rbcL, matK, and atpA) [with placeholder taxa in parenthesis]. Voucher specimen infor
mationisgivenonlyfornewlyobtainedsequences,indicatedbyboldaccessionnumbers.Herbarium
abbreviationsarefromHolmgrenandHolmgren(1998).indicatesmissingsequencedata.
Acoraceae: Acorus calamus L., AJ879453, AB040154, AF039256. Alismataceae: Alisma plantago
aquatica L., L08759, AF542573, . Caldesia oligococca (F.Muell.) Buchanan, AY277799, AY952427,
AY277800. Hydrocleys nymphoides (Humb. & Bonpl. ex Willd.) Buchenau, U80716, AB002580, .
Limnocharis flava (L.) Buchenau, U80717, AB088778, . Sagittaria latifolia Willd. , , AY299832.
Aponogetonaceae: Aponogeton fenestralis (Pers.) Hook.f., AB088808, AB088779, . Araceae:
Arisaema triphyllum (L.) Torr., AJ005629, AF3877428 (A. tortuosum (Wall.) Schott), AY299717.
Gymnostachys anceps R.Br., AB088806, AB040177, AF039244. Orontium aquaticum L., AJ005632,
AF543744, AY299816. Butomaceae: Butomus umbellatus L., U80685, AY870364, AY299733.
Cymodoceaceae: Amphibolis antarctica (Labill.) Asch., U80686, , . Cymodocea serrulata (R.Br.)
Asch.&Magnus,U80715,,AY277801.Haloduleuninervis(Forssk.)Boiss.,AY952436,AY952424,.
Hydrocharitaceae: Hydrocharis dubia (Blume) Backer, AB004892, AB002572, . Limnobium
laevigatum(Humb.&Bonpl.exWilld.)Heine,AB004894,AB002574,.NajasmarinaL.,U80705,,
. Ottelia acuminata (Gagnep.) Dandy, AY952435, AY952432, AY277802. Stratiotes aloides L.,
U80709,AB002576,.VallisneriaamericanaMichx.,U03726,,.Juncaginaceae:Lilaeascilloides
(Poir.) Hauman, USA, California, Moore s.n. (MJG), U80715, GQ452345, GQ452348. Maundia
triglochinoides F.Muell., Australia, S. Jacobs 9453 (MJG, NSW), GQ452330, GQ452347, GQ452349.
TetronciummagellanicumWilld.,Argentina,A.Vogels.n.(MJG),GQ452337,GQ452346,
GQ452351.TriglochinbarrelieriLoisel.,Italy,C.Uhinks.n.(MJG),GQ452331,GQ452342,GQ452352.
Triglochin elongata Buchenau, South Africa, P. Vargas 537PV00 (MJG), GQ452332, GQ452343,
GQ452353. Triglochin laxiflora Guss., Italy, S. von Mering s.n. (MJG), , , GQ452354. Triglochin
maritimaL.,Turkey,D.Albach&F.zgokce912(MJG),GQ452333,GQ452339,GQ452355.Triglochin
palustris L., Russia, P. Schnswetter & A.Tribsch T145 (WU), GQ452334, GQ452340, . Triglochin
rheophilaAston,Australia,S.Jacobs9392(MJG,NSW),GQ452335,GQ452344,GQ452356.Triglochin
striata Ruiz & Pav., Australia, N. Schmalz s.n. (MJG), GQ452336, GQ452341, GQ452357.
Posidoniaceae: Posidonia oceanica (L.) Delile, U80719, , . Potamogetonaceae: Potamogeton
distinctus A.Benn., AB088809, AB088780, AY299829 (P. natans L.). Ruppiaceae: Ruppia maritima L.,
U03729,,.Scheuchzeriaceae:ScheuchzeriapalustrisL.,C.Uhinks.n.(MJG),U03728,GQ452338,
GQ452350. Tofieldiaceae: Pleea tenuifolia Michx., AJ131774, AF465301, AY299827. Tofieldia
calyculata (L.) Wahlenb., AB183410, AB183403, AY299851. Zosteraceae: Heterozostera tasmanica
(M.Martens ex Asch.) Hartog, U80730, AB096171, . Phyllospadix torreyi S.Watson, U80731,
AB096172 (P. iwatensis Makino), . Zostera marina L., U80734, , . Z. noltii Hornem., U80733,
AB096170,.

Appendices
172
Appendices
173
APPENDIX2.
VoucherinformationandGenBankaccessionnumbersforsequencesanalysedinthisstudy.
Taxon Voucher[Collector,collector
no.,herbariumcode]
Origin ITS psbA
trnH
matK
T.sp.A(Floraof
Australia)
Jacobs4147a(NSW) Australia x
T.barrelieriLoisel. v.Merings.n.[2006](MJG) Italy,Sardinia x x x
T.barrelieriLoisel. Spain,Mallorca
T.barrelieriLoisel. Uhinks.n.[2005](MJG) Italy,Sicily x
T.barrelieriLoisel. Deils.n.[1990](FB) Spain x
T.barrelieriLoisel. Mateoetal.4824/95 (RNG) Morocco x x
T.barrelieriLoisel. Sekegin&Sokoloff [2004]
(NW)
Cyprus x
T.barrelieriLoisel. Westberg17/01/Gr (MJG) Greece x
T.barrelieriLoisel. Poelt&Poelt207 (M) Portugal x
T.buchenauiKcke,
Mering&Kadereit
Kcke&Steffen[2006]
(MJG)
SouthAfrica
T.bulbosaL. Mucina&Jakubowsky
040806/02(MJG)
SouthAfrica x
T.bulbosaL. Kcke&Mucina030406/
22(MJG)
SouthAfrica x
T.bulbosaL. Kcke&Mucina
140406/01(MJG)
SouthAfrica x
T.bulbosaL. Mucina&Jakubowsky
040806/05(MJG)
SouthAfrica x
T.bulbosaL. Kcke&Mucina
140406/33(MJG)
SouthAfrica x
T.bulbosaL. Kcke&Steffen190406/02
(MJG)
SouthAfrica x
T.bulbosaL. Kcke&Steffen190406/28
(MJG)
SouthAfrica x
T.bulbosaL.ssp.bulbosa Linder7076(MJG) SouthAfrica x x
T.bulbosaL.ssp.calcicola
Mucina251008/05 (MJG) SouthAfrica x x
Mucina&Merunkova
280908/04(MJG)
SouthAfrica x x
MerunkovaKM50/34 (MJG) SouthAfrica x
T.bulbosaL.ssp.
quarcicolaMering,Kcke
&Kadereit
Schmiedel124811 (MJG) SouthAfrica x x
T.calcitrapaHook. Wilson1860a(B) Australia x x
T.calcitrapaHook. Greuter18564(B) Australia x
T.calcitrapaHook. Michell&Rister1769 (B) Australia x
T.centrocarpaHook. Greuter20973(B) Australia x x
T.cf.bulbosa Wrz04.10.06.01 (STU) Australia x x
Appendices
174
T.compactaAdamson Nauds.n.(MJG) SouthAfrica x
T.compactaAdamson Goldblatt1792(C) SouthAfrica x
T.concinnaBurttDavy Rittenhouse1937 (OSC) USA,Oregon x x
T.concinnaBurttDavy Ackermann95(B) Chile x
T.concinnassp.debilis
(M.E.Jones)J.T.Howell
Halse2383(C) USA x
Tiehm11558(OSC) USA,Nevada x x
Wright1807(C) USA,Oregon x
T.elongataBuchenau Vargas537PV00 (MJG) SouthAfrica x x x
T.gaspensisLieth&
D.Lve
Teeges.n.[2005](MJG) USA x
T.hexagonaJ.M.Black Latz8458(NSW) Australia x x
T.laxifloraGuss. v.Merings.n.(MJG)
Sard5129/131_Sardinia
Italy,Sardinia x x
T.laxifloraGuss. Mifsuds.n.[2007] (MJG) Malta x x
T.laxifloraGuss. Rosellos.n.(MJG) Spain,Mallorca x
T.laxifloraGuss. Rosellos.n.(MJG) Spain,Mallorca x
T.maritimaL. Lambrachts.n.(MJG) Germany x
T.maritimaL. v.Merings.n.[2006](MJG) Germany x
T.maritimaL. Hhns.n.[2005](MJG) Hungary x x
T.maritimaL. seeds(BGGreifswald) Mongolia x
T.maritimaL. Nesbitt1178(RNG) Turkey x x
T.maritimaL. Mehregans.n.[2002](MJG) Iran x
T.maritimaL. Strausss.n.(GFW) Mongolia x x
T.maritimaL. Ito&Tanaka5403 (MJG) Japan x x
T.maritimaL. Masudas.n.[2006](MJG) Japan x
T.maritimaL. Itoetal.2983(MJG) SouthKorea x
T.maritimaL. Larson11383(OSC) USA,SouthDakota x x
T.maritimaL. Porsild&Weber22933 (C) USA,Colorado x x
T.maritimaL. Tiehm10714(OSC) USA,Nevada x x
T.maritimaL. Coope35(OSC) USA,Oregon x
T.maritimaL. Pierce1843(C) USA x
T.maritimaL. DQ786518 x
T.maritimaL. Garneaux&Tailleuxs.n.(O) Canada,Quebec x
T.maritimassp.asiatica
Kitag.
Ignatov32(MW) Russia,Sakhalin x x
T.milneiHorn Bidgood,Mbago&Vollesen
2603(P)
Tanzania x x
T.palustrisL. Franklin7426(OSC) USA,Utah x
T.palustrisL. Strauss[2002](GFW) Mongolia x x
T.palustrisL. Schnswetter&Tribsch
T145(WU)
Russia x x
T.palustrisL. deLange[1995](AK) NewZealand x x
T.palustrisL. Zndorf21382(JE) Chile x x
T.palustrisL. Schmalzs.n.[1997] (MJG) Germany x x
T.palustrisL. cult.,BGMainz Germany x
Appendices
175
T.scilloides(Poir.)Mering
&Kadereit
Moores.n.[2006](MJG) California x x x
T.scilloides(Poir.)Mering
&Kadereit
Feuerer9524a(HBG) Bolivia x
T.striataRuiz&Pav. GlassonPD90/17 (OSC) NewZealand x x
T.striataRuiz&Pav. MacmillanBH98/32 (CHR) NewZealand x
T.striataRuiz&Pav. Gibbons314(C) Australia x x
T.striataRuiz&Pav. Schmalzs.n.(MJG) Australia x x x
T.striataRuiz&Pav. Pullen4322(C) Australia x
T.striataRuiz&Pav. MyndelPedersens.n.(C) Brazil x
T.striataRuiz&Pav. SAfrcon0705064 SouthAfrica x
T.striataRuiz&Pav. ManitzRCH71(JE) Chile x x
T.striataRuiz&Pav. Burnett389(OSC) USA,Washington x
T.striataRuiz&Pav. Leonard2109(C) USA,NorthCarolina x x
Cycnogetonmultifructum
(Aston)Mering&
Kadereit
JacobsSJ9397(NSW) Australia x x
Cycnogetonrheophilum
(Aston)Mering&
Kadereit
JacobsSJ9392(NSW) Australia x x x

Appendices
176

Appendices
177
APPENDIX3.
PCR information: primer sequences and origin, PCR ingredients and conditions. Notes: ITS, internal
transcribedspacer;BSA,bovineserumalbumine.
ThecompleteITSregion(ITS1,the5.8gene,ITS2)wasamplifiedasasinglepieceusingthestandard
primers18Sand28S(MuirandSchltterer,1999).ForsomeaccessionswithlowDNAquality,ITSwas
amplifiedintwooverlappingfragmentsusingtheprimerpairsITSA,ITSC,andITSB,ITSD(Blattner,
1999).
Marker Primersequence53 Source PCRingredients PCRconditions

ITS 18S:CCTTMTCATYTA
GAGGAAGGAG
Muiretal.2001 MgCl
2
0.48mM
Polymerase0.04
U/LdNTPs0.2
mM
Pretreatment:94C,
60s
Denaturation:94C,
18s
Annealing:52C,30s
ITS4:TCCTCCGCT
TATTGATATGC
Whiteetal.
1990
Primer0.8mM
BSA0.1mg/ml
Elongation:72C,60s
Posttreatment:52C,
78s,72C,8min
Cycles:35

psbA
trnH
psbAF:GTTATGCAT
GAACGTAATGCTC
Sangetal.1997 MgCl
2
0.48mM
Polymerase0.04
U/L
Pretreatment:94C,
60s
Denaturation:94C,
20s
Annealing:52C,30s
dNTPs0.2mM Elongation:72C,60s
Posttreatment:52C,
80s,72C,8min
Cycles:35
trnHR:CGCGCATGG
TGGATTCACAAATC
Sangetal.1997 Primer0.8mM
BSA0.8mg/ml

matK seevonMering&
Kadereit2010

Appendices
178
Appendices
179
APPENDIX4.
ListofstudiedspecimensofTetronciummagellanciumWilld.
ARGENTINA: PROVINCIA DE RO NEGRO: Departamento Bariloche: Parque Nacional Nahuel Huapi, Laguna Ortiz
Basualdo,terrenopantanoso,07.01.1952,O.Boelcke5366&M.N.Correa(SIimage!).ParqueNacionalNahuel
Huapi,PampaLinda,mllinAmer,highraisedbogaboveLagoMascardi,alt.1000m,08.12.1973,V.Markgraf
s.n. (P). Parque Nacional Nahuel Huapi, between Puerto Blest and Lago Frias, alt. 750 m, 1819.12.1978, S.
Laegaard 12533 (K). PROVINCIA SANTA CRUZ: Dpto. Lago Argentino: Brazo Norte, valle de la Cristina, ac de los
Cipresas,alestelagoPearson,alt.450m,17.2.1953,F.B.Vervoorst4496(NY,P).BrazoOnelli,mallnalNWdel
puesto abandonado Neumann, alt. 350 m, 26.2.1953, F. B. Vervoorst 4564 (NY). Mayo Glacier, Shipton
ExpeditiontoPatagonia(LagoArgentino)195859,insmallboginclearinginNothofagusbetuloideswood,near
the front of the glacier, 49 8051 S, 7273 30 W, alt. c. 400 ft., 30.01.1959, P. W. James 1500 (BM, SI
image!). PROVINCIA DE TIERRA DEL FUEGO, ANTRTIDA E ISLAS DEL ATLNTICO SUR: Departamento Ro Grande: Source
grande de Ro Grande. Alt. 400450 m, 06.03.1896, N. Alboff s.n. (SI image!). Nahe Lago Verde an Ruta B,
09.01.1989, J. Poelt s.n. (M). Departamento Ushuaia: St. Vincents Bay [Buen Thetis or Thetis Bay], Captain
Cooks first voyage, H.M.S. Endeavour, 17681771, in collibus subalpinus, 14.01.1769, J. Banks & D. Solander
s.n. (BM). Ushuaia, in a bog in the Nothofagus pumilio forest, above the town, alt. c. 300 m, 01.02.1940, R.
Santesson 396 (K, P). Lago Victoria, turbera, 07.11.1965, Luis Mendoza s.n. (BAB image!). Estancia Harberton,
Moat Bay, alt. c. 60100 ft., hill behind lake, growing in Astelia bog, 01.01.1966, R. N. P. Goodall 247 (RNG).
Estancia Harberton, Harberton swamp, alt. c. 60100 ft., in Sphagnum, deeply buried, only 23 ins. Showing,
01.02.1966,R.N.P.Goodall300(RNG).EstanciaHarberton,Harbertonswamp,NWendnearfence,elev.060
m,plantsgrowingingreenandreddishSphagnum,Sphagnumverywetandeasilybrokenapart,showingabove
moss, 28.12.1966, R. N. P. Goodall 443 (BAB image!, NY, RNG, SI image!). Estancia Harberton, Harberton
swamp, elev. 060 m, growing in Sphagnum roots, have haustoria of Nanodea mucosa, 29.04.1967, R. N. P.
Goodall795(P).BahaAguirre,c.5kmEofPuertoEspagnol,5453S6554W,3080m,boggygroundwith
Astelia.14.02.1968,D.M.Moore1831(K,RNG).RoLashifashajvalley,c.1kmNofLagunaVictoria,5448S
67 27 W, Sphagnum swamp, 01.03.1968, D. M. Moore 2082 (BAB image!, RNG). Estancia Harberton,
CambaceresBay,inswamp.03.12.1968,R.N.P.Goodall1882(RNG).LomaLargaForte,900ft.,inopenburned
woods,notgrazed, 22.01.1968, R. N. P. Goodall1309 (B). Hill to theNorth of abandoned settlement at Baha
Thetis, top of hill is Astelia formation with numerous small pools, each edged with Tetroncium, sterile at this
time,20.11.1969,R.N.P.Goodall2311(RNG,SIimage!).TierradelFuegoaustralis,RanchoHambre,5445S,
67 54 W, turbal sphagnoso, abierto, copiosamente, 140 m, 16.01.1970. H. Roivainen s.n. (RNG). Lapataia,
growing in Sphagnum swamp surrounding Laguna Negra, to the West of Ro Lapataia, 02.03.1970, R. N. P.
Goodall2424 (BAB image!, SI image!). Lapataia, LagunaNegra, in wet Sphagnum swamp,06.12.1970,R.N. P.
Goodall 2634 (RNG). Pink Mountain, edge of mountain torrent at SE edge of mountain and in Sphagnum
swampatbaseofmountain,2000to1500ft.,11.03.1971,R.N.P.Goodall3700(SGO).MitrePennsulaofIsla
Grande,BahaBuenSuceso,5448S6520W,14.10.1971,T.R.Dudley,R.N.P.Goodall&G.Crow272(BAB
image!). Tra Lapataia ed il lago Roca (ovest di Ushuaia), torbiera nella parte orientale della Laguna Negra, sui
cuscini di Sphagnum, Spedizione Scientifica Italiana Mares AMF in Patagonia, Terra del Fuoco ed Antartide
organizzatadalGruppoRicercheScientificheTecnicheSubacqueodiFirenze,23.01.1974,R.E.G.PichiSermolli
&P.Bizzarri7539(K,P,SIimage!).Ushuaia,WegzumNationalpark/caminoalParqueNacional,27.12.1976,P.
Seibert183etal./T.B.P.A.2213(BABimage!,M).AtPassoGaribaldi,alt.
750800m,26.31.01.1979,S.Laegaard13296(K).CordondelToroatHost.Alakush,WofUshuaia,alt.600m,
01.04.02.1979, S. Laegaard 13338 (K). Isla de los Estados, Baha Franklin, entra Caleta Croix y lago Gaona,
5452S6441W,abundantelocalmenteencharcosenturberadominadaporAstelia,10.12.1999,F.Biganzoli
699(SIimage!).
CHILE: IX REGIN DE LA ARAUCANA: Provincia de Malleco: Parque Nacional de Nahuelbuta, centro del parquet,
3743S 7302W, 1250 m, 28.12.1968, M. Ricardi 5697 & C. Marticorena 1858 (CONC image!, SI image!). X
REGIN DE LOS LAGOS: Provincia de Chilo: Isla Grande de Chilo, altiplanicie central (campaos), January 1902,
Appendices
180
M. Espinosa s.n. (SGO). Cordillera San Pedro, wet hollow in campaa, 15.11.1958, E. J. Godley 474a (BM).
CordilleraSanPedro,wethollowsinopencampaa,18.11.1958,E.J.Godley503(SGO).IslaGrandedeChilo,
Cordillera de Piuchn, 4222S 7359W, 640 m, 12.02.1983, C. Villagrn 4907 (CONC image!). Llicaldad, en
turbera esfagnosa, 4229S 735038W, 495 m, 18.02.2010, S. Pfanzelt 455 (CONC image!). Provincia de
Palena:5.9kmNjct.rds.toPalenaandChaitn,VillaSta.Lucia,boginNothofagusforest,560m,March1985,
T. F. Stuessy et al. 7170 (CONC image!). XI REGIN AYSN DEL GENERAL CARLOS IBEZ DEL CAMPO: Provincia de
Capitn Prat: [Messier Channel], Halt Bay [Baha Halt], 21.03.1868, R. O. Cunningham s. n. (K). [Sarmiento
Channel],PuertoBueno,08.12.1868,R.O.Cunninghams.n.(K).GlaciarHammick,4850S7413W,comn
enpantanos,30m,28.12.1967,S.F.Anliot6033(SGO).VillaOHiggins,CuencadelRoMosco,turbera,sub
cuenca del Ro Claro, 4827S 7228W, 785 m, 20.03.2003, N. Garca 21 (CONC image!). XII REGIN DE
MAGALLANES Y DE LA ANTRTICA CHILENA: Provincia de ltima Esperanza: [Sarmiento Channel], Puerto Bueno
(Puertobono),ExpditiondelaMagicienne,18761879,lieuxhumidesettourbeux,15.02.1877,L.Savatiers.n.
(P 2 sheets). [Wellington Island], Eden [Puerto Edn], Expdition de la Magicienne 18761879, 24.01.1879, L.
Savatiers.n.(K,P3sheets).CanalSmith,February1900,R.[s.c.,s.n.](SGO).CanalSmith[CanalSmyth,Smyth
Channel],January1924,M.Gusindes.n.(M).PuertoEden,WellingtonIs.,lowlandbog,nearsea,06.12.1958,E.
J. Godley 631a, 632b (BM). Baha del Indio, Lote San Isidro, Ro Yumbel, interior, en turbales herbaceous,
25.02.1973, E. Pisano V. 3968 (RNG). Seno Unin, N side, Ancn Sin Salida, 95 m, 529S 7321W, Y201, bog
withscatteredtrees,10.01.1976,O.Dollenz,D.M.Moore,E.PisanoV.&A.A.Saenz/T.B.P.A.876(RNG).Seno
Unin, 5225S 7335W, 13.01.1976, Ulrich Eskuche / T.B.P.A. 695 (BAB image!). Isla Piazzi, Caleta Ocasin,
Abra Leackeys Retreat, 150 m, 5144S 741W, S231, rocky hill summit, wet seepage area, 16.01.1976, O.
Dollenz,D.M.Moore,E.PisanoV.&A.A.Saenz/T.B.P.A.1003(BABimage!,RNG).IslaPiazzi,CaletaOcasin,
Abra Leackeys Retreat, 80m, 5144S, 741W, S231, coastal bog, Donatia/Tetroncium dominant, 18.01.1976,
O.Dollenz,D.M.Moore,E.PisanoV.&A.A.Saenz/T.B.P.A.1069(BABimage!,RNG).IslaRennelNorte,Canal
Smyth,5154S7412W,W231,bog,24.01.1976,O.Dollenz,D.M.Moore,E.PisanoV.&A.A.Saenz/T.B.P.A.
1152(BABimage!,RNG).IslaVidalGormaz,SenoNantuel,BahaMaraAnglica,30m,5153S7441W,W251,
Donatiabog,01.02.1976,O.Dollenz,D.M.Moore,E.PisanoV.&A.A.Saenz/T.B.P.A.1237(BABimage!,RNG).
IslaVidalGormaz,SenoNantuel,EsteroLobos,5153S7441W,W251,Donatiabog,04.02.1976,O.Dollenz,D.
M.Moore,E.PisanoV.&A.A.Saenz/T.B.P.A.1395(RNG).IslaVirtudes,CanalElis,PuertoVirtudes,5133S
7454W,P261,bog,poolwithorganicbottommaterial,10.02.1976,O.Dollenz,D.M.Moore,E.PisanoV.&A.
A.Saenz/T.B.P.A.1477(BABimage!,RNG).SenoltimaEsperanza,LagoAzul,Eside,5127S7318W,N201,
dryish bog, hummocks with some water between, 10.01.1977, D. M. Moore & E. Pisano V. / T.B.P.A. 1541
(RNG). Seno ltima Esperanza, Puerto Bella Vista, S shore, 5131S 7316W, P201, bog above coastal forest,
higher ground at margin of bog, 11.01.1977, D. M. Moore & E. Pisano V. / T.B.P.A. 1599 (RNG). Seno ltima
Esperanza, Puerto Toro, Ro Serrano, penin. W side estuary, 5125S 735W, N191, Sphagnum bog with
scattered trees of Pilgerodendron, N. antarctica & N. betuloides, 16.01.1977, D. M. Moore & E. Pisano V. /
T.B.P.A. 1733(RNG). Seno ltima Esperanza, Lote Sanchez, costaE RoSerrano,5126S 735W,N192,turbal
esfagnoso, en cojines de Sphagnum magellanicum, 26.01.1977, E. Pisano V. / T.B.P.A. 2039 (RNG). Pennsula
Roca, Seno Resi, ca. 175 m, 5151S 732W, W191, urbal interior, comn, 22.01.1978, E. Pisano V. / T.B.P.A.
2811 (BAB image!, RNG). Puerto Bella Vista, 5130S 7315W, 06.12.1979, F. Roig, O. Dollenz & E. Mndez /
T.B.P.A.5139Censo38(BABimage!).PuertoBellaVista,5130S7315W,09.12.1979,F.Roig,O.Dollenz&E.
Mndez / T.B.P.A. 5300 Censo 77 (BAB image!). W coast of Brunswick Peninsula, where Puerto Prat / Los
Canellos road meets, 05.04.1985, B. J. Wallace 089/85 (SGO). Fiordo Peel, Ro Murtillar, 5027S 7337W, en
turbales musgosos en afloramientos rocosos, 18.11.1985, E. Pisano V. 5925 (RNG). Fiordo Peel, Ro Murtillar,
5027S7337W,formatapicesextensosensitiosmuyhmedosenturbalesesfagnosos,19.11.1985,E.Pisano
V. 5948 (RNG). Fiordo Peel, Ro al E del cerro Aguilera, 5030S 7344W, en turbales esfagnosos sobre
afloramientosrocosos,sitiosinundados,15.12.1985,E.PisanoV.6070(RNG).ProvinciadeMagallanes:Dtroit
de Magellan, s. loc., s. d., Lman s.n. (P). Port Famine [Puerto (del) Hambre], Capt. Kings Voyage to South
America in H.M.S. Adventure (Kings Voyage) 18261830, [J.] Anderson s.n. (BM, K). Dtroit de Magellan, Port
Gallant (Ports Gallant, Famine, etc.), Voyage de lAstrolabe et de la Zle 18381840, H. Jacquinot & [J.]
Hombron s.n. (P). Dtroit de Magellan, Port Famine, Voyage de lAstrolabe et de la Zle, 18381840, [E.] Le
Guillous.n.(P).DtroitdeMagellan,BaieSt.NicholasetBougainville[St.NicholasBay],VoyagedelAstrolabe
Appendices
181
et de la Zle, 18381840, [E.] Le Guillou s.n. (P). Port Gallant [Puerto Gallant], ExtraTropical South America
Survey of H. M. S. Nassau 18661869, 19.04.1869, R. O. Cunningham s. n. (K, NY). Isla Dawson, Laguna Buen
Pastor,23.12.1910,A.Benoves.n.(SIimage!).IslaDawson,BahaLomas,23.12.1910,A.Benoves.n.(SIimage!).
PuntaArenas,turberas,Jan.March1917,G.Bonarelli38(SIimage!2sheets,ZT).[PuntaArenas],LagunaLynch,
turbera,Jan.March1917,[G.]Bonarelli38a(SIimage!).SfromPuntaArenasalongWsideofStraits,c.4kmN
of Fuerte Bulnes, wet boggy depression among Nothofagus antarcticus woodland, abundant on bare peat in
Sphagnumbog,assoc.Droserauniflora,23.03.1964,D.M.Moore1132(AASimage!,K,RNG,SGO).FiordoSilva
Palma, Angostura Titus,costa opuesta a ex aserradero,en turbales esfagnosos, 08.01.1973, E. Pisano V.3771
(RNG).LagunaParrillar,costaE,enturbalesesfagnosos,01.02.1973,E.PisanoV.3934(RNG).EstanciaSkyring,
Punkt046,5225S7152W,25.02.2001,A.Vogels.n.(MJG).ProvinciadeTierradelFuego:Pantanoscercadel
Rio Fontaine, Expeditio suecia 19071909, 01.03.1908, C. Skottsberg s.n. (SGO). Provincia de la Antrtica
Chilena: Tierra del Fuego, Orange Harbor, U.S. ExploringExpedition (Wilkes Expedition) 18381842,[January
April 1839], s. coll., s.n. (K, US image!). Hermite Island [Isla Hermite], Cape Horn, Antarctic Expedition 1839
1843,[20.09.07.11.1842],J.D.Hookers.n.(K,P2sheets).Terredefeu,leHoste[=IslaHoste],BaieOrange,
MissionduCapeHorn,28(?).07.1883,P.Hariots.n.(P).TerredeFeu,s.loc.,18901891,Rousson&Willemss.n.
(P). Terre de Feu, s. loc., 1892, Willems & Rousson s.n. (LY image!). Canal Beagle [Beagle Channel], Rimolino
Sumpf, February 1922, M. Gusinde s.n. (BR, K, M). Puerto Williams, Navarino Is., lowland Sphagnum bog,
08.01.1959,E.J.Godley924(SGO),E.J.Godley924a(BM),E.J.Godley924b(SGO).PuertoToro,IslaNavarino,
en turbales, 17.01.1972. E. Pisano V. 3460 (RNG). Isla Wollaston, Caleta Lientur, 5544S 6719W, en turbal
musgoso, escaso, 17.02.1980, E. Pisano V. 5006 (RNG). Arch. Cabo de Hornos, Isla Grevy, Rada Norte, Ba.
Gretton, en sitios pantanosos de turbales pulvinados, 16.01.1982, E. Pisano V. 5385 (RNG). Isla Hoste, Caleta
Awaiakirrh,ensitiosinundadosdeturbalesesfagnosos,23.01.1982,E.PisanoV.5446(RNG).lePicton[Picton
Island],CaletaBanner,5501S6656W,500mauSduhameau,michemindulac,alt.20m,30.12.1998,L.
Gautier3414(M).
FALKLAND ISLANDS (ISLAS MALVINAS):EAST FALKLAND: s.loc.,s.d.[1842],J.D.Hookers.n.(K).CanopusHill,Eastof
Stanley, 51 41S 57 47 W, assoc. Juncus scheuchzerioides and Cortaderia pilosa, male and female plants
growing about 25 m apart, 14.12.2006, R. W. Woods s.n. & M. Morrison (MJG). Cape Pembroke, close to
Stanley airport, acid grassland, associated with grass, Gunnera, Carex fuscula, small fern & christmas bush,
30.01.2009,MilleniumSeedBank(T.Heller&L.Taylor)MSBFI25(Kimage!).WESTFALKLAND:s.loc.,s.d.,[W.]
Wrights.n.(K).BetweenTealRiverandGoringHouse,Chartres,justNofRockyinlet,19(17?).12.1949,W.J.L.
Sladen Fa125/49 (AAS image!, BM).Roy Cove, Sharp Peak, N side of flat ground by sea, wet patch of ground,
assoc.Uniciniabrevicaulis,Cortaderia,Oreobolus,13.02.1964,D.M.Moore866(K,RNG).
GOUGHISLAND:CrestofridgebetweenGlen&1
st
Guleh,frequentonwetplacesandinhighlevelbogsfrom1200
ft. upwards, alt. 1500 ft., 01.12.1955, N. M. Wace 22 (BM). Base of Hags Tooth, alt. 1500 ft., frequent
especiallyinwetplaces,onshallowpeatoverlyingrocks,07.01.1956,N.M.Wace76(BM,BOL).XSt.Gonydale
RaisedBog,alt.1500ft,18.01.1956,N.M.Wace99(BM).Tafelkop,inwethighaltitudemire/grassland,4020
S954W,500masl,21.09.2006,R.M.Wanlesss.n.(MJG).
Note: Specimens with the abbreviation T.B.P.A. and a number were collected during the project Transecta
BotnicadePatagoniaAustral(Boelcke&al.1985).

Appendices
182

Appendices
183
APPENDIX5.
ListofstudiedherbariumspecimensofMaundiatriglochinoidesF.Muell.
AUSTRALIA. Queensland: South East Queensland [Moreton]: Moreton Bay, s. d., F. Mueller s. n. (K).
Slacks Creek, Logan River, s. d., N. Michael s. n. (BRI). Bald Hills Road, in shallow water, 272'S
1530'E,s.d.,S.T.Blake20053(K).BetweenPetrieandRedcliffe,infreshwatercreek,08x1959,S.
T. Blake 21028 (BRI). Woodford, in shallow shelving edge of One Mile Creek, 265'S 1524'E, 16 iii
1960, S. T. Blake 21205 (BRI, K). One Mile Creek at Woodford, N of George Street and opposite
Nicklaus Street, growing in narrow, recently flooded creek bed, now without free water, 2657'S
15247'E, 28 v 1997, H. I. Aston 2883 & T. Spokes (BRI). New South Wales: Wyong district, Porters
Creek Wetland, c. 1.5 km n of Watanobbi, 3315'25''S 15126'00''E, water channel in area of moist
dense woodland of Melaleuca ericifolia, 13 ii 2009, B.G. Briggs 10003 (NSW image!). Kogarah
Swamp, ca. 7 miles SW of Sydney, 17 i 1903; J. H. Camfield s.n. (BRI, K, NSW). Sans Souci, 3400'S
15107'E,18i1903,J.H.Camfield(NSW).Approx.1kmWofPacificHighway,betweenTuggerahand
Wyong, 3318'S 15125'E, small creek, 12 xii 1978, S. W. L. Jacobs 3461 (NSW, 2 sheets). Wyong,
swamp in centre of race course, 3317'S 15126'E, 12 xii 1978, S. W. L. Jacobs 3464, S. W. L. Jacobs
3465 (NSW). 7 km along Colletts Crossing Road S from WooliPillar Valley road, 2950'S 15312'E,
lagoon with Melaleuca quinquenervia in woodland of scribbly gum, bloodwood, stringybark,
Casuarina littoralis, 20 xii 1978, K. L. Wilson 4001 (NSW). Tucabia district, swamp near Upper
Coldstream, 2937'S 15307'E, 25 xi 1979, R. Pressey s. n. (NSW). Moffats Swamp, 2 km E of
RingwoodRoad,Medowie,3245'S15153'E,inandbesidedraininswampbelowsewagetreatment
works, with Eleocharis sphacelata, 22 xii 1979, K. L. Wilson 3088 (NSW). 1 km along Yellow Cutting
Road from Newfoundland Road, Newfoundland State Forest, 2955'S 15309'E, small stream
bordered by wetforestin Euc.forest,dominant insmallpool,20xii1981,K.L.Wilson3993(NSW).
Collombatti Creek, 10 km NNW of Kempsey, 3059'10''S 15249'50''E, 10 xii 1983, R. Pressey 30
(NSW). Tuggerah, on Gadlock Rd., N of Johnson Rd., 3319'S 15125'E, pond in swampy area beside
road,30xi1990,S.Papassotiriou13&S.W.L.Jacobs(NSW).PortersCreekWetland,Wyong,entry
to wetland from unnamed short road running NE off Fishburn Rd and just NW of Augusta Close,
Watanobbi. Entry point into swamp from Railway Rd (dirt track on W side of rail line). 3315'36.7''S
15126'11.4''E,elev.14m,3xii2008,L.C.StanbergLS80&G.Sainty(MJG).

Appendices
184
Appendices
185
APPENDIX6.
InteractiveidentificationkeysforTriglochininEuropepublishedontheinternetplatformOffene
Naturfhrer.FortechnicaldetailsseeHagedornetal.(2010).
IdentificationkeyforthegenusTriglochininEurope(inGerman),publishedonlineathttp://www.
offenenaturfuehrer.de/wiki/Die_Gattung_Triglochin_(Dreizack,_Juncaginaceae)_in_Europa_
(Sabine_von_Mering)
vonMeringS.2011:DieGattungTriglochin(Dreizack,Juncaginaceae)inEuropa.Publishedathttp:/
/www.offenenaturfuehrer.de/wiki/Die_Gattung_Triglochin_(Dreizack,_Juncaginaceae)_in_
Europa_(Sabine_von_Mering).DieseArbeitisteineOriginalarbeit,dieerstmalighierpubliziertist.
AnotherinteractiveidentificationkeyforTriglochininCentralEurope(alsoinGerman)ispublished
onlineathttp://offenenaturfuehrer.de/web/Triglochin_%28Mitteleuropa%29
Appendices
186
Appendices
187
Appendices
188
189
SUMMARY
I investigated the systematics, phylogeny and biogeographical history of Juncaginaceae, a small

familyoftheearlydivergingmonocotorderAlismataleswhichcomprisesabout30speciesofannual
andperennialherbs.Awiderangeofmethodsfromclassicaltaxonomytomolecularsystematicand
biogeographicapproacheswasused.
In Chapter 1, a phylogenetic analysis of the family and members of Alismatales was conducted to
clarify the circumscription of Juncaginaceae and intrafamilial relationships. For the first time, all
accepted genera and those associated with the family in the past were analysed together. Phyloge
neticanalysisofthreemolecularmarkers(rbcL,matK,andatpA)showedthatJuncaginaceaearenot
monophyletic. As a consequence the family is recircumscribed to exclude Maundia which is pro
posed to belong to a separate family Maundiaceae, reducing Juncaginaceae to include Tetroncium,
Cycnogeton and Triglochin. Tetroncium is weakly supported as sister to the rest of the family. The
reinstated Cycnogeton (formerly included in Triglochin) is highly supported as sister to Triglochin
s.str.LilaeaisnestedwithinTriglochins.str.andhighlysupportedassistertotheT.bulbosacomplex.
The results of the molecular analysis are discussed in combination with morphological characters, a
keytothegeneraofthefamilyisgiven,andseveralnewcombinationsaremade.
In Chapter 2, phylogenetic relationships in Triglochin were investigated. A specieslevel phylogeny

was constructed based on molecular data obtained from nuclear (ITS, internal transcribed spacer)
andchloroplastsequencedata(psbAtrnH,matK).Basedonthephylogenyofthegroup,divergence
timeswereestimatedandancestraldistributionareasreconstructed.ThemonophylyofTriglochinis
confirmed and relationships between the major lineages of the genus were resolved. A clade
comprisingtheMediterranean/AfricanT.bulbosacomplexandtheAmericanT.scilloides(=Lilaeas.)
is sister to the rest of the genus which contains two main clades. In the first, the widespread T.
striata is sister to a clade comprising annual Triglochin species from Australia. The second clade
comprisesT.palustrisassistertotheT.maritimacomplex,ofwhichthelatterisfurtherdividedinto
a Eurasian and an American subclade. Diversification in Triglochin began in the Miocene or
Oligocene, and most disjunctions in Triglochin were dated to the Miocene. Taxonomic diversity in
some clades is strongly linked to habitat shifts and can not be observed in old but ecologically
invariablelineagessuchasthenonmonophyleticT.maritima.
Chapter 3 is a collaborative revision of the Triglochin bulbosa complex, a monophyletic group from
the Mediterranean region and Africa. One new species, Triglochin buchenaui, and two new
subspecies, T. bulbosa subsp. calcicola and subsp. quarcicola, from South Africa were described.
Furthermore, two taxa were elevated to species rank and two reinstated. Altogether, seven species
and four subspecies are recognised. An identification key, detailed descriptions and accounts of the
ecologyanddistributionofthetaxaareprovided.AnIUCNconservationstatusisproposedforeach
taxon.
Chapter 4 deals with the monotypic Tetroncium from southern South America. Tetroncium
magellanicum is the only dioecious species in the family. The taxonomic history of the species is
described,typematerialistraced,andalectotypeforthenameisdesignated.Basedonanextensive
study of herbarium specimens and literature, a detailed description of the species and notes on its
ecology and conservation status are provided. A detailed map showing the known distribution area
ofT.magellanicumispresented.
In Chapter 5, the flower structure of the rare Australian endemic Maundia triglochinoides
(Maundiaceae,seeChapter1)wasstudiedinacollaborativeproject.AsthemorphologyofMaundia
is poorly known and some characters were described differently in the literature, inflorescences,
flowers and fruits were studied using serial mictrotome sections and scanning electron microscopy.
The phylogenetic placement, affinities to other taxa, and the evolution of certain characters are
discussed.AsMaundiaexhibitsamosaicofcharactersofotherfamiliesoftepaloidcoreAlismatales,
itssegregationasaseparatefamilyseemsplausible.
190
191
ZUSAMMENFASSUNG
Die Systematik, Phylogenie und Biogeographie der Juncaginaceae, einer kleinen Familie aus der
MonokotylenOrdnungAlismatalesmitetwa30einodermehrjhrigenkrautigenArtenwurdeunter
sucht. Verschiedene Anstze von der klassischen Revision bis hin zu molekularsystematischen und
biogeographischenMethodenwurdengenutzt.
Kapitel 1 beschreibt die Ergebnisse einer phylogenetischen Analyse der Familie und Vertreter der
Alismatales, um die Abgrenzung der Juncaginaceae und Gattungsbeziehungen zu klren. Erstmals
wurden alle akzeptierten Gattungen und in der Vergangenheit mit der Familie assoziierte Taxa
gemeinsam analysiert. Die phylogenetische Analyse von drei molekularen Markern (rbcL, matK, und
atpA)zeigt,dassdieJuncaginaceaenichtmonophyletischsind.DieFamiliewurdeneuumschrieben:
nach Ausgliederung von Maundia, die zur separaten Familie Maundiaceae gestellt wurde, umfassen
die Juncaginaceae Tetroncium, Cycnogeton und Triglochin. Tetroncium ist Schwester zum Rest der
Familie, allerdings nur mit schwacher Untersttzung. Cycnogeton wurde wieder in Gattungsrang
erhoben (vorher in Triglochin eingeschlossen) und ist mit groer Untersttzung Schwester zu
Triglochin s.str. Lilaea fllt in Triglochin s. str. und ist gut untersttzt als Schwester des T. bulbosa
Komplexes. Die Ergebnisse der molekularen Analysen werden zusammen mit morphologischen
Merkmalendiskutiert.AuerdemwerdeneinSchlsselzudenGattungenderFamilievorgelegtsowie
einige neue Kombinationen gemacht. In Kapitel 2 werden die phylogenetischen Beziehungen inner
halb von Triglochin untersucht. Eine Phylogenie auf Artniveau wurde mit nukleren (ITS, internal
transcribedspacer)undplastidrenSequenzdaten(psbAtrnH,matK)erstellt.BasierendaufderPhy
logenie wurden eine Altersbestimmung der Gruppe durchgefhrt und die Ursprungsgebiete rekon
struiert. Die Monophylie von Triglochin wurde besttigt und die Beziehungen zwischen den Haupt
linien der Gattung aufgeklrt. Eine Gruppe bestehend aus dem T. bulbosaKomplex (Mittelmeer,
Afrika)undderinAmerikaverbreitetenT.scilloides(=Lilaeas.)istSchwesterzumRestderGattung,
der aus zwei Hauptlinien besteht. In der ersten ist die weitverbreitete T. striata Schwester zu einer
Gruppe australischer einjhriger TriglochinArten. In der zweiten Hauptlinie besteht ein Schwester
gruppenverhltnis zwischen T. palustris und dem T. maritimaKomplex, wobei letzterer in eine
eurasischeundeineamerikanischeLiniezerfllt.DiversifikationinTriglochinbegannimMioznoder
Oligozn, die meisten Disjunktionen wurden in das Miozn datiert. Die taxonomische Diversitt in
einigen Gruppen ist stark mit Habitatwechseln verknpft. Viel geringere taxonomische Diversitt
wurdeinaltenaberkologischwenigervariablenLinienwiedernichtmonophyletischenT.maritima
beobachtet. Kapitel 3 ist eine gemeinschaftliche Revision des Triglochin bulbosaKomplexes, einer
monophyletischen Gruppe aus dem Mittelmeergebiet und Afrika. Aus Sdafrika wurden eine neue
Art, Triglochin buchenaui, und zwei neue Unterarten, T. bulbosa subspp. calcicola und quarcicola,
beschrieben.AuerdemwurdenzweiTaxaindenArtrangerhobenundzweiweiterewiederalsArten
besttigt.InsgesamtwerdensiebenArtenundvierUnterartenanerkannt.EinBestimmungsschlssel,
detaillierteBeschreibungenundAngabenzurkologieundVerbreitungderTaxawerdenprsentiert.
Ein Gefhrdungsstatus nach IUCNKriterien wird fr alle Taxa vorgeschlagen. In Kapitel 4 wird die
monotypische Gattung Tetroncium aus dem sdlichen Sdamerika untersucht. Tetroncium
magellanicum ist die einzige dizische Art der Familie. Die taxonomische Geschichte der Art wird
beschrieben, Typusmaterial ausfindig gemacht und ein Lectotypus fr den Namen designiert.
Basierend auf Bearbeitungen von Herbarmaterial sowie Literaturangaben werden die Art detailliert
beschrieben und Angaben zur kologie und dem Gefhrdungsstand gemacht. Eine Punktverbrei
tungskarte zeigt das bekannte Areal von T. magellanicum. In Kapitel 5 wird die Bltenmorphologie
von Maundia triglochinoides (Maundiaceae, siehe Kapitel 1) in einem Gemeinschaftsprojekt unter
sucht. Infloreszenzen, Blten und Frchte wurden mittels Serienschnitten und Rasterelektronen
mikroskopie analysiert, um widersprchliche Literaturangaben zu klren. Die Beziehungen zu
anderen Taxa und die Evolution ausgewhlter Merkmale werden diskutiert. Da Maundia Merkmale
verschiedener Familien der tepaloiden KernAlismatales vereint, erscheint eine Abgrenzung als
separateFamilieplausibel.

2

Transféré par

Informations du document

Description originale:

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

2

Transféré par

Droits d'auteur :

Formats disponibles

Systematics,phylogenyandbiogeography

As an earlydiverging monocot lineage, Alismatales are one key to the understanding of

The family Juncaginaceae was established at the beginning of the 19

In Chapter 1 a phylogenetic analysis of the family and members of Alismatales was

Chapter 3 is an taxonomical treatment carried out in collaboration with colleagues. The

Scheuchzeria Scheuchzeria Maundia Maundia Lilaea Tetroncium Maundia Cycnogeton

sect.Eutriglochin subg.Eutriglochin Triglochin Triglochin

The objective of this study is to provide a comprehensive phylogenetic analysis of

Taxon SamplingAltogether, nine species currently recognized as members of Juncagi

Molecular Data Phylogenetic AnalysisRelationships within Juncaginaceae and the

Dataset rbcL matK atpA rbcL+matK rbcL+matK+atpA

Character Lilaea Triglochins.s. Triglochinproceracomplex

Circumscription of Juncaginaceae and Intrafamilial RelationshipsThe circumscription of

With the removal of Maundia from Juncaginaceae, a recircumscription of the family is

A specieslevel phylogeny is presented for Triglochin, the largest genus of Juncaginaceae

Triglochin L. is an almost cosmopolitan genus with between 25 and 30 currently recognized

Biogeographic analysis To reconstruct ancestral distribution areas of Triglochin (and Juncagina

T. striata/Australian annual Triglochin Triglochin striata is recovered as sister to the lineage of

Taxon Distribution Habitat Lifeform Floweringand

Triglochinelongata Buchenau SouthAfrica upperandsupratidalzonesofestuarinesaltmarshes

Triglochinmuelleri Buchenau Australia coastalplaininsand,loamorclayofwinterwetflats

perennial Aug. Feb. 2n=64

Node Description rbcL/matK rbcL/matK ITS/psbAtrnH

Triglochin bulbosa L. was described by Linnaeus in 1771 based on collections by Thunberg

Distribution. TheTriglochinbulbosacomplexisnativetothe Mediterraneanregion,extendingalong

Taxonomic remarks. Triglochin elongata is similar to T. buchenaui in having rhizomes instead of

The distribution area of T. magellanicum comprises mainly southern South America

Gough Island is recognized as a nature reserve and an ordinance entitled 'Conservation of

Fig. 7. Maundia triglochinoides. Threedimensional reconstructions of floral vasculature. Fusion bundles =

Detailed observations on the flower groundplan in Maundia were published by Markgraf

Phylogenetic relationships within Juncaginaceae as proposed by this study (Chapter 1) have

Important achievements presented in this thesis comprise the recircumscription of the

Marker Primersequence53 Source PCRingredients PCRconditions

I investigated the systematics, phylogeny and biogeographical history of Juncaginaceae, a small

In Chapter 2, phylogenetic relationships in Triglochin were investigated. A specieslevel phylogeny

Vous aimerez peut-être aussi