AmericanJournal ofBotany80(5): 500-516. 1993. ANATOMICALLY PRESERVED GLOSSOPTERIS STEMS WITH ATTACHED LEAVES FROM THE CENTRAL TRANSANTARCTIC MOUNTAINS, ANTARCTICA' KATHLEEN B. PIGG2,4 AND THOMAS N. TAYLOR3 2Department ofBotany,ArizonaStateUniversity, Tempe,Arizona85287-1601;and 3Department ofPlantBiologyand ByrdPolarResearchCenter,The Ohio StateUniversity, 1735Neil Avenue,Columbus,Ohio 43210 StemsandbudsofGlossopteris skaarensis PiggandbudsofG. schopfii PiggfromthePermianSkaarRidgelocalityin the centralTransantarctic Mountains, Antarctica demonstrate thefirst anatomically preserved knownwithstem/ glossopterids leafattachment. StemsofG. skaarensis are 1-12 mmin diameter (X = 3.1 mm)witha broadpith,poorlydefined primary xylem, anda zoneofsecondary xylemup to6 mmthick.Pycnoxylic woodconforming toAraucarioxylon Krausis composed oftracheids withuni-to biseriateoval to hexagonalborderedpitson radialwalls,uniseriate raysone to a fewcellshigh, andcupressoid totaxodioidcross-field Stemshavea narrow pitting. zoneofsecondary phloem,aerenchymatous cortexwith scattered sclereids, and sometimes a narrowperiderm. Two wedge-shaped leaftraceseachbifurcate to formfourstrandsin thebase of each petiole.Smallaxillarybranchesare vascularized by doublebranchtracesthatfuseat themarginof the mainaxis.BudsofG. skaarensis haveleaveswithnarrow laterallaminaeanda thickened midribcontaining a widelacuna, delicatevascularstrands, and a prominent In contrast, hypodermis. budsof G. schopfii have uniformly thickleaveswith prominent, circular vascularbundlesheaths. Theseanatomical detailsareusedtoreconstruct individual typesofglossopterid plants,providing newinformation towardunderstanding theecologyandevolution ofthisimportant groupofPermianseed plants. Throughoutthe Permian, Gondwana was dominated by gymnospermousplantsthatbore theentire,reticulateveined leaves of the genus GlossopterisBrongniart.Such leaves representthe dominantelementin Permian compression-impressionflorasof Gondwana (e.g., Chandra and Surange, 1979; Anderson and Anderson, 1985; McLoughlin, 1990b). Anatomically preservedglossopterid leaves are also knownin rarelyoccurringpermineralized chertsin Antarctica(Schopf,1970; Pigg,1988, 1990; Pigg and Taylor, 1990; Taylorand Taylor, 1990) and Australia (Gould and Delevoryas, 1977; McLoughlin, 1990a; Pigg and McLoughlin, 1992). The low diversityGondwana florasalso containwood ofAraucarioxylonKraus, rooting structuresof the genus VertebrariaRoyle, pollen organs assignable to ArberiellaPant and Nautiyal, striatebisaccate pollen of the Protohaploxypinustype,and a variety organs(e.g.,Schopf, ofseeds and ovule-and pollen-bearing 1973, 1976; Lacey, van Dijk, and Gordon-Gray, 1975; Gould and Delevoryas,1977; Retallackand Dilcher, 1988; Taylor and Taylor, 1990). The repeatedco-occurrenceof these assemblages at numerous localities has suggested that these disarticulatedorgans representa single group of plants,the glossopterids. Most reconstructions ofthe Glossopteris planthave been based on disarticulatedorgans (e.g., Plumstead, 1958; Gould and Delevoryas, 1977; Pant, 1977; Retallack and Dilcher, 1988). Unfortunately, however,reportsof connectionsbetweenorgansare relativelyrare.Glossopterid leaves preservedas compression/impression fossilshave been occasionally reportedin attachmentto compressed vegetativeaxes (Table 1). CompressedGlossopterisleaves have also been foundattachedto a varietyofreproductive structures (e.g., Thomas, 1958; Holmes, 1974, 1990; Lacey,van Dijk, and Gordon-Gray,1975; Surangeand Chandra, 1975; Schopf,1976; Rigby,1978a; White,1978; Anderson and Anderson, 1983, 1985; Pant and Nautiyal, 1984b; McLoughlin, 1990b). Occurrencesthat demonstratethe connectionof a given reproductivestructureto particularleaf and wood typesare valuable in providing a criticalbasis forsortingout thetaxonomyand evolution ofthegroup.In a fewexamples ofcompressedstemswith attached leaves, the stems are anatomically preserved (Etheridge, 1895, 1899; Pant and Singh, 1974; White, 22 July1992;revision accepted14January 1978; Pant and Nautiyal, 1984a), but these specimens 'Receivedforpublication 1993. have not yet provided detailed informationabout leaf/ Unit6des Scienceet Meyer-Berthaud, The authorsthankBrigitte wood relationships.Isolated petrifiedstemsand wood are OhioUniduLanguedoc, Montpellier, MaryLouiseTrivett, Techniques knownfromnumerousGondwana localities(e.g.,Seward, EdithL. Taylor,TheOhioStateUniversity, Columbus, Athens, versity, 1914; Kriiusel,Maithy,and Maheshwari, 1961; Kriusel, disofTexas,Austin,forhelpful and TheodoreDelevoryas, University Lee Wilcox, 1962; Maheshwari, 1972; Prasad, 1982, 1986; Mussa, cussionand readingofan earlierdraftofthemanuscript; 1986; Mussa and Coimbra, 1987; Taylor and Taylor, and RobertW. Roberson,ArizonaState The Ohio StateUniversity reconstruc- 1990; Giraud, 1991). They very fromsmall stems (e.g., Tempe,forassistancewiththree-dimensional University, tions;RobertJones,The AustralianMuseum,Sydney,foraccess to Mussa, 1986) to large forestscontainingtreeswith subGlossopteris linearis,and talbragarense, specimensof Blechnoxylon stantialtrunks(e.g., Francis et al., 1991; Taylor, Taylor, and walkomii;and KristieL. Butlerforphotographic Gangamopteris inpartbyFacultyGrant- and Cuneo, 1992). Althoughthese reportsindicate that wasfunded Thisresearch assistance. technical andNSF grantsBSR-9006625 at least some glossopteridswere large woody trees,they in-AidB-600,ArizonaStateUniversity provide littleinformationabout the systematicsor evoto KBP and DPP-8815976to TNT. 4 Author forcorrespondence (FAX: 602-965-6899). lution withinthe group. 500 May 1993] PIGG AND TAYLOR-GLOSSOPTERIS MATERIALS AND METHODS Permineralized peat in the central Transantarctic Mountains, Antarcticacontains anatomicallypreserved woodystemsand smallerbuds withtheleaves Glossopteris skaarensis Pigg and buds of G. schopfiiPigg (Pigg, 1988, 1990). Specimens were recovered froma site on Skaar Ridge (Global Positioning System 84049'1 18"S 163?20'370"E) (Schopf, 1970; Taylor, Taylor, and Collinson, 1989; Taylor and Taylor, 1990) and are regarded as Late Permian(Farabee, Taylor,and Taylor, 1991). The materialwas studiedwithserialpeels usinga modification of the cellulose acetate peel technique (Joy,Willis, and Lacey, 1956), etchingthe silicate matrixwith 40% hydrofluoricacid forseveral minutes(Basinger and Rothwell, 1977) and neutralizingwitha saturatedsolution of sodium bicarbonate.Specimens were cut frompeels and mountedon standardmicroscopeslides in xylene-soluble Coverbond. Axes were mapped and measured with the IBM PC-Based Three-Dimensional ReconstructionSystem (HVEM-3D), Version 1.2 (Royer, 1988). All preparations are deposited in the Paleobotanical Collections, The Ohio StateUniversity,and bear acquisitionnumbers 14092, 14094, 14102, 14103, 14128-14214, 1425714372, 14396-14403, 18070, 18072, 18106-18206, 18208-18223, 18233-18241, 18256. SYSTEMATICS Genus: Glossopteris Brongniart Species: G. skaarensis Pigg sp. amplif.Pigg & Taylor Pigg,K. B. 1990. Rev. Palaeobot. Palynol. 66: 105-127. Amplifieddiagnosis: Vegetativeleaffeaturesas delimited in Pigg,1990. Stems 1-12 mm in diam (X= 3.1 mm) withbroad pithup to 1.1 mm across. Primaryxylemwith helical to scalariformwall thickenings.Secondaryxylem up to 6 mm thick;pycnoxylic;assignable to Araucarioxylon Kraus; tracheidswithuni- to biseriate,oval to hexagonal borderedpits on radial walls; uniseriaterays one to a few cells high; cross-fieldpittingcupressoid-taxodioid. Secondaryphloemca. fiveto eightcellsthick;cortex aerenchymatous-parenchymatous, occasionally with clustersof sclereids;peridermup to eightcells thick;leaf in thepetiole;axillarybranchtraces tracestwo,bifurcating two, fusingat base of branch. Holotype: Specimen on 465 C3 Side and C4 Side, Pigg, 4, Plates 1990: Figs. 5, 6 and Plate VI, 1; Plates VII, 1, 3,_ VIII, 1-4. Slides 14098-14101; 14230-14248 in-thePaleobotanical Collections,The Ohio State University. Paratypes: Specimen G2. 1, 463 E Bot Fragment,Fig. 4a and Plate V, 1 of Pigg, 1990; Specimen G2.4, 452 Fragment,Fig. 4c and Plate V, 4 of Pigg, 1990; Specimen G2.5, 452 Fragment,Plate V, 4, of Pigg, 1990; Specimen a in 484 C Bot no. lg, Plate VI, 5 of Pigg, 1990. Slide 14097 in the Paleobotanical Collections,The Ohio State University. Collectinglocality: Skaar Ridge, Beardmore Glacier, central Transantarctic Mountains, 84?47'S, 16301 5'E, BuckleyIsland Quadrangle(Barrettand Elliot,1973) (GPS 84049'118"S 163020'370"E). Stratigraphicoccurrence: Upper Buckley Formation, Late Permian. STEMS 501 DESCRIPTION Stems and buds of Glossopterisskaarensis-The descriptionis based on thestudyof 152 specimensofwoody stems,of which 12 were studiedin detail (Figs. 1-14, 2224), and fiveleafybuds (Figs. 15-17, 25-30). Some ofthe largerstemshave partiallypreservedextraxylary tissues, leaf bases, and axillarybranches(Figs. 1-5). These specimens are assigned to G. skaarensis on the basis of leaf trace and corticalanatomythatis identicalto thatof the vascular tissue and mesophyllof mature,isolated leaves of G. skaarensis (Pigg, 1990). The smalleraxes are buds that produced numerousleaves with G. skaarensis anatomy.Woody stemsof G. skaarensisare 1-12 mm in diam (X = 3.1 mm, 152 stemsmeasured) and typicallypossess a broad pith up to 1.1 mm across (Figs. 1-5). Tissue comprisingthe pith is variable: the pith may be hollow, septate.When tissolidlyparenchymatous,or irregularly sue of the pith is preserved the parenchymatouscells generallycontain dark contents.The marginof the pith is usually scalloped or disruptedwith cavities occurring presumablyin thepositionof theprimary intermittently, vascular bundles,and sometimesin association withleaf tracedivergence(Figs. 3-5). Primaryvascular strandsare difficult to distinguishin transversesection. However, a markedchange in tracheidsize and the patternof pitting can be recognizedin longitudinalsection frominner to outerstelarmargin.This reflectsthepositionofa primary xylemzone (Fig. 12). Primaryxylemtracheidsare up to 8.8 Am in diam and have helical or scalariformpitting patterns. Outside the primaryxylemthe secondaryxylem may be up to 6 mm thickon the largestspecimens.The wood may appear as a continuous cylinder(Figs. 1, 2) or, at other levels, be dissected radially by the divergenceof several leaf traces (Figs. 3, 5, 22-24). When leaf traces are present,the cylinderis generallydissectedby at least fivetraces ratherthan by only one or two. Preservation of the wood is variable, with numerous small areas of well-preservedcells scatteredamonglargerareas ofpoorly preservedtissue. Because of this,several featuresof the wood are obscured,includingdetailsoftheprimaryxylem bundles. It is also difficult to determinewhethergrowth rings occur in this wood. There is some suggestionof periodicchangesin tracheiddiameterin some ofthebetter specimens (e.g., Fig. 1, lower right). Secondary xylem is similar to AraucarioxylonKraus (Lepekhina and Yatsenko-Khmelevsky,1966; Maheshwari, 1972; Pant and Singh, 1987). Tracheids are 10-15 Amin diam and possess uni-to biseriate,oval to hexagonal borderedpits (5 x 5,m) only on theirradial walls (Figs. 12-14). On some tracheids,widelyseparatedoval-shaped pits are presentin combinationsof uniseriateto biseriate arrangement(Figs. 12, 13). On otherstheyare crowded, biseriate,alternate,and more hexagonal in outline (Fig. 13). The wood contains numerousuniseriate,parenchymatous rays,each one to a fewcells high(Fig. 14). Crossfieldpittingis of the cupressoid to taxodioid type(Figs. 10-11). To the outside of the secondaryxylem the secondary phloem occurs as a zone fiveto eightcells thickof poorly preserved,radiallyaligned cells. The vascular cambium is not preserved.The primarycortexis up to 5 mm thick 502 [Vol. 80 AMERICAN JOURNAL OF BOTANY TABLE 1. Occurrencesof glossopteridleaves attachedto stemsa Taxon Preservation Age Locality Gl brownianabc* Impression Filicitesd Impression S. poljphyllae Impression S. (?) longifuliae Impression Permian Australia Permian India Permian India Permian India Be. voodtmazson)il Permian Impression Glossopteris19" Impression,petrifiedaxis VertebrariaG.browniana5 Impression Vertebraria/GI. communisi Impression BL. talbragarens ek** Impression,petrifiedaxis Glossopteris. Impression V"australSmn' Impression G. browniana-"n Impression Terminal budg0 Impression S. Iongicaulisp Impression cf. GI. angusttfblia. Impression India Permian Australia Permian S. Africa Permian India Permian Australia Permian Australia Permian Australia Permian S. Africa Permian Australia Triassic S. Africa Triassic S. Africa -- 3 x 1.1 Sessile - 1.2 x 0.2 Permian 9.3 x 1.8 Permian - GZ. recMurvaY Impression G. c 'lopteroides Impression 6 1.0-4.3 x 0.1-0.3 Smooth, some leaf scars Transversenodes 6.0 x 0.8 Transversenodes 4.5-6 x 2.5-4.5 - 8.5 x 1.8 G. indicas Permian Transversearticulations 15.2 x 1.8 Oval leaf scars 3.7-5.2 x 0.7 Transverseridges 1.5 x 1 Transverseridges Petiolate 1.8-2.5 x ? 1.5 x 1-2 Sessile 7 x 1.5 Petiolate 10-12 x 2 Petiolate Sessile 9 x 1.5 - S. Africa Permian Australia Triassic S. Africa Permian India Permian India Permian India Permian India Permian India 20 Subopposite 7 0.7-2.1 x 0.5-1 Petiolate 11 x 2.5 Helical? I Cluster 16 Cluster 7 Cluster Up to 7 x 0.4 Longitudinalstriations ? x 0.5 Longitudinalstriations 6.4 x ? Longitudinalstriations Longitudinalstriations Longitudinalstriations 14 x ? - 4 Whorl or tighthelix 2-8 per node Whorl 6 Apparentwhorls Apparentwhorls 4 per node Apparentwhorls 2 3 - 9.5 x 2.2 Longitudinalstriations 35.2 x ? India 8.5 x 1.7 Sessile 12.8x ? - -2 - Permian- India 3-8 "Verticel" 4 10 x 0.70.9 - 7 x 3.5 Petiolate 9.8 x 3.8 Petiolate 14 x 4 Petiolate 15.4 x 6.8 Petiolate 8.7 x 3.2 Sessile Helical in tufts 5.8 x 0.3-0.4 Longitudinalstriations 40 x 6 - - Helical, close whorl 8 Singly 1 per node Close tuft 7 Whorl or tighthelix 4 Overlapping,tighthelix 10-15 Verticelor tighthelix 4 per node "Whorls' 5-7 per node India Permian India Permian 7 - 7 x 1.1 Sessile 2.5 x 0.8 Petiolate 17.3 x 3.1 Sessile Brazil Permian S. Africa S. Africa 7.5 x 0.6 Longitudinalstriations 0.7 x 0.25 - - Permian Impression G. obovata var. attenuataS Impression G. walkomiPt Impression Go. verticillatal Impression Gl searsolensisv Compression/Cuticle Gl. maculatav Compression/Cuticle Gl. sastriiv Compression/Cuticle Gl. oldhamiiv Compression/Cuticle GI. panduratav Impression Gl. taenioides Impression GI. agustifoliav Impression Attachmentof specimen Leaftoial or leaves/node - Vertebrariaf Impression,cast of axis Gl. angustifolias Impression Stem: Length x width(cm) Surfae Leaf: lngth x width(cm) Petiolate or sessile? I May 1993] TABLE 1. PIGG AND TAYLOR-GLOSSOPTERIS 503 STEMS Continued Taxon Preservation G. cyclopteroides var. cordiferav Impression ArberialeeukuilensisImpression LidgettonialidgettonioidesImpression Gl. skaarensis' Petrifiedstem and leaves Age Locality Permian India Permian S. Africa Permian S. Africa Permian Antarctic Leaf: Length x width(cm) Petiolate or sessile? -- 10 x 4 14 x 1.8 (3.5) x 2.7 Stem: Length x width(cm) Surface - 4 _ 8 - - - 8 - a Informationobtained fromsources indicated. Abbreviationsforgenericnames: Be GI = Glossopteris;Go = Gontriglossa;S = Sagenopteris;V = Vertebraria. bDana, 1849*. c Rigby,Maheshwari and Schopf, 1980. d Bunbury,1861. e Feistmantel,1881. I Etheridge,1895. gWhite, 1978. h Zeiller, 1896. Oldham, 1897. Etheridge,1899. k Pant and Nautiyal, 1984a. Etheridge,1904. - Seward, 1910. n Walton and Wilson, 1932. OWalkom, 1928. P DuToit, 1927. q Thomas, 1952. r Dolianiti, 1954. s Plumstead, 1958. t Rigby, 1967. u Anderson and Anderson, 1983. v Pant and Singh, 1974. Anderson and Anderson, 1985. x This paper. * Not accepted; ** Anatomicallypreserved. to aeren(Figs. 1, 2, 4) and rangesfromparenchymatous parenchywithan innerzone oflarger-celled chymatous, more ofsmaller, bya hypodermis matouscellssurrounded compactcells(Figs. 1, 2). Occasionallysmallclustersof A periderm sclereidsare embeddedin theparenchyma. up to eightcellsthickoccursin someaxes. Preservation poor. tissuesis generally ofextraxylary leaftracesdivergefromthe stele Two wedge-shaped to formfour and enterthepetiolewheretheybifurcate strands(Figs. 2, 3, 5, 22-24). It is unclearwhetherthe of the same symtracesoriginatefromthe bifurcation fromtwoadjacentcaulinebundles. podiumor separately Distalto itslevelofattachment to thestempetiolebases in are 1.5 x 0.58 mm in diam and oval to triangular outline(Figs. 2, 24). At thislevel the vascularbundles ringsoftangentially aresurrounded byseveralconcentric cellsthatcomprisea bundlesheath. elongateparenchyma parenchyma Thegroundtissueis composedoflarge-celled at hypodermis in the centralareas and a smaller-celled theperiphery (Fig. 2). xylemthebranchtraces Atthemarginofthesecondary appearas paired,elongateareasaccompaniedby a large amountofsecondary xylem(Fig.9). Theybecomeovoid intransverse sectionandca. 7 10,tmindiamatthemargin fuseinto oftheparentaxis (Figs.7, 8). Theyeventually a singleradialstrandca. 1.8 x 0.6 mmindiamthatenters Attachmentof specimen Leaf total or leaves/node = Belemnopteris;Bi = Blechnoxylon;G = Gangamopteris; theaxillary branch(Figs.4, 6). Axillary branches areabout one-fifth thesize ofthemajorstemaxis (Fig. 5). In additionto maturewoodyaxes thatbear axillary branches,small,leafybudsofG. skaarensisalso occurin thechert.The mostextensive budis sectionedat thelevel of thestemaxis and bearsat least24 helicallyarranged leaveswiththeanatomycharacteristic ofmatureG.skaarensisleaves (Pigg,1990; Figs. 15-17, 25-30). This bud containsan axis 1.2mmindiameter atthemostproximal level witha steleca. 54 ,umin diameterconsisting of a broad,hollowpithsurrounded bya ringofpoorlydefined vascularbundles.The steleis surrounded by a partially preserved, parenchymatous cortexthatextendsintolobelike leafbases (Fig. 17). The relativepositionsof individual diverging leaves suggeststhatthephyllotaxy approachestwo-fifths (Figs.25-30). The steleis too poorly preservedto providedetailsabouttheprimary vascular architecture, butatmostlevelsfivetoninepoorlydefined, individualbundlesappearto be present. Leaveson thebudsareup to 540 ,umlong.Theyappear first as lobed leafbases extending fromthestemmargin and becomeoval-triangular in transverse sectiondistal to theirpointofattachment (Figs.15, 17,27-30). Leaves are characterized by a thickenedmidribup to 300 ,um acrossthatcontainsa prominent lacuna,andconsiderably thinner laterallaminae(Figs. 15-17,25, 26). Mesophyll 504 AMERICAN JOURNAL OF BOTANY [Vol. 80 is poorlypreservedbut appearssimpleand parenchy- thatthenodalanatomyofG. skaarensiswouldbe ofthe matous,lackingpalisadeandspongy layers.Evenonthese unilacunar,one- or two-trace type.The petiolebecame Although smallleavesthecuticleis thick(Fig. 17). oval-shaped,distalto thelevelofattachment. it is not knownwhetherG. skaarensisleaves had long Buds of Glossopterisschopfii-Five small buds withG. sessile.Isolated werenotentirely petioles,theyapparently laminarwings, slight leavesarepreserved distalto thelevelofthestele proximalsectionsofleaves,showing schopfii (Figs. 18-21,31, 32). The innermost, smallestleavesare are comparablein size and anatomyto petiolebases at250 x 31 ymandhavea prominent, keeledcentralregion tachedto thestems(compareP1.IV, 4 ofPigg,1990with butonlyslightly developedlaterallaminae(Fig.19).Leaves Fig. 2, thismanuscript). Branchtracesin G. skaarensis to the outsideof thesehave a uniformthickness, are originateas double vascularstrandsthatfusepriorto is likethatin some comparableto matureleavesin width,and have prom- entering theaxillarybud.Thispattern inentlacunaeinthepositionofthevascularbundles(Figs. cordaites(Mesoxylon Traverse,1950).Stems thompsonii, 18, 31). Leaves tendto overlapone another,enclosing producedaxillarybranchesfairlyfrequently. up plantwereoblanceolate, theaxis.Tissuesoftheleaflaminaearenotwellpreserved. LeavesoftheG. skaarensis to 2.8 cm wide,and probablynot over 10 cm in length midriband a coarsepolygonal-meshed witha prominent DISCUSSION venationpattern(Pigg,1990). Theyhad a thickhypovasradiallysymmetrical Glossopterisskaarensis stems with attached leavesdermisand delicatemesophyll, midribareas,and thinlamina. Glossopteris skaarensisstemswithattachedleaves are cularbundles,thickened relatively small,branching woodyaxes.Whenthesestems A thickcuticlecoveredepidermalcellswithsinuousanare mappedin thematrix, sunkenstomataontheabaxial mostofthembelongto one of ticlinalwallsandscattered, beakcellswithprominent fourlargeclusterscontaining fiveto 12 axes each.Based surface. Five to sixsubsidiary on three-dimensional thestoma(Pigg,1990). reconstructions, some oftheindi- likepapillaesurrounded beencomparedmost vidual sectionsprovedto be froma singleaxis thatwas has previously Leafmorphology leaf impressionswith twistedin thematrixso thatit was sectionedmorethan closelywith otherfragmentary once. The flexibility oftheaxes thatenablesthemto be coarse-meshed venation,includingG. conspicua,G. retwistedin the matrix,alongwiththe relatively and severalspeciesof Belemnopteris simple tifera,G. shirleyi, extraxylary tissues,and smallamountofperiderm, 1977;Banerjee,1978; sup- (Schopf,1976;PantandChoudhury, thattheseaxeswereslenderstems Rigby,1978b,1983;Lacey,van Dijk,and Gordon-Gray, portstheinterpretation or branches.The lack of both adventitiousrootsand 1975; Pigg,1990). Anatomicaldetailshave previously asymmetrical withleavesfromtheBowenBasin, xylarytissuessuggeststhattheymaynot beencomparedbriefly havebeenrhizomes, vines,orlianas.Whether theseleaf- Queensland(Gould and Delevoryas,1977; Pigg,1990), bearingaxes weremainstemsor,morelikely,thelateral but recentstudysuggestsleaves fromthetwo localities 1992). branchesofa largertree,is notknown. notequivalent(PiggandMcLoughlin, areprobably Based on theanatomydescribedabove,it is believed skaarensisleaves can also now be comGlossopteris Pigg.1-3,5. Serialsectionsfroma singleaxis,withFig. 1 beingthemostdistal.Bars= 1 mm.1. Transverse skaarensis Figs.1-5. Glossopteris cortex(C). 451 D botr No. ofsecondary xylem(X), andaerenchymatous pith(P), solidcylinder ofwoodystemshowing internode sectionthrough 54, x 27. 2. Moreproximalsectionof same stemwithpetiole(at lowerright)attachedat a stillmoreproximallevelin Fig. 3. At thislevelthe same sectionthrough see Fig.25. 451 D botr No. 31, x 27. See barin Fig. 1. 3. Transverse vascularsupplyofthepetioleconsistsoffourstrands, appearance Thepetiolehasthecharacteristic (arrows). bytwostrands petiole(PT) is vascularized level.Heretheattached atmoreproximal specimen see Fig. 24. 451 D bot r No. 4, x 27. See barin Fig. 1. 4. Obliquesectionofstemwithaxillarybranch(AX) ofisolatedleavesof G. skaarensis, stemat levelofnodewithseveralleaf 451 D botr No. 54, x 12.5. Mostproximalsectionofstemin Figs.1-3 showing and petiole(PT) (at right). tracesoccurring at margin(arrows)and severalattachedpetioles(PT), see Fig.23. 451 D boti No. 3, x 20. Tracesariseas doubletracesin thestem(Figs. skaarensisPigg.6-9. Seriesshowingaxillarybranchtracederivation. Figs.6-14. Glossopteris proximalto axillarybranch.451 D botl No. 7-9) and fusein thecortex(Fig.6). Bar = 0.5 mm.6. Fusedbranchtraceat levelofthestemcortex, tracesofthesamebranchtracepriorto fusion.451 D bott No. 6, 4, x 33. 7. Moreproximallevelofmainstem,showingthetwocontributing x 33. See bar in Fig. 6. 8. Pairedbranchtracesat themarginofsecondary xylem(arrows).451 D bott No. 18, x 33. See bar in Fig. 6. 9. More xylem.451 D botl No. 27, x 33. See barin Fig.6. Figs. pairedbranchtraces(arrows)inthesecondary proximalsectionofsamespecimenshowing 10. 451 D botl No. 54, cross-field pitting. taxodioidto cupressoid 10-14.Wood anatomy.Bars= 0.05 mm.Figs.10, 11. Radial sectionshowing xylem fromprimary (left)to secondary x 450. 11. 451 D bott No. 97, x 450. See barin Fig. 10. 12. Radial sectionofwood showingtransition pitson radialwallsoftracheids ofuni-and biseriate distribution 451 D bottNo. 97, x 450. See barin Fig. 10. 13. Radialsectionshowing (right). wallsand uniseriate rays.Notepits lackingpitson tangential sectionwithtracheids and ray(at bottom).451 E topNo. 74, x 404. 14. Tangential section(arrow).451 E topNo. 14, x 404. See barin Fig. 13. in transverse sectionofsmallaxis 15. Mid-longitudinal Bars= 0.5 mm. 15-17.G. skaarensis. and G. schopfii. skaarensis Figs. 15-21. Budsof Glossopteris sectionof smallaxis showingdetailof leaves.Note (A) showingattachedleaves(arrows).451 D top a No. 6, x 31. 16. Obliquelongitudinal sectionofsmallaxis withattachedleafbases, midribarea (MR) and narrowlaterallaminae.533 B topH No. 2, x 42. 17. Transverse prominent sectionofbud 18. Transverse in Figs.25-30. 451 D bot0 No. 32, x 50. Figs.18-21.G. schopfii. and poorlydefinedstele.Thisspecimenis figured positionsof vascularbundles,and lacunaein leaves(arrows)thatrepresent leaves.Note prominent above thelevelof thestemwithencircling youngleavesin sectionofbud showing in Fig. 32. 484 B I bota No. 4, x 32. 19. Transverse laterallaminae.Thisspecimenis figured overlapping in Fig. 33. 484 B bot6 No. leaves.Thisspecimenis figured overlapping sectionofbud showing section.484 B botc No. 1, x 36. 20. Transverse leaves.484 A bota No. 15, x 18. sectionofbudwithoverlapping 1, x 42. 21. Obliquetransverse May 1993] PIGG AND TAYLOR-GLOSSOPTERIS ew V 505 STEMS ~~ ~ ~ ~ ~ ~~~- ~~ ~ ~ ~ ~ j ~ ~ A, 4-C~~~ .0-~ *Ma ~ ~ ~ ' WP kAN19 ~ ~ ? 506 AMERICANJOURNALOF BOTANY - ~~~~ ~~~~~~~~~~~~~~ -W df~ Id N- f As~~~~~~~~~~ [Vol. 80 May 1993] PIGG AND TAYLOR-GLOSSOPTERIS STEMS 507 A'4*~~~~~~~~~~~~~~~~~~~~~~4 IN ~ , __ '4~~~~~ 2i)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 508 AMERICAN JOURNAL OF BOTANY [Vol. 80 detailstoothercoarseparedinmorphology andcuticular meshedleaves attachedto axes. Theyare similarto G. including sastriiand G. maculatain a numberoffe-atures leafand meshshape,presenceofa petiole,sizeofabaxial positionand irregular epidermalcells,and hypostomatic of stomata(Pantand Singh,1974; Tables 2, orientation 3). In particular, G. skaarensisand G. maculatasharea similarmeshsize,whilemeshesof G. sastriiare larger. Glossopteris skaarensisand G. sastriibothhavea ringof orbeakedpapillaesurroundsubsidiary cellswitharching, ing each stoma.Of the threetaxa, G. skaarensisis the smallestleaf(estimated to be ca. 10 cm longand 2.7 cm wide),withbothG. maculata(9.8 x 3.8 cm)and G.sastrii (14 x 4 cm) somewhatlarger.Midribwidthand size of withlaminasize,while meshesare positively correlated veindensitydecreases. ofthesemediumIt is interesting thatseveralfeatures inleavesare positively correlated, sized,broad-meshed cludinglanceolateshape,thickcuticles,and sunkenstostudmata,withsubsidiary, beaklikepapillae.In general, leavesindicate iesofcuticular structure amongglossopterid a widerangeoftypes(Pigg,1988).To datetherehasbeen littlecorrelation betweencuticleand variousmorphologhave ical features.Instead,subtlecuticulardifferences been used to establishadditionalspeciesof Glossopteris becomebetter (e.g.,Pantand Singh,1974).As affinities betweencuticlesofanatomunderstood, thecorrelation whether icallypreservedleaves may help to determine therearetrendsin cuticleanatomywithinthesubgroups ofglossopterids as those ofcomparabletaxonomicutility of,say,theconifers. Wood of G. skaarensisconformsto the formgenus is ofAraucarioxylon Kraus.The history Araucarioxylon tothewoodgenusDadoxylon complexandtiedintimately whichalso servesas a formtaxonforPaleozoic Endlicher, and Mesozoic pycnoxylic wood witharaucariancrossfieldpitting(Pant and Singh,1987; Giraud,1991). Althoughsome authors(e.g., Kriusel, Maithy,and Mathat 1964) have suggested heshwari,1961; Vogellehner, ofDashouldbe regarded as a synonym Araucarioxylon Dadoxylonas havingwelldoxylon,otherscharacterize bundlesand a nonseptate definedendarchprimary pith and Araucarioxylon and severalothergeneraforwood of theprimaryxylem characters thatlacksthedefining andMaheshwari, bundlesandpith(Kriiusel, 1961; Maithy, 1966; MaheshLepekhinaand Yatsenko-Khmelevsky, wari, 1972; Pant and Singh,1987). Additionalcriteria thesetwogenerainclude thathavebeenusedtodistinguish and paleogeographic distribution, typeof stratigraphic and raystructure. tracheidpitting and cross-field pitting, theimportance (1972) stressed Forexample,Maheshwari in G. skaarensis. ofraystructure Figs.22-24. Seriesshowingleaftracedivergence that the name and concluded Dadoxylon xylemstriped,and by solid areas,secondary Leaf tracesrepresented shouldbe used forwood withmultiseriate rays,while outercorticalandleafbasesbyclearareas.Arrowsshowpositionsofa topartly biseriate woodswithuniseriate xylemraysshould x 20. Bar = 1 mm. pairedleaftraceas itentersthepetiole.All figures However,Prasad(1982) xylem,severalleaf be assignedto Araucarioxylon. 22. Basalmostsectionshowingdissectedsecondary Redrawnfrom451 leafbases.Matureleafat right. tracesandtriangular can be variablethroughout cautionedthatraystructure D bot r, No. 3 (see Fig. 5). 23. More distalsectionwithseveralleaf a woodystemand cannotbe used as a majormeansof tracesin cortex.Redrawnfrom451 D bot r, No. 4 (see Fig. 3). 24. Pant and Singh(1987) have progenericdelimitation. two xylemcylinder, Distalmostsectionshowingcompletesecondary ofthetaxonomic vided further among problems summary petioles(lowerrightand upperleft).Redrawnfrom451 D bot r, No. a newclassification LowerGondwanawoodsand offered 31 (see Fig.2). forthem. woodoftheDadoxylon We nowknowthatpycnoxylic and/orAraucarioxylon typesoccursin at leastthefollow- May 1993] PIGG AND TAYLOR-GLOSSOPTERIS STEMS 509 10~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Fig. 30 is the most proximal level, Fig. 25 the most distal. Leaves are Figs. 25-30. Series showingleaf arrangementin bud of G. skaarensis. phyllotaxy,exceptforperhapsleaf no. 7. Specimen numberedin orde'rof theirpresumedproduction.This numberedsequence followsa two-fifths becomes crushedat Fig. 27 and more distally.All figuresx 44. Bar = 0.5 mm. 25. Redrawn from451 D bot 0, No. 41. 26. Redrawn from451 D bot 0, No. 40. 27. Redrawn from451 D bot 0, No. 39. 28. Redrawn from451 D bot 0, No. 37. 29. Redrawnfrom451 D bot 0, No. 36. 30. Redrawn from451 D bot 0, No. 32. 510 AMERICAN JOURNAL OF BOTANY [Vol. 80 7~~~~~~~~~~~~~~~ 12 Figs. 31, 32. Leaf arrangementin buds of G. schopfii.Bars = 0.5 mm. 31. Redrawn from484 B bot f No. 4. x 42. 32. Redrawn from484 BL bot No. 1. x77. ing groups: cordaites, conifers,Paleozoic and Mesozoic seed ferns,includingglossopterids.The potentialconfusion resultingfromthissituationis illustratedin a recently published table where the genus Dadoxylon is indicated as belongingto the Glossopteridales,while Araucarioxylon remains listed as wood of the Coniferales(Prasad, 1986; see Giraud, 1991). Consideringthe numerousAraucariox.lon stems that reportedlylack details of the primaryxylem,the poorly definedprimaryxylembundles of G. skaarensis may be a real characterof this plant and not merelythe resultof limitedpreservation.Ifso, thenat leastsome glossopterids may have been similarto some Paleozoic gymnosperms that lack well-developed primarybodies such as Heterangium kentuckhvensis (Pigg,Taylor, and Stockey, 1987) and Mesox-vlonpriapi (Trivettand Rothwell, 1985). In comparison withthe various species of Araucarioxylonand othergeneraofGondwana Permianwoods, wood of G. sk-aarensisis extremelysimple. For example, the raystructureconsistingof uni-biseriaterays,one to a few cells high,is one ofthe simplestpossible structuresof this wood type.This wood has a less complex structurethan, forinstance,thattypicalof the V1ertebraria axes foundin the same matrix.The wood is also unlikewood ofAntarcticoxylon,described by Seward (1914) and recentlyreevaluated (Meyer-Berthaudand Taylor, 1991) in known charactersincludingsize and ray structure.Glossopteris compressionswithattachedleaves showa similartracheid anatomy (Pant and Singh, 1974). Some have suggestedthat typeof cross-fieldpittingis taxonomically significantamong many gymnosperms, theconifers(Phillips,1948; Jane,1970). Wood particularly of G. skaarensis shows cross-fieldpits that appear to be eitherof the cupressoid or taxodioid type(Figs. 10, 11). These two types of bordered pits resemble one another superficially,but in the taxodioid type the apertureis May 1993] PIGG AND TAYLOR-GLOSSOPTERIS 511 STEMS TABLE2. Leaf morphology ofglossopterid leaveswithcuticlethatarefoundin attachment tostems Taxon Gl. skaarensisa Gl. maculatab Gl.sastriib Gl. searsolensisb Gl. oldhamiib a b Morphology: Shape Apex Midrib width (mm) Oblanceolate Retuse 1.1 Broadlypolygonal Lanceolate Up to 2 Broadly polygonal Up to 3 Polygonal Up to 4 - 3.5 Long polygonal Obtuse Lanceolate Rounded Lanceolate Spatulate Obtuse, pointed TABLE 58 40 12-40, X = 25 17-50, X = 31 61-94, X= 71 Veryacute 25-42 30-70 70-88 Vein concentration: Center(cm2) Lateral (cm2) 13.7 19 5-13, X = 9 13-20, X= 17 4-9, X= 6 10-20,X= 15 12-20 16-24 10 23 resemble the poorly preserved secondary xylem of our stems. The assignmentof Araguainorachisas a gymnosperm, and possibly a glossopteridstem, ratherthan a seed fernrachis,seems more likely. Other stems from the Skaar Ridge matrix and from Australiansilicifiedpeats currently understudyresemble specimens described by Mussa (1986) fromthe Permian of Brazil as Paulistoxyloninflatum,Piracicaboxylon agrestinum, and Austroscleromedulloxylon geraldinii. Previousstudiesofglossopteridstem/leafattachmentAlthoughspecimens of Glossopteris are most commonly found as detached leaves, attachmentof leaves to stems has been reportedin numerousinstances(Table 1). One of the earliestreports(Dana, 1849) has been regardedas based on poorly preservedmaterialthat cannot be evaluated (Rigby,Maheshwari,and Schopf,1980). In general, glossopteridleaves have been reportedattachedto three different types of axis: 1) Vertebraria-like axes; 2) axes withrhomboidalleafscars;and 3) axes witheithersmooth stems or shallow, longitudinalstriationsof wrinkleson the surfaces.Other specimens show clustersof leaves at- 3. Cuticular featuresofglossopterid leavesfoundin attachment tostemsv Taxon Epidermal cells (lower) (I x w Mm) Epidermal cells: Intercostalshape Shape over veins Gl. skaarensisb 24 x 45 Irregular,sinuous margin Elongate GI. maculatac 88 x 38 Irregular,sinuous margin Gl. sastriie 85 x 23 Irregular,sinuous margin Elongate,straightmargin 105 x 38 Polygonal,sinuous margin Narrow,rectanguloid Gl. searsolensisc Gl. oldhamiic - Irregular,straightmargin Informationobtained fromsources indicated. Generic abbreviationGI Pigg, 1990 and this paper. c Pant and Singh, 1974. b 2.9 x 0.06 2.9 x 0.06 2.5 x 1.5 1.6 x 0.6 5.2 x 1.85 5.8 x 0.9 4.4 x 0.65 5.6 x 0.6 7 x 1 5 x 0.5 Vein angle (): With midrib With margin GI = Glossopteris. Pigg, 1990 and this paper. Pant and Singh, 1974. largerthantheborderand theborderwalls slope outward, ratherthaninward,as in thecupressoidtype(Jane,1970). Cupressoid cross-fieldpits have been considered characteristicof modem araucarian wood, and of the fossil (Prasad, 1982). Whetherthe trangenus Araucarioxylon sitional morphologyseen in the cross-fieldpits of G. skaarensisis significantis unclear,but the classification systembased on cross-fieldpittingdeveloped by Phillips (1948) was establishedon data fromextantconifers.Like stomatal and cuticularstructures,wood charactersmay eventuallyturnout to be less definitiveamong the seed ferns. The G. skaarensisstems also resemble specimens described fromthe Permian of Brazil as Araguainorachis simplissima(Mussa and Coimbra, 1987). The resemblance is based on the lack of pronounced primaryvascular bundles,a broad pithregionwithsimilarlyvariable anatomy,and generalpatternsofpreservationofthewood. are considerablylargerthan our Axes ofAraguainorachis material,on the orderof 6 cm in diam. Mussa and Coimbra (1987) interpretedthis fossil as part of a seed fern frondrachis.The tissuestheyinterpretas corticalremains a Mesh size (cm): Near midrib Near margin Mesh shape Stomata: Distribution Orientation Density (mm2) Hypostomatic Irregular 90.6 Hypostomatic Irregular 36 Hypostomatic Irregular 63 Hypostomatic Irregular 157 Irregular Frequent Glossopteris. Subsidiarycells: Number/Stoma Morphology Width (Jm): Guard cells Stoma 5-6 Papillate 34 x 9 9 Ordinary - 43 x 18 24 x 5 6-7 Papillate 21 x 7 - 39 x 21 Papillate 52.5 x 25 - 512 AMERICAN JOURNAL OF BOTANY [Vol. 80 tachedto shortaxis segmentsthatlack distinguishing 1978). White(1978) suggestedthatall threeaxes bore structures. 1881;Plumstead, 1958;Pant,1967), reproductive features (Feistmantel, whilestillothersare reportedas clustersof leaves that Observation ofthesespecimens byoneofus (KBP) fails lack a preservedstem(e.g.,Plumstead,1958; Anderson toverify White'sinterpretation oftheleaftransition series and Anderson,1985). and fertile natureofthesespecimens.Although a variety thought to be attachedto ofleafmorphologies, Glossopteris leaveswerefirst including smallscales,arefoundin thatshowed thesurrounding Vertebraria axesonthebasisoftwospecimens matrixand in otherspecimensfromthe ridge,pre- Mudgeelocality,detailsofthecentralregionofthespeca leafwithitspetioleadjacentto a transverse axis (Zeiller,1896). imenareobscureand theattachment sumablya node, on a Vertebraria ofleaveswiththese AdditionalspecimensweredescribedbyOldham(1897), variedmorphologies to theG. linearisaxis are notclear Arber(1902), and Dolianiti(1954) as attachedto Ver- (Fig. 33 ofWhite,1978). The "terminalbud" originally longitu- describedbyWalkom(1928) is certainly tebrariaon thebasis of shallow,discontinuous moreplausible dinalwrinkles and deepergroovesthatanastomoseat the as a shootapex bearingscalelikeleaves(Walkom,1928; areas,ortrans- White,1978),butagain,no clearevidencehas beenprenodeto formnarrowelongaterectangular ofGlos- sentedthatthisaxisbearsfertile theseputativeattachments versesepta.Generally leaves.Squamellaampla axes have beendiscounted White,thelargewoodyaxis withlenticular sopterisleavesto Vertebraria scars,is conby most authors(Waltonand Wilson, 1932; Thomas, siderablylargerthanboththeG. linearis(Mudgee)axis anatomicalstud- and Walkom's(1928) axis,and bearslargerleafscarslike 1952;PantandSingh,1974).Moreover, therootingnatureofthese thoseof G. linearisthatare foundproximalto attached ies of Vertebraria underscore axes(WaltonandWilson,1932;SurangeandMaheshwari, vegetativeleaves (Fig. 33 of White,1978, at bottom). 1962;Schopf,1965;Pantand Singh,1968;Gould,1975). Consequently, thereis no basis forinterpreting thisaxis StemswithrhomboidalleafscarsincludetwoAustra- as a cone. Thereis, likewise,no reasonto assumethe leavesde- Mudgeespecimenrepresents lian plants.One is a stembearingGlossopteris a "wholeplant"as asserted scribedby Etheridge (1895), fromMudgee,New South byEtheridge (1895) andWhite(1978),as itcouldas likely linearisby be a branchofa largerwoodyplant. Wales,thatwas laterassignedto Glossopteris White(1978). The secondis a smallplant,Blechnoxylon A secondAustraliangenus,Blechnoxylon talbragarfromtheTalbragarDistrictofNew South ense,is based on impressions of a small,delicateplate talbragarense, Wales (Etheridge,1899), thathas been redescribedas withsomedetailsofinternal anatomy(Etheridge, 1899). byPantand Nautiyal(1984a). Althoughoriginally Glossopteris talbragarense describedas a fern,Blechnoxylon is Bothoftheseplantshave anatomically stems. clearlya seed plantwithsecondaryxylem,nodes and preserved witha petrified internodes, The Mudgeespecimen,an impression ofnarrowdichotomously and clusters veined leavesattached leaveswithdenselydistributed trichomes (Seward,1910). axis,showsa crownofeightGlossopteris toa stoutaxis 15.2cmlongx 1.8cmwide.On thesurface As withG. linearis, resectioning and studyofxylemanatare crowdedrhomboidalleafbase scars.The stemanat- omyofBlechnoxylon mayadd to a clearerunderstanding butlikeG. skaarensisstemsit ofitsrelationships. Pantand Nautiyal(1984a) suggested omyis poorlypreserved, has a broadpith,indistinct primary bundles,and a wide itas a speciesofGlossopteris (G. talbragarense). Although withtheglossopterids zoneofdensesecondary xylem(White,1978).In contrast morphological similarities certainly to theG. skaarensisstems,thepithin theMudgeespec- exist,observationsof the materialby one of us (KBP) somecellswith failedto confirmthe reticulatenatureof the venation imenappearsheterogeneous, containing Detailsofthesecondary describedby Pantand Nautiyal(1984a), althougha few dark,possiblyresinouscontents. information aboutwood mesophyll cellsare preserved thatcouldhave beenmisxylemarenotknown,butfurther further thatthe structure prep- takenforvenation.Etheridge (1899) suggested might possiblybe obtainedthrough smallpyriform bodiesborneby some oftheaxes might aration. natureofthe represent reproductive structures, buttheycouldas likely Etheridge (1895) assertedtheglossopterid mor- be scaleleavesofsmallvegetative itlackedtheVertebraria leaves.Further invesalthough Mudgeespecimen, at thetimeto characterize Glossopteris tigationof thistaxon,and the discoveryof additional phologythought itsrelationship to theglossopterids. stems.This specimenwas laterstudiedbyWhite(1978) material, mayclarify it as G. linearisMcCoybased on itssim- At present,however,the lack of a reticulatevenation who identified leaves of thatspecies. pattern andlackofknowledge about ilaritieswithdetachedvegetative typicalofGlossopteris ofvarying its reproductive On thebasisoffurther associatedleaffragments structures taxonomic precludesfurther series assignment. Whitereconstructed a transitional morphologies, wereborneon theMudgeeaxis The moretypicaloccurrence ofglossopterid leavesatofleavesthatshethought extendedfrom tachedtosmoothstemsorstemswithdelicatelongitudinal (Fig. 35 of White,1978). The transition vegetativeG. linearisleaves through"gangamopterid" striationsor wrinkleshas been reportedby numerous re- authors(Bunbury,1861; Feistmantel, leaves to "scale fronds"to fertilemicrosporophylls 1881;Pant,1967; toSquamellaWhite,a genusshecreatedforisolated Rigby,1967; Pant and Singh,1974; Andersonand Anferred instemmorphology microsporophylls (White,1978). White(1978) also as- derson,1983,1985).Differences may and/orpositionon signeda secondaxisfromMudgeebearingscalelikeleaves be explainedin thecontextofgrowth to thegenusSquamella (Fig. 20 of White,1978). This theplant,ratherthanas evidenceofdifferent stemtypes, describedbyWalkom(1928) as or even species.For example,specimenswithlenticular specimenwas originally ofa larger scars,liketheaxisfromMudgee,shownarrowinternodes, a "terminal bud."A thirdspecimen, consisting scarswas whilein otherspecimens(e.g.,Dolianiti,1954) the inlenticular woodyaxisbearinghelicallyarranged assignedto Squamella ampla White(Fig. 31 of White, dividualleavesare morewidelyspaced.Variationin in- May 1993] PIGG AND TAYLOR-GLOSSOPTERIS ofshortand a combination ternodallengthcouldreflect longshoots,as suggestedby some authors(e.g.,Plumstead,1958; Pantand Singh,1974),or merelyvariation one wouldexpectalonga branchin a treewithproleptic branching. Reconstructionsof Glossopteris-Previous reconstruc- have been suggestedby tionsof the plant Glossopteris Edwards(1935),Seward(1941),Plumstead(1958),Rigby (1967, 1969),Pant(1977),Pantand Singh(1974),Gould and Delevoryas(1977),and Retallackand Dilcher(1988, Etheridge(1895) ofsporiferum"). as "Dictyopteridium as a small of Glossopteris feredthe firstreconstruction woody plant,based on the Mudgee specimenthathe the entireplant.White(1978) has thoughtrepresented on thisspecimen.Edwards also based a reconstruction on Oldham's(1897) (1935),whobasedhisreconstruction and Glossopteris and Seward(1941),describing material, plantas a theglossopterid reconstructed Gangamopteris, thatGlossopteris smallshrub.Plumstead(1958) suggested was a largewoodytreewithdeciduousleaves bornein on branchesoflarge whorlson shortshootsthatoccurred Dadoxylontrunks.Rigby(1967) consideredGangamopofa singlelargetree, teriswalkomiito be smallfragments borne to oppositely bearingshortshootsand alternately theGlossopterisl leaveson longshoots.He reconstructed plantas a tall treewithwhorledsmall Gangamopteris in tufts(Rigby,1969),and as a smallunleavesgrowing branchedtreewitha crownof largeleaves at the top, similarto some moderncycads.A plantwithlongand by Pant and Singh shortshootswas also hypothesized (1974). Gould and Delevoryas(1977) providedperhapsthe to date, reconstruction Glossopteris mostcomprehensive and compressed specimens.Furbased on bothpetrified was suggested of thisreconstruction by therrefinement of Retallackand Dilcher(1988) in theirreconstruction Thisnameis invalid,however,since "Dictyopteridium." betweenthepetrified similarities theysuggested although GouldandDeDictyopteridium, material andcompressed name the anatomicallyprelevoryasdid not formally servedseed-bearing megasporophylls. in theGlossopteris Thereare severalcommonfeatures to date.Theseincludethefolreconstructions presented plantswerearlowingsuggestions:1) the Glossopteris borescent;2) the presenceof longand shortshoots;3) leaveswerebornein whorlsortighthelices;and 4) leaves weredeciduous.The G. skaarensismaterialallowsus to discussand reexaminethebases forthesevariousideas. thematerialwe is unclearwhether 1) Arborescence-It theentirestemofa small, seefromSkaarRidgerepresents woodyplant,or thedistal,possiblydeciduous,branches form.If theseaxes are branches, ofa larger,arborescent thengiventhe lack of largestemsin the silicifiedpeat to thelargerparent thatattachments itis unlikely matrix, by detailsof stemcan be readilyresolved.Correlation wood anatomymaybe possibleifjuvenilewood borne appreciablyfrommaturestem by twigsdoes not differ and Stockey, wood(e.g.,Carlquist,1975;Cevallos-Ferriz suchas tra1990). Furtheranalysisof wood characters cheidlengthwhereavailablemayalso be helpfulin dethepositionofa givenspecimenon theplant termining STEMS 513 plantsfor (e.g.,Carlquist,1975). So far,all glossopterid is knownare at leastmarwhichanyinternalstructure thetinystemsofrBlechnoxylon ginallywoody,including (Etheridge,1899). One would expectgymnospermous bifacialvascularcambium, stemsto have a functioning oftheecologicalhabitoftheplant. regardless 2) Long and shortshoots-A numberof researchers borelongandshortshoots thatGlossopteris havesuggested (e.g.,Plumstead,1958;Pantand Singh,1974;Gouldand Delevoryas,1977;RetallackandDilcher,1988).Presumaretypably,thewhorlsofleavesin whichglossopterids icallyfoundwould be the shortshoots,whilethe less encountered,oppositelyor alternatelyarfrequently ranged,individualleavesoccuron longshoots,including Glossopteris themainaxis(e.g.,Dolianiti,1954).Whether boretrueshortshootslikethoseof Ginkgo,or whether lengthand elongationmerely thevariationin internodal suchas one wouldfindin an extant seasonality reflected hardwoodtree,is notclear.It is also unclear temperate boreits reproductive like Ginkgo,Glossopteris whether, withvegetative on spur shoots,intermixed structures leaves. Glos3) Whorledvs. helicalleafarrangement-Both possessanatomicalevskaarensisand G. schopfii sopteris in a two-fifths idencethatleaveswerehelicallyarranged The suggestion bysomeofa whorledarrangephyllotaxy. mentbased on compressedspecimensmaybe theresult of relatively large,broad leaves attachedalonga small Thereis no compelling areaon shortcrowdedinternodes. boretheirleavesin whorled, evidencethatglossopterids decussate,or othermorespecializedarrangements. 4) Deciduousness-Evidencefordeciduousleaves is based on depositionalratherthandirectanatomicalfealeaves were thatglossopterid tures.The firstsuggestion byPlumstead(1958)whoobserved deciduouswasoffered foundindensemats leavesweretypically thatcompressed of "autumnalbanks."LaterRetallack(1980) correlated and not in varvedstratawithfall/winter, leafoccurrence layers.In G. skaarensisthe petiolesatspring/summer tachedto the stemare presentbeyondthe level where theyare separatefromthe stem,at a level moredistal thanthepositionofan abscissionzone.Thequestiondoes not seem to be one of whetherleaves abscised,since leavesalso produceabscissionlayersuponleaf evergreen suchas Isoetes, fall(in contrastto a fewpteridophytes, therewas a thatdo not).Rather,thequestionis whether with thatwascorrelated relatedto seasonality periodicity olderleaveswereshedindividually leafdrop,orwhether (Chabot as newoneswereadded,as in manyevergreens observationmaybe and Hicks, 1982). One interesting in shalescontaining compressed notedhere:frequently littlevarileaves the glossopteridleaves demonstrate ability,exceptforsize. This may suggestthattotalleaf perhapsinresponsetoa comperiodically, dropoccurred frost stresses binationofphysiological stress, (e.g.,drought, leafdropofthistypein experience etc.).Extantconifers responseto variousair pollutants(F. Telewski,Buffalo and Erie CountyBotanicalGardens,personalcommunication;1992). In some plants,suchas Ginkgo,leaves may be lost in a totalleafdrop eventat the end of a 514 [Vol. 80 AMERICAN JOURNAL OF BOTANY growing season,ratherthanthemoregradualautumnal leafloss typicalofangiosperms. Leafsenescence, theabscissionprocess,and theactualleafdropdo notalways occurtogether in angiosperms, as senescentleavesoften remainon temperate treeswellintothewinter.It maybe thatthe physiological mechanismsassociatedwithleaf intheglossopterids replacement differed fromthosemore typicalof angiosperms. It is also possiblethatwhorlsof glossopterid leavesabscisedas entireshootsystems, like thosefoundin extantconiferssuchas Taxodium.This wouldaccountforthecommonoccurrence offossils showinga disarticulated branchwithleavesattached. The presenceof glossopterid scale leaves,oftenscatteredamongmaturevegetative leavesthroughout thematrix,has been well documentedby manyauthors(e.g., Lacey,van Dijk, and Gordon-Gray, 1975;White,1978). Whilesome scaleshave beendemonstrated to be fertile microsporophylls, othersmaybe vegetativebud scales. The questionof leafpolymorphism in Glossopteris has notyetbeenfullyaddressed. Fromtheevidencethathas beenassembledto dateit is clearthatthe conceptof Glossopteris encompassesa varietyofdifferent typesofplants.Previousreconstructionsofthe"Glossopteris plant"havebeenbroadlybased on limitedinformation. Continuedstudy,particularly of anatomically preserved specimenssuchas G. skaarensis, willincreasetheresolution withwhichwe can viewglossopteridsand allow us to understand moreclearlythe diversity present.Greaterknowledgeof thediversity of thisinteresting groupwillenableus toaddressmoreclearlythebroaderphylogenetic andecologicalquestionsabout theseplantsthatdominated thePermianGondwanalandscape. LITERATURE CITED of Glossopteris.Monograph No. 2. Birbal Sahni Instituteof Palaeobotany. Lucknow, India. 291 pp. DANA,J. D. 1849. Fossil plants.Appendix 1. United StatesExploring Expedition, duringthe years 1838, 1839, 1840, 1842 under the command of Charles Wilkes, U. S. N. 10 (Geology). C. Sherman, Philadelphia, PA. DOLLANITI, E. 1954. A florado Gondwana inferiorem Santa Catarina, IV 0 Genero Vertebraria. Notas preliminarese Estudos. Ministerio da Agricultura, Departmento Nacionalda ProduVdo Mineral,Divisdo de Geologiae Mineralogia, Rio de Janerio(Brasil)81: 1-5. DuToIT, A. L. 1927. The fossilfloraof theUpper Karoo beds. Annals oftheSouthAfrican Museum22: 290-420. EDwARDs, W. N. 1935. A guide to the fossil plants in the British Museum (Natural History),2d ed. BritishMuseum (Natural History),London. ETHERIDGE, R. 1895. On the mode of attachmentof the leaves or frondsto the caudex in Glossopteris;withremarkson the relation of thegenus to its allies. ProceedingsoftheLinnean SocietyofNew SouthWales9: 228-258. 1899. On a fern(Blechnoxylontalbragarense),withsecondary wood, forminga newgenus,fromthecoal measuresoftheTalbragar District,New South Wales. Records of The AustralianMuseum 3: 135-146. 1904. Sub-reniform-ovate leaves of Glossopteris,withfurther remarkson the attachmentof its leaves. Records of the Geological SurveyofNewSouthWales7: 315-318. FARABEE,M. J.,E. L. TAYLOR, ANDT. N. TAYLOR. 1991. LatePermian palynomorphsfromtheBuckleyFormation,centralTransantarctic Mountains, Antarctica. ReviewofPalaeobotany andPalynology 69: 353-368. FEISTMANTEL, 0. 1881. The floraof the Damuda-Panchet divisions, in The Fossil Flora of the Gondwana system.Memoirs ofthe Geo- logicalSurveyofIndia,Palaeontologica Indica3: 78-149. FRANcis,J. E., K. J. WOOLFE,M. J. ARNOT, AND P. J. BARRETrr.1991. Permian fossilforestsand climates of Allan Hills, Antarctica.Abstracts,EighthInternationalSymposium on Gondwana, Subcommission on Gondwana Stratigraphy (IUGS), June1991, University of Tasmania, Hobart, Tasmania. p. 32 (Abstract). GIRAUD,B. 1991. Les especes du genreDadoxylon depuis 1962: leur r6partitionet leur 6volution du Permien a la findu M6sozoique. ReviewofPalaeobotany and Palynology 67: 13-40. GOULD, R. E. 1975. A preliminaryreporton petrifiedaxes of Vertebrariafromthe PermianofeasternAustralia.In K. S. W. Campbell ANDERSON,J.M.,ANDH.M.ANDERSON. 1983. Palaeofloraofsouthern [ed.],Gondwana geology,109-115. AustralianNational University, AfricaMoltenoFormation (Triassic).A. A. Balkema,Rotterdam. Canberra. ofsouthern Africa.Prodromus , AND --- . 1985. Palaeoflora ofSouthAfrican megafloras Devonianto LowerCretaceous. A. A. evi, AND T. DELEVORYAS. 1977. The biologyof Glosopteris: dence frompetrifiedseed-bearingand pollen-bearingorgans.AlBalkema,Rotterdam. 1902. On theClarkecollection cheringa1: 387-399. offossilplantsfromNew SouthWales.Quarterly Journal oftheGeological of the GlossopteriSocietyofLondon HOLMES,W. B. K. 1974. On some fructifications dales from the Upper Permian of N. S. W. Proceedings of the 58: 1-27. LinneanSocietyofNewSouthWales98: 131-141. BANERJEE,M. 1978. GenusGlossopterisBrongniartand itsstratigraphic 1990. AustroglossawalkomiiHolmes, a glossopteridovulate in thePalaeozoicsofIndia.Part1. A revisional significance study fructification fromthe Late Permian of New South Wales. ProofsomespeciesofthegenusGlossopteris. BulletinoftheBotanical ceedingsofthe 3rd InternationalOrganizationofPaleobotanyConSocietyofBengal32: 81-125. ference,1988. Melbourne. BARRETT, P. J.,AND D. H. ELLIOT. 1973. Reconnaissance geologicmap oftheBuckleyIslandQuadrangle. Transantarctic Ant- JANE, F. W. 1970. The structureofwood, 2d ed. revisedby K. Wilson Mountains, and D. J. B. White. Adam & Charles Black, London. arctica.UnitedStatesGeologicalSurveyMap A-3. UnitedStates Joy, K. W., A. J. WILLIS,AND W. S. LACEY. 1956. A rapid cellulose GeologicalSurvey,Washington, DC. acetatepeel methodin paleobotany.Annals ofBotany20: 635-637. BASINGER,J.F.,ANDG. W. ROTHWELL. 1977. Anatomically preserved KRAUSEL,R. 1962. Antarcticfossilwood. Appendix. In E. P. PlumplantsfromtheMiddleEocene(AllenbyFormation) ofBritish Costead, Fossil florasof Antarctica.Transantarcticexpedition 1955lumbia.CanadianJournalofBotany55: 1984-1990. 1956, Science Report (Geology), 133-154. Trans-AntarcticExpeoffossilplantsfrom BUNBuRY, C. J. F. 1861. Noteson a collection dition Committee,London. Nigpur,centralIndia.Quarterly JournaloftheGeologicalSociety ofLondon17: 325-349. , P. K. MAITHY,AND H. K. MAHESHWARI.1961. GymnosperS. 1975. Ecologicalstrategies mous woods with primarystructuresfrom Gondwana rocks-a of xylemevolution.UniCARLQuIsT, versity ofCalifornia Press,Berkeley, CA. review.ThePalaeobotanist 10: 97-107. S. R. S., AND R. A. STocyEy. 1990. Vegetative LACEY,W. S., D. E. VAN DuK, AND K. D. GoRDoN-GRAY. 1975. Fossil CEvALLOs-FERRIZ, remainsof the Magnoliaceaefromthe Princetonchert(Middle plants fromthe Upper Permian in the Mooi River District,Natal, Eocene)ofBritish Columbia.CanadianJournal 1327ofBotany68: South Africa.Annals of theNatal Museum 22: 349-420. 1339. LEPEKHINA, V. G., AND A. A. YATsENKo-KHMELEvsKY.1966. ClasCHABOT, B. F., AND D. J.HicKs. 1982. The ecology sification ofleaflifespans. and nomenclatureofwoodsofPalaeozoic pycnoxylic plants. AnnualReviewofEcologyand Systematics Taxon5: 66-70. 13: 229-259. Indianspecies MAHESHwARI, H. K. 1972. Permianwood fromAntarcticaand revision CHANDRA, S., ANDK. R. SURANGE. 1979. Revisionofthe ARBER, E. N. May 1993] PIGG AND TAYLOR-GLOSSOPTERIS STEMS 515 of some Lower Gondwana wood taxa. Palaeontographica 138B: 143. McLoUGHLIN, S. 1990a. Palaeobotany and palaeoenvironmentsof Permianstrata,Bowen Basin, Queensland. Ph. D. dissertation,Universityof Queensland. St. Lucia, Queensland. and leaves 1990b. Some Permian glossopteridfructifications fromthe Bowen Basin, Queensland, Australia.Review of Palaeo- to MiddleTriassicmegaG. 1980. Late Carboniferous RETALLACK, Survey Basin.Bulletin oftheGeological fromtheSydney fossilfloras ofNewSouthWales26: 384-430. of selectedseed , AND D. L. DILCHER. 1988. Reconstructions ferns. AnnalsoftheMissouriBotanicalGarden75: 1010-1057. walkomiisp. nov. Recordsof RIGBY,J. F. 1967. On Gangamopteris Museum27: 175-182. TheAustralian B., AND T. N. TAYLOR. 1991. A probable conifer MEYER-BERTHAUD, fromthe Triassic of Antarctica.Rewithpodocarpacean affinities Geosciencias 27: 3-13. and othercycadopsidfructifi. 1978a. Permianglossopterid cationsfromQueensland.GeologicalSurveyofQueenslandPubPaper41, 1-21.QueenslandDepartlication367 Palaeontological mentofMines,Brisbane. GeoinQueensland. . 1978b. ThePermianplantBelemnopteris Pa367 Palaeontological logicalSurveyofQueenslandPublication ofMines,Brisbane. per42, 25-27. QueenslandDepartment a florain biostratigraphy: 1983. The roleoftheGlossopteris intheReidsDome Beds.PermianGeology assessment preliminary QueensofQueensland, 221-229.GeologicalSocietyofAustralia, landDivision.Brisbane,Queensland. , H. K. MAHESHwARI, AND J. M. SCHOPF. 1980. Revisionof PermianplantscollectedbyJ.D. Dana during1839-1840in Australia.GeologicalSurveyof QueenslandPublication376 PaleonofMines,BristologicalPaper47, 1-25.QueenslandDepartment bane. reconstruction ROYER, S. M. 1988. IBM PC-basedthree-dimensional for system(HVEM-3D). User'smanual,version1.2. Laboratory CelofMolecular, Department Microscopy, HighVoltageElectron ofColorado,Boulder, Biology, University lularandDevelopmental 62: 11-40. botanyand Palynology 67: 179-198. and Palynology viewofPalaeobotany MUSSA,D. 1986. Eustelos gondwanicosde medulas diafragmadase a Paleobotdnica e Palinologia na America sua posiciio estratigrafica. de Universidade de Geociencias, do Sul- 1985-BoletimInstituto Sdo Paulo 17: 11-26. , AND A. M. CoiMBRA. 1987. Novas perspectivascomparaciio entreas tafofloraspermianas (de lenhos) das bacias do Paranaiba e do Parana. Anais do X Congressobrasileirode paleontologia,1925 de julho, 1987, Rio de Janeiro,Brazil, 901-923. Sociedade Brasileirade Paleontologia,Rio de Janeiro,Brazil. with part of the OLDHAM,R. D. 1897. On a plant of Glossopteris rhizome attached,and on the structureof Vertebraria.Records of theGeologicalSurveyofIndia 30: 45-50. PANT,D. D., 1967. On thestemand attachmentof Glossopterisleaves. MemorialVolume)17: (PanchananMaheshwari Phytomorphology 351-359. The JournaloftheIndian 1977. The plantof Glossopteris. BotanicalSociety56: 1-23. Feist, AND A. CHOUDHURY. 1977. On thegenusBelemnopteris 164B: 153-166, mantel.Palaeontographica talbra, AND D. D. NAuLrTYAL.1984a. Noteson Glossopteris * 1969. The Lower Gondwana scene. Boletim Paranaense de CO. In J. B. SCHOPF, J. M. 1965. Anatomyof theaxis in Vertebraria. oftheAntarctic. Antarctic garense(EtheridgeJr.)comb. nov. and some otherAustralianglosHadley[ed.],Geologyand Paleontology sopterids.Phyta 4, 5: 33-37. number1299,217-288.American ResearchSeries6, Publication . 1984b. On the morphologyand structureof OtUnionoftheNationalAcademyofSciences-National Geophysical , AND of Glossopteris.Patokaria zeillerisp. nov., a femalefructification DC. ResearchCouncil,Washington, laeontographica193B: 127-152. 1970. Petrified peatfroma Permiancoal bed in Antarctica. indica Science169:274-277. , AND R. S. SINGH. 1968. The structureof Vertebraria Royle. Palaeontology 11: 643-653. fromPermianand plantassemblages 1973. The contrasting . 1974. On thestemandattachment ofGlossopteris In A. Loganand L. V. continents. Triassicdepositsin southern , AND and Gangamopterisleaves. Part II-Structural features.PalaeonHills [eds.],The Permianand TriassicSystemsand theirmutual CalGeology, tographica147B: 42-73. 379-397.CanadianSocietyforPetroleum boundary, gary,Alberta. , AND V. K. SINGH. 1987. Xylotomyof some woods fromRaniin glosstructures offertile ganj Formation (Permian), Raniganj Coalfield, India. Palaeontointerpretation . 1976. Morphologic graphica 203B: 1-82. E. W. 1948. Identificationof softwoodsby theirmicroscopic PHILLIPS, Products ResearchBulletin22: 1-56. structure. Forestry PiGG,K. B. 1988. AnatomicallypreservedGlossopterisandDicroidium fromthecentralTransantarcticMountains.Ph. D. dissertation,The Ohio State University.Columbus, OH. . 1990. Anatomicallypreserved Glossopterisfoliage fromthe centralTransantarcticMountains.ReviewofPalaeobotanyand Palynology66: 105-127. AND S. McLouGHLIN. 1992. Comparison ofanatomicallypreservedglossopteridleaves fromthe Bowen Basin, Queensland, and Sydney Basin, New South Wales, Australia. Resumes des communications,Abstracts,Organisation internationalede Paleobotanique IVeme Conference,Paris, p. 123 (Abstract). and Glossopteris , AND T. N. TAYLOR 1990. Permineralized DicroidiumfromAntarctica.In T. N. Taylorand E. L. Taylor [eds.], of Gondwana, Antarcticpaleobiology:its role in thereconstruction New York, NY. 164-172. Springer-Verlag, , AND R. A. STocKEY. 1987. Paleozoic seed ferns:Het--- , sp. nov., fromtheUpper Carboniferousof erangiumkentuckyensis North America. AmericanJournalof Botany 74: 1184-1204. E. P. 1958. The habitofgrowthofGlossopteridae.TransPLUMSTEAD, actionsand Proceedings oftheGeologicalSocietyofSouthAfrica 61: 81-96. M. N. V. 1982. An annotatedsynopsisof Indian Palaeozoic PRASAD, woods.ReviewofPalaeobotanyand Palynology gymnospermous 38: 119-156. 1986. XylotaphofloraoftheKamthi Formation,Indian Lower importanceof its Gondwana, withremarkson the biostratigraphic taphoflora.Palaeontographica 201 B: 111-134. sopteridgymnosperms.ReviewofPalaeobotanyand Palynology21: 25-64. A. C. 1910. Fossilplants,vol.2. Cambridge Press, University SEWARD, Cambridge. Museum fossilplants.BulletinoftheBritish 1914. Antarctic Geology1: 1-49. (NaturalHistory), . 1941. Plant lifethroughtheages. CambridgeUniversityPress, Cambridge. 1975. Morphologyof the gymSURANGE, K. R., AND S. CHANDRA. of the Glossopterisfloraand theirrelanospermousfructifications 149B: 153-180. Palaeontographica tionship. flora , AND H. K. MAHESHwARI 1962. Studiesin the Glossopteris of India. 11. Some observations on Vertebrariafromthe Lower Gondwanas of India. The Palaeobotanist 9: 61-67. preservedPermTAYLOR,E. L., AND T. N. TAYLOR. 1990. Structurally ian and Triassic florasfromAntarctica.In T. N. Taylor and E. L. Taylor [eds.], Antarcticpaleobiology:its role in the reconstruction New York,NY. ofGondwana,149-163.Springer-Verlag, AND J.W. COLLINSON.1989. Depositional settingand ofPermianandTriassicpermineralized peatfromthe paleobotany Antarctica. Peatandcoal:origin, Transantarctic central Mountains, facies,and depositionalmodels. InternationalJournalof Coal Geology 12: 657-679. AND N. R. CuNEo. 1992. The presentis not the key to the past: a polar forestfromthe Permian of Antarctica.Science 257: 1675-1679. withwhorledleaves. ThePalaeoTHOMAS,H. H. 1952. A Glossopteris botanist1: 435-438. Bulletin a newtypeoffertile Glossopteris. 1958. Lidgettonia, Museum(NaturalHistory), Geology3: 179-189. oftheBritish 516 AMERICAN JOURNAL OF BOTANY A. 1950. The primaryvascularbody ofMesoxylonthompTRAVERSE, sonii, a new American cordaitalean.AmericanJournalof Botany 37: 318-325. TRIVErr,M. L., AND G. W. ROTHWELL. 1985. Morphology,systematics,and paleoecologyofPaleozoic fossilplants:Mesoxylonpriapi, sp. nov. (Cordaitales). SystematicBotany 10: 205-223. VOGELLEHNER, D. 1964. Zur Nomenklaturder fossilenHoltzgattung Dadoxylon Endlicher 1874. Taxon 13: 233-237. WALKOM,A. B. 1928. Notes on some additions to the Glossopteris florain New South Wales. Proceedingsof the Linnean Society of New South Wales 35: 555-564. [Vol. 80 WALTON,J., ANDJ. A. R. WILSON. 1932. On the structureof Verte- 52: 200-207. oftheRoyalSocietyofEdinburgh braria.Proceedings WHITE, M. E. 1978. Reproductivestructuresof the Glossopteridales in the plant fossilcollection of The AustralianMuseum. Records Museum31: 473-505. ofTheAustralian ZEILLER, R. 1896. Etude sur quelques plantes fossiles,en particulier Des environsde Johannesburg(TransVertebraria et Glossopteris. duFrance3rdSeries24: 349-377. vaal)Bulletin SocieteGeologique
© Copyright 2026 Paperzz