the microfossil record of early land plants - david moore...fossil record. the fossil record of...

The microfossil record of early land plants

Charles H Wellman1 and Jane Gray21Centre for Palynology University of Shecurreneld Dainton Building Brook Hill Shecurreneld S3 7HF UK

2Department of Biology University of Oregon Eugene OR 97403-1210 USA

Dispersed microfossils (spores and phytodebris) provide the earliest evidence for land plants They arecentrst reported from the Llanvirn (Mid-Ordovician) More or less identical assemblages occur from theLlanvirn (Mid-Ordovician) to the late Llandovery (Early Silurian) suggesting a period of relative stasissome 40 Myr in duration Various lines of evidence suggest that these early dispersed microfossils derivefrom parent plants that were bryophyte-like if not in fact bryophytes In the late Llandovery (late EarlySilurian) there was a major change in the nature of dispersed spore assemblages as the separated productsof dyads (hilate monads) and tetrads (trilete spores) became relatively abundant The inception of triletespores probably represents the appearance of vascular plants or their immediate progenitors A little laterin time in the Wenlock (early Late Silurian) the earliest unequivocal land plant megafossils occur Theyare represented by rhyniophytoids It is only from the Late Silurian onwards that the microfossilmegafossil record can be integrated and utilized in interpretation of the poundora Dispersed microfossils arepreserved in vast numbers in a variety of environments and have a reasonable spatial and temporalfossil record The fossil record of plant megafossils by comparison is poor and biased with only a dozenor so known pre-Devonian assemblages In this paper the early land plant microfossil record and itsinterpretation are reviewed New discoveries novel techniques and fresh lines of inquiry are outlined anddiscussed

Keywords early land plants early embryophytes bryophytes tracheophytes spores phytodebris

1 INTRODUCTION

The earliest fossil evidence for embryophytes consistsentirely of dispersed microfossils These oldest generallyaccepted microfossils are from the Llanvirn (Mid-Ordovician) but it is not until the Wenlock (LateSilurian) some 40 Myr later that the oldest unequivocalland plant megafossils occur The early land plant micro-fossil record consists of dispersed spores and phytodebris(fragments of cuticles and tubes) Such remains arepreserved in large numbers in a variety of environments(continental and marine) because (i) they were com-posed of recalcitrant material and readily fossilized(ii) they were produced in vast numbers and (iii) theyhad the potential for long-distance dispersal by wind andwater due to their small size Consequently they have anexcellent fossil record with good spatial and temporalrepresentation However the dispersed microfossils aredicurrencult to interpret because they represent discrete partsof whole organisms and phytodebris is fragmentaryNonetheless comparisons with purportedly homologousstructures in extant plants and to a certain extentyounger fossil plants provide compelling evidence thatthey derive from land plants As the earliest evidence forland plants they are utilized in interpretation of thenature and acurrennities of the plants from which they deriveFurthermore their excellent fossil record documents theearly evolution and diversicentcation of land plants andprovides information on the composition and distribution

of early terrestrial vegetation In this review the micro-fossil record for early embryophytes is documented andinterpretation of these remains discussed

2 DISPERSED SPORES

(a) The fossil recordThe earliest dispersed embryophyte spores that appear

creditable are reported from the early Llanvirn (Vavrdova1984) of Bohemia (ie Perunica the palaeogeographicterm applied to the Cambro-Devonian of CentralBohemia A J Boucot personal communication) andfrom the mid-Llanvirn of Saudi Arabia (Strother et al1996) well-separated palaeogeographically from PerunicaApart from Vavrdovarsquos limited material (she only reportsspore tetrads) more or less identical spore assemblageshave been reported from the Llanvirn (Mid-Ordovician)to Llandovery (Early Silurian) (Gray 1985 1991Richardson 1988 Strother 1991 Wellman 1996 Steemans1999) These early spores have unusual concentgurationsmorphology in comparison with mature spores of extantembryophytes (centgures 2 and 3) and are sometimestermed cryptosporesrsquo a name originally used to repoundecttheir unfamiliar appearance and lack of knowledgeregarding the nature of their producers (see Richardson1996a and references therein) Cryptospores comprisemonads and permanently united dyads and tetrads thatare either naked or enclosed within a thin laevigate orvariously ornamented envelope The nature of cohesionbetween spores in permanently united dyads and tetradsis unclear (Wellman 1996) They are termed unfused if

Phil Trans R Soc Lond B (2000) 355 717^732 717 copy 2000 The Royal Society

Deceased 9 January 2000

there is a supercentcial line of attachment between thespores and cohesion probably results from localizedexospore links or bridges rather than large-scale fusionThey are termed fused if there is no discernible line ofattachment and cohesion probably results from fusionover most or all of the contact area Cohesion may alsoresult from enclosure within a tight-centtting envelope Thecomposition of cryptospores (wall and envelope) isunknown but their preservation in ancient depositssuggests they constitute sporopollenin or a sporopollenin-type macromolecule Since there seems little question thatthese structures are spores the term cryptospore has notaxonomic usefulness but we retain the term here fordescriptive convenience

Reports of OrdovicianÊarly Silurian dispersed sporesare relatively few but are stratigraphically and geogra-phically widespread and indicate that cryptosporeassemblages are remarkably constant in composition

(both temporally and spatially) throughout this interval(Gray 1985 1988 1991 Richardson 1996a Strother et al1996 Wellman 1996 Steemans 1999) These data suggestthat the vegetation was widespread but of limited diver-sity with little evolutionary change (at least in the spores)during an interval some 40 Myr in duration

Over much of the globe a major change in the natureof spore assemblages is reported in the late Llandovery(late Early Silurian) (Gray 1985 1991 Gray et al 1986Richardson 1988 1996a Burgess 1991 Wellman 1996Steemans 1999) (centgure 1) While naked monads dyadsand tetrads continued to dominate spore assemblages theabundance of envelope-enclosed forms is seriouslyreduced and in places they appear to virtually disappearAt the same time hilate monads and trilete spores centrstbecame widespread and occur in relative abundanceHilate monads comprise single spores with a circularcontact area on their proximal surface and clearly

718 C HWellman and J Gray The microfossil record of early land plants

PhilTrans R Soc Lond B (2000)

Ordovician

envelope-enclosedtetrads

envelope-encloseddyads

envelope-enclosedmonads

naked tetrads

naked dyads

naked monads

hilate monads

trilete sporesL

lanvirn

Llandeilo

Caradoc

469

464

443

439

430

424

411409

396

Ashgill

Llandovery

Wenlock

Ludlow

Lochkovian

Pragian

epochstage

Pridoli

Silurian Devonian period

Figure 1 Stratigraphical rangechart for early land plant sporemorphotypes Time-scale fromHarland et al (1989) Regardingannotation adjacent to the strati-graphic column the small `Cooksoniarsquoindicates the age of the earliestunequivocal land plant megafossilsthe asterisks indicate the age ofimportant plant mesofossilassemblages (ie LudfordLane-Pridoli North Brown CleeHill-Lochkovian) The dashedtime-line indicates the position ofthe major change in the nature ofdispersed spore assemblages in theLlandovery The width of range barsprovides a very basic indication ofrelative abundance to illustrate grosschanges in abundance through timeNote that range bar width has noquantitative implications The taperand question marks at the base ofrange bars extending into theLlanvirn indicate the age of theearliest known spore assemblagesand do not necessarily coincide withtheir centrst appearance nor necessarilywith changes in abundance

The microfossil record of early land plants C HWellman and J Gray 719

Phil Trans R Soc Lond B (2000)

(a) (b) (c) (d ) (e) ( f )

(g) (h) (i)( j)

(k)

(l) (m) (n) (o)

( p)

(q) (r) (s)

(t) (u) (v) (w)

Figure 2 LM images of dispersed early land plant spores (a^k) from the type area for the Caradoc Shropshire England(Caradoc Late Ordovician) (Wellman 1996) (l^w) from the Stonehaven Group of the north-east Midland Valley of Scotland(late Wenlock Late Silurian) (Wellman 1993b) (ab) Naked unfused permanent tetrad (pound 1500) (cd ) naked unfusedpermanent dyad (pound 1500) (e f ) naked fused permanent dyad (pound 1500) (g h) envelope-enclosed permanent tetrad (pound 1500)(i j) envelope-enclosed permanent dyad (pound 1500) (k) envelope-enclosed monad (pound 1500) (l) naked fused permanent tetrad(pound 1000) (mô) naked monad (pound 1000) ( pq) laevigate hilate monad (pound 1000) (rs) ornamented hilate monad (pound 1000)(tu) laevigate trilete spore (pound 1000) (vw) ornamented trilete spore (pound 1000)

formed in dyads that dissociated prior to dispersal(centgures 2 and 3) Similarly trilete spores comprise singlespores with a distinct trilete scar and three contact areason their proximal surface and clearly formed in tetradsthat dissociated prior to dispersal (centgures 2 and 3) Disso-ciation of the products of (presumably) meiosis suggests amajor change in plant reproductive strategy (see Gray1985Wellman et al 1998a Edwards et al 1999)

Both trilete spore and hilate monad abundance anddiversity increased throughout the Late Silurian It waspreviously considered that their centrst appearance inabundance and associated changes in dispersed sporeassemblages was near-contemporaneous on a globalscale However recent reports from Ashgill (UpperOrdovician) Lower Llandovery (Early Silurian) depositsfrom Turkey and Saudi Arabia support an early appear-ance of both hilate monads and trilete spores in thisregion of northeast Gondwana (Steemans et al 19962000 Wellman et al 2000a) However the hilate monadsand trilete spores occur in low numbers and do not

coincide with a decrease in the abundance of envelope-enclosed cryptospores which in this region persist inrelative abundance at least into the Late Silurian

Sculptured forms of both hilate monads and triletespores (centgure 2) appear in the late Wenlock (LateSilurian) and structural^sculptural innovations ensued asboth groups proliferated throughout the remainder of theSilurian and earliest Devonian (eg Richardson ampMcGregor 1986 Burgess amp Richardson 1995 Steemans1999) Again the appearance of sculptured hilate monadsand trilete spores wasbelieved to be near-contemporaneouson a global scale but recent reports from Saudi Arabiaindicate that ornamented hilate monads may occurearlier (in the Llandovery) in this region (Steemans et al2000) Other morphotypes (naked dyads and tetrads)persisted in the hilate monadtrilete spore-dominatedassemblages but as an increasingly insignicentcant elementIn the earliest Devonian hilate monad numbers began todecline (possibly rapidly) until cryptospores became aminor component of spore assemblages that were



(a) (b)(c) (d )

(e)

( f )

(g) (h) (i)

( j) (k) (l) (m)

Figure 3 SEM images of dispersed early land plant spores (aî) From the type area for the Caradoc Shropshire England(Caradoc Late Ordovician) (Wellman 1996) ( j^m) from the Silurian inliers along the southern margin of the Midland Valleyof Scotland (early Wenlock Late Silurian) (Wellman amp Richardson 1993) (a) Naked unfused permanent tetrad (pound 2000)(bc) naked unfused permanent dyad (pound 1500) (de) naked fused permanent dyad (pound 1500) ( f^h) envelope-enclosed perma-nent dyad (pound 1500) (i) envelope-enclosed permanent monad (pound 1500) ( j) naked fused permanent tetrad (pound 1210) (k) loosetetrad (pound 1070) (l) laevigate trilete spore (pound 1380) (m) laevigate hilate monad (pound 1230)

dominated by a bewildering variety of trilete spores (egRichardson amp McGregor 1986) Cryptospores persist inspore assemblages at least until the end of the Lochkovian(Early Devonian) Their upper stratigraphic range ispoorly documented but they are seldom reported frompost-Lochkovian strata

For the Late SilurianÊarly Devonian interval there isconsiderable variation in the calibre of reporting for thedispersed spore record with diiexclerent time intervalsexhibiting substantial diiexclerences in the number andquality of reports and the extent of their palaeogeogra-phical coverageThis creates problems when attempting toidentify palaeophytogeographical variations which arefurther enhanced by dicurrenculties associated with recog-nizing the eiexclects of small-scale variation dependent onlocal environmental control (eg substrate altitude etc)subtle facies eiexclects and non-contemporaneity due to incor-rect correlation of strata Nonetheless certain patterns arebeginning to emerge and are summarized below Recentdetailed reviews are provided by Richardson (1996a)Steemans (1999) and Edwards amp Wellman (2000)

For the Late Silurian there is acceptable reporting ofspore assemblages from the palaeocontinents LaurentiaAvalonia and Baltica which were closely associated andconstituted part of a large cluster of land that straddledthe Equator and from northern Gondwana which layfurther to the south and was separated by a relativelylarge ocean Reporting from elsewhere is sparseSequences of spore assemblages from Laurentia Avaloniaand Baltica are notably similar in composition suggestingthat this land mass contained a poundora representing a singlepalaeophytogeographical realm and subtle diiexclerencesprobably repoundect small-scale variation related to localizedvariations in environmental conditions However se-quences of spore assemblages from northern Gondwana(eg Spain north Africa and Brazil J Gray unpublisheddata) although sharing many elements in common withthose from Laurentia Avalonia and Baltica also exhibitmarked diiexclerences suggesting they belong to a diiexclerentpalaeophytogeographical realm (Richardson amp Ioannides1973 Rodriguez 1983 Tekbali amp Wood 1991)

By earliest Devonian times Laurentia Avalonia andBaltica were all essentially part of the same largecontinent (the Òld Red Sandstone continentrsquo) which wasnow not far removed from Gondwana the distancebetween them having diminished through the LateSilurian The vast majority of spore assemblages knownfrom the Lochkovian (Early Devonian) is from the ÒldRed Sandstone continentrsquo and northern Gondwana withfew reported assemblages from elsewhere One mightexpect diiexclerences between the sequences of spore assem-blages from these two regions to become less apparent asthe two palaeophytogeographical realms moved closertogether and permitted interaction between their biotaand preliminary research seems to indicate that this wasthe case However comparisons between Lochkovianspore assemblages are dicurrencult because localized variationappears to have become more prevalent by this timehampering the identicentcation of larger-scale variationThe small-scale variation probably repoundects a combinationof eiexclects including variation due to localized diiexclerencesin environmental conditions and facies eiexclects which aredicurrencult to disentangle For example in the Òld Red

Sandstone continentrsquo there are diiexclerences between coevalspore assemblages from upland intermontane andlowland poundoodplain deposits (Richardson et al 1984Wellman 1993a Edwards 1996 Wellman amp Richardson1996 Wellman et al 2000b) It is unclear to what extentthese repoundect subtle facies eiexclects or variation (small- orlarge-scale) in the distribution of poundora due to variation inenvironment (altitude climate substrate etc) Howeverit is clear that cryptospores are far more abundant in thelowland poundoodplain deposits than in the upland inter-montaine deposits possibly because the former were adamper environment more suited to the cryptospore-producing plants (eg Wellman amp Richardson 1996Wellman et al 2000b)

(b) Acurrennities(i) Introduction

Evidence for the acurrennities of early higher land plantspores (cryptospores and trilete spores) is derived fromfour main sources (i) occurrence (ie depositional envir-onment) of the dispersed fossil spores (ii) inferencesbased on comparison with the spores of extant land plants(size and morphology) (iii) studies of land plant fossilspreserving in situ spores (iv) analysis of spore wall ultra-structure There is abundant evidence suggesting thattrilete spores represent the reproductive propagules ofland plants (see Gray 1985 1991 and references therein)However evidence of similar acurrennities for the crypto-spores particularly the earliest occurrences has beenmore centercely contested (eg Banks 1975) Nonethelessrecent centndings provide convincing evidence for higherland plant acurrennities and there are now few who disagreewith such an origin (see Gray 1985 1991 Strother 1991Richardson 1992 Taylor 1996 Edwards et al 1995a 1998)The evidence for higher land plantembryophyte acurrennitiesis summarized below

(ii) Occurrence of early land plant sporesIt has long been noted that early land plant spores are

distributed in a similar range of depositional environmentsin which the sporespollen of extant land plants occur andwith similar abundances Their occurrence in continentaland nearshore marine deposits (with abundances usuallydecreasing oiexclshore) is wholly consistent with their repre-senting the subaerially released spores of land plantswhich were transported to their sites of deposition throughthe actions of wind and water However while there arenumerous examples of spore assemblages derived fromcontinental deposits from the Late Silurian and EarlyDevonian few examples exist for the OrdovicianÊarlySilurian interval These centndings are almost certainly anartefact of the stratigraphical record the OrdovicianÊarly Silurian was a time of persistently high sea levelsand fewer continental deposits are known and those thatdo exist possess geological characteristics unsuitable forthe preservation of organic-walled microfossils (eg unsui-table lithologies andor high thermal maturity) Theearliest known spore assemblages preserved in continentaldeposits are from the Ashgill (Gray 1988) Llandovery(Pratt et al 1978 Johnson 1985 Gray 1988) and Wenlock(Strother amp Traverse 1979 Wellman 1993b Wellman ampRichardson 1993) All of these reported assemblagesinclude cryptospores and trilete spores except those from



the Ashgill palaeokarst on Manitoulin Island Ontario(Gray 1988) the early Llandovery Tuscarora Formationin the Millerstown section Pennsylvania (Gray 1988) andthe early Llandovery Tuscarora Formation at the MillHall locality Pennsylvania (Johnson 1985) which are theoldest and appear to contains only cryptospores

(iii) Comparisons with extant embryophyte sporesEarly higher land plantembryophyte spores are

similar to the reproductive propagules of extant landplants in terms of size gross morphology and possessionof a thick sporopollenin spore wall (regarded as asynapomorphy for embryophytes) Sporopollenin wallsmay have multiple functions (Graham amp Gray 2000) onebeing to protect propagules during transport followingsubaerial release (eg Blackmore amp Barnes 1987) Thusthe possession of such walls in early higher land plantspores provides excellent evidence that they werefunctionally similar to their modern counterpartsFurthermore the small size of early land plant spores iswithin the range of subaerially dispersed spores producedby extant free-sporing plants Based largely on analogywith the reproductive propagules of extant embryophytesGray (1985 1991 and references therein) has arguedpersuasively that the obligate spore tetrad of the fossilrecord is an ancestral character in embryophytes and thatsuch tetrads derive from land plants at a bryophyte mostlikely hepatic grade of organization She notes thatamong extant free-sporing embryophytes only hepaticsregularly produce permanent tetrads as mature sporessome of which are contained within an envelope similarto those enclosing certain fossil spore tetrads but that atetrad regularly occurs in the spore ontogeny of embryo-phytes The acurrennities of hilate monads and dyads aremore equivocal primarily because such morphologies donot have an obvious modern counterpart (Wellman et al1998a) either in mature spores or in spore ontogenyDyads rarely occur in extant (non-angiosperm) embryo-phytes and only through meiotic abnormalities (Fanninget al 1991 Gray 1993 Richardson 1996b Wellman et al1998ab) The abundance of dyads in early land plantspore assemblages indicates that they were commonlyproduced and are therefore probably not the products ofmeiotic abnormalities Their occurrence is most comfort-ably explained by invoking successive meiosis withseparation occurring following the centrst meiotic divisionand sporopollenin deposition on the products of thesecond division It has been noted that monads dyadsand tetrads often have identical envelopes and someauthors have suggested that they are closely relatedperhaps even deriving from a single species (Johnson1985 Richardson 1988 1992 Strother 1991 Hemsley1994) However centxing acurrennity on the basis of a singlecharacter is problematic Trilete spores dissociated fromjuvenile tetrads have a clear counterpart among extantembryophytes where their production is widespreadamong free-sporing tracheophytes and also occurssporadically among bryophytes (eg Gray 1985)

(iv) Fossil plants preserving in situ sporesStudies of in situ spores supply the only direct link

between the dispersed spore and plant megafossilrecords and are critical to our understanding of the

acurrennities of dispersed spore types Unfortunatelyhowever the plant megafossil record for the OrdovicianÊarly Devonian is relatively poor (see Edwards 1990Edwards amp Wellman 2000) Plant megafossils arepractically unknown until the Late Silurian probablybecause the vast majority of plants believed to be at abryophyte-like grade of organization lacked the appro-priate recalcitrant tissues suitable for preservationHence there are no in situ spore records for the centrst 50million or so years of higher land plant evolution Whenthey eventually appear land plant megafossils areinitially very rare becoming progressively morecommon in younger sediments (Edwards amp Wellman2000) They are usually preserved as coalicented compres-sions and in this mode of preservation in situ spores aregenerally absent or when present too poorly preservedto permit comparisons with better-preserved dispersedspores Occasionally however exceptional preservationof plant megafossils preserves in situ spores in sucurrencientdetail to enable such comparisons (Allen 1980 Gensel1980) Suitable preservation occurs at the Ludford Lane(Late Silurian-Pridoli) and North Brown Clee Hill (EarlyDevonian-Lochkovian) localities from the Welsh Border-land and the record of in situ early land plant spores isbased primarily on material from these localities (seereviews by Fanning et al 1991 Edwards 1996 this issueEdwards amp Richardson 1996) At both localities extre-mely small plant fragments (sometimes referred to asmesofossils sensu Hughes (1994)) are preserved as rela-tively uncompressed coalicentcations that preserve exquisitecellular detail (Edwards 1996)

The record of early land plants with in situ spores hasmost recently been reviewed by Edwards (1996 thisissue) Edwards amp Richardson (1996) and Edwards ampWellman (2000) It is noteworthy that most cryptosporemorphotypes have been recovered in situ (naked andenvelope-enclosed permanent tetrads and dyads andhilate monads) However it is unclear if the parent plantsrepresent relict populations and provide a true repoundectionof earlier cryptospore-producing plants or if the crypto-spores are plesiomorphic in more advanced plants orperhaps even arose due to convergence (Gray 1991Edwards this issue) It must be borne in mind that thefossils occur some 65 Myr after the earliest reported cryp-tospores from the Llanvirn (Mid-Ordovician)

Interpretation of the parent plants is not alwaysstraightforward as the mesofossils are fragmentary Usuallyonly terminal parts of the axes ( sect sporangia) arepreserved and cellular detail is variable Furthermoremany of the `rhyniophytoidrsquo plants preserve unusual char-acter combinations confusing considerations of acurrennitiesHowever trilete spores have been recovered from therhyniophyte Cooksonia pertoni (Fanning et al 1988) whichis demonstrably a true tracheophyte (Edwards et al 1992)and the vast majority of trilete spore-producers appear tohave constituted plants with bifurcating axes terminalsporangia and often stomata Interestingly some dyadsand tetrads derive from plants with bifurcating axesspor-angia (Edwards et al 1995a 1999 Wellman et al 1998a) acharacter not represented among extant bryophytes (seeEdwards this issue) Another interesting observation isthe presence of stomata on plants containing in situ hilatemonads (Habgood 2000 Edwards this issue) Stomata



are absent from liverworts but present in hornwortsmosses and vascular plants (although losses are notuncommon in these groups and are generally consideredto be related to ecological factors and functional require-ments) (see review in Kenrick amp Crane 1997)

Consideration of dispersed spore assemblages fromLudford Lane and North Brown Clee Hill allows us togauge the extent to which the in situ record repoundects thedispersed record Permanent tetrads and dyads (includingenvelope-enclosed forms) were relatively minor compo-nents of dispersed spore assemblages dominated by hilatemonads and trilete spores Studies of in situ spores repoundectthis situation in that specimens with tetrads and dyads areextremely rare but those containing hilate monads andtrilete spores are abundant However not all trilete sporemorphotypes have been discovered in situ In situ patinatespores and in situ Emphanisporites-type spores are extremelyuncommon (Edwards amp Richardson 2000) althoughboth spore types are abundant in dispersed spore assem-blages and even coprolites from the mesofossil-bearinghorizons These centndings may be a consequence of(i) preservational artefact (the missing spore types mayderive from plants with low preservation potential)(ii) palaeogeographical eiexclects (the missing spore typesmay derive from plants from outside of the depositionalbasinoumlan unlikely scenario as the spores are preserved incoprolites at this locality) (iii) facies eiexclects (the missingspore types may derive from larger plants not representedamong the assemblages of small highly sorted meso-fossils)

(v) Spore wall ultrastructureIt is well established that analysis of spore wall ultra-

structure characters can be extremely procenttable whenattempting to ascertain the phylogenetic relationships ofextant land plants and similar research has been extra-polated back in time and is now routinely undertaken onfossil spores (eg Kurmann amp Doyle 1994) Such researchis also of paramount importance in studies of spore walldevelopment Recently there has been a surge of interestin wall ultrastructure in early higher land plantembryo-phyte spores and it is hoped that exploitation of thispotentially extremely rich data source will providecharacters useful in ascertaining the acurrennities of theseancient plants and shed light on the nature of spore walldevelopment

To date studies of wall ultrastructure in early landplant spores are in their infancy and two principal lines ofinquiry have been explored Some of the earlier crypto-spores have been studied based on analysis of isolateddispersed spores from the Late Ordovician (Ashgill)Êarly Silurian (Llandovery) of Ohio USA (Taylor1995ab 1996 1997) Later cryptospores and trilete sporeshave been studied based on analysis of in situ spores excep-tionally preserved in mesofossils from the latest Silurian(Ludford Lane) and earliest Devonian (North BrownClee Hill) localities in the Welsh Borderland (Rogerson etal 1993 Edwards et al 1995b 1996a 1999 Wellman 1999Wellman et al 1998ab) Studies on Late SilurianÊarlyDevonian dispersed spores remains an unexploited butpotentially extremely useful data source Nonethelessmany early land plant spore morphotypes have now beenultrastructurally examined Taylor has studied early

examples of naked and envelope-enclosed tetrads (Taylor1995b 1996 1997) and naked and envelope-encloseddyads (Taylor 1995a 1996 1997) Later tetrads (Edwardset al 1999) dyads (Wellman et al 1998a) and hilatemonads (Wellman et al 1998b) have also been examinedas has a variety of trilete spore taxa (Rogerson et al 1993Edwards et al 1995b 1996a Wellman 1999)

In terms of ascertaining phylogenetic relationshipscentndings to date are rather dicurrencult to interpret with noclear patterns emerging regarding the relationshipsbetween diiexclerent mesofossil taxa in situ sporemorphology and wall ultrastructure This is most likely aconsequence of the frailty of the database as studies areextremely limited to date and there are major gaps inour knowledge However interpretation of wallultrastructure in early embryophyte spores is alsoproblematical due to a number of technical andtheoretical factors These include (i) diiexclerent workersemploy slightly diiexclering techniques each with itsassociated artefacts which may be dicurrencult to recognize(ii) diagenetic eiexclects may vary rendering comparisonsproblematical particularly when comparing extant withfossil plants or fossils from diiexclerent localities withdiiexclering diagenetic histories (iii) unlike extant plantswhere complete ontogenetic sequences can be studiedfossils usually preserve only a particular ontogenetic stateand although they usually represent mature forms theexact stage of maturation may be unclear (iv) it is dicurren-cult to detect convergence resulting from similarities indevelopmental processes (v) it is uncertain to whatextent the spore wall ultrastructure characters of extantplants diiexcler from such distant ancestors Nonethelessthese problems are not insurmountable If the databasecontinues to improve at its current rate `noisersquo created bythe above-mentioned problems will probably be centlteredout by an abundance of reliable data It is likely that inthe future spore wall ultrastructural studies will play anincreasingly important role in phylogenetic analysis ofearly land plants Taylor (1997) recently summarized hiscentndings and proposed a tentative hypothesis for evolu-tionary relationships among early cryptospore producersHe suggests that at least two separate lineages occur butgoes on to say that `the phylogenetic relationship betweenthese groups and to more recent land plants remainsuncertainrsquo although he has suggested possible hepaticacurrennities for some of the dyads (eg Taylor 1995a) Thequestion of lineages among permanent tetrads has beendiscussed by Gray (1991) based on envelope characters

In terms of understanding spore wall development inearly land plants studies of spore wall ultrastructure haveresulted in a number of important centndings Lamellae havebeen recognized in trilete spores (Wellman 1999) dyads(Taylor 1995a) and hilate monads (Wellman et al 1998b)including the presence of typical white-line-centredlamellae in the latter Such centndings provide the earliestfossil evidence for the antiquity of such structures andprovide further evidence that sporopollenin deposition onthese structures is the most primitive mode of sporopol-lenin deposition among land plants (eg Blackmore ampBarnes 1987) Important information has also beenprovided on the nature of cryptospore envelopes (Taylor1996 1997 Wellman et al 1998a Edwards et al 1999) andjunctions between units in permanent tetrads and dyads



(Taylor 1995a 1996 1997 Wellman et al 1998a Edwards etal 1999) For example Edwards et al (1999) and Habgood(2000) have recently suggested based on evidence derivedfrom ultrastructural studies of fossil material that crypto-spore envelopes may have been tapetally derived apossibility previously discussed by Gray (1991)

(vi) ConclusionsStudies on early land plant spores are providing an

important source of information in attempts to under-stand the acurrennities and phylogenetic relationships of earlyhigher land plants This is particularly true for theearliest land plants which left no megafossil record as weare dependent on the dispersed microfossil record as ouronly source of information Analysis of these earliest landplant spores suggests that the producers were bryophyte-like if not bryophytes (Gray 1985 1991) It is anticipatedthat our understanding of early land plant spores willcontinue to improve as further localities are discoveredand more work is undertaken on these and pre-existinglocalities particularly if these searches should turn upbody fossils of the spore producers Additionally it isprobable that some of the identicented gaps in knowledgewill be centlled following further research on in situ sporesand spore wall ultrastructure which may shed furtherlight on the evolutionary relationships of early landplants Researchers are also utilizing diiexclerent methods ofphylogenetic analysis on the primary data set Forexample Kenrick amp Crane (1997) have recently under-taken a detailed cladistic analysis of early land plantswhich incorporates some of the data derived from studiesof early land plant spores At this juncture it is interestingto note that certain recent cladistic analyses (based onmorphological andor molecular data) indicate that the`bryophytesrsquo are paraphyletic with respect to the tracheo-phytes but the three major lineages of `bryophytesrsquo (liver-worts hornworts and mosses) are all monophyletic andoccupy basal positions within the embryophytes with oneof these lineages (probably the mosses) sister group to thetracheophytes (see Mishler et al 1994 Kenrick amp Crane(1997) and references therein) Thus it might be anticipatedas previously suggested (Gray 1985 1991 and other refer-ences) that the earliest land plants were bryophyte-like(cryptospore-producers) with tracheophytes (triletespore-producers) appearing somewhat later However itmust be borne in mind that early land plant classicentcationand phylogenetic interpretation are in a state of poundux andthe biological integrity and evolutionary relationships ofmany higher taxa are uncertain (eg Kenrick amp Crane1997) Furthermore many early fossil plants such as therhyniophytoids are of unknown acurrennity and it is possiblethat the diversity of higher taxa is masked

(c) Observations on the dispersed spore fossil record(i) Completeness and integrity

The early land plant dispersed record varies in terms ofthe quantity and quality of reports and their palaeogeo-graphical coverage for diiexclerent stratigraphical levels Twogeneralizations can be made (i) as one moves up columnthe dispersed spore record becomes better known(ii) there is an overrepresentation of reports from easternNorth America (Laurentia) western Europe (LaurentiaAvalonia and Baltica) and North AfricaÂrabia (northern

Gondwana) with other areas either under-represented orsimply not represented at all Obviously this createsproblems for interpretation of the dispersed spore recordparticularly when attempting to identify patterns ofevolution and palaeophytogeographical variation on aglobal scale

(ii) Patterns of evolution and palaeophytogeographyComments about evolutionary and palaeophytogeogra-

phical patterns are necessarily highly speculative In partthis is due to gaps and probable biases associated with thedispersed spore record and in part it is due to the state ofpoundux in palaeogeographical models Palaeophytogeogra-phical speculations here and elsewhere in this paper arebased on the Scotese amp McKerrow (1990) model Itshould be realized that observations based on any onemodel including this one will probably be subject tomajor revision in the future

Their centrst known appearance in the Llanvirn (Mid-Ordovician) appears to have been followed over much ofthe globe by an extended period of stasis of ca 40 Myrwith little evolutionary development until the lateLlandovery (Early Silurian) During this period the vastmajority of dispersed spore assemblages consist entirely ofcryptospores believed to derive from plants at a bryo-phyte-like (hepatic) grade of organization if not in factbryophytes A notable exception is the early occurrence oftrilete spores in northern Gondwana (see p 4)

Gray et al (1992) suggest that variation in the palaeo-geographical distribution of envelope-enclosed permanenttetrads (manifested in type of envelope ornament)provides the centrst evidence for phytogeographical diiexcleren-tiation among early land plants during the OrdovicianÊarly Silurian They suggest that a cool MalvinokaiexclricRealm and a warmer extra-Malvinokaiexclric Realmexisted However certain reported data apparentlyconpoundict with their centndings (see discussion in Wellman1996 p130 and Steemans et al 2000) More palaeogeo-graphically extensive reporting is required to clarify thismatter

It has been highlighted that many of the early occur-rences of cryptospores are from high latitudes (Gray et al1992 Edwards 1998) in areas associated with ice caps forat least some of this time As there is little variation inspore assemblages over this interval it has been suggestedthat the cryptospore-producers were little aiexclected byclimate changes associated with the glaciation and couldrapidly reinvade previously glaciated areas (Richardson1996a Edwards 1998) Such centndings are consistent withthe interpretation of Gray (1984 1985) that the crypto-spores-producers were at a bryophyte-like grade oforganization with life history strategies that included anecophysiological tolerance to desiccation and a shortvegetative lifestyle and thus could tolerate wide extremesof climate

There was a major event in the late Llandovery that inmany areas saw the near disappearance of envelope-enclosed cryptospores which apparently coincided withthe appearance of hilate monads and trilete spores inrelative abundance (cf Gray et al 1986) It has beensuggested that the inception of trilete spores may repoundectthe centrst appearance of tracheophytes (rhyniophytes) (egGray 1985 Edwards amp Wellman 2000) The acurrennities of



hilate monads are more controversial Some appear toderive from rhyniophytoids (a group erected for plants ofuncertain acurrennities) some may represent tracheophyteswhile others may represent non-vascular plants perhapsincluding `primitiversquo groups ancestral to the vascularplants Other cryptospore morphotypes remained aspersistent although not abundant elements of dispersedspore assemblages attesting to the continuation of certaincryptospore-producing plants although there was clearlya serious reduction in the importance of the forms produ-cing envelope-enclosed spores As more evidence becomesavailable the possibility that trilete spores centrst appearedin northern Gondwana seems more likely fuelling specu-lation that this may have been a centre of originrsquo fortracheophytes

Following the turnover in the nature of dispersedspore assemblages noted above trilete spores and hilatemonads had a major diversicentcation which probablyrepoundects a diversicentcation in the early tracheophytes (andpossibly also their putative ancestorsrsquo that may occuramong the rhyniophytoids) At about this time (lateWenlockêarly Ludlow) we centnd the centrst evidence forsignicentcant palaeophytogeographical diiexclerentiation (seep 5) which is not surprising as this was a time whensome of the larger land masses were widely separatedEventually towards the end of the Lochkovian hilatemonad abundance diminishes leaving a poundora dominatedby trilete spore-producing plants with cryptospore-producing plants being a minor component It may besignicentcant that this patternoumla reduction in palaeophy-tographical diiexclerentiationoumlhas been documented inareas that indicate a reduction in continental separationbut that paradoxically show increased provincialism inmarine invertebrates and vertebrates (A J Boucotpersonal communication) The post-Lochkovian (EarlyDevonian) paucity of cryptospores attests to large-scaleextinction among some of the `lineagesrsquo at bryophytelevel of organization and possibly some early tracheo-phytes andor their putative ancestors included in therhyniophytoids It relates as well to tetrad dissociationwithin at least one `lineagersquo of tetrad producers (seediscussion in Gray 1985 1991)

(iii) Measures of diversityThere is often a reluctance to use dispersed spore diver-

sity as a proxy for land plantsrsquo diversity because ofconcerns that the former may not necessarily repoundect thelatter It is frequently noted based on observations ofboth extant and fossil plants that (i) the same taxonmay produce a variety of diiexclerent spore morphologiesleading to an overestimate in diversity (ii) similar sporesare produced by diiexclerent plant groups resulting in anunderestimate of diversity

There is no in situ record for the earliest cryptosporeproducers and hence it is dicurrencult to test whether or notdispersed spore diversity repoundects true diversity Moreoverseveral potential pitfalls have been noted First it is clearthat many taxa are morphologically simple and it ishighly likely that such similar forms could have beenproduced by a number of plant types (ie homoplasyabounds) but it would be dicurrencult or impossible to distin-guish between their dispersed spores Second it has beennoted that diiexclerent morphotypes (monads dyads and

tetrads) possess identical envelopes and it has beensuggested that they might have been produced by thesame parent plant taxon (Johnson 1985 Richardson 19881992 Strother 1991) Third envelope-enclosed crypto-spores may potentially lose their envelopes during trans-portation or diagenesis producing forms that aresecondarily naked and cannot be distinguished fromthose that never had an envelope (Gray 1991)

The occurrence of early land plants containing in situspores from the latest Silurianêarliest Devonian allowsus to examine critically the relationship between thedispersed spore and plant megafossil record at this time(eg Fanning et al 1991) Findings suggest that sporeformation was usually simultaneous within a singlesporangium producing spores of only one type (seesummaries in Fanning et al 1991 Edwards 1996 Edwardsamp Richardson 1996) However Fanning et al (1988)found that morphologically identical plant megafossilsmay produce diiexclerent spore types and suggested thatthis repoundected reticulate evolution where evolution of sporemorphology was rapid compared with that in themorphologically (although not necessarily anatomically)simple plants Furthermore Wellman et al (1998b) foundthat supercentcially morphologically similar spores wereproduced by diiexclerent plant types Both of these centndingscould lead to discrepancies in diversity data derived fromcounts of dispersed spore `speciesrsquo However it is clearthat the supposedly morphologically similar megafossilsspores can usually be distinguished if analysed inadequate detail

There have been few sucurrenciently detailed investigationsof early land plant dispersed spore diversity published todate Counts based on spore genera are totally inadequatedue to the ad hoc methods utilized in the creation of taxaof this taxonomic rank Counts based on species arealmost certainly more reliable Steemans (1999) providescounts of genera and species for both cryptospores andtrilete spores for the interval Caradoc (Late Ordovician)Lochkovian (Early Devonian) based on selectedpublications from geographically dispersed localities Hiscentndings seem to concentrm the general diversity changesnoted in the synthesis of the early land plant spore recordprovided above Similarly preliminary counts based onspecies abundances in geographically isolated sequences(the Òld Red Sandstone continentrsquo and northernGondwana) (C H Wellman unpublished data) concentrmthese general trends However one must bear in mind(i) potential distorting eiexclects as noted above particu-larly for the older records (OrdovicianÊarly Silurian)where there are no in situ records available to testdispersed sporeparent plant relationships (ii) variablespatial andor temporal coverage due to diiexclerences in theavailability andor integrity of data

3 DISPERSED PHYTODEBRIS

Enigmatic dispersed fragments (phytodebris) believedto derive from embryophytes andor fungi have longbeen known from the OrdovicianEarly Devonian andhave provided an important contribution to our under-standing of early land plants and terrestrial ecosystems(eg Gray 1985 Sherwood-Pike amp Gray 1985 Genselet al 1991 Edwards amp Wellman 1996) They consist



primarily of fragments of cuticle tubular structures(aseptate) and centlaments (septate) which occur isolated orin complex associations (centgures 4 and 5) The acurrennities ofmany forms are controversial as they lack a convincingmodern counterpart although recent advances have gonea long way towards clarifying their biological relation-ships They undoubtedly derive from non-marineorganisms because they occur in continental deposits andhave a similar distribution to equivalent (analogous orhomologous) fragments derived from extant land plants(cuticles conducting tissues) and fungi (centlaments)

(a) The fossil recordThe earliest fragments of cuticle are reported from the

Caradoc (Late Ordovician) (Gray et al 1982) and theyare relatively abundant in the Llandovery (EarlySilurian) By the Wenlock (Late Silurian) ornamentedcuticles and forms clearly derived from tracheophytesare present The fossil record of OrdovicianÊarlyDevonian dispersed cuticles has recently been reviewedby Gray (1985) Gensel et al (1991) Edwards amp Wellman(1996) and Edwards et al (1996b)

The earliest tubular structures are ornamented formsfrom the early Llanvirn of Bohemia (Vavrdova 1984) andsmooth forms from the Ashgill of southern Britain(Burgess amp Edwards 1991) A proliferation of diiexclerenttypes of tubular structure (including types with externalornament and others with internal annular or spiral

thickenings) occurs in the Wenlock (Late Silurian) andsimilarly diverse forms exist until at least the EarlyDevonian In addition to variation in their internal andexternal ornament tubular structures vary in presenceabsence of branching and nature of terminal structures (ifpresent) They commonly occur in complex associations

The earliest reported centlaments are from the Llandovery(Early Silurian) and they have been reported sporadicallythroughout the Silurian and Lower Devonian The centla-ments are usually branched sometimes with poundask-shapedprotuberances and the septa may or may not be perforate(centgure 5c) The occurrence of OrdovicianEarly Devoniantubular structures and centlaments is reviewed by Gray(1985) Burgess amp Edwards (1991) Gensel et al (1991)Wellman (1995) and Edwards amp Wellman (1996)

(b) AcurrennitiesThe acurrennities of dispersed phytodebris are conjectural

and aroused much controversy in the past Early sugges-tions that they derived from land plants andor fungiwere based largely on inferences following comparisonswith similar structures in extant forms (summarized inGray 1985) However caution was advised (eg Banks1975) and it was even suggested that some forms mayderive from marine organisms although such an originhas now been rejected following their recovery fromcontinental deposits More recently the early land plantmegafossilmesofossil record has turned up a number of



Ordovician Silurian

rsquonematophytersquo cuticle

non-banded tubes

banded tubes

higher land-plant cuticle

stomata

tracheids

Llanvirn

Llandeilo

Caradoc

469

464

443

439

430

424

411409

396

Ashgill

Llandovery

Wenlock

Ludlow

Lochkovian

Pragian

epochstage

Pridoli

Devonian period

Figure 4 Stratigraphical rangechart for early land plantphytodebris Time-scale fromHarland et al (1989) Regardingannotation adjacent to thestratigraphic column the small`Cooksoniarsquo indicates the age of theearliest unequivocal land plantmegafossils the asterisks indicatethe age of important plant mesofossilassemblages (ie LudfordLane-Pridoli North Brown CleeHill-Lochkovian) The dashedtime-line indicates the position ofthe major change in the nature ofdispersed spore assemblages in theLlandovery

interesting fossils relevant to the debate Other avenues ofresearch such as geochemical analysis have also beenexplored In the following account important contribu-tions to our understanding of the acurrennities of phytodebrisare summarized

(i) Inferences based on similar structures in extant organismsIt has long been recognized that the dispersed cuticles

are morphologically similar to the cuticles of extant landplants However it is not until the Wenlock (LateSilurian) that cuticles can be unequivocally assigned totracheophytes These show marked alignment ofepidermal cells often with stomata that clearly derivefrom axes and forms with well-decentned outlines occasion-ally with attached spores that derive from theirsporangia However the earlier forms which persist untilat least the Early Devonian are more enigmatic Theylack stomata and have cellular patterns unlike those in

extant land plants However cuticle appears to be anadaptation for protection in subaerially exposed plantsand its presence therefore suggests that at least part of theparent plant was subaerially exposed at least some of thetime

The dispersed tubular structures have less convincinganalogues among structures present in extant land plantsIt has long been noted that forms with internal annularspiral thickenings resemble the tracheids of extanttracheophytes (centgure 5f ) They are often referred todescriptively as tracheid-like tubes in the older literature(Gray amp Boucot 1977 1979) However these tubes clearlydiiexcler from tracheids in terms of structure Nonetheless ithas frequently been suggested that some of the diiexclerentforms of tubular structures may have functioned asconducting cells even if they are not strictly homologouswith the tracheids of tracheophytes Gray amp Boucot (1977pp164^168 1979 p 62) have thoroughly reviewed the



(a)

(c) (d ) (e) ( f )

(g) (h) (i)

(b)

Figure 5 LM and SEM images of nematophytes and dispersed phytodebris (abhi) courtesy of Professor Dianne Edwards fromthe Ditton Group of North Brown Clee Hill Shropshire England (Lochkovian Early Devonian) (Edwards 1996) (c f ) from theLower Old Red Sandstone of Lorne Scotland (earliest Lochkovian Early Devonian) (Wellman 1995) (deg) from the Silurianof the Arabian peninsula (a) Nematophyte (pound 65) (b) close up of (a) revealing tubular organization including banded andlaevigate forms (pound 1075) (c) fungal hypha (pound 320) note poundask-shaped protuberance and internal septae (d ) laevigate tube(pound 320) (e) tube with external ornament (pound 320) ( f ) banded tube (pound 250) (g) dispersed cuticle (pound 870) (h) Tortilicaulissporangium with banded tubes attached via an amorphous centlm which in places forms pustules ( pound 1000) (see Edwards et al1996b) (i) Close up of (h) illustrating an internally thickened tube attached via the amorphous centlm with pustules (pound 550)

question of the possible relationship of these tracheid-liketubes to other organisms and concluded as we do herethat most could be attributed to nematophytes

Some of the dispersed centlaments bear a striking resem-blance to the hyphae of extant fungi and it has beensuggested that they provide evidence for terrestrial fungiof ascomycete acurrennity (Sherwood-Pike amp Gray 1985)Additional more compelling evidence is in the form ofassociated spores including multiseptate spores that aredicurrencult to interpret as anything but ascomycetaceousascospores or the conidial stage of an ascomycete(Sherwood-Pike amp Gray 1985)

Recently Kroken et al (1996) suggested that some ofthe dispersed cuticles and tubular structures may repre-sent the fragmentary remains of fossil bryophytes Theirhypothesis is based on observations on the fragments thatsurvive after extant bryophytes are treated with high-temperature acid hydrolysis (ie recalcitrant fragmentsthat one would expect to survive in the fossil record) Thefragments bear supercentcial resemblance to certain earlyland plant phytodebris although there are a number ofdiiexclerences relating to the size and symmetry of thefossils This novel avenue of research is exciting and mayin the future prove extremely rewarding One should bearin mind that this research is in its infancy and to dateonly a few extant bryophyte taxa have been examined

(ii) The plant megafossilmesofossil recordThe early land plant megafossil record consists of

unequivocal tracheophytes (rhyniophytes zosterophyllslycopsids etc) rhyniophytoids of uncertain status inaddition to the enigmatic nematophytes It has beendemonstrated that higher land plants are the source ofsome of the dispersed cuticles since at least the Wenlock(Late Silurian) a not unsurprising centnding consideringthe abundance of fragmented land plant cuticle that isincorporated into present-day sediments However it is tothe enigmatic nematophytes that many have turned intheir search for the source of many of the other dispersedremains

Nematophytes are a group of putative land plants ofuncertain acurrennity with somatic organization based on acomplex association of tubes (often including tubes withinternal annular or spiral thickenings) and believed byLang to possess a cuticular covering (Lang 1937 see alsoGray 1984) They are known principally through ratheruninformative coalicented compressions (eg Lang 1937Strother 1988) However uncompressed fragments areabundant in the recently discovered exceptionallypreserved mesofossil assemblages from the Welsh Border-land and these specimens are providing valuable newinformation on the morphology and anatomy of theseorganisms (centgure 5ab) It is clear that many of thedispersed tubular structures (particularly laevigate andinternally thickened forms) are similar to those in nema-tophytes particularly as some of the dispersed formsregularly occur in associations similar to those observedin the megafossilsmesofossils However the covering ofnematophytes is dicurrencult to examine in the megafossilsmesofossils and there is some doubt as to whether it is thesource of dispersed fragments of cuticle Nonethelessrecent chemical analyses demonstrate that the dispersed`nematophytersquo cuticle does not have the same chemical

composition as that from axial higher land plants addingweight to the suggestion that they are fundamentallydiiexclerent and raising queries concerning their acurrennitiesand function (Edwards et al 1996b)

It has been suggested that some of the nematophytes(Prototaxites) were terrestrial fungi (eg Hueber 1996)Recent evidence supporting this hypothesis is thediscovery of internally thickened tubes seemingly growingon (Edwards et al 1996b 1998) and within the tissues(Edwards amp Richardson 2000) of higher land plants(centgure 5hi) It is possible that these tubes were eitherattacking the plant while alive (pathogens) or dead(decomposers)

Another interesting possibility is that some of thenematophytes represent lichens The reported `lichenizedcyanobacteriarsquo in the Rhynie Chert (Early Devonian)(Taylor et al 1995 1997) suggests great antiquity forthese organisms However Poinar et al (2000) state thatthere is no indication that the Devonian associationwhich involves a zygomycete and a cyanobacterium is astable one of the type demanded in the physiologicalinteraction of a lichen symbiosis Indeed Taylor et al(1997 p1003) note that the fossil `lacks the structuralorganization of the modern lichen thallusrsquo Moreover theresemblance of the `mycobiontrsquo to Zygomycota isadditional cause for scepticism since the mycobionts ofextant lichens are usually ascomycetes or rarely basidio-mycetes No extant true lichen involves a zygomycete(Poinar et al 2000)

Fungal or lichen acurrennities for nematophytes mustremain in the realm of speculation although eitherpossibility might help explain the anomalous chemicalcomposition of cuticlersquo attributed to this taxon

(iii) ConclusionsIt is clear that the acurrennities of many forms of dispersed

phytodebris remain conjectural While many of theyounger dispersed cuticles derive from higher land plantsa nematophyte origin seems most plausible for some ofthe remainder The majority of tubular structures prob-ably also derives from nematophytes and some dispersedcentlaments doubtless derive from terrestrial fungiHowever the acurrennities of the nematophytes remain highlycontroversial although some see evidence for fungal orlichen acurrennities (Hueber 1996 Edwards et al 1996b 1998Edwards amp Richardson 2000)

(c) Observations on the dispersed phytodebris fossilrecord

It is clear from the early land plant microfossil andmegafossil record that nematophytes were an importantelement of early terrestrial ecosystems (cf Gray amp Boucot1977 1979 Gray 1984) both in terms of abundance anddiversity They include a variety of forms includingthalloid (eg Nematothallus) and axial (eg Prototaxites) Italso seems likely that they included by far the largestorganisms in early terrestrial ecosystems (specimens ofPrototaxites over 1m wide have been reported) Recentwork on latest Silurianêarliest Devonian exceptionallypreserved mesofossil assemblages from the Welsh Border-land is providing further details of nematophyte structureand revealing a diverse array of new forms Howeverreproductive structures remain unknown and the



acurrennities and mode of life of these bizarre tissuesorganisms remain obscure

4 CONCLUSIONS

Reconstruction and interpretation of the early terres-trial vegetation is hampered by the inadequacies of theearly land plant fossil record which is incomplete andbiased This is particularly the case for the earliest landplants as there is no megafossil record and we rely ondispersed microfossils which provide more limitedevidence as to the nature and acurrennities of the producersHowever centnds of new fossils continue to provide evidencethat increases our understanding of early land plants andterrestrial ecosystems An example is the wealth of newinformation derived from the reports of Ordovician^Lower Silurian microfossil assemblages from SaudiArabia that is forcing us to reconsider previous inter-pretations regarding the evolution and palaeophyto-geography of early land plants Similarly a wealth of newdata is coming to light based on work on the latestSilurianêarliest Devonian exceptionally preserved meso-fossil assemblages from the Welsh Borderland Work onthe mesofossils is demonstrating the vast diversity of earlyland plants while examination of dispersed spore assem-blages from the same localities indicates that many of thespore-producers have not yet been identicented and truediversity is in fact even larger It is imperative that wecontinue to locate new material (microfossil mesofossiland megafossil) and centll in the gaps in the temporalspatial record of early land plants For example there areas yet no reports of Upper Silurian^Lower Devoniandispersed spore assemblages from Australasia a continentfamed for the anomalous early occurrence of relativelylarge lycopsids (Garratt 1978)

In addition to new material new techniques and novellines of enquiry are also being explored There are anumber of notable examples Chemical analysis of fossilmaterial such as that on dispersed cuticles performed byEdwards et al (1996b) is producing exciting resultsalthough identicentcation and interpretation of diageneticeiexclects is currently proving frustrating Comparisonsbetween dispersed phytodebris and fragments of extantbryophytes that are likely to survive in the fossil record(eg Kroken et al 1996) may eventually prove fruitfulStudies of wall ultrastructure in early land plant sporesare still in their infancy but are already providing awealth of new information (eg Taylor 1997 Wellman etal 1998)

We would like to thank A J Boucot who read a number ofversions of this manuscript and oiexclered many helpfulsuggestions and Dianne Edwards who generously allowed us toreproduce photographs of her fossil material

REFERENCES

Allen K C 1980 A review of in situ Late Silurian andDevonian spores Rev Palaeobot Palynol 29 253^270

Banks H P 1975 The oldest vascular land plants a note ofcaution Rev Palaeobot Palynol 20 13^25

Blackmore S amp Barnes S H 1987 Embryophyte spore wallsorigin homologies and development Cladistics 3 185^195

Burgess N D 1991 Silurian cryptospores and miospores fromthe type Llandovery area south-west Wales Palaeontology 34575^599

Burgess N D amp Edwards D 1991 Classicentcation of uppermostOrdovicianto Lower Devonian tubular and centlamentous maceralsfromtheAnglo-Welsh Basin BotJ Linn Soc10641^66

Burgess N D amp Richardson J B 1995 Late Wenlock to earlyPridoli cryptospores and miospores from south and southwestWales Great Britain Palaeontographica B 236 1^44

Edwards D 1990 Constraints on Silurian and Early Devonianphytogeographic analysis based on megafossils In Palaeozoicpalaeogeography and biogeography (ed W S McKerrow amp C RScotese) pp 233^242 Memoir No12 LondonThe GeologicalSociety

Edwards D 1996 New insights into early land ecosystems aglimpse of a Lilliputian world Rev Palaeobot Palynol 90159^174

Edwards D 1998 Climate signals in Palaeozoic land plantsPhilTrans R Soc Lond B 353 141^157

Edwards D amp Richardson J B 1996 Review of in situ spores inearly land plants In Palynology principles and applications vol 1Principles (ed J Jansonius amp D C McGregor) pp 391^407Salt Lake City UT American Association of StratigraphicPalynologists Foundation Publishers Press

Edwards D amp Richardson J B 2000 Progress in recon-structing vegetation on the Old Red Sandstone continent twoEmphanisporites producers from the Lochkovian of the WelshBorderland In New perspectives on the Old Red Sandstone (edP F Friend amp B P J Williams) Geological Society SpecialPublication

Edwards D amp Wellman C H 1996 Older plant macerals(excluding spores) In Palynology principles and applications vol1Principles (ed J Jansonius amp D C McGregor) pp 383^387Salt Lake City UT American Association of StratigraphicPalynologists Foundation Publishers Press

Edwards D amp Wellman C H 2000 Embryophytes on landthe Ordovician to Lochkovian (Lower Devonian) record InPlants invade the land (ed D Edwards amp P G Gensel) (In thepress)

Edwards D Davies K L amp Axe L 1992 A vascularconducting strand in the early land plant Cooksonia Nature357 683^685

Edwards D Ducket J G amp Richardson J B 1995a Hepaticcharacters in the earliest land plants Nature 374 635^636

Edwards D Davies K L Richardson J B amp Axe L 1995bThe ultrastructure of spores of Cooksonia pertoni Palaeontology38 153^168

Edwards D Davies K L Richardson J B Wellman C Hamp Axe L 1996a Ultrastructure of Synorisporites downtonensisand Retusotriletes cf coronadus in spore masses from the Pridoliof the Welsh Borderland Palaeontology 39 783^800

Edwards D Abbott G D amp Raven J A 1996b Cuticles ofearly land plants a palaeoecophysiological evaluation InPlant cuticlesoumlan integrated functional approach (ed G Kersteins)pp1^31 Oxford BIOS Scienticentc Publishers

Edwards D Wellman C H amp Axe L 1998 The fossil recordof early land plants and interrelationships between primitiveembryophytes too little too late In Bryology for the 21st century(ed J W Bates N W Ashton amp J G Duckett) pp15^43Maney Publishing and British Bryological Society

Edwards D Wellman C H amp Axe L 1999 Tetrads insporangia and spore masses from the Upper Silurian andLower Devonian of the Welsh Borderland Bot J Linn Soc130 111^115

Fanning U Richardson J B amp Edwards D1988 Cryptic evolu-tion in an early land plant EvolTrends Plants 213^24

Fanning U Richardson J B amp Edwards D 1991 A review ofin situ spores in Silurian land plants In Pollen and spores



patterns of diversicentcation (ed S Blackmore amp S H Barnes)pp 25^47 The Systematics Association Special VolumeNo 44 Oxford Clarendon Press

Garratt M J 1978 New evidence for a Silurian (Ludlow age)for the earliest Baragwanathia poundora Alcheringa 2 217^224

Gensel P G 1980 Devonian in situ spores a survey and discus-sion Rev Palaeobot Palynol 30 101^132

Gensel P G Johnson N G amp Strother P K 1991 Early landplant debris (Hookerrsquos `waifs and straysrsquo) Palaios 5 520^547

Graham L E amp Gray J 2000 The origin morphology andecophysiology of early embryophytes neontological andpaleontological perspectives In Plants invade the land (ed DEdwards amp P G Gensel) (In the press)

Gray J 1984 Ordovician^Silurian land plants the inter-dependence of ecology and evolution Special Papers inPalaeontology 32 281^295

Gray J 1985 The microfossil record of early land plantsadvances in understanding of early terrestrialization 1970^1984 PhilTrans R Soc Lond B 309 167^195

Gray J 1988 Land plant spores and the Ordovician^Silurianboundary Bull Br Mus Nat Hist (Geol) 43 351^358

Gray J 1991 Tetrahedraletes Nodospora and the crossrsquo tetrad anaccretion of myth In Pollen and spores patterns of diversicentcation(ed S Blackmore amp S H Barnes) pp 49^87 The SystematicsAssociation Special Volume No 44 Oxford Clarendon Press

Gray J 1993 Major Paleozoic land plant evolutionary bio-events Palaeogeogr Palaeoclimatol Palaeoecol 104 153^169

Gray J amp Boucot A J 1977 Early vascular land plants proofand conjecture Lethaia 10 145^174

Gray J amp Boucot A J 1979 The Devonian land plantProtosalvinia Lethaia 12 57^63

Gray J Massa D amp Boucot A J 1982 Caradocian land plantmicrofossils from Libya Geology 10 197^201

Gray J Theron J N amp Boucot A J 1986 Age of theCedarberg Formation South Africa and early land plantevolution Geol Mag 123 445^454

Gray J Boucot A J GrahnY amp Himes G 1992 A new recordof early Silurian land plant spores from the Parana BasinParaguay (Malvinokaiexclric Realm) Geol Mag 129 741^752

Habgood K S 2000 Two cryptospore-bearing land plantsfrom the Lower Devonian (Lochkovian) of the WelshBorderland Bot J Linn Soc (In the press)

Harland W B Armstrong R L Cox A V Craig L ESmith A G amp Smith D G 1989 A geologic timescaleCambridge University Press

Hemsley A R 1994 The origin of the land plant sporophytean interpolation scenario Biol Rev 69 263^273

Hueber F M 1996 A solution to the enigma of Prototaxites ThePalaeontological Society Special Publication 8 Sixth NorthAmerican Paleontological Convention Abstracts

Hughes N F 1994 The enigma of angiosperm origins CambridgeUniversity Palaeobiology Series no 1 Cambridge UniversityPress

Johnson N G 1985 Early Silurian palynomorphs from theTuscarora Formation in central Pennsylvania and their paleo-botanical and geological signicentcance Rev Palaeobot Palynol45 307^360

Kenrick P K amp Crane P R 1997 The origin and early diversicentca-tion of land plants Washington and London SmithsonianInstitution Press

Kroken S B Graham L E amp Cook M E 1996 Occurrenceand evolutionary signicentcance of resistant cell walls in charo-phytes and bryophytes Am J Bot 83 1241^1254

Kurmann M H amp Doyle J A 1994 Ultrastructure of fossil sporesand pollen its bearing on relationships among fossil and living groupsKew The Royal Botanic Gardens

Lang W H 1937 On the plant-remains from the Downtonian ofEngland and Wales PhilTrans R Soc Lond B 227 245^291

Mishler B D Lewis L A Buchheim M A RenzagliaK S Garbary D J Delwiche C F Zechman F WKantz T S amp Chapman R L 1994 Phylogenetic relation-ships of the green algaersquo and `bryophytesrsquo Ann Missouri BotGdn 81 451^483

Poinar Jr G O Peterson E B amp Platt J L 2000 Fossillichens in the New World amber (Parmelia) Lichenologist (Inthe press)

Pratt L M Phillips T L amp Dennison J M 1978 Evidence ofnon-vascular land plants from the early Silurian (Llandoverian)of Virginia USA Rev Palaeobot Palynol 25 121^149

Richardson J B 1988 Late Ordovician and Early Siluriancryptospores and miospores from northeast Libya InSubsurface palynostratigraphy of northeast Libya (ed A El-ArnautiB Owens amp B Thusu) pp 89^109 Benghazi LibyaGaryounis University Publications

Richardson J B 1992 Origin and evolution of the earliest landplants In Major events in the history of life (ed W J Schopf )pp 95^118 Boston MA Jones and Bartlett Publishers

Richardson J B 1996a Lower and middle Palaeozoic records ofterrestrial palynomorphs In Palynology principles andapplications vol 2 Applications (ed J Jansonius amp D CMcGregor) pp 555^574 Salt Lake City UT AmericanAssociation of Stratigraphic Palynologists FoundationPublishers Press

Richardson J B 1996b Abnormal spores and possible inter-specicentc hybridization as a factor in the evolution of EarlyDevonian land plants Rev Palaeobot Palynol 93 333^340

Richardson J B amp Ioannides N 1973 Silurian palynomorphsfrom the Tanezzuft and Acacus Formations TripolitaniaNorth Africa Micropalaeontology 19 201^252

Richardson J B amp McGregor D C 1986 Silurian andDevonian spore zones of the Old Red Sandstone Continentand adjacent regions Geol Surv Can Bull 364 1^79

Richardson J B Ford J H amp Parker F 1984 Miosporescorrelation and age of some Scottish Lower Old RedSandstone sediments from the Strathmore region (Fife andAngus) J Micropalaeont 3 109^124

Rodriguez R M 1983 Palynologia de las Formaciones del Siluro-superior-Devonico inferiorde la Cordillera Cantabrica Publ UnivLeon (Espamiddotia) Oviedo Gracentcas Summa

Rogerson E C W Edwards D Davies K L amp RichardsonJ B 1993 Identicentcation of in situ spores in a Silurian Cooksoniafrom the Welsh Borderland Spec Pap Palaeont 49 17^30

Scotese C R amp McKerrow W S 1990 Revised maps andintroduction In Palaeozoic palaeogeography and biogeography (edW S McKerrow amp C R Scotese) Memoir No 12 LondonGeological Society of London

Sherwood-Pike M A amp Gray J 1985 Silurian fungal remainsprobable records of the Class Ascomycetes Lethaia 18 1^20

Steemans P 1999 Paleodiversicentcation des spores et des crypto-spores de LrsquoOrdovicien au Devonien inferieur Geobios 32341^352

Steemans P Le Herisse A amp Bozdogan N 1996 Ordovicianand Silurian cryptospores and miospores from southeasternTurkey Rev Palaeobot Palynol 93 35^76

Steemans P Higgs K amp Wellman C H 2000 Cryptosporesand trilete spores from the Llandovery NYYM-2 boreholeSaudi Arabia Special GeoArabia Publication no1 pp 92^115Bahrain Gulf PetroLink

Strother P K 1988 New species of Nematothallus from theSilurian Bloomsburg Formation of Pennsylvania J Paleontol62 967^982

Strother P K 1991 A classicentcation schema for the cryptosporesPalynology 15 219^236

Strother P K amp Traverse A 1979 Plant microfossils fromLlandoverian and Wenlockian rocks of PennsylvaniaPalynology 3 1^21



Strother P K Al-Hajri S amp Traverse A 1996 New evidencefor land plants from the lower Middle Ordovician of SaudiArabia Geology 24 55^58

Taylor T N Hass H Remy W amp Kerp H 1995 The oldestfossil lichen Nature 378 244

Taylor T N Hass H amp Kerp H 1997 A cyanolichen from theLower Devonian Rhynie Chert Am J Bot 84 992^1004

Taylor W A 1995a Spores in earliest land plants Nature 373391^392

Taylor W A 1995b Ultrastructure of Tetrahedraletes medinensis(Strother and Traverse) Wellman and Richardson from theUpper Ordovician of southern Ohio Rev Palaeobot Palynol 85183^187

Taylor W A 1996 Ultrastructure of lower Paleozoic dyads fromsouthern Ohio Rev Palaeobot Palynol 92 269^279

Taylor W A 1997 Ultrastructure of lower Paleozoic dyadsfrom southern Ohio II Dyadospora murusattenuata functionaland evolutionary considerations Rev Palaeobot Palynol 971^8

Tekbali A O amp Wood G D 1991 Silurian spores acri-tarchs and chitinozoans from the Bani Walid Borehole ofthe Ghadames Basin Northwest Libya Geol Libya 41243^1273

Vavrdova M 1984 Some plant microfossils of possible terrestrialorigin from the Ordovician of Central Bohemia VestnikUstredniho Ustavu Geologickeho 3 165^170

Wellman C H 1993a A Lower Devonian sporomorph assem-blage from the Midland Valley of Scotland Trans R SocEdinb Earth Sci 84 117^136

Wellman C H 1993b A land plant microfossil assemblage ofMid Silurian age from the Stonehaven Group Scotland JMicropalaeont 12 47^66

Wellman C H 1995 `Phytodebrisrsquo from Scottish Silurian andLower Devonian continental deposits Rev Palaeobot Palynol84 255^279

Wellman C H 1996 Cryptospores from the type area of theCaradoc Series in southern Britain Spec Pap Palaeontol 55103^136

Wellman C H 1999 Sporangia containing Scylaspora from theLower Devonian of the Welsh Borderland Palaeontology 4267^81

Wellman C H amp Richardson J B 1993 Terrestrial plantmicrofossils from Silurian inliers of the Midland Valley ofScotland Palaeontology 36 155^193

Wellman C H amp Richardson J B 1996 Sporomorph assem-blages from the `Lower Old Red Sandstonersquo of LorneScotland Spec Pap Palaeont 55 41^101

Wellman C H Edwards D amp Axe L 1998a Permanent dyadsin sporangia and spore masses from the Lower Devonian of theWelsh Borderland Bot J Linn Soc 127 117^147

Wellman C H Edwards D amp Axe L 1998b Ultrastructure oflaevigate hilate spores in sporangia and spore masses from theUpper Silurian and Lower Devonian of the Welsh BorderlandPhilTrans R Soc Lond B 353 1983^2004

Wellman C H Higgs K T amp Steemans P 2000a Sporeassemblages in the Silurian sequence in borehole HWYH-151from Saudi Arabia Special GeoArabia Publication no1pp116^133 Bahrain Gulf PetroLink

Wellman C H Habgood K Jenkins G amp Richardson J B2000b A new plant assemblage (microfossil and megafossil)from the Lower Old Red Sandstone of the Anglo-WelshBasin its implications for the palaeoecology of early terres-trial ecosystems Rev Palaeobot Palynol (In the press)

DiscussionP Kenrick (Department of Palaeontology The Natural HistoryMuseum London) Modern phylogenetic research clearlyshows that land plants are a monophyletic group This

implies a single point of origin Based on the fossil recordof dispersed spores can we pinpoint the origin of landplants to a specicentc Ordovician landmass

Is it possible to say anything about how the coloniza-tion of the various Ordovician landmasses (LaurentiaBaltica Gondwana) proceeded following the origin ofland plants

C Wellman and J Gray As noted in our paper theearliest known dispersed spores interpreted as derivedfrom embryophytes are from the Llanvirn (Mid-Ordovician) of Bohemia and Saudi Arabia By theCaradoc similar spores are known from Libya andsouthern Britain and similar Ashgill assemblages havebeen reported from South Africa LibyaTurkey BohemiaGermany southern Britain eastern North America andChina Some of these Late Ordovician assemblages arepreserved in continental deposits that theoretically shouldprovide better control on their provenance

While it is tempting to attach evolutionary signicentcanceto the early occurrences in Bohemia and Saudi Arabiathe Ordovician database outlined above is too limitedfor us to speculate on the location of a `single point oforiginrsquo for higher land plantsembryophytes if indeed a`centrst appearancersquo of dispersed spores can be assumedunder any circumstances to correspond to à single pointof originrsquo The best premise with fossils especially poorlyknown small fossils with a wide dispersal potential is toassume that `centrst appearancersquo may have little to do with`point of originrsquo

Anything approaching representative palaeogeographicreporting of dispersed embryophyte spores is centrst achievedin the Ashgill By this time however more or less identicaldispersed spore assemblages are widespread Againbecause of the limited database it is presently impossibleto comment on how the colonization of the various Ordo-vician land masses proceeded following the origin of landplantsWe suspect that colonization would have been rapiddue to the considerable dispersal potential of these sporesrelated to their small size thick protective wall and theircapability to reproduce from a single dispersal propagule(see discussion in Gray 1985)

W G Chaloner (Department of Geology Royal HollowayUniversity of London Surrey UK) Do any specimens of theOrdovician òbligate tetradsrsquo that you have shown us giveany indication of where they germinated Did the processof germination force the members of the tetrad apart (iedid they germinate through the proximal face) or do anyof the tetrads show signs of an aperture on the exposed(distal) faces of the spores

As a further but closely related question do you knowhow the tetrads of the Marchantialean liverworts (withwhich Gray has compared these obligate tetrads) germi-nate Does the site of germination in the fossil tetrads (ifany evidence for this exists) give support for a liverwortacurrennity

C Wellman and J Gray As far as we are aware none ofthe dispersed obligate tetrads from the Ordovician^Lower Devonian provide any indication of where theygerminated This however is not surprising Most if notall of these dispersed spores probably represent `failedrsquo



spores ie they were not transported into an environ-ment suitable for germination Very rarely do dispersedearly land plant spores exhibit evidence of germinationalthough certain trilete spores occasionally exhibit gapingsutures that may indicate that the gametophyte emergedthrough this suture and very occasionally germinatingspores are exquisitely preserved in exceptionallypreserved poundoras for example in the Rhynie Chert (Lyon1957 Bhutta 1973 Remy amp Hass 1996)

Gray (1985) suggested that `for spore-producing landplants the obligate spore tetrad appears to be the primi-tive reproductive unit the single spore split from thetetrad at maturity derivativersquo and that `distal dehiscenceis primitive proximal dehiscence derivativersquo She thor-oughly reviewed the literature on extant tetrad-producingbryophytes and germination in their spore tetrads and werefer you to that source for details some summarizedhere In obligate tetrads produced by living liverwortsgermination is via the distal (ie free) surface (Duthie ampGarside 1936 Proskauer 1954 Schuster 1981) Interest-ingly in a number of extant bryophyte spores that aredispersed as monads (ie following dissociation of thetetrad) dehiscence is also via the distal surface (Proskauer1961) This is even the case for some species that producetrilete spores for example spores of most species of Riccia(Duthie amp Garside 1936) the endemic Indian Stephenson-iella (Mehra amp Kachroo 1952) and other taxa

With regard to this debate there are a number ofother pertinent observations Permanent dyads andtetrads are often preserved with the distal wallcollapsed suggesting that the distal wall was thinSections of such spores concentrm that this is often thecase (Wellman et al 1998 Edwards et al 1999) Perhapsthe distal wall of these spores was thin in order to facil-itate dehiscence Interestingly Taylor (1997) sectioneddispersed permanent dyads from latest Ordovicianêarliest Silurian deposits of Ohio and noted simplebreaksregions of thinning that occurred on the distalsurface of the spores comprising the dyads He

suggested that they might represent `primitive suturesrsquoalthough we consider this interpretation dubious andsuspect that these structures may be artefactual Wherewe do have evidence for germinating spores in the fossilrecord (ie the Rhynie Chert) the situation is notalways as simple as one might expect Remy amp Hass(1996) note that the plant Aglaophyton major producedtrilete spores that in certain horizons are fairlyfrequently preserved in the process of germinating Thespores are usually oriented with the proximal surfacefacing up and the gametophyte emerging through thetrilete mark such that the spore wall splits into threeregular valves Occasionally however the spores areoriented with the distal surface facing up and the wallsin these spores often show irregular ruptures or areruptured into irregular parts

ReferencesBhutta A A 1973 On the spores (including germinating spores)

of Horneophyton (Hornea) lignieri (Kidston amp Lang) Barghoornamp Darrah (1938) PakistanJ Bot 5 45^55

Duthie A V amp Garside S 1936 Studies in South AfricanRicciaceae I Three annual species R plana Taylor R cupili-fera sp nov and R curtisii T P JamesTrans R Soc S Afr 2493^133

Lyon A G 1957 Germinating spores in the Rhynie ChertNature 180 12^19

Mehra P N amp Kachroo P 1952 Sporeling germination studiesin Marchantiales II Stephensoniella brevipedunculata Kash TheBryologist 55 59^64

Proskauer J 1954 On Sphaerocarpos stipitatus and the genusSphaerocarpos J Linn Soc Bot 55 143^157

Proskauer J 1961 On Carrpos I Phytomorphology11 359^378Remy W amp Hass H 1996 New information on gametophytes

and sporophytes of Aglaophyton major and inferences aboutpossible environmental adaptations Rev Palaeobot Palynol 90175^193

Schuster R M 1981 Paleoecology origin distribution throughtime and evolution of Hepaticae and Anthocerotae InPaleobotany paleoecology and evolution (ed K J Niklas)pp129^191 New York Praeger



there is a supercentcial line of attachment between thespores and cohesion probably results from localizedexospore links or bridges rather than large-scale fusionThey are termed fused if there is no discernible line ofattachment and cohesion probably results from fusionover most or all of the contact area Cohesion may alsoresult from enclosure within a tight-centtting envelope Thecomposition of cryptospores (wall and envelope) isunknown but their preservation in ancient depositssuggests they constitute sporopollenin or a sporopollenin-type macromolecule Since there seems little question thatthese structures are spores the term cryptospore has notaxonomic usefulness but we retain the term here fordescriptive convenience

Reports of Ordovician^Early Silurian dispersed sporesare relatively few but are stratigraphically and geogra-phically widespread and indicate that cryptosporeassemblages are remarkably constant in composition

(both temporally and spatially) throughout this interval(Gray 1985 1988 1991 Richardson 1996a Strother et al1996 Wellman 1996 Steemans 1999) These data suggestthat the vegetation was widespread but of limited diver-sity with little evolutionary change (at least in the spores)during an interval some 40 Myr in duration

Over much of the globe a major change in the natureof spore assemblages is reported in the late Llandovery(late Early Silurian) (Gray 1985 1991 Gray et al 1986Richardson 1988 1996a Burgess 1991 Wellman 1996Steemans 1999) (centgure 1) While naked monads dyadsand tetrads continued to dominate spore assemblages theabundance of envelope-enclosed forms is seriouslyreduced and in places they appear to virtually disappearAt the same time hilate monads and trilete spores centrstbecame widespread and occur in relative abundanceHilate monads comprise single spores with a circularcontact area on their proximal surface and clearly



Ordovician

envelope-enclosedtetrads

envelope-encloseddyads

envelope-enclosedmonads

naked tetrads

naked dyads

naked monads

hilate monads

trilete sporesL

lanvirn

Llandeilo

Caradoc

469

464

443

439

430

424

411409

396

Ashgill

Llandovery

Wenlock

Ludlow

Lochkovian

Pragian

epochstage

Pridoli

Silurian Devonian period

Figure 1 Stratigraphical rangechart for early land plant sporemorphotypes Time-scale fromHarland et al (1989) Regardingannotation adjacent to the strati-graphic column the small `Cooksoniarsquoindicates the age of the earliestunequivocal land plant megafossilsthe asterisks indicate the age ofimportant plant mesofossilassemblages (ie LudfordLane-Pridoli North Brown CleeHill-Lochkovian) The dashedtime-line indicates the position ofthe major change in the nature ofdispersed spore assemblages in theLlandovery The width of range barsprovides a very basic indication ofrelative abundance to illustrate grosschanges in abundance through timeNote that range bar width has noquantitative implications The taperand question marks at the base ofrange bars extending into theLlanvirn indicate the age of theearliest known spore assemblagesand do not necessarily coincide withtheir centrst appearance nor necessarilywith changes in abundance



(a) (b) (c) (d ) (e) ( f )

(g) (h) (i)( j)

(k)

(l) (m) (n) (o)

( p)

(q) (r) (s)

(t) (u) (v) (w)








(a) (b)(c) (d )

(e)

( f )

(g) (h) (i)

( j) (k) (l) (m)





































































Ordovician Silurian


non-banded tubes

banded tubes


stomata

tracheids

Llanvirn

Llandeilo

Caradoc

469

464

443

439

430

424

411409

396

Ashgill

Llandovery

Wenlock

Ludlow

Lochkovian

Pragian

epochstage

Pridoli

Devonian period









(a)

(c) (d ) (e) ( f )

(g) (h) (i)

(b)



















4 CONCLUSIONS




REFERENCES
























































































































(a) (b) (c) (d ) (e) ( f )

(g) (h) (i)( j)

(k)

(l) (m) (n) (o)

( p)

(q) (r) (s)

(t) (u) (v) (w)








(a) (b)(c) (d )

(e)

( f )

(g) (h) (i)

( j) (k) (l) (m)





































































Ordovician Silurian


non-banded tubes

banded tubes


stomata

tracheids

Llanvirn

Llandeilo

Caradoc

469

464

443

439

430

424

411409

396

Ashgill

Llandovery

Wenlock

Ludlow

Lochkovian

Pragian

epochstage

Pridoli

Devonian period









(a)

(c) (d ) (e) ( f )

(g) (h) (i)

(b)



















4 CONCLUSIONS




REFERENCES




























































































































(a) (b)(c) (d )

(e)

( f )

(g) (h) (i)

( j) (k) (l) (m)





































































Ordovician Silurian


non-banded tubes

banded tubes


stomata

tracheids

Llanvirn

Llandeilo

Caradoc

469

464

443

439

430

424

411409

396

Ashgill

Llandovery

Wenlock

Ludlow

Lochkovian

Pragian

epochstage

Pridoli

Devonian period









(a)

(c) (d ) (e) ( f )

(g) (h) (i)

(b)



















4 CONCLUSIONS




REFERENCES

























































































































































































Ordovician Silurian


non-banded tubes

banded tubes


stomata

tracheids

Llanvirn

Llandeilo

Caradoc

469

464

443

439

430

424

411409

396

Ashgill

Llandovery

Wenlock

Ludlow

Lochkovian

Pragian

epochstage

Pridoli

Devonian period









(a)

(c) (d ) (e) ( f )

(g) (h) (i)

(b)



















4 CONCLUSIONS




REFERENCES





























































































































(a)

(c) (d ) (e) ( f )

(g) (h) (i)

(b)



















4 CONCLUSIONS




REFERENCES







































































































































4 CONCLUSIONS




REFERENCES






















































































































the microfossil record of early land plants - david moore...fossil record. the fossil record of...

Documents