The Origin and Early Evolution of Plants on Land PDF

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/242879569 The origin and early evolution of plants on Land Article in Nature · September 1997 DOI: 10.1038/37918 CITATIONS...

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/242879569 The origin and early evolution of plants on Land Article in Nature · September 1997 DOI: 10.1038/37918 CITATIONS READS 1,471 31,909 2 authors: Paul Kenrick Peter R. Crane Natural History Museum, London Oak Spring Garden Foundation 190 PUBLICATIONS 8,351 CITATIONS 315 PUBLICATIONS 21,836 CITATIONS SEE PROFILE SEE PROFILE All content following this page was uploaded by Paul Kenrick on 22 December 2014. The user has requested enhancement of the downloaded file.  review article The origin and early evolution of plants on land Paul Kenrick & Peter R. Crane........................................................................................................................................................................................................................................................ The origin and early evolution of land plants in the mid-Palaeozoic era, between about 480 and 360 million years ago, was an important event in the history of life, with far-reaching consequences for the evolution of terrestrial organisms and global environments. A recent surge of interest, catalysed by palaeobotanical discoveries and advances in the systematics of living plants, provides a revised perspective on the evolution of early land plants and suggests new directions for future research. The origin and early diversification of land plants marks an interval Eoembryophytic (mid-Ordovician [early Llanvirn: !476 Myr] to of unparalleled innovation in the history of plant life. From a simple Early Silurian [late Llandovery: !432 Myr])3. Spore tetrads (com- plant body consisting of only a few cells, land plants (liverworts, prising four membrane-bound spores; Fig. 2d) appear over a broad hornworts, mosses and vascular plants) evolved an elaborate two- geographic area in the mid-Ordovician and provide the first good phase life cycle and an extraordinary array of complex organs and evidence of land plants3,26,29. The combination of a decay-resistant tissue systems. Specialized sexual organs (gametangia), stems with wall (implying the presence of sporopollenin) and tetrahedral an intricate fluid transport mechanism (vascular tissue), structural configuration (implying haploid meiotic products) is diagnostic tissues (such as wood), epidermal structures for respiratory gas of land plants. The precise relationships of the spore producers exchange (stomates), leaves and roots of various kinds, diverse within land plants are controversial, but evidence of tetrads and spore-bearing organs (sporangia), seeds and the tree habit had all other spore types (such as dyads) in Late Silurian and Devonian evolved by the end of the Devonian period. These and other megafossils16,30, as well as data on spore wall ultrastructure25 and the innovations led to the initial assembly of plant-dominated terres- structure of fossil cuticles31, support previous suggestions of a land trial ecosystems, and had a great effect on the global environment. flora of liverwort-like plants (Fig. 1c)3. Some early spores and Early ideas on the origin of land plants were based on living cuticles may also represent extinct transitional lineages between groups, but since the discovery of exceptionally well-preserved fossil charophycean algae (Fig. 1a, b) and liverworts (Box 1), but precise plants in the Early Devonian Rhynie Chert, research has focused understanding of their affinities is hindered by the dearth of almost exclusively on the fossil record of vascular plants1,2. During associated megafossils. the 1970s, syntheses of palaeobotanical and stratigraphic data Eotracheophytic (Early Silurian [latest Llandovery: !432 Myr] to emphasized the Late Silurian and Devonian periods as the critical Early Devonian [mid Lochkovian: !402 Myr])3. The early Silurian interval during which the initial diversification of vascular plants (latest Llandovery) marks the beginning of a decline in diversity of occurred1,2, and identified a group of simple fossils (rhyniophytes, tetrads and a rise to dominance of individually dispersed, simple such as Cooksonia and Rhynia) as the likely ancestral forms2. They spores, which are found in several basal land plant groups (such as also supported earlier hypotheses of two main lines of evolution: hornworts, some mosses, and early vascular plants)3. Although one comprising clubmosses (Fig. 1f) and extinct relatives, the other tetrads remain dominant in some Early Devonian localities from including all other living vascular plants (ferns, horsetails and seed northwestern Europe32, the elaboration of simple spores and turn- plants; Fig. 1g–j) and related fossils1,2. During the past two decades, over of spore ‘species’3 provide evidence of increasing land plant the discovery of fossil spores from as far back as the mid-Ordovician diversity and vegetational change. Although spores have been period3, improved knowledge of living green algae4,5, renewed observed in Silurian megafossils, the affinities of most dispersed interest in the phylogenetic position of other relevant groups such forms remain unknown, indicating that substantial land plant as mosses and liverworts5, and advances in molecular systema- diversity is currently undetected in the megafossil record30. tics5–14, together with unexpected new data on the structure and The earliest unequivocal land plant megafossils are from the mid- biology of Silurian and Devonian fossils15–25, have provided a Silurian of northern Europe33, and lowermost Upper Silurian of broader perspective on the origin of a land flora26. These new data Bolivia34 and Australia35, and the uppermost Silurian of north- indicate that the early diversification of land plants substantially western China36. Early assemblages include clubmosses (such as pre-dates the Late Silurian to Early Devonian, and suggest that the Baragwanathia) and related early fossils (such as zosterophylls, main basal lineages originated over a period of more than 100 some species of Cooksonia), and various other plants of uncertain million years (Myr). affinity (such as Salopella and Hedeia; Fig. 3). These data document an influx into land plant communities of diverse but generally small Patterns in the early fossil record (usually less than 10 cm tall) organisms related to vascular plants Evidence on the origin and diversification of land plants has come (Fig. 3). Exceptions to the generally small size include the clubmoss mainly from dispersed spores and megafossils. Gray recognized Baragwanathia37 and the large and much-branched Pinnatiramosus three new plant-based epochs (Eoembryophytic, Eotracheophytic from the early Silurian of China38. The habit and branching of and Eutracheophytic) spanning the origin and early establishment Pinnatiramosus is similar to that of green algae in the Caulerpales, of land plants: each is characterized by the relative abundance of but the presence of tracheid-like tubes is inconsistent with this spore types and megafossils3. This synthesis highlights diversifica- interpretation39. Additional details, including conclusive data on tion and floral change in the Ordovician and Silurian3,27,28, and reproductive structures, are needed to clarify the relationships of emphasizes a major discrepancy between evidence from spores and this enigmatic plant. megafossils: unequivocal land plant megafossils are first recognized Data from northern Europe, Siberia, Podolia (southwestern in the fossil record roughly 50 Myr after the appearance of land plant Ukraine), Libya, Vietnam, Bolivia, Australia and Xinjiang and spores. Yunnan (China) document increasing land plant diversity into Nature © Macmillan Publishers Ltd 1997 NATURE | VOL 389 | 4 SEPTEMBER 1997 33 review article the base of the Devonian33–36,40. These fossils, together with the among land plants, including the capacity to produce sporo- relative chronology implicit in current hypotheses of relationship, pollenin, cutin, phenolic compounds and the glycolate oxidase imply a minimum mid-Silurian origin for several important vas- pathway4,48. However, the absence of well-developed sporophytes, cular plant groups (Box 1; Fig. 4). gametophytes with sexual organs of land plant type, cuticle and Eutracheophytic (Early Devonian [late Lochkovian: !398 Myr] to non-motile, airborne, sporopollenin-walled spores suggests that mid-Permian [!256 Myr])3. In the Early Devonian (late Lochko- these innovations evolved during the transition to the land4,18. In vian) the diversity of spores and megafossils increased contrast to animal groups, the entire multicellular diploid phase of dramatically29,40–42. Early assemblages include the classic floras the plant life cycle probably evolved in a terrestrial setting. from the Rhynie Chert20,43,44, the Gaspé Peninsula of eastern The transition from an aqueous to a gaseous medium exposed Canada43,44, New York State43,44, the Rhine Valley of Germany45, plants to new physical conditions that resulted in key physiological Belgium46, Australia35 and Yunnan Province (China)33, which docu- and structural changes. Important metabolic pathways leading to ment a substantial increase in vascular plant diversity, including the lignins, flavenoids, cutins and plant hormones in vascular plants appearance and early diversification of many important living probably arose from pre-existing elements of primary metabolism groups. in charophycean algae and bryophytes4. Although the evolution of these pathways is poorly understood, possible phenolic precursors Building a land plant have been detected in charophycean algae4,31, and elements of auxin Phylogenetic studies favour a single origin of land plants from metabolism have been recognized in mosses and hornworts49. charophycean green algae (Box 1). Based on the ecology of living Phylogenetic studies predict that early land plants were small and species, a freshwater origin of land plants seems likely, but direct morphologically simple, and this hypothesis is borne out by fossil evidence from the fossil record is inconclusive as mid-Palaeozoic evidence (Fig. 3). Early fossils bear a strong resemblance to the charophytes are found in both freshwater and, more commonly, simple spore-producing phase of living mosses and liverworts (Fig. marine facies47. Living charophycean algae (Fig. 1a, b) possess 1d, e and 5)16,26,50,51, and these similarities extend to the anatomical several biosynthetic attributes that are expressed more fully details of the spore-bearing organs and the vascular system19. The Box 1 Relationships among land plants Spermatopsids (seed plants) Filicopsids (ferns) Land plants (embryophytes) are most closely related to the Charophyceae, a Psilotaceae (whisk ferns) small group of predominantly freshwater green algae, within which either Eutracheophytes Equisetopsids (horsetails) Euphyllophytes (All living vascular plants) and many other extinct taxa Coleochaetales (!15 living species; Fig.1a) or Charales (!400 living species; Fig. 1b), or a group containing both, is sister group to land plants4,5,10,12,74. Tracheophytes Psilophyton dawsonii † (Vascular plants) 4,5,26,75 Lycopsids (clubmosses) Land-plant monophyly is supported by comparative morphology and Lycophytes Zosterophylls † gene sequences (18S rRNA, mitochondrial DNA: cox III)12,14. Relationships Cooksonia pertonii † among the major basal living groups are uncertain4,5,26,76,77, but the best- Rhynia gwynne-vaughanii † Rhyniopsids † supported hypothesis resolves liverworts (Fig. 1c) as basal and either Stockmansella langii † mosses (Fig. 1e) or hornworts (Fig. 1d) as the living sister group to vascular Aglaophyton major † Embryophytes ‘Protracheophytes’ † plants (tracheophytes)4,5,13,14,26,75. Less parsimonious hypotheses recognize Horneophytopsids † (Land plants) bryophyte monophyly and either a sister-group relationship with vascular Bryopsida (mosses) plants26 or an origin from within basal vascular plants14,76,78. Anthocerotopsida (hornworts) ‘Bryophytes’ Among vascular plants, living ferns (Fig. 1g), horsetails (Fig. 1i) and seed Marchantiopsida (liverworts) Coleochaetales plants (Fig. 1j) (euphyllophytes) are the sister group to clubmosses 13,14,26,75,79 Charales (Fig.1f). Euphyllophyte monophyly is strongly supported by compara- Chaetosphaeridium tive morphology26 and a unique 30-kb inversion in the chloroplast genome8, as 13 14 well as sequence data from 18S rRNA and mitochondrial DNA (cox III). These data also provide evidence that the enigmatic Psilotaceae (Fig. 1h) (a group of simple plants once thought to be living relicts of the earliest vascular plants) are more closely related to the fern–seed plant lineage than to basal vascular plants (clubmosses or the extinct rhyniophytes). Within vascular extinct Cooksonia and similar early fossils (such as Tortilicaulis, Uskiella, plants, molecular and morphological assessments of phylogeny at the level Caia15–17,81) (Fig. 3) suggests that simple early land plants (once grouped as of orders and below give similar results11, but at deeper levels (for example, rhyniophytes1) are an unnatural assemblage26. Some Cooksonia species may the divergence of major groups of ferns, horsetails and seed plants) phylo- be among the precursors to vascular plants (protracheophytes), whereas genetic resolution is poor. These difficulties highlight the problems of others are vascular plants apparently allied to the clubmoss lineage26. 7,80 approaches based solely on living species. Combined analyses of mole- Clubmosses emerge from a poorly resolved grade of extinct cular sequences from multiple loci, and large-scale structural characteristics Zosterophyllum-like plants (Fig. 4), although most zosterophylls form a of the genome (such as introns and inversions), may be more useful in monophyletic group26. Within clubmosses, early leafy herbaceous fossils assessing deep phylogenetic patterns. such as Baragwanathia and Asteroxylon are basal26,82, and living Lycopodia- Megafossils fill some of the substantial morphological ‘gaps’ among living ceae are resolved as sister group to a calde that comprises the extinct groups. Phylogenetic analyses19,26 interpolate two Early Devonian Rhynie herbaceous Protolepidodendrales, living Selaginella and the predominantly Chert plants, Aglaophyton and Horneophyton, between bryophytes and arborescent carboniferous lepidodendrids, including living Isoetes26,82 (Fig. 4). basal vascular plants as they possess some features unique to vascular Euphyliophytes make up more than 99% of living vascular plants and plants (a branched, nutritionally independent sporophyte) but also retain exhibit much greater diversity than lycophytes. Relationships among basal bryophyte-like characteristics (terminal sporangia, columella in euphyllophytes are still poorly understood26. Further progress requires a Horneophyton, and the absence of leaves, roots and tracheids with well- better understanding of the relationships of several fossil groups of uncertain defined thickenings). The discovery of previously unrecognized diversity in status (such as Trimerophytina, Cladoxylales and Zygopteridales)26,79. Nature © Macmillan Publishers Ltd 1997 34 NATURE | VOL 389 | 4 SEPTEMBER 1997  review article Figure 1 Morphological diversity among basal living land plants and potential sule); magnification × 4.5 (photograph courtesy of W. Burger). f, Huperzia land-plant sister groups. a, Coleochaete orbicularis (Charophyceae) gameto- (clubmoss) sporophyte with leaves showing sessile yellow sporangia; magnifi- phyte; magnification × 75 (photograph courtesy of L. E. Graham). b, Chara cation × 0.8. g, Dicranopteris (fern) sporophyte showing leaves with circinate (Charophyoceae) gametophyte; magnification × 1.5 (photograph courtesy of M. vernation; magnification × 0.08. h, Psilotum (whisk fern) sporophyte with reduced Feist). c, Riccia (liverwort) gametophyte showing sporangia (black) embedded in leaves and spherical synangia (three fused sporangia); magnification × 0.4. i, the thallus; magnification × 5 (photograph courtesy of A. N. Drinnan). d, Equisetum (horsetail) sporophyte with whorled branches, reduced leaves, and a Anthoceros (hornwort) gametophyte showing unbranched sporophytes; magni- terminal cone; magnification × 0.4. j, Cycas (seed plant) sporophyte showing fication × 2.5 (photograph courtesy of A. N. Drinnan). e, Mnium (moss) leaves and terminal cone with seeds; magnification × 0.05 (photograph courtesy gametophyte showing unbranched sporophytes with terminal sporangia (cap- of W. Burger). Figure 2 a, Longitudinal section of part of a silicified early fossil gametophyte are from the Remy Collection (slides 200, 90 and 330), Abteilung Paläobotanik, (Kidstonophyton discoides from the Rhynie Chert). Antheridia (male sexual Westfälische Wilhelms-Universität, Münster, Germany (photographs courtesy of organs) are located on the upper surface of the branch; magnification × 3.4. b, H. Hass and H. Kerp). d, Scanning electron micrograph of Tetrahedraletes Longitudinal section of antheridium of Lyonophyton rhyniensis from the Rhynie medinensis showing a spore tetrad of possible liverwort affinity from the Late Chert; magnification × 40. c, Longitudinal section of archegonium (female sexual Ordovician (photograph courtesy of W. A. Taylor); magnification × 670. organ) of Langiophyton mackiei from the Rhynie Chert; magnification × 80. a–c Figure 3 Sporophyte diversity in Early Devonian rhyniophyte fossils. a, Cooksonia nification × 30. d, Transverse section of sporangium showing thick wall and pertonii apiculispora: sporophyte (incomplete proximally) with terminal central spore mass; magnification × 70. e, Details of epidermis at rim of spor- sporangium15; magnification × 15. b, Tortilicaulis offaeus: sporophyte (incomplete angium; magnification × 45. f, Stomate with two reniform guard cells (stippled); proximally) with terminal sporangium81; magnification × 40. c. Tortilicaulis offaeus: magnification × 120. sporophyte (incomplete proximally) with terminal bifurcating sporangium81; mag- Nature © Macmillan Publishers Ltd 1997 NATURE | VOL 389 | 4 SEPTEMBER 1997 35 review article fossil record also documents significant differences from living Box 2 Early evolution of the land plant life cycle groups, particularly in life cycles and the early evolution of the Land-plant life cycles are characterized by alternating multicellular sexual phase (Box 2). sexual (haploid gametophyte, n) and asexual phases (diploid sporo- In common with some animal groups, internalization of vital phyte, 2n). Phylogenetic studies indicate that land plants inherited a functions and organs (such as gas exchange surfaces and sexual multicellular gametophyte from their algal ancestors but that the organs), combined with the development of impermeable exterior sporophyte evolved during the transition to the land. Most megafossils surfaces, seem to have been primary responses to life on land. are sporophytes, and until recently there was no direct early fossil Together, these changes resulted in more highly differentiated plants evidence for the gametophyte phase. Recent discoveries of gameto- with stomates, multicellular sexual and spore-bearing organs, phytes in the Rhynie Chert (Early Devonian, 380–408 Myr) have shed water-conducting and other tissue systems52–54. Morphological new light on the evolution of land–plant life cycles18,20. differentiation occurred in both phases of the life cycle (gameto- Early gametophytes (a in figure) are more complex than in living phyte and sporophyte), but there was subsequently a dramatic plants and have branched stems bearing sexual organs on terminal cup- reduction in the gametophyte and a great increase in sporophyte or shield-shaped structures (Fig. 2a). Archegonia (female gametangia) complexity among vascular plants (Box 2). Apical growth and are flask-shaped with a neck canal and egg chamber, and are sunken as branching coupled with delayed initiation of spore-bearing organs in hornworts and most vascular plants (Fig. 2c). Antheridia (male were important innovations of vascular plants that led to a more gametangia) are roughly spherical, sessile or with a poorly-defined complex architectural framework on which subsequent morpholo- stalk, and superficial (Fig. 2b). Gametophytes are very similar to gical diversification was based. The fossil record clearly shows that associated sporophytes, and shared anatomical features (water- many vascular-plant organs can be interpreted in terms of mod- conducting tissues, epidermal patterns, and stomates) have been used ification (especially duplication and sterilization) of basic structural to link corresponding elements of the life cycle18,20. Our provisional units such as the spore-bearing tissues and the stem26,54. In ferns and reconstruction of the life cycle of an early vascular plant is based on seed plants, much morphological diversity is clearly attributable to information from anatomically preserved plants and contemporaneous modifications of branching systems into a variety of leaf-like organs, compression fossils. whereas the relatively conservative clubmoss bauplan has a dearth of The similarities between gametophyte and sporophyte in early fossil organ systems that can be interpreted as modified branches. In both vascular plants contrast strongly with the marked dissimilarities typical lineages, however, meristem dormancy and abortion were early of living land plants (b in figure). The phylogenetic position of fossils innovations, providing evidence of hormonal control and substan- suggests that, after the development of a simple, unbranched, ‘parasitic’ tial phenotypic flexibility21,26 sporophyte among early land colonizers at the bryophyte grade (such as mosses) there was elaboration of both gametophyte and sporophyte in Early terrestrial ecosystems vascular plants. The implications for interpreting life cycles in living The advent of land plants had important consequences for energy vascular plants18,26 are shown. The small, simple, often subterranean and nutrient fluxes among terrestrial and freshwater ecosystems29,55 and saprophytic gametophytes of living clubmosses (such as Lyco- and hence for the evolution of animal groups that live in these podiaceae) and ferns (such as Psiloataceae, Stromatopteridaceae, habitats. The vegetational changes of the Silurian and Devonian also Ophioglossaceae) result from morphological loss. Phylogenetic evi- had a major impact on the atmosphere and other aspects of the dence indicates that gametophyte reduction was independent in club- global environment. The evolution of roots is thought to have been mosses and the fern–seed plant lineage. These data provide a new an important factor in the reduction of atmospheric CO2 concen- interpretation of the gametophyte morphology of living clubmosses trations through increased weathering of Ca–Mg silicate minerals (Lycopodiaceae)18. brought about by mechanical disruption and soil acidification 56,57. Accelerated weathering has also been linked to the formation of Devonian and Early Carboniferous marine black shales58, but this requires further investigation in view of similar deposits earlier and later in the geological record. Root-like impressions have been recognized in Late Silurian palaeosols59, but the earliest unequivocal evidence comes from Early Devonian vascular plants26, which have modified prostrate stems bearing rhizoids resembling those of living bryophytes. More substantial roots capable of anchoring large trees evolved independently in several groups during the Middle to Late Devonian. A further series of innovations in vascular plants, including the biosynthesis of lignin and the origin of lateral meristems (cam- bium), were critical to the development of large plants, and these developments may have been stimulated by competition for light. Trees evolved independently in several major groups, resulting in stratified forest communities by the end of the Middle Devonian and the production of large amounts of highly decay-resistant organic material (in the form of lignified wood). The early evolution of lignin-decomposing fungi (some Ascomycetes, and Basidio- mycetes) is still poorly understood24, but these groups would have been essential for recycling much of the organic carbon. The earliest land plants probably encountered terrestrial ecosys- tems that had been occupied by bacteria and protists60,61, algae4, lichens23,62 and fungi24 since the Late Proterozoic. A variety of enigmatic plants (such as Protosalvinia44,63) were also present, and some of the largest elements (Prototaxites ‘trunks’ !69 cm in diameter) may have been fungi64. Such organisms, or perhaps some rhyniophytes16, may be the source of the microscopic tubular Nature © Macmillan Publishers Ltd 1997 36 NATURE | VOL 389 | 4 SEPTEMBER 1997  review article Figure 4 Simplified phylogenetic tree showing the minimum stratigraphic ranges remain to be confirmed44. Note that megafossil evidence for vascular plants of selected groups based on megafossils (thick bars) and their minimum implied precedes megafossil evidence of bryophytes and charophycean algae. Confir- range extensions (thin lines). Also illustrated alongside time scale are minimum mation that the Early Devonian Sporogonites is a plant at the bryophyte grade age estimates for the appearance of certain important land-plant features (from could help to reduce this discrepancy. Tre, Tremadoc; Arg, Arenig; Lln, Llanvirn; the bottom: spore tetrads, cuticles, single trilete spores, megafossils and Llo, Llandeilo; Crd, Caradoc; Ash, Ashgill; Lly, Llandovery; Wen, Wenlock; Lud, stomates). The first unequivocal record of charophycean algae is based on Ludlow; Pri, Pridoli; Lok, Lochkovian (Gedinnian); Prg, Pragian (Siegenian); Ems, calcified charalean oogonia (female sexual organs) from the Late Silurian (Pridoli, Emsian; Eif, Eifelian; Giv, Givetian; Frs, Frasnian; Fam, Famennian; Tou, Tournai- !410 Myr)83 and distinctive gametophytes from the Early Devonian Rhynie Chert44. sian; Vis, Visean; Spk, Serpukhovian; Bsh, Bashkirian; Mos, Moscovian; Kas, Proposed similarities between living Coleochaete and Early Devonian Parka Kasimovian; Gze, Gzelian. Figure 5 Diversity of water-conducting cells (tracheids) in early land plants (median longitudinal section through cells, basal and proximal end walls not shown; cells are !20–40 "m diameter). a, Top, bryophyte hydroid; bottom, details of hydroid wall showing distribution of plasmodesmata-derived micropores (10– 50 nm diameter; stipple)84. b, Top, S-type tracheid (fossil) of Rhyniopsida; bottom, details of S-type cell wall showing distribution of plasmodesmata-derived micro- pores (stipple) and ‘spongy’ interior to thickenings19. c, Top, G-type tracheid (fossil) of basal extinct eutracheophytes, which closely resemble the tracheids of some living vascular plants; bottom, details of G-type cell wall showing pores distributed between thickenings19. d, Top, scalariform pitted P-type tracheid (fossil) typical of trimerophyte grade plants (euphyllophytes); bottom, details of P-type cell wall showing pit chambers and sheet with pores that extends over pit apertures26. Nature © Macmillan Publishers Ltd 1997 NATURE | VOL 389 | 4 SEPTEMBER 1997 37 review article fragments commonly extracted from Silurian and Early Devonian should be expected to confirm the underlying unity and relative sediments28. These tubes are often associated with cellular cuticular simplicity of developmental processes in land plants. ! fragments (Nematothallus and Cosmochlaina) that may represent fragmented cuticular material from bryophyte-like plants31. The Paul Kenrick is at the Swedish Museum of Natural History, Box 50007, S-104 05, discovery of fungal arbusculae in Early Devonian megafossils22 Stockholm, Sweden; Peter R. Crane is at The Field Museum, Roosevelt Road at confirms earlier suggestions that endomycorrhizal associations Lake Shore Drive, Chicago, Illinois 60605, and the Department of the Geophysical Sciences, University of Chicago, USA. were an important innovation in the colonization of the land65. In contrast to megascopic plants, which appear to have colonized 1. Banks, H. P. Reclassification of Psilophyta. Taxon 24, 401–413 (1975). 2. Chaloner, W. G. & Sheerin, A. in The Devonian System (eds House, M. R., Scrutton, C. T. & Bassett, the land only once, many animal groups made the transition to M. G.) 145–161 (The Palaeontological Association, London, 1979). terrestrial existence independently and overcame the problems of 3. Gray, J. Major Paleozoic land plant evolutionary bio-events. Palaeogeog. Palaeoclimatol. Palaeocol. 104, 153–169 (1993). water relations in different ways52,66,67. Early evidence for terrestrial 4. Graham, L. E. Origin of Land Plants (Wiley, New York, 1993). animals is sparse29,67–69, but by the Early Devonian exquisitely 5. Mishler, B. D. et al. Phylogenetic relationships of the ‘‘green algae’’ and ‘‘bryophytes’’. Ann. MO Bot. preserved arthropod faunas are known from several localities in Gard. 81, 451–483 (1994). 6. Manhart, J. R. & Palmer, J. G. The gain of two chloroplast tRNA introns marks the green algal North America, Germany and the United Kingdom29,66,67. These ancestors of land plants. Nature 345, 268–270 (1990). faunas document the appearance of diverse arthropod communities 7. Manhart, J. R. Phylogenetic analysis of green plant rbcL sequences. Mol. Phylogenet. Evol. 3, 114–127 (1994). including centipedes, millipedes, trigonotarbids and their living 8. Raubeson, L. A. & Jansen, R. K. Chloroplast DNA evidence on the ancient evolutionary split in relatives spiders, pseudoscorpians, mites (orbatids and endeostig- vascular land plants. Science 255, 1697–1699 (1992). matids), arthropleurids (extinct arthropods), archaeognathans 9. Chapman, R. L. & Buchheim, M. A. Ribosomal RNA gene sequences: analysis and significance in the phylogeny and taxonomy of green algae. Crit. Rev. Plant Sci. 10, 343–368 (1991). (primitive wingless insects), collembolans and possibly bristletails. 10. McCourt, R. M., Karol, K. G., Guerlesquin, M. & Feist, M. Phylogeny of extant genera in the family Available evidence indicates that these animals were mainly pre- Characeae (Charales, Charophyceae) based on rbcL sequences and morphology. Am. J. Bot. 83, 125– 131 (1996). dators and detritivores and, until the appearance of vertebrate 11. Pryer, K. M., Smith, A. R. & Skog, J. E. Phylogenetic relationships of extant ferns based on evidence herbivores in the latest Palaeozoic, most energy flow into animal from morphology and rbcL sequences. Am. Fern J. 85, 205–282 (1995). components of early terrestrial ecosystems was probably through 12. Kranz, H. D. et al. The origin of land plants: phylogenetic relationships among charophytes, bryophytes, and vascular plants inferred from complete small-subunit ribosomal RNA gene the decomposer pathway rather than direct herbivory29. Indirect sequences. J. Mol. Evol. 41, 74–84 (1995). evidence for herbivory comes from wound responses in the tissues 13. Kranz, H. D. & Huss, V. A. R. Molecular evolution of pteridophytes and their relationships to seed plants: evidence from complete 18S rRNA gene sequences. Plant Syst. Evol. 202, 1–11 (1996). of some fossil plants70,71, and perhaps also from fossil faecal pellets 14. Hiesel, R., von Haeseler, A. & Brennicke, A. Plant mitochondrial nucleic acid sequences as a tool for containing abundant spores70,72. phylogenetic analysis. Proc. Natl Acad. Sci. USA 91, 634–638 (1994). 15. Edwards, D., Davies, K. L. & Axe, L. A vascular conducting strand in the early land plant Cooksonia. Nature 357, 683–685 (1992). Future directions 16. Edwards, D., Duckett, J. G. & Richardson, J. B. Hepatic characters in the earliest land plants. Nature The fossil record of spores, combined with phylogenetic studies, 374, 635–636 (1995). 17. Fanning, U., Edwards, D. & Richardson, J. B. A diverse assemblage of early land plants from the Lower indicates that groups related to living bryophytes were early colo- Devonian of the Welsh Borderland. Bot. J. Linn. Soc. 109, 161–188 (1992). nisers of the land, and suggests that several major lineages of 18. Kenrick, P. Alternation of generations in land plants: new phylogenetic and morphological evidence. vascular plant had already evolved by the mid-Silurian. Megafossils Biol. Rev. 69, 293–330 (1994). 19. Kenrick, P. & Crane, P. R. Water-conducting cells in early fossil land plants: implications for the early of land plants, however, appear much later, and in these assemblages evolution of tracheophytes. Bot. Gaz. 152, 335–356 (1991). there is a conspicuous bias toward the recognition and perhaps 20. Remy, W. Gensel, P. G. & Hass, H. The gametophyte generation of some early Devonian land plants. Int. J. Plant Sci. 154, 35–58 (1993). representation of vascular plants. The most important source of 21. Remy, W. & Hass, H. New information on gametophytes and sporophytes of Aglaophyton major and data on early megafossils has been the northern European inferences about possible environmental adaptations. Rev. Palaeobot. Palynol. 90, 175–194 (1996). (Laurussian) region, but the appearance of megafossils in this 22. Remy, W., Taylor, T. N., Hass, H. & Kerp, H. Four hundred-million-year-old vesicular arbuscular mycorrhizae. Proc. Nat Acad. Sci. USA 91, 11841–11843 (1994). area coincides with facies changes driven by a widespread marine 23. Stein, W. E., Harmon, G. D. & Hueber, F. M. in International Workshop on the Biology and Evolutionary regression28,73, and all Silurian land-plant megafossils are from Implications on Early Devonian Plants (Westfälische Wilhelms-Universität Münster, Germany, 1994). 24. Taylor, T. N. & Osborne, J. M. The Importance of fungi in shaping the paleoecosystem. Rev. Palaeobot. marine sediments33. It seems likely that the onset of continental Palynol. 90, 249–262 (1996). conditions in the Devonian of northern Europe allowed megafossils 25. Taylor, W. A. Ultrastructure of lower Paleozoic dyads from southern Ohio. Rev. Palaeobot. Palynol. 92, to be preserved at a time when vascular plants were well established 269–280 (1996). 26. Kenrick, P. & Crane, P. R. The Origin and Early Diversification of Land Plants: A Cladistic Study but still diversifying. The rapid appearance of vascular plants in this (Smithsonian Institution Press, Washington DC, 1997). region40–42 owes as much to changing geological conditions as to 27. Gray, J. The microfossil record of early land plants: advances in understanding of early terrestrializa- tion, 1970–1984 Phil. Trans. R. Soc. Lond. B 309, 167–195 (1985). rapid biological diversification27,28. Intensified sampling in areas 28. Gray, J. & Boucot, A. J. Early vascular land plants: proof and conjecture. Lethaia 10, 145–174 (1977). that are remote from these regional events is therefore a high 29. DiMichele, W. A. et al. in Terrestrial Ecosystems Through Time: Evolutionary Paleoecology of Terrestrial priority. Plants and Animals (ed. Behrensmeyer, A. K.) 205–325 (Univ. Chicago Press, 1992). 30. Fanning, U., Richardson, J. B. & Edwards, D. in Pollen and Spores (eds Blackmore, S. & Barnes, S. H.) Palaeobotanical evidence shows that the major groups of living 25–47 (Clarendon, Oxford, 1991). land plants are relicts, even though much modern species diversity 31. Kroken, S. B., Graham, L. E. & Cook, M. E. Occurrence and evolutionary significance of resistant cell walls in charophytes and bryophytes. Am. J. Bot. 83, 1241–1254 (1996). within these groups may have evolved more recently. Data from the 32. Wellman, C. H. & Richardson, J. B. Sporomorph assemblages from the ‘Lower Old Red Sandstone’ of fossil record are therefore especially important for clarifying homo- Lorne, Scotland. Special Papers Palaeontol. 55, 41–101 (1996). logies among major organ systems which may otherwise be difficult 33. Edwards, D. in Palaeozoic Palaeogeography and Biogeography (eds McKerrow, W. S. & Scotese, C. R.) 233–242 (Geological Society, London, 1990). to detect as a result of morphological divergence and extinction. 34. Morel, E., Edwards, D. & Iñiquez Rodriguez, M. The first record of Cooksonia from South America in Such combined studies of living and fossil plants provide an the Silurian rocks of Bolivia. Geol. Mag. 132, 449–452 (1995). 35. Tims, J. D. & Chambers, T. C. Rhyniophytina and Trimerophytina from the early land flora of Victoria, improved basis for comparative studies of plant development. Australia. Palaeontology 27, 265–279 (1984). They indicate, for example, that the ontogeny of leaves and spore- 36. Cai, C.-Y., Dou, Y.-W. & Edwards, D. New observations on a Pridoli plant assemblage from north bearing organs in clubmosses are likely to share substantial simi- Xinjiang, northwest China, with comments on its evolutionary and palaeographical significance. Geol. Mag. 130, 155–170 (1993). larities, but are unlikely to exhibit common features with leaves in 37. Hueber, F. M. Thoughts on the early lycopsids and zosterophylls. Ann. MO Bot. Gard. 79, 474–499 seed plants, ferns and horsetails. They also suggest that fundamental (1992). 38. Cai, C. et al. An early Silurian vascular plant. Nature 379, 592 (1996). features of land plants, such as the spore-bearing organs, stems, 39. Geng, B.-Y. Anatomy and morphology of Pinnatiramosus, a new plant from the Middle Silurian stomates and sexual organs, are each under the same kind of (Wenlockian) of China. Acta Bot. Sin. 28, 664–670 (1986). developmental control in all main groups. To explore these issues 40. Raymond, A. & Metz, C. Laurussian land-plant diversity during the Silurian and Devonian: mass extinction, sampling bias, or both? Paleobiology 21, 74–91 (1995). further, data are needed on the molecular basis of plant develop- 41. Edwards, D. & Davies, M. S. in Major evolutionary radiations (eds Taylor, P. D. & Larwood, G. P.) 351– ment from a broader selection of land plants than are currently 376 (Clarendon, Oxford, 1990). 42. Knoll, A. H., Niklas, K. J., Gensel, P. G. & Tiffney, B. H. Character diversification and patterns of under study. In the context of a more complete understanding of evolution in early vascular plants. Paleobiology 10, 34–47 (1984). plant diversity than that provided by living plants alone, such data 43. Gensel, P. G. & Andrews, H. N. Plant Life in the Devonian (Praeger, New York, 1984). Nature © Macmillan Publishers Ltd 1997 38 NATURE | VOL 389 | 4 SEPTEMBER 1997  review article 44. Taylor, T. N. & Taylor, E. L. The Biology and Evolution of Fossil Plants (Prentice Hall, New 68. Gray, J. & Boucot, A. J. Early Silurian nonmarine animal remains and the nature of the early Jersey, 1993). continental ecosystem. Acta Palaeontol. Pol. 38, 303–328 (1994). 45. Schweitzer, H.-J. Die Unterdevonflora des Rheinlandes. Palaeontographica B 189, 1–138 (1983). 69. Retallack, G. J. & Feakes, C. R. Trace fossil evidence for Late Ordovician animals on land. Science 235, 46. Gerrienne, P. Inventaire des végétaux éodévoniens de Belgique. Ann. Soc. Géol. Belg. 116, 105–117 61–63 (1987). (1993). 70. Scott, A. C., Stephenson, J. & Chaloner, W. G. Interaction and coevolution of plants and arthropods 47. Tappan, H. N. The Paleobiology of Plant Protists (Freeman, San Francisco, 1980). during the Palaeozoic and Mesozoic. Phil. Trans. R. Soc. Lond. B 336, 129–165 (1992). 48. Raven, J. Plant responses to high O2 concentrations: relevance to previous high O2 episodes. 71. Banks, H. P. & Colthart, B. J. Plant-animal-fungal interactions in early Devonian trimerophytes from Palaeogreg. Palaeoclimatol. Palaeocol. 97, 19–38 (1991). Gaspé, Canada. Am. J. Bot. 80, 992–1001 (1993). 49. Sztein, A. E., Cohen, J. D., Slovin, J. P. & Cooke, T. J. Auxin metabolism in representative land plants. 72. Edwards, D., Seldon, P. A., Richardson, J. B. & Axe, L. Coprolites as evidence for plant-animal Am. J. Bot. 82, 1514–1521 (1995). interaction in Siluro-Devonian terrestrial ecosystems. Nature 377, 329–331 (1995). 50. Edwards, D. New insights into early land ecosystems: a glimpse of a Lilliputian world. Rev. Palaeobot. 73. Allen, J. R. L. Marine to fresh water: the sedimentology of the interrupted environmental transition Palynol. 90, 159–174 (1996). (Ludlow-Siegenian) in the Anglo-Welsh region. Phil. Trans. R. Soc. Lond. B 309, 85–104 (1985). 51. Edwards, D., Fanning, U. & Richardson, J. B. Stomata and sterome in early land plants. Nature 323, 74. Melkonian, M. & Surek, B. Phylogeny of the Chlorophyta: congruence between ultrastructural and 438–440 (1986). molecular evidence. Bull. Soc. Zool. Fr. 120, 191–208 (1995). 52. Raven, J. A. Comparative physiology of plant and arthropod land adaptation. Phil. Trans. R. Soc. Lond. 75. Bremer, K., Humphries, C. J., Mishler, B. D. & Churchill, S. P. On cladistic relationships in green B 309, 273–288 (1985). plants. Taxon 36, 339–349 (1987). 53. Raven, J. A. The evolution of vascular plants in relation to quantitative functioning of dead water- 76. Garbary, D. J., Renzaglia, K. S. & Duckett, J. G. The phylogeny of land plants: a cladistic analysis based conducting cells and stomata. Biol. Rev. 68, 337–363 (1993). on male gametogenesis. Plant Syst. Evol. 188, 237–269 (1993). 54. Niklas, K. J. Plant Allometry: The Scaling of Form and Process. (Univ. Chicago Press, 1994). 77. Capesius, I. A molecular phylogeny of bryophytes based on the nuclear encoded 18S rRNA genes. J. 55. Beerbower, R. in Geological Factors and the Evolution of Plants (ed. Tiffney, B. H.) 47–92 (Yale Univ. Plant Physiol. 146, 59–63 (1995). Press, New Haven, CT, 1985). 78. Taylor, T. N. The origin of land plants: some answers, more questions. Taxon 37, 805–833 (1988). 56. Berner, R. A. GEOCARB II: a revised model of atmospheric CO2 over Phanerozoic time. Am. J. Sci. 79. Rothwell, G. W. in Pteridiology in Perspective (eds Camus, J. M., Gibby, M. & Johns, R. J.) (Royal 294, 56–91 (1994). Botanic Gardens, Kew) (in the press). 57. Mora, C. I., Driese, S. G. & Colarusso, L. A. Middle to Late Paleozoic atmospheric CO2 levels from soil 80. Albert, V. A. et al. Functional constraints and rbcL evidence for land plant phylogeny. Ann. MO Bot. carbonate and organic matter. Science 271, 1105–1107 (1996). Gard. 81, 534–567 (1994). 58. Algeo, T. J., Berner, R., Maynard, J. B. & Scheckler, S. E. Late Devonian oceanic anoxic events and biotic 81. Edwards, D., Fanning, U. & Richardson, J. B. Lower Devonian coalified sporangia from Shropshire: crises: ‘‘rooted’’ in the evolution of vascular land plants? GSA Today 5, 45, 64–66 (1995). Salopella Edwards & Richardson and Tortilicaulis Edwards. Bot. J. Linn. Soc. 116, 89–110 (1994). 59. Retallack, G. J. in Paleosols: their Recognition and Interpretation (ed. Wright, V. P.) (Blackwell, Oxford, 82. Bateman, R. M., DiMichele, W. A. & Willard, D. A. Experimental cladistic analysis of anatomically 1986). preserved lycopsids from the Carboniferous of Euramerica: an essay on paleobotanical phylogenetics. 60. Knoll, A. H. The early evolution of eukaryotes: a geological perspective. Science 256, 622–627 (1992). Ann. MO Bot. Gard. 79, 500–559 (1992). 61. Bengtson, S. (ed) Early life on Earth. (Columbia Univ. Press, New York, 1994). 83. Feist, M. & Grambast-Fessard, N. in Calcareous Algae and Stromatolites (ed. Riding, R.) 189–203 62. Taylor, T. N., Hass, H., Remy, W. & Kerp, H. The oldest fossil lichen. Nature 378, 244 (1995). (Springer, Berlin, 1991). 63. Hemsley, A. R. in Ultrastructure of Fossil Spores and Pollen (eds Kurmann, M. H. & Doyle, J. A.) 1–21 84. Hébant, C. in Bryophyte Systematics (eds Clarke, G. C. S. & Duckett, J. G.) 365–383 (Academic, (Royal Botanic Gardens, Kew, 1994). London, 1979). 64. Hueber, F. M. in International Workshop on the Biology and Evolutionary Implications of Early Devonian Plants (Westfälische Wilhelms-Universität, Münster, 1994). Acknowledgements. We thank W. G. Chaloner, D. Edwards, J. A. Raven, P. S. Herendeen, E. M. Friis, S. Bengtson and especially J. Gray for criticisms of earlier drafts of this manuscript; W. Burger, J. Cattel, A. N. 65. Simon, L., Bousquet, J., Léveque, C. & Lalonde, M. Origin and diversification of endomycorrhizal Drinnan, M. Feist, L. E. Graham, H. Haas, H. Kerp, W. A. Taylor and P. Lidmark for assistance with fungi with vascular plants. Nature 363, 67–69 (1993). illustrations. This work was supported in part by the Swedish Natural Science Research Council (NFR) 66. Selden, P. A. & Edwards, D. in Evolution and the Fossil Record (eds Allen, K. C. & Briggs, D. E. G.) 122– and the National Science Foundation. 152 (Belhaven, London, 1989). 67. Gray, J. & Shear, W. Early life on land. Am. Sci. 80, 444–456 (1992). Correspondence should be addressed to P.K. (e-mail: [email protected]). Nature © Macmillan Publishers Ltd 1997 NATURE | VOL 389 | 4 SEPTEMBER 1997 39 View publication stats