马尾松雌球果的发生和早期发育研究

采用常规石蜡制片技术对马尾松雌球果的发生和早期发育进行了研究。结果表明：雌球果原基发生时间为10月中旬,不同的树龄和着生部位,其发生时间不同。雌球果原基与营养茎端在外部形态及内部细胞组织学分区结构有明显差异。营养茎端外形扁平,内部顶端分生组织结构有顶端原始细胞区、中央母细胞区、形成层状过渡区、周围分生组织区及肋状分生组织区5个明显的分区;而雌球果原基外形呈圆锥状,内部结构只有套层和髓区。12月初,最初的苞片原基在雌球果原基的鳞片的叶腋处产生,之后其由基部向顶部连续发生。翌年1月初,在苞片原基的叶腋处,珠鳞原基发生,发生方向亦为向顶发育。2月底,苞片体积不再发生变化,珠鳞膨大端的基部的近轴面分化出2个倒生胚珠。从雌球果原基发生到胚珠分化历时4个多月。亚热带的冬季气候对马尾松雌球果的生长发育没有明显的抑制作用。     

MORE EVIDENCE FOR CONIFER DIVERSITY IN THE UPPER TRIASSIC OF NORTH CAROLINA

An ovulate strobilus from the Upper Triassic Deep River Basin, North Carolina, has helically arranged, loosely aggregated, elongated, spatulate bracts with axillary ovule-bearing appendages with about 8–10 ovules attached in two lateral rows, with outwardly directed micropyles. The axillary ovuliferous appendage is homologous with the voltzialean fertile dwarf shoot, but probably not directly evolved from it. More credible is a suggested origin from a completely fertile axillary appendage such as that of the Lower Permian Trichopitys. The occurrence of this cone, Metridiostrobus palissyaeoides, gen. and sp. nov., along with Compsostrobus neotericus and Voltzia andrewsii, reflects considerable diversity among conifer ovulate cones during the Upper Triassic.     

OVULE INVERSION IN THE EARLIEST CONIFERS

Recent investigations of ovulate conifer cones from southern Europe and midcontinent North America have independently documented that certain Paleozoic walchians have inverted ovules, rather than the erect ovules previously thought to characterize the most primitive conifers. Reinvestigation and consideration of other walchian conifers, including Moyliostrobus and Lebachia piniformis (sensu Florin), reveals that they also had inverted ovules. These different patterns of ovule orientation demonstrate that the nature and the polarity of the character states are dramatically different than popularly believed, and the shift from megasporophyll to ovuliferous cone-scale occurred within the Paleozoic walchians.     

Nageiaceae—A New Gymnosperm Family

A new monotypic gymnosperm family, Nageiaceae D. Z. Fu, is separated from Podocarpaceae. It is characterized by having multinerved leaves without costae, and primitive shoot-like female reproductive organs (female strobili). The new family contains a single genus consisting of 2 sections, 5 species and is distributed along the western coast of the Pacific, from low coastal mountains of eastern and southern Asia to the Phillipines and Papua New Guinea. The first species in the Nageiaceae was described as an angiosperm, Myrica nagi Thunb. (1784), but it was soon recognized to be a gymnosperm belonging to a new genus, and was renamed as Nageia japonica Gaert. (1788). The generic name, Nageia, however, has seldom been used, and the members of Nageia have generally been treated as an isolated section of Podocarpus in the Podocarpaceae. When revising the Podocarpaceae, De Laubenfels (1969) established a new genus Decussocarpus based on Nageia, but several years later (1987) he revived the old generic name, Nageia. Page ( 1988,1990)considered Nageia to be a valid generic name and redefined it as a natural genus. The distinctive,broadly lanceolate, multinerved leaves (without costae) of Nageia are rather unusual in gymnosperms,only being similar to those of Agathis in the Araucariaceae, their leaves are also similar to each other in anatomy. For example, there are many single vascular bundles arranged parallelly, between which occur sclerenchyma cells in the mesophyll. Apparently,leaves in Nageia are rather similar both externally and internally to paleogymnosperm cordaitean leaves, and sclerenchyma cells found in Nageia might be the remains of straps between veins in cordaitean leaves. In addition to leaf characters, the large and nearly round pith of the young shoot in Nageia appears to be a reminiscent of the large pith in cordaitean stem. The female reproductive organs (female strobili ) in Nageia are shoot-like. The female strobilus has a sterile terminal bud, and several opposite or subopposite sterile scaly bracts on its axis; two opposite megasporophylls are found near the axis apex and both have an anatropous ovule which is almost entirely covered by the megasporophyll; a bract is partly adnate to the lower back of the megasporophyll;mature arillate seeds are 1-2 or occasionally 3 in number; the axis becomes woody when the seeds mature, but in some species (N. wallichiana) the upper part of the axis becomes fleshy (in the shape of a receptacle), in which no distinct boundary was found between the fleshy receptacle and the woody part, and both have the same scaly bracts or traces. Many characters in Nageia are distinctly different from those in Podocarpus. Leaves in the Podocarpaceae have distinct midribs; in Podocarpus, the reproductive organ, which was generally thought to be similar to that in Nageia, has no terminal bud, and its bract is entirely free from the lower back of the megasporophyll, the fleshy receptacle is derived from both the axis and the sterile bracts (except the lowest two), and the female strobilus at the seed stage has a secondary stalk. The multinerved leaf in Nageia can rarely be found in most of the living gymnosperms except in some rather isolated groups, such as Araucariaceae, Ephedraceae,Ginkgoaceae and Welwitschiaceae. Paleobotanical evidence shows that multinerved leaves have been found in all of the geological ages from the Paleozoic to the present, and such a shoot-like female reproductive organ as in Nageia was found in some paleogymnosperms. It is very difficult to determine the systematic positions of these fossil plants because of lacks adequate material of reproductive organs or even lack of complete vegetative organs. The vascular system and leaf characters of gymnosperms are considered to be very conservative, and the fact that the common leaf shape and venation exist in both fossil and living gymnosperms could imply that there exists a multinerved-leaved evolutionary line ( M-line ) in gymnosperms, which could be traced back to the paleogymnosperm cordaitean plants or even older ones with multinerved leaves. The different types of the female strobili (female reproductive organs) of living gymnosperms, regardless of having one or only several seeds without a typical cone or many seeds with a cone, might have been derived from shoot-like or spikelike female reproductive organs possessed by their common ancestor.The fossil eviden ce shows that the typical cone similar to those of living gymnosperms first appeared in the Jurassic, much later than the single-seeded fossil plant without cones. The seed fossil appeared in the late Devonian Period. It is very difficult to infer the relationships among living gymnosperms, which are hardly derived from one another. But an analysis of the strobili, including the axis structure and position, number, morphology and degree of adnation of the phyllomes on them, would be helpful to the study of their phylogeny. It is evident, therefore, that the gymnosperms with leaves having a midrib might also have a rather long evolutionary course,but no transition between the midrib and multinerved patterns of leaf venation has so far been found in both living and fossil plants. Finally, it is noteworthy that the Nageiaceae are distributed along the western coast of the Pacific, where many primitive representatives, both in gymnosperms and angiosperms, still survive. This would be advantageous to the consideration of Nageiaceae as a primitive representative, or a descendant of fhe paleogymnos-perm cordaitean plants.     

A Contribution to the Embryology of Ginkgo with a Discussion on the Affinity of the Ginkgoales

The ancestry of Ginkgoales or Ginkgophyta can be traced back to the Paleozoic. But today this order is represented only by a single species, Ginkgo biloba. Seward (1838) considered Ginkgo as one of the wonders of the world; it has persisted with little change until the present through a long succession of ages when the earth was inhabited by animals and plants for the most part far removed, in kind as in time, from their living descendents. Ginkgo is generally believed to be native to China, but so far it has not been found in wild state. A number of studies concerning the embryogeny of Ginkgo have been reported, on the basis of the materials accumulated during the past years and supplemented in 1978–1980, the embryogeny of Ginkgo is here described. Finally the phylogeny of Ginkgoales is discussed by comparing with the embryogeny of other groups of the living gymnosperms. Pollination usually takes place from the end of April to the first days of May and fertilization occurs around August 16–20 in the suburbs of Peking. Thus, the interval between pollination and fertilization is a few days less than four months. The embryo of Ginkgo is generally considered as suspensorless. In authors’ opinion, however, the somewhat elongated and much enlarged cells at the micropylar end of the embryo may be considered as the reduced suspensor cells though they are not the typical ones. There is no distinct demarcation between the proembryo and the young embryo in Ginkgo. In comparison with the embryogenesis in Coniferales, the tissue differentiation of the late embryo of Ginkgo is rather indistinct. Many authors such as Chamberlain (1935), Florin (1949), Delevoryas (1963), Sporne (1965) and others divide the gymnosperms into two major groups, Cycadophyta and Coniferophyta. From the point of view of morphological structure there are many significant common features shared by Ginkgoales and Coniferales. For example, Ginkgo possesses long shoot and short shoot while some conifers also have long and short shoots. The anatomical features of the stem of Ginkgo such as the well-developed secondary xylem, relatively small pith and the presence of bordered pits on the tracheids are also similar to those of the Coniferales. Not only Ginkgo has its characteristic leaf shape and the dichotomous venation but also they are quite different from the fronds of cycads. From the reproductive structure, on the other hand, Ginkgo and Cycadales are rather similar: both of them having one sulcate pollen grains, the pollen tube being of haustorial nature, the sperms being released from the base and not from the tip of the pollen tube, the presence of mulficiliate and rather large sperms, development of large female gametophyte bearing archegonia with exceptionally large eggs, more divisions of the free nuclear stage in the proembryo and less distinct differentiation of the ate embryo. All these features are primitive embryological characters. Thus, from the point of view of embryology the Ginkgo is close related to the Cycadales rather than to the Coniferales. Since Ginkgophyta are related to Cycadophyta in reproductive structures on one hand and similar to Coniferophyta in morphological and anatomical characters of the vegetative organs the Ginkgophyta are related to Cycadophyta in reproductive structures on one hand and similar to Coniferophyta in morphological and anatomical characters of the vegetative organs on the other hand, it indicates the interrelationship among these groups is clearly shown. Recently the discovery and studies on the progymnospermopsida (Beck, 1976) are worth notice, this fossil group contains the plants with certain characters of some conifers and certain characters of some cycads and this kind of plants bearing both homosporous and heterosporous forms are similar to ferns. They link the gymnosperms to the ferns. The present evidence, therefore, shows that the gymnosperms are very probably monophyletic and the progymnosperms might be the ancester of all the gymnosperms. The embryological characteristics supports the monophyletic origin of the gymnosperms.     

Ginkgo biloba: The Ovuliferous Leaf and Its Phylogenetic Implication

This paper reports the occurrence of the ovuliferous leaves on a Maidenhair tree (Ginkgo biloba)growing in Longdong, Yiyuan County, Shandong Province. This phenomenon is discussed in connection with the history of Ginkgoales. It is believed that this discovery serves as an illustration of the hypothesis that the secondary apical position of ovules (Ginkgoales) is a modification of the marginal position (Pteridospermidae). Again, it probably leads to the assumption that, in Ginkgo, “the foliage leaves” bear some features of “thesporophylls” and they are derived from telomes.     

Nystroemiaceae,a new family of Permian gymnosperms from China with an unusual combination of features

Nystroemiaceae is proposed as a new family of gymnosperms from the Permian of Cathaysia that adds to the diversity of gymnosperms known from this critical time in seed plant evolution. This family is characterized by bifurcating and highly branched pinnate ovuliferous organs bearing bicornute ovules (seeds) and entire leaves with anastomozing veins that are born on complex and modern-looking branching systems with clear axillary branching. The reconstruction is based on numerous large specimens from two localities in North China, in which the different plant parts are attached to each other. The ovulate structures show some apparently plesiomorphic (primitive) character states more typical of early seed plants, whereas the leaves and branches show the clearly apomorphic (derived) character states of broad-leaved gymnosperms.     

Morphological and ontogenetic studies on southern African podocarps. Shoot apex morphology and ovuliferous cone initiation

STOFFBERG, E., 1991. Morphological and ontogenetic studies on southern African podocarps. Shoot apex morphology and ovuliferous cone initiation. Four species of Podocarpus indigenous to southern Africa were investigated. The morphology of the primordium of the female cone is compared with that of the shoot apex. Rhythmic growth occurs in Podocarpus. The external morphology of bud scales protecting dormant shoot apices is described and illustrated. Female strobili of the three species of section Podocarpus studied are initiated in the axils of euphylls during the spring growth flush as laterally flattened triangular structures. The axillary position of a female cone indicates that it is a modified shoot. The first two cone bracts (prophylls) are formed approximately at right angles to the subtending bract (one or both are fertile), while the 3rd and 4th bracts originate on the anterior and posterior sides of the strobilus respectively. Two to four bracts per cone are formed, not in pairs–the phyllotaxis is spiral. In P.falcatus primordia of female strobili and vegetative branches could be distinguished only after emergence of the seed scale complex. Based on cell differentiation, well-defined cytological zones can be distinguished in the shoot apex and it is classified as being of the Abies-Cryptomeria-lype. Meristematic zones of cone primordia and vegetative branches are basically similar, although the former are less well defined. No gradual transition from a vegetative to a reproductive apex could be identified and it would seem that the fate of axillary buds are determined at the time of their origination or even before.     

Implications of fossil conifers for the phylogenetic relationships of living families

Fossils have played a central role in our understanding of the evolution of conifers. Interpretation of the seed cone as a compound strobilus and the homologies of the ovuliferous scales of modern conifers with the axillary dwarf shoot of Pennsylvanian forms are based on fossils. Similarly, early evolutionary trends involving the reduction, fusion, and planation of the fertile and sterile elements of the axillary dwarf shoot, leading to structures characteristic of modern families, are documented in Late Permian and Triassic conifers. However, a phylogeny elucidating the derivation of modern families from fossil forms based on shared derived features has been elusive. The present cladistic treatment using 11 characters of ovulate cones and one of pollen grains suggests three phylogenetic groups of Late Paleozoic conifers, represented loosely by the Emporicaceae, Utrechtiaceae, and Majonicaceae of Mapes and Rothwell. The Taxaceae appears to have diverged from ancestors within the Utrechtiaceae, whereas the other modern families owe their origins to the Majonicaceae. The origin of the Taxodiaceae appears to have been biphyletic.Taxodium, Cupressus andSciadopitys are strongly linked toDolmitia of the Majonicaceae, butCryptomeria, Cunninghamia andAraucaria are grouped together and diverge basal to the former taxa.Pinus branches from a position basal to the known genera of the Majonicaceae and all modern families except the Taxaceae.Podocarpus also diverges basal toMajonica but may share an ancestor with this genus;Cepahalotaxus diverges basal to theDolmitiaPseudovoltzia subclade but distal toMajonica. Similarly, the Cheirolepidiaceae originated from basal members of the Majonicaceae and shows no close phylogenetic relationship with any modern family. Except for a strong linkage betweenCycadocarpidium and theAraucariaCunninghamia subclade, genera of the Voltziaceae appear to have branched more or less independently from within the Majonicaceae and show no strong affinity with modern conifers. Thus differences between modern conifer families are due mainly to their divergence from different Paleozoic ancestors.     

Morphological and molecular data on the origin of angiosperms: on a way to a synthesis

Molecular phylogenetic data have drastically changed the views on the phylogeny of higher plants. All the extant gymnosperms were asserted as a monophyletic group opposed to the highly isolated angiosperms. The 'Anthophyte Theory' was thus rejected. The identification and analysis of gymnosperm orthologues of genes regulating flower development in angiosperms resulted in the formulation of the 'Mostly Male Theory' of the evolutionary origin of flower; this theory does not contradict the concept of monophyly of all the extant gymnosperms. The Mostly Male Theory assumes that the origin of angiosperms was caused by a loss of the Needle family gene that effected ovuliferous (female) organs and the translocation of the ovules onto the adaxial side of some of the (male) leafy microsporangiophores. Having acquired ovules, the former microsporangiophores started evolving into the carpels. The prerequisite bisexual design of the ancestral fructification thus becomes unnecessary. Indeed, this assumption suggests the deriving of Angiosperms from any gymnosperm plant with leafy microsporangiophores. The problem of carpel origin has subsequently changed to some degree into the problem of the origin of the bitegmic anatropous ovule presumably inherent in ancestral Angiosperms. The Mostly Male Theory consideredeither Corystospermataceae (= Umkomasiaceae) or Caytoniaceae to be the forerunners of such an ovule. Yet the capsules of Corystospermataceae distinctly differ from angiosperm ovules in the locations of their adaxial/abaxial sides, while Caytoniaceae had no leafy microsporangiophores. This inconsistency suggests that functions of the Needle family regulatory genes in Gymnosperms should be much better understood to appraise properly both the possibilities and the consequences of their hypothetical loss by the emerging angiosperms. Moreover, the extant gymnosperm groups are actually held as monophyletic and contrasted to Angiosperms on the basis of analysing the unrepresentative scant remnants of these, mostly extinct, taxa. Therefore, traditional botanical and paleobotanical data should not be rejected. In any case, Meyen's idea angiosperms origin from Bennettitales is worth being retained as a hypothesis to be tested with new results of both paleobotany and molecular biology.     

Morphological "primary homology" and expression of AG-subfamily MADS-box genes in pines, podocarps, and yews

The morphological variation among reproductive organs of extant gymnosperms is remarkable, especially among conifers. Several hypotheses concerning morphological homology between various conifer reproductive organs have been put forward, in particular in relation to the pine ovuliferous scale. Here, we use the expression patterns of orthologs of the ABC-model MADS-box gene AGAMOUS (AG) for testing morphological homology hypotheses related to organs of the conifer female cone. To this end, we first developed a tailored 3'RACE procedure that allows reliable amplification of partial sequences highly similar to gymnosperm-derived members of the AG-subfamily of MADS-box genes. Expression patterns of two novel conifer AG orthologs cloned with this procedure-namely PodAG and TgAG, obtained from the podocarp Podocarpus reichei and the yew Taxus globosa, respectively-are then further characterized in the morphologically divergent female cones of these species. The expression patterns of PodAG and TgAG are compared with those of DAL2, a previously discovered Picea abies (Pinaceae) AG ortholog. By treating the expression patterns of DAL2, PodAG, and TgAG as character states mapped onto currently accepted cladogram topologies, we suggest that the epimatium-that is, the podocarp female cone organ previously postulated as a "modified" ovuliferous scale-and the canonical Pinaceae ovuliferous scale can be legitimally conceptualized as "primary homologs." Character state mapping for TgAG suggests in turn that the aril of Taxaceae should be considered as a different type of organ. This work demonstrates how the interaction between developmental-genetic data and formal cladistic theory could fruitfully contribute to gymnosperm systematics.     

Morphology of the seed-scale complex in Picea abies (L.) Karst. (Pinaceae)

The seed-scale is made up of only the two prophylls of the axillary shoot of the bract. These bear one ovule each dorsally. In contradistinction to Pinus , ovules may come to bear hypertrophied, foliarized integuments, but the latter are less developed as the scale is more deeply split into its two components. With total splitting, ovules disappear. This fact, along with previous observations on Pinus , is taken to mean that ovules are but dorsal appendages of the scale components, with no homologous structure in the vegetative state. Florin's view of ovules being, or topping, discrete sporophylls cannot be supported. Our interpretation, on the contrary, is in accordance with Schweitzer's more recent observations.     

银杏营养器官解剖结构的观察

为了揭示不同树龄银杏的根、茎、叶解剖结构以及内生菌分布情况,本研究采用石蜡切片法对银杏(Ginkgo biloba L.)根、茎、叶显微解剖结构进行了观察。结果显示:(1)一年生银杏幼根不含树脂道,内生菌含量低,而皮层中含有大量蛋白细胞;多年生银杏老根含有较多树脂道,皮层细胞中含有大量内生菌并有针晶物质分布,未发现蛋白细胞。(2)一年生银杏幼茎有明显的角质层,皮层分布有大量蛋白细胞,韧皮射线及髓部发达,其中髓由大量薄壁细胞构成并且有蛋白细胞分布,未观察到树脂道但有簇晶物质存在。(3)多年生银杏叶片海绵组织疏松,具有树脂道,叶肉细胞含有簇晶物质;气孔下陷并具有耐旱的结构特点。本结果可为研究不同树龄银杏对环境的适应性变化以及内生菌特点提供参考。     

FERTILE CONIFEROPHYTE REMAINS FROM THE LATE TRIASSIC DEEP RIVER BASIN,NORTH CAROLINA

Seed cones (Compsostrobus neotericus gen. et sp. nov.), pollen cones, and vegetative remains of coniferophytes occur in Upper Triassic rocks of the Deep River Basin (Pekin Formation) of Central North Carolina. Seed cones have spatulate ovuliferous scales, each with two ovules and subtended by an elongated bract with an attenuate tip. Cuticle of seed cones resembles that of leaves on vegetative axes. Slender leaves are borne along two sides of the axis. Pollen cones have helically arranged microsporophylls, each with two abaxial sporangia bearing pollen grains of the Alisporites type. Seed cones, pollen cones, and vegetative remains suggest a coniferophyte very modern in aspect.     

Reconstruction of the Jurassic conifer Sewardiodendron laxum (Taxodiaceae)

Compressed seed cones and pollen cones of Sewardiodendron laxum are described from the Middle Jurassic of Yima, Henan, central China. They are either organically attached to or associated with leafy shoots. Seed cones are terminally borne. Each cone is ovate to elongated, up to 6.5 cm long and 3.5 cm wide, and consists of a stout axis and numerous helically arranged bract-scale complexes. The bract protrudes beyond and is partially fused with the reduced ovuliferous scale. The ovuliferous scale bears approximately six inverted, small, and flattened seeds. Pollen cones are borne in terminal clusters. Microsporophylls are helically arranged, each bearing three abaxial, basally fused pollen sacs. Pollen is assaccate, rounded, and with an inconspicuous pore. Morphological, structural, and cuticular features of seed cones, pollen cones, and leafy shoots of S. laxum are compared with those of fossil and extant conifers. S. laxum is included in Taxodiaceae and believed to have its closest affinities with a Mesozoic conifer Elatides and a group of Cunninghamia-like conifers. It is reconstructed as a half-evergreen tree that lived in a humid, warm-temperate climate.     

CONE AND OVULE DEVELOPMENT IN SCIADOPITYS (TAXODIACEAE-CONIFERALES)

Ontogeny of seed cones of Sciadopitys, with special reference to the ovule-supporting structure, is studied in material collected in Japan and Massachusetts. Cones are initiated as lateral or terminal structures in summer and complete the formation of most organs before winter. Bract development is well advanced before ovule-supporting structures are initiated. Continued cone development involves the formation of a narrow ridge of tissue in the axil of each fertile bract. This ridge develops a series of nine (but up to 12) apical lobes in centrifugal order, each of which is the primordium of a future tooth on the ovuliferous scale. Ovules are initiated as outgrowths of the adaxial surface of each lobe so that there is a one-to-one ratio between lobes and ovules. Intercalary extension of the ovuliferous scale itself (distally) and the common base of the bract and ovuliferous scale (proximally) greatly extends the complex. The ovuliferous scale eventually exceeds the subtending bract and its apex becomes recurved. Bracts each have a single trace, but each ovuliferous scale has a pair of traces that proliferate distally to irrigate ovule and scale lobe. Intercalary growth results in recurvature of the ovule trace. The organization of the cone is directly comparable with certain Permian fossils. Sciadopitys also seems unique within the Taxodiaceae in its centrifugal development of the ovule-supporting complex.     

Evolution of Reproductive Organs in Land Plants

LEAFY gene is the positive regulator of the MADS-box genes in flower primordia. The number of MADS-box genes presumably increased by gene duplications before the divergence of ferns and seed plants. Most MADS-box genes in ferns are expressed similarly in both vegetative and reproductive organs, while in gymnosperms, some MADS-box genes are specifically expressed in reproductive organs. This suggests that (1) the increase in the number of MADS-box genes and (2) the subsequent recruitment of some MADS-box genes as homeotic selector genes were important for the evolution of complex reproductive organs. The phylogenetic tree including both angiosperm and gymnosperm MADS-box genes indicates the loss of the A-function genes in the gymnosperm lineage, which is presumably related to the absence of perianths in extant gymnosperms. Comparison of expression patterns of orthologous MADS-box genes in angiosperms, Gnetales, and conifers supports the sister relationship of Gnetales and conifers over that of Gnetales and angiosperms predicted by phylogenetic trees based on amino acid and nucleotide sequences. Received 30 July 1999/ Accepted in revised form 9 September 1999     

An embryological study and systematic significance of the primitive gymnosperm Ginkgo biloba

Ginkgo biloba L.is considered one of the most ancient seed plants,with several primitive features of plant reproductive process.However,the phylogenetic position of Ginkgo and its relationship with other extant seed plants remain unclear.To gain a better understanding of these issues,we observed the embryological development of G.biloba using semi-thin sections and scanning electron microscopy.In late August,the zygote moved from the end of the micropylar to the middle of the archegonium,and mitosis resulted in many free nuclei distributed randomly in the archegonium.Afterwards,the cell wall was formed and the proembryo began to differentiate into the embryonal region and the underdeveloped presuspensor region.In early October,the embryo differentiated into two cotyledons,plumule,hypocotyl,radicle,and suspensor tissues.Subsequently,the two cotyledons grew rapidly,but the undeveloped suspensor began to degenerate and gradually disappear,indicating that the embryo had begun to mature.During early embryo development,the main supply of nutrients was carbohydrate in the cells of the jacket,tentpole,and surrounding endosperm,whereas endosperm provided nutrients during embryo maturation.Our results indicate that Ginkgo is extremely similar to cycads in terms of embryology but more similar to conifers in macromorphology and vegetative anatomy,suggesting that the Ginkgo lineage may have an intermediate phylogenetic position between cycads and conifers.