cone and ovule development in cunninghamia and taiwania

American Journal of Botany 90(1): 8–16. 2003.

CONE AND OVULE DEVELOPMENT IN CUNNINGHAMIA AND TAIWANIA (CUPRESSACEAE SENSU LATO) AND ITS SIGNIFICANCE FOR CONIFER EVOLUTION1

ALJOS FARJON2,4

AND

SOL ORTIZ GARCIA3

Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AB, UK; and 3Instituto de Ecologia, Laboratorio de Gene´tica Molecular y Evolucioń, Universidad Nacional Autońoma de Me´xico, Apt. Postal 70-275, Me´xico D.F., C.P. 04510, Mexico

2

We examined the early developmental stages of the seed cones and seeds of two conifer genera, Cunninghamia and Taiwania, using scanning electron microscope (SEM) images of freshly collected material. In recent similar studies, these two taxa were not described. The present paper aims to fill that gap. Both genera appear to have features crucial to the understanding of the evolution of the cupressaceous cone, characteristic of the families Cupressaceae and Taxodiaceae, and provide further evidence for the need to merge these families. These features are: the ovuliferous scale in Cunninghamia develops as a small lobe with each of three ovules; in Taiwania these lobes are absent, but a small ridge could be a vestige of them. In neither of these two genera does an ovuliferous scale develop to maturity and only limited intercalary growth transforms the bracts, of which only their width and final shape distinguishes them from sterile leaves. Thus, the bracts, not the ovuliferous scales, form the mature cone in these two genera. This trend is continued in more derived genera of Cupressaceae. Another key extant taxon that has helped to elucidate the evolution of this type of conifer cone is Sciadopitys; similar studies have already been done on this genus, and we compared our findings to them. We also considered certain fossil Mesozoic conifer cones, which shed further light on the evolution of the cupressaceous cone. The evidence from these various genera strongly indicates that recently reconstructed phylogenies of gymnosperms based on molecular evidence from extant taxa do not reflect the evolution that actually happened. Such studies need to take into account nonmolecular evidence, as detailed here. Key words: conifers; Cunninghamia; Cupressaceae; evolution; Mesozoic ovuliferous cones; ontogeny; phylogeny; Sciadopitys; Taiwania; Taxodiaceae.

Takaso and Tomlinson’s (1989, 1990, 1992) studies of early stages in the development of the ovuliferous cone of several taxa in Taxodiaceae have given us a much better understanding of their comparative structures. They have also raised questions about the interpretation of the cones’ parts in relation to theories about the evolution of the female conifer cone. Surveying the morphology of the early stages of the ovuliferous cones of all genera in Cupressaceae sensu lato (s.l.) (including Taxodiaceae), we found that Cunninghamia shares many characters with Taiwania and to a lesser extent with Athrotaxis (A. Farjon and S. Ortiz Garcia, unpublished data). The early development of the ovuliferous cone in the genus Sciadopitys, formerly in Taxodiaceae but now classified in its own family Sciadopityaceae (Hayata, 1931), has been described in the same manner (Takaso and Tomlinson, 1991). Although the genera Cunninghamia and Taiwania were not included in Takaso and Tomlinson’s studies, they are interesting taxa because they appear to have characters in the ovuliferous cone that are intermediate between those of Sciadopitys and other members of Cupressaceae s.l. The cone and ovule development in the genus Athrotaxis was recently described by Jagel (2002). In all these studies detailed serial observations of the earliest stages using scanning electron microscopy (SEM) have been vital to understanding homology, and we adopted this method

in our own studies of conifer cone development. No precedent to our observations was found in the literature, although Hida (1957) presented small drawings of two early stages in Cunninghamia that concur with our SEM images. She only gave a drawing of a mature scale of Taiwania; the text of the paper is in Japanese. Some studies have also investigated anatomical details of the early stages of cone development (e.g., Takaso and Tomlinson, 1992), but on the whole the histology obtained from longitudinal sections gives little extra information, and we have therefore not included this in our study. The genus Cunninghamia, as presently circumscribed (Farjon, 2001), comprises only two closely related species: C. konishii Hayata, in Taiwan and Vietnam, and C. lanceolata (Lamb.) Hook., widespread in mainland China. The differences between the two species are largely quantitative and pertain entirely to leaf characters. The smaller and more compact seed cones of C. konishii with their shorter, more obtuse scales are, as detailed below, merely a consequence of different leaf shapes and therefore wholly dependent characters. For this reason the present study is restricted to C. lanceolata. The taxonomic position of Cunninghamia has been debated from time to time, even after its inclusion in Taxodiaceae (by Pilger in 1926) became the generally accepted standard. Earlier authors on conifer cone morphology (e.g., Strasburger, 1872; Cˇelakovsky´, 1890) or general plant classification (e.g., Bentham and Hooker, 1880) treated Cunninghamia as a member of Araucariaceae, on the basis of an observed fusion of the basal parts of vascular bundles of bract and ‘‘Achselknospenbu¨ndel’’ (i.e., a vascular system thought to supply an axillary bud) and an apparent reduction of the seed scale relative to the bract and their partial fusion. As in Agathis and Araucaria, the bract, homologous to a foliage leaf from below the cone, appeared to constitute the bulk of the cone scale. Similarity with Agathis

Manuscript received 19 April 2002; revision accepted 8 August 2002. The authors wish to thank the following for their assistance with obtaining the plant material for this study: the curator and staff of the Royal Botanic Gardens, Kew, United Kingdom; Yang Jeng-Chuan, director of the Taiwan Forestry Research Institute, Taipei; and P. B. Tomlinson, Harvard University, USA. The second author wishes to thank the Royal Botanic Gardens, Kew, for support of her work at Kew from January to December 2000 under the Kew Latin American Research Fellowship Programme. 4 Author for reprint requests (e-mail: [email protected]). 1

8

January 2003] TABLE 1.

FARJON

AND

ORTIZ GARCIA—CONE

9

AND OVULE DEVELOPMENT

Species examined, with provenances, dates, and figure numbers. Provenance

Cunninghamia lanceolata (Lamb.) Hook. Kew Gardens, Acc. no. 1973-16525 Kew Gardens, Acc. no. 1973-16525 Kew Gardens, Acc. no. 1978-6254 Kew Gardens, Acc. no. 134-76.00965 Kew Gardens, Acc. no. 1978-6254 Taiwania cryptomerioides Hayata Experimental Forest, National Taiwan University, Taipei, Taiwan Courtesy P. B. Tomlinson, Harvard University, Cambridge, Massachusetts, USA

and Araucaria in the early stages of cone development was also observed by Aase (1915), who emphasized vascular anatomy in the tradition of Van Tieghem (1869) and Radais (1894) but also included observations on very young cones. ‘‘The low cushion behind the ovule in Agathis australis suggests the complete fusion of a scale to a large bract; a similar fusion is nearing its completion in Cunninghamia.’’ (Aase, 1915, p. 308). Radais (1894) did not classify Cunninghamia in Araucariaceae, but instead recognized Taxodiaceae, with (Sciadopitys [Athrotaxis, Cunninghamia {Cryptomeria, Taxodium, Sequoia}]) as its generic relationships. Hirmer (1936) grouped Sciadopitys, Cunninghamia, and Athrotaxis on the basis of their shared considerable differentiation between bract and scale, with the bract remaining prominent. Satake (1934), in another study of cone scale vascularization, disagreed and separated Sciadopitys from Taxodiaceae, but agreed with a grouping of Athrotaxis and Cunninghamia, which he suggested could constitute a family, Cunninghamiaceae. The emphasis on differences in cone morphology had been taken even further by Hayata (1932), who split the Taxodiaceae into four families, one being Cunninghamiaceae, which included Athrotaxis and Cunninghamia. The genus Taiwania is known from a single species with a remarkably disjunct distribution; the only currently known and verified indigenous populations occur in the border mountains between Myanmar (Burma) and China (Yunnan province), in the Hoang Lien Son mountain range in northern Vietnam (discovered by Vietnamese botanists, October 2001, indigenity verified by the first author, April 2002) and in Taiwan. The Chinese populations were well known to botanists starting in the early decades of the 20th century and have sometimes been regarded as different species (Gaussen, 1939; Li, 1986). Hayata (1906, 1907) considered the cones of Taiwania to be most similar in structure to those of Cunninghamia, although he saw a minute bract subtending the cone scale, which is in fact absent in both genera. He correctly observed the similarities in ovule position and the differences in the number of ovules on each scale (‘‘bract’’ in our terminology). On closer examination, Hayata (1907) also found striking similarities with Athrotaxis, even though he had seen only illustrations of this Tasmanian conifer. He therefore concluded that his new genus, Taiwania, ‘‘should be placed between Cunninghamia and Athrotaxis’’ (p. 26) with a relationship closest to Cunninghamia. He was certain that all three genera belonged in Taxodiaceae. In two recent phylogenetic analyses based on molecular data (Gadek et al., 2000; Kusumi et al., 2000) Cunninghamia, Taiwania, and Athrotaxis formed three basal branches in all par-

Date

18 25 9 25 4

Figure

Jan 2000 Jan 2000 Feb 2000 Feb 2000 Apr 2000

Fig. 3 Figs. 1, 2, 4 Fig. 5 Fig. 6 Fig. 7

28 Jan 2000

Figs. 8–11

unknown

Fig. 12

simony trees. Phylogenetic analysis of morphology (Farjon et al., 2002; A. Farjon, unpublished data) gave very similar results, with Cunninghamia and Taiwania forming one clade and Athrotaxis another in a basal polytomy. In the context of the basal phylogenetic position of these taxa, the description of seed cone ontogeny in Cunninghamia and Taiwania using SEM may help increase the understanding of the evolution of these conifers. MATERIALS AND METHODS Young seed cones of progressive stages were obtained from living trees in cultivation at the Royal Botanic Gardens, Kew, UK (C. lanceolata; for accession numbers and dates see Table 1) and at the Experimental Forest of National Taiwan University, Taipei, Taiwan (T. cryptomerioides, a 21-m-tall tree planted in 1966). Additional Taiwania material was obtained from P. B. Tomlinson at Harvard University, Cambridge, Massachusetts, USA. We collected the Kew material at intervals of 10–14 d from 18 January 2000 to 6 April 2000 and on 19 November 2000 (very early stages); the Taiwan material was collected on 28 January 2000 by personnel of the Experimental Forest and immediately shipped to Kew. All material was collected fresh, fixed in formaldehyde acetonic acid (FAA), and kept in 75% methylated spirit until the young cones were dissected. The Harvard material came from spirit-preserved specimens collected earlier from a cultivated tree. Young cones with some or most of the bracts removed were dehydrated in an ethanol series up to 100%, critical point dried, mounted with double-sided adhesive tape on stubs, and coated with platinum in a sputter coater (Balzers SCD 050, Liechtenstein). Observations of the samples were made on an S-2400 scanning electron microscope (Hitachi Ltd., Tokyo, Japan) in the herbarium at Kew at 18 kV with a standard tilt of 308 under moderate magnifications.

RESULTS Cunninghamia lanceolata (Lamb.) Hook—Both species in the genus Cunninghamia are tall trees with a strong capacity to coppice, sprouting new shoots from numerous dormant buds in the cambium. The lateral foliage branches are shed entirely; leaves do not fall separately. The phyllotaxis is spiral (helical), and the linear-lanceolate leaves are shortly decurrent and slightly constricted at the base, then twisted to spread pectinately, curved, flattened, 30–60 3 3–5(–6) mm (12–30 3 2–3 mm in C. konishii), with two bands of stomata on the underside, serrulate margins, and a pungent apex. In C. konishii there are also two continuous and distinct lines of stomata on the upper side; in C. lanceolata sometimes two intermittent lines are present. The leaves on branchlets just below the seed cones are much shorter, not pectinately spreading, and merge into the scales of full-grown cones. The cones are ovoid-glo-

10

AMERICAN JOURNAL

bose and 25–40 3 25–35 mm (15–25 3 15–20 mm in C. konishii). The numerous cone scales are attached by a pedicellate base to a slender axis, spirally arranged, imbricate, broadly triangular, coriaceous, with serrulate margins and a cuspidate apex. There are usually two inverted, winged seeds per scale, attached to a proximal ridge of intercalary growth. Ovule initiation in the trees cultivated at Kew occurs from October to November. The earliest stages of the cones we observed still had two to three parastichies of spirally arranged bract primordia situated around a low conical cone apex (Fig. 1). Later, in January, these had grown into bracts with serrulate margins and a rostrate apex (Fig. 2), and the ultimate bracts curved over the cone apex. At their base (with exception of the ultimate few bracts) ovule primordia had developed (Figs. 2, 3). The lower and middle fertile bracts bear three ovules, the smaller upper bracts two or none, and the lowest and the ultimate bracts are sterile. The primordia differentiate from cells at the base of each fertile bract, which seem to form a faint ridge (Fig. 3); in all instances the primordia come away with the bract if it is broken off from the cone axis (Figs. 2–4). The 2–3 ovules on a bract develop virtually simultaneously in January at Kew (Fig. 3). On the cone axis, a very short basal portion of the bract remains, as evidenced by the leaf-type cell structure and the single resin canal (Fig. 2). Stages of ovule differentiation progress from apex to base of the cone (Fig. 2); soon a further differentiation becomes visible (Fig. 4). Starting with the outer two ovules, a distal and slightly flattened rim emerges, which slightly later becomes apparent on the central ovule as well (Fig. 5). The central ovule in Fig. 5 also shows the emergence of the nucellus. In late February, three separate distal lobes have formed, one to each ovule (Fig. 6). The ovules have a well-defined integument and nucellus. At this stage, they are erect and positioned above the narrowed base of the bract (broken off in Fig. 6). At the pollination stage (late March–early April in Kew), the pedicellate base of the bract is further elongated, taking the ovules away from the base (Fig. 7). The ovules begin to turn their micropyle inward toward the cone axis because of this growth and especially because of the growth of the lobes, which now have fringed margins (Fig. 7). Later, the lobes will merge and form a ridge of thickened intercalary growth on the proximal adaxial side of the growing cone scale, pushing the seeds in an inverse position (not shown). The lateral ridges of the ovules (Fig. 7) will widen into the seed wings; the middle ovule aborts, and a maximum of two seeds develop on each cone scale. Taiwania cryptomerioides Hayata—Trees of this species can grow to 65–70 m tall. The foliage consists of two types of leaves: a protracted juvenile stage with falcate-subulate leaves that eventually becomes a mature stage with smaller (3–6 3 1.2–3 mm) scale leaves. Both types of leaves are spirally arranged, shortly decurrent and imbricate, but while the juvenile leaves spread out from the shoot, the mature leaves are more appressed, with incurved apices. In ovuliferous cones, the leaves (bracts) are appressed, with serrulate margins and an acute, slightly keeled apex (Figs. 8, 9). The bracts are attached by a broad base to the cone axis and, like the sterile leaves, have a wide resin canal (Figs. 9, 10). Stomata are visible on the adaxial side (Figs. 10, 11); they also occur near the base of the bract on the abaxial side. The time of ovule initiation is not exactly known; our material collected in Taiwan on 28 January 2000 (Figs. 8–11) had

OF

BOTANY

[Vol. 90

ovules in an early stage of development. Presumably, ovule primordia are initiated in late summer, as is the case in most taxa (Takaso and Tomlinson, 1989, 1990, 1992; A. Farjon and S. Ortiz Garcia, unpublished data). There are normally two simultaneously arising ovules on the base of each bract, but sometimes only a single ovule arises (Fig. 11). The ultimate bracts of the cone are sterile. The ovule primordia soon differentiate into nucellus and integument (Fig. 10), but during their development no corresponding lobes are formed, as in Cunninghamia. In Fig. 11, the groove and slightly bulging side of the single ovule could be a vestige of such a lobe, or perhaps it is a remnant of the second, abortive ovule, as it is too close to the first. In all bracts we examined at the pollination stage, lobes associated with ovules were completely absent (Fig. 12). The ovules remain close to the cone axis because no basal elongation of the bract occurred at this stage; their orientation remains virtually erect. With further growth of the bracts to form the cone scales, the young seeds become inverted as their points of attachment move outward and eventually they lie subapically on the scale (Hayata, 1907, pl. 1; A. Farjon, unpublished data). This change in position is presumably effected by a differential growth rate of different portions of the scale, although we were not able to observe this process. A narrowed, pedicellate base of the bract does not form, as it does in Cunninghamia (and Athrotaxis). A slight rim or ridge separates the proximal abaxial surface of the scale from the distal one, and each surface has a different texture. The rim is located at the same distance from the base of the scale as are the seed attachment points on the adaxial side. Our results for Taiwania have to be considered preliminary because we were limited to the material that was sent to us. This material does not represent all the stages of development, particularly not the earliest stages. DISCUSSION Systematics—The relationships of Taiwania with other genera were discussed by Hayata (1906, 1907), who concluded that ‘‘Taiwania should be placed between Cunninghamia and Athrotaxis’’ (1907, p. 26) and that its cones better resemble those of Cunninghamia and its leaves are similar to those of Athrotaxis. Sorger (1925) came to a similar conclusion, but gave more weight to the similarities in leaf morphology and anatomy shared by Taiwania and Athrotaxis and placed Taiwania closest to Athrotaxis. Modern phylogenetic reconstructions of Cupressaceae s.l. (Brunsfeld et al., 1994; Tsumura et al., 1995; Gadek et al., 2000; Kusumi et al., 2000; Farjon et al., 2002) that included all three genera placed at least Cunninghamia and Taiwania, and sometimes also Athrotaxis, in basal, but usually separate, clades in the consensus trees. A more limited morphological study by Yu and Fu (1996), including only taxa traditionally placed in Taxodiaceae, appeared to confirm the relationships between these three genera that had been inferred by the earlier authors. Karyotype analysis (Schlarbaum and Tsuchiya, 1984) revealed a general similarity between Cunninghamia and Taiwania, with ‘‘the specific chromosome type in Taiwania likely [to be] derived from [that] in Cunninghamia spp.’’ (p. 180). We therefore accept as a hypothesis that these two genera represent basal, and possibly related, clades in the phylogeny of Cupressaceae. The identity of the sister taxon to Cupressaceae s.l. is more uncertain and is probably extinct. Sciadopitys was included in Taxodiaceae by Pilger (1926) but considered distinct by Florin

January 2003]

FARJON

AND

ORTIZ GARCIA—CONE


11

Figs. 1–6. Ovuliferous cone development in Cunninghamia lanceolata. 1. Cone apex with bract primordia. 2. Cone with removed bracts; arrows point to ovule primordia. 3. Ovule primordia at the base of the bract. 4. Ovule primordia with initiation of lobes (a). 5. Ovule primordia with initiation of lobes (a) and differentiation of nucellus (b) and integument. 6. Young ovules with well-differentiated lobe (a), nucellus (b), and integument (c). Scale bars: 1, 6 5 500 mm; 2 5 1 mm; 3–5 5 200 mm.

(1922) and formally segregated by Hayata (1931) who classified it in its own family, a treatment that has become more widely accepted only in the last decade (Price and Lowenstein, 1989; Page, 1990). Some unique morphological characters separate this taxon from other conifers, most notably its ‘‘double needles,’’ which are shoot homologues with anatomical and morphological characters of both stems and leaves (Roth,

1962). Sciadopitys also has a distinct embryology (Lawson, 1910) and karyotype. The latter, with its diploid number 2n 5 20, is considered the only instance of aneuploidy known in conifers (Schlarbaum and Tsuchiya, 1985) and links Sciadopitys again with Cupressaceae and Taxodiaceae (2n 5 22). However, Li (1988) concluded that the cytological data for Sciadopitys indicate that, rather than having a karyotype de-

12

AMERICAN JOURNAL

OF

BOTANY

[Vol. 90

Figs. 7–12. Ovuliferous cone development in Cunninghamia lanceolata (7) and Taiwania cryptomerioides (8–12). 7. Development at pollination stage, with fringed lobes. 8. Cone with fertile bracts and sterile apical bracts. 9. Cone with bracts removed; arrows point to ovule primordia. 10. Bract with two ovules without lobe development. 11. Bract with single ovule and a slight differentiation beside it (a), which is separated from the integument (b). 12. Development at pollination stage, lacking lobes. Scale bars: 7–9 5 1 mm; 10, 12 5 200 mm; 11 5 100 mm.

rived from this lineage, it is more primitive than those of Taxodiaceae. Phylogenetic analyses are also inconclusive. Three molecular analyses (Brunsfeld et al., 1994; Chaw et al., 1997; Stefanovic´ et al., 1998), each using a different gene, placed taxads (Amentotaxus, Cephalotaxus, Taxus, Torreya) closer to Cupressaceae and Taxodiaceae than Sciadopitys. A fourth analysis (Tsumura et al., 1995) using restriction endonuclease

fragment length polymorphism (RFLP) placed Sciadopitys as sister to the Pinaceae. Evolution—The ontogenies of the seed cones in Cunninghamia and Taiwania demonstrate the vestigial nature of an ovuliferous scale, reduced to small distal lobes associated with the ovules in Cunninghamia and apparently absent in Taiwan-

January 2003]

FARJON

AND

ORTIZ GARCIA—CONE

ia. Takaso and Tomlinson (1991) studied the ontogeny of the seed cone in Sciadopitys. In this taxon, the ovule primordia are preceded by a rim at the adaxial base of the bract, which develops a lobed outer margin that is shortly followed by ovule primordia. There is a one-to-one relationship between ovules and lobes, as in Cunninghamia, but both are more numerous (usually 8–9 and up to 14) and the lobes continue to grow, fusing into a single scale proximally. As in Cunninghamia, the basal portion of the bract elongates, displacing the ovules away from the cone axis. Subsequent intercalary growth soon inverts the ovules, but in Sciadopitys this growth is much more pronounced and occurs roughly equally in both the bract and the lobed axillary scale, both elongating and thickening them. Both structures become largely fused in the mature bract-scale complex. In Cunninghamia, the lobes also fuse with each other and with the basal portion of the bract, but growth then ceases there, whereas it continues in the bract, resulting in minimal thickening. In Taiwania, in the absence of lobes, growth is restricted to the bracts, but some differentiation remains between proximal and distal portions, with the division marked just above the ovules and causing their inversion. From the ontogenies of the seed cones of some other taxa more or less basal in the Cupressaceae s.l., it is known that the initiation of an ovuliferous scale is either delayed until after the pollination stage of the ovules, as in Athrotaxis (Jagel, 2002), or that the ovuliferous scale has virtually disappeared, as in the genera Metasequoia, Sequoia, and Sequoiadendron (Takaso and Tomlinson, 1992; A. Farjon and S. Ortiz Garcia, unpublished data). The ovules originate at the base of the bracts, and secondary intercalary growth mostly forms the cone scales. In Athrotaxis, a ridge develops distal to the ovules on the lower adaxial side of the bract, which may be homologous with the lobes of Cunninghamia, but this ridge only becomes apparent when the ovules are well developed and is sometimes absent. In later stages, intercalary growth, starting in this zone, forms a bulging adaxial outgrowth as well as a proximal thickening of the abaxial side of the bract. The histogenesis of the bract and ovule at this stage in some related taxa fails to demonstrate any vascular traces that indicate the presence of an ovuliferous scale; instead there appears an ovular trace (Takaso and Tomlinson, 1992, Figs. 48, 54, 60). We concur with these authors that vascular traces appear as required and do not think they are the harbingers of (vestigial) structures. In A. selaginoides D. Don and A. laxifolia Hook., the bract apex that did not swell during growth is relatively large; in A. cupressoides D. Don, it is small and the final morphology of the cone approaches that of Sequoia. In taxa still higher in the phylogenetic tree of the Cupressaceae, the ovules originate axillary to the bracts or even on the cone apex (Jagel and Stu¨tzel, 2001; Farjon and Ortiz Garcia, 2002). Here any vestiges of a separate ovuliferous scale have disappeared completely. Palaeobotany can give further insight into the evolution of the seed cone of Cupressaceae s.l. as it appears in its basal Recent taxa. From the large fossil record (see for reviews, e.g., Stewart and Rothwell [1993] and Ohsawa [1994]) we will select some examples to demonstrate the possible evolution of a type of seed cone scale in the two genera studied here. In the Upper Permian conifer Pseudovoltzia liebeana (Geinitz) Florin, an acute bract subtended a five-lobed, partially fused ovuliferous scale that bore three inverted seeds on the largest lobes (Schweitzer, 1963; Clement-Westerhof, 1987). The lobes


13

of the ovuliferous scale exceeded the bract and fusion was only partial near the base (Fig. 13A). A conifer of Lower Triassic age, Voltziopsis (Fig. 13B) had four to five lobes, each with an inverted seed near its point of proximal fusion, subtended by a bifurcated bract (Townrow, 1967; Clement-Westerhof, 1988). This fossil genus in particular may fit the role of progenitor of the cone present in Sciadopitys (Takaso and Tomlinson, 1991) (Fig. 13C, D). The Triassic conifer Cycadocarpidium is another type of cone scale demonstrating this development (Miller, 1988), although it may not be closely related to Cupressaceae s.l. It had a single large acute bract and an ovuliferous scale partly fused with it, with 2–3 partly fused appendages but only 1–5 ovules. The species Cycadocarpidium pilosum Grauvogel-Stamm (Fig. 13F), after Grauvogel-Stamm (1978), had three ligulate appendages reminiscent of those in Cryptomeria and 2–3 inverted seeds. Its cones apparently disintegrated on the trees, as numerous separate bract-scale complexes have been found in the sediment. The fossil conifer Elatides williamsonii T. M. Harris (Fig. 13E) from the middle Jurassic (Harris, 1943) had cone scales similar to those of Cunninghamia (Fig. 13G, H) and Taiwania (Fig. 13I, J). The scales were thin and stomatiferous abaxially, and they had a similar difference of scale texture in their proximal and distal parts, divided by a low ridge, below which were six inverted seeds. The leaves of this conifer were similar to those of Cryptomeria (probably the juvenile stage) and the mature leaves of Taiwania. Another species, E. harrisii Z. Y. Zhou of Lower Cretaceous age from China (Zhou, 1987), had leaves and a terminal cone similar to Cunninghamia, but the thin cone scales ended in long caudate cusps. The cuticle of the scales was similar to that of the leaves, including the stomata. There were probably three seeds per scale, but nothing certain is known about their position. The genus Elatides extends back to the Jurassic (Schweitzer and Kirchner, 1996). The genus Cunninghamiostrobus dates back to the Early Cretaceous (Ohsawa, 1994) or Late Cretaceous (Ohana and Kimura, 1995) but is better known from the middle Tertiary; its cones are also similar to those of Cunninghamia. Phylogenies that include fossil conifers (Miller, 1988, 1999; Ohsawa, 1997) are limited by a paucity of comparable data, with synapomorphies often strongly dependent on interpretations of incompletely preserved specimens. Owing to limits of character availability, analyses can be carried out only with limited taxon sampling. Of the ‘‘basal conifers,’’ Pseudovoltzia and Cycadocarpidium showed close relationships with Sciadopitys and Cunninghamia in those consensus cladograms that included both Recent and at least one of these fossil taxa (Miller, 1999). Thus, the Majonicaceae from the Permian and Triassic could perhaps have contained a conifer that was the common ancestor of Sciadopitys as well as the taxodiaceous Cupressaceae. Several fossils resembling Sciadopitys from the Jurassic and Cretaceous are now attributed to the fossil family Miroviaceae (Bose and Manum, 1991; Manum, van Konijnenburg-van Cittert, and Wilde, 2000) but could well be closely related to this genus. Sciadopitys itself may also have been present from these periods (Manum, 1987) and was certainly present in the Late Cretaceous (Christophel, 1973; Saiki, 1992) and probably represents the last survivor of a split between its immediate ancestor and a lineage leading to the taxodiaceous Cupressaceae. We do not, of course, imply with the examples given in Fig. 13 that there is a phylogenetic tree that includes all these taxa; we only want to demonstrate an evolutionary trend in the development of certain ovuliferous conifer cones.

14

AMERICAN JOURNAL

OF

BOTANY

[Vol. 90

Fig. 13. The gradual modification and final reduction of the ovuliferous scale in a hypothetical evolution of Upper Permian (Pseudovoltzia) and Mesozoic (Voltziopsis, Triassic; Elatides, Middle Jurassic; Cycadocarpidium, Triassic) conifers to Recent conifers. The fossil conifers are shown in mature stages; the Recent conifers in juvenile (C, G, I) and mature stages; b 5 bract. (A). Pseudovoltzia (after Schweitzer, 1963). (B). Voltziopsis (after Clement-Westerhof, 1988). (C, D). Sciadopitys. (E). Elatides (interpreted from illustrations in Harris, 1943). (F). Cycadocarpidium (after Grauvogel-Stamm, 1978). (G, H). Cunninghamia. (I, J). Taiwania. Sizes are variable and not to scale.

With such conifers as Elatides, Cunninghamiostrobus, and indeed Cunninghamia and Taiwania, the evolution of a retarded ovuliferous scale set the stage for the extraordinary diversity of seed cone types seen in Cupressaceae today. No longer bound by an axillary appendage, ovules moved to the cone axis and the bracts took on the form and function of sometimes highly specialized cone scales. One of the consequences of this evolution was a different mode of seed wing ontogeny. In Araucariaceae and Pinaceae, the other extant conifer families

with winged seeds, the wing is derived from the seed scale epidermis. In the absence of a seed scale, wings had to be formed by other tissue. In Cupressaceae s.l., they are outgrowths of the seed integument; this transformation has already taken place in Sciadopitys, which still has an ovuliferous scale, and is present also in Cunninghamia. The transformation of the seed wing from the ovuliferous scale to the ovule integument is a synapomorphy. It places Sciadopitys in a close phylogenetic relationship with Cupressaceae s.l. The sister-

January 2003]

FARJON

AND

ORTIZ GARCIA—CONE

group relationship of taxads (Cephalotaxaceae, Taxaceae) with Cupressaceae, as inferred from several gene sequences (see references above), is, in view of this evidence and what is understood from palaeobotany (Miller, 1999), very doubtful. Recently, Axsmith, Taylor and Taylor (1998), using one of these papers (Chaw et al., 1997) as an example, stressed that ‘‘interpreting phylogeny and character evolution goes well beyond molecular techniques and tree construction’’ (p. 107). The present study appears to underline their observation, especially for groups of plants that are characterized by a profusion of extinct taxa. LITERATURE CITED AASE, H. C. 1915. Vascular anatomy of the megasporophylls of conifers. Botanical Gazette 60: 277–315. AXSMITH, B. J., E. L. TAYLOR, AND T. N. TAYLOR. 1998. The limitations of molecular systematics: a palaeobotanical perspective. Taxon 47: 105– 108. BENTHAM, G., AND J. D. HOOKER. 1880. Ordo CLXV. Coniferae. In G. Bentham and J. D. Hooker [eds.], Genera Plantarum 3 (1): 420–442. Reeve, London, UK. BOSE, M. N., AND S. B. MANUM. 1991. Additions to the family Miroviaceae (Coniferae) from the Lower Cretaceous of West Greenland and Germany: Mirovia groenlandica n. sp., Tritaenia crassa (Seward) comb. nov., and Tritaenia linkii Ma¨gdefrau et Rudolf emend. Polar Research 9: 9–20. BRUNSFELD, S. J., P. S. SOLTIS, D. E. SOLTIS, P. A. GADEK, C. J. QUINN, D. D. STRENGE, AND T. A. RANKER. 1994. Phylogenetic relationships among the genera of Taxodiaceae and Cupressaceae: evidence from rbcL sequences. Systematic Botany 19: 253–262. CˇELAKOVSKY´, L. J. 1890. Die Gymnospermen: eine morphologisch-phylogenetische Studie. Abhandlungen der Ko¨nigliche Bo¨hmische Gesellschaft fu¨r die Wissenschaft (Prag) 4: 1–148. CHAW, S. M., A. ZHARKIKH, H. M. SUNG, T. C. LAU, AND W. H. LI. 1997. Molecular phylogeny of extant gymnosperms and seed plant evolution: analysis of nuclear 18S rRNA sequences. Molecular Biology and Evolution 14: 56–68. CHRISTOPHEL, D. C. 1973. Sciadopitophyllum canadense gen. et spec. nov.: a new conifer from western Alberta. American Journal of Botany 60: 61–66. CLEMENT-WESTERHOF, J. A. 1987. The Majonicaceae, a new family of Late Permian conifers. [Aspects of Permian palaeobotany and palynology VII]. Review of Palaeobotany and Palynology 52: 375–402. CLEMENT-WESTERHOF, J. A. 1988. Morphology and phylogeny of Palaeozoic conifers. In C. B. Beck [ed.], Origin and evolution of gymnosperms. Columbia University Press, New York, New York, USA. FARJON, A. 2001. World checklist and bibliography of conifers, 2nd ed. Royal Botanic Gardens, Kew, Richmond, Surrey, UK. FARJON, A., N. T. HIEP, D. K. HARDER, P. K. LOC, AND L. AVERYANOV. 2002. A new genus and species in Cupressaceae (Coniferales) from northern Vietnam, Xanthocyparis vietnamensis. Novon 12: 179–189. FARJON, A., AND S. ORTIZ GARCIA. 2002. Towards the minimal conifer cone: ontogeny and trends in Cupressus, Juniperus and Microbiota (Cupressaceae s.s.). Botanische Jahrbu¨cher fu¨r die Systematik 124: 129–147. FLORIN, C. R. 1922. On the geological history of the Sciadopitineae. Svensk Botanisk Tidskrift 16: 260–270. GADEK, P. A., D. L. ALPERS, M. M. HESLEWOOD, AND C. J. QUINN. 2000. Relationships within Cupressaceae sensu lato: a combined morphological and molecular approach. American Journal of Botany 87: 1044–1057. GAUSSEN, H. 1939. Une nouvelle espe`ce de Taiwania, T. flousiana. Travaux du Laboratoire Forestier Toulouse (tome) 1, (vol.) 3, (art.) 2: 1–9. GRAUVOGEL-STAMM, L. 1978. La flore du gre`s a` Voltzia (Bundsandstein supe´rieur) des Vosges du Nord (France): morphologie, anatomie, interpre´tations phylogeńique et paleógeógraphique. Sciences Geólogiques, Me´moire 50: 1–225. HARRIS, T. M. 1943. The fossil conifer Elatides williamsonii. Annals of Botany (London), new series, 7: 325–339. HAYATA, B. 1906. On Taiwania, a new genus of Coniferae from the island of Formosa. Journal of the Linnean Society of London, Botany 37: 330– 331.


15

HAYATA, B. 1907. On Taiwania and its affinity to other genera. Botanical Magazine (Tokyo) 21: 21–27. HAYATA, B. 1931. The Sciadopityaceae represented by Sciadopitys verticillata Sieb. et Zucc., an endemic species of Japan. Botanical Magazine (Tokyo) 45: 567–569. HAYATA, B. 1932. The Taxodiaceae should be divided into several distinct families, i.e. the Limnopityaceae, Cryptomeriaceae, Taiwaniaceae, and the Cunninghamiaceae; and further Tetraclinis should represent a distinct family, the Tetraclinaceae. Botanical Magazine (Tokyo) 46: 24–27. HIDA, M. 1957. The comparative study of Taxodiaceae from the standpoint of the development of the cone scale. Botanical Magazine (Tokyo) 70: 44–51 (in Japanese, with English abstract). HIRMER, M. 1936. Entwicklungsgeschichte und vergleichende Morphologie des weiblichen Blu¨tenzapfens der Coniferen. [Die Blu¨ten der Coniferen I]. Bibliotheca Botanica 114: 1–100. JAGEL, A. 2002. Morphologische und morphogenetische Untersuchungen zur Systematik und Evolution der Cupressaceae s. l. (Zypressengewa¨chse). Ph.D. dissertation, Ruhr-University, Bochum, Germany. JAGEL, A., AND T. STU¨TZEL. 2001. Zur Abgrenzung von Chamaecyparis Spach und Cupressus L. (Cupressaceae) und die systematische Stellung von Cupressus nootkatensis D. Don [5 Chamaecyparis nootkatensis (D. Don) Spach]. Feddes Repertorium 112: 179–229. KUSUMI, J., Y. TSUMURA, H. YOSHIMARU, AND H. TACHIDA. 2000. Phylogenetic relationships in Taxodiaceae and Cupressaceae sensu stricto based on matK gene, chlL gene, trnL-trnF IGS region, and trnL intron sequences. American Journal of Botany 87: 1480–1488. LAWSON, A. A. 1910. The gametophytes and embryo of Sciadopitys verticillata. Annals of Botany (London) 24: 403–422. LI, L. C. 1986. Cytological investigation of Taiwania flousiana Gaussen (Taxodiaceae). Acta Phytotaxonomica Sinica 24: 376–381 (in Chinese, with English abstract). LI, L. C. 1988. The systematic position of Sciadopitys Sieb. et Zucc. based on cytological data. Guihaia 8: 135–141. MANUM, S. B. 1987. Mesozoic Sciadopitys-like leaves with observation on four species from the Jurassic of Andøya, Northern Norway, and emendation of Sciadopityoides Sveshnikova. Review of Palaeobotany and Palynology 51: 145–168. MANUM, S. B., J. H. A. VAN KONIJNENBURG-VAN CITTERT, AND V. WILDE. 2000. Tritaenia Ma¨gdefrau et Rudolf, Mesozoic ‘‘Sciadopitys-like’’ leaves in mass accumulations. Review of Palaeobotany and Palynology 109: 255–269. MILLER, C. N. 1988. The origin of modern conifer families. In C. B. Beck [ed.], Origin and evolution of gymnosperms. Columbia University Press, New York, New York, USA. MILLER, C. N. 1999. Implications of fossil conifers for the phylogenetic relationships of living families. Botanical Review 65: 239–277. OHANA, T., AND T. KIMURA. 1995. Further observations of Cunninghamiostrobus yubariensis Stopes and Fujii from the Upper Yezo Group (Upper Cretaceous), Hokkaido, Japan. Transactions and Proceedings of the Palaeontological Society of Japan, new series, 178: 122–141. OHSAWA, T. 1994. Anatomy and relationships of petrified seed cones of the Cupressaceae, Taxodiaceae, and Sciadopityaceae. Journal of Plant Research 107: 503–512. OHSAWA, T. 1997. Phylogenetic reconstruction of some conifer families: role and significance of permineralized cone records. In K. Iwatsuki and P. H. Raven [eds.], Evolution and diversification of land plants. SpringerVerlag, Tokyo, Japan. PAGE, C. N. 1990. Coniferophytina. In K. Kubitzki [ed.], The families and genera of vascular plants, vol. 1, K. U. Kramer and P. S. Green [eds.], Pteridophytes and gymnosperms, 282–361. Springer-Verlag, Berlin, Germany. PILGER, R. 1926. Phylogenie und Systematik der Coniferae. In A. Engler and K. Prantl [eds.], Die natu¨rlichen Pflanzenfamilien, 2nd ed., vol. 13, Gymnospermae, 121–403. Engelmann, Leipzig, Germany. PRICE, R. A., AND J. M. LOWENSTEIN. 1989. An immunological comparison of the Sciadopityaceae, Taxodiaceae, and Cupressaceae. Systematic Botany 14: 141–149. RADAIS, M. 1894. L’anatomie compareé du fruit des conife`res. Annales des Sciences Naturelles Botaniques, series 7, 19: 165–368. ROTH, I. 1962. Histogenese und morphologische Deutung der Doppelnadeln von Sciadopitys. Flora 152: 1–23. SAIKI, K. 1992. A new sciadopityaceous seed cone from the Upper Cretaceous of Hokkaido, Japan. American Journal of Botany 79: 989–995.

16

AMERICAN JOURNAL

SATAKE, Y. 1934. On the systematic importance of the course of vascular bundles in the cone scales of the Japanese Taxodiaceae (preliminary report). Botanical Magazine (Tokyo) 48: 186–20 (in Japanese with English summary). SCHLARBAUM, S. E., AND T. TSUCHIYA. 1984. The chromosomes of Cunninghamia konishii, C. lanceolata and Taiwania cryptomerioides (Taxodiaceae). Plant Systematics and Evolution 145: 169–181. SCHLARBAUM, S. E., AND T. TSUCHIYA. 1985. Karyological derivation of Sciadopitys verticillata Sieb. et Zucc. from a pro-taxodiaceous ancestor. Botanical Gazette (Crawfordsville) 146: 264–267. SCHWEITZER, H. J. 1963. Die weibliche Zapfen von Pseudovoltzia liebeana und seine Bedeutung fu¨r die Phylogenie der Koniferen. Palaeontographica Abteilung B, Palaeophytologie 113: 1–29. SCHWEITZER, H. J., AND M. KIRCHNER. 1996. Die Rhaeto-Jurassischen Floren des Irans und Afghanistans: 9. Coniferophyta. Palaeontographica Abteilung B, Palaeophytologie 238: 77–139. SORGER, O. 1925. Die systematische Stellung von Taiwania cryptomerioides ¨ sterreichische Botanische Zeitschrift 74: 81–102. Hayata. O STEFANOVIC´, S., M. JAGER, J. DEUTSCH, J. BROUTIN, AND M. MASSELOT. 1998. Phylogenetic relationships of conifers inferred from partial 28S rRNA gene sequences. American Journal of Botany 85: 688–697. STEWART, W. N., AND G. W. ROTHWELL. 1993. Palaeobotany and the evolution of plants, 2nd ed. Cambridge University Press, Cambridge, UK. STRASBURGER, E. 1872. Die Coniferen und die Gnetaceen: eine morphologische Studie. Fischer, Jena, Germany.

OF

BOTANY

[Vol. 90

TAKASO, T., AND P. B. TOMLINSON. 1989. Aspects of cone and ovule ontogeny in Cryptomeria (Taxodiaceae). American Journal of Botany 76: 692– 705. TAKASO, T., AND P. B. TOMLINSON. 1990. Cone and ovule ontogeny in Taxodium and Glyptostrobus (Taxodiaceae-Coniferales). American Journal of Botany 77: 1209–1221. TAKASO, T., AND P. B. TOMLINSON. 1991. Cone and ovule development in Sciadopitys (Taxodiaceae-Coniferales). American Journal of Botany 78: 417–428. TAKASO, T., AND P. B. TOMLINSON. 1992. Seed cone and ovule ontogeny in Metasequoia, Sequoia and Sequoiadendron (Taxodiaceae-Coniferales). Botanical Journal of the Linnean Society 109: 15–37. TOWNROW, J. A. 1967. On Voltziopsis, a southern conifer of Lower Triassic age. Papers and Proceedings of the Royal Society of Tasmania 101: 173– 188. TSUMURA, Y., K. YOSHIMURA, N. TOMARU, AND K. OHBA. 1995. Molecular phylogeny of conifers using RFLP analysis of PCR-amplified specific chloroplast genes. Theoretical and Applied Genetics 91: 1222–1236. VAN TIEGHEM, P. 1869. Anatomie compareé de la fleur femelle et du fruit des Cycadeés, des Conife`res et des Gne´taceés. Annales des Sciences Naturelles, series 5, Botanique 10: 269–304. YU, Y. F., AND L. K. FU. 1996. Phylogenetic analysis of the family Taxodiaceae. Acta Phytotaxonomica Sinica 34: 124–141 (in Chinese, with English abstract). ZHOU, Z. Y. 1987. Elatides harrisii sp. nov. from the Lower Cretaceous of Liaoning, China. Review of Palaeobotany and Palynology 51: 189–204.