Am. J. Bot. Join the BSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (14)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schönenberger, J.
Right arrow Articles by Friis, E. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Schönenberger, J.
Right arrow Articles by Friis, E. M.
Agricola
Right arrow Articles by Schönenberger, J.
Right arrow Articles by Friis, E. M.
(American Journal of Botany. 2001;88:467-480.)
© 2001 Botanical Society of America, Inc.

Fossil flowers of ericalean affinity from the Late Cretaceous of Southern Sweden1

Jürg Schönenberger2 and Else Marie Friis

Department of Palaeobotany, Swedish Museum of Natural History, Box 50007, SE-104 05 Stockholm, Sweden

Received for publication December 7, 1999. Accepted for publication May 4, 2000.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 FORMAL DESCRIPTION
 LITERATURE CITED
 
Charcoalified fossil flowers of a new genus and species (Paradinandra suecica) with affinities to Ericales s.l. (sensu lato) are described from the Late Cretaceous (Santonian-Campanian) from Southern Sweden. The flowers are pentamerous, hypogynous, and actinomorphic. Aestivation of sepals and petals is imbricate-quincuncial. The androecium consists of an outer whorl with single episepalous stamens and an inner whorl with paired epipetalous stamens. Pollen is small and probably tricolpate. Three carpels form a syncarpous ovary with numerous campylotropous ovules on parietal placentae. The styles are free for most of their length. The structure of mature fruits and seeds is unknown. Clear distinction of sepals and petals, possible dehiscence of anthers by restricted slits, presence of a nectary, and the general floral construction (salverform corolla) with a canalized access to the floral center clearly indicate insect pollination of the fossil flowers. Comparisons with extant taxa demonstrate that Paradinandra suecica shares many similarities with Ericales s.l. and in particular with members of Ternstroemiaceae, Theaceae, and Actinidiaceae. However, it is neither identical to any one genus of these families nor to any of the previously described ericalean taxa from the Cretaceous and thus provides further evidence of the diversity of Cretaceous ericalean plants.

Key Words: Ericales s.l. • flowers • fossils • insect pollination • Late Cretaceous • Paradinandra suecica • Sweden • Ternstroemiaceae


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 FORMAL DESCRIPTION
 LITERATURE CITED
 
During the last 20 yr our knowledge about angiosperm fossils from the Cretaceous has increased at a rapid rate (for a review see Crane, Friis, and Pedersen, 1995 ). By the Late Cretaceous, floral construction and architecture had reached an evolutionary level almost comparable to that of many recent angiosperms. This allows us in many cases to compare these fossils with living angiosperms and draw direct conclusions not only about their systematic relationships, but also about floral biology and pollination (e.g., Basinger and Dilcher, 1984 ; Crepet, Friis, and Nixon, 1991 ).

The ongoing study of plant fossils from Late Cretaceous (Santonian-Campanian) deposits from Åsen in southern Sweden has revealed many extremely well-preserved fossil flowers, fruits, and seeds. Specimens belonging to the following major lineages of angiosperm evolution have so far been described: Chloranthaceae (Chloranthistemon—Crane, Friis, and Pedersen, 1989 ; Eklund, Friis, and Pedersen, 1997 ), Proteales (Platananthus—Friis, Crane, and Pedersen, 1988 ), Saxifragales (Archamamelis—Endress and Friis, 1991 ), Fagales (Antiquocarya, Caryanthus, Manningia—Friis, 1983 ), Hydrangeaceae (?) (Scandianthus—Friis and Skarby, 1982), Ericales (Actinocalyx—Friis, 1985a ), and Escalloniaceae (?) (Silvianthemum—Friis, 1990 ).

The present study describes fossil floral structures including floral buds and flower fragments from the Åsen locality. Organization and construction of these flowers clearly indicate that they were insect pollinated. The specimens reflect a syndrome of features that supports a placement in the Ericales s.l. sensu APG (1998) . Delimitation and systematic position of the Ericales sensu Cronquist (1981) have undergone many changes lately. Not only are they now placed within the asterid clade, but they are also circumscribed in a broader sense including several taxa previously referred to different dilleniid orders (Kron, Chase, and Hills, 1991 ; Anderberg, 1992 ; Downie and Palmer, 1992 ; Hufford, 1992 ; Olmstead et al., 1992 ; Anderberg, 1993 ; Chase et al., 1993 ; Judd and Kron, 1993 ; Olmstead et al., 1993 ; Swensen and Chase, 1995 ; Morton et al., 1996 ; Soltis et al., 1997 ; Nandi, Chase, and Endress, 1998 ).

Angiosperm fossils from the Tertiary are often so similar to modern groups that they can be placed in modern genera. Ericales s.l. have an extensive fossil record from that period of time (e.g., Collinson, 1978 ; Friis, 1985b ; Manchester, 1994 ). Several ericalean taxa have also been reported from the Late Cretaceous, although the details of their relationships within the group are often less sure (Actinocalyx with sympetalous flowers—Friis, 1985a ; fruits and seeds of different ericalean affinities—Knobloch and Mai, 1986 ; Paleoenkianthus with awned anthers—Nixon and Crepet, 1993 ; stamens and flowers of possible ericalean affinities, not yet formally described— Crepet and Nixon, 1996 ; Parasaurauia with free styles that emerge from an apical depression in the ovary—Keller, Herendeen, and Crane, 1996 ; flowers and fruits of possible ericalean affinities, not yet formally described—Herendeen et al., 1999 ; Takahashi, Crane, and Ando, 1999a ).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 FORMAL DESCRIPTION
 LITERATURE CITED
 
The fossils examined in this study were isolated from deposits of Late Santonian—Early Campanian age exposed in the Höganäs AB kaolin quarry (now used as a garbage dump) at the Åsen locality in the Kristianstad Basin, Scania (56°9' N, 14°30' E). The sediments consist of unconsolidated sands and clays, from which the plant fossils were extracted by sieving of bulk samples in water. More details on the site and standard preparation methods for the material are given in Friis, Crane, and Pedersen (1988) .

The material studied in this paper occurs in two samples (Åsen 2 and Åsen 3; collected by S. Lindbom). The fossils are charcoalified (fusanized), with their original three-dimensional form intact or rarely slightly compressed. A discussion of this type of preservation is given in Friis, Crane, and Pedersen (1988) . The fossil material comprises a few more or less complete floral buds [S105998-01/S105998-02, S106015-01–S106015-04, S106016, S106194-01, S107080; all specimens and preparations of fossil material are deposited in the collections of the Swedish Museum of Natural History (S)], a few ovaries with androecium and perianth sometimes partially preserved (S106021, S106022, S106253, S106254, S106256), as well as many abscised corolla tubes with adhering filaments and in few cases also with anthers attached (S106006–S106014, S106378, S107076–S107079). All the fossils are astonishingly small (<3.5 mm in length). However, this is not only true for most other floral remains from the same locality (e.g., Friis and Skarby, 1981 ; Friis, 1984, 1985c ), but it is also the case for most mesofossils from other Cretaceous deposits in Europe, North America, and central and eastern Asia (e.g., Nixon and Crepet, 1993 ; Friis, Pedersen, and Crane, 1994 ; Eklund and Kvacek, 1998 ; Frumin and Friis, 1999 ; Herendeen et al., 1999 ; Takahashi, Crane, and Ando, 1999a ).

In preparation for scanning electron microscopy (SEM) specimens were mounted on SEM stubs, sputter coated with Au, and then studied with a Phillips 515 scanning microscope. Two of the buds (S105998-01 and S106015-01) were carefully dissected with the help of fine needles and tweezers under a stereomicroscope. The different organs then were remounted (on separate stub S105598-02 and stubs S106015-02–S106015-04, respectively), sputter coated, and then studied in detail with scanning electron microscopy. Two of the buds (S106194-01, section slides S106194-02–S106194-53; S107080-01, section slides S107080-02–S107080-35) and two corolla tubes (S107081-01, section slides S107081-02–S107081-19; S107082-01, section slides S107082-02–S107082-21) were embedded in 2-hydroxyethyl methacrylate (Kulzer's Technovit 7100; Heraeus Kulzer GmbH, Philipp-Reis-Strasse 8/13, D-61273 Wehrheim/Ts., Germany) and cut at 3–6 µm with a rotary microtome. This is a widely used routine method for anatomical and histological investigations of animal and plant tissues. For description of embedding method and products used see Igersheim (1993) and, in particular, Igersheim and Cichocki (1996) , who describe embedding and sectioning of charcoal specimens in detail.

Parts of the section series were drawn with a drawing tube. Shape, size, and arrangement of floral organs were reconstructed from the sections of the best-preserved bud, which is also the holotype (S106194-01).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 FORMAL DESCRIPTION
 LITERATURE CITED
 
The following description of the flowers is mainly based on two of the floral buds (S106015-01-03, SEM, Fig. 4; S106194, SEM, Figs. 2, 3 and microtome sections). Both are extraordinarily well preserved and allow detailed study of morphology and anatomy. One of the buds (S106015) represents a slightly older pre-anthetic stage, which is indicated by its size and by the presence of immature pollen grains in the anthers. In the sectioned specimen no pollen grains were found.



View larger version (158K):
[in this window]
[in a new window]
  		Figs. 1–7. Paradinandra suecica. Floral morphology and transverse section of sepal. 1. Pedunculate flower in axil of subtending bract. S106021. 2. Floral bud. Arrowhead indicates scale-like structures in axil of bracteole. S106194-01. 3. Same bud as in Fig. 2 , from side. 4. Floral bud. Sepals partly broken off. S106015-01. 5. Part of sepal with unicellular trichomes. S106015-04. 6. Papillate epidermis of sepal. S106015-04. 7. Transverse section of sepal from bud in Figs. 2 and 3 . Star indicates prominent median vein; arrowheads point to papillate epidermal cells. S106194-29. Scale bars in Figs. 1–4 = 1 mm; in Figs. 5, 7 = 100 µm; in Fig. 6 = 10 µm. Figure Abbreviations: B, bracteole; G, gynoecium; iS, incomplete septum; L, locule; N, nectary; P, petal; S, sepal; sB, subtending bract.

 
Floral organization and morphology
The flowers are borne in the axils of subtending bracts. The peduncle is short and rather thick (Fig. 1). Four to six small, scale-like structures are present in the axils of the two imbricate bracteoles, which are at least half as long as the whole bud (Figs. 2, 3). The flowers are pentamerous and the organ whorls alternate. The flowers are hypogynous, actinomorphic, and bisexual. The buds are ~3.5 mm long and ~1.2 mm wide.

Calyx
The sepals are completely free from each other and were probably initiated in a spiral sequence. This is indicated by their different insertion levels on the floral axis (Fig. 38m) and a divergence angle of about 137° between subsequent organs, which results in a quincuncial-imbricate aestivation (two sepals overlap their neighbors, two are overlapped by their neighbors, and one is half-overlapped). The sepals are broadly lanceolate with obtuse apices (reconstruction from section series, Figs. 37, 38). Abaxially, the sepals have unicellular, simple trichomes up to 0.5 mm long (Figs. 2, 5) concentrated along a well-developed median vein, which is prominent in the upper third of the sepal (Fig. 7). Each sepal has four or six additional veins extending only along a part of the sepal's length (Fig. 37). The central area of each sepal, where the veins are found, is made up of several cell layers, whereas the lateral parts are very thin and membranaceous. The abaxial epidermis is slightly papillate (Figs. 6, 7).



View larger version (56K):
[in this window]
[in a new window]
  		Fig. 38. Paradinandra suecica. Line drawings (a–m) of transverse section series (S106194-02–S106194-53) of floral bud (S106194-01) and floral diagram (n). Colors indicate different organs: gray = bracteoles, green = sepals, red = petals, yellow = episepalous stamens (light yellow = filaments of episepalous stamens), lilac = epipetalous stamens (light lilac = filaments of epipetalous stamens), light blue = nectary, blue = gynoecium. Scale bar = 1 mm

 


View larger version (45K):
[in this window]
[in a new window]
  		Fig. 37. Paradinandra suecica. Series of transverse sections (S106194-02–S106194-53) of floral bud (S106194-01) and schematic diagram showing relative sizes and shapes of reconstructed floral organs, levels a–i corresponding to figured sections a–i. Scale bar = 1 mm

 
Corolla
The corolla is thought to be sympetalous, but this cannot be demonstrated unequivocally neither from sections of buds (Fig. 37h, i) and abscised corolla tubes (Fig. 20), nor from specimens studied with the SEM (Figs. 18, 19). The sections indicate a short united zone at the very base of the corolla. Union of the petal bases is also suggested by the fact that abscised corollas are preserved as distinct tubular structures. However, for most of its length the corolla, which opens incompletely during anthesis, consists of free petal lobes. The petals are mitriform with their broadest part shortly below their middle (reconstructed from sectioned bud; Fig. 37). Petal aestivation in the bud is imbricate-quincuncial (Figs. 37, 38).



View larger version (174K):
[in this window]
[in a new window]
  		Figs. 18–27. Paradinandra suecica. Floral morphology and transverse section of corolla tube. 18. Abscised corolla tube from side. S107078-04. 19. Base of abscised corolla tube. S106378-05. 20. Transverse section of corolla tube and filaments. S107082-07. 21. Apical view of corolla tube with attached filaments and anthers. S106007. 22. Apical view of corolla tube with attached filaments. Arrowhead indicates anther within tube. S106010. 23. Close-up of anther within corolla in Fig. 22 . 24–26. Tricolpate (?) pollen grains on adaxial surface of corolla tube. S107078-04. 27. Close-up of slightly foveolate pollen exine. S107078- 04. Scale bars in Figs. 18–22 = 1 mm; in Fig. 23 = 100 µm; in Figs. 24–26 = 10 µm; and in Fig. 27 = 1 µm

 
Androecium
All specimens studied have 15 stamens. The filaments are united at their very base and are probably fused to the base of the corolla (Figs. 8, 37i). Also in this case it is difficult to establish with certainty whether the organs are really united or whether they only adhere to each other. The stamens probably were initiated in two whorls as is indicated by their arrangement in the bud (Figs. 37, 38). The outer episepalous whorl consists of five stamens and the inner epipetalous whorl consists of ten paired (doubled) stamens. The filaments are massive at their base and taper towards the point at which the anther is attached (Figs. 8–10, 21). The anthers are basifixed and the joint between filament and connective is relatively thin (Figs. 15, 16). The anthers are dithecal, tetrasporangiate, and X-shaped, i.e., the two thecae are basally and apically separated (Figs. 11, 15–17, 23). The type of dehiscence is not completely clear as no well-preserved anthetic anthers were found. In pre-anthetic anthers a furrow that runs along nearly the whole length of the thecae is present (Fig. 12). However, enlarged epidermal cells, which are especially prominent at the distal ends of the thecae, may indicate dehiscence by restricted apical slits or pores (Figs. 12, 13).



View larger version (171K):
[in this window]
[in a new window]
  		Figs. 8–17. Paradinandra suecica. Floral morphology and transverse sections of stamens. 8. Part of corolla tube with attached filaments, anthers broken off. S105998-02. 9. Part of corolla tube with adhering filaments. Note elongate filament. S106015-04. 10. Stamen with attached anther. S106015-03. 11. X- shaped anther, from abaxial side. S106015-03. 12. Anther from side with pore-like opening at apical end. S106015-03. 13. Close-up of apical end of theca showing enlarged cells along stomium. S106015-03. 14. Immature pollen in opened anther. S106015-03. 15. Transverse section of one of the five shorter epipetalous stamens at the level of the joint between anther and filament (arrowhead). Pollen sacs are collapsed. S106194-17. 16. Transverse section of episepalous stamen at the level of joint between anther and filament (arrowhead). Pollen sacs are collapsed. S106194-21. 17. Transverse section of epipetalous stamen at the level of the connective. Anther is tetrasporangiate and pollen sacs are collapsed. S106194-19. Scale bars in Figs. 8, 9 = 0.5 mm; in Figs. 10–13, 15–17 = 100 µm; in Fig. 14 = 10 µm

 
The anthers are stacked on three levels in the bud (reconstructed from section series). All the anthers are more or less of the same size, whereas the filaments differ markedly in length (Fig. 37). The five filaments of the outer stamen whorl are on average 0.75 mm long and the filaments are upright. One stamen of each stamen pair of the inner whorl has a slightly longer filament than its neighbor. The pattern is irregular, i.e., short and long stamens of this whorl do not always alternate. The five longer filaments are on average 0.55 mm long, and they are upright as well. The filaments of the five shorter stamens of the inner whorl are on average 0.45 mm long and thus only slightly shorter. However, the upper parts of these shortest filaments are bent outwards in the bud in a way that the anthers of these five stamens are positioned on an even lower level. This results in a stacking of the 15 anthers on three levels within the bud (Figs. 37, 38d–h). Because no complete anthetic flowers were found, it is not possible to establish whether this difference in stamen length is retained during later development or whether it is simply a way of dense packing of the floral organs within the bud, which is lost later in ontogeny. In abscised corolla tubes with attached filaments the filaments are normally broken off at different levels, which makes it difficult to recognize any pattern of anther placement. At least in some specimens there is evidence of some filaments being longer than others (Figs. 21, 22).

Pollen
Immature pollen grains were found within the anthers of one of the buds (Fig. 14). Mature pollen grains are present on many of the abscised corolla tubes. The pollen grains are similar in all specimens studied and are always restricted to the inside of the corolla tubes. Similar pollen has not been observed on other fossil remains from the same sample, and it is therefore most likely that pollen and corolla tubes correspond to the same flower. The pollen was studied using SEM only, and details of pollen ultrastructure and apertures are therefore not fully explored. The pollen is small (10–14 µm in length) and probably tricolpate, with long colpi. The exine is finely foveolate (Figs. 24–27).

Nectary
A shallow nectary disc is present around the base of the ovary. It is emarginate as it slightly extends towards the narrow spaces between the filaments (Figs. 28, 31, 32, 37h, i).



View larger version (161K):
[in this window]
[in a new window]
  		Figs. 28–36. Paradinandra suecica. Floral morphology and transverse sections of gynoecium. 28. Gynoecium with three styles and nectary disc. S106015- 02. 29. Immature stigma. S106015-02. 30. Part of style. Arrowheads point to ventral slit. S106015-02. 31. Apical view of gynoecium, styles broken off. S106254. 32. Opened ovary showing parietal placentation. Star indicates open space in center of ovary (same specimen as in Fig. 31 ). 33. Base of styles. Arrowhead points to cavity in apical part of ovary. S106015-02. 34. Lateral view of opened ovary with numerous ovules. Star indicates open space in center of ovary. 35. Campylotropous ovules with reticulate surface. S106254. 36. Schematic drawing and series of transverse sections showing cavities in apical part of ovary. Lines connecting Figs. 36b–e indicate plane of schematic longitudinal section in Fig. 36a . Arrowheads indicate one of three cavities in apical part of ovary. (a) Schematic longitudinal section of gynoecium. Lines b–e indicate approximate levels of transverse section in Figs. 36b–e . (b) Uppermost transverse section. Styles free from each other; line. (c) Section through united styles with three cavities. (d) Section through top of ovary. Three cavities are visible. (e) Lowermost section through apical parts of locule (appearing three-loculed on top of ovary). Arrowhead indicates one of the cavities (the sections are not exactly transversal, therefore only one of the three cavities is present at this level) S106194-15/16. Scale bar in Fig. 28 = 0.5 mm; in Figs. 29, 30 = 10 µm; in Figs. 31–34, 36 = 100 µm; and in Fig. 35 = 50 µm

 
Gynoecium
The gynoecium is tricarpellate and syncarpous (Figs. 28, 31). The three styles are elongate and slightly flattened (Figs. 28, 37c–e). They are free over most of their length; only their very bases seem to be united (Fig. 36c). An inconspicuous ventral slit is present over the whole length of the styles (Fig. 30). The stigma, as seen in the bud, is small and simple; no papillae were found (Fig. 29). At the top of the ovary three small, apically open cavities are present. They extend from the basal (united) part of the styles down into the incomplete septa between the apical ends of the locules (Figs. 33, 36a–e). The dorsal vascular bundles of the carpels are prominent on the outside of the globose ovary (Fig. 31). The carpel margins fail to meet in the center over most of the length of the ovary. The ovary therefore is unilocular and the placentation is parietal (Figs. 32, 34, 37g). Numerous ovules are borne on the intruding margins of the carpels. The ovules are curved (campylotropous) and have a reticulate surface (Figs. 32, 34, 35). On some of the ovaries scattered unicellular trichomes are present.

Structure of mature fruits and seeds is unknown.

Floral construction
The general shape of the flowers is salverform. They are tubular in the lower part and spreading distally (Fig. 18). The upper part consists of the free petal lobes. The access to the floral center and the nectary is canalized and further narrowed by the fleshy filaments (Figs. 8, 9, 21).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
FORMAL DESCRIPTION
 LITERATURE CITED
 
Systematic affinity and comparison with extant plants
The combination of floral characters and the general appearance of the flowers are, on first sight, not clearly indicative of any extant angiosperm family. However, two of the characters, the stamen arrangement with a single episepalous stamen and epipetalous stamen pairs and the campylotropous ovules, are rather distinctive. A comparable stamen arrangement occurs in ~20 eudicot families (Ronse Decraene and Smets, 1996 ), while campylotropy is found, predominantly or as an exception, in >65 families belonging to different orders of both monocots and dicots (Bouman and Boesewinkel, 1991 ). The combination of the two characters is, however, not very common among angiosperm families, but does occur in the Ericales s.l. Epipetalous stamen pairs are present in Ebenaceae, Sapotaceae, Styracaceae, and Ternstroemiaceae (Ronse Decraene and Smets, 1996 ). The same stamen arrangement as observed in the fossil flowers is also present as the basic androecial pattern in young flowers of Saurauia subspinosa in the Actinidiaceae (Brown, 1935 ). Campylotropous ovules occur in Diapensiaceae, Ericaceae, Myrsinaceae, Primulaceae, Theophrastaceae, and Ternstroemiaceae. Other floral characters such as the presence of two bracteoles, free or basally united petals, a compound ovary with almost completely free styles together with the general appearance of the flowers indicate a systematic position of the fossil flowers near the Theaceae, Ternstroemiaceae, and Actinidiaceae (Table 1). These three families were traditionally placed in the Theales within the subclass Dilleniidae (Cronquist, 1981 ). Recent molecular studies indicate that the Theales in this sense are not monophyletic (Swensen and Chase, 1995 ; Morton et al., 1996 ). Two of the theaceous subfamilies, Theoideae and Ternstroemioideae, fall into the Ericales s.l. and are raised to family level. They form two independent clades that are not sister taxa and are most closely related to the core Ericales (Morton et al., 1996 ). The Actinidiaceae are also referred to the Ericales s.l. in analyses of morphological and molecular data (Chase et al., 1993 ; Olmstead et al., 1993 ; Morton et al., 1997 ; Soltis et al., 1997 ; Nandi, Chase, and Endress, 1998 ) and are probably the basal clade of the core Ericales (Anderberg, 1992, 1993 ; Judd and Kron, 1993 ; Kron, 1996 ; Morton et al., 1996 ). Morphologically, the flowers of Theaceae, Ternstroemiaceae, and basal core Ericales, including the Actinidiaceae, share many similarities among each other (Judd et al., 1999 ) and with the fossil flowers. The Ternstroemiaceae have representatives with both epipetalous stamen pairs (Visnea—Payer, 1857 ; Corner, 1946 ; Ronse Decraene and Smets, 1996 ) and campylotropous ovules (Keng, 1962 ; Tsou, 1995 ) and are of particular interest for comparison with the fossil material.


View this table:
[in this window]
[in a new window]
  		Table 1. Comparison of floral characters of Paradinandra suecica with extant taxa.a

 
Within Ternstroemiaceae two main subgroups are recognized (Melchior, 1925 ; Keng, 1962 ). They have been variously named, but, to avoid confusion, we simply refer to them here as the Ternstroemia group and the Adinandra group, respectively, based on their name-giving genus. Melchior (1925) separated the two groups based mainly on the arrangement of the perianth organs. In the Ternstroemia group sepals and petals are opposite, whereas in the Adinandra group they alternate. According to Keng (1962) species of the Ternstroemia group have filaments that are often much shorter than the anthers and have only two to few ovules per carpel, whereas in the Adinandra group filaments are often longer than anthers and there are numerous ovules per carpel. The fossil flower with its alternating sepals and petals, filaments longer than anthers, and numerous ovules per carpel, fits well in the Adinandra group (Table 1).

Following Keng (1962) , the Adinandra group consists of ten genera, of which four are monotypic. Among these is the genus Visnea, which also has epipetalous stamen pairs (Payer, 1857 ; Corner, 1946 ; Ronse Decraene and Smets, 1996 ) in a pattern similar to that of the fossil flowers. The genus Cleyera differs from other taxa of this group in having clearly parietal placentation (Keng, 1962 ) and anthers, which open by restricted apical slits (Melchior, 1925 ). These characters are also present in the fossil flowers.

There are thus many important morphological similarities between the flowers of the Ternstroemiaceae (especially the Adinandra group) and the fossil flowers. However, there are also important differences: (1) The filaments of the fossil flowers are several times longer than the anthers, which contrasts with the condition in most Ternstroemiaceae, where filaments are only slightly longer or shorter than the anthers (Walter Judd, personal communication, University of Florida, Gainsville, Florida, USA). Regarding this feature the fossils are more similar to Actinidiaceae or Theaceae. (2) The pollen observed in the fossils is probably tricolpate, not tricolporate, as is the general condition in the Actinidiaceae, Theaceae, and Ternstroemiaceae (Erdtman, 1952 ; Cronquist, 1981 ). Ferguson and Davison (1986) reported tricolpate pollen in Actinidia deliciosa and Schmid (1978) described the pollen of Actinidia chinensis as tricolpate or possibly tricolporate with obscure germinal pores. Most other Ericales s.l. have tricolporate pollen, but there are a few families where tricolpate grains occur: Balsaminaceae, Lecythidaceae, and Primulaceae (Erdtman, 1952 ). (3) The presence of a disc-shaped, intrastaminal nectary distinguishes the fossil flowers from Ternstroemiaceae where nectaries have not been reported. In the Theaceae, either the base of the filaments or the base of the ovary is rarely nectariferous (Judd et al., 1999 ). Within Actinidiaceae, the corolla base of Saurauia subspinosa is nectariferous (Brown, 1935 ). In many other families of the Ericales s.l. intrastaminal nectary discs are common. The absence of a nectary in most extant Actinidiaceae, Theaceae, and Ternstroemiaceae may be a secondary feature induced by the evolution of their prevailingly polystaminate pollen flowers, in which pollen, and not nectar, is the reward for the pollinators. Vogel (1978) mentioned Theaceae as one of the families that still exhibits nectarogenous transitions to pollen flowers. (4) The salverform shape of the fossil flowers as observed in the abscised corolla tubes is another feature, which distinguishes the fossils from Actinidiaceae, Theaceae, and Ternstroemiaceae. The flowers of these families are often bowl shaped. Sepals, petals, and stamens are spreading at anthesis, and the floral center is open. In the fossil flowers access to the floral center (and the nectary) is canalized by the narrowed tubular part of the corolla and the thick filament bases. In the genus Ternstroemia there are at least two species with conical corolla tubes, which are correlated with buzz pollination (Bittrich, Amaral, and Melo, 1993 ). Flowers with corolla tubes and canalized access to the floral center evolved also in other families of the Ericales s.l., e.g., Ericaceae, Polemoniaceae, and Primulaceae.

The incomplete understanding of corolla and androecium structure in the fossil flowers also poses some problems for evaluations of its systematic affinities. If correctly interpreted, the petals are united at their very base in pre-anthetic stages. The abscised corollas are clearly tubular, and it is possible that the united part of the corolla is longer in anthetic flowers. It is, however, difficult to establish even from transverse sections whether the petals are really united or just closely adhering to each other. It is also not clear whether the filament bases are free or fused to the corolla. Interestingly, both features are variable in extant members of the Ternstroemiaceae, Theaceae, and Actinidiaceae (Table 1). In the general systematic literature (e.g., Cronquist, 1981 ; Takhtajan, 1997 ), the petals of these families are described either as distinct or shortly connate, and the stamens are often described as free or adnate to the base of the petals. In his treatment of Theaceae (including Ternstroemiaceae), Melchior (1925) reported that petals are rarely totally free from each other. In some genera (e.g., Pyrenaria and Anneslea of Theaceae; Ternstroemia and Adinandra of Ternstroemiaceae) petals appear to be constantly united to different degrees, while other genera apparently also include species with free petals (e.g., Stewartia and Gordonia of Theaceae; Eurya of Ternstroemiaceae). In Actinidiaceae the petals are free or basally united. In Actinidia, as well as in Saurauia, both conditions seem to be present (Gilg and Werdermann, 1925 ; Dickison, 1972 ; Takhtajan, 1997 ). There is also considerable variation in the degree of fusion of the stamens with the corolla in the three families. In some genera the filaments are constantly free from the corolla (e.g., Anneslea, Actinidia), in others they are fused to the corolla to various degrees (e.g., Adinandra), while in some genera both conditions are present (e.g., Stewartia, Saurauia).

Another feature of the fossil flowers that is difficult to interpret is the type of anther dehiscence, which we infer to be by short, apical slits. Similar dehiscence is found in Cleyera of the Adinandra group (Melchior, 1925 ). Ternstroemia crassifolia also has similar poricidal anthers (Ramirez et al., 1990 ), while all other Ternstroemiaceae and Theaceae apparently have anthers dehiscing by longitudinal slits. The anthers of Actinidiaceae also open by pores or short slits. Cronquist (1981) states that the openings are morphologically basal and only seemingly apical due to inversion of the anthers during floral ontogeny as is the case in all core Ericales (Anderberg, 1992 ; Judd et al., 1999 ).

The three cavities in the apical part of the ovary and the short, united part of the styles have not been described for any extant member of the Ericales s.l. Some Actinidiaceae have ovaries with an apical depression from which the styles arise. Without detailed morphological and ontogenetical studies it is, however, impossible to conclude whether these structures are homologous.

Pollination
Clear distinction of sepals and petals, possible dehiscence of anthers by restricted slits, presence of a nectary, and the general floral construction (salverform corolla) with a canalized access to the floral center, clearly indicate that the fossil flowers were insect pollinated. A salverform corolla is especially prominent among flowers pollinated by Lepidoptera, but there are also many salverform flowers that are bee pollinated (Endress, 1994 ). Both Lepidoptera and Hymenoptera were probably already present during the Late Cretaceous (for overviews see Crepet, Friis, and Nixon, 1991 ; Grimaldi, 1999 ). The small size of the flowers, their relatively short floral tubes, and the presence of a nectary might indicate pollination by nectar-foraging butterflies or smaller bees rather than by larger bees or Lepidoptera with a longer proboscis such as hawk moths.

Extant members of Actinidiaceae, Theaceae, and Ternstroemiaceae are mainly pollinated by pollen-foraging bees, but thrips and flies have also been reported as pollinators (Knuth, Appel, and Loew, 1904 ). Different species are adapted to buzz pollination by bees (Bittrich, Amaral, and Melo, 1993 ; Proctor, Yeo, and Lack, 1996 ). Buzz pollination is often correlated with anthers opening by apical pores or restricted slits and usually rather small pollen with a smooth exine (Buchmann, 1983 ; Endress, 1994 ). The pollen of the fossil flowers is very small, and the exine is only slightly foveolate. Further, the anthers probably opened by restricted apical slits. The presence of a distinct nectary and the salverform shape of the flowers, however, do not suggest buzz pollination for the fossil flowers. Floral nectar is almost never produced in buzz-pollinated flowers, although some exceptions have been reported in the Ericaceae (e.g., Arctostaphylos and Vaccinium; Buchmann, 1983 ).

Final conclusions on specific pollinators or pollination mechanism in the fossil flowers cannot be drawn based on the information currently available.

All characters taken together, the flowers from the Late Cretaceous described here most closely resemble those of the Adinandra group in Ternstroemiaceae. However, the fossil species do not belong to any of the extant genera described for this family. It probably represents a separate, now extinct phylogenetic lineage within the Ericales s.l., most likely among Ternstroemiaceae, Theaceae, and basal core Ericales.

Fossil record of Ericales s.l
Different lineages of Ericales s.l. were clearly established during the Late Cretaceous and so far four distinct ericalean genera, including the new taxon in this study, have been formally described from the Late Cretaceous (Table 2). In addition to these taxa, which represent perhaps four different families in the Ericales s.l., a number of unnamed, putatively ericalean taxa have been reported from the Late Cretaceous. They were mainly found in mesofossil floras from Europe and eastern North America. Recently possible ericalean flowers have also been reported from coniacian deposits in Japan, also indicating a wide geographical distribution of ericalean plants in the Cretaceous.


View this table:
[in this window]
[in a new window]
  		Table 2. Comparison of age, systematic affinity, and floral characters of Cretaceous taxa referred to Ericales.a

 
The general organization of the flowers of most Cretaceous taxa referred to the Ericales follows a common pattern: the flowers are small, bisexual, actinomorphic, and hypogynous; the pentamerous perianth consists of sepals and petals; the androecium consists of five to 20 (often multiples of five) stamens (eight in Palaeoenkianthus sayrevillensis), which are arranged in whorls; and the syncarpous gynoecium is often trimerous (tetramerous in P. sayrevillensis) and contains numerous ovules. However, beyond this general floral "bauplan" morphology and structure of the floral organs are extremely diverse in fossil ericalean taxa (Table 2) and probably were paralleled by a broad diversity of pollinators and pollination mechanisms.

The flowers of Actinocalyx bohrii (Friis, 1985a ) were found at the same locality in southern Sweden as the fossils described here. Actinocalyx shares many similarities with extant Diapensiaceae and is in some respects also similar to the flowers described here (Table 2). Its flowers were probably insect pollinated as is indicated by their sympetaly and the canalized access to the floral center. From the same locality a few other so far unnamed flowers, fruits, and seeds are thought to belong also to the Ericales s.l., i.e., to Ternstroemiaceae or Theaceae (Friis, 1984 ).

Close in age to the fossils from the Åsen locality is Parasaurauia allonensis and a few other flowers and fruits with possible ericalean affinity (Keller, Herendeen, and Crane, 1996 ; Herendeen et al., 1999 ) from the late Santonian Allon locality of central Georgia in North America. The floral characters of Parasaurauia allonensis (Table 2) are most similar to those of extant Actinidiaceae. Within this family the flowers of P. allonensis are most closely comparable to those of Saurauia, from which they differ mainly in the number of stamens (ten in the fossil vs. 15 to numerous in Saurauia).

A variety of fossil taxa are reported from the Late Cretaceous (Turonian) of New Jersey (Crepet, 1996 ; Crepet and Nixon, 1996 ; Weeks, Nixon, and Crepet, 1996 ). Preliminary estimates suggest that there are at least 15 species and several genera with affinities to modern Ericales from that locality (Crepet, 1996 ). So far only one species has been named and described in detail: Palaeoenkianthus sayrevillensis, a species compared to basal Ericaceae, probably near extant Enkianthus (Nixon and Crepet, 1993 ). The floral features of P. sayrevillensis (Table 2), including sympetaly, awned anthers, and probably viscin threads, indicate a complex insect-pollination mechanism.

Knobloch and Mai (1986) described a variety of fossils from the Late Cretaceous of central Europe. Among others, fruits and/or seeds possibly related to Ternstroemiaceae (e.g., Eurya, Visnea), Actinidiaceae (Saurauia) and Ericaceae (e.g., Leucothoe) are present. However, as most of these fossils are poorly preserved and hardly any floral characters are present, they are difficult to compare with other fossils and flowers of extant plants. Their systematic affinities therefore remain to be established with certainty.

Recently flowers with capsular fruits and ribbed seeds, which possibly belong to the Ericales s.l., have been reported from the Coniacian of northeast Japan (Takahashi, Crane, and Ando, 1999b ). So far no detailed description or comparison with extant plants has been provided and their systematic affinity remains uncertain.

The obvious relative abundance of reported fossils from Late Cretaceous floras raises the question whether ericalean species were relatively more abundant during that time than they are today. This may well be the case. However, currently only a small percentage of the whole angiosperm diversity occurring in Late Cretaceous assemblages has been described. Therefore, information on the systematic composition of Late Cretaceous floras and the relative abundance of specific groups remains highly inconclusive.

The morphological and systematical diversity, which is present in the ericalean fossil record from the Late Cretaceous and from the Tertiary, where the record of Ericales s.l. is extensive (e.g., Collinson, 1978 ; Friis, 1985b ; Manchester, 1994 ), indicates that Ericales s.l. must have undergone a considerable differentiation already during the Cretaceous. Combined with the modern conception of asterid affinity (e.g., APG, 1998 ), the fossil record of Ericales s.l. implies that at least some basal groups of the asterids originated early in the diversification of the higher eudicots.


   FORMAL DESCRIPTION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 FORMAL DESCRIPTION
LITERATURE CITED
 
Order
Ericales s.l.

Family
incertae sedis

Paradinandra, gen. nov.

The name Paradinandra refers to the similarities of the fossil flowers with those of different species of the Adinandra group [Adinandrieae sensu Keng (1962) ].

Generic diagnosis
Flower small, in axil of subtending bract, short pedunculate, bibracteolate, actinomorphic, hypogynous, pentamerous; calyx probably initiated in a spiral, all other floral parts whorled, whorls alternating; calyx of five distinct and imbricate sepals; corolla of five basally united (?) petals, imbricate; androecium of 15 stamens, diplostemon, inner whorl with stamen pairs, filaments basally united and fused to the corolla (?); anthers basifixed, x-shaped, dithecal, tetrasporangiate, probably opening by restricted apical slits; pollen small, tricolpate (?) with long colpi, exine finely foveolate; intrastaminal nectary disc; gynoecium trimerous, syncarpous, superior; with three almost completely free elongate styles; ovary unilocular, placentation parietal (ovary with incomplete septa); ovules numerous, campylotropous, surface reticulate.

Comments on the new genus
Mature fruits and seeds as well as most vegetative parts are unknown.

Type species
Paradinandra suecica, spec. nov.

The specific epithet refers to Sweden where the fossil material was collected.

Specific diagnosis
As for the genus with the following additions: small scales present in the axils of the bracteoles, sepals broadly lanceolate, aestivation quincuncial, abaxial epidermis slightly papillate, with a prominent middle vein and simple trichomes abaxially, margins membraneous; petals mitriform, aestivation quincuncial; anthers stacked on three levels in the bud; filaments massive basally and tapering apically; thecae with enlarged epidermal cells along the stomium; styles with ventral slits; ovary with simple trichomes in older stages.

Dimensions
Measurements given for sectioned bud (holotype, S106194-01): peduncle ~0.6 mm long; flower ~3.5 mm long and ~1.2 mm wide (bud, including bracteoles); sepals 2.7–3 x 1.2–1.5 mm; petals 1.8–2 x 1.2–1.3 mm; filaments 0.45–0.75 mm long; anthers ~0.27 x 0.16 mm; pollen 10.5–14 µm in polar length; ovary ~0.4 mm in diameter; styles ~0.65 mm long.

Holotype
S106194-01 (sample Åsen 2), microtome sections S106194-02–53, Figs. 2, 3, 7, 15–17, 36–38.

Type locality
Höganäs AB quarry at Åsen near Axeltorp, Scania, Sweden.

Type stratum
Fluviatile-lacustrine sequence.

Age
Late Cretaceous (Late Santonian-Early Campanian).


   FOOTNOTES
 
1 The authors thank Stig Lindbom for collecting the fossil material; Anton Igersheim and Yvonne Arremo for technical assistance; Arne Anderberg, Alex Bernhard, Peter Crane, Michael Donoghue, Walter Judd, Peter K. Endress, Maria von Balthazar, and an anonymous reviewer for valuable comments on the manuscript. This work was supported by the Swedish Natural Science Research Council. Back

2 Author for reprint requests, current address: Institute of Systematic Botany, University of Zurich, Zollikerstrasse 107, CH-8008 Zurich, Switzerland (e-mail: jsberger{at}systbot.unizh.ch ) Back


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 FORMAL DESCRIPTION
 LITERATURE CITED
 
Anderberg, A. A. 1992 The circumscription of the Ericales, and their cladistic relationships to other families of "higher" dicotyledons. Systematic Botany 17: 660–675[CrossRef][ISI]

———. 1993 Cladistic interrelationships and major clades of the Ericales. Plant Systematics and Evolution 184: 207–231[CrossRef][ISI]

APG (The Angiosperm Phylogeny Group). 1998 An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85: 531–553[CrossRef][ISI]

Basinger, J. F., and D. L. Dilcher. 1984 Ancient bisexual flowers. Science 224: 511–513[Abstract/Free Full Text]

Bittrich, V., C. E. Amaral, and G. A. R. Melo. 1993 Pollination biology of Ternstroemia laevigata and T. dentata (Theaceae). Plant Systematics and Evolution 185: 1–6

Bouman, F., and F. D. Boesewinkel. 1991 The campylotropous ovules and seeds, their structure and functions. Botanische Jahrbücher für Systematik 113: 255–270

Brown, E. G. S. 1935 The floral mechanism of Saurauia subspinosa Anth. Transactions and Proceedings of the Botanical Society of Edinburgh 31: 485–497

Buchmann, S. L. 1983 Buzz pollination in angiosperms. In C. E. Jones and R. J. Little [eds.], Handbook of Experimental Pollination Biology, 73– 113. Scientific and Academic Editions, New York, New York, USA

Chase, M. W., et al. 1993 Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528–580[CrossRef][ISI]

Collinson, M. E. 1978 Rhododendron seeds from the Palaeocene of southern England. Botanical Journal of the Linnean Society 76: 195–205

Corner, E. J. H. 1946 Centrifugal stamens. Journal of the Arnold Arboretum 27: 423–437

Crane, P. R., E. M. Friis, and K. R. Pedersen. 1989 Reproductive structure and function in Cretaceous Chloranthaceae. Plant Systematics and Evolution 165: 211–226[CrossRef][ISI]

———, ———, and ———. 1995 The origin and early diversification of angiosperms. Nature 374: 27–33[CrossRef]

Crepet, W. L. 1996 Timing in the evolution of derived floral characters: Upper Cretaceous (Turonian) taxa with tricolpate and tricolpate-derived pollen. Review of Palaeobotany and Palynology 90: 339–359

Crepet, W. L., E. M. Friis, and K. C. Nixon. 1991 Fossil evidence for the evolution of biotic pollination. Philosophical Transactions of the Royal Society of London B 333: 187–195[CrossRef]

———, and K. C. Nixon. 1996 The fossil history of stamens. In W. G. D'Arcy and R. C. Keating [eds.], The anther—form, function and phylogeny, 25–57. Cambridge University Press, Cambridge, UK

Cronquist, A. 1981 An integrated system of classification of flowering plants. Columbia University Press, New York, New York, USA

Dickison, W. C. 1972 Observations on the floral morphology of some species of Saurauia, Actinidia, and Clematoclethra. Journal of the Elisha Mitchell Scientific Society 88: 43–54

Downie, S. R., and J. D. Palmer. 1992 Restriction site mapping of the chloroplast DNA inverted repeat: a molecular phylogeny of the Asteridae. Annals of the Missouri Botanical Garden 79: 266–283[CrossRef][ISI]

Eklund, H., E. M. Friis, and K. R. Pedersen. 1997 Chloranthaceous floral structures from the Late Cretaceous of Sweden. Plant Systematics and Evolution 207: 13–42[CrossRef][ISI]

———, and J. Kvacek. 1998 Lauraceous inflorescences and flowers from the Cenomanian of Bohemia (Czech Republic, Central Europe). International Journal of Plant Sciences 159: 668–686[CrossRef]

Endress, P. K. 1994 Diversity and evolutionary biology of tropical flowers. Cambridge University Press, Cambridge, UK

———, and E. M. Friis. 1991 Archamamelis, hamamelidalean flowers from the Upper Cretaceous of Sweden. Plant Systematics and Evolution 175: 101–114[CrossRef][ISI]

Erdtman, G. 1952 Pollen morphology and plant taxonomy. Angiosperms. Almqvist & Wiksell, Stockholm, Sweden

Ferguson, A. R., and R. M. Davison. 1986 Actinidia deliciosa. In A. H. Halevy [ed.], Handbook of flowering, vol. 5, 1–14. CRC Press, Boca Raton, Florida, USA

Friis, E. M. 1983 Upper Cretaceous (Senonian) floral structures of juglandalean affinity containing Normapolles pollen. Review of Palaeobotany and Palynology 39: 161–188[CrossRef][ISI]

———. 1984 Preliminary report on Upper Cretaceous angiosperm reproductive organs from Sweden and their level of organization. Annals of the Missouri Botanical Garden 71: 403–418[CrossRef][ISI]

———. 1985a Actinocalyx gen.nov., sympetalous angiosperm flowers from the Upper Cretaceous of southern Sweden. Review of Palaeobotany and Palynology 45: 171–183

———. 1985b Angiosperm fruits and seeds from the Middle Miocene of Jutland (Denmark). Biologiske Skrifter, Det Kongelige Danske Videnskabernes Selskab 24: 1–165

———. 1985c Structure and function in Late Cretaceous angiosperm flowers. Biologiske Skrifter, Det Kongelige Danske Videnskabernes Selskab 25: 1–37

———. 1990 Silvianthemum suecicum gen. et sp. nov., a new saxifragalean flower from the Late Cretaceous of Sweden. Biologiske Skrifter, Det Kongelige Danske Videnskabernes Selskab 36: 1–35

———, P. R. Crane, and K. R. Pedersen. 1988 Reproductive structures of Cretaceous Platanaceae. Biologiske Skrifter, Det Kongelige Danske Videnskabernes Selskab 31: 1–55

———, K. R. Pedersen, and P. R. Crane. 1994 Angiosperm floral structures from the Early Cretaceous of Portugal. Plant Systematics and Evolution 8 (Supplement): 31–49

———, and A. Skarby. 1981 Structurally preserved angiosperm flowers from the Upper Cretaceous of southern Sweden. Nature 291: 485–486

———, and ———. 1982 Scandianthus gen. nov., angiosperm flowers of saxifragalean affinity from the Upper Cretaceous of southern Sweden. Annals of Botany, New Series 50: 569–583

Frumin, S., and E. M. Friis. 1999 Magnoliid reproductive organs from the Cenomanian-Turonian of north-western Kazakhstan: Magnoliaceae and Illiciaceae. Plant Systematics and Evolution 216: 265–288[CrossRef][ISI]

Gilg, E., and E. Werdermann. 1925 Actinidiaceae. In A. Engler [ed.], Die natürlichen Pflanzenfamilien, vol. 21, 36–47. W. Engelmann, Leipzig, Germany

Grimaldi, D. 1999 The co-radiations of pollinating insects and angiosperms in the Cretaceous. Annals of the Missouri Botanical Garden 86: 373– 406[CrossRef][ISI]

Herendeen, P. S., S. Magallón-Puebla, R. Lupia, P. R. Crane, and J. Kobylinska. 1999 A preliminary conspectus of the Allon flora from the Late Cretaceous (Late Santonian) of central Georgia, U.S.A. Annals of the Missouri Botanical Garden 86: 407–471[CrossRef][ISI]

Hufford, L. 1992 Rosidae and their relationships to other monmagnoliid dicotyledons: a phylogenetic analysis using morphological and chemical data. Annals of the Missouri Botanical Garden 79: 218–248[CrossRef][ISI]

Igersheim, A. 1993 The character states of the Caribbean monotypic endemic Strumpfia (Rubiaceae). Nordic Journal of Botany 13: 545–559[ISI]

———, and O. Cichocki. 1996 A simple method for microtome sectioning of prehistoric charcoal specimens, embedded in 2-hydroxyethyl methacrylate (HEMA). Review of Palaeobotany and Palynology 92: 389–393[CrossRef]

Judd, W. S., C. S. Campbell, E. A. Kellogg, and P. F. Stevens. 1999 Plant systematics: a phylogenetic approach. Sinauer, Sunderland, Massachusetts, USA

———, and K. A. Kron. 1993 Circumscription of Ericaceae (Ericales) as determined by preliminary cladistic analyses based on morphological, anatomical, and embryological features. Brittonia 45: 99–114

Keller, J. A., P. S. Herendeen, and P. R. Crane. 1996 Fossil flowers and fruits of the Actinidiaceae from the Campanian (Late Cretaceous) of Georgia. American Journal of Botany 83: 528–541[CrossRef][ISI]

Keng, H. 1962 Comparative morphological studies in Theaceae. University of California Publications in Botany 33: 269–384

Knobloch, E., and D. H. Mai. 1986 Monographie der Früchte und Samen in der Kreide von Mitteleuropa. Rozpravy ústredního ústavu geologické ho, Praha 47: 1–219

Knuth, P., O. Appel, and E. Loew. 1904 Handbuch der Blütenbiologie. W. Engelmann, Leipzig, Germany

Kron, K. 1996 Phylogenetic relationships of Empetraceae, Epacridaceae, Ericaceae, Monotropaceae, and Pyrolaceae: evidence from nuclear ribosomal 18s sequence data. Annals of Botany 77: 293–303[Abstract/Free Full Text]

———, M. W. Chase, and H. G. Hills. 1991 Phylogenetic relationships of Ericaceae, Empetraceae, and Epacridaceae based on rbcL sequence data and their position within the Dilleniidae. American Journal of Botany 78: 197–198 (Abstract)

Lechner, S. 1915 Anatomische Untersuchungen über die Gattungen Actinidia, Saurauia, Clethra und Clematoclethra mit besonderer Berückstchtigung ihrer Stellung im System. Beihefte zum Botanischen Centralblatt 32: 431–467

Manchester, S. R. 1994 Fruits and seeds of the Middle Eocene Nut Beds Flora, Clarno Formation, Oregon. Palaeontographica Americana 58: 1– 205

Melchior, H. 1925 Theaceae. In A. Engler [ed.], Die natürlichen Pflanzenfamilien, vol. 21, 109–154. W. Engelmann, Leipzig, Germany

Morton, C. M., M. W. Chase, K. A. Kron, and S. M. Swensen. 1996 A molecular evaluation of the monophyly of the order Ebenales based upon rbcL sequence data. Systematic Botany 21: 567–586[CrossRef][ISI]

———, S. A. Mori, G. T. Prance, K. G. Karol, and M. W. Chase. 1997 Phylogenetic relationships of Lecythidaceae: a cladistic analysis using rbcL and morphological data. American Journal of Botany 84: 530–540[Abstract]

Nandi, O. I., M. W. Chase, and P. K. Endress. 1998 A combined cladistic analysis of angiosperms using rbcL and non-molecular data sets. Annals of the Missouri Botanical Garden 85: 137–212

Nixon, K. C., and W. L. Crepet. 1993 Late Cretaceous fossil flowers of ericalean affinity. American Journal of Botany 80: 616–623[CrossRef][ISI]

Olmstead, R. G., B. Bremer, K. M. Scott, and J. D. Palmer. 1993 A parsimony analysis of the Asteridae sensu lato based on rbcL sequences. Annals of the Missouri Botanical Garden 80: 700–722[CrossRef][ISI]

———, H. J. Michaels, K. M. Scott, and J. D. Palmer. 1992 Monophyly of the Asteridae and identification of their major lineages inferred from DNA sequences of rbcL. Annals of the Missouri Botanical Garden 79: 249–265

Payer, J.-B. 1857 Traité d'organogénie comparée de la fleur. Masson, Paris, France

Proctor, M., P. Yeo, and A. Lack. 1996 The natural history of pollination. Harper Collins, London, UK

Ramirez, N., C. Gil, O. Hokche, A. Seres, and Y. Brito. 1990 Biologia floral de una comunidad arbustica tropical en la Guyana Venezolana. Annual of the Missouri Botanical Garden 77: 383–397

Ronse Decraene, L. P., and E. F. Smets. 1996 The morphological variation and systematic value of stamen pairs in the Magnoliatae. Feddes Repertorium 107: 1–17

Schmid, R. 1978 Reproductive anatomy of Actinidia chinensis (Actinidiaceae). Botanische Jahrbücher für Systematik 100: 149–195

Soltis, D. E., et al. 1997 Angiosperm phylogeny inferred from 18S ribosomal DNA sequences. Annals of the Missouri Botanical Garden 84: 1–49

Swensen, S. M., and M. W. Chase. 1995 Relationships of Theales sensu Cronquist from the perspective of cladistic analysis of rbcL sequence data. American Journal of Botany 82: 166 (Abstract)

Takahashi, M., P. R. Crane, and H. Ando. 1999a Esqueiria futabensis sp. nov.; a new angiosperm flower from the Upper Cretaceous (lower Coniacian) of northeastern Honshu, Japan. Paleontological Research 3: 81–87

———, ———, and ———. 1999b Fossil flowers and associated plant fossils from the Kamikitaba locality (Ashizawa Formation, Futaba Group, lower Coniacian, Upper Cretaceous) of Northeast Japan. Journal of Plant Research 112: 187–206[CrossRef][ISI]

Takhtajan, A. 1997 Diversity and classification of flowering plants. Columbia University Press, New York

Tsou, C.-H. 1995 Embryology of Theaceae—anther and ovule development of Adinandra, Cleyera and Eurya. Journal of Plant Research 108: 77– 86[CrossRef][ISI]

Vogel, S. 1978 Evolutionary shifts from reward to deception in pollen flowers. In A. J. Richards [ed.], The pollination of flowers by insects, 89–96. Academic Press, London, UK

Weeks, A., K. C. Nixon, and W. L. Crepet. 1996 A new dilleniid fossil flower from the Late Cretaceous of New Jersey. American Journal of Botany 83: 113 (Abstract)




This article has been cited by other articles:


Home page