Journal of Integrative Plant Biology 2009
The Earliest Normal Flower from Liaoning
Province, China
∗
Xin Wang1,2 and Shaolin Zheng3
(1 Nanjing Institute of Geology and Palaeontology, the Chinese Academy of Sciences, Nanjing 210008, China;
2
3
Fairylake Botanical Garden, Shenzhen 518004, China;
Shenyang Institute of Geology and Mineral Resources, Shenyang 110032, China)
Abstract
The early evolution of angiosperms has been a focus of intensive research for more than a century. The Yixian Formation in
western Liaoning yields one of the earliest angiosperm macrofloras. Despite multitudes of angiosperm fossils uncovered,
including Archaefructus and Sinocarpus, no bona fide normal flower has been dated to 125 Ma (mega-annum) or older. Here
we report Callianthus dilae gen. et sp. nov. from the Yixian Formation (Early Cretaceous) in western Liaoning, China as the
earliest normal flower known to date. The flower demonstrates a typical floral organization, including tepals, androecium,
and gynoecium. The tepals are spatulate with parallel veins. The stamens have a slender filament, a globular anther,
bristles at the anther apex, and in situ round-triangular pollen grains. The gynoecium is composed of two stylate carpels
enclosed in a fleshy envelope, and develops into a “hip” when mature. Since the well-accepted history of angiosperms
is not much longer than 125 Ma, Callianthus together with Chaoyangia, Archaefructus and Sinocarpus from the Yixian
Formation demonstrate a surprisingly high diversity of angiosperms, implying a history of angiosperms much longer than
currently accepted.
Key words: angiosperm; bisexual; Callianthus; Early Cretaceous; flower; Liaoning; perianth; Yixian Formation.
Wang X, Zheng S (2009). The earliest normal flower from Liaoning Province, China. J. Integr. Plant Biol. doi: 10.1111/j.1744-7909.2009.00838.x
Available online at www.jipb.net
The discovery of Archaefructus Sun, Dilcher, Zheng et Zhou
(Sun et al. 1998, 2001, 2002; Ji et al. 2004) has made
the Yixian Formation (125 Ma ago, Early Cretaceous) well
known worldwide. Associated with Archaefructus are more
angiosperms: Sinocarpus Leng et Friis (Leng and Friis 2003,
2006; or Hyrcantha Krassilov et Vachrameev, Dilcher et al.
2007) and Chaoyangia Duan (Duan 1998), and more animals
(e.g. Wang et al. 2000b; Li and Luo 2006) from the Formation.
Received 2 Dec. 2008 Accepted 8 Mar. 2009
Supported by the National Natural Science Foundation of China (40772006,
40372008, 40632010 and J0630967), Resources Platform of Mineral Rock
However, a bisexual flower with a perianth, typical of most
angiosperms, has never before been found. Here we report
a normal flower, Callianthus dilae gen. et sp. nov., showing
gynoecium, androecium, and tepals from the Yixian Formation.
The tepals are spatulate and parallel veined, with a long claw
and a round tip. The stamens have a filament and a globular
anther, with bristles at the anther apex. Round-triangular pollen
grains are found in the anther. Two stylate carpels are enclosed
in a fleshy envelope, forming a “hip” when mature. Callianthus
and other early angiosperms documented in the Yixian Formation demonstrate a diversity of angiosperms unexpectedly high
for their initiating stage and help to further elucidate the early
radiation of angiosperms.
and Fossil, and the Scientific Research Foundation for the Returned Overseas
Chinese Scholars, State Education Ministry. This paper is a contribution to
IGCP 506.
∗
Results
Author for correspondence.
Tel: +86 26 8328 2266;
Fax: +86 26 8328 2140;
Callianthus Wang et Zheng gen. nov.
E-mail: <[email protected]>.
Type species: Callianthus dilae Wang et Zheng gen. et sp. nov.
Diagnosis: Flower small, bisexual, with a perianth, hypogynous, with a slender pedicel. Tepals in two cycles, spatulate,
C 2009 Institute of Botany, the Chinese Academy of Sciences
doi: 10.1111/j.1744-7909.2009.00838.x
2 Journal of Integrative Plant Biology
2009
parallel veined, with a long claw and a round tip. Stamen
composed of a filament and a globular anther, with numerous
bristles at the apex. In situ pollen grains round-triangular.
Fleshy envelope enclosing two separate carpels. Each carpel
composed of a hemi-globular ovary and a papillate style.
Fructification hip-like, including two facing fruits, with persistent
styles.
Etymology: Calli- for beautiful in Greek, -anthus for flower in
Greek.
Remark: There are three fossil taxa, including Spanomera
Drinnan, Crane, Friis et Pedersen (Drinnan et al. 1991), Lusicarpus Pedersen, Balthazar, Crane et Friis (Pedersen et al. 2007)
and Erenia Krassilov (Krassilov 1982), sharing some similarity
with Callianthus. However, the following comparisons justify the
new genus Callianthus.
Spanomera is an inflorescence of unisexual flowers, related to
Buxaceae, from the mid-Cretaceous of North America (Drinnan
et al. 1991). The gynoecium is bicarpellate, like Callianthus.
However, its unisexuality, lack of evident style, lack of fleshy
envelope surrounding the carpels, and lack of spatulate tepals
distinguish it from Callianthus.
Lusicarpus is a pistillate bicarpellate flower, related to Buxaceae, from the Early Cretaceous of Portugal (Pedersen et al.
2007). Like Callianthus, its gynoecium includes two carpels.
However, its stout style, striate tricolpate pollen grains on the
stigma, lack of fleshy envelope surrounding the carpels, lack
of stamen, and lack of spatulate tepals distinguish it from
Callianthus.
Erenia stenoptera Krassilov is a fossil taxon established
based on material from the Early Cretaceous in Mongolia
(Krassilov 1982). It is a small (2 × 2 mm), stalked, winged fruit
with bilocular endocarp and sessile stigma (Krassilov 1982).
Although its smooth membranous wing and elliptical bilocular
endocarp might be said to be more or less comparable to the
fleshy envelope and two fruits in the “hip” of Callianthus, its
characteristic “funnel-shaped, sessile” stigma contrasts strongly
with the divergent papillate styles (stigma) in Callianthus, and
this character alone is enough to distinguish Erenia from Callianthus. Furthermore, Callianthus is distinguished from Erenia
in its larger size, distinct stamens and tepals, and lack of “stalkavoiding” wing. The consistent differences between Callianthus
(one complete flower and six specimens examined) and Erenia
suggest that they are distinct fossil plants.
The “hip” will be used to describe the fructification hereafter;
however, this term is used in a slightly altered sense: a typical hip
is derived from carpels enclosed by a hypanthium (Harris and
Harris 2000) rather than by a fleshy envelope as in Callianthus.
Callianthus dilae Wang et Zheng gen. et sp. nov.
Diagnosis: Currently the same as that of the genus.
Description: Holotype: Flower small, bisexual, with a perianth, hypogynous, pedicellate, 6.9 mm high, 7.3 mm wide
(Figure 1A,B). Pedicel up to 1.8 mm long and 0.35 mm wide
(Figure 1A,B). Four tepals and two stamens are seen attached
to the pedicel (Figure 1A,B,I). Tepals are distinct, spatulate,
with a long claw and a round tip, up to 6.5 mm long and 0.9 mm
wide in the distal portion, in two cycles (Figure 1A,B,D–E,I).
Each tepal has two major parallel veins in the distal portion
(Figures 1D, 2A). A stoma is seen on a tepal, with stomatal
aperture 1–2 × 7–8 μm (Figure 2C,F). Stamens are attached
just above the inner cycle of the tepals by a thin filament, which
is about 1.2 mm long and 0.19 mm wide (Figure 1A,B,F,G,I).
The anther is attached to the terminal of the filament, globular,
about 0.5 mm wide, and with numerous bristles up to 0.8 mm
long and 60–65 μm wide at the apex (Figures 1A,B,F,G, 2B).
Pollen grains in situ are compressed into various shapes, but
two of them appear round-triangular, 28–32 μm in diameter
(Figure 3A–G). Similar pollen grains have been seen three
times in the anther region on the transfers (Figure 3F,G). Two
stylate carpels are base-fixed in a cup-shaped fleshy envelope,
which has a rough surface (Figures 1A,B,H,I, 2E). The fleshy
envelope is widest at the middle (about 4.2 mm wide) and about
3.75 mm wide at the top, 0.6–1.6 mm thick, with a 0.4 mm-high
raised ring close to the margin (Figure 1A,B, H). Each carpel
is separated from the adjacent one almost to its bottom by a
gap about 0.3 mm wide (Figure 1A–C). Each carpel includes an
apical style and basal ovary (Figure 1A,B). The ovary is hemiglobular, about 3.1 mm high and 1.4 mm thick (Figure 1A,B).
The style is short, slightly curved, papillate, and more than
1 mm long and about 0.2 mm wide (Figures 1A,B,H, 2D).
The papillae on the style are probably conical-shaped, tapering
distally, at least 5 μm long, covering the whole length of the style
(Figure 2D).
Further specimens: The fructifications look like the gynoecium in morphology and dimensions, with a hip-like organization (Figure 4A,D,H–I). The “hip” includes two fruits and a
fleshy envelope, about 4.0–5.8 mm high and 4.0–5.5 mm wide
(Figure 4A,D,H–I). The fleshy envelope surrounds two facing
fruits, and has a raised ring at the top (Figure 4A,D,H). Each
fruit is hemi-globular, with a persistent style (Figure 4A,D). The
style is always more than 1 mm long and about 0.2 mm wide
(Figure 4A,B,D,F). Stamens and tepals are missing the fructifications (Figure 4A–I). There are traces of vascular bundles
entering the fruits and the envelope, respectively, at the bottom
(Figure 4D,E). Each fruit is 2.9–3.5 mm high, 1.3–1.7 mm thick,
about 3.5 mm wide, separated by a gap in between, probably
with a dorsal ridge (Figure 4A,D,H,I).
Holotype: PB21047a&b.
Further specimen: PB18320, PB21091a&b, PB21092.
Etymology: dila- for the collector of the most complete
specimen (PB21047), Mr. Dila Chen; -e, Latin suffix.
Type locality: Huangbanjigou, Shangyuan, Beipiao, Liaoning, China (41◦ 12 N, 119◦ 22 E).
Stratigraphic horizon: the Yixian Formation, equivalent to
the Barremian, Lower Cretaceous (125 Ma).
Earliest Normal Flower from China
3
4 Journal of Integrative Plant Biology
2009
Depository: The specimens are deposited in the Palaeobotanical Collection, Nanjing Institute of Geology and Palaeontology, the Chinese Academy of Sciences, Nanjing, China.
Remarks: “Erenia stenoptera Krassilov” described by Wu
(Plate XVI, figures 5,5a, 1999; figure 243, 2003) from the
holotype locality of Callianthus is apparently different from the
holotype of Erenia in its lack of a “funnel-shaped, sessile”
stigma, and lack of a “stalk-avoiding” wing. This fossil (shown
in Figure 4A–C) should rather be assigned to C. dilae because
the former shares, besides the same locality, almost exactly
the same morphology with the gynoecium of the latter. This
idea is supported by further specimens with almost identical
morphology (Figure 4D–I). Thus it can be inferred that the “hip”
of C. dilae falls off from the pedicel when mature.
The central dark materials in Figure 1A,B are interpreted
as two carpels because of: (i) the presence of two papillate
styles (unlike the glabrous micropylar tube of Gnetales); (ii)
their hemi-globular forms and central position in the flower;
(iii) the physical connection with stamens and perianth; (iv) the
presence of stamens, which are missing when mature, as in
Figure 4A–I; (v) their positions and morphology corresponding
to two fruits in the fructification (Figure 4A–I); and (vi) two carpels
per “hip” rather than one ovuliferous unit per outer integument in
Gnetales. Degaging fails to reveal the existence of a third style.
The smooth connection (Figures 1H,4A,B,D,F) between the two
styles also implies the nonexistence of a third style.
The distinction between the style and the stigma is not obvious
in Callianthus since the papillae are scattered all over the style,
it is possible that the whole style functions as a stigma.
There is a vertical mark in Figure 4I (upper arrow), which
appears narrower, much more inconspicuous and different from
the gap between carpels/fruits. Since the dark line on the dorsal
of the carpel in Figure 1B (double white arrow) may represent a
dorsal vascular bundle, we interpret the mark in Figure 4I as a
dorsal ridge on the fruit.
The word “tepal” is preferred because no significant differentiation is seen in the perianth, although the tepals look more
like typical petals rather than sepals. Only four tepals are visible
in the specimens (Figure 1A,B), but the actual total number of
tepals may be more. Likewise, the number of stamens may be
more than two.
The dimensions of the papillae on the styles appear to vary
greatly even in the same picture (Figure 2D). A plausible
explanation, which is adopted here, is that the papillae are
conical in form and that Figure 2D demonstrates cross sections
of the papillae of different orientations at different levels.
Comparing Figure 4A,B with Figure 4D,F,I indicates that
the fructifications are preserved in different orientations. In
Figure 4A,B the bedding plane appears parallel to the plane
of the styles, as suggested by the wide space between the two
divergent styles. In Figure 4D,F the fructification appears to be
slightly rotated around its vertical axis since the spaces between
the styles and fruits are highly compressed and the two styles
are on different levels (Figure 4F showing both styles clearly
is a composite image of two originals focusing on different
levels). In Figure 4I the fructification appears to be rotated about
90 degrees from the position in Figure 4A,B since the fructification top has eclipsed the styles almost completely and the figure
shows a ridge on the fruit not seen in Figure 4A,B. The absence
of styles in Figure 4G,I implies that the fructification has a raised
ring at its top that eclipses the style; this interpretation is favored
by raised shoulders seen in Figures 1A,B,4A,B,D. The constant
presence of a fleshy envelope in the flower and all fructifications preserved in various orientations (Figures 1A,B,4A,B,D,F)
←
Figure 1. The flower and some of its details. All stereomicroscope photographs. Specimen number PB21047a for A, C, D and F, PB21047b
for B, E and G-I.
(A) A general view of the flower. Note four tepals (black arrows) and two stamens (white arrows). Bar, 2 mm.
(B) The counterpart of the flower in Figure 1A. Note the styles (black arrows), possible dorsal vascular bundle of the carpel (white double arrow), and
fleshy envelope (white arrow). Bar, 2 mm.
(C) A detailed view of the gap between the carpels in the flower shown in Figure 1A. Note the outlines of the gap (arrows). Bar, 0.2 mm.
(D) The top portion of a tepal in Figure 1A. Note the two parallel veins (white arrows). Bar, 1 mm.
(E) A complete tepal exposed after degaging. Note the spatulate shape, long claw (white arrow), and round tip. Bar, 1 mm.
(F) One of the anthers enlarged from Figure 1A (left white arrow). Note the filament at the base (white double arrow), globular anther (white arrow),
and bristles at the apex (black arrows). Bar, 0.5 mm.
(G) The counterpart of the anther shown in Figure 1F, enlarged from Figure 1B,E. Note the globular anther (white arrow) and bristles at the apex
(black arrows). Bar, 0.5 mm.
(H) The divergent styles. Note the deployment of the styles, and the relationship between the fleshy envelope (black arrow) and the carpels (white
arrow). Bar, 1 mm.
(I) The arrangement of the floral parts. Note the spatial relationship among the pedicel (central bottom), tepals (black arrows) in two cycles, filaments
(white arrows), fleshy envelope (white double arrows), and the vascular bundle to the carpels (black double arrows). Bar, 1 mm.
Earliest Normal Flower from China
5
Figure 2. Some details of the flower. All are scanning electron microscopy (SEM) photographs. Specimen number PB21047b.
(A) Parallel veins in a tepal. The outline of the tepal is marked by the white lines. Bar, 0.1 mm.
(B) Two of the bristles (arrows) on the anther apex. Note the fleshy envelope (fe) eclipsing the bristles. Bar, 10 μm.
(C) Tepal cuticle with a stomatal aperture (white arrow). Bar, 5 μm.
(D) Surface of the style, from a broken cellulose transfer made from PB21047b. Note the slender papillae (black arrow) and the bigger scars left by
papilla bases on the style (white arrow). Bar, 10 μm.
(E) Clusters of trichomes on the surface of the fleshy envelope (white arrow). The outside of the fleshy envelope is toward the top. Bar, 10 μm.
(F) A stoma on a tepal. The black triangle in Figure 1B indicates the approximate position of the stoma on the tepal. Bar, 5 μm.
indicates that the organ is originally radially symmetrical, and
that the fleshy envelope surrounds the carpels.
One recent hot topic generating discussion in angiosperm
systematics centers on Hydatellaceae, in which the reproductive
organ is a highly reduced inflorescence composed of several
highly simplified flowers (Rudall et al. 2007; Saarela et al.
2007). We cannot exclude the possibility of an inflorescence
of a large terminal bicarpellate pistillate flower plus small lateral
staminate flowers in axils of the inner cycle bracts in Callianthus
completely. However, considering the early age of this fossil,
lack of bract just below the gynoecium, and little chance for
the supposed reduction or simplification to occur, we hereafter
ignore this alternative for the time being and hope future
discovery will shed light on this problem.
Discussion
The Yixian Formation in Liaoning, China is well known for its fossils of early angiosperms, including Chaoyangia, Archaefructus
and Sinocarpus (Duan 1998; Sun et al. 1998, 2001, 2002; Leng
and Friis 2003, 2006; Ji et al. 2004). It is rich in fossils of various
animals and plants (Wang et al. 2000a,b, 2003, 2004; Li and Luo
2006; Yao et al. 2006; Liu et al. 2007; among others). Among
the plants, there are Bryophyta, Lycopodiales, Equisetales, Filicales, Pteridospermae, Cycadales, Bennettitales, Ginkgoales,
Czekanowskiales, Coniferales, and Angiospermae (Wu 1999;
Sun et al. 2001; Zheng et al. 2005). In the past there used to
be a controversy over the age of the Yixian Formation (Sun
et al. 1998, 2001, 2002; Swisher et al. 1998; Leng and Friis
2003, 2006; Ji et al. 2004; Friis et al. 2005, 2006), but now there
seems to be, at least for the time being, a general consensus
that it is about 125 Ma old (the Early Cretaceous, Dilcher et al.
2007).
Angiosperms are defined by quite a few characters, including reticulate leaf venation, vessel elements in wood, double
fertilization, enclosed ovules/seeds, double fertilization, pollen
tube, and tectate-columellate pollen wall structure (Eames 1961;
Bierhorst 1971; Hu 1998; Wang et al. 2007a,b). Although
the flower has been a focus for palaeobotanical research for
decades, there is no strict consensus on its definition. Friis
et al. (2006) gave the following description or definition: “The
6 Journal of Integrative Plant Biology
2009
Figure 3. In situ pollen grains from the anther. All are scanning electron microscopy (SEM) photographs. Specimen number PB21047b.
(A) A piece of the sediment removed from the anther. Note that clusters of pollen grains (arrows) are visible on the surface of the sediment even
without chemical processing. The inset shows a fragment obtained by degaging. This SEM image was taken from the rectangle in the inset. Note the
top darker region in the inset is a mark left by the tepal. Bar, 50 μm.
(B) The pollen cluster indicated by the lowest arrow in Figure 3A. Bar, 10 μm.
(C) The pollen grain indicated by the black arrow in Figure 3B. Bar, 10 μm.
(D) The sculpture of the pollen grain, enlarged from the rectangular region in Figure 3C. Bar, 1 μm.
(E) The pollen grain indicated by the white arrow in Figure 3B. Note the round-triangular shape of the pollen grain, and its angles (arrows) and
sculpture. Bar, 10 μm.
(F) A round-triangular pollen grain in the anther, on a transfer. Bar, 10 μm.
(G) The sculpture of the pollen grain shown in Figure 3F. Bar, 1 μm.
Earliest Normal Flower from China
7
Figure 4. More isolated “hips”. All stereomicroscope photographs. Specimen number PB18320 for Figure 4A–C, PB21091a for Figure 4D–F, PB21092
for Figure 4G,I, PB21091b for Figure 4H.
(A) A “hip” including two fruits with persistent styles surrounded by a fleshy envelope. Note the wide space between the styles, and the gap between
the fruits (arrow). Bar, 1 mm.
(B) Persistent styles (black arrows) and a piece of yellowish cuticle (white arrow) in the gap between the two fruits. Bar, 0.5 mm.
(C) A detailed view of the “hip”, from the lower right of Figure 4A. Note the gap (arrow) in the middle. Bar, 0.5 mm.
8 Journal of Integrative Plant Biology
2009
angiosperm flower is formed by carpels (pistillate organs) and
stamens (staminate organs) that are often surrounded by a
perianth.” Although its accuracy and completeness are still
debatable, this definition does reflect the impression of a normal
flower in the minds of the general public. Until now, there has
been no well-accepted normal flower (bisexual, with a perianth)
from the Yixian Formation or older strata, since Sinocarpus is
fruits associated with leaves (Leng and Friis 2003, 2006) and
Archaefructus has no perianth (Sun et al. 1998, 2001, 2002).
Callianthus, however, meets the criterion of a flower given
by Friis et al. (2006). Besides this, we assign Callianthus to
angiosperms because: (i) its bisexuality distinguishes it from all
known gymnosperms except some Bennettitales and Gnetales;
(ii) the lack of interseminal scales, lack of a dome-shaped receptacle as well as lack of fleshy pollen organs, and the presence of
two divergent styles distinguish Callianthus from Bennettitales;
and (iii) two divergent papillate styles, a pedicellate solitary
flower, two rather than one fruit surrounded by the fleshy
envelope, and spatulate tepals in Callianthus contrast strongly
with a single ovuliferous unit with glabrous micropylar tube
sessile in bract axil, single seed surrounded by outer integument
in bract axil, and a triangular bract in Gnetales (Biswas and Johri
1997; Yang et al. 2003; Yang 2007). These characters clearly
distinguish Callianthus from Gnetales: (iv) the typical flower-like
arrangement of floral parts seen only in angiosperms is present
in Callianthus; and (v) seed surrounded by a fleshy envelope is
seen in Ginkgoaceae, Taxaceae and Podocarpaceae, and seed
in Gnetales is surrounded by an outer integument (Chamberlain
1957; Bierhorst 1971; Tomlinson et al. 1991; Tomlinson 1992;
Biswas and Johri 1997; Doyle 1998; Cope 1998; Tomlinson
and Takaso 2002). These appear similar to the fructification
organization in Callianthus. However, the presence of two
(rather than one) fruits per envelope, two stamens, two divergent
styles per flower/fructification, and several tepals in cycles in
a single flower, and their deployment in the flower distinguish
Callianthus clearly from these gymnosperms. Considering all of
these aspects – the similarities shared with angiosperms and
differences to gymnosperms – Callianthus should be placed in
angiosperms. Therefore, Callianthus is the oldest normal flower
that is bisexual and with a perianth.
As an early angiosperm, the following characters of Callianthus are concordant with conjectures of primitive angiosperms based on the analysis of living angiosperms: bisexuality, small size, undifferentiated perianth, superior ovary,
distinct tepals, moderate or low number of floral parts, distinct
stamens, and medium-sized pollen grains (Doyle and Endress
2000; Endress 2001). Doyle (2008) thinks that “the ancestral
flower had a perianth, more than one stamen, and more than
one carpel”. It appears that the morphology of Callianthus favors
these generalizations. Callianthus with tepals in cycles may represent the early-derived flowers with whorled arrangement, as
suggested by Soltis et al. (2000). Callianthus also demonstrates
a certain similarity to the eudicot mesofossils from the Early
Cretaceous in Portugal and North America in that they all have
small flowers, with few parts and undifferentiated tepals (Friis
et al. 2006). However, there are several unexpected features in
Callianthus, including the fleshy envelope, stamen, and round
triangular pollen grains.
An interesting character of Callianthus is the fleshy envelope.
Since there is no fleshy envelope enclosing carpels in plants
from the previous Cretaceous fossil record (Dilcher 1979; Friis
et al. 2006), the presence of a fleshy envelope in Callianthus
appears unique from a fossil perspective. Rarely in extant
angiosperms are carpels enclosed by a structure, such as the
floral roof in some Laurales (Heywood 1979; Endress 1980)
and expanded receptacle in Nelumbo (Nelumbaceae) (Hayes
et al. 2000). However, the tepals and/or stamens are inserted
on the outer surface of the floral roof in Laurales (Endress
1980), unlike the situation in Callianthus where the stamens
and tepals are distinct and arranged below the fleshy envelope
(Figure 1A,B,F,G,I). The expanded receptacle in Nelumbo and
fleshy envelope in Callianthus are similar to each other in their
fleshy nature and spatial relationship relative to the carpels,
stamens and perianth. In addition, the flowers in Nelumbo
and Callianthus also share the following features: possible
long pedicel, bisexual flower, evident tepals, parallel veins
←
Figure 4. (D) Another “hip” including two fruits with persistent styles surrounded by a fleshy envelope. Note the separation (arrow) in the
middle, and the space between two styles narrower than in A and B. Bar, 1 mm.
(E) A detailed view of the bottom of the “hip” in D. Note the two vascular bundles (triangles) to the envelope and two (arrows) to the fruits. Bar, 0.2 mm.
(F) A detailed view of the two styles, pieced together from two original images focusing on different levels. Note the border (arrow) between the fruit
and the envelope. Bar, 0.2 mm.
(G) A detailed view of the envelope in I. Note the outline of the envelope top (black line). The prick on the top may be related to the style that is not
visible in this specimen. Bar, 0.2 mm.
(H) A “hip” partially covered by sediments. Note the gap (arrow) in the middle, and the surrounding fleshy envelope of light color. Pieced together from
two original images. Bar, 1 mm.
(I) A “hip” with its top portion of the envelope preserved. Note the lack of apical depression and the dorsal ridge (upper arrow) in the fruit, and a
vascular bundle (lower arrow) at the bottom. Bar, 1 mm.
Earliest Normal Flower from China
Figure 5. A sketch of the flower. Note the pedicel (1), outer tepal (2),
inner tepal (7), filament (3), bristles (6) at the apex of the anther (5),
vascular bundle (4) to the carpels, gap (8) between two carpels (9),
fleshy envelope (10) around the carpels, papillate style (11), and a ring
on the fleshy envelope (12).
in tepals, distinct stamens and tepals, and enclosed carpels.
However, their differences are also obvious: Nelumbo has many
apocarpous carpels with sessile stigmas individually surrounded
by the receptacle (Hayes et al. 2000), while Callianthus has
only two paired stylate carpels enclosed by the fleshy envelope,
plus Nelumbo has tricolpate pollen while Callianthus has nontricolpate pollen. It should be borne in mind that the above
comparison may be superficial, and the similarities shared
between Callianthus and these living plants may be simply a
result of convergence or parallel evolution, and mean little in
phylogeny. It is premature to correlate Callianthus directly with
any living taxa for the time being, considering so little is known
about early fossil angiosperms.
The typical anthers of angiosperms have four pollen sacs and
are borne on filaments (Eames 1961; Friis et al. 2006). This is
also taken as a character used to identify a fossil angiosperm
(Friis et al. 2006). However, an exception to this rule does exist:
for example, Eames (1961) has mentioned anthers with two or
only one pollen sac. The above criterion for angiosperm anthers
based on generalization of living angiosperms should be applied
to early angiosperms with caution. There is much information
missing about Callianthus anthers, such as the anatomy of the
pollen sac wall, and the occurrence of subdivision in the anther.
The currently available data restrict us from further evaluation on
the anther of Callianthus. Bristles at the top of anthers are rare or
never seen in angiosperms. Although there are some reports of
anther appendages in some angiosperms (e.g. Melastomaceae,
Eames 1961), these can be easily distinguished from the bristles
in Callianthus by morphology, number, and spatial relationship
relative to the anther. Therefore the bristles on the top of the
9
anther in Callianthus cannot find their counterpart in extant
angiosperms.
Among the in situ pollen grains, two demonstrate a triangular
profile (Figure 3E,F). Triangular pollen grains may suggest
triaperturate pollen. If this was true, Callianthus might be more or
less related to eudicots. However, scanning electron microscopy
(SEM) cannot reveal details about the aperture of the pollen and
only two of the pollen grains appear so. Consequently, we are
not certain about the aperture of the pollen in Callianthus. We
cannot eliminate the possibility of trichotomosulcate aperture in
Figure 3E,F. Trichotomosulcate pollen grains, which are thought
to be transitional between monosulcate and tricolpate, have
been seen in basal eudicots, monocots and magnoliids (Wilson
1964; Harley 1990; Rudall et al. 1997; Sampson 2000; Furness
et al. 2002; Harley 2004). In addition, triangular pollen has
been seen in basal eudicots (Wilson 1964) and at least 27
genera of monocots (Harley 2004). For example, Agrostocrinum
scabrum (Hemerocallidaceae) has round triangular pollen grain
with trichotomosulcate aperture (Figure 3C,F, Harley 2004),
which is similar to those of Callianthus (Figure 3E,F). This
complicated situation cautions us against prematurely relating
Callianthus to eudicots. Apparently, the in situ pollen grains of
Callianthus alone do not provide enough information to resolve
its affinity in angiosperms.
The tepals of Callianthus demonstrate parallel venation
(Figures 1D, 2A). Usually a perianth element is thought to be
comparable to a leaf. If leaves of Callianthus, which are missing
in this case, share venation patterns with that of the tepals, it
may suggest that Callianthus appears to be related to monocots rather than eudicots. However, a similar venation pattern
has also been seen in Gnetales (Ephedra and Welwitschia)
(Biswas and Johri 1997; Yang et al. 2005) and Chaoyangia
(an Early Cretaceous fossil plant from western Liaoning with
a controversial phylogenetic position) (Duan 1998; Sun et al.
1998). Patience and further effort are required before making a
conclusion on the systematic position of Callianthus.
Bicarpellate gynoecium is a feature that has been frequently
seen in basal eudicots (Buxaceae, Papaveraceae, Gunneraceae, Hamamelidaceae, Menispermaceae, Ranunculaceae,
Circaeasteraceae, Sabiaceae, Chenopodiaceae, and Daphniphyllaceae) (Chu et al. 1991; Drinnan et al. 1991, 1994;
Takhtajan 1997; Judd et al. 1999; Zhang et al. 2004), suggesting
a possible eudicot affinity for Callianthus. However, it should
be borne in mind that bicarpellate gynoecium is also present
in Winteraceae and core eudicots (Brassicaceae, Salicaceae,
Solanales, Lamiales) (Drinnan et al. 1991; Judd et al. 1999;
Zhang et al. 2004). Suckleya (Chenopodiaceae) is especially
similar to Callianthus in divergent styles (Chu et al. 1991).
However, all of these living taxa can be easily distinguished
from Callianthus by the assemblage of style, stamen, perianth
morphology, and the lack of a fleshy envelope (Chu et al.
1991; Judd et al. 1999; Zhang et al. 2004). Drinnan et al.
(1994) has pointed out that the fossils on the stem lineage
10 Journal of Integrative Plant Biology
2009
leading to eudicots may have only two carpels. In addition,
the lack of differentiated sepal or petals and a few-parted,
cyclic floral architecture in Callianthus are also thought to be
primitive in eudicots (Drinnan et al. 1991, 1994). In the fossil
record, Callianthus is similar to Spanomera and Silucarpus from
the Cretaceous in bicarpellate gynoecium, the two latter are
related to Buxales (Drinnan et al. 1991; Pedersen et al. 2007). If
really related to Spanomera, Silucarpus or Nelumbo, Callianthus
would extend the fossil record of eudicots and provide support
for the position of Drinnan et al. (1991, 1994) owing to its
earlier age. Considering the early fossil record of tricolpate
pollen in the Early Cretaceous (Brenner 1976; Hughes 1994;
Harley 2004), the possible relationship of Sinocarpus (from the
same formation) to eudicots, and the above similarities shared
between Callianthus and eudicots, a future confirmation of a
relationship between Callianthus and eudicots would not be
surprising.
The stamens and carpels of Callianthus are different from
those of Archaefructus from the same locality (Sun et al 1998,
2002; Ji et al. 2004). The same can be said for Sinocarpus (Leng
and Friis 2003, 2006). This situation implies that our current
understanding on the early angiosperms is far from complete,
and that there is no model or pattern yet, at least in the Yixian
Formation, for the early angiosperms to follow. The increased
diversity of angiosperms in the Yixian Formation (Duan 1998;
Sun et al. 1998, 2001, 2002; Leng and Friis 2003, 2006; Ji
et al. 2004) and the early record of eudicots (Brenner 1976;
Drinnan et al. 1994; Pedersen et al. 2007) imply the existence
of angiosperms before the Barremian. This implication is not
only in agreement with the insect and pollen record (Ren 1998;
Wang et al. 2000a), but is also concordant with the recent report
of a Jurassic angiosperm (Wang et al. 2007a,b; Wang 2009;
Wang and Wang, in press).
The above discussion argues for, or is based on, the placement of Callianthus in angiosperms. If the discussion is correct,
Callianthus is of importance because it, along with Chaoyangia
Archaefructus and Sinocarpus (Hyrcantha, Dilcher et al. 2007),
is among the earliest angiosperms and thus sheds new light on
the origin and early radiation of angiosperms. However, even
if the above discussion was incorrect, the botanical value of
Callianthus would not shrink at all: it would be the monotype of
a new isolated class in seed plants, and thus would provide
new fuel for study on evolution, diversity and phylogeny of
seed plants. However, as stated above, the latter is very risky
in that the floral organization and details of the floral parts in
Callianthus have no counterpart, even remotely, in any known
gymnosperm.
in two facing slabs, embedded in grayish-yellow, thick, muddy
siltstones (PB21047, Figures 1–3). Further materials were from
similar rocks (PB18320, PB21091a&b, PB21092, Figure 4A–I)
at the same locality, Huangbanjigou Village, Shangyuan Town,
Beipiao City, Liaoning Province, China. The specimens were
observed and photographed using a Leica MZ-16A stereomicroscope with a digital camera (Figures 1A–I, 4A–I) and a Leo
1530 VP scanning electron microscope at Nanjing Institute of
Geology and Palaeontology, Nanjing, China. Some of the parts
of the flower were removed and coated with gold, then observed
using the SEM (Figures 2C, 3A–E). A fragment from the anther
region was found to contain in situ pollen grains (Figure 3A–E).
Cellulose transfers were made on specimen PB21047b. The
transfers were cleaned with hydrofluoric, coated with gold,
and observed using the SEM (Figures 2A,B, D–F, 3F,G). All
related information was used together to reconstruct the flower
(Figure 5).
Acknowledgements
We thank Dr Zhiyan Zhou for his support during this research, Mr
Gang Han for his generous support of contributing the valuable
specimens (PB21091, PB21092), Mr Yan Chen for collecting
the specimen (PB21047), Dr Shunqing Wu for the access to
specimen PB18320, Drs Shuren Zhang, Weiming Wang, and
Jianguo Li, Ms Chunzhao Wang, Mr Erjun Zhuo, Ms Zhiqin
Wang, Mr Xiting Cheng, and Ms Cuiling He for their help during
this research, and Ms Margaret Joyner for her help with English.
We thank Dr Michael Frohlich, James A. Doyle, and Peter Crane
for their help and constructive suggestions for improving the
manuscript. We also thank two anonymous referees for their
time and efforts reviewing and improving this manuscript.
References
Bierhorst DW (1971). Morphology of Vascular Plants. Macmillan, New
York.
Biswas C, Johri BM (1997). The Gymnosperms. Springer-Verlag, New
York.
Brenner GJ (1976). Middle Cretaceous floral province and early migrations of angiosperms. In: Beck CB, ed. Origin and Early Evolution of
Angiosperms. Columbia University Press, New York. pp. 23–47.
Chamberlain CJ (1957). Gymnosperms, Structure and Evolution. Johnson Reprint Corporation, New York.
Chu GL, Stutz HC, Sanderson SC (1991). Morphology and taxonomic
position of Suckleya suckleyana (Chenopodiaceae). Am. J. Bot. 78,
63–68.
Materials and Methods
Cope EA (1998). Taxaceae: the genera and cultivated species. Bot.
The specimens of Callianthus reported here included two flowers and four isolated fructifications. The flowers were found
Dilcher DL (1979). Early angiosperm reproduction: an introductory
Rev. 64, 291–322.
report. Rev. Palaeobot. Palyn. 27, 291–328.
Earliest Normal Flower from China
Dilcher DL, Sun G, Ji Q, Li H (2007). An early infructescence Hyrcantha
decussata (comb. nov.) from the Yixian Formation in northeastern
China. Proc. Natl. Acad. Sci. USA 104, 9370–9374.
Doyle JA (1998). Molecules, morphology, fossils, and the relationship
of angiosperms and Gnetales. Mol. Phyl. Evol. 9, 448–462.
Doyle JA (2008). Integrating molecular phylogenetic and paleobotanical
evidence on origin of the flower. Intl. J. Plant Sci. 169, 816–843.
11
Systematics: A Phylogenetic Approach. Sinauer Associates Inc.,
Sunderland.
Krassilov VA (1982). Early Cretaceous flora of Mongolia. Palaeontographica Abt B 181, 1–43.
Leng Q, Friis EM (2003). Sinocarpus decussatus gen. et sp. nov., a new
angiosperm with basally syncarpous fruits from the Yixian Formation
of Northeast China. Plant Syst. Evol. 241, 77–88.
Doyle JA, Endress PK (2000). Morphological phylogenetic analysis
Leng Q, Friis EM (2006). Angiosperm leaves associated with Sinocar-
of basal angiosperms: comparison and combination with molecular
pus Leng et Friis infructescences from the Yixian Formation (mid-
data. Intl. J. Plant Sci. 161, S121–153.
Drinnan AN, Crane PR, Friis EM, Pedersen KR (1991). Angiosperm
flowers and tricolpate pollen of buxaceous affinity from the Potomac
Group (mid-Cretaceous) of eastern North America. Am. J. Bot. 78,
153–176.
Drinnan AN, Crane PR, Hoot SB (1994). Patterns of floral evolution
in the early diversification on non-magnoliid dicotyledons (eudicots).
Plant Syst. Evol. 8, 93–122.
Duan S (1998). The oldest angiosperm – a tricarpous female reproductive fossil from western Liaoning Province, NE China. Sci. Chin. D.
41, 14–20.
Eames AJ (1961). Morphology of the Angiosperms. McGraw-Hill, New
York.
Endress PK (1980). Ontogeny, function and evolution of extreme floral
construction in the Monimiaceae. Plant Syst. Evol. 134, 79–120.
Endress PK (2001). The flowers in extant basal angiosperms and
inferences on ancestral flowers. Intl. J. Plant Sci. 162, 1111–1140.
Friis EM, Pedersen KR, Crane PR (2005). When earth started blooming: insights from the fossil record. Curr. Opin. Plant Biol. 8, 5–12.
Friis EM, Pedersen KR, Crane PR (2006). Cretaceous angiosperm
flowers: innovation and evolution in plant reproduction. Palaeogeo.
Palaeoclim. Palaeoecol. 232, 251–293.
Furness CA, Rudall PJ, Sampson FB (2002). Evolution of microsporogenesis in angiosperms. Int. J. Plant Sci. 163, 235–260.
Harley MM (1990). Occurrence of simple, tectate, monosulcate or
trichotomosulcate pollen grains within the Palmae. Rev. Palaeobot.
Palyn. 64, 137–147.
Harley MM (2004). Triaperturate pollen in the monocotyledons: configurations and conjectures. Plant Syst. Evol. 247, 75–122.
Harris JG, Harris MW (2000). Plant Identification Terminology: an
Illustrated Glossary. Spring Lake Publishing, Spring Lake.
Hayes V, Schneider EL, Carlquist S (2000). Floral development of
Nelumbo nucifera (Nelumbonaceae). Intl. J. Plant Sci. 161, S183–
S191.
Heywood VH (ed.) (1979). Flowering Plants of the World. Oxford
University Press, Oxford.
Early Cretaceous) of NE China. Plant Syst. Evol. 262, 173–187.
Li G, Luo ZX (2006). A Cretaceous symmetrodont therian with some
monotreme-like postcranial features. Nature 439, 195–200.
Liu M, Lu W, Ren D (2007). A new fossil mordellid (Coleoptera:
Tenebrionoidea: Mordellidae) from the Yixian Formation of Western
Liaoning Province, China. Zootaxa 1415, 49–56.
Pedersen KR, Von Balthazar M, Crane PR, Friis EM (2007). Early
Cretaceous floral structures and in situ tricolpate-striate pollen: new
early eudicots from Portugal. Grana 46, 176–196.
Ren D (1998). Flower-associated Brachycera flies as fossil evidences
for Jurassic angiosperm origins. Science 280, 85–88.
Rudall PJ, Furness CA, Chase MW, Fay MF (1997). Microsporogenesis and pollen sulcus type in Asparagales (Lilianae). Can. J. Bot. 75,
408–430.
Rudall PJ, Sokoloff DD, Remizowa MV, Conran JG, Davis JI, Macfarlane TD et al. (2007). Morphology of Hydatellaceae, an anomalous
aquatic family recently recognized as an early-divergent angiosperm
lineage. Am. J. Bot. 94 1073–1092.
Saarela JM, Rai HS, Doyle JA, Endress PK, Mathews S, Marchant
AD et al. (2007). Hydatellaceae identified as a new branch near the
base of the angiosperm phylogenetic tree. Nature 446, 312–315.
Sampson FB (2000). Pollen diversity in some modern Magnoliids. Intl.
J. Plant Sci. 161, S193–S210.
Soltis PS, Soltis DE, Zanis MJ, Kim S (2000). Basal lineages of
angiosperms: relationships and implications for floral evolution. Intl.
J. Plant Sci. 161, S97–S107.
Sun G, Dilcher DL, Zheng SL, Zhou ZK (1998). In search of the first
flower: a Jurassic angiosperm, Archaefructus, from Northeast China.
Science 282, 1692–1695.
Sun G, Ji Q, Dilcher DL, Zheng S, Nixon KC, Wang X (2002).
Archaefructaceae: a new basal angiosperm family. Science 296,
899–904.
Sun G, Zheng S, Dilcher D, Wang Y, Mei S (2001). Early Angiosperms
and their Associated Plants from Western Liaoning, China. Shanghai
Technology & Education Press, Shanghai.
Swisher CC, Wang YQ, Wang XL, Xu X, Wang Y (1998).
40
Ar/39 Ar
Hu SY (1998). Centenary on S.G. Nawaschin’s discovery of double
dating of the lower Yixian Fm, Liaoning Province, northeastern China.
fertilization: retrospects and prospects. Acta Bot. Sin. 40, 1–13.
Abstract in the 9th International Conference on Geochronology, Cos-
Hughes N (1994). The Enigma of Angiosperm Origins. Cambridge
mochronology and Isotope Geology (August 20–26, 1998), Beijing,
University Press, Cambridge.
Ji Q, Li H, Bowe M, Liu Y, Taylor DW (2004) Early Cretaceous
Archaefructus eoflora sp. nov. with bisexual flowers from Beipaio,
Western Liaoning, China. Acta Geol. Sin. 78, 883–896.
Judd WS, Campbell SC, Kellogg EA, Stevens PF (1999). Plant
China. Chin. Sci. Bull. 43, 125.
Takhtajan A (1997). Diversity and Classification of Flowering Plants.
Columbia University Press, New York.
Tomlinson PB (1992). Aspects of cone morphology and development
in Podocarpaceae (Coniferales). Intl. J. Plant Sci. 153, 572–588.
12 Journal of Integrative Plant Biology
2009
Tomlinson PB, Braggins JE, Rattenbury JA (1991). Pollination drop in
Wang XL, Wang YQ, Zhang FC, Zhang JY, Zhou ZH, Jin F et al.
relation to cone morphology in Podocarpaceae: a novel reproductive
(2000b). Vertebrate biostratigraphy of the Lower Cretaceous Yixian
mechanism. Am. J. Bot. 78, 1289–1303.
Formation in Lingyuan, western Liaoning and its neighbouring south-
Tomlinson PB, Takaso T (2002). Seed cone structure in conifers in
ern Nei Mongol (Inner Mongolia), China. Vertebr. PalAs. 38, 81–99.
relation to development and pollination: a biological approach. Can.
Wilson TK (1964). Comparative morphology of the Canellaceae. III.
J. Bot. 80, 1250–1273.
Wang WL, Zhang H, Zhang LJ, Zheng SL, Yang FL, Li ZT et al.
(2003) Stratigraphic sequence of the Yixian Formation of YixianBeipiao region, Liaoning – a study and establishment of stratotype
of the Yixian stage. J. Stratigr. 27, 227–232.
Wang WL, Zhang H, Zhang LJ, Zheng S-L, Yang FL, Li ZT et al.
(2004). Standard Sections of Tuchengzi Stage and Yixian Stage
and their Stratigraphy, Palaeontology and Tectonic-volcanic Actions.
Geological Publishing House, Beijing.
Wang X, Duan S, Geng B, Cui J, Yang Y (2007a). Schmeissneria: a
missing link to angiosperms? BMC Evol. Biol. 7, 14.
Wang X, Duan S, Geng B, Cui J, Yang Y (2007b). Is Juras-
Pollen. Bot. Gaz. 125, 192–197.
Wu S-Q (1999). A preliminary study of the Jehol flora from the western
Liaoning. Palaeoworld 11, 7–37.
Wu S-Q (2003). Land plants. In: Chang MM, Chen PJ, Wang YQ, Wang
Y, Miao DS, eds. The Jehol Biota. Shanghai Scientific & Technical
Publishers, Shanghai. pp. 165–177.
Yang Y, Fu DZ, Zhu G (2003). A new species of Ephedra (Ephedraceae)
from China. Novon 13, 153–155.
Yang Y (2007). Asymmetrical development of biovulate cones resulting
in uniovulate cones in Ephedra rhytiodosperma (Ephedraceae).
Plant Syst. Evol. 264, 175–182.
Yang Y, Geng B-Y, Dilcher DL, Chen ZD, Lott TA (2005). Morphology
sic Schmeissneria an angiosperm? Acta Palaeont. Sin. 46, 486–
and affinities of an early Cretaceous Ephedra (Ephedraceae) from
490.
China. Am. J. Bot. 92, 231–241.
Wang X (2009). New fossils and new hope for the origin of
Yao Y, Cai W, Ren D (2006). Fossil flower bugs (Heteroptera: Cimi-
angiosperms—Dedicated to the 200th anniversary of Charles Dar-
comorpha: Cimicoidea) from the Late Jurassic of Northeast China,
win’s birthday. In: Pontarotti P, ed. Evolutionary Biology from Concepts to Application II. Springer-Verlag, New York.
Wang X, Wang S (in press). Xingxueanthus, an angiospermous gymnosperm from the Jurassic? Acta Geol. Sin.
Wang X, Ren D, Wang Y (2000a). First discovery of angiospermous
pollen from Yixian Formation. Acta Geol. Sin. 74, 265–272.
including a new family, Vetanthocoridae. Zootaxa 1360, 1–40.
Zhang H, Huang Y, Miao R, Ye C, Liao W, Jin J (2004). Systematics
of Spermatophyta. Science Press, Beijing.
Zheng S, Li N, Li Y, Zhang W, Bian X (2005). A new genus of
fossil cycads Yixianophyllum gen. nov. from the Late Jurassic Yixian
Formation, Western Liaoning, China. Acta Geol. Sin. 79, 582–592.
(Handling editor: Mengxiang Sun)