Micromorphology of the Seed Envelope of Ephedra L. (Gnetales) and Its Relevance for the Timing of Evolutionary Events

Abstract

Micromorphology of the seed envelope of Ephedra (Gnetales) is known to be variable, but variation patterns have never been systematically documented. We test the usefulness of this feature for species determination and subclade delimitation in Ephedra and investigate the relationship of this character to infrageneric evolutionary patterns. Most species have a basically smooth seed envelope, which in some species appears slightly striate or reticulate due to convex or depressed outer periclinal cell walls. Ephedra rhytidosperma from China and Ephedra torreyana from North America have transverse lamellae formed by the epidermis. A papillate surface is found in respective close relatives of these two species. Micromorphology of the seed envelope is generally not useful for species identification or subclade delineation. The amount of variation is low, and intraspecific variation, which in some cases seems to be correlated with hybridization and/or introgression, complicates species recognition. Furthermore, parallel evolution of similar micromorphological patterns in unrelated subclades of Ephedra is evident and cannot be explained by similar seed dispersal mechanisms. The Asian species with transverse lamellae or papillae on the seed are dispersed by frugivores whereas similar American species are anemochoric. Transverse ridges occur in several Early Cretaceous fossil seeds with affinity to Ephedra. However, our results indicate that the resemblance between these fossils and extant taxa with similar features is superficial and convergent. In line with other recent studies, we find that Cretaceous ephedroids are extinct stem relatives to the extant clade.

Full text

MICROMORPHOLOGY OF THE SEED ENVELOPE OF EPHEDRA L. (GNETALES) AND ITS
RELEVANCE FOR THE TIMING OF EVOLUTIONARY EVENTS
Stefanie M. Ickert-Bond
1;
* and Catarina Rydin
y
*University of Alaska Museum of the North Herbarium (ALA), Department of Biology and Wildlife and Institute of Arctic Biology,
University of Alaska Fairbanks, 907 Yukon Drive, P.O. Box 756960, Fairbanks, Alaska 99775-6960, U.S.A., and School of
Life Sciences, Arizona State University, P.O. Box 874501, Tempe, Arizona 85287-4501, U.S.A.; and
y
University of
Zurich, Institute of Systematic Botany, Zollikerstrasse 107, CH-8008 Zurich, Switzerland, and
Department of Botany, Stockholm University, SE-106 91 Stockholm, Sweden
Micromorphology of the seed envelope of Ephedra (Gnetales) is known to be variable, but variation patterns
have never been systematically documented. We test the usefulness of this feature for species determination and
subclade delimitation in Ephedra and investigate the relationship of this character to infrageneric evolutionary
patterns. Most species have a basically smooth seed envelope, which in some species appears slightly striate or
reticulate due to convex or depressed outer periclinal cell walls. Ephedra rhytidosperma from China and
Ephedra torreyana from North America have transverse lamellae formed by the epidermis. A papillate surface
is found in respective close relatives of these two species. Micromorphology of the seed envelope is generally
not useful for species identification or subclade delineation. The amount of variation is low, and intraspecific
variation, which in some cases seems to be correlated with hybridization and/or introgression, complicates
species recognition. Furthermore, parallel evolution of similar micromorphological patterns in unrelated
subclades of Ephedra is evident and cannot be explained by similar seed dispersal mechanisms. The Asian
species with transverse lamellae or papillae on the seed are dispersed by frugivores whereas similar American
species are anemochoric. Transverse ridges occur in several Early Cretaceous fossil seeds with affinity to
Ephedra. However, our results indicate that the resemblance between these fossils and extant taxa with similar
features is superficial and convergent. In line with other recent studies, we find that Cretaceous ephedroids are
extinct stem relatives to the extant clade.
Keywords: Early Cretaceous, Ephedra, fossils, Gnetales, micromorphology, scanning electron microscopy, seed
envelope, systematics.
Introduction
Ephedra L. (Gnetales) comprises ;50 species inhabiting dry,
often sandy or rocky, subtropical to temperate areas of the
Northern Hemisphere and South America (Kubitzki 1990).
They are shrubs or herbs, similar in gross morphology, with op-
posite or whorled phyllotaxis. Ephedra species are usually dioe-
cious, and ovulate cones have fleshy or dry bracts. The ovules
are surrounded by an outer structure unique to Gnetales, the
seed envelope. The apical part of the integument forms a micro-
pylar tube that extends beyond the seed envelope and serves as
the pollen-receiving area. Both morphological variation and
molecular sequence divergence have been of limited value in
understanding evolution in Ephedra, and phylogenetic relation-
ships in the clade have long remained elusive. However, recent
studies using additional molecular data (Ickert-Bond and Woj-
ciechowski 2004; Rydin and Korall 2009) provide a robust
framework for investigations of character evolution.
Several recent discoveries document a surprisingly rich diver-
sity of ephedran species in the Early Cretaceous. Well-preserved
reproductive shoots bearing ovulate (and probably also pollen)
cones are recorded, for example, from the Yixian Formation of
western Liaoning, China (Guo and Wu 2000; Sun et al. 2001;
Yang et al. 2005; Rydin et al. 2006b), the Manlaj locality in
Mongolia (Krassilov 1982), and the Crato Formation, Brazil
(Mohr et al. 2004). Furthermore, well-preserved coalified seeds
are described from Buarcos, Portugal, and the Patuxent Forma-
tion of Virginia (Rydin et al. 2004; Rydin et al. 2006a).
Ephedra archaeorhytidosperma (Yang et al. 2005) is a com-
pression fossil collected from the Jianshangou Bed in the
lower part of the Yixian Formation and is of Early Cretaceous
age (Aptian; see summary in Zhou 2006). The fossils from the
Yixian Formation are of particular value in showing intercon-
nections between vegetative and reproductive parts. However,
they generally lack anatomical preservation, and it is often dif-
ficult to assess their phylogenetic position in detail. Features
diagnostic for extant Ephedra (i.e., Ephedra-type pollen and
papillae on the seed envelope; see also Rydin et al. 2006a), are
typically not preserved, but the general habit of E. archaeo-
rhytidosperma clearly indicates that it is an ephedroid plant.
One distinctive feature of this fossil is the presence of nu-
merous small transverse ridges on the surface of the seed enve-
lope (Yang et al. 2005). A similar seed surface pattern is also
found in the extant Chinese species Ephedra rhytidosperma
1
Author for correspondence; e-mail: smickertbond@alaska.edu.
Manuscript received June 2010; revised manuscript received September 2010.
36
Int. J. Plant Sci. 172(1):36–48. 2011.
Ó 2011 by The University of Chicago. All rights reserved.
1058-5893/2011/17201-0002$15.00 DOI: 10.1086/657299

(Yang et al. 2005; Yang 2007), which is endemic to the Helan
Mountains between Nei Mongol and Ningxia in China (Fu
et al. 1999). Yang et al. (2005) suggested an affinity between
E. rhytidosperma and E. archaeorhytidosperma for several
reasons, among them the transverse laminar ridges. However,
variation patterns in seed surface micromorphology have not
been documented for extant species of Ephedra, and the dis-
tribution of this feature is thus unknown.
Here we document variation in micromorphological pat-
terns of the seed envelope of extant Ephedra.Wetestthe
taxonomic usefulness of seed envelope patterns for species
determination a nd subclade delineation in Ephedra and in-
vestigate e volutionar y patterns of this feature. We also as-
sess the phylogenetic position of the Early Cretaceous fossil
E. archaeorhytidosperma.
Material and Methods
We investigated 117 specimens representing 48 living taxa
from both the Old World and the New World (table 1). In
general, two to five (or more) accessions of each species were
examined, but for 13 species, only one specimen of each was
available to us (table 1). Material at the pollination stage of
development and mature seeds were studied. Ovulate cones
were fixed in 70% ethanol or in FAA and dehydrated under
vacuum in an ethanol series (70%, 80%, 96%, 100%). Her-
barium material was softened using bis(2-ethylhexyl) sulpho-
succinate sodium salt (following Erbar 1995) for 7–10 d at
room temperature and then placed in 70% ethanol and dehy-
drated under vacuum as described above.
For SEM studies, specimens were postfixed in 2% osmium
tetroxide (Fluka), dehydrated in an ethanol series (70%, 80%,
90%, 100%) and acetone, critical-point dried, mounted on
stubs, and sputter-coated with gold. Seeds were examined
with an SEM, either an Amray 1000A (Arizona State Univer-
sity), an ISI-SR-50 (University of Alaska Fairbanks), or a Hita-
chi S-4000 (University of Zurich).
For serial sections, specimens were embedded in Kulzer’s
Technovit 7100 (2-hydroethyl methacrylate), following the
procedures outlined by Igersheim and Cichocki (1996) and
sectioned on a Microm HM 355 rotary microtome with a con-
ventional knife D. The sections were 6–7 mm thick and were
stained with ruthenium red and toluidine blue for 2 þ 2 min-
utes and mounted in Histomount. Mature, highly sclerified
seeds were embedded in Bio-Plastic resin (Ward’s Natural Sci-
ence, Rochester, NY) and sectioned on a lapidary saw. The
sections were mounted on slides, ground down to a minimal
thickness, and photographed using reflective and transmitted
light microscopy (Benedict et al. 2008). Permanent slides are
deposited at the Department of Systematic Botany, Stockholm
University (SUNIV), and the ASU Fossil Plant Collections,
Arizona State University.
Results
General Morphology
The shape of Ephedra seeds varies between species and
sometimes also within species (table 1). In longitudinal sec-
tion, they may be lanceolate (fig. 1A; e.g., E. torreyana), ellip-
tic (fig. 1B ; E. foeminea), ovate (fig. 1C; E. equisetina), or
oblong (fig. 1D; E. transitoria). In transverse section, they are
rounded or angled and the angles are sometimes prominent
(e.g., in E. rhytidosperma). In one-seeded ovulate cones, the
adaxial side of the seed is convex at midlength of the ovule/
seed (e.g., E. equisetina, E. californica); in two-seeded cones,
it is generally flat (e.g., E. sinica, E. pedunculata); and in
three-seeded cones it has a median longitudinal ridge (e.g., E.
multiflora, E. przewalskii). Two abaxial, lateral furrows may
be present in some species (e.g., E. sinica, E. pachyclada, E.
triandra). Mature seeds are typically larger than those at the
pollination stage of development, and their seed envelopes
have longer epidermal cells, but no other differences in shape
or surface patterns are observed (table 1).
SEM Observations of the Seed Envelope
Epidermal cells are rectangular and arranged parallel to
the longitudinal axis of the ovule or seed. The cells are usu-
ally ;20–30 m m wide. Cell length is variable within and be-
tween species (table 1), ranging in mature seeds from 50 to
500 mm at midlength of the seed. Cells are shorter in the api-
cal region of the seed and in the seed envelope of ovules at
the pollination stage of development.
Most species have a basically smooth seed envelope surface
(figs. 1B,1D,3A). Epidermal cell boundaries are usually
clearly visible, and periclinal walls are flat (fig. 1E), depressed
(fig. 1F), or convex (fig. 1G). Sometimes (e.g., in E. pachy-
clada), cell boundaries are less obvious. Features are usually
distinct but may be variable within species or even within
a single seed (table 1). For example, some but not all speci -
mens of E. aspera and E. fasciculata consistently have an
indistinctly coarse surface pattern, which obscures cell bound-
aries (not shown). Variation is also observed in the shape of
the end walls of cells; they are either straight or oblique and
variously raised above the general epidermal cell surface (fig.
1E–1G). The shape of end walls of cells is usually highly vari-
able within a single seed (e.g., fig. 1G).
In two unrelated species, E. rhytidosperma from East Asia
(fig. 2C,2G,2K) and E. torreyana from North America (figs.
2B,2F,2J,3C, 4), periclinal walls of adjacent epidermal cells
grow outward and are united to form transverse lamellae on
both the adaxial and the abaxial sides of the seed envelope.
The outgrowths are typically 100–500 mm wide in both spe-
cies and are formed by a few cells to up to 20 adjacent cells in
E. torreyana and up to 40 cells in E. rhytidosperma. The la-
mellae are formed by the cells of the epidermis; the mesophyll
is not involved in forming the lamellar outgrowths (fig. 3C).
In two specimens of E. torreyana, the lamellae are ‘‘papillate,’’
that is, less prominent and formed by one to several adjacent
papillae (fig. 1J). Ephedra multiflora has transverse ridges (fig.
1I) of the same width as those of E. torreyana and E. rhytido-
sperma, but the ridges are less prominently lamellar, which
gives an overall wavy pattern rather than a lamellar one. In
addition, weak tendencies to form tiny transverse ridges on at
least parts of the seed envelope were observed in several unre-
lated species, for example, in one specimen of E. trifurca
and
in E. pachyclada and E. alata.
A few species in two unrelated clades have a prominently
papillate seed surface (figs. 1H,2I,2L, 4). The papillae are
37
ICKERT-BOND & RYDIN—MICROMORPHOLOGY OF THE EPHEDRA SEED ENVELOPE

Table 1
Voucher Information of Ephedra Specimens Examined and Seed Characters Studied (Based on 10 Measurements)
Clade, specimen ID Species Locality Voucher
Seed shape in
longitudinal
section; apex shape
General
epidermal
pattern
Sculpt
size (mm) Cell length (mm)
Periclinal
cell walls ITS
a
Mediterranean:
UAF E. foeminea Forssk. Italy 1526: 1983 (PE
seed bank)
b
Elliptic; acute Smooth ... 80–200 Variable ...
130 E. foeminea Forssk. Greece Rydin 130 (Z)
b
Narrowly
elliptic; acute
Smooth ... 120–200 Convex GU968546
152 E. foeminea Forssk. Dalmatia Freitag 19.807 (KAS)
c
Elliptic; acute Smooth ... 100–180 Convex GU968551
159 E. foeminea Forssk. Greece Fries C-7619 (S)
c
Narrowly
elliptic; acute
Smooth ... 100–180 Convex ...
UAF E. alata Decne. Algeria Cosson s.n. (MO)
b
Ovate; acuminate Smooth (partly
weakly
transversely
lamellar)
... Unclear Flat ...
128 E. alata Decne. Algeria Anderberg 481 (S)
c
Ovate; acuminate Smooth (partly
weakly
transversely
lamellar)
100–300 100–180 Flat or convex ...
147 E. alata Decne. Algeria Cosson C-311 (S)
b
Ovate; acuminate Smooth (partly
weakly
transversely
lamellar)
;300 Unclear Flat ...
082 E. altissima Desf. Tunisia Botan SU 18 (S)
c
Elliptic; acute Smooth ... 40–150 Convex AY755773
132 E. altissima Desf. Morocco Freitag 35.001 (KAS)
b
Ovate; acute Smooth ... 150–200 Convex ...
080 E. aphylla Forssk. Libya Anderberg 853 (S)
c
Elliptic; acute Smooth ... 30–60 Convex AY755771
124 E. aphylla Forssk. Palestine Kramer 4727 (Z)
c
Narrowly
elliptic; acute
Smooth ... 100–200 Convex GU968544
154 E. aphylla Forssk. Israel Amdursky 402 (S)
c
Ovate; acute Smooth ... 150–300 Variable GU968552
101 E. fragilis Desf. Morocco Jonsell 5412 (UPS)
c
Narrowly
oblong; acute
Smooth ... 50–100 Variable FJ958014
109 E. fragilis Desf. Morocco Denk s.n. (S)
c
Narrowly
elliptic; acute
Smooth ... 70–140 Convex FJ958019
120 E. fragilis Desf. Hispaniola Freitag 328–40 (Z)
c
Narrowly
elliptic; acute
Smooth ... 50–150 Convex ...
162 E. major Host ssp. major Spain Ipse 71/677E (Z)
c
Narrowly
oblong; acute
Smooth ... 120–220 Convex GU968553
166 E. major Host ssp. major Algeria Hofmann
013–1971 (Z)
c
Narrowly
oblong; acute
Smooth ... 120–150 Convex GU968557
167 E. major Host ssp. major Algeria Juillet 94 (Z)
c
Narrowly
oblong; acute
Smooth ... 80–180 Convex GU968558
Clade A:
146 E. ciliata Fisch. et C.A. Mey. Morocco Balls B2487 (S)
c
Ovate; acute Smooth ... 70–180 Variable GU968548
153 E. ciliata Fisch. et C.A. Mey. Turkmenistan Androssov 3367 (S)
c
Elliptic; acute Smooth ... 80–200 Flat or depressed ...
096 E. foliata Boiss. et C.A. Mey. Somalia Thulin 10745 (UPS)
c
Elliptic; acute Smooth ... 80–180 Flat or convex FJ958010
North America:
ASU E. antisyphilitica Berl. ex C.A. Mey. Texas Ickert-Bond 900 (ASU)
b
Elliptic; obtuse Smooth ... 110–140 Depressed AY599148
UAF E. aspera Engelm. ex S. Watson Arizona Rose 40086 (MO)
b
Ovate; acute Smooth ... 50–100 Depressed ...
ASU E. aspera Engelm. ex S. Watson Texas Correll 23971 (NY)
b
Ovate; acute Indistinctly coarse ... Unclear Convex ...
UAF E. aspera Engelm. ex S. Watson California Faulkner 545 (UCR)
b
Ovate; acute Indistinctly coarse ... 200–280 Convex ...
38

UAF E. aspera Engelm. ex S. Watson Big Bend Basin
Road, Texas
Ickert-Bond 895 (ASU)
b
Ovate; acute Papillate 5–10 160–270 Flat; papillate ...
UAF E. californica S. Watson California Hendrickson 8616 (ASU)
b
Ovate; acute Smooth ... Unclear Convex ...
ASU E. californica S. Watson California Jepson 20690 (NY)
b
Ovate; obtuse Smooth ... 130–300 Convex ...
161 E. californica S. Watson California Tidestrom 9692 (S)
c
Ovate; acute Smooth ... 80–170 Convex ...
UAF E. clokeyi Cutler Arizona Jones s.n. (MO)
b
Narrowly
elliptic; acute
Smooth ... 100–180 Flat ...
ASU E. compacta Rose Mexico Correll & Johnson
19900 (NY)
b
Oblong; acute Smooth ... 70–180 Depressed ...
155 E. compacta Rose Mexico Purpus s.n. (S)
c
Oblong; acute Smooth ... 60–120 Depressed ...
UAF E. coryi E.L. Reed Texas Warnock 10713 (TEX)
b
Elliptic; obtuse Smooth ... 90–150 Convex ...
UAF E. coryi E.L. Reed Texas Correll 32762 (TEX)
c
Elliptic; obtuse Smooth ... 60–130 Convex ...
UAF E. cutleri Peebles Utah Holmgren 12744 (ASU)
b
Ovate; acute Smooth ... 60–130 Flat ...
ASU E. cutleri Peebles Arizona Ickert-Bond 692 (ASU)
b
Oblong-unclear;
acute
Smooth ... 120–260 Flat ...
UAF E. cutleri Peebles Arizona Ickert-Bond 996 (ASU)
b
Oblong-ovate;
acute
Smooth ... 130–240 Flat or convex ...
UAF
E. fasciculata A. Nelson
Arizona Ickert-Bond 513 (ASU)
b
Ovate;
acuminate
Indistinctly coarse ... 70–140 Flat or convex AY599180
ASU
E. fasciculata A. Nelson
Arizona Ickert-Bond 813 (ASU)
b
Ovate;
acuminate
Smooth ... 120–160 Flat or convex ...
UAF
E. fasciculata A. Nelson
Arizona Ickert-Bond 541 (ASU)
b
Ovate;
acuminate
Indistinctly coarse ... 50–120 Convex ...
UAF
E. fasciculata A. Nelson
Arizona Ickert-Bond s.n. (ASU)
b
Ovate;
acuminate
Smooth ... 170–240 Convex ...
UAF E. funerea Coville et Morton Arizona Ickert-Bond 573 (ASU)
b
Lanceolate;
acuminate
Papillate 10–25 Unclear Flat; papillate ...
UAF E. funerea Coville et Morton Arizona Baker 13971 (ASU)
b
Lanceolate;
acuminate
Papillate 5–10 Unclear Flat, papillate ...
ASU E. funerea Coville et Morton California Wolf 10599 (RSA, NY)
b
Lanceolate;
acuminate
Papillate 5–10 150–360 Flat; papillate ...
UAF E. funerea Coville et Morton California Sanders 9049 (UCR)
b
Lanceolate;
acuminate
Papillate 5–10 Unclear Flat; papillate ...
UAF E. nevadensis S. Watson Nevada Gierisch 4722 (ARIZ)
b
Ovate; acute Smooth ... 160–410 Flat or convex ...
122 E. nevadensis S. Watson n.a. Polovado s.n. (Z)
c
Ovate; acute Smooth ... 50–120 Convex ...
UAF E. pedunculata Engelm. ex S. Watson Texas Parks 3199 (MO)
b
Oblong; acute Smooth ... 100–250 Variable ...
ASU E. pedunculata Engelm. ex S. Watson Texas Palmer 1291 (NY)
b
Oblong; acute Smooth ... 180–350 Variable ...
UAF E. torreyana S. Watson Arizona Ickert-Bond 666 (ASU)
b
Lanceolate;
acuminate
Transversely
lamellar
50–520 150–350 Flat or convex ...
ASU E. torreyana S. Watson Colorado Franklin 3368 (NY)
b
Lanceolate;
acuminate
Transversely
lamellar
120–580 150–450 Flat or convex ...
S E. torreyana S. Watson Nevada Clokey 8224 (S)
c
Lanceolate;
acuminate
Transversely
lamellar
100–500 100–300 Flat or convex ...
UAF E. torreyana S. Watson Arizona Ickert-Bond 998 (ASU)
b
Lanceolate;
acuminate
Papillate 30–50 600–670 Convex;
papillate
...
126 E. torreyana S. Watson New Mexico Porter & Porter 8998 (S)
c
Lanceolate;
acuminate
Papillate 20–30 100–500 Convex;
papillate
...
UAF E. trifurca Torrey ex S. Watson Arizona Skjot-Pedersen s.n. (PE)
b
Lanceolate;
acuminate
Smooth ... 120–200 Convex ...
UAF E. trifurca Torrey ex S. Watson Arizona Ickert-Bond 577 (ASU)
b
Lanceolate;
acuminate
Smooth ... 140–220 Convex ...
ASU E. trifurca Torrey ex S. Watson Arizona Ickert-Bond 753 (ASU)
b
Lanceolate;
acuminate
Smooth (partly
weakly
transversely
lamellar)
... 130–170 Convex AY599164
39

Table 1
(Continued )
Clade, specimen ID Species Locality Voucher
Seed shape in
longitudinal
section; apex shape
General
epidermal
pattern
Sculpt
size (mm) Cell length (mm)
Periclinal
cell walls ITS
a
253 E. trifurca Torrey ex S. Watson Arizona Goodding 2268 (S)
b
Lanceolate;
acuminate
Smooth ... 150–320 Convex ...
254 E. trifurca Torrey ex S. Watson Arizona Nelson & Nelson 1290 (S)
b
Lanceolate;
acuminate
Smooth ... 170–240 Convex ...
ASU E. viridis Coville California Parish 2975 (NY)
b
Oblong-ovate;
acute
Smooth ... 190–280 Convex ...
091 E. viridis Coville Utah Holmgren et al. 1826 (UPS)
c
Ovate; acute Smooth ... 25–60 Convex FJ958005
South America:
ASU E. americana Humb. et
Bonpl. ex Willd.
Ecuador Juncosa 2257 (NY)
b
Elliptic; acute Smooth ... 190–400 Depressed ...
127 E. americana
Humb. et Bonpl. ex Willd.
Argentina Novara 8219 (S)
c
Elliptic; acute Smooth ... 120–300 Convex GU968545
UAF E. boelkei Roeg. Argentina Ickert-Bond 1252 (ASU)
b
Ovate; acute Smooth ... 100–250 Convex AY599175
ASU E. breana Phil. Chile Ickert-Bond 1233 (ASU)
b
Elliptic; rounded Smooth ... 140–240 Depressed ...
025 E. chilensis K. Presl n.a. Chase 10140 (K)
c
Elliptic; acute Smooth ... 40–80 Convex AY755744
ASU E. chilensis K. Presl. Argentina Jostasato 4333 (NY, ARIZ)
b
Elliptic; acute Smooth ... 120–220 Convex ...
075 E. chilensis K. Presl. Chile Forbes 49.0542 (UC)
c
Elliptic; acute Smooth ... 30–100 Convex AY755767
123 E. chilensis K. Presl. Chile Gay 400 (Z)
c
Elliptic; acute Smooth ... 40–140 Convex GU968543
UAF E. multiflora Phil. ex Stapf Chile Ickert-Bond 1211 (ASU)
b
Ovate;
acuminate
Transversely
wavy
120–480 120–470 Flat AY599173
UAF E. multiflora Phil. ex Stapf Chile Ickert-Bond 1231 (ASU)
b
Ovate;
acuminate
Transversely
wavy
100–500 100–500 Flat ...
UAF E. ochreata Miers Argentina Ickert-Bond 1253 (ASU)
b
Narrowly
ovate; obtuse
Smooth ... 110–220 Flat or convex ...
ASU E. ochreata Miers Argentina Ickert-Bond 1257 (ASU)
b
Narrowly
ovate; obtuse
Smooth ... 100–180 Depressed ...
UAF E. rupestris Benth. Ecuador Ickert-Bond 1100 (ASU)
b
Elliptic; obtuse Smooth ... 50–150 Depressed AY599167
073 E. rupestris Benth. Ecuador Ornduff 9675 (UC)
c
Elliptic; obtuse Smooth ... 25–60 Depressed AY755765
ASU E. triandra Tul. Argentina Capitanelli 584 (ASU)
b
Oblong;
narrowly acute
Smooth ... 80–190 Depressed ...
UAF E. triandra Tul. Argentina R. Leal 15981 (ASU)
b
Oblong;
narrowly acute
Smooth ... 150–400 Depressed ...
UAF E. tweediana Fisch. ex C.A. Mey. Uruguay Herter 1010 (MO)
b
Elliptic; obtuse Smooth ... 120–200 Depressed ...
076 E. tweediana Fisch. ex C.A. Mey. Argentina Forbes 66.0742 (UC)
c
Elliptic; obtuse Smooth ... 30–100 Variable AY755768
Asia H:
002 E. likiangensis Florin n.a. Rydin 03–926 (S)
c
Elliptic; acute Smooth ... 20–45 Convex AY755739
157 E. likiangensis Florin n.a. Cult. 1988–844 (K)
c
Oblong; obtuse Smooth ... 40–110 Convex ...
158 E. likiangensis Florin n.a. Cult. 18480.000 (K)
c
Oblong; acute Smooth ... 90–150 Convex ...
007 E. minuta Florin Sikang,
China
Rydin 03–930 (S)
c
Elliptic; acute Smooth ... 40–80 Convex AY755742
063 E. minuta Florin n.a. Rydin 04–486 (S)
c
Elliptic;
acute-obtuse
Smooth ... 30–50 Convex AY755756
Asia K:
UAF E. equisetina Bunge Mount Helan,
China
Yang 2004003 (PE)
b
Ovate; acute Papillate and
transversely
lamellar
5–10; 0–200 Unclear Flat or convex;
papillate
...
071 E. equisetina Bunge Georgia Merello et al. 2241 (MO)
c
Elliptic; acute Papillate 2–4 20–80 Flat or convex;
papillate
AY755763
40

125 E. equisetina Bunge Turkmenistan Sintensis 666 (S)
c
Narrowly
ovate; acute
Smooth ... 80–200 Convex ...
142 E. equisetina Bunge Russian Altai Freitag 05.2008 (KAS)
c
Ovate; acute Smooth ... 100–200 Convex ...
234 E. equisetina Bunge Turkmenistan Lipsky 2610 (S)
c
Ovate; acute Smooth ... 40–100 Convex GU968572
249 E. equisetina Bunge Turkestan Moldengauer 22 (S)
c
Ovate; acute Smooth ... 75–200 Convex ...
250 E. equisetina Bunge Turkmenistan Lipsky 3653 (S)
b
Ovate; acute Smooth ... 100–180 Convex ...
251 E. equisetina Bunge Turkmenistan Lipsky 2587 (S)
c
Ovate; acute Smooth ... 100–200 Convex ...
252 E. equisetina Bunge Kopet Dag
Mountain
Lipsky 2124 (S)
c
Ovate; acute Papillate 4–10 70–150 Convex ...
148 E. gerardiana Wall. ex Florin Almora, India Parker 2099 (S)
c
Oblong; acute Smooth ... 50–150 Depressed ...
UAF E. major Host Morocco Lewalle 9642 (MO)
b
Oblong; acute Papillate 4–8 40–100 Flat; papillate ...
163 E. major Host Spain Montserrat 319171 (Z)
c
Elliptic; acute Papillate 4–10 100–130 Convex; papillate GU968554
164 E. major Host Transcaucasia Grossheim s.n. (Z)
c
Elliptic; acute Papillate 2–8 90–150 Convex; papillate GU968555
165 E. major Host Herzegovina Baenitz s.n. (Z)
c
Elliptic; acute Papillate 2–8 100–175 Convex; papillate;
with apical warty
projections
GU968556
169 E. major Host France Zogg & Gassner
8388 (Z)
c
Elliptic; acute Papillate 2–8 120–180 Convex; papillate GU968559
156 E. pachyclada Boiss. Hissar,
Turkestan
Regel s.n. (S)
c
Elliptic; rounded Smooth (at
mid-length
with weak
transverse
lamellae)
10–35 Unclear Flat or convex;
papillate; with
apical warty
projections
...
UAF E. rhytidosperma Pachom. Mount Helan,
China
Yang 20060620 (PE)
b
Obovate-elliptic;
acute
Transversely
lamellar
100–500 Unclear Convex ...
138 E. saxatilis (Stapf) Royle ex Florin n.a. Cult. 1947–2603 (K)
c
Oblong; obtuse Smooth ... 50–130 Flat or convex ...
144 E. saxatilis (Stapf) Royle ex Florin Nepal Freitag 098–38–74–84
(KAS)
c
Oblong; obtuse Smooth ... 100–200 Flat or convex ...
Asia M:
140 E. distachya L. n.a. Cult. 46126.000 (K)
c
Elliptic; acute Smooth ... 15–40 Convex ...
143 E. sarcocarpa Aitch. et Hemsl. Iran Freitag 13.988 (KAS)
c
Ovate; acute Smooth ... 100–170 Flat ...
UAF E. strobilacea Bunge Asia Media Collector unknown
(PE seed bank no.
0679: 1961)
b
Lanceolate;
narrowly
acute
Smooth ... 40–100 Depressed ...
150 E. strobilacea Bunge Turkmenistan Androssov 1900 (S)
c
Lanceolate;
narrowly
acute
Smooth ... 100–240 Flat or convex GU968549
111 E. transitoria Riedl Saudi Arabia Collenette 9095 B (E)
c
Narrowly
oblong; acute
Smooth ... 70–130 Flat FJ958021
Asia N:
173 E. fedtschenkoae Paulsen Xinjiang, China Zhu Taiyan
650764 (N)
c
Narrow
elliptic; acute
Smooth ... 60–100 Flat or convex ...
UAF E. intermedia Schrenk ex C.A. Mey. Gansu, China Yang LZ060707
(PE)
b
Ovate; acute Smooth ... 100–200 Flat ...
006 E. intermedia Schrenk
ex C.A. Mey.
Tien-Shan
Mountains,
Asia Media
Rydin
03–925 (S)
c
Elliptic; acute Smooth ... 60–170 Flat or convex AY755741
092 E. lomatolepis Schrenk Kazakhstan Baitulin et al.
s.n. (UPS)
c
Ovate; acute Smooth 10–40 Unclear Flat or convex;
with rare
warty projections
FJ958006
102 E. lomatolepis Schrenk Tschu-Ili
Mountains,
Turkestan
Titow 488 (S)
c
Ovate; acute Smooth 10–20 100–170 Flat; with rare
warty projections
FJ958015
41

Table 1
(Continued )
Clade, specimen ID Species Locality Voucher
Seed shape in
longitudinal
section; apex shape
General
epidermal
pattern
Sculpt
size (mm) Cell length (mm)
Periclinal
cell walls ITS
a
174 E. lomatolepis Schrenk Pakistan Bosshard et al.
803.24 (Z)
c
Elliptic; acute Smooth ... 70–190 Flat or depressed GU968562
UAF E. regeliana Florin Xinjiang, China K.C. Kuan
1067 (PE)
b
Narrow
elliptic; acute
Smooth ... 50–100 Flat or depressed ...
UAF E. sinica Stapf Hebei, China Unknown coll.
s.n. (herb. no.
00015747,
PE)
b
Ovate; acute Smooth ... 40–150 Flat ...
UAF E. sinica Stapf Inner Mongolia,
China
Chu 20060801
(PE)
b
Ovate; acute Smooth ... 70–150 Flat ...
151 E. sinica Stapf Inner Mongolia,
China
Eriksson
05–9020 (S)
c
Ovate; acute Smooth ... Unclear Depressed GU968550
a
The internal transcribed spacer of the nuclear ribosomal DNA (ITS) is available for some of the vouchers and has been included in phylogenetic analyses in published studies (Ickert-
Bond and Wojciechowski 2004; Rydin et al. 2004, 2010; Rydin and Korall 2009). n.a. ¼ information not available.
b
Material ¼ mature seeds.
c
Material in pollination stage of development.
42

Fig. 1 Seed shapes in longitudinal outline (A–D) and surface patterns of the seed envelope (E–L). Scale bars: A–D, 2 mm; E–G, I–L, 100 mm;
H,20mm. A, Seed lanceolate in Ephedra torreyana (Franklin 3368 [NY]); B, seed elliptic in Ephedra foeminea (1526 [PE seed bank]); C, seed
ovate in Ephedra multiflora (Ickert-Bond 1211 [ASU]); D, seed oblong in Ephedra transitoria (Collenette 9095B [E], at pollination stage of
development, i.e., smaller than the others); E, outer periclinal cell walls flat in Ephedra sarcocarpa (Freitag 13.988 [KAS]); F, outer periclinal cell
walls depressed in Ephedra regeliana (K.C. Kuan 1067 [PE]); G, outer periclinal cell walls convex in Ephedra trifurca (Ickert-Bond 577 [ASU]); H,
43

formed by the outer periclinal walls of the epidermis (fig.
3B). There are generally ;2–8 papillae per cell and each pa-
pilla is 2–10 mm across. As with the transverse lamellae, the
cells of the mesophyll are never involved in forming the pa-
pillae. A papillate seed surface is present in the Old World
species E. equisetina and E. major (fig. 2D,2H,2L) and in
the New World species E. funerea (fig. 2A,2E,2I). Investi-
gated specimens of E. major and E. funerea consistently have
a papillate seed envelope, but in E. equisetina there is sub-
stantial intraspecific variation (table 1). Several investigated
outer periclinal cell walls papillate in Ephedra equisetina (Yang 2004003 [PE]); I, ‘‘wavy’’ transverse ridges in E. multiflora (Ickert-Bond 1231
[ASU]); J, ‘‘papillate lamellae’’ in E. torreyana (Porter & Porter 8998 [S]); K, ‘‘ridge-like papillae’’ in E. equisetina (Yang 2004003 [PE]); L, ‘‘wart-
like projections’’ in Ephedra pachyclada (Regel s.n. [S]).
Fig. 2 Micromorphological patterns of the seed envelope: transverse lamellae (B–C, F–G, J–K) and papillae (A, D–E, H–I, L). Scale bars: A–D,2
mm; E–H, J–K, 100 mm; I, L,20mm. A, E, I, Papillate surface of the North American species Ephedra funerea (Sanders 9049 [UCR]); B, F, J, transverse
lamellar surface of the North American species Ephedra torreyana (Ickert-Bond 666 [ASU]); C, G, K, transverse lamellar surface of the Asian species
Ephedra rhytidosperma (Yang 20060620 [PE]); D, H, L, papillate surface of the Eurasian species Ephedra major (Lewalle 9642 [MO]).
44
INTERNATIONAL JOURNAL OF PLANT SCIENCES

specimens of E. equisetina lack papillae altogether and have
smooth seed envelopes (not shown). Seeds of E. equisetina
from Mount Helan, China, have transverse lamellae similar
to those in E. rhytidosperma and E. torreyana but with pa-
pillae on the lamellae (fig. 1K).
The specimens of E. pachyclada and E. lomatolepis and one
specimen of E. major (table 1) have wartlike projections on
the surface (fig. 1L). In E. pachyclada and E. major, they are
present only in the apical region of the seed envelope. In E. lo-
matolepis, they are rare but may occur over the entire seed
surface. Each projection is ;10–40 mm across. Some overarch
cell boundaries, and some appear to have collapsed.
In several unrelated species (e.g., E. alata, E. californica, E.
equisetina, E. major ssp. major, E. intermedia, and E. pachy-
clada), we observed stomata in the apical region of the seed
envelope. They have the same shape and structure as those of
cone bracts and leaves; that is, there are no obvious subsidiary
cells, and the guard cells are sunken to the level of the base of
the epidermal cell layer.
Discussion
Comparative Structural Evaluation and
Systematic Implications
Generally, phyllotaxy determines the number of ovules/seeds
per cone (Takaso 1985; Rydin et al. 2010), but a reduction
in ovule number has taken place in one-seeded cones in some
decussate-leaved (e.g., Ephedra antisyphilitica, E. equisetina)
and trimerous-leaved taxa (e.g., E. trifurca). The shape of the
ovule/seed in transverse section at its midlength is largely de-
termined by pressure from other organs, that is, the number of
ovules/seeds in the cone, whereas transverse shape in the apical
region of the ovule/seed is determined by the number of vascu-
lar bundles in the seed envelope (Rydin et al. 2010).
The majority of Ephedra species have smooth seed enve-
lopes with minor variation in the micromorphological pat-
terns, detectable only under SEM. The outer periclinal cell
wall of the epidermis can be convex or depressed, which re-
sults in striate or reticulate patterns of the seed envelope. Taxa
of the Mediterranean species complex (fig. 4) are often striate,
whereas many Asian species have flat periclinal walls, but var-
iation is extensive within species.
Transverse lamellae on the surface of the seed envelope, vis-
ible with the naked eye, are present in two distantly related
species: E. rhytidosperma from China and E. torreyana from
North America. In both species, the lamellae are between 100
and 500 mm wide (variable in width within a single seed) and
are formed by the outer epidermis. As seen in transverse sec-
tions of the seed envelopes of E. torreyana (fig. 3C) and of E.
rhytidosperma (Yang 2002), the lamellae are formed by exten-
sions of the outer tangential cell wall of one to several adja-
cent epidermal cells. The inner tangential cell wall of these
cells is in line with that of the other epidermal cells, and the
mesophyll is not involved in the formation of the lamellae (fig.
3C; Yang 2002; Rydin et al. 2010). A similar pattern, but one
that is less prominently lamellar, is observed in E. multiflora
from South America, and weak ridges are seen in parts of
seeds of several other species (table 1).
From our survey it is evident that a similar lamellate micro-
morphological pattern has evolved independently in several
clades in Ephedra (fig. 4). It does not appear to have evolved
in response to a common dispersal syndrome or other ecologi-
cal factors. Ephedra rhytidosperma has fleshy cone bracts at
seed maturity, and seeds are probably dispersed by birds or liz-
ards. In contrast, E. torreyana has dry winged bracts at seed
maturity, and seeds are dispersed by the wind (Hollander
et al. 2010). We do not find any other shared ecological trait
that could explain this parallelism.
A papillate surface of the seed envelope has arisen (at least)
twice in the Asian clade (E. equisetina, E. major) and once in
the North American clade (E. funerea; fig. 4). Like the species
with transversely lamellar seeds, there is no ecological correla-
tion between species with papillate seeds. Ephedra equisetina
and E. major have fleshy cone bracts, whereas E. funerea has
dry winged bracts. Interestingly, however, there is some sup-
port for an evolutionary link between papillae and transverse
Fig. 3 Seed envelope in transverse section. Scale bars: A, 900 mm;
B, 300 mm; C,30mm. A, Smooth envelope of Ephedra ochreata
(Ickert-Bond 1257 [ASU]); B, papillate envelope of Ephedra funerea
(Baker 13971 [ASU]); C, transverse lamella in Ephedra torreyana
(Porter and Porter 8998 [S]).
45
ICKERT-BOND & RYDIN—MICROMORPHOLOGY OF THE EPHEDRA SEED ENVELOPE

Fig. 4 Seed shape in longitudinal section (squares) and micromorphological patterns of the seed envelope (circles) mapped on a phylogeny of
Ephedra (phylogeny redrawn, with permission, from work by Rydin and Korall [2009]).

lamellae. Ephedra funerea (papillate seeds) is sister to one of
the two accessions assigned to E. torreyana (transverse lamel-
lar), and similarly, E. equisetina and E. major (papillate seeds)
belong in the same clade as E. rhytidosperma (transverse la-
mellar; fig. 4; Rydin and Korall 2009). The wartlike projec-
tions observed in a few species (E. pachyclada [fig. 1L], E.
lomatolepis, and one specimen of E. major) are typically
larger than the papillae of E. equisetina, E. funerea, and E.
major but very thin, and some appear to have collapsed. The
development and histology of these structures are not known
to us; they are not observed in serial sections (Rydin et al.
2010). They may be the remains of collapsed papillae but
could also potentially be an artifact caused, for example, by
material degradation.
Further, there are intermediate forms of lamellae and papil-
lae. Two specimens of E. torreyana have papillae-like lamellae
as well as distinct papillae (fig. 1J), indicating that the lamellae
are formed by fused papillae. Hybridization and/or introgres-
sion may also be responsible for some seed surface patterns.
Wendt (1993) found clear indications of hybridization between
E. torreyana var. powelliorum and E. aspera at the type local-
ity of E. torreyana var. powelliorum in Big Bend National
Park, Texas, and our data support this interpretation. Our
specimen of E. aspera from the Big Bend Basin (see table 1) has
a papillate seed envelope (not shown), whereas all other speci-
mens of E. aspera investigated here lack papillae. Similarly, in
one sample from Mount Helan, China, identified as E. equise-
tina, we found a pattern of transverse lamellae with papillae
on the lamellae (fig. 1K), thus combining features otherwise
characteristic of E. rhytidosperma and E. equisetina, respec-
tively. The specimen could have originated from hybridization,
since these two species occur in close proximity at Mount Helan.
Species determinations/delimitations in Ephedra are often
uncertain or incorrect (e.g., Freitag and Maier-Stolte 1994;
Ickert-Bond and Wojciechowski 2004; Rydin and Korall
2009). Morphological variation is limited, and variable char-
acters, such as growth habit, pollen morphology, and leaf and
cone morphology, show substantial parallelism and/or intra-
specific variation (Foster 1972; El-Ghazaly and Rowley 1997;
Ickert-Bond et al. 2003; Ickert-Bond and Wojciechowski
2004; Huang et al. 2005). Micromorphology of the seed enve-
lope exhibits a similar pattern. Because of a relatively low
amount of variation, evident parallel evolution of similar fea-
tures, and intraspecific variation of traits, micromorphology
of the seed envelope of Ephedra is generally not useful for spe-
cies identification or subclade delimitation.
The Fossil Record and the Age of the Extant Clade
The surface sculpturing of the seed envelope of extant E.
rhytidosperma is indeed very similar in size and shape to that
described for the Early Cretaceous compression fossil E. ar-
chaeorhytidosperma (Yang et al. 2005). However, from our
study it is clear that similar patterns of ridged seeds are not
unique to these two species but have evolved independently at
least twice in the extant clade. Moreover, the micromorpho-
logical patterns (e.g., the transverse lamellae) of extant species
are formed by the epidermis alone (this study; Yang 2002; Ry-
din et al. 2010). In the fossil, however, it is unlikely that the
ridged pattern of the seed envelope would have been preserved
if formed only by a thin and soft tissue, such as the epidermis.
Although epidermal cell patterns can often be preserved in
fossil cuticles, these are rarely present in plant fossils from the
Yixian Formation (e.g., Zhou et al. 2003; Rydin et al. 2006b)
due to rapid decay during the initial fossilization process and
replacement of organic tissue with pyrite microcrystallines
(Leng and Yang 2003).
A preserved ‘‘ridged’’ pattern, formed by sclerenchymatous
tissue of the mesophyll, occurs in other Early Cretaceous fos-
sils with a probable gnetalean affinity (Rydin et al. 2006a;
Friis et al. 2007, 2009). The mesofossils E. portugallica Ry-
din, Pedersen, Crane et Friis and E. drewriensis Rydin, Peder-
sen, Crane et Friis (Rydin et al. 2006a) appear to have had
smooth seed envelopes, but other seeds of uncertain affinity
(e.g., fig. 7G,7H in Rydin et al. 2006a) have transverse lamel-
lae. The epidermis and perhaps also parts of the mesophyll are
typically abraded in these fossils (but sometimes preserved in
small areas in the apical-most region; C. Rydin, personal ob-
servation).
Based on comparative anatomy and histology, Rydin et al.
(2010) found that the Cretaceous mesofossils E. portugallica
and E. drewriensis (Rydin et al. 2004, 2006a) are extinct
members of the stem lineage of Ephedra. They hypothesized
a reduction of the vasculature of the seed envelope from prob-
ably four vascular bundles in the fossils to three in ancestral
members of the crown group and to two in some extant
forms. This morphological transformation series was thought
to reflect the evolutionary origin of the ephedran seed enve-
lope from a pair of cone bracts. Using the conclusions in
Rydin et al. (2010) to assess the phylogenetic position of E.
archaeorhytidosperma would also support the fossil being
a stem relative of the extant clade. In Ephedra, the number of
vascular bundles is typically strongly correlated with the api-
cal, transverse shape of the seed envelope (Rydin et al. 2010).
While the seeds of E. archaeorhytidosperma appear to have
four angles (C. Rydin, personal observation), that is, four vas-
cular bundles in the seed envelope, extant E. rhytidosperma is
nested within the Asian clade, in which species have two vas-
cular bundles in the seed envelope (Rydin et al. 2010). It is in
fact even difficult to firmly assign E. archaeorhytidosperma to
the ephedran lineage, since synapomorphies of the extant
clade (Ephedra-type pollen and apical papillae on the seed en-
velope) have not been observed in the fossil.
Thus, in spite of a striking and indisputable similarity with
extant Ephedra, the fossils appear to have been components
of a Cretaceous diversity that became largely extinct toward
the end of the period. Findings in several previous studies lend
support to this hypothesis. Crane and Lidgard (1989) demon-
strate a substantial decline in diversity and abundance of
ephedroid pollen toward the early Late Cretaceous. Phyloge-
netic and temporal analyses (Ickert-Bond et al. 2009; Rydin
and Korall 2009) and comparative anatomy/morphology
(Rydin et al. 2010; and this study) concomitantly support that
extant diversity is the result of a (second) radiation in Ephe-
dra, which presumably began in the Paleogene.
Acknowledgments
We thank the curators of the ARIZ, ASU, E, K, KAS, MO,
N, NY, PE, S, TEX, UC, UCR, UPS, and Z herbaria for access
47
ICKERT-BOND & RYDIN—MICROMORPHOLOGY OF THE EPHEDRA SEED ENVELOPE

to plants; James A. Doyle and an anonymous reviewer for
valuable comments on the text; and John Benedict, Monte
Garroutte, Zachary Meyers, and Yong Yang for technical as-
sistance. This work was supported by grants from the Swedish
Research Council to C. Rydin and a National Science Founda-
tion grant (Collaborative Research: Gymnosperms on the Tree
of Life: Resolving the Phylogeny of Seed Plants, NSF-
0629657) to S. M. Ickert-Bond.
Literature Cited
Benedict JC, KB Pigg, ML Devore 2008 Hamawilsonia boglei gen. et
sp. nov. (Hamamelidaceae) from the late Paleocene Almont flora of
central North Dakota. Int J Plant Sci 169:687–700.
Crane PR, SH Lidgard 1989 Paleolatitudinal gradients and temporal
trends in Cretaceous floristic diversity. Science 246:675–678.
El-Ghazaly G, JR Rowley 1997 Pollen wall of Ephedra foliata.
Palynology 21:7–18.
Erbar C 1995 On the floral development of Sphenoclea zeylanica
(Sphenocleaceae, Campanulales): SEM investigations on herbarium
material. Bot Jahrb Syst 117:469–483.
Foster AS 1972 Venation patterns in the leaves of Ephedra. J Arnold
Arbor Harv Univ 53:364–378.
Freitag H, M Maier-Stolte 1994 Characterization of areas, Ephedra-
ceae. Pages 5–16 in K Browicz, ed. Chorology of trees and shrubs in
south-west Asia and adjacent regions. Vol 10. Polish Academy of
Sciences, Institute of Dendrology, Kornik.
Friis EM, PR Crane, KR Pedersen, S Bengtson, PCJ Donoghue, GW
Grimm, M Stampanoni 2007 Phase-contrast x-ray microtomogra-
phy links Cretaceous seeds with Gnetales and Bennettitales. Nature
450:549–552.
Friis EM, KR Pedersen, PR Crane 2009 Early Cretaceous mesofossils
from Portugal and eastern North America related to the Bennettitales-
Erdtmanithecales-Gnetales group. Am J Bot 96:252–283.
Fu LK, YF Yu, H Riedl 1999 Ephedraceae. Page 98 in CY Wu, P
Raven, eds. Flora of China. Vol 4. Science, Beijing.
Guo S, X Wu 2000 Ephedrites from latest Jurassic Yixian Formation
in western Liaoning, northeast China. Acta Palaeontol Sin 39:
81–91.
Hollander JL, SB Vander Wall, JG Baguley 2010 Evolution of seed
dispersal in North American Ephedra. Evol Ecol 24:333–345.
Huang J, DE Giannasi, RA Price 2005 Phylogenetic relationships in
Ephedra (Ephedraceae) inferred from chloroplast and nuclear DNA
sequences. Mol Phylogenet Evol 35:48–59.
Ickert-Bond SM, C Rydin, SS Renner 2009 A fossil-calibrated
relaxed clock for Ephedra indicates an Oligocene age for the
divergence of Asian and New World clades and Miocene dispersal
into South America. J Syst Evol 47:444–456.
Ickert-Bond SM, JJ Skvarla, WF Chissoe 2003 Pollen dimorphism in
Ephedra L. (Ephedraceae). Rev Palaeobot Palynol 124:325–334.
Ickert-Bond SM, MF Wojciechowski 2004 Phylogenetic relationships
in Ephedra (Gnetales): evidence from nuclear and chloroplast DNA
sequence data. Syst Bot 29:834–849.
Igersheim A, O Cichocki 1996 A simple method for microtome
sectioning of prehistoric charcoal specimens, embedded in 2-
hydroxyethyl methacrylate (HEMA). Rev Palaeobot Palynol 92:
389–393.
Krassilov VA 1982 Early Cretaceous flora of Mongolia. Palaeontogr
B 181:1–43.
Kubitzki K 1990 Ephedraceae. Pages 379–382 in K Kubitzki, ed.
The families and genera of vascular plants. I. Pteridophytes and
Gymnosperms. Springer, Berlin.
Leng Q, H Yang 2003 Pyrite framboids associated with the
Mesozoic Jehol biota in northeastern China: implications for
microenvironment during e arly fossilization. Prog Nat Sci 13:206–
212.
Mohr BAR, ME Bernardes-de-Oliveira, AMF Barreto, MC Castro-
Fernandes 2004 Gnetophyte preservation and diversity in the
Early Cretaceous Crato Formation (Brazil). 7th International
Organisation of Palaeobotany Conference, Bariloche, Argentina,
March 21–26. Abstracts, p 81.
Rydin C, A Khodabandeh, PK Endress 2010 The female reproductive
unit of Ephedra (Gnetales): comparative morphology and evolu-
tionary perspectives. Bot J Linn Soc 163:387–430.
Rydin C, P Korall 2009 Evolutionary relationsh ips in Ephedra
(Gnetales), with implications for seed plant phylogeny. Int J Plant
Sci 170:1031–1043.
Rydin C, KR Pedersen, PR Crane, EM Friis 2006a Former diversity
of Ephedra (Gnetales): evidence from Early Cretaceous seeds from
Portugal and North America. Ann Bot 98:123–140.
Rydin C, KR Pedersen, EM Friis 2004 On the evolutionary history of
Ephedra: Cretaceous fossils and extant molecules. Proc Natl Acad
Sci USA 101:16571–16576.
Rydin C, S Wu, EM Friis 2006b Liaoxia (Gnetales): ephedroids from
the Early Cretaceous Yixian Formation in Liaoning, northeastern
China. Plant Syst Evol 262:239–265.
Sun G, S Zheng, DL Dilcher, Y Wang, S Mei 2001 Early angiosperms
and their associated plants from western Liaoning, China. Shanghai
Scientific and Technological Education, Shanghai.
Takaso T 1985 A developmental study of the integument in
gymnosperms. 3. Ephedra distachya L. and E. equisetina Bge.
Acta Bot Neerl 34:33–48.
Wendt T 1993 A new variety of Ephedra torreyana (Ephedraceae)
from west Texas and Chihuahua, with notes on hybridization in the
E. torreyana complex. Phytologia 74:141–150.
Yang Y 2002 Systematics and evolution of Ephedra L. (Ephedraceae)
from China. PhD diss. Chinese Academy of Sciences, Beijing.
——— 2007 Asymmetrical development of biovulate cones resulting
in uniovulate cones in Ephedra rhytidosperma (Ephedraceae). Plant
Syst Evol 264:175–182.
Yang Y, B Geng, DL Dilcher, Z Chen, TA Lott 2005 Morphology and
affinities of an Early Cretaceous Ephedra (Ephedraceae) from
China. Am J Bot 92:231–241.
Zhou Z 2006 Evolutionary radiation of the Jehol biota: chronolog-
ical and ecological perspectives. Geol J 41:377–393.
Zhou Z, PM Barrett, J Hilton 2003 An exceptionally preserved
Lower Cretaceous ecosystem. Nature 421:807–813.
48
INTERNATIONAL JOURNAL OF PLANT SCIENCES