| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Paleobotany |
Department of Biological Sciences, University of Alberta, Edmonton, Alberta, T6G 2E9 Canada
Received for publication February 7, 2007. Accepted for publication July 26, 2007.
ABSTRACT
Investigations of small permineralized flowers from the Middle Eocene Princeton Chert, British Columbia, Canada have revealed that they represent an extinct species of Saururus. Over 100 flowers and one partial inflorescence were studied, and numerous minute perianthless flowers are borne in an indeterminate raceme. Each flower is subtended by a bract, and flowers and bracts are borne at the end of a common stalk. Five stamens are basally adnate to the carpels. Pollen is frequently found in situ in the anthers. Examined under SEM and TEM, pollen grains are minute (6–11 µm), monosulcate, boat-shaped-elliptic, with punctate sculpturing and a granulate aperture membrane. The gynoecium is composed of four basally connate, lobed carpels with recurved styles and a single ovule per carpel. Flower structure and pollen are indicative of Saururaceae (Piperales), and in phylogenetic analyses using morphological characters, the fossils are sister to extant Saururus. The fossil flowers are described here as Saururus tuckerae sp. nov. These fossil specimens add to the otherwise sparse fossil record of Piperales, represent the oldest fossils of Saururaceae as well as the first North American fossil specimens of this family, and provide the first evidence of saururaceous pollen in the fossil record.
Key Words: Eocene • fossil flowers • magnoliids • Piperales • pollen • Princeton Chert • Saururaceae • Saururus
Piperales today are a diverse, speciose group of early-divergent angiosperms and are important for understanding early angiosperm evolution (Wanke et al., 2007
). However, there are few fossils confidently assigned to the order, inhibiting a deeper understanding of evolutionary trends. As circumscribed by APG (2003)
, Piperales include the families Aristolochiaceae, Hydnoraceae, Lactoridaceae, Piperaceae, and Saururaceae. Saururaceae are a small family consisting of four genera and six extant species: Anemopsis californica Hook. et Arnott, Gymnotheca chinensis Decaisne, G. involucrata Pei, Houttuynia cordata Thunb., Saururus cernuus L., and S. chinensis (Lour.) Baill. (Wu and Kubitzki, 1993
). These herbaceous, rhizomatous plants tend to inhabit moist to wet environments (Wu and Kubitzki, 1993
; Liang, 1995
; Xia and Brach, 1999
). Within Saururaceae, two species (S. cernuus and A. californica) are found in North America and four species (S. chinensis, H. cordata, G. chinensis, and G. involucrata) in eastern Asia (Wu and Kubitzki, 1993
). The family is likely monophyletic and is sister to Piperaceae in the order Piperales (Tucker and Douglas, 1996
; Meng et al., 2002
, 2003
; Jaramillo et al., 2004
; Neinhuis et al., 2005
). Flowers of both Piperaceae and Saururaceae are distinct from other piperalean flowers in being small and lacking a perianth. Saururaceous flowers are subtended by a bract and typically have three (Houttuynia, Anemopsis) or four (Saururus, Gymnotheca) carpels and three (Anemopsis) or six (Houttuynia, Saururus, and Gymnotheca) stamens (Wu and Kubitzki, 1993
). However, some variation in the numbers of carpels and stamens within flowers of an infloresence has been described (e.g., Raju, 1961
; Tucker, 1975
, 1981
, 1985
; Liang and Tucker, 1989
; Liang, 1994
).
To date only two fossil species are known for Saururaceae, and there are few confirmed reports of fossil Piperaceae. Fruits and seeds of Saururus bilobatus (Nikitin) Mai have been recognized from the Late Eocene to Pliocene of Europe and Siberia (Mai and Walther, 1978
; Friis, 1985
; Stuchlik et al., 1990
; Lesiak, 1994
), and seeds of Houttuynia bavarica Mai (Mai, 1999
) from the Miocene of Germany. These are known from gross morphology and have not been anatomically examined. No fossil vegetative or pollen remains have been firmly placed in the family Saururaceae. However, we recently recognized that fossil flowers from the Middle Eocene Princeton Chert, previously thought to have affinities to Alismatales (Currah and Stockey, 1991
; Stockey, 1994
, 2001
, 2006
; Pigg and Stockey, 1996
; Smith and Stockey, 2004
, 2005
), have many similarities to saururaceous flowers. Here, we describe this flower type and demonstrate that while these fossil flowers are similar to certain alismatids, floral and inflorescence structure and anther contents suggest that this flower type is related to Saururaceae (Piperales).
MATERIALS AND METHODS
Fossil material
Fossils were collected from the Princeton Chert outcrop, located on the east bank of the Similkameen River, 8.4 km south of the town of Princeton, British Columbia (Boneham, 1968
). The outcrop consists of interbedded layers of chert and coal, with occasional ash layers (Stockey, 1983
). There are about 49 major chert layers, but these split and anastomose to make approximately 70 individual beds ranging in thickness from 1 to 50 cm (Smith et al., 2006
). The Princeton Chert is part of the Princeton Group, Allenby Formation (Boneham, 1968
). A Middle Eocene age has been determined based on freshwater fish (Wilson, 1977
, 1982
), mammals (Russell, 1935
; Gazin, 1953
) and K-Ar dating (Hills and Baadsgaard, 1967
). The ash of Layer #22 is currently dated at 48.7 million yr (H. Baadsgaard, University of Alberta, personal communication, 1999).
Chert blocks were cut into slabs and studied using the cellulose acetate peel technique (Joy et al., 1956
) modified for concentrated (48%) hydrofluoric acid (Basinger and Rothwell, 1977
; Basinger, 1981
). Peels were mounted on microscope slides using Eukitt (O. Kindler, GmbH, Freiburg, Germany) xylene-soluble mounting medium. Images were captured with a PowerPhase digital scanning camera (Phase One, A/S, Fredriksberg, Denmark) and a MicroLumina digital scanning camera (Leaf Systems, Bedford, Massachusetts, USA). Photographs were processed with Adobe (San Jose, California, USA) Photoshop CS. Three-dimensional reconstructions were done using photos of serial sections (taken with a Nikon [Tokyo, Japan] Coolpix 5400) and the computer visualization software AMIRA 3.1.1 (TGS Software, San Diego, California, USA).
Electron microscopy
Scanning electron microscopy (SEM) of the fossil pollen was done using the back side of deeply etched peels. Peel sections were mounted on double-sided tape on stubs and covered with 150 Å gold using a Nanotek SEMprep II sputter coater (Prestwich, Manchester, UK). Samples were observed using a JEOL (Tokyo, Japan) 6301F (field emission scanning electron microscope). Extant pollen from herbarium sheets was examined in the same way.
Fossil pollen was prepared for transmission electron microscopy (TEM) by dissolving the acetate matrix in two changes of acetone and demineralizing in concentrated (48%) hydrofluoric acid, with distilled water rinses. Pollen from extant Saururus was sampled from herbarium material (S. cernuus: ALTA 5509, E. H. Moss s. n., 31 July 1914; S. chinensis: LSU 72527, Liang Hanxing 8709, 15 June 1987) and fixed in FPA overnight. Fossil and extant pollen was placed in 2% OsO4 for 2.0–3.5 h, rinsed again in distilled water, and embedded in Spurr's (1969)
resin, following an ethanol–propylene oxide dehydration series for fossil material and an ethanol–acetone dehydration series for extant material. Sections were cut at 60–80 nm using a diamond knife, collected on grids, and stained using uranyl actetate and lead citrate. Sections were observed using a Philips/FEI (Hillsboro, Oregon, USA) Morgagni 268 TEM.
Phylogenetics
A morphological cladistic analysis of the Princeton fossils and Piperaceae/Saururaceae was performed, which assumes a priori assignment of the fossil to Piperales. It is beyond the scope of this paper to do a broader analysis including all early-divergent angiosperms, but this should be undertaken in the future. The data matrices of Tucker et al. (1993)
, Tucker and Douglas (1996)
, and Meng et al. (2003)
were modified, and the fossil data were added to form a new morphological data set, that was then analyzed phylogenetically (Table 1, Appendix 1). Molecular data (e.g., Nickrent et al., 2002
; Neinhuis et al., 2005
) have recognized Piperaceae/Saururaceae as sister to a clade of Lactoridaceae, Hydnoraceae, and Aristolochiaceae. Analysis of the relationships of the fossil taxon among saururaceous taxa used Lactoris Phil. (Lactoridaceae), Aristolochia L., and Asarum L. (Aristolochiaceae) as outgroups. The ingroup consisted of Piper L., Peperomia Ruiz & Pavon, and Zippelia Blume for Piperaceae, and Saururus chinensis, S. cernuus, Gymnotheca, Anemopsis californica, and Houttuynia cordata for Saururaceae. Piper and Peperomia were coded as generic placeholders using polymorphic characters. Where inferences about the ancestral states could be made (e.g., stamen and carpel numbers; see Jaramillo and Manos, 2001
), these states were used to represent the genus, rather than also including more derived states, and thus some characters are monomorphic. Anemopsis and Houttuynia are monotypic, and both species of Saururus were used. Although Gymnotheca has two species, they do not vary in the floral characters coded in the matrix, and so they were treated as one terminal taxon. The only previously known fossils, fruits and seeds of Saururus bilobatus and seeds of Houttuynia bavarica Mai, were not included because of a lack of scorable characters.
|
). Heuristic searches were performed using 1000 random addition replicates with tree-bisection-reconnection (TBR) branch swapping and MULTREES on. Characters were unordered and equally weighted. Analyses were done with or without the fossil taxon. Branch support was estimated using bootstrap analyses (Felsenstein, 1985
), with 100 bootstrap replicates and the same search criteria as heuristic searches. Character state reconstructions were done using MacClade 4.08 (Maddison and Maddison, 2005
). RESULTS
Systematics
Order—Piperales Dumort
Family
Saururaceae Martynov
Genus
Saururus L.
Species
Saururus tuckerae S.Y.Smith & Stockey sp. nov.
Specific diagnosis
Inflorescence a raceme, at least 2.9 mm long and 1.0 mm diam. Flowers ca. 0.8 mm diam. Subtending bract, cup-shaped, ovate; flower-bract stalk and pedicel present. Perianth absent. Stamens five, adnate to carpels, up to 0.8 mm long, tetrasporangiate, with latrorse longitudinal dehiscence. Pollen 6–11 µm, monosulcate, boat-shaped-elliptic; sculpturing punctate. Carpels four, basally connate, tapering at apex, up to 1.2 mm long, 0.4 mm wide. Styles one per carpel, recurved. Seeds one per carpel, attached marginally near base.
Holotype
P1631 Bbot a (Figs. 1–5).
|
|
Type locality
Princeton Chert, east bank of the Similkameen River, ca. 8.4 km south of Princeton, British Columbia, Canada. Princeton Map Sheet 92 H/7 (1 : 50 000) UTM 10U FK 783724.
Stratigraphy and age
Princeton Group, Allenby Fm.; Middle Eocene
Description
Inflorescence and floral morphology
A single specimen representing the apical portion of an inflorescence and several hundred isolated flowers have been found in the chert. The inflorescence is a raceme, the preserved portion 2.9 mm long and ca. 1.0 mm in diameter (Figs. 1, 6). Flowers at the apex are very immature, and their bracts are larger than the androecium and gynoecium (Figs. 1, 3). Those at the base are more mature, with well-developed anthers and poorly developed carpels (Figs. 1, 2, 4, 5). A cup-shaped bract subtends each flower, up to 0.4 mm long, 0.8 mm wide, and 48 µm thick.
|
Androecium
There are five stamens per flower, although some cross sections show only four (Figs. 7–9). The small flower size makes reconstruction difficult because much of it may be lost in the saw cut. However, five complete flowers have been reconstructed based on consecutive sections, and all have five stamens: one adaxial median stamen and two pairs of lateral stamens. Stamens are adnate to the base of the carpel and tend to be about the same height as the carpels, at least 0.8 mm long (Figs. 10, 11). Anthers are tetrasporangiate, and most specimens retain pollen inside (Figs. 8, 9, 12). Older dehisced anthers have a longitudinal, latrorse dehiscence pattern (Fig. 13). Thickenings in the endothecium are apparent, especially toward the inside of the anther.
Pollen found within the anthers is minute, about 6–11 µm in diameter, monosulcate, boat-shaped-elliptic, and under light microscopy, appears to be psilate (Fig. 15). Scanning electron microscopy reveals that the pollen grains have punctate (or perforate, both sensu Punt et al., 1994
) sculpturing and small granula on the aperture membrane (Figs. 21, 22). Transmission electron microscopy shows that the aperture membrane is thin and the pollen wall is tectate-columellate, up to 380 nm thick (Figs. 26, 27). The tectum measures 75–130 nm thick, and the foot layer is 100–175 nm thick (Fig. 29). There are perforations in the tectum where the puncta occur on the surface (Fig. 29). Columellae are irregularly spaced in section view and 95–115 nm high (Fig. 29).
|
|
Gynoecium
Each fossil flower has four carpels, which are basally connate (Fig. 30). Carpels are wider at the base (up to 0.4 mm in diam), taper near the top (ca. 0.1 mm in diam), and are up to 1.2 mm long (Figs. 30–33). Each carpel has two lateral lobes that are apparent in cross section (Figs. 8, 30, 31). Cells of the carpel wall are generally small (Figs. 30–34), but the innermost layer is composed of large, thin-walled cells (Figs. 30, 34, 35). There is one recurved style per carpel (Figs. 32, 33), and the stigmatic surface appears to be papillate (Fig. 33, arrow). Each carpel is uniloculate with a single ovule (Figs. 30, 33, 34). Ovule attachment is marginal, toward the base of the carpel, and helically thickened tracheary elements are preserved in the funiculus of some specimens (Figs. 34, 35). There is no evidence of embryos or other internal tissues. Fungal hyphae are occasionally found inside the ovule or along the cell walls of the carpel. Spherical fungal structures up to 100 µm in diameter also occur in the outermost layers of the carpels in some specimens. No meiospores or conidia were found in these structures, and their exact nature is unclear.
|
|
) for the single most parsimonious tree with the fossil included (Fig. 36; Appendix 2). Piperaceae are supported by three synapomorphies: presence of a sessile stigma (character 20), one ovule per gynoecium (character 22), and basal placentation (character 23). Saururaceae are supported by having boat-shaped pollen (character 14). A clade of Anemopsis + Houttuynia is supported by two characters showing homoplasy: the presence of sessile flowers (character 4) and three carpels (character 17). The presence of a flower-bract stalk (character 5) and four carpels (character 16; this is homoplasious) support the Gymnotheca-Saururus clade. The Saururus clade is supported by three characters: basally connate carpels (character 18), 1–2 ovules per carpel (character 22), and marginal placentation (character 23). Only the first is nonhomoplasious. The clade formed by extant Saururus is supported by the presence of trichomes on the bract (character 6). DISCUSSION
Affinities of the fossil flowers
These fossil flowers have previously been thought to represent an undescribed alismatid taxon (Currah and Stockey, 1991
; Stockey, 1994
, 2001
, 2006
; Pigg and Stockey, 1996
; Smith and Stockey, 2004
, 2005
). Morphologically, the fossil flowers resemble some families of Alismatales (particularly Aponogetonaceae, Juncaginaceae, and Potamogetonaceae) and Piperales (Saururaceae), all of which have minute flowers borne on spikes (or racemes). Alismatid flowers, however, generally have tepals, trimerous flowers, and free carpels and stamens (Dahlgren et al., 1985
), unlike the fossil flowers described here. The zygomorphic flowers of Aponogetonaceae typically have two tepals; six stamens with longitudinal, extrorse dehiscence; and three free carpels, each with 2–12 ovules and a short style; but some species are variable in numbers of parts (Dahlgren et al., 1985
; van Bruggen, 1998
). Flowers of Potamogetonaceae are actinomorphic, usually with four tepals (adnate to the androecium), four stamens with longitudinal extrorse dehiscence, and four free carpels with a short style and single ovule (Haynes et al., 1998b
). The genus Maundia F. Muell. (Juncaginaceae) has four weakly connate carpels; but Maundia flowers have no bract, 2–4 tepals, up to eight stamens, and carpels without a style (Haynes et al., 1998a
).
Flower structure of the fossils is most similar to that seen in Saururaceae. Typical saururaceous flowers are minute, are borne on a spike or raceme, have a bract and no perianth, three or six stamens and three or four carpels (Liang and Tucker, 1990
; Wu and Kubitzki, 1993
; Igersheim and Endress, 1998
). Saururus and Gymnotheca flowers have a stalk bearing both bract and flower—the "flower-bract stalk" (sensu Liang and Tucker, 1990
). Stamens have a strong degree of adnation to carpels, except in Saururus cernuus, where stamens are free. In all species of Saururaceae, carpels are connate at least at the base, except S. cernuus, which is apocarpous (Liang and Tucker, 1990
; Igersheim and Endress, 1998
). Mature fossil flowers have a flower-bract stalk, a bract, no perianth, and five stamens basally adnate to a four-carpellate, basally connate gynoecium.
While variations on the typical floral plan might allow the inclusion of the fossil in Aponogetonaceae, Potamogetonaceae, Juncaginaceae, or Saururaceae, pollen morphology is more diagnostic of Saururaceae. Pollen of Aponogetonaceae is 21–45 µm in diameter, ellipsoidal, monosulcate, with per-reticulate exine sculpturing and supratectectal spinules (Erdtman, 1952
; Thanikaimoni, 1985
; van Bruggen, 1998
). In Potamogetonaceae, pollen is inaperturate, ellipsoid to spheroidal, 20–30 µm in diameter, with homobrochate exine sculpturing (Haynes et al., 1998b
; Erdtman, 1952
). Pollen grains in Maundia (Juncaginaceae) are globose, inaperturate, 27–30 µm in diameter (Erdtman, 1952
; Hope, 2006
). Thus, pollen grains in these alismatids are at least twice as large as the fossil pollen grains, with reticulate (rather than punctate in the fossil pollen) sculpturing. Pollen of Potamogetonaceae and Juncaginaceae is inaperturate, whereas the fossil pollen is monosulcate.
The pollen of the fossil taxon is a key feature for placing it within Saururaceae. Saururaceous pollen is characterized as minute (mostly <15 µm), boat-shaped-elliptic to globose, monosulcate with granula on the aperture membrane, and punctate sculpturing (Erdtman, 1952
; Walker, 1976
; Xi, 1980
; Takahashi, 1986
; Grayum, 1992
; Liang, 1992
; Pontieri and Sage, 1999
; Sampson, 2000
). The fossil pollen has the same features as pollen of Saururaceae. Although the fossil grains were originally thought to be fungal spores (Currah and Stockey, 1991
; LePage et al., 1994
), it is now clear that size, shape, and structure of the fossil pollen are clearly characteristic of Saururaceae.
Within Saururaceae, flowers of Saururus in particular are similar to the fossil flowers. Saururus flowers are developed on a racemose inflorescence. Flowers of Saururus have a flower-bract stalk, which is longer in S. chinensis than in S. cernuus (Liang and Tucker, 1990
) and diverges at a low angle from the inflorescence axis, and flowers are pedicellate. Flowers of Saururus chinensis have basally fused carpels (Raju, 1961
; Tucker, 1976
; Liang and Tucker, 1990
), like those in the fossil taxon, but S. cernuus is apocarpous. In Saururus flowers, there are six stamens initiated in pairs (Tucker, 1975
). In S. cernuus, the stamens have long filaments, are distinct from the gynoecium, and overtop the carpels at maturity (Raju, 1961
; Liang and Tucker, 1990
). In S. chinensis, stamens are fused partway up the gynoecium, and filaments are shorter and thicker than in S. cernuus (Raju, 1961
; Liang and Tucker, 1990
). Stamen features of the fossil flowers are most similar to those of S. chinensis. The flowers of Saururus are reported to be protogynous, with the stigma being receptive prior to anthesis, as are many magnoliids (Thien et al., 1994
, 2000
). It is difficult to determine exactly whether the fossil is protogynous or protandrous (with the stamens maturing before carpels are receptive). However, the fossil flowers are similar to those of extant Saururus in having pre-anthesis stage anthers overtopping the carpels and, after dehiscence, carpels that are somewhat taller than stamens. Extant Saururus is self-incompatible (Pontieri and Sage, 1999
), so it is possible that the stigmas are receptive before the stamens dehisce, without reducing the chances of outcrossing.
There are several differences between the fossil flowers and those of extant Saururus. The fossil flowers are smaller in size, about 0.8 mm in diameter compared to 1.4 mm diameter in S. chinensis and 1.7 mm diameter in S. cernuus (Liang and Tucker, 1990
). Flowers of extant Saururus have trichomes on the bracts and inflorescence axis, but no trichomes are seen in the fossil material. Bracts are more ovate and cup-shaped in the fossil than the narrower, elongate bracts seen in extant Saururus (Liang and Tucker, 1990
). Stamens are five in the fossil and typically six in extant Saururus. The fossil taxon also differs from extant Saururus in certain features of the pollen grain: grains of S. cernuus and S. chinensis are typically larger than those of the fossil, have fewer and smaller puncta than the fossil, and the puncta have raised edges not seen in the fossil. In addition, TEM shows that the pollen wall of the fossil specimens has a more open and regularly spaced columellate layer than in extant Saururus. Saururus cernuus has a thicker ectexine than in the fossil grains. Thus, these fossil flowers clearly fit in Saururaceae and are most similar to Saururus. However, the differences in flower size, presence/absence of trichomes, number of stamens, and pollen features between the fossil specimens and extant Saururus species warrant the description of a new species, Saururus tuckerae sp. nov.
Phylogenetics
In recent molecular analyses, Piperales are usually found to be sister to Canellales (= Winterales), in a clade with Laurales and Magnoliales: the magnoliid clade (Qiu et al., 1999
, 2000
, 2005
; Graham and Olmstead, 2000
; Nickrent et al., 2002
; Zanis et al., 2002
; APG, 2003
; Borsch et al., 2003
; Hilu et al., 2003
; Soltis and Soltis, 2004
; Graham et al., 2006
). Based on molecular data, Saururaceae are consistently accepted as a monophyletic group within the Piperales, which as circumscribed by APG (2003)
consist of Aristolochiaceae, Hydnoraceae, Lactoridaceae, Piperaceae, and Saururaceae. Although relationships within the order are still somewhat uncertain (APG, 2003
), Saururaceae and Piperaceae are always found as sister groups. Nuclear and mitochondrial data support inclusion of Hydnoraceae in this order, but delineating the relationships between Hydnoraceae, Lactoridaceae, and Aristolochiaceae requires further study (Gonzalez and Rudall, 2001
; Nickrent et al., 2002
; APG, 2003
).
Several previous studies have examined relationships within Saururaceae. Morphological data were used by Tucker et al. (1993)
, who analyzed the data in different ways. Their results found three recurrent hypotheses of relationships in Saururaceae: (1) Saururus sister to the rest of Saururaceae and Piperaceae (Saururaceae not monophyletic); (2) Saururus sister to the rest of Saururaceae, and Gymnotheca sister to Anemopsis + Houttuynia; (3) Saururus + Gymnotheca sister to Anemopsis + Houttuynia. The second topology is supported by morphological analyses (Tucker and Douglas, 1996
) and either atpB or 18S data (Jaramillo et al., 2004
). A fourth topology has resulted from analyses using nuclear genes, with Anemopsis basal (Meng et al., 2001
, 2003
). Meng et al. (2003)
further resolved the tree to Saururus + Gymnotheca sister to Houttuynia. This topology was not seen in morphological analyses or other analyses with molecular data.
However, there is growing support for the third topology [(Saururus + Gymnotheca), (Anemopsis + Houttuynia)]. Molecular analyses, using data from the plastid rbcL, atpB, matK, trnL-trnF regions and mitochondrial matR and nuclear 18S genes, tend to support this topology (Meng et al., 2002
, 2003
; Jaramillo et al., 2004
; Neinhuis et al., 2005
; Wanke et al., 2007
), although (as one might expect) relationships vary when one genus is excluded (e.g., Nickrent et al., 2002
; Qiu et al., 2005
). The morphological and combined molecular–morphological analyses by Meng et al. (2003)
also resulted in a single tree with this topology. Furthermore, the analyses using morphological data from our study, both with and without the fossil taxon included, resulted in a single most parsimonious tree with this topology. The fossil taxon, S. tuckerae, is always found in a clade with extant Saururus, supporting the placement of the fossil within this genus. For confidently resolving relationships between saururaceous taxa, we need to include G. involucrata and G. chinensis in addition to the more commonly used taxa Anemopsis, Houttuynia, and Saururus in future phylogenetic studies.
Variability of flower structure
Although Saururaceae are often described as having three or six stamens and three or four carpels, flowers with other numbers of parts have been documented. In their investigation of Gymnotheca,Liang and Tucker (1989)
noted the presence of abnormal flowers; some had five or seven stamens rather than six, or three carpels instead of the usual four. Further investigations by Liang (1994)
found that Gymnotheca flowers had anywhere from 4–8 stamens on three- or four-carpellate flowers. Liang (1994)
examined 381 flowers and found only 79% had the "typical state" of four carpels and six stamens; nearly 11% had four carpels and seven stamens, while 6% had four carpels and five stamens, as seen in the fossil flowers. In Anemopsis, flowers usually have six stamens and three carpels. Abnormal Anemopsis flowers include those with five stamens, two carpels, or unisexual staminate flowers (Tucker, 1985
). In Houttuynia, flowers at the apex of the inflorescence are unisexual and reduced (Raju, 1961
; Tucker, 1981
). Saururus flowers are also reported to have variable numbers of parts, from five to eight stamens and three or four carpels (Raju, 1961
; Tucker, 1975
). Tucker (1975)
noted that these flowers tend to be those found closest to the apex of the inflorescence. Although abnormal stamen numbers occur on the same inflorescence as typical flowers in extant taxa, with the limited data currently available for the fossils, we cannot say if this is the same in S. tuckerae. While the fossil taxon at first seems to have an unusual stamen number compared to other Saururaceae, as determined from the five complete flowers studied, five stamens are known from abnormal extant saururaceous flowers.
Ontogenetic sequences of stamen development in Piperaceae and Saururaceae are well understood; these are summarized in Jaramillo et al. (2004)
, and for Saururaceae in Liang (1994)
and Hufford (1997)
. All Piperaceae as well as Anemopsis and Houttuynia start with the initiation of two lateral stamen primordia. In contrast, the first-initiated primordia in Saururus are two median stamen primordia (Tucker, 1975
; Liang, 1994
; Jaramillo et al., 2004
). Gymnotheca initially has one adaxial median stamen primordium (Liang and Tucker, 1989
; Liang, 1994
; Jaramillo et al., 2004
). The development of stamen primordia in Gymnotheca, however, then becomes similar to that in Anemopsis, Houttuynia, Piper, and Zippelia (Jaramillo et al., 2004
). The sequence of stamen initiation in the fossil flowers is not clear. However, a five-staminate condition could be arrived at by arresting development in earlier stages; Anemopsis, Gymnotheca, and Zippelia all go through a stage with five stamen primordia, and therefore it is not a developmentally difficult state at which to arrive; thus the fossil taxon would show a derived stamen character state.
Pseudanthia
Inflorescences of Saururaceae have been described as pseudanthia because the showy basal bracts in Anemopsis, Gymnotheca involucrata, and Houttuynia give the whole inflorescence the appearance of a flower (Classen-Bockhoff, 1990
; Liang and Tucker, 1990
). Raju (1961)
first proposed that individual flowers of Saururaceae represent pseudanthia, with each small flower representing a reduced inflorescence. This scenario has been suggested for many taxa, including those in the monocot families Triuridaceae, Aponogetonaceae, Potamogetonaceae, Scheuchzeriaceae, and Juncaginaceae (Burger, 1977
; Posluszny et al., 1986
; Rudall, 2003
). Raju's (1961)
evidence for pseudanthial saururaceous flowers, such as spirally arranged floral organs and lack of stamens on the last-formed carpel, has been refuted by the work of Tucker (1976)
, Omori (1982)
, and Liang and Tucker (1990)
. Tucker (1976)
and Liang and Tucker (1990)
provide other lines of argument against the saururaceous flower being a pseudanthium. For example, Saururus flowers are bilateral, not radial, as are most inflorescences (Tucker, 1976
). The properties of the floral apex differ from the inflorescence apex as well; the floral apex produces dorsiventrally organized organs, while the inflorescence apex produces laterally organized organs and can produce atypical flowers or peloria (e.g., in Houttuynia and Anemopsis) (Liang and Tucker, 1990
). In addition, no branched inflorescences have been observed in Saururaceae (Tucker, 1976
); although floral parts show plasticity, the inflorescence structure does not, as one might expect if the flowers were actually reduced inflorescences.
Conclusions
Saururaceae have an interesting distribution, with four species native to eastern Asia (Gymnotheca chinensis, G. involucrata, Houttuynia cordata, and Saururus chinensis), one to western North America (Anemopsis californica), and one to eastern North America (S. cernuus). These plants prefer moist or wetland habitats (Wu and Kubitzki, 1993
; Xia and Brach, 1999
), a characteristic likely shared by the fossil plant described here. The presence of Saururus at Princeton provides yet another piece of evidence that the Princeton Chert preserved a wetland environment as shown by Cevallos-Ferriz et al. (1991)
. Saururus tuckerae is always found associated with Decodon allenbyensisCevallos-Ferriz and Stockey (1988)
. Layer #43 of the Princeton Chert likely preserves a marginal area of a small lake or pond, with Decodon and Saururus growing along the edge of and out into the water. Today, Decodon and Saururus do occasionally co-occur in swampy areas of the southeastern United States (Bennett, 2001
; J. Richard Abbott, University of Florida, Gainesville, personal communication, 2006), and these environments might represent a close modern analog to the Middle Eocene floral assemblage of the Princeton Chert.
The fossil record of Saururaceae, like Piperales in general, is sparse. There is no palynological record of Saururaceae (see Muller, 1981
), and Song et al. (2004)
suggest piperaceous pollen is neglected because it is geologically insignificant and the grains are minute in size. The same could be said for saururaceous pollen. If more fossil pollen records like the one presented here were found for these two groups, we could better understand the past geographic distribution and stratigraphic occurrence of this magnoliid group. The only previously described fossil Saururaceae are seeds of Houttuynia bavarica Mai, from the Lower Miocene of Germany (Mai, 1999
), and fruits and seeds of Saururus bilobatus (Nikitin ex Dorofeev) Mai, from the Upper Eocene to Pliocene of Europe and Siberia (Mai and Walther, 1978
; Friis, 1985
; Stuchlik et al., 1990
; Lesiak, 1994
). Thus, the inflorescence and flowers of S. tuckerae represent the oldest macrofossils, the only recognized fossil pollen, and the first North American record for Saururaceae.
Early Cretaceous fruits and associated pollen described as Appomattoxia ancistrophoraFriis, Pedersen and Crane (1995)
have been suggested to have affinities to Piperaceae, Saururaceae, Chloranthaceae, or Circaeaster (Circaeasteraceae, Ranunculales). Although it is similar in having a thick nexine and sculptured sulcus, Appomattoxia pollen is quite distinct from that of Saururaceae in having a continuous (not perforate) tectum, a granular to columellate infratectum, and verrucate to finely echinate (rather than smooth) tectal sculpturing (Friis et al., 1995
). Until we better understand relationships among extant magnoliids or find floral and vegetative material of Appomattoxia, relationships of this fossil will remain elusive (Friis et al., 1995
).
Another fossil that has been compared closely to Saururaceae, as well as to Piperaceae, Chloranthaceae, Disocoreaceae, and Smilacaceae, comes from the Aptian of Australia (Taylor and Hickey, 1990
). It is known from leaves with attached lateral pistillate inflorescences (Taylor and Hickey, 1990
). While the fossil resembles Saururaceae (among other families) in leaf characters and in having small apetalate flowers subtended by a bract and arranged in a spike, the presence of bracteoles (like Chloranthaceae) and single-carpeled flowers with truncate stigmas are very dissimilar (Taylor and Hickey, 1990
).
The occurrence of fossil Saururus in western North America, in combination with the fossil fruit record of Europe, shows that ancestral Saururus (and perhaps all Saururaceae) were once widespread. The cooling climate of the middle Paleogene may explain why Saururaceae became extinct from Europe and other areas, while it survived in southeast Asia and eastern North America where the climate remained relatively humid and subtropical (Liang, 1995
). Developing better search patterns and, especially, looking for minute pollen grains with the characteristic structure of Saururaceae, will help provide unequivocal evidence of the former distribution and evolutionary history for Saururaceae. The Princeton fossils show that Saururaceae, and thus Piperales, were well developed by the Middle Eocene. Specimens of Saururus tuckerae sp. nov. represent the first fossil saururaceous flowers, the first fossil pollen record, and the first North American fossil species for the family and thus will help to elucidate the evolutionary of Saururaceae and Piperales in general.
APPENDIX 1.
Characters used in phylogenetic analysis of fossil and extant Saururaceae.
).
).
).
).
).
), both Saururaceae and Piperaceae may be interpreted as having globose pollen because that of Saururaceae is somewhat intermediate in shape.
).
).
|
1 The authors thank S. Tucker (University of California, Santa Barbara), H. Nishida (Chuo University), and the Louisiana State University Herbarium (LSU) for providing extant material for comparison; M. Collinson (Royal Holloway University of London), J. A. Doyle (University of California, Davis), E. M. Friis (Swedish Museum of Natural History), S. W. Graham (University of British Columbia), A. Jaramillo (Universidade Federal do Rio de Janeiro), D. Sokoloff (Moscow State University), and R. Zetter (University of Vienna) for helpful discussions; R. Mandryk, R. Bhagnatar, and J. Scott (Advanced Microscopy Facility, University of Alberta) and G. Braybrook and R. Nair (University of Alberta) for aid in sectioning material and electron microscopy. The manuscript was improved thanks to comments from P. Rudall (Royal Botanic Gardens, Kew) and an anonymous reviewer. This work was supported in part by NSERC grant A-6908 to R.A.S., NSERC Canada Graduate Scholarship (CGS), and Alberta Ingenuity Studentship to S.Y.S. A version of this manuscript was submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the University of Alberta by S.Y.S. 
3 Author for correspondence and reprints (ruth.stockey{at}ualberta.ca
)
101 2 Current address: Department of Geology, Royal Holloway University of London, Egham, Surrey, TW20 0EX England
LITERATURE CITED
Angiosperm Phylogeny Group (APG).. 2003. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141: 399-436..[CrossRef][ISI]
Basinger J. F.. 1981. The vegetative body of Metasequoia milleri from the Middle Eocene of southern British Columbia. Canadian Journal of Botany 59: 2379-2410..
Basinger J. F. Rothwell G. W.. 1977. Anatomically preserved plants from the Middle Eocene (Allenby Formation) of British Columbia. Canadian Journal of Botany 55: 1984-1990..
Bennett D. J.. 2001. Propagation protocol for production of container Saururus cernuus L. plants. Native Plants Journal 2: 44-45..
Bernardello G. Anderson G. J. Lopez S P. Cleland M. A. Stuessy T. F. Crawford D. J.. 1999. Reproductive biology of Lactoris fernandeziana (Lactoridaceae). American Journal of Botany 86: 829-840..
Boneham R. F.. 1968. Palynology of three Tertiary coal basins in south central British Columbia. Ph.D. dissertation, University of Michigan, Ann Arbor, Michigan, USA..
Borsch T. Hilu K. W. Quandt D. Wilde V. Neinhuis C. Barthlott W.. 2003. Noncoding plastid trnT-trnF sequences reveal a well resolved phylogeny of basal angiosperms. Journal of Evolutionary Biology 16: 558-576..[CrossRef][ISI][Medline]
Buddell G. F. II Thieret J. W.. 1997. 6. Saururaceae. In Flora of North America Editorial Committee, [ed.], Flora of North America north of Mexico, vol. 3. Magnoliophyta: Magnoliidae and Hamamelidae, 37-38. Oxford University Press, New York, New York, USA..
Burger W. C.. 1977. The Piperales and the monocots. Alternate hypotheses for the origin of monocotyledonous flowers. Botanical Review 43: 345-393..
Cevallos-Ferriz S. R. S. Stockey R. A.. 1988. Permineralized fruits and seeds from the Princeton chert (Middle Eocene) of British Columbia: Lythraceae. Canadian Journal of Botany 66: 303-312..
Cevallos-Ferriz S. R. S. Stockey R. A. Pigg K. B.. 1991. The Princeton chert: evidence for in situ aquatic plants. Review of Palaeobotany and Palynology 70: 173-185..[CrossRef][ISI]
Classen-Bockhoff R.. 1990. Pattern analysis in pseudanthia. Plant Systematics and Evolution 171: 57-88..[CrossRef][ISI]
Currah R. S. Stockey R. A.. 1991. A fossil smut fungus from the anthers of an Eocene angiosperm. Nature 350: 698-699..[CrossRef]
Dahlgren R. M. T. Clifford H. T. Yeo P. F.. 1985. The families of the monocotyledons. Springer-Verlag, New York, New York, USA..
Doyle J. A.. 2005. Early evolution of angiosperm pollen as inferred from molecular and morphological phylogenetic analyses. Grana 44: 227-251..[CrossRef][ISI]
Erdtman G.. 1952. Pollen morphology and plant taxonomy. Angiosperms. Almqvist & Wiksell, Stockholm, Sweden..
Felsenstein J.. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791..[CrossRef][ISI]
Friis E. M.. 1985. Angiosperm fruits and seeds from the Middle Miocene of Jutland (Denmark). Det Kongelige Danske Videnskaberne Selskab Biologiske Skrifter 24: 1-165..
Friis E. M. Pedersen K. R. Crane P. R.. 1995. Appomattoxia ancistrophora gen. et sp. nov., a new Early Cretaceous plant with similarities to Circeaster and extant Magnoliidae. American Journal of Botany 82: 933-943..[CrossRef][ISI]
Gazin C. L.. 1953. The Tillodontia: an early Tertiary order of mammals. Smithsonian Miscellaneous Collections 121: 1-110..[Medline]
Gonzalez F. Rudall P.. 2001. The questionable affinities of Lactoris: evidence from branching pattern, inflorescence morphology, and stipule development. American Journal of Botany 88: 2143-2150..
Graham S. W. Olmstead R. G.. 2000. Utility of 17 chloroplast genes for inferring the phylogeny of the basal angiosperms. American Journal of Botany 87: 1712-1730..
Graham S. W. Zgurski J. M. McPherson M. A. Cherniawsky D. M. Saarela J. M. Horne E. S. C. Smith S. Y. Wong W. A. O'Brien H. E. Biron V. L. Pires J. C. Olmstead R. G. Chase M. W. Rai H. S.. 2006. Robust inference of monocot deep phylogeny using an expanded multigene plastid data set. In J. T. Columbus, E. A. Friar, J. M. Porter, L. M. Prince, M. G. Simpson, [eds.], Monocots: comparative biology and evolution (excluding Poales), 3-20. Rancho Santa Ana Botanic Garden, Claremont, California, USA..
Grayum M. H.. 1992. Comparative external pollen ultrastructure of the Araceae and putatively related taxa. Monographs in Systematic Botany from the Missouri Botanical Garden 43: 1-167..
Haynes R. R. Les D. H. Holm-Nielsen L. B.. 1998a. Juncaginaceae. In K. Kubitzki, H. Huber, P. J. Rudall, P. S. Stevens, T. Stützel, [eds.], The families and genera of vascular plants, vol. IV. Flowering plants. Monocotyledons: Alismatanae and Commelinanae (except Gramineae), 260-263. Springer-Verlag, Berlin, Germany..
Haynes R. R. Les D. H. Holm-Nielsen L. B.. 1998b. Potamogetonaceae. In K. Kubitzki, H. Huber, P. J. Rudall, P. S. Stevens, T. Stützel, [eds.], The families and genera of vascular plants, vol. IV. Flowering plants. Monocotyledons: Alismatanae and Commelinanae (except Gramineae), 408-415. Springer-Verlag, Berlin, Germany..
Hills L. V. Baadsgaard H.. 1967. Potassium–argon dating of some lower Tertiary strata in British Columbia. Bulletin of Canadian Petroleum Geology 15: 138-149..[Abstract]
Hilu K. W. Borsch T. Müller K. Soltis D. E. Soltis P. S. Savolainen V. Chase M. W. Powell M. P. Alice L. A. Evans R. Sauquet H. Neinhuis C. Slotta T. A. B. Rohwer J. G. Campbell C. S. Chatrou L. W.. 2003. Angiosperm phylogeny based on matK sequence information. American Journal of Botany 90: 1758-1776..
Hope G.. 2006. Australian National University Pollen Database, website: http://www.geo.arizona.edu/palynology/sem/anu.html [accessed 6 April 2006]..
Huber H.. 1993. Aristolochiaceae. In K. Kubitzki, J. G. Rohwer, V. Bittrich, [eds.], The families and genera of vascular plants, vol. II. Flowering plants. Dicotyledons: magnoliid, hamamelid and caryophyllid families, 129-137. Springer-Verlag, Berlin, Germany..
Hufford L.. 1997. The roles of ontogenetic evolution in the origins of floral homoplasies. International Journal of Plant Sciences 158: S65-S80..[CrossRef][ISI]
Igersheim A. Endress P. K.. 1998. Gynoecium diversity and systematics of the paleoherbs. Botanical Journal of the Linnean Society 127: 289-370..[CrossRef][ISI]
Jaramillo M. A. Manos P. S.. 2001. Phylogeny and patterns of floral diversity in the genus Piper (Piperaceae). American Journal of Botany 88: 706-716..
Jaramillo M. A. Manos P. S. Zimmer E. A.. 2004. Phylogenetic relationships of the perianthless Piperales: reconstructing the evolution of floral development. International Journal of Plant Sciences 165: 403-416..[CrossRef][ISI]
Joy K. W. Willis A. J. Lacey W. S.. 1956. A rapid cellulose peel technique in paleobotany. Annals of Botany, New Series 20: 635-637..
Kelly L. M.. 2001. Taxonomy of Asarum section Asarum (Aristolochiaceae). Systematic Botany 26: 17-53..[ISI]
Kubitzki K.. 1993. Lactoridaceae. In K. Kubitzki, J. G. Rohwer, V. Bittrich, [eds.], The families and genera of vascular plants, vol. II. Flowering plants. Dicotyledons: magnoliid, hamamelid and caryophyllid families, 359-361. Springer-Verlag, Berlin, Germany..
Lei L.-G. Liang H.-X.. 1999. Variations in floral development in Peperomia (Piperaceae) and their taxonomic implications. Botanical Journal of the Linnean Society 131: 423-431..[CrossRef][ISI]
LePage B. A. Currah R. S. Stockey R. A.. 1994. The fossil fungi of the Princeton Chert. International Journal of Plant Sciences 155: 828-836..[CrossRef][ISI]
Lesiak M. A.. 1994. Plant macrofossils from the Middle Miocene of Lipnica Mala (Orawa-Nowy Targ Basin, Poland). Acta Palaeobotanica 34: 27-81..
Liang H.-X.. 1992. Study on the pollen morphology of Saururaceae. Acta Botanica Yunnanica 14: 401-404 1 plate [in Chinese with English summary]..
Liang H.-X.. 1994. On the systematic significance of floral organogenesis in Saururaceae. Acta Phytotaxonomica Sinica 32: 425-432 [in Chinese with English summary]..
Liang H.-X.. 1995. On the evolution and distribution in Saururaceae. Acta Botanica Yunnanica 17: 255-267 [in Chinese with English summary]..
Liang H.-X. Tucker S. C.. 1989. Floral development in Gymnotheca chinensis (Saururaceae). American Journal of Botany 76: 806-819..[CrossRef][ISI]
Liang H.-X. Tucker S. C.. 1990. Comparative study of the floral vasculature in Saururaceae. American Journal of Botany 77: 607-623..[CrossRef][ISI]
Liang H.-X. Tucker S. C.. 1995. Floral ontogeny of Zippelia begoniaefolia and its familial affinities: Saururaceae or Piperaceae?. American Journal of Botany 82: 681-689..[CrossRef]
