«There have been many proposals of candidates for the ancestors or closest relatives of angiosperms. Some of the currently more frequently cited ...»
Suggested Angiosperm Ancestors
There have been many proposals of candidates for the ancestors or closest relatives
of angiosperms. Some of the currently more frequently cited examples are introduced here. Although none of them has been conﬁrmed to be closely related to
angiosperms, a comparison between them and angiosperms helps to identify where
the problem and gaps in knowledge are. It is these candidates and their possible
relationships to angiosperms that compose the foundation on which the current systematics of seed plants is based. Understanding them is also helpful to make a balanced judgment of the point of view in this book.
At one time or another almost all gymnosperms, and even ferns, have been proposed either as angiosperm ancestors or as their close relatives by various scholars based on various reasons in the past century (Maheshwari 2007). Even today some systematic botanists still favor some of these views. There is currently no consensus as to which of the several fossil taxa most appropriately bridges the gap between angiosperms and gymnosperms, and most of the dawn angiosperms documented later in this book appear to fall well within the scope of angiosperms.
The suggested angiosperm ancestors or close relatives of angiosperms therefore still very much deserve our attention with regard to understanding angiosperm origin. They can help us to trace the development of the science, and constitute the background from which this book originates. Here I brieﬂy introduce Gnetales, Gigantopteris, Sanmiguelia, Leptostrobus, Caytonia, Bennettitales, Umkomasia, Problematospermum, Dirhopalostachys, Ktalenia, and Pentoxylales, as examples among many (Fig. 2.1), and discuss their similarities to as well as differences from angiosperms. To a few taxa new information will be added.
2.1 Gnetales Among living plants the Gnetales (Ephedra, Gnetum, and Welwitschia) are a group considered currently by many to be most closely related to the angiosperms. Gnetum lives today in tropical forests, while Ephedra and Welwitschia are dry-climate or desert plants. These three genera in Gnetales are fairly isolated from each other although they share many synapomorphies, including multiple axillary buds, X. Wang, The Dawn Angiosperms, Lecture Notes in Earth Sciences 121, DOI 10.1007/978-3-642-01161-0_2, C Springer-Verlag Berlin Heidelberg 2010 6 2 Suggested Angiosperm Ancestors Fig. 2.1 Approximate temporal distribution of the taxa discussed in this chapter. Dashed lines indicate uncertainty opposite and decussate phyllotaxy, vessel elements, circular bordered pits in protoxylem, a terminal ovule with two integuments, lack of archegonia, ribbed pollen (except for Gnetum) and anastomose (except for Ephedra) (Eames 1952; Zhang and Xi 1983; Crane 1996; Ickert-Bond et al. 2003; Maheshwari 2007). A micropylar tube is another common feature shared by these three genera, uniquely in living gymnosperms (Bierhorst 1971). Recent studies indicate that the micropylar tube is a feature seen in the Bennettitales-Erdtmanithecales-Gnetales clade (Friis et al.
2009). Gnetales appear to have had their greatest diversity in the past, and Ephedralike pollen alone once accounted for up to 10–20% of palynoﬂoral assemblage in northern Gondwana Province in the Middle Cretaceous (Brenner 1976). Gnetalean pollen grains also possibly occurred in the Permian (Delevoryas 1962; Wang 2004).
Recent more megafossils most likely related to Gnetales have been found from the Early Cretaceous in South America and China (Rydin et al. 2003, 2004, 2006a; Tao and Yang 2003; Dilcher et al. 2005; Yang et al. 2005; Guo et al. 2009; Wang and Zheng 2010). Gnetales are frequently associated with other anthophytes including angiosperms in phylogenetic analyses (Thompson 1916; Crane 1985). The Gnetales
are characterized by a suite of characters allying them closely to the angiosperms:
eudicot-like venation, relict bisexuality, two integuments, pollen tube, vessel elements, and “endosperm” development after fertilization (Arber and Parkin 1908;
2.2 Gigantopteriales 7 Eames 1952; Carlquist 1996; Chamberlain 1957; Martens 1971; Friedman 1990, 1991, 1992a, b; Biswas and Johri 1997; Doyle 1998; Yang et al. 2000; Rydin and Friis 2005). In addition, double fertilization, a phenomenon formerly thought restricted to angiosperms, is also found in Ephedra (Chamberlain 1957; Martens 1971; Friedman 1990, 1991, 1992a; Yang et al. 2000; Friedman and Williams 2004; Raghavan 2005). Despite all these similarities, however, there are still big gaps between Gnetales and angiosperms, for example, in terms of reproduction: in Gnetales the pollen grains are captured by a fertilization droplet and drawn in it to the ovule surrounded by integuments while in angiosperms pollen grains typically germinate on the stigma and sperms are conveyed to the ovule via the pollen tube (Chamberlain 1957; Eames 1961; Bierhorst 1971; Friedman 1992a, 1993; Biswas and Johri 1997; Friedman and Barrett 2008). Moreover, there are molecular data suggesting that Gnetales may actually be more closely related to Pinaceae than to angiosperms (Soltis et al. 2002; Qiu et al. 2007; Rydin and Korall 2009; For more, see Chap. 8).
The gigantopterids (Fig. 2.2) are an enigmatic plant group from the Lower Permian to Triassic of southeastern Asia and southern North America. Their stems and cuticle have been studied anatomically (Yao and Crane 1986; Li et al. 1996; Li and Taylor 1998, 1999; Wang 1999), but reproductive organs remain elusive in spite of the reconstruction based on various fossil materials (Li and Yao 1983; Li 1992).
Gigantopterid megaphylls are characterized by pinnate venation, with tertiary anastomosing veins and giving rise to higher order veins that may anastomose again and form meshes. Their leaf organization is so similar to angiosperms that Glasspool Fig. 2.2 Leaf morphology, venation and vessel elements of Gigantopterids. A Leaf morphology of Gigantonoclea (IBCAS). B Venation of Gigantonoclea rosulata (PB4969, NIGPAS). C Vessel elements of Vasovinea tianii (Courtesy of Dr. Hongqi Li) 8 2 Suggested Angiosperm Ancestors et al. (2004) prefer to describe them using angiospermous terms although they rule out any relationship between them and angiosperms. These foliar features however were used by Asama (1982) to suggest that angiosperms in his view are derived from gigantopterids. The most intriguing feature of gigantopterids is undoubtedly that they are similar to angiosperms not only in leaf morphology and physiognomy but also in vessel elements in their wood (Li et al. 1994, 1996; Li and Taylor 1998, 1999). Oleanane, a chemical species formerly found only in extant angiosperms, has also been found in Gigantopterids (Taylor et al.
2006a). This discovery suggests a possible relationship among Gigantopterids and angiosperms as well as Bennettitales, as they are the only groups containing this chemical species (Taylor et al. 2006a). However, the hypothesized connection between Gigantopteridales and angiosperms is now largely discounted due to large time gap and lack of accurate information regarding their reproductive organs. The similarities between Gigantopterids and angiosperms may well represent large-scale convergence or parallelism (Glasspool et al. 2004).
Glossopteridales (Fig. 2.3) were mainly distributed on the Gondwanan continents (Taylor 1996; Biswas and Johri 1997; Taylor et al. 2007), although there are few, perhaps specious, reports from the northern hemisphere (Delevoryas 1969).
They thrived from the Late Carboniferous to the Triassic (Middle Jurassic?) (Delevoryas 1969; Taylor 1996; Biswas and Johri 1997; Taylor et al. 2007).
Common ovulate structures connected or associated with the leaves, Glossopteris, include Lidgettonia, Denkania, Scutum, Ottokaria and Dictyopteridium, and the pollen organs Eretmonia and Glossotheca with their bisaccate striate pollen of Protohaploxypinus-type (Taylor and Taylor 2009). Dadoxylon is the wood, and Vertebraria is the root (Biswas and Johri 1997). Glossopteris leaves are tongueshaped, with an entire margin, a distinct midrib, and reticulate venation. In Glossopteridales, both pollen and ovulate structures are borne on the adaxial surface of the Glossopteris leaf. Unitegmic orthotropous ovules are attached either directly to the adaxial surface of a megasporophyll (Fig. 2.3b, c) or in stalked uniovulate cupules borne on a branching system (Nishida et al. 2007; Taylor et al. 2007; Taylor and Taylor 2009). Pollen sacs develop in pedicellate clusters that arise from the Fig. 2.3 Leaf and reproductive organ of Glossopterids. A Leaf.
B Axis with a megasporophyll. C Cross section of cupule showing adaxial arrangement of seeds partially inrolled by the cupule
2.5 Leptostrobus 9 midvein of a modiﬁed leaf. The glossopterids have been suggested as potential angiosperm ancestors (Retallack and Dilcher 1981b). Theoretically, the glossopterid vegetative leaf could be homologous to an angiosperm’s carpel, and the megasporophyll to the outer integument (Retallack and Dilcher 1981b; Doyle 2008). In some Glossopteridales, the margins of the megasporophyll are laterally inrolled (Nishida et al. 2007; Taylor and Taylor 2009; Fig. 2.3c), much like an under-developed conduplicate carpel of angiosperms. Among all the alternative hypotheses on carpel origin, the glossopterid-based theory is the only one that does not need to derive any carpel part de novo, and thus would be the least troublesome in morphological terms (Retallack and Dilcher 1981b). However, this interpretation is open to debate due to the differences in pollen organs, pollen grains, leaf features, and age gap between Glossopteridales and angiosperms (Retallack and Dilcher 1981b; Taylor and Taylor 2009). Moreover, the provenance of stamens and perianth are further challenges for this interpretation. Meanwhile, it has also been suggested that the Glossopteridales are the ancestors of Caytoniales based on leaf venation, pollen grains and seed structure (Krassilov 1977b).
2.4 Sanmiguelia Sanmiguelia sensu lato is an enigmatic plant with large palm-like, pleated leaves and is found from the Middle to Upper Triassic of Colorado and Texas, USA (Brown 1956; Ash 1976; Tidwell et al. 1977; Cornet 1986, 1989b). The reconstructed plant includes the leaves (Sanmiguelia), female inﬂorescence (Axelrodia), and male inﬂorescence (Synangispadixis). Axelrodia includes two kinds of ﬂowers with “carpels” bearing apical “stigmas” and enclosing pairs of basal ovules. Synangispadixis lacks a perianth and bears hundreds of spirally arranged microsporophylls with monocolpate pollen grains. Cornet (1989b) described the transmitting tissue, cotyledons, and developmental pattern in the fossil. Despite his and others’ work, its phylogenetic position remains however both enigmatic and isolated (Friis et al. 2006).
Sanmiguelia apparently is not closely related to any known gymnosperm or fern.
It demonstrates certain similarities to monocots, such as leaf venation, ovule/seed developmental pattern, and leaf morphology. However, its relationship to other groups of plants, including angiosperms, cannot be assured until more fossils bridging the gaps between Sanmiguelia and other plants are found.
Leptostrobus (Czekanowskiales) is widely distributed in the Triassic to Cretaceous of the Laurasian continents and Australia (Liu et al. 2006b). It consists of an axis bearing numerous short-stalked, spirally arranged bivalved capsules containing many seeds (Krassilov 1977a; Liu et al. 2006b; Fig. 2.4). The capsule valves have papillate ﬂanges (or lips), which may have functioned like stigmatic bands (Krassilov 1977a, Fig. 2.4c). Each valve bears 3–5 seeds (Liu et al. 2006b;
10 2 Suggested Angiosperm Ancestors Fig. 2.4 Reproductive organ of Czekanowskiales.
A Leptostrobus, showing capsules attached to an axis.
B Longitudinal section of capsule showing two identical facing valves forming a capsule. C Interior view of a valve of the capsule showing seeds and ﬂange Figs. 2.4b, c). The ﬂange is not seen in Leptostrobus species from earlier ages, therefore its presence in younger species of the genus may be derived (Krassilov 1977a). Its leaf is Phoenicopsis-like. Krassilov (1977a) related it to monocots based on its leaf morphology and cuticular features, although he admitted that it was hard to imagine that the coalescence of the valves could result in any known angiosperm carpel.
2.6 Caytonia Caytonia is a cupulate female organ ﬁrst recognized by Thomas (1925) from the Middle Jurassic of England. More materials of Caytoniales have been subsequently found in strata ranging from the Upper Triassic to Lower Cretaceous of Greenland, Poland, Canada, Siberia, Australia, Antarctic, Japan, Sweden (Harris 1933, 1940, 1964; Reymanowna 1970, 1973; Krassilov 1977a; Nixon et al. 1994; Barbacka and Boka 2000a; b; Taylor et al. 2006b), and China (Wang 2010; Fig. 2.5). Although never found physically attached, their association is so strong that it has been widely assumed that the related leaf is Sagenopteris. The male organ is assumed to be Caytonanthus with in situ monosulcate bisaccate pollen grains, Vitreisporites (Harris 1964; Taylor et al. 2006b; Taylor and Taylor 2009). Caytoniales have an axis bearing stalked, rounded, helmet-like cupules. Each cupule is recurved, with a lip-like projection near its base, and contains 8–30 orthotropous unitegmic ovules arranged in curved rows (Nixon et al. 1994; Taylor and Taylor 2009, Wang 2010).
The cupule rim and cupule stalk form a cupule opening (Nixon et al. 1994; Fig. 2.5).
The micropyles of the ovules are connected to the cupule opening via canals (Harris 1933; Reymanowna 1970, 1973). Because Caytonia encloses its seeds completely, Thomas (1925) initially thought that it was an angiosperm and that its cupule was equivalent to the carpel of angiosperms. Its Jurassic age also made it a perfect candidate for angiosperm ancestry (Knowlton 1925; Thomas 1925). However, later research, particularly by Harris, indicates that before fertilization the ovules of Caytoniales are exposed to the outside through canals, that the fertilization in Caytoniales is completed by drawing pollen grains through the canals to the ovules presumably in exuded ﬂuid (a typical gymnospermous way). The seeds are
2.6 Caytonia 11 Fig. 2.5 Paracaytonia from the Yixian Formation (Early Cretaceous) of Liaoning, China.