TABLE IS UPDATED PERIODICALLY
Based on scanty paleontologic evidence the basal arthropod orders Collembola (springtails), Archaeognatha (bristletails), and Zygentoma (firebrats or silverfish) probably co-occurred with the earliest divisions of vascular plants (Rhyniophyta, Trimerophytophyta, and Zosterophyllophyta; Table 1, page 350, Labandeira 1998; Table 1, page 415, Labandeira 2006). Myriapods (centipedes and millipedes) probably also fed on detritus associated with Silurian and early Devonian algae, fungi, and the first vascular plants (Labandeira 1998).
Spores were probably consumed by the earliest arthropods and other detritivores. The Silurian and early Devonian Periods are the points in geologic time when "Assemblage 1: Primitive Vascular Plants and Unknown Arthropods" once existed (Labandeira 2000).
Ponomarenko (2009) reviews the contributions of Paleozoic arthropods toward development of terrestrial biomes. Coal ball permineralizations provide an extensive record of oribatid mite detrivory, mite body fossil occurrences, and evidence of insect borers, which were indigenous to Carboniferous forested swamps of North America (Labandeira et al. 1997).
As the Earth slowly warmed following the early Paleozoic icehouse (DECARB), "Assemblage 2: Ferns, Seed Plants and the Paleozoic Insect Fauna" became discernable in the fossil record (Labandeira 1998, 2000, 2006). Rocks of the Carboniferous and Permian Periods from widely separated localities yield fossils of Ephemeroptera (mayflies), Palaeodictyoptera (extinct), Plecoptera (stoneflies), Dermaptera (earwigs), Grylloblattodea (ice-crawlers), Caloneurida (extinct), Orthoptera (crickets, grasshoppers, katydids), Dictyoptera (cockroaches, mantises, termites), Lophioneurida (including psocids and bark lice), Hemiptera (plant lice, true bugs), and Coleoptera (beetles and weevils) (Table 3, above; Table 4.24, page 146, Grimaldi and Engel 2005).
Pollen- and pollenkitt-eating insects of Paleozoic Assemblage 2 included several extinct families of Grylloblattodea; hypoperlids, and psocids. Gut contents of these insects consisted of palynomorphs assignable to extinct voltzialean conifers, cordaitaleans, glossopterids, gnetaleans, and peltaspermaleans (Labandeira 2000).
Paleozoic phytophagy was not only confined to ovules, pollen, pollenkitt, seeds, spores, and wood. Gigantopterid leaves were also eaten by insects (Beck and Labandeira 1998, Glasspool et al. 2003). Foliage representing the remains of several vascular plant groups was evidently part of the diet of Permo-Carboniferous insects (Labandeira 1998, 2006).
The resurgence of terrestrial animal and plant life after recovery from the end-Permian extinction (EPE) probably led to evolution of the first seed plant pollinators (Cornet 1989). Labandeira (2000) identifies these (and other) insect-plant mutualisms as "Assemblage 3: Seed Plants and the Earlier Phase of Modern Insect Fauna," which are based on his review of paleontologic evidence from insect mouthparts and gut contents, and seed plant reproductive organs.
Data summarized in Table 3 (above) show the extinction of the Palaeodictyoptera and Caloneurida coupled with the appearance of the five new insect orders: Emboidea (appearing in the early Cretaceous Period), Thysanoptera, Hymenoptera (ants, bees, and wasps), Diptera (flies, Figure 306, page 230-231, Rasnitsyn 2002), and Lepidoptera.
There was an explosion of evolutionary diversity of Hymenoptera following the EPE, beginning with the appearance of the basal family Xyelidae in the Triassic Period (Figure 331, page 244-245, Rasnitsyn 2002). Sawflies are often preserved in the rock record together with their pollen loads (Krassilov et al. 2003).
Labandeira's "Assemblage 4: Angiosperms and the Later Phase of the Modern Insect Fauna" is discussed in the third essay on Mesozoic Paleobiodiversity of Angiosperms, which is located elsewhere on this web site.
Adaptive radiation of at least some of the pollen- and nectar-feeding insect orders following the EPE is probably attributable to evolutionary-development of innovative mouthpart designs. Specific feeding adaptations were first seen in fossil flies (labellar modifications for sponging of pollen and seed plant rewards), moths (siphons for nectar and secretions), and wasps (clasping structures used to collect pollen) (Labandeira 2000).
The first essay on the Origin of Flowering Plants paints a picture that global catastrophe might have played a role in coevolution of insects and seed plants. Molecular coevolution between arthropod and gymnosperm developmental tool kits was suggested as a way to explain evo-devo of holometabolous (larvae-forming) insects and angiosperms that resided in Permo-Carboniferous and Permo-Triassic shrub lifeboats (compartments). Several questions may be posed with these ideas in mind:
What did external biotic (host plant hormonal and secretory, or fungal and microbial) and environmental (hypercapnia, hypoxia temperature extreme) factors have to do (if anything) with evolutionary-development of larval moult cycles and innovative mouthpart designs of phytophagous insects?
Did the evolution of holometabolus insects from hemimetabolous ancestors occur as a result of developing global hypoxia as a selective force during the late Devonian-early Carboniferous ice-house and following the end-Guadalupian extinction of the late Permian Period?
Grimaldi and Engel (2005, page 333) raise the key question, "how did they [Holometabola] evolve?
The below graphic is the same crude plot of changing oxygen levels in Earth's atmosphere over geologic time based on GEOCARB III (Berner and Kothavala 2001), and other work (Beerling and Berner 2002, Huey and Ward 2005, Ward et al. 2006), which was introduced in the first essay.
I superimposed an evolutionary tree showing the major groups of Hexapoda listed in Table 3, above. The arthropod classification scheme is based on Grimaldi and Engel (2005).
Three of the four insect herbivore expansion phases outlined by Labandeira (2006) are superimposed on the graph. Red symbols denote points of divergence of the Class Insecta from other invertebrates, the point when pterygote (wing-forming) insects separated from wingless orders, the point of divergence of Neoptera (wing-folders) from Pterygota, and cladogenesis of the Holometabola (larvae-forming moulting insects).
The preceding graph is based on data from Grimaldi and Engel (Figure 4.24, page 146, 2005), with additional data added from GEOCARB III (Berner and Kothavala 2001), GEOCARBSULF (Berner 2006), and Labandeira (2006).
Arthropod stem groups. Evolutionary-developmental (evo-devo), genetic, molecular phylogenetic, and rearing studies on extant relatives of ancient arthropod assemblages include published research on 28S rRNA genes of early hexapods (Dell'Ampio et al. 2009), reproductive biological and molecular investigations of Collembola (Xiong et al. 2008, Tully and Ferrière 2008), research on hemocyanin proteins of Zygentoma (Pick et al. 2008), and improvement of phylogenetic inference using modified molecular substitution models (von Reumont et al. 2009).
Permian phytophagous insects are of interest because their occurrence in deep-time corresponds with early radiation of seed plants around the point in geologic time when molecular systematists tell us that the angiosperms split from the gymnosperms.
Hemimetabolous insects. Hemimetabolous insects including Dictyoptera (cockroaches and termites), Grylloblattida (ice crawlers), Hemiptera (aphids and true bugs), and Orthoptera (crickets and grasshoppers) are among the most poorly understood hexapod groups from a phylogenetic perspective.
Important pieces of work published in the recent literature are studies by Shao et al. (2003) on mitochondrial genes in hemipteroids, Fenn et al. (2008) on the mitochondrial genomes of Orthoptera, Murienne et al. (2008) on New Caledonian endemic dictyopterans assignable to Blattaria (cockroaches), and Urban and Cryan (2009) on phylogenetics of Fulgoridae (lanternflies). A supertree of Dictyoptera, which was recently computed by R. B. Davis et al. (2009), illustrates the evolutionary success of termites (isopterans).
A review of neotenous development in termites is available (Korb and Hartfelder 2008).
The Arkhangelsk and Perm regions of western Asia and eastern Europe and Triassic beds in South Africa continue to yield fossils of extinct insect orders including the Grylloblattida (Aristov 2009 [two papers], Aristov et al. 2009).
Triassic and Jurassic sediments of China continue to yield new species of Hemiptera assignable to Aphidomorpha (Y. Hong et al. 2009), and Procercopidae and Tettigarctidae (B. Wang and H. Zhang 2009). Early Cretaceous amber deposits in western Europe have recently yielded whole body fossils of bugs assignable to Neazoniidae (Szwedo 2009).
Panorpida and the origin of insect pollination. Permo-Carboniferous panorpoids were the ancestors of holometabolous Antliophora (Diptera [flies] and Mecoptera [scorpionflies]). When included with the Lepidoptera (moths and butterflies), Siphonoptera (fleas), and Trichoptera (caddisflies), the dipterans and mecopterids comprise the Panorpida (Grimaldi and Engel 2005). According to Grimaldi and Engel (page 469, Figure 12.1, 2005) Panorpida, which are a sister group to the Hymenoptera (ants, bees, and wasps), diverged more than 290 million years ago (blue symbols on the above graph), coincident with the angiosperm-gymnosperm split.
Based upon an analysis of preserved head and mouthparts certain scorpionflies probably fed on ovular secretions of extinct seed plant groups of the anthophyte clade (Ren et al. 2009). The study by Ren and coworkers suggests that modes of biotic coevolution of gymnosperms and pollinators were in place by the Jurassic Period of the Mesozoic Era.
Validity of the biotic pollination coevolution hypothesis has been challenged by several cogent arguments (Gorelick 2001).
Palynivores and pollinators: the "big five." The final sub-section of this chapter is a brief survey of highpoints and milestones in molecular systematic and paleontologic research on the five orders of insects that eat pollen and pollinate seed plants. The "big five" (page 241, Labandeira 2000) are Coleoptera, Diptera, Hymenoptera, Lepidoptera, and Thysanoptera.
Coleoptera. Of these five orders, the Coleoptera contain about as many species as flowering plants. Expressed sequence tag (EST) databases are being used to understand the big picture in beetle molecular systematics (J. Hughes et al. 2006). Key papers by Hunt et al. (2007) and Hunt and Vogler (2008) provide a synthesis on the Mesozoic and Cenozoic superradiation of major lineages of beetles. Use of nuclear genes that code for alpha-spectrin, arginine kinase, CAD, enolase, PEPCK, RNA polymerase II, topoisomerase, and wingless protein show promise in deducing beetle phylogeny (Wild and Maddison 2008).
The oldest beetle is known from the Carboniferous Period (Béthoux 2009).
Tremendous progress has been made in the last decade on unraveling and better understanding complex phylogenetic relationships of beetles (Farrell 1998, Bernhardt 1999). Staphylinid beetles, one of the largest groups of phytophagous coleopterans, represented by Libanoeuaesthetus (Euaesthetinae) is known from the early Cretaceous (Lefebvre et al. 2005).
Paleobiology and fossil history of the Euaesthetinae and Steninae (Staphylinidae) is reviewed by Clarke and Chatzimanolis (2009).
Molecular phylogenetic studies are underway in several labs including analyses of rRNA variation across 167 taxa of Chrysomelidae (Gómez-Zurita et al. 2008), comparative 18S and 28S rDNA sequence analysis of Chrysomelidae and Curculionoidea (Marvaldi et al. 2009), rRNA ITS2 sequence variation research in Meligethinae (Trizzino et al. 2009), 16S rDNA-based investigations on Thylacosterninae (Vahtera et al. 2009), and nuclear gene studies of Carabidae and Trachypachidae (Maddison et al. 2009).
It is widely accepted that certain phytophagous insect groups diversified together with their flowering plants hosts (Wilf et al. 2000, Grimaldi and Engel 2005, Crepet and Niklas 2009). Increasingly detailed studies however, demonstrate that beetle and weevil diversification is out-of-sync with radiations of angiosperm families and genera (Gómez-Zurita et al. 2007, McKenna et al. 2009).
Molecular phylogenies of chrysomelid leaf beetles suggest a younger age of the group than previously realized (Gómez-Zurita and Galián 2005, Gómez-Zurita et al. 2007). These findings contradict earlier proposals on a coevolutionary origin of angiosperms during the Cretaceous Period (Farrell 1998, Grimaldi 1999, Wilf et al. 2000).
Morphologically-based phylogenies of beetles yield important clues on the timing of early macroevolutionary branching events in Coleoptera (Beutel et al. 2008), and yield insight into microevolutionary insect antagonist/plant host dynamics (Borghuis et al. 2009).
Diptera. The molecular systematics and paleobiology of Diptera is reviewed by Wiegmann et al. (2003) and Labandeira (2005), respectively. Flies and the origin of flowering plants are discussed by Labandeira (1998) and Ren (1998). Fossilized mouthparts of brachyceran flies identified by Ren (1998) may indicate a close association with angiosperms: "... it is equally likely that basal brachyceran lineages of flies were pollinating anthophytes other than angiosperms ..." (page 58, Labandeira 1998).
Conrad Labandeira (2005) offers one of the more complete bibliographies on the paleobiology of dipterans.
Recent work on the molecular systematics of flies includes studies by Bertone et al. (2008) and Menguai et al. (2008).
Hymenoptera. The origin and diversification of angiosperms might be intertwined with early Mesozoic diversification of bees (Danforth et al. 2006). Danforth and coworkers (2006) have applied DNA sequence and morphological data toward a phylogeny of Andrenidae, Colletidae, Halictidae, and Stentrididae bees.
André Nel and coworkers continue to describe new groups of bees and wasps that provide phylogenetic insight on Hymenoptera to minimum age map early branch-points computed from molecular data (Michez et al. 2009, Perrichot et al. 2009). African bees assignable to Anthophila are reviewed by Kuhlmann (2009).
Genomic studies by Dearden et al. (2006) reveal that honey bees and fruit flies are the most widely-diverged species of Holometabola. Colletid, halictid, megachilid, melittid bee and wasp phylogenies have been computed from genetic and/or morphological data (Patiny et al. 2008, Praz et al. 2008, Almeida and Danforth 2009, Danforth et al. 2006).
Papers published by Brady et al. (2006), Moreau et al. (2006), Bacci et al. (2009), and J. A. Russell et al. (2009) are gateways to the vast literature on the evolution of herbivory and phylogenetic relationships of ants.
Lepidoptera. Butterflies and moths figure prominently in classic papers on coevolution and radiation of angiosperms with insects (Raven 1977, Labandeira et al. 1994). However, despite the possible importance of lepidopterans toward the explosive radiation of basal angiosperms, eudicots, monocots, rosids, and asterids, relatively little is known of phylogenetic relationships of butterfly and moth lineages in deep-time (see Figure 297, page 224-225, Rasnitsyn 2002).
Milestones in ongoing research on the phylogenetic systematics of families and tribes of moths and butterflies (Lepidoptera) include studies by J. S. Miller et al. (1997), Lopez-Vaamonde et al. (2006), Regier et al. (2008), Silva-Brandão et al. (2008), Warren et al. (2008), Kawahara et al. (2009), Pohl et al. (2009), and Wahlberg et al. (2009 [two papers]).
Thysanoptera. Oldest members of aeolothripids and thripids (Thysanoptera) have been reported from lower Cretaceous sediments of Transbaikal, Asia (Shmakov 2009). Early Cretaceous sediments have yielded a novel family of Thysanoptera described as Moundthripidae by P. Nel et al. (2007).
There are many taxonomic papers in the literature on the paleoentomology of Baltic and Rovno amber deposits, among others. Cenozoic arthropod paleobiology is beyond the scope of the present work.
I now outline a Paleozoic and Mesozoic fossil history of the major seed plant orders necessary to understand the paleobotany of angiosperm origins.
Gigantopterids:
One of the missing seed plant groups in the many published phylogenetic reconstructions of seed plants is the abundant and morphologically diverse group of Carboniferous and Permian pteridosperms generically-termed gigantopterids.
A review by Booi et al. (2009) supports a prevailing view among paleobotanists that gigantopterids are a polyphyletic group composed of several late Paleozoic families of seed plants incertae cedis characterized by large megaphylls with reticulate venation.
Interestingly, gigantopterids classified in the orders Gigantopteridales and Gigantonomiales are two of several extinct gymnosperm orders discussed in literature as possible ancestors of flowering plants (T. N. Taylor et al. 2009).
I lump Paleozoic spermopteroid seed plants Archaeocycas whitei (page 8, Mamay 1976), Eophyllogonium cathayense (Mei et al. 1992), Phasmatocycas bridwellii (Axsmith et al. 2003), and Sobernheimia jonkeri (Kerp 1983) with gigantopterids for now.
Potential importance of gigantopteroid seed plant fossils in deciphering the ancestry of flowering plants and paraphyletic clades of gymnosperms reflects my choice of the unusual title for this web site. The term "gigantopteroid" was first adopted by Mamay et al. (1984) to pigeon-hole the youngest Permian plant megafossils known at that time.
The reproductive anatomy, taxonomy, and evolutionary relationships of enigmatic gigantopteroid seed plants is problematic (Booi et al. 2009). Presence of woody leaf midribs and cuticles on large leaves up to 30 centimeters in length (Mamay et al. 1988), and sparse reproductive material (X. Li and Z. Yao 1983), suggests that gigantopteroids are gymnospermous (T. N. Taylor et al. 2009).
Little is known of the whole plant morphology of Carboniferous and Permian gigantopterids although some paleobotanists have deduced a vine habit for these gymnosperms incertae cedis (H. Li et al. 1994, H. Li and D. W. Taylor 1999). A possible connection with the enigmatic pteridosperm order Peltaspermales has been recently suggested (T. N. Taylor et al. 2009).
The place and time to begin a survey of gigantopterids is in Cathaysian rocks of the Permian Period of China (H. Xilin et al. 1996). The Upper Paleozoic Era consists of two geologic periods: Carboniferous (Mississippian and Pennsylvanian) and Permian.
In Paleozoic times gigantopterids were a diverse group of probably unrelated vascular plants constituting one of several terrestrial vegetation types of the biogeographic provinces of Angara, Cathaysia, and North America (Read and Mamay 1964, X. Li and Z. Yao 1982, Mamay et al. 1984).
The kodachrome to the right is a nearly complete leaf of Delnortea abbottiae (USNM 364419). The image was captured on film the day the rock slab was unearthed and collected from the Lower Permian (Leonardian) Cathedral Mountain Formation.
Gigantopterids had several morphologic features: woody midribs, erect vernation, abscission layering at the base of petioles, vessels, and libriform fibers (X. Li and Z. Yao 1983, Mamay et al. 1988, H. Li and Tian 1990, H. Li et al. 1994, H. Li et al. 1996, R. Weber 1997, H. Li and D. W. Taylor 1998, 1999), remarkably similar to characters found in bennettitaleans, gnetophytes, and certain extant flowering plants.
Reproduction in gigantopterids is incompletely known (X. Li and Z. Yao 1983, R. Weber 1997).
Taxon-specific biomarkers (TSBs) known as oleonone triterpanes are documented from gigantopterid fragments preserved in coal ball permineralizations, and in leaf compressions (Moldowan and Jacobson 2002). Oleananes are also TSBs for angiosperms and Bennettitales. Oleonone triterpanes have been isolated from several gigantopterid leaf specimens, fossilized bennettitalean foliage, pyrolized extant angiosperm plant material, and from leaf compressions of flowering plants (D. W. Taylor et al. 2006).
Were these seed plants forerunners of anthophytes (including flowering plants) or eudicot angiosperm analogs that diminished in numbers at the close of the Permian Period?
Known gross morphological characters of gigantopterids, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--unknown, possibly shrub- and/or vine-like (H. Li et al. 1994, H. Li and D. W. Taylor 1999), stems possess a bifacial cambium with vessels (H. Li and D. W. Taylor 1998, 1999); additional paleontologic data are needed to reconstruct whole plants and nodal anatomy
REPRODUCTIVE MODULES--inadequate data, possibly phyllospermous (X. Li and Z. Yao 1983, Mamay et al. 1988, R. Weber 1997), details of ovule attachment to laminar microsporophylls are needed, internal anatomy of ovules unknown, anatomy of microsporangia and pollen are unknown; anatomical and developmental details of sexual reproduction unknown, permineralizations are in need of discovery and study
LEAVES--dicot or Gnetum-like with simple with flaring petioles and four orders of reticulate venation (Mamay et al. 1988) or Calophyllum-like (R. Weber 1997); woody midribs, and abscission zones (Mamay et al. 1988); or compound pinnate with spines, reticulate venation, waxy cuticles, sclerenchyma, secretory ducts, and certain conducting tissues. Attachment details of leaves to stems are unclear (Z.-Q. Yao and P. R. Crane 1986, H. Li and Tan 1990, H. Li et al. 1994, H. Li et al. 1996, H. Li and D. W. Taylor 1998, H. Li and D. W. Taylor 1999, Z.-Q. Wang 1999, Glasspool et al. 2004, Z.-Q. Yao and Liu 2004, Booi et al. 2009)
Despite the abundance of Cathaysian gigantopterid compressions and impressions summarized by X. Li and Z. Yao (1982), X. Li and Z. Yao (1983), H. Xilin et al. (1996), Z.-Q. Wang (1999), Glasspool et al. (2004), and Booi et al. (2009), study of heretofore undiscovered permineralized leaf fossils would yield more useful information
PHYTOPHAGOUS ASSOCIATE(S)--arthropods belonging to the Caloneurodea, Orthoptera, Protorthoptera (Beck and Labandeira 1998); vertebrate coprolites require discovery and study, known leaf bite- and chew marks need analysis, more fossils are needed for study
PLANT IDENTIFICATION--Cathaysiopteris, Gigantopteridium, Zeilleriopteris (Labandeira 1998); Gigantonoclea hallei, Gigantonoclea lagrelii (Glasspool et al. 2003)
HOST SEED PLANT ORGAN(S) BEING EATEN--leaves (Beck and Labandeira 1998, Glasspool et al. 2003), permineralizations of plant tissue, tetrapod coprolites, and insect frass require discovery and study
Detached pieces of taeniopteroid-type Paleozoic leaves could belong to gigantopteroid plants having dimorphic foliage, as inferred from poorly preserved compressions and impressions discussed in a progress report by X. Li and Yao (page 321, 1983), and later illustrated by X. Li and Yao (page 26 and Plate 4, 1983), Mei et al. (page 101, Plate 2, 1992), and R. Weber (page 232, Plate 3, 1997).
Clearly, detailed studies of permineralized material of fertile gigantopteroid leaves, when unearthed, thin-sectioned, and described, would shed light on these enigmatic forms and help us to decipher their evolutionary relationships with sympatric Paleozoic congeners.
Delnortea and American gigantopteroids. Permian rocks of North- and South America yield several species of gigantopteroids including Cathaysiopteris yochelsonii (Mamay 1986), Delnortea abbottiae (Mamay et al. 1986), Evolsonia texana (Mamay 1989), Gigantonoclea sp. (Mamay 1986, 1988), Gigantopteridium americanum (Koidzumi 1936), Lonesomia mexicana (R. Weber 1997), and Zeilleropteris wattii (Mamay 1986). However, the late Paleozoic floral zones of the Permian of the western hemisphere including Alaska (Mamay and Read 1984), though dominated at least in part, by Gigantopteridaceae, are paleofloras without Gigantopteris (Mamay et al. 1988).
Large leaf compressions and permineralizations including a permineralized leaf midrib (see left) of Lower Permian (Leonardian) plants were described about 20 years ago (Mamay et al. 1986, 1988). Delnortea abbottiae is now known from North American and South American sedimentary beds (Ricardi et al. 1999).
Leaves of Delnortea abbottiae are up to 30 cm long however, most of the specimens are fragmentary and the form of the whole plant is a mystery. Rock layers yielding Delnortea leaves also possessed retuse taeniopterid fragments (Mamay et al. 1984), which I now assign provisionally to Lonesomia mexicana (R. Weber 1997).
Pictured to the left is a piece of a partially permineralized Delnortea abbottiae specimen, which was photographed before the midrib was thin-sectioned and prepared. The thin-sectioned samples of USNM 372427 are illustrated by Mamay et al. (page 1418, Figures 24-28, 1988).
The phrase on page 760 of T. N. Taylor et al. (2009), "most specimens of D. abbottiae are 3-5 cm long," is misleading. Fewer than 20 tiny, 3-5 cm long leaves were found in the uppermost two decimeters of Section IV, Unit 5 of the Cathedral Mountains Formation just below a three meter thick layer of conglomerate representing Unit 6, which was devoid of leaf fossils but yielded a permineralized log of Dadoxylon.
One leaf specimen yielded ovoid concretions on the distal edge of the lamina, and bite marks were seen on another fossilized leaf. Microscopic study of limonitic permineralizations of Delnortea abbottiae reveal a pattern of secondary growth from a vascular cambium; a developmental syndrome often seen in seed plants (Mamay et al. 1988).
Mamay's suggestion that the stratigraphic occurrence of Delnortea in Upper Leonardian rocks of the Cathedral Mountain Formation may lead to a better understanding of Permian floral zones is supported by discovery of Delnortea from the Artinskian of northwestern South America (Ricardi et al. 1999).
Following the initial report of the paleobotanical discovery (Mamay et al. 1984), and later anatomical studies of Delnortea abbottiae by Mamay et al. (1988), paleontologists elucidated "dicot-like" leaf anatomy and found vessel elements in several other Cathaysian gigantopterids (H. Li et al. 1994, H. Li et al. 1996, H. Li and D. W. Taylor 1998, H. Li and D. W. Taylor 1999).
Phasmatocycas and other spermopterids. Phasmatocycas is an emerging group of ovule-bearing taeniopteroid leaves from Carboniferous (Pennsylvanian) rocks of interior North America (Axsmith et al. 2003) once thought to be allied with the cycads (Mamay 1976). Classification of detached fossilized Phasmatocycas leaves into a specific group of gymnosperms is impossible at the present time.
The Phasmatocycas-like Paleozoic seed plants Archaeocycas whitei (page 8, Mamay 1976), Eophyllogonium cathayense (Mei et al. 1992), Phasmatocycas bridwellii (Axsmith et al. 2003), and Sobernheimia jonkeri (Kerp 1983) might be ovulate pieces of bisexual gigantopteroid proanthostrobili with the male parts missing. I regard all four of these species as a whole new group of gigantopteroid Paleozoic seed plants possibly allied to Gigantopteridales, and do not include them in the Cycadales where T. N. Taylor et al. (pages 709-713, 2009) discusses them, which is in line with Axsmith et al. (2003).
Are Phasmatocycas leaves including ovule-bearing megasporophylls and microsporangia-bearing microsporophylls, the detached remains of gigantopterid plants with two kinds of leaves and gigantic petaloid organs attached to bisexual cone axes?
Further, did certain detached retuse-tipped taeniopteroid leaves belong to fertile axes of gigantopterid plants having dimorphic foliage, as inferred by poorly preserved compressions and impressions discussed in a progress report by X. Li and Yao (page 321, 1983) and later illustrated by X. Li and Yao (page 26 and Plate 4, 1983), Mei et al. (page 101, Plate 2, 1992), and R. Weber (page 232, Plate 3, 1997)?
Fertile material of taeniopterids from Pennsylvanian rocks of interior North America was first described as Spermopteris (Cridland and Morris 1960). The appearance of the whole plant to which spermopterid leaves were attached is a mystery.
Spermopterids may be detached pieces of early cycadophytes or some other unknown gymnospermous gigantopteroid shrub, tree, or vine. Critical permineralizations that unambiguously demonstrate diagnostic anatomy of cuticles, epidermal patterns (including stomatal complexes and cutin nanoridges), and ovulate position on abaxial or adaxial leaf surfaces are unknown. Therefore, taxonomic assignment of spermopterids to a specific seed plant order is unsupported by lack of fossil evidence.
To the left is a photograph of a possible immature ectopic ovule attached to a retuse megasporophyll (or the leaf was damaged before fossilization, and the object is a "tear") that we provisionally assign to Lonesomia mexicana (R. Weber 1997) or Phasmatocycas.
What is a taeniopterid? It is an common name which refers to often abundant Paleozoic foliage that resembles the leaves of extant Calophyllum (Clusiaceae, Theales, Dilleniidae) or Musa (Musaceae, Zingiberales, Zingiberidae).
Fossilized leaf remains of Taeniopteris are generally not attached to a stem or rachis, thus, in at least some forms it is not known whether the fossil fragments represent pieces of simple or compound leaves, or one leaf-type of fossil plants having dimorphic leaves.
Taeniopterid leaf compressions and impressions are common in terrestrial and deltaic sedimentary deposits of the Paleozoic. Taeniopterids often co-occur with fossilized remains of detached net-veined leaves (or leaflets) of gigantopterids, gigantopteroids, and glossopterids.
Pictured to the right is an example of a laminar microsporophyll that I provisionally assign to Phasmatocycas. Elongate rice-shaped structures on the adaxial leaf surface may be pollen-bearing sacs. The fossil I collected is the only known microsporophyll of a gigantopteroid, and is left without a complete diagnosis of the organ and the mother plant.
The images are 280 million year old permineralizations photographed by the author in 1982 a couple days after the fossils were excavated from the bedding plane of Unit 5 of Section IV of the uppermost Cathedral Mountain Formation, Del Norte Mountains, North America. These specimens and others are deposited in the United States National Museum (Mamay et al. 1988). The fossilized leaf imaged to the right is actual size.
The taeniopterid microsporophyll shown to the right (if the rice-grained lumps on the permineralization are proven to be pollen-containing sacs) could be the developmental product of B-class MADS-box genes such as expression of AP3 or PI homologs along the lines suggested by D. E. Soltis et al. (2007), but in a much older i.e. 280 million year old Paleozoic gymnosperm.
It is unknown whether megasporophylls and microsporophylls of Lonesomia mexicana or Phasmatocycas were attached to a massive bisexual cone axis of the same plant or to separate female and male individuals.
Paleozoic spermopterids are a relatively unknown group. In Phasmatocycas bridwellii, ovules located on the lower (abaxial) surface of leaves were attached by stalks to leaf midribs but not to the leaf edges as suggested by Cridland and Morris (1960).
In 1983 Professor Hans Kerp described Sobernheimia jonkeri (page 713, Figure 17.23, T. N. Taylor et al. 2009) from detached megasporophylls. Were ovule-bearing leaves of Sobernheimia jonkeri parts of enormous bisexual gigantopteroid proanthostrobili with microsporophylls and perianth segments shed and therefore, evading understanding of whole reproductive axes?
The image to the left is the distal portion of an undescribed, retuse Phasmatocycas ovulate leaf that bears resemblance to sterile taeniopterid leaves of Lonesomia mexicana (Plate 3, Figs. 1-3, page 232, Weber 1997), which is a gigantopterid according to Professor Weber.
The two lumps shown on this image are probably ovules. The fossils were collected from exposures of the lower Permian (Leonardian) Cathedral Mountain Formation located in the Del Norte Mountains of southwestern North America. This specimen is deposited in the United States National Museum (Mamay et al. 1988).
Were ovulate Phasmatocycas bridwellii leaves (Axsmith et al. 2003) detached pieces of a large bisexual cone axis, i.e. the "proanthostrobilus" (page 817, J. A. Doyle 2008) of an unknown Paleozoic gymnospermous gigantopterid tree or shrub with the pollen-bearing leaves (microsporophylls) and strap-shaped petaloid organs of the perianth missing?
To answer this question is one of many paleobotanical challenges that require a coordinated attack strategy to be devised by developmental biologists and paleontologists.
"... Nonetheless, it is possible that co-expression of AP3- and PI-homologs [B-class MADS-box genes, see Chapter on Genetics Considerations] mediated the evolutionary innovation of animal-attractive, petal-like organs well before the appearance of flowers."
The preceding passage is quoted from page 361 of D. E. Soltis, H. Ma, M. W. Frohlich, P. S. Soltis, V. A. Albert, D. G. Oppenheimer, N. S. Altman, C. dePamphilis, and J. Leebens-Mack (2007), The floral genome: an evolutionary history of gene duplication and shifting patterns of gene expression. Trends in Plant Science 12(8): 358-367. The phrase in brackets [] is mine.
Howe and Cantrill (2001) describe paleosols from the Albian of Antarctica having lenses of fossilized, detached Taeniopteris daintreei leaves, Carnoconites crantwelli ovulate organs, and Pentoxylon stems. Did these organs belong to a shrub-like Pentoxylon plant?
Paleozoic spermopterids, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--unknown, Phasmatocycas bridwellii was possibly shrub-like (Axsmith et al. 2003); additional paleontologic data are needed to reconstruct whole plants and nodal anatomy
REPRODUCTIVE MODULES--Phasmatocycas: the modules are phyllospermous (Cridland and Morris 1960, Mamay 1973, Axsmith et al. 2003), ovules attached to midribs on the lower (abaxial) surface of laminar megasporophylls
Were detached sporophylls of Phasmatocycas, including the undescribed microsporophyll pictured above and sterile taeniopterid leaves (perianth parts), pieces of a massive “proanthostrobilus” sensu J. A. Doyle (page 817, 2008), which was attached to a gigantopteroid shrub having dimorphic leaves?
The only known spermopterid male specimen (illustrated above) suggests placement of rice-grained shaped pollen sacs on the upper (adaxial) surface of the microsporophyll; anatomical and developmental details of sexual reproduction unknown, permineralizations are in need of discovery and study
LEAVES--taeniopteroid with a stout multi-stranded midrib, the lateral veins parallel resembling Clusiaceae or Musaceae (see above), whole leaves unknown but probably simple (Cridland and Morris 1960, Mamay 1973, Axsmith et al. 2003); permineralizations with preserved leaf anatomy are needed for study
PHYTOPHAGOUS ASSOCIATE(S)--arthropods belonging to the Caloneurodea, Orthoptera, Protorthoptera (Beck and Labandeira 1998); vertebrate coprolites require discovery and study; fossilized invertebrate exoskeletons and guts are needed for study
PLANT IDENTIFICATION(S)--Taeniopteris (Labandeira 1998); definitive anatomical data are needed for precise taxonomic and nomenclatural placement of the form genus into a known seed plant order, family, and genus
HOST SEED PLANT ORGAN(S) BEING EATEN--leaves (Beck and Labandeira 1998); additional fossils require discovery and study
Asian gigantopterids. Asian gigantopterids were distinct from glossopterids, extinct Permo-Triassic seed plants whose stratigraphic distribution has been used to support Wegener's Theory of Continental Drift. In 1982 X.-G. Li and Z.-Q. Yao reviewed the work to that date on the Cathaysian flora in Asia. A compilation is available in H. Xilin et al. (1996) that summarizes research on the Permian coal floras of Jiangxi Province, China.
The most recent review of the leaf form genera of Jambi gigantopterids is by Booi et al. (2009).
Gigantopterids from Permian rocks of Asia were first described by Schenck as Megalopteris nicotianaefolia from poorly preserved fossil impressions (Glasspool et al. 2004). Morphotype genera indigenous to Asian Paleozoic rocks are Aculeovinea, Cathaysiopteridium, Cathaysiopteris, Cardioglossum, Gigantonoclea, Gigantonomia, Gigantopteridium, Gigantopteris, Gigantotheca, Gothanopteris, Linophyllum, Neogigantopteridium, Palaeogoniopteris, Progigantopteris, Trinerviopteris, Vasovinea, and Zeilleropteris (H. Li et al. 1994, Booi et al. 2009, among others).
Fossilized connections of the leaf impressions with whole plants and reproductive structures of gigantopterids are exceedingly rare or undescribed. Permian Gigantopteris of China unearthed and studied by X.-G. Li and Z.-Q. Yao in 1983, yield a rare glimpse of fertile material: ovules and pollen-bearing organs were attached to leaves and leaf-midribs, but the anatomy and placement of the connections is indeterminate. A relatively recent report of reproductive structures found preserved in a bedding plane in close association with gigantopterid leaves only adds to the mystery of these plants and their gymnosperm associates (Mei et al. 1992).
Permineralized gigantopterid foliage and stems belonging to Aculeovinea yunguiensis, Gigantonoclea guizhouensis, and Vasovinea tianii (H. Li et al. 1994, Z.-Q. Wang 1999, H. Li and D. W. Taylor 1999), possess angiosperm-like vessels and libriform fibers. Leaves of Chinese gigantopterids with waxy cuticles have been described (Z.-Q. Yao and Crane 1986).
Based on the anatomy of vessel-containing permineralized stems H. Li et al. (1994) and H. Li and D. W. Taylor (1999) proposed that gigantopterids were vines.
Was the twining habit and innovative vessel-containing secondary xylem of Paleozoic gigantopterids attributable to the same biomechanical properties and developmental plasticity seen in extant tropical eudicots (Ménard et al. 2009)?
Do xylem patterns sensu Carlquist (2009) in gigantopterid leaf midribs and stems offer clues on a potential neotenous origin of angiosperms?
In 1992 Mei et al. described Eophyllogonium cathayense, an enigmatic seed plant from the Permian of China. Seed-bearing taeniopterid leaves with reticulate venation were found in the same bedding plane as sterile gigantopterid leaves assignable to Gigantonoclea acuminatiloba and Gigantopteris dictyophylloides. Was Eophyllogonium cathayense attached to a gigantopteroid plant with dimorphic leaves?
The Paleozoic fossil Sobernheimia jonkeri (Kerp 1983; page 713, Figure 17.23, T. N. Taylor et al. 2009) might also be detached megasporophylls of a massive bisexual gigantopteroid proanthostrobilus with the microsporophylls and perianth also shed and therefore, evading a complete diagnosis of the whole fertile SAM.
Paleobotanists are better understanding the anatomy, morphology, and systematics of Cathaysian gigantopterids, but more work is needed to elucidate relationships with the detached and fragmentary fossilized remains of other Paleozoic gymnosperms including Phasmatocycas.
Cycadales:
A Paleozoic origin of cycads is demonstrable based on the work of T. N. Taylor (1969) but the evolutionary relationships of some early fossil forms (e.g. Phasmatocycas, Mamay 1969, 1973, 1976) is open to reinterpretation (Axsmith et al. 2003). Nilssoniales are often treated as an evolutionary branch distinct from the cycads (Krassilov 1997). I include the group here, which is in line with T. N. Taylor et al. (2009).
A living specimen of Cycas revoluta (Cycadaceae, Cycadales) is pictured to the right. The brownish fertile leaves that bear bright red seeds are termed megasporophylls. Fertile leaves and green vegetative megaphylls are produced spirally on the monopodial SAM. The SAMs of some cycads contain colonies of beetles and weevils (Norstog et al. 1986, Norstog and Fawcett 1989).
The cycad literature is extensive covering the Carboniferous, Permian, Triassic, Jurassic, Cretaceous, and Paleogene periods (Harris 1961, T. N. Taylor 1969, Delevoryas and Hope 1976, Smoot et al. 1985, Krassilov and Bugdaeva 1988, Zhu et al. 1994, Bremer et al. 2003, Klavins et al. 2003, Archangelsky and Villar de Seoane 2004, Krassilov and Doludenko 2004, Hermsen et al. 2006, Watson and Ash 2006, Pott et al. 2007, and Hermsen et al. 2009).
Molecular phylogenetic analyses of cycadophytes are available (Hill et al. 2003, Zgurski et al. 2008). Volume 70, Number 2 of The Botanical Review (2004), compiles some of the research on these seed plants.
Paleozoic cycads (e.g. Crossozamia and Lasiostrobus) are hermaphroditic cycads known from both China and North America (Norstog and Nicholls 1997, L. Liu and Z. Yao 2002). In many of the known fossil plant localities, cycad fossils (with preserved reproductive structures) are found with sterile leaf fragment impressions referable to the form genus Taeniopteris.
The Paleozoic spermopterid seed plants Archaeocycas whitei (page 8, Mamay 1976), Eophyllogonium cathayense (Mei et al. 1992), Phasmatocycas bridwellii (Axsmith et al. 2003), and Sobernheimia jonkeri (Kerp 1983) might be ovulate pieces of bisexual gigantopteroid proanthostrobili with the male parts missing. I regard all four of these species as a whole new group of gigantopteroid Paleozoic seed plants and do not include them in the Cycadales where T. N. Taylor et al. (pages 709-713, 2009) discusses them, which is in line with Axsmith et al. (2003).
Triassic cycads have been reported (Ash 1985, Delevoryas and Hope 1971, 1976; Ash 2001, Rozynek 2008, X. Wang et al. 2009) including recent Antarctic fossil finds described as Antarcticycas schopfii (Smoot et al. 1985, Hermsen et al. 2006) and Delemaya spinulosa (Klavins et al. 2003). Hermsen et al. (2006) report on the anatomy of stems and leaves of these fossils together with a discussion of the evolution of the group.
Jurassic cycads are well-known principally of the work by Harris (1932, 1941, 1954, 1961, 1964). Harris (1961) and Tidwell (2002) are keys to the literature on the fossil history and paleobiology of cycads. Enigmatic fossils from the Jurassic Period, possibly transitional between earlier gymnosperms and later cycads include Baruligyna disticha from the Callovian of Georgia in western Asia (Krassilov and Doludenko 2004); a close relative of Semionogyna, from the Lower Cretaceous of Transbaikalia, Russia (Krassilov and Bugdaeva 1988).
The image to the left is a male cone of Cycas revoluta (Cycadaceae, Cycadophyta) from a plant in cultivation, photographed by the author.
Known gross morphological characters of some of cycads, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--monopodial shrubs with the main stem sheathed in helically arranged cataphylls (Smoot et al. 1985, Hermsen et al. 2006)
REPRODUCTIVE MODULES--cones (Harris 1941, Zhu et al. (1994), Klavins et al. 2003), ovule-bearing leaves (megasporophylls) are incompletely known from compressions and impressions (Zhu et al. 1994); a bipinnate microsporophyll has been described as Androcycas santucci (Watson and Ash 2006)
Permineralized pollen cones with pollen are described as Delemaya spinulosa (Klavins et al. 2003); additional permineralizations of reproductive material are needed to better understand pollination biology and reproductive anatomy. Is Delemaya spinulosa the male cone of Antarcticycas schopfii?
LEAVES--cuticles (Harris 1932); simple and strap-shaped leaves which are associated with ovules have been described as Archaeocycas (Mamay 1976); Nilssoniopteris (Toshihiro Yamada et al. 2009); and compound pinnate leaves e.g. Zamites tidwellii (Ash 2001) are well-known from Mesozoic and Cenozoic rocks
PHYTOPHAGOUS ASSOCIATE(S)--arthropods belonging to the Caloneurodea, Orthoptera, Protorthoptera (Beck and Labandeira 1998), insect frass not assignable to a particular species is known from permineralizations of Antarcticycas (Hermsen et al. 2006); vertebrate coprolites require discovery and study; fossilized invertebrate exoskeletons and permineralized insect guts are needed for study
PLANT IDENTIFICATION(S)--Taeniopteris (Labandeira 1998); definitive anatomical data are needed for precise taxonomic and nomenclatural placement of the form genus into a family and genus belonging to the Cycadales
Antarcticycas schopfii is the host plant for unknown insect phytophagous associates (Hermsen et al. 2006); important insights have been gained as summarized by Hermsen et al. (2006), but much more paleobotanical research is needed
HOST SEED PLANT ORGAN(S) BEING EATEN--leaves (Beck and Labandeira 1998) and cataphylls (Hermsen et al. 2006), more permineralizations require discovery and study
Several studies on the evolutionary relationships, genetics, molecular systematics, phytophagous insect associates, pollination ecology, and taxonomy of extant cycads are published by Stevenson (1980, 1981), Norstog et al. (1986), Norstog and Fawcett (1989), Waggoner (2001), Terry et al. (2004), P. Zhang et al. (2004), Azuma and Kono (2006), Sass et al. (2007), and Wu et al. (2007)
Additional work on cycads accomplished during the last decade includes papers written by Terry et al. (2007), Cabrera-Toledo et al. (2008), S.-M. Chaw et al. (2008), D. A. Downie et al. (2008), González et al. (2008), González-Astorga et al. (2008), Rozynek (2008), Schutzman et al. (2008), R. Singh and Radha (2008), A. S. Taylor et al. (2008), Chiang et al. (2009), Lindstrom (2009), Pinares et al. (2009), Proches and Johnson (2009), Toshihiro Yamada et al. (2009), and X. Wang et al. (2009).
Cycads have figured prominently in classic papers on the origin of angiosperms (Arber and Parkin 1907, E. Anderson 1934, Crane 1985). Further, modern syntheses (e.g. D. E. Soltis et al. 2007) suggest that whole genome duplications were important in the early evolution of the angiosperm stem group. Yet, "it is noteworthy that polyploidy is absent in some ancient plant lineages, such as the cycads" ... (page 376, Crepet and Niklas 2009).
Glossopteridales:
One of the dominant vegetation types of the southern reaches of Pangaea during the Permian Period consisted of small to large trees belonging to a group of seed ferns known as glossopterids. Following the Late Carboniferous and early Permian ice age, mesic forests of glossopterids spread poleward. Gradual warming at the South Pole led to replacement of Glossopteris forests by stands of Dicroidium. Change in the composition of overstory trees might have altered understory shrubs and herbs possibly contributing to a decline in biodiversity of herbivorous dicynodonts (Tiffney 1992, Zavada and Mentis 1992).
Mary E. White (1986) reviews the fossil history of glossopterids. Several morphotype genera of glossopterids circumscribe detached fossil leaves (Eretmonia, Glossotheca, and Kendostrobus, among others), isolated pollen sacs (Arberiella, Lithangium, and Polytheca), ovule-bearing leaves (e.g. Scutum), compound ovulate structures (Lidgettonia etc.), detached seeds (Pterygospermum and Stephanostoma), fossil leaves (Belemopteris, Gangamopteris, Glossopteris, and Rhabtotaenia, among others), and underground parts (Vertebraria).
Plumstead (1973) presents an illustrated discussion of the Glossopteris flora, the paleogeography of Gondwana, and Wegener's Theory of Continental Drift. Glossopteridales and the possible relationships of glossopterids with angiosperms are discussed by Retallack and Dilcher (1981) and E. L. Taylor and T. N. Taylor (1992).
Key articles on glossopterids are published by Rigby (1967), Surange and Maheshwari (1970), Delevoryas and Gould (1971), Maheshwari (1972), Rigby (1972), Surange and Chandra (1972, 1973, 1975), Holmes (1973), Delevoryas and Person (1975), Chandra and Surange (1976), Schopf (1976), Pant and Choudhury (1977), Gould and Delevoryas (1977), M. E. White (1978), Rigby (1978), Pant and Nautiyal (1984), Pant and Nautiyal (1984), Pant (1987), and Pigg et al. (1987).
More recent works are Pigg (1990), Pigg and Taylor (1990), McLoughlin (1990), Rigby and Chandra (1990), Pigg and Trivett (1994), Berthelin et al. (2004), Nishida et al. (2004, 2007), Tewari (2007), Prevec et al. (2008), Cariglino et al. (2009), Decombeix et al. (2009), Rydberg (2009), and E. L. Taylor and T. N. Taylor (2009).
Pigg and T. N. Taylor (1993), Iannuzzi (2000), and Pigg and Nishida (2006) compile particularly complete bibliographies on the fossil history of glossopterids.
The image to the left is a plate showing the morphology of some glossopterids. It is Figure 7 from Peter R. Crane (1985), Phylogenetic analysis of seed plants and the origin of angiosperms, Annals of the Missouri Botanical Garden 72: 716-796, reprinted with permission of the Missouri Botanical Garden and Peter Crane.
"Figure 7. Morphology of glossopterids. -A. Ottokaria megasporophyll and associated leaf, redrawn from Pant (1977a, fig. 10E-G), orientation of megasporophyll based on Pant and Nautiyal (1984); × 1. -B. Lidgettonia africana megasporophyll, based on Thomas (1958), Surange and Chandra (1975, text-fig. 1D), Schopf (1976, fig. 8D); × 1.5. -C. Eretmonia microsporophyll, redrawn from Surange and Chandra (1975, text-fig. 1D); × 2. -D. "Glossopteris" (?Dictyopteridium) megasporophyll in axil of vegetative leaf, redrawn from Gould and Delevoryas (1977, fig. 1d), note that details of sporophyll attachment and sporophyll orientation are uncertain, see text for discussion; × 1. -E. Glossopteris sastrii leaves borne on a shoot, based on Pant and Singh (1974, text fig. 2B-D); × 0.5. -F. Pterygospermum raniganjense platyspermic ovule based on Pant and Nautiyal 1960, text-fig. 3A); × 25. -G. Pollen grain from micropyle of P. raniganjense, redrawn from Pant and Nautiyal (1960, text-fig. 4G); × 550."
Known gross morphological characters of glossopterids, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--trees and shrubs (Pant and Singh 1974, White 1986)
REPRODUCTIVE MODULES--ovule-bearing leaves (megasporophylls) are of two types: those of Section Megafructi produce either stalked or sessile ovules on upper surfaces of megasporophylls known as regular leaves (Crane 1985, M. E. White 1986, E. L. Taylor and T. N. Taylor 1992). In Microfructi, the ovules are attached to scale-leaves termed microphylls (M. E. White 1986)
Ovules of Glossopteris have been described as Homevaleia gouldii (Nishida et al. 2007). Preserved sperm, pollen tubes, and ovules provide evidence of zooidogamy in Glossopteris (Nishida et al. 2003, Nishida et al. 2004)
Gondwanan beds have yielded ovulate fructifications assignable to Arberia (Rigby 1972). Ovuliferous fructifications have been restudied and placed in Bifariala (Prevec et al. 2008)
Pollen-bearing (male) organs of glossopterids are attached to the upper surfaces of scales described as Eretmonia, Glossotheca, and Squamella (M. E. White 1986). All three form genera display Arberiella microsporangia that produce pollen (M. E. White 1986). Fossilized peltate discs bearing taeniate pollen have been described from the late Permian of Australia by Rigby and Chandra (1990)
LEAVES--simple, lanceolate, of the Glossopteris and Gangamopteris-type (Crane 1985, White 1986, Tewari 2007)
PHYTOPHAGOUS ASSOCIATE(S)--wood-boring Coleoptera (Zavada and Mentis 1992, Weaver et al. 1997, Labandeira 1998). Caloneurids, orthopterans, and protorthopterans are external leaf-feeders on Glossopteris foliage (Labandeira 1998). Hypoperlids and grylloblattids feed on pollen inside of Protohaploxypinus microsporangia (Labandeira 1998)
PLANT IDENTIFICATION(S)--Glossopteris, Protohaploxypinus
HOST SEED PLANT ORGAN(S) BEING EATEN--bark, leaves, microsporangia, pollen, and wood
Some workers suggest that surviving lineages of Paleozoic glossopterids are represented in Mesozoic floras by corystosperms and even Caytoniales. Thomas N. Taylor et al. (page 598, 2009) offer the most up-to-date review of glossopterids.
Gnetales:
The Gnetales is one of the orders of Paleozoic seed plants widely regarded as a sister group to the flowering plants (Arber and Parkin 1907, J. A. Doyle and Donoghue 1986, Cornet 1996, Krassilov 1997, J. A. Doyle 2006).
Based on molecular phylogenetic studies, a close relationship of conifers with Gnetales is supported (Y.-L. Qiu et al. 1999, Winter et al. 1999, Magallón and Sanderson 2002, Burleigh and Mathews 2004, Braukmann et al. 2009). Gnetales continue to pose a challenge in many phylogenetic analyses of flowering plants and their unknown seed plant ancestors and congeners (Sean W. Graham and Ihles 2009).
The anthophyte hypothesis was rejected by Donoghue and J. A. Doyle in 2000.
Cones identifiable to Gnetales are well documented from Permian rocks (Z.-Q. Wang 2004) despite J. A. Doyle’s assertion that "Gnetales and angiosperms are not known until the Mesozoic-late Triassic for probable stem relatives of Gnetales ..." (page 818, J. A. Doyle 2008).
Details of the fossil history of the group are summarized by Crane (1996), Rydin et al. (2004), Z.-Q. Wang (2004), Rydin et al. (2006), T. N. Taylor et al. (2009), and Tekleva and Krassilov (2009).
The main body of research on the evolutionary relationships, biology, phylogeny, and reproductive biology of the Gnetales is published in several papers by Carlquist (1996), Carmichael and Friedman (1995, 1996, 1998), Crane (1996), J. A. Doyle (1996), Endress (1996), Friedman (1996), Friedman and Carmichael (1996), Hufford (1996), Price (1996), and Krassilov (2009).
Debates on competing hypotheses linking gnetophytes with angiosperms are reviewed by Friis et al. (2007), Krassilov (2009), and Rothwell et al. (2009). Paleobotanical evidence (Z.-Q. Wang 2004) and molecular phylogenetic studies reviewed by Braukmann et al. (2009) support the prevailing gnepine hypothesis.
Another lab (page 209, Rydin et al. 2002) points out apparent discrepancies between the gne-pine hypothesis and findings derived from studies of anatomy, morphology and paleobotany of conifers and gnetophytes constituting " ... the main impediment for reaching a consensus on seed plant phylogeny."
Perplexingly, the existence of the late Paleozoic fossil Palaeognetaleana (Z.-Q. Wang 2004) adds a whole new dimension to the gne-pine hypothesis and relationships of Gnetales with walchian conifers and spermopterids at the time of the supposed angiosperm-gymnosperm split.
The image to the right is an attached fossil flower-like organ of Eoantha zerikhinii, an anthophytic gnetalean from the Baisian Assemblage, early Cretaceous Period, Transbaikalia, Russia. Four ovuliphores are visible together with a perianth of linear bracts. Each ovule contain a pollen chamber filled with Ephedrites-type pollen (this picture is from an original image provided by Professor Valentin Krassilov, posted here with his permission).
Known gross morphological characters of Gnetales, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--shrubs and small trees (Krassilov 1997)
REPRODUCTIVE MODULES--cupulate pre-flowers (Krassilov 1997); flower-like parts arranged in a compound cone with whorled, partially fused microsporophylls; or one or more ovules surrounded by enveloping bracts (Stewart and Rothwell 1993), reduced ovuliphores in male strobili function as nectar-producing organs (Krassilov 1997). Ovules are borne on the tips of stems; each ovule is enclosed by a bracteole. Ovules are enclosed by a fused integument and nucellus. The micropyle is biseriate and tubular. Ingrowths of cells close the micropyle and a pollen chamber is present (Rothwell et al. 2009)
LEAVES--linear and flattened in Drewria, "dicot"-like in Gnetum, strap-like in Welwitschia, or reduced to scales in Ephedra (Stewart and Rothwell 1993). Leaves are opposite and not pinnate (Rothwell et al. 2009)
PHYTOPHAGOUS ASSOCIATE(S)--Spathoxyela, a xyelid sawfly (see Dmitriev and Ponomarenko 2002)
PLANT IDENTIFICATION(S)--Baisianthus (see Dmitriev and Ponomarenko 2002)
HOST SEED PLANT ORGAN(S) BEING EATEN--pollen-bearing flower-like organs of Baisianthus (see Dmitriev and Ponomarenko 2002)
Some of the research on the, anatomy, ecology, evolutionary-developmental biology, fossil history, and molecular systematics of gnetophytes is published by Van Konijnenburg-Van Cittert (1992), Osborn et al. (1993), J. A. Doyle (1998), Pham and Sinha (2003), Rydin et al. (2002), Mundry and Stützel (2004), Won and Renner (2005), Hajibabaei et al. (2006), Rydin et al. (2006), Friis et al. (2007), C.-S. Wu et al. (2007), H.-M. Liu et al. (2008), McCoy et al. (2008), and T. Yamada et al. (2008).
Recent papers of potential evolutionary and phylogenetic significance are Braukmann et al. (2009), Friis et al. (2009), S.-X. Guo et al. (2009), Hollander and Vander Wall (2009), Krassilov (2009), Kunzmann et al. (2009), Rydin and Korall (2009), Tekleva and Krassilov (2009), and C.-S. Wu et al. (2009).
The most reduced cpDNAs are known from extant Gnetales specifically Ephedra equisetina, Gnetum parvifolium, and Welwitschia mirabilis (C.-S. Wu et al. 2009).
Vojnovskyales:
In Carboniferous and Permian times another intriguing group of shrub-like seed plants with palm-like leaves, leafy short shoots, and bisexual cone axes appears in the stratigraphic column (Krassilov and Burago 1981, Rothwell et al. 1996, Naugolnykh 2001). Several species have been described from Permian rocks all over the world. Perhaps the best known species are Sergeia neubergii (Rothwell et al. 1996) and Vojnovskya paradoxa (Mamay 1976).
Vojnovskyaleans bear close resemblance in many details of leaf, reproductive, and stem anatomy to Triassic Sanmiguelia and the later pentoxylaleans (Crane 1985, Krassilov 1997, Naugolnykh 2001).
Mamay (page 295, 1976) provoked interesting speculation on Maekawa's 1962 proposal casting Vojnovskya as a "presumable ancestor of angiosperms." It is equally interesting that some workers have drawn a connection between foliar material of Vojnovskya paradoxa and Sanmiguelia lewisii (page 779, Crane 1985).
Another view held by Mamay is that the Vojnovskyales might be remotely related to the Bennettitales or Cordaitales, or represent "some bizarre, short-lived group of late Paleozoic gymnosperms that attained a sparse but geographically broad distribution" (page 295, Mamay 1976).
Several groups of vojnovskyaleans had architectural adaptations possibly exploited by Paleozoic insects. Intriguing fossilized evidence in Paleozoic rocks of whole plant organs of Sergeia neuburgii (Vojnovskyaceae, Vojnovskyales) is published on page 1072 of Rothwell et al. (1996). Unfortunately these fossils cannot be easily screened for invertebrate remains possibly intercalated between the helically arranged leaf permineralizations, because the fossil was transported with other turbidites some distance from the shoreline (Rothwell et al. 1996).
The image to the left consists of four drawings showing the fructification morphology of certain vojnovskyaleans. It is Figure 1 from S. V. Naugolnykh (2001), Morphology and systematics of representatives of Vojnovskyales, Paleontological Journal 35(5): 545-556, reprinted with written permission of Pleiades Publishing, Inc. I thank Serge V. Naugolnykh and the Paleontological Journal for this contribution.
"Figure 1. Fructification morphology of representatives of Vojnovskyales: (a, c, d) Paravojnovskya (al. Gaussia) imbricata (Naug.) Naug. et Doweld, (a) specimen no. 3773(11)/326(92), (c, d) the specimen from the collection by Vaulev (Perm Regional Museum), (b) proposed arrangement of fructifications of P. imbricata on a fertile shoot, in axils of scale-form bracts of Nephropsis (Sulcinephropsis). The locality of Chekarda-1, the Kungurian, Lower Permian of the Middle Fore-Urals. Scale bar is 1 cm."
Known gross morphological characters of Vojnovskyales, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--possibly shrub-like (Mamay 1975, 1976) with scaly, bracetose, and leafy short shoots often fertile in the developing SAMs (Krassilov 1997). Permineralizations are needed to better understand the anatomy of short shoots
REPRODUCTIVE MODULES--globose head-like bisexual strobilus (e.g. Gaussia, Krassilov 1997) composed of basally attached ovule-bearing leaves (megasporophylls) and terminal microsporophylls each with two microsporangia attached to the distal end of each leaf; the fossils are incompletely known from compressions and impressions (Mamay 1976, Krassilov 1997). Shortened, bracteose, cone-like structures subtended by umbrella-shaped leaves that contain either pollen or stalked seeds were described by Naugolnykh (2001)
LEAVES--simple and flabelliform in Sandrewia texana (Mamay 1975, 1976), and lanceolate in other genera (Krassilov 1997); the leaves of Permian Vojnovskya paradoxa resemble Sanmiguelia lewisii, a Triassic plant
PHYTOPHAGOUS ASSOCIATE(S)--unknown; existing specimens should be restudied for the presence of insect remains and traces
PLANT IDENTIFICATION(S)--not applicable
HOST SEED PLANT ORGAN(S) BEING EATEN--not applicable
The reproductive biology of the Vojnovskyales is not clearly understood, but some research progress has been made (Rothwell et al. 1996, Naugolnykh 2001). Additional discoveries of permineralized sexual structures are needed to better understand these seed plants and to reanalyze them in a phylogenetic context.
"Since conifers extend back to the Late Carboniferous, this implies that the line leading to angiosperms goes back this far too - an apparent conflict with the stratigraphic record (Axsmith et al., 1998; Doyle 1998a)."
This preceding passage is quoted from page 172 of J. A. Doyle (2001), Significance of molecular phylogenetic analyses for paleobotanical investigations on the origin of angiosperms, The Palaeobotanist 50: 167-188.
Seed Plants of the Mesozoic Era:
One approach toward our better understanding of flowering plant cladogenesis from the lineage(s) containing gymnospermous seed plant ancestors is to list the known fossil seed plant groups of the Mesozoic Era and to compare and contrast them with respect to morphological characters most often associated with angiosperms.
The image to the right is a drawing of a detached spherical ovule-bearing Vardekloeftia sulcata head with densely packed ovules and interseminal scales, from a bennettitalean bush. It is Figure 9B from Peter R. Crane (1985), Phylogenetic analysis of seed plants and the origin of angiosperms, Annals of the Missouri Botanical Garden 72: 716-796, reprinted with permission of the Missouri Botanical Garden and Peter Crane.
"Figure 9. Morphology of the Vardekloeftia and Bennetticarpus plants. -B. V. sulcata, spherical head composed of ovules and interseminal scales, based on Harris (1932b, pl. 15, fig. 1, pl. 17, figs. 1, 2, pl. 18); × 2."
Did sauropods and smaller herbivorous reptiles feed on these fleshy Vardekloeftia seed heads? Were the ovules and scales of this plant filled with antiherbivory poisons such as polyacetylenes and toxic cycasins?
"In sum, from the plant's point of view, the age of dinosaurs was not an extension of Permian herbivory, nor a duplication of the present. While smaller herbivores duplicated some aspects of the preceding and following time, the immense herbivores imposed a unique selective force on the physiognomy and life history strategies of Mesozoic plants. Initially, these forces were met by a limited diversity of genetic lineages of plants, but by the end of the Mesozoic, gymnosperms with efficient vegetative growth and abilities to recover from damage (= angiosperms) had evolved. Much of the morphology and biology of Mesozoic plants should be considered in the light of this substantial herbivore pressure."
The preceding quotation is from page 94 of Tiffney (1992), The role of vertebrate herbivory in the evolution of land plants, The Palaeobotanist 41: 87-97.
The first occurrence of angiosperm-like palynomorphs in the stratigraphic record is from the early Triassic Period (Hochuli and Feist-Burkhardt 2004), but the anatomy and morphology of the seed plant(s) which shed such pollen is a mystery. Fossilized remains of Triassic seed plants which have flowering plant characteristics are very rare and often poorly preserved.
I now present a brief survey of several enigmatic and sometimes spectacular seed plant finds.
Sanmiguelia. Of all the enigmatic seed plants of the early Mesozoic Era, Sanmiguelia lewisii has attracted the most attention by paleobotanists. This rather common Triassic fossil of southwestern North America is remarkably similar to the Paleozoic Vojnovskyales (page 779, Crane 1985), another group of angiosperm-like seed plants from Asiatic and North American Paleozoic rocks (Mamay 1976).
Sanmiguelia lewisii is an innovative Triassic plant having palm-like leaves (Brown 1956, Ash 1976) with flowers and angiosperm-like reproductive modules not unlike the monocotyledonous angiosperm Veratrum (Cornet 1986, 1989). Detailed studies of the reproductive morphology of Sanmiguelia have been published (Cornet 1989). Additional permineralized fossil material of Sanmiguelia is probably needed to better understand the anatomy of reproduction and whole plant morphology.
Morphological details of Sanmiguelia, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--Sanmiguelia lewisii was a woody herb (Cornet 1986, 1989), or small shrub. Better preserved fossilized material and more anatomical data are needed to better understand the anatomy and morphology of whole plants in relation to Vojnovskyales and the angiosperms Joinvillea and Veratrum
REPRODUCTIVE MODULES--ovuliferous inflorescences (first described as Axelrodia burgeri), polleniferous inflorescences (named Synangispadixis tidwellii), flowers with ovuliferous units and polleniferous units, megasporophylls as carpels, synangia as anthers, bracts, bitegmic ovules (Cornet 1986, 1989)
The reproductive biology of Sanmiguelia is not clearly understood, but some research progress has been made (Cornet 1986, 1989). Additional discoveries of permineralized sexual structures are needed to better understand the evolutionary position of these seed plants in relation to other anthophytes (including monocotyledonous flowering plants) and older vojnovskyaleans
LEAVES--simple, alternate, clasping and flabelliform to lanceolate resembling extant Veratrum californicum (Liliaceae, Liliales, Liliidae) (Cornet 1989)
PHYTOPHAGOUS ASSOCIATE(S)--bite marks and trails of anther debris left by unknown insects (Cornet 1989); additional specimens await discovery to be studied for the presence of insect remains and traces
PLANT IDENTIFICATION(S)--Sanmiguelia lewisii (Cornet 1989)
HOST SEED PLANT ORGAN(S) BEING EATEN--leaf tissue and pollen? (Cornet 1989)
Enigmatic seed plants of the Mesozoic. The early Mesozoic fossil record of seed plants is extremely sparse and poorly known. Too few paleobotanists specialize in Triassic and Jurassic paleobotany, and inadequate resources are applied toward finding early Mesozoic insect and plant fossil transitional series.
Several enigmatic genera are known from Mesozoic rocks. These poorly known fossil forms include Baisia (Krassilov 1997), Cycandra (pollen cones resembling fossil cycads or Nilssoniales), cupules of Dirhopalostachys (reminiscent of basal flowering plants such as Lesqueria elocata), Doylea tetrahedrasperma (a corystospermous cupule or carpel-like organ, Stockey and Rothwell 2009), Fraxinopsis (Axsmith et al. 1997), Furcula (Harris 1932), Irania (with both androclades and gynoclades), Leptostrobus (a czekanowskialean with stigmatic cupules), Pannaulika triassica (Cornet 1993), and Raunsgaardispermum lusitanicum (palynomorphs resembling Bennettitales and Gnetales) (Mendes et al. 2008).
The most recent discovery is Schmeissneria, an intriguing seed plant incertae cedis from the earliest Jurassic Era (Lias interval) of Europe, and Middle Jurassic of Asia (X. Wang et al. 2007).
Bennettitales. By Triassic time, an intriguing group of monopodial seed plant shrubs, known as cycadeoids appears in the stratigraphic column (Crepet 1972, Crepet 1974, Crane 1986, Stewart and Rothwell 1993, Nishida 1994, Barbacka 2000, Rothwell et al. 2009, T. N. Taylor et al. 2009). These shrubs were cycad or palm-like in overall aspect. Reproductive structures were borne on modified leaves, clustered in flower-like strobili. Classified in the taxonomic order Bennettitales, cycadeoids were thought by Arber and Parkin (1907) as likely candidates as ancestors of modern flowering plants (pages 739-740, T. N. Taylor et al. 2009).
Paleobotanists Crepet, Rothwell, and Stockey offer a detailed literature review with data on the anatomy and morphology of the Bennettitales including implications of recent studies on the anthophyte hypothesis of angiosperm origins (Rothwell et al. 2009).
There are two or three main groups of cycadeoids often treated at the family level in the taxonomic hierarchy. Cycadeoidaceae studied by Wieland and Delevoryas include Cycadeoidea, fossilized remains of squat, shrub-like trunks bearing leaves, cones, and pollen recovered from Cretaceous rocks of North America and Monanthesia (see Delevoryas 1962). The other group classified in the Williamsoniaceae includes Williamsonia sewardiana, described from the Jurassic period of India.
Detailed studies of bennettitaleans by paleobotanists have been underway for several decades. This work includes Bose (1968), Delevoryas (1968), Harris (1969), Crepet and Delevoryas (1972), Crepet (1972, 1974), T. N. Taylor (1973), Harris (1974), Sharma (1974, 1976, 1977), Crane (1986), Pedersen et al. (1989), Delevoryas (1991), Nishida (1994), Osborn and Taylor (1995), Saiki and Yoshida (1999), Barbacka (2000), G. Sun et al. (page 165, 2001), Rothwell and Stockey (2002), Stockey and Rothwell (2003), Watson and Lydon (2004), Boyd (2004), Friis et al. (2007), Pott et al. (2007 [two papers]), Going et al. (2007), and Y.-D. Wang et al. (2008).
The most recent articles pertaining to cycadeoids in the scientific literature are by Crane and Herendeen (2009), Pott and McLoughlin (2009), Rothwell et al. (2009), Yamada (2009), and Zavialova et al. (2009).
Form genera representing detached fossilized organs of bennettitaleans include Anomozamites (leaves), Bucklandia (stems and leaf bases), Cycadeoidea (branches, cones, leaves, synangia, and trunks), Cycadeoidella (Nishida 1994), Cycadolepis (cone bracts and scales, Crane and Herendeen 2009), Dictyozamites (leaves), Exesipollenites (pollen), Ischnophyton (stems and fronds), Neozamites (leaves), Nilssoniopteris vittata (leaves, Crane and Herendeen 2009), Otozamites (leaves), Pseudocycas (leaves), Pterophyllum (leaves), Ptilophyllum (leaves), Rehezamites (leaves), Tyrmia (leaves), Weltrichia (pollen cones), Williamsonia (ovulate cones), and Zamites (leaves).
The plate reproduced on the left shows reconstructed, detached organs of bennettitaleans. It is Figure 11 from Peter R. Crane (1985), Phylogenetic analysis of seed plants and the origin of angiosperms, Annals of the Missouri Botanical Garden 72: 716-796, reprinted with permission of the Missouri Botanical Garden and Peter Crane.
"Figure 11. Morphology of Bennettitales. -A. Williamsonia coronata, longitudinal half-section through 'flower,' based on Harris (1944, fig. 1); × 3. -B. Nilssoniopteris vittata, redrawn from Harris (1969, fig. 32A), an unusual specimen showing two lateral pinnae at the base; × 1. -C. N. vittata, detail of stoma showing paracytic guard cells and sinuous anticlinal flanges, based on Harris (1969, fig. 32D); × 1,000. -D. W. coronata, surface of immature 'gynoecium' showing micropyles and interseminal scales, based on Harris (1969, fig. 62F); × 50. -E. Cycadeoidea sp., longitudinal half section through 'flower,' based on Wieland (1906, fig. 88) and Crepet (1974, pl. 61, fig. 21); × 1.5."
Several potentially interesting points of discussion on the assumed relationship of phytophagous insect associates with cycadeioids may be found in Crepet (1974) and T. N. Taylor et al. (page 732 and 741, 2009).
"It is possible that predation pressure was not only responsible for the phylogenetic 'closing' of the microsporophylls, but was also responsible for the position of the cones on the trunks of Cycadeoidea which offers additional protection."
The above quotation is from page 160 of Crepet (1974), "Investigations of the North American cycadeoids: the reproductive biology of Cycadeoidea," Palaeontographica Abt. B: 148.
Known gross morphological characters of bennettitaleans, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--shrubs (Crepet 1974, Rothwell et al. 2009)
REPRODUCTIVE MODULES--simple cones with helically arranged bracts, ovule-bearing leaves (megasporophylls), interseminal scales, and pollen-bearing leaves (microsporophylls). Nectar glands, ovules, pollen, and seeds are present. Ovules are born on the distal ends of megasporophylls. The nucellus is free from the integument and a tubular micropyle is absent. A pollen chamber is absent and the micropyle is closed by a nucellar plug (Rothwell et al. 2009)
Fragments of Williamsoniella coronata consist of ovulate parts (with monosulcate pollen adhering to some surfaces), fragments of pollen-producing organs, scale leaves, and interseminal scales of a larger bisexual structure (Crane and Herendeen 2009)
LEAVES--simple and strap-shaped or compound pinnate and palm-like (Boyd 2000, G. Sun et al., pages 188-190, 2001), xeromorphic with thick cuticles (Villar de Seoane 2001)
PHYTOPHAGOUS ASSOCIATE(S)--indeterminate beetles, helid and nemonychid weevils, possible pollinating parandrexid or protoceline chrysomelid beetles, unidentified nectar-feeders, and insect eggs (Crepet 1974, Dmitriev and Ponomarenko 2002, Pott et al. 2008)
PLANT IDENTIFICATION(S)--bisexual fructifications of Cycadeoidea with burrows (mines), insect traces in various Mesozoic fructifications (Crepet 1974, Dmitriev and Ponomarenko 2002)
HOST SEED PLANT ORGAN(S) BEING EATEN--leaf tissues, ovules, and pollen?
Were peculiar modifications of female and male bennettitalean organs, specifically hairs and scales, able to deliver toxic alkaloid and polyacetylene toxins to potential vertebrate herbivores? Did extinct cycadeoids possess these antiherbivory chemical warfare agents?
Caytoniales. The enigmatic pteridosperm order Caytoniales first appeared in the Triassic Period. By Jurassic and Cretaceous time, caytonialeans comprised a minor but important floristic element indigenous to terrestrial biomes on the northern landmasses (Krassilov 1977, 1997).
Caytonia is a sister group to angiosperms in many emerging phylogenetic reconstructions (D. E. Soltis et al. 2005, J. A. Doyle 2006, J. A. Doyle 2008, Sean W. Graham and Iles 2009).
Despite pivotal discussions on Caytoniales and the origin of the angiosperm carpel (Thomas 1925, Harris 1940, Harris 1951, J. A. Doyle 1978, J. A. Doyle 2006), the group is one of the least known of Mesozoic seed plant clades (Stockey and Rothwell 2009, T. N. Taylor et al. 2009)
The image to the right and the figure legend below in quotation marks is from page 753 of Peter R. Crane (1985), Phylogenetic analysis of seed plants and the origin of angiosperms, Annals of the Missouri Botanical Garden 72: 716-796, reprinted with permission of the Missouri Botanical Garden and Peter Crane.
"Figure 17. Morphology of the Caytonia plant. -A. Sagenopteris colpodes, based on Thomas (1925, pl. 15, fig. 50); × 0.75.
-B. S. colpodes, detail of leaflet attachment and venation, based on Harris (1964, fig. 2H); × 4.
-C. Caytonanthus arberi, based on Harris (1941, pl. 2, fig. 3); × 7.
-D. Pollen from C. arberi, based on Townrow (1962b, fig. 3d, e); × 1,200.
-E. Caytonia nathorstii megasporophyll, based on Harris (1964, fig. 10-A-C); × 5.
-F. Caytonia "cupule" containing seeds, based on Reymanowna (1973, particularly text fig. 12E, F); × 12.5.
-G. Caytonia "cupule," longitudinal section, based on Reymanowna (1973, particularly text fig. 12E, F); × 12.5.
-H. C. nathorstii ovule longitudinal section, redrawn from Harris (1958, fig. 7); × 110."
Several form genera were described by Harris (1940, 1941, 1951, 1964, 1971), including Amphorispermum (seeds), Caytonanthus (pollen-bearing organs), Caytonia (cupules), Ktalenia (cupules), Sagenopteris (leaves), and Vitreisporites (pollen). Barbacka and Boka (2000) provide additional details.
General evolutionary relationships of caytonialeans with other seed plant groups are discussed by Thomas (1925), Harris (1940, 1951), Krassilov (1997), J. A. Doyle (2006), Stockey and Rothwell (2009), and E. L. Taylor and T. N. Taylor (2009). Krassilov (1997) draws a possible evolutionary connection between caytonialeans and ranunculid flowering plants. Detailed discussion of the morphology and anatomy of this group of anthophytes, homology of characters, phylogenetic relationships with other seed plants, and the origin of angiosperms may be found in Krassilov (1977) and J. A. Doyle (2006).
Interestingly, certain Asian paleobotanists are drawing a possible evolutionary connection between Leptostrobus (Czekanowskiales) and Caytoniales or Irania (X.-Q. Liu et al. 2006). This is surprising since the Caytoniales are sister to flowering plants in some phylogenetic reconstructions (J. A. Doyle 2006) and Czekanowskiales are often considered antecedents of Ginkgo (Ginkgoales).
If paleobotanical evidence and homology reassessment places Leptostrobus in the Caytoniales, and phylogenetic reconstructions support a close relationship with Ginkgo then recently published phylogenetic reconstructions of seed plants (including Amborella and Hydatellaceae) might be incorrect. The only solution to this problem is to find and describe permineralized reproductive and vegetative material of Caytonia, Irania, and Leptostrobus and reassess homology of critical morphological characters defining these groups.
Further, T. N. Taylor et al. (page 637, 2009) illuminate similarities between the cupules of Petriellaea (Petriellales) and Caytonia. Known gross morphological characters of the Caytoniales, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--possibly shrub-like (Reymanowna 1974, Krassilov 1997). Coalifications, compressions, permineralizations, and petrifactions are needed to elucidate the form of whole plants
REPRODUCTIVE MODULES--cupules, eight to 30 bitegmic ovules per cupule, and saccate pollen
The reproductive biology of caytonialeans is not clearly understood, but some research progress has been made (Krassilov 1997). Additional discoveries of permineralized sexual structures are needed to better understand these seed plants
LEAVES--compound palmate, the pinnae are lanceolate and glossopteroid in general aspect; leaves have been classified in the form genus Sagenopteris (Krassilov 1997)
PHYTOPHAGOUS ASSOCIATE(S)--indeterminate insect traces on foliage (Dmitriev and Ponomarenko 2002)
PLANT IDENTIFICATION(S)--Sagenopteris
HOST SEED PLANT ORGAN(S) BEING EATEN--leaves (Dmitriev and Ponomarenko 2002)
Corystospermales. Corystosperms were once a dominant forest vegetation type of Gondwana. The Dicroidium flora probably replaced the Glossopteris flora of the southern latitudes (White 1986) while many glossopterids were probably extirpated by the end-Permian extinction (EPE). The Corystospermales represented by the Paleogene Tasmanian fossil Komlopteris cenozoicus, is the only group of seed ferns that survived the K-T asteroid impact (McLoughlin et al. 2008).
The pteridosperm order Corystospermales is receiving more interest based upon recovery of permineralized reproductive material from Antarctica, the Middle East, and North America, including, for example, preserved pollen organs described as Pteruchus; cupules, pollen, and other detached remains (Pigg et al. 1993, Osborn and Taylor 1993, Yao et al. 1995, Axsmith et al. 2000, Kerp et al. 2006, T. N. Taylor et al. 2009). Compression fossils of ovulate axes described as Umkomasia are known from late Permian beds in India (S. K. Chandra et al. 2008), the Triassic Molteno Formation of South Africa (Axsmith et al. 2000), Upper Triassic of northern Asia (Shuqin et al. 2008), and from Jurassic deposits.
Corystospermales thought by some workers as possible angiosperm antecedents (Frohlich 2002) are not sister to the branch leading to flowering plants, but cluster instead in another clade with peltasperms, Gnetales, conifers, and Ginkgo (D. E. Soltis et al. 2008).
Several form genera of corystosperms are known including Alisporites (pollen), Dicroidium (foliage), among other leaf morphotypes; Karibacarpon (detached ovules), Pilophorosperma (ovules and associated leaves), Pteroma (ovulate organs), Pteruchus (pollen organs and associated leaves), Rhexoxylon (permineralized wood), and Spermatocodon (ovulate organs).
The image to the left and figure legend is from page 755 of Peter R. Crane (1985), Phylogenetic analysis of seed plants and the origin of angiosperms, Annals of the Missouri Botanical Garden 72: 716-796, reprinted with permission of the Missouri Botanical Garden and Peter Crane.
"Figure 18. Morphology of corystosperms. -A. Pachypteris papillosa stem with leaves, based on Harris (1983a, fig. 2): × 0.25 [transposed caption letters G, H, and I were corrected by J. M. Miller in 2008].
-B. Pteroma thomasii synangium, abaxial view, redrawn from Harris (1964, fig. 66B); × 0.25.
-C. P. thomasii, lateral view, based on Harris (1964, fig. 66A-G, I); × 2.
-D. Dicroidium odontopteroides, redrawn from Thomas (1933, fig. 49a); × 0.75.
-E. Umkomasia macleanii, redrawn from Thomas (1933, fig. 1, pl. 26, fig. 56); × 2.5.
-F. Corystosperm ovule based on Thomas (1933, fig. 33c); × 4.
-G. [I]. Rhexoxylon, transverse section of stem, based on Archangelsky and Brett (1961, fig. 2A); × 0.5.
-H. [G]. Pollen of Pteruchus africanus, redrawn from Townrow (1962a, fig. 1A-D); × 2.5.
-I. [H]. P. africanus, based on Townrow (1962a, fig. 1A-D); × 2.5."
Most of the form genera of corystosperms are now known to be attached to massive trees (Dicroidium odontopteroides) that once formed a prevalent forest type on southern portions of Pangaea (Axsmith et al. 2000). The phylogenetic position of corystosperms as a group, is problematic (Klavins et al. 2002), but a position within the anthophyte evolutionary clade is demonstrable.
Known gross morphological characters of the Corystospermales, and a list of phytophagous animal associates, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--shrubs and trees (M. E. White 1984)
REPRODUCTIVE MODULES--cupules or carpel-like organs of Doylea tetrahedrasperma (Stockey and Rothwell 2009)
The reproductive biology of corystosperms is not clearly understood, but some research progress has been made (Stockey and Rothwell 2009). Additional discoveries of permineralized sexual structures are needed to better understand these seed plants
LEAVES--compound pinnate in general aspect (M. E. White 1984)
PHYTOPHAGOUS ASSOCIATE(S)--poorly understood
PLANT IDENTIFICATION(S)--not applicable
HOST SEED PLANT ORGAN(S) BEING EATEN--leaves (Dmitriev and Ponomarenko 2002)
Frohlich in his update of the mostly male theory (MMT) of flower origins raises the interesting suggestion that early Mesozoic corystosperms might be flowering plant ancestors (Frohlich 2002). However, despite an earlier assertion that the morphology of Pteroma supports MMT, there are problems in comparing the ovule morphology of flowering plants with corystosperms (J. A. Doyle 2008).
Corystosperms represented by the Paleogene fossil Komlopteris cenozoicus, apparently survived the K-T asteroid impact, and co-occur with angiosperms, conifers, cycads, marine dinoflagellates, and ferns in a Tasmanian paleoflorule (McLoughlin et al. 2008). Some of the recent review papers on corystosperms include Axsmith et al. 2000, Klavins et al. (2002), Axsmith et al. (2007), and Stockey and Rothwell (2009).
Pentoxylales. Taeniopteroid foliage, fragments of short-shoots, and reproductive structures were recovered from Jurassic sediments in the Rajmahal Hills of the Indian sub-continent and described as a new group of seed plants by Birbil Sahni in 1948. The fossil history and morphology of the group is reviewed by T. N. Taylor et al. (pages 768-773, 2009)
Leaves attached to fragments of shoots of the whole plant Pentoxylon sahnii were first described as the form genus Nipaniophyllum.
Taeniopterid leaves which were first described from several Australasian Jurassic fossil plant localities as Taeniopteris spatulata, are now known to be attached to whole plants referable to Pentoxylon. Detached ovulate axes were initially described as Carnoconites but now too, belong to Pentoxylon (Sahni 1948, Wesley 1963, among others).
The image to the right is Figure 19 Peter R. Crane (1985), Phylogenetic analysis of seed plants and the origin of angiosperms, Annals of the Missouri Botanical Garden 72: 716-796, reprinted with permission of the Missouri Botanical Garden and Peter Crane.
"Figure 19. Morphology of Pentoxylon plants. -A. Carnoconites cranwelliae, ovulate heads, based on Harris (1962, text-fig. 2B, fig. 1): × 2.5 [transposed caption letters B and C were corrected by J. M. Miller in 2008].
-B. [C]. Carnoconites, longitudinal section through ovule, based on Sahni (1948, fig. 21); × 10.
-C. [B]. Nipaniophyllum raoi, redrawn from Sahni (1948, fig. 34 a,b); × 1.
-D. Pentoxylon sahnii, transverse section of stem showing vascular strands, based on Sahni (1948, fig. 9); × 8.
-E. Sahnia microsporangiate 'flower,' based on Vishnu-Mittre (1953, fig. 11) and Bose et al. (in press [1985]); × 2.5."
Following in the footsteps of Sahni (1948), Vishnu-Mittre (1953) described pollen-bearing "flowers" of the Pentoxylales. Rao (1976, 1981) provides additional discussion of the relationships of the Pentoxylales with other gymnosperms.
The anatomy, paleoecology, and taxonomy of pentoxylaleans indigenous to the Antarctic polar forest of Cretaceous times was reviewed by Howe and Cantrill (2001). In Mesozoic times these plants were probably shrub-like not unlike some of the bennettitaleans. Additional discussion of the biodiversity and paleobiogeography of this enigmatic seed plant group may be found in Wesley (1973) and T. N. Taylor et al. (2009).
Known gross morphological characters of the Pentoxylales, references, and future research needs are summarized below.
WHOLE PLANT MORPHOLOGY--shrubs and trees? (T. N. Taylor et al. (2009)
REPRODUCTIVE MODULES--ovulate clusters are described as Carnoconites. Howe and Cantrill (2001) provide a detailed discussion of the homologies of Carnoconites ovule-bearing spur shoots, which are almost certainly evolutionarily reduced megastrobili lacking laminar megasporophylls.
Pollen-bearing cones and microsporophylls are absent and reduced to naked, stalked microsporangia arising from spur shoot apexes illustrated above and in Figure 19.25, page 771, T. N. Taylor et al. (2009)
LEAVES--detached leaves have been described as Taeniopteris daintreei (see page 782, Figure 2, Howe and Cantrill 2001)
PHYTOPHAGOUS ASSOCIATE(S)--unknown
PLANT IDENTIFICATION(S)--not applicable
HOST SEED PLANT ORGAN(S) BEING EATEN--not applicable
There are few, if any anecdotal accounts of phytophagous animal associates of pentoxylaleans. Data are needed from study of existing museum specimens.
Phylogenetic Considerations:
This final chapter considers phylogenetic inference as a tool to shed light on the origin of angiosperms, evolutionary-development (evo-devo) of perianth morphology sensu Hileman and Irish (2009), and evolution of hermaphroditic cone axes (Baum and Hileman 2006, Theißen and Melzer 2007).
The cartoon was drawn by Sul Ross State University geology student Mark Munday in 1981. The artwork by Mark, which is furnished David Rohr, Ph.D., and Mark's help at the delnortea beds is gratefully acknowledged.
I assessed homologies of traditional seed plant characters, polarized these characters, and conducted two preliminary analyses from respective data matrices (Tables 4 and 5). Two separate phylogenies (one analysis includes two Devonian progymnosperm taxa) were computed using a standard maximum parsimony algorithm.
An ontogenetic emphasis is considered in some aspects of my thinking, which is not unlike alternative ideas on character polarization in the absence of outgroups (Nelson 1978) and the concepts of biological and process-based homology (G. P. Wagner 1989, Laubichler 2000, Brigandt 2003).
I determined character polarities and rationale for assigning one-to-one correspondence along the lines proposed by Stevens (1984) and Zelditch and Fink (1996), but also considered a nested approach (H. E. Schneider et al. 2002).
The prevailing view of molecular systematists, morphologists, and paleobotanists is that seed plants form a monophyletic group derived from early Devonian vascular land plant stock (Crane 1985, Rothwell and Serbet 1994, Hilton and Bateman 2006, J. A. Doyle 2006).
I do not agree with past ideas on a monophyletic origin of seed plants. Based on models of fertile SAM organization derived from evo-devo research (Baum and Hileman 2006, Theißen and Melzer 2007), pteridosperms might not constitute "the backbone of seed-plant phylogeny" (title, Hilton and Bateman 2006).
The present review of biochemical models involving floral homeotic protein quartets, evolutionary-developmental (evo-devo) studies of extant model vascular plant model organisms, and molecular phylogenetic analyses of homeodomain proteins and cis-acting homeotic transcription factors suggests that several (at least two) lines of seed plant evolution might have existed 300 to 400 m.y.a., during the Devonian Period.
This conclusion is based on studies of the molecular evolution of KNOX/ARP homeodomain proteins (Beerling and Fleming 2007, Rosin and Kramer 2009), evolution of LEAFY [LFY] genes following whole genome duplications [WGDs] (Mouradov et al. 1998, Shindo et al. 2001, Vázquez-Lobo et al. 2007), and molecular evolution of homeotic MIKC-Type MADS-box protein quartets (Theißen and Melzer 2007).
Modifications to the transcriptional machinery of seed plant SAMs could potentially lead to drastic changes in morphology of organs and reproductive modules, with profound implications toward traditional ideas of homology of seed plant organs. For example, carpels, cupules, fertiligers, integuments, microsporophylls, megasporophylls, ovules, and strobili are important in cladistic analyses of seed plants.
This means that in disparate seed plant groups, traditional ideas and supposed homologies of organs such as ovular integuments (origin by evo-devo splitting of tissues or by incorporation of foliar organs), and position of ovules on the megasporophyll whether adaxial or abaxial (Figures 13 and 14, J. A. Doyle 2006), are called into question.
Further, location and canalization of sporophylls in relation to the SAM or lateral meristems might have evo-devo significance.
Can angiosperm carpels share homology with the cupules of extinct seed plants if they were derived by completely different evo-devo programs of cis-acting transcriptional regulation?
Molecular phylogenetics is one of the most important tools in providing insight into the timing of the origin of angiosperms, evolution of the perianth, and diversification rates of the crown group of flowering plants (J. A. Doyle 2001, D. E. Soltis and P. S. Soltis 2003, Magallón and Sanderson 2005, Bateman et al. 2006, Sanderson and McMahon 2007, Hileman and Irish 2009, Magallón and Castillo 2009), despite a persistent long branch that separates extant flowering plants from gymnosperms (Sean W. Graham and Ihles 2009).
Gene duplications may shed light on long branch attraction and other sources of errors during computation of molecular phylogenetic trees according to a critical analysis that contrasts "bottom-up" and "top-down" approaches (Bateman et al. 2006). Polyploidy including whole-genome duplications (WGDs) is one of driving forces behind diversification of the flowering plant clade (Sanderson and McMahon 2007, D. E. Soltis et al. 2009, P. M. Soltis et al. 2009).
A critique of phylogenetic approaches toward a better understanding of the origin of angiosperms and evolutionary development of the ancestral flower by Bateman et al. (2006) suggests that too much reliance is placed on "empirical rigour" of large nucleotide and indel data matrices derived from very few extant plant species. Further, the review by Bateman, Hilton and Rudall states, "whereas morphology-based approaches increase the number of phylogenetically informative taxa (including fossils) at the expense of accessing only a restricted spectrum of phenotypic characters" (abstract, Bateman et al. 2006).
Bateman et al. (2006) provide a detailed discussion of several questions posed in their review:
"What is a flower?
Which are the (other) benchmark studies in floral evolution?
Which are the benchmark studies in reconstructing seed-plant phylogeny?
Which kinds of phylogeny are of greatest value?
How significant are optimization and outgroups choice?
What is the best way to choose among the plethora of phylogenies?
What is the best way to assess the quality of a phylogeny?
Do some nodes in the land-plant phylogeny merit particular emphasis?
Why is it important that optimal taxon sampling dissects long branches?
What are the effects of culling taxa and/or characters?
What are the pros and cons of monophyly?
Which key characters best define the flower?
Why are palaeobotanists obsessed with extinct gymnosperms?
Are clades such as angiosperms best defined using taxa or characters?
Do some categories of character merit particular emphasis?
What is the preferred research programme for the next decade?"
Top-down and bottom-up approaches in unraveling the ancestry of seed plants and origin of the flower are reviewed by Specht and Bartlett (2009) and Rudall and Bateman (2010).
Major milestone papers in molecular and morphological phylogenetics of vascular plants including monilophytes, seed plants, and angiosperms are Martin et al. (1989), K. H. Wolfe et al. (1989), Bousquet et al. (1992), Hamby and Zimmer (1992), Chase et al. (1993), Martin et al. (1993), J. A. Doyle et al. (1994), Savard et al. (1994), B. G. Baldwin et al. (1995), Barraclough et al. (1996), Goremykin et al. (1996), Sytsma and Baum (1996), Goremykin et al. (1997), D. E. Soltis et al. (1997), Chase and Albert (1998), J. A. Doyle (1998), Kuzoff et al. (1998), Nandi et al. (1998), P. S. Soltis and D. E. Soltis et al. (1998), Wolf et al. (1998), Mathews and Donoghue (1999), Parkinson et al. (1999), Samigullin et al. (1999), and P. S. Soltis et al. (1999).
Key papers on the molecular phylogenetics of lignophytes including angiosperms as a whole, published during the last ten years include Barkman et al. (2000), Chaw et al. (2000), J. A. Doyle and Endress (2000), Savolainen et al. (2000), Wikström et al. (2001), Sanderson and J. A. Doyle (2001), Magallón and Sanderson (2002), Rydin et al. (2002), P. S. Soltis et al. (2002), Sanderson et al. (2004), Magallón and Sanderson (2005), Y. L. Qiu et al. (2005), Jansen et al. (2007), D. E. Soltis et al. (2008), Braukmann et al. (2009), Mathews (2009), H. E. Schneider et al. (2009), Tillich et al. (2009), and de la Torre-Bárcena et al. (2009).
Reviews on the origin of angiosperms and the flower by J. A. Doyle (1994, 2006, 2008), D. E. Soltis et al. (2008), Specht and Bartlett (2009), and Rudall and Bateman (2010), outline progress on the use of phylogenetics to resolve the big picture of seed plant evolution. Figure 1 on page 5 of D. E. Soltis et al. (2008) or Figure 1 on page 219 of Specht and Bartlett (2009) are places to begin discussion of top-down phylogenetic inference on the origin and relationships of stem- and crown-group flowering plants.
The graphic below represents a phylogenetic reconstruction of seed plants which was redrawn from J. A. Doyle (2008). Professor Doyle's tree represents the most parsimonious cladogram of his several phylogenetic reconstructions. I rotated the original drawing 45º to the right and left off other details including the number of steps and labels of higher orders of formal and informal classification. A detailed discussion of the analysis, character states, and the original data set may be found in the original paper (J. A. Doyle 2008).
The preceding graphic is redrawn from Fig. 6 on page 827 of J. A. Doyle (2008), Integrating molecular phylogenetic and paleobotanical evidence on origin of the flower, International Journal of Plant Sciences 169(7): 816-843.
Some of the seed fern groups are colored indigo brown on the dendrogram shown above. Corystosperms and glossopterids appear in blue typescript. Bennettitaleans are denoted by the reddish-brown type-face. Pentoxylales are colored pink on the chart. Gnetophytes, once regarded as a sister group to flowering plants (J. A. Doyle and Donoghue 1986, 1987; Donoghue and J. A. Doyle 2000) are displayed as purple letters. Cycadales and Caytonia are shown on the graphic in green letters. Common groups of conifers and the ginkgos appear in brown type. Finally, some of the critical fossil groups of flowering plants and extant angiosperms are depicted in red type.
Important reviews of seed plant phylogeny including angiosperms by Crane (1985), J. A. Doyle and Donoghue (1986), J. A. Doyle and Donoghue (1987), Loconte and Stevenson (1990), J. A. Doyle and Donoghue (1993), Nixon et al. (1994), Rothwell and Serbet (1994), J. A. Doyle (1998), C. N. Miller, Jr. (1999), J. A. Doyle (2001), J. A. Doyle (2006), Hilton and Bateman (2006), D. W. Taylor et al. (2006), Erbar (2007), J. A. Doyle (2008), Chase and Reveal (2009), J. A. Doyle (2009), Mathews (2009), and Rothwell et al. (2009), overlook several groups of possibly important Paleozoic stem group gymnosperms.
Two overlooked Permo-Carboniferous seed plant groups: Vojnovskyales (Mamay 1976) and gigantopterids (page 162, G. Sun et al. 2001) have been discussed as possible angiosperm ancestors.
Vojnovskyales constitute a small order of hermaphroditic gymnosperms with monocot-like foliage known from rocks of Carboniferous Period (Mamay 1976, Rothwell et al. 1996, Naugolnykh 2001). Interestingly, these seed plants with their bisexual cone axes existed during the interval in deep time when molecular phylogenetic data suggest that the angiosperms and gymnosperms first diverged (Bowe et al. 2000).
Albeit, Mamay's review (page 295, 1976) was uncomplimentary however, the taxon in question (Vojnovskya), the original publication by Professor Maekawa (in Japanese), and sister genera of Vojnovskyales should be evaluated and incorporated into phylogenetic analyses and discussions of seed plant evolution. Rothwell et al. (1996) is one of the only papers that evaluates Vojnovskyales from a phylogenetic perspective.
Oleanane-containing Permian seed plants (Moldowan and Jacobson 2002) known as gigantopterids, are missing from almost all phylogenetic trees published in the literature, with the exception of D. W. Taylor et al. (2006). This is understandable because fossilized reproductive material (X.-G. Li and Z.-Q. Yao 1983) is controversial, fragmentary, and inadequately described as morphotype genera. The original papers by Asama (1960, 1982) and H. D. Zhang (cited in G. Sun et al. 2001) detailing an origin of flowering plants from gigantopterids will require translation, discussion, and reanalysis in a coevolutionary and phylogenetic context, in my opinion.
Hypothetical gigantopteroid proanthostrobili. In the first essay on the Origin of Angiosperms I discussed a recent paper by Rothwell et al. (2009), revisited the main hypotheses on the origin of flowering plants and the flower, and proposed a novel hypothesis based on molecular coevolution of insect and seed plant developmental tool kits and cis-regulatory modules.
I also stated that a few elements of the anthophyte hypothesis (J. A. Doyle and Donoghue 1986) and related classic proposals by Arber and Parkin (1907), Asama's Growth Retardation Theory (Asama 1982), Takhtajan's ideas on a neotenous origin of angiosperms (1976), the LEAFY (LFY) gene duplication hypothesis by Albert et al. (2002), and out-of-male and out-of-female hypotheses involving floral quartets in compartments of meristematic cells of developing shoot apical meristems (SAMs) of bisexual cone axes (Theißen Saedler 2001, Becker and Theißen 2003, Baum and Hileman 2006, Theißen and Melzer 2007), dovetail with- and potentially supported a coevolutionary hypothesis on the origin of flowering plants.
Rothwell et al. assert (page 317, 2009) that missing links and "transformational series" in the fossil record in support of the anthophyte hypothesis have yet to be discovered. These workers however, neglect to discuss important clues from detailed anatomical studies of pieces of detached fertile Paleozoic leaves once thought to be related to cycads (Axsmith et al. 2003), or compare evolutionarily precocious and innovative leaf anatomy (Mamay et al. 1988), and venation density patterns (Boyce et al. 2009) of oleanane-containing Permian gigantopterids (D. W. Taylor et al. 2006) to Bennettitales, Gnetales, and eudicot flowering plants.
While the whole plant morphology of Permo-Carboniferous gigantopterid and Phasmatocycas gymnosperms remains a mystery detached pieces of ovule-bearing Phasmatocycas bridwellii megasporophylls, undescribed microsporangia-bearing leaves (preceding section), and gigantopteroid-like petaloid organs (discussed herein), might be fragments of much larger proanthostrobili with a well-defined perianth and elongated or conical Magnolia-like receptacle.
The diagram below is a reconstructed longitudinal section of the short shoot and floral organ of a hypothetical gigantopteroid seed plant. A 25-centimeter broad bracteopetaloid perianth consisting of retuse Lonesomia mexicana-like taeniopteroid tepals is diagrammed. The hermaphroditic spur shoot attached to a small tree, pachycaulous shrub, or vine-like mother plant might have been subtended by clasping gigantopterid leaves clothing a long shoot or vegetative branch.
Spirally-arranged tepals would be attached to the base of a shortened hermaphroditic cone axis illustrated above. Just inside a spiral of several tepals, a third ring might have consisted of several spirally arranged ultraviolet-absorbing and/or colored andropetals, which are enclosing an inner ring of microsporangium-bearing leaves (microsporophylls). The leaf-like microsporophylls would clasp an innermost ring of Phasmatocycas bridwellii-like ovule-bearing megasporophylls.
Are unambiguous fossils of whole bisexual gigantopteroid proanthostrobili known from Paleozoic rocks? No.
However there are five examples of detached pieces and whole plant fragments known from Carboniferous and Permian sediments which have been described and illustrated by Axsmith et al. (page 1586, Figures 1-5, 2003), X. Li and Yao (page 26 and Plate 4, 1983), Mei et al. (page 101, Plate 2, 1992), R. Weber (page 232, Plate 3, 1997), and T. N. Taylor et al. (page 713, Figure 17.23, 2009). A color image of a Phasmatocycas bridwellii distal branch (I interpret this fossil as an incomplete proanthostrobilus) is provided by T. N. Taylor et al. (page 710, Figure 17.15, 2009).
The Paleozoic fossil Sobernheimia jonkeri (Kerp 1983; page 713, Figure 17.23, T. N. Taylor et al. 2009) might also be detached megasporophylls of a massive bisexual gigantopteroid proanthostrobilus with the microsporophylls and perianth also shed and therefore, evading a complete diagnosis of the whole fertile SAM.
While discussing fossilized "distinct clusters" of fertile and sterile taeniopterid leaves on page 1591 Axsmith et al. (page 1586, Figure 1, 2003) introduce the descriptor "flushes" to the reader.
Could the term "flush" be replaced with the horticultural term "blossom" when describing an anthocyanic hypothetical bisexual proanthostrobilus of a Paleozoic seed plant?
Further, attachment points of some Paleozoic taeniopterid laminae are remarkably similar to tepal bases of ANITA grade basal angiosperms (page 847, Figure 2, Endress 2008). Unfortunately, it is not clear from Reinhard Weber’s 1997 paper where taeniopterid laminae were attached to the whole gigantopterid plant.
Interestingly, Paleozoic rock layers that yield detached gigantopterid leaves with clasping leaf bases are often accompanied by taeniopterid leaves possessing wedge-shaped bases. The possibility should be entertained that both kinds of leaves came from the same deciduous plant. Taphonomic data are needed in most cases when unearthing and cataloging leaf-bearing Permian rock beds. Paleontologists should also be on the lookout for stem fragments with preserved leaf bases in gigantopterid-bearing sedimentary layers. Oleanane data are needed in most cases.
The hypothetical gigantopteroid proanthostrobilus is potentially equivalent to the ancestral angiosperm protoflower with all attendant biological and evolutionary implications expressed in the Strobilus Theory of the Angiosperm Fructification (page 37 and Figure 1 on page 44, Arber and Parkin 1907); and by Crane (1985), J. A. Doyle and Donoghue (1986), J. A. Doyle (2008), Endress (2008), Endress and J. A. Doyle (2009), Hileman and Irish (2009), and Sage et al. (2009).
While discussing salient findings of key papers published during the last decade of the 20th Century, the anthophyte hypothesis, and refutation of the idea that Gnetales are related to angiosperms, Friis et al. (2005) assert:
"As a result, the relationships of angiosperms remain uncertain and the prospects for understanding the origin of characteristic angiosperm features, such as the carpel and stamen, are as remote as ever."
I disagree with this opinion.
Models of transcriptional regulation based upon our understanding of LEAFY, NEEDLY, and MIKC-type MADS-box gene expression (Baum and Hileman 2006), and molecular phylogenetic studies of 300-million year duplications of MIKC-type MADS-box genes, support a prediction of a Permo-Carboniferous gigantopteroid proanthostrobilus.
Further, our knowledge of carpel, floral, and ovular transcriptional regulators in extant angiosperm model organisms does not preclude derivation of evo-devo models that explain curling, inrolling, and fusion in 300 million-year-old ovule-bearing gigantopteroid Phasmatocycas bridwellii leaves to form carpels, ovaries, and pistils.
Therefore, 160 million years of neotenic evolution to include condensation of hypothetical gigantopteroid proanthostrobili is the most simple evo-devo process to explain the origin of reproductive organs in Mesozoic crown group angiosperms and extant basal Amborellanae, Austrobaileyanae, Nymphaeanae, and Magnolianae.
"... Equally puzzling is that despite intense interest in the origins of seed plants and angiosperms throughout the entire last century, few have looked at the problems from a life cycle evolutionary developmental perspective, with perhaps one exception (Takhtajan 1976), who alluded to neoteny as one of the possible mechanisms contributing to the origin of angiosperms."
The above quotation is from page 296 of Y.-L. Qiu (2008), Phylogeny and evolution of charophytic algae and land plants. Journal of Systematics and Evolution 46(3): 287-306.
Did Earth "start blooming" during the Permian or Carboniferous Period?
Further, based on molecular phylogenetic studies of MYB transcription factors in seed plants and arthropod olfactory and nuclear receptors and vision-related proteins indicating a conserved evolutionary history, it would not be unreasonable to propose that the hypothetical proanthostrobilus illustrated above might have attracted (or repelled) large phytophagous insects including paleodictyopterans. An interesting geochemical study might involve screening for oleonone triterpanes among detached fragments of bisexual cone axes in the same bedding plane.
"... one of the biggest remaining challenges facing evolutionary biologists is ascertaining the fossil lineages that represent the closest relatives of angiosperms."
The preceding phrase is from page 4 of D. E. Soltis et al. (2008), Origin and early evolution of angiosperms, Pp. 3-25 In: C. D. Schlichting and T. A. Mousseau (eds.), Annals of the New York Academy of Sciences, Volume 1133 Issue, The Year in Evolutionary Biology 2008. New York: The New York Academy of Sciences, 203 pp.
Detailed paleobotanical studies of the anatomy and morphology of obscure, poorly studied Paleozoic and Mesozoic gymnosperms are needed to fill-in the considerable gaps in existing data sets before phylogenetic analyses can be calibrated with minimum fossil ages (Gandolfo et al. 2008). Yet, one recent paper proclaims:
"The oracle whether the angiosperms arose via rapid accumulation of the synapomorphies that characterize flowering plants (the carpel, double fertilization, flower) or through gradual accumulation of these traits over longer time (the 'transitional-combinational theory', Stuessy 2004) remains unanswered although results from 'evo-devo' research support the possibility of a more or less sudden flower origin. The coincidence between the origin and diversification of the class E genes, the duplication event in the class B genes, the decoupling of C and D function and the origin of angiosperms (Fig. 9) suggests that these genes are involved in the processes that made possible the morphological invention of the flower."
The preceding passage is quoted from page 127 of Erbar (2007), Current opinions in flower development and the evo-devo approach in plant phylogeny, Plant Systematics and Evolution 269: 107-132.
Possessing a perianth of taeniopteroid tepals and andropetals attached to the short shoot of a massive proanthostrobilus subtended by a vegetative gigantopterid leaf (thus the whole plant had dimorphic leaves), the hypothetical gigantopteroid would have been a completely different kind of Paleozoic gymnosperm, perhaps indigenous to upland or marginal environments away from coal swamps of the period, which were dominated by giant lycophytes, horsetails, and medullosan pteridosperms.
When supported by paleobotanical evidence, how would the existence of 300 million-year-old proanthostrobili with a differentiated perianth of tepals (andropetals or bracteopetals) dovetail with a "Mosaic Theory for the Evolution of the Dimorphic Perianth" (page 3571, Warner et al. 2009)?
Some of these hypothetical gymnosperms with proanthostrobili might possibly represent fertile natural intergeneric hybrids arising in long extinct populations of sympatric gigantopteroids and Vojnovskyales.
We should not forget that classical paleobotanists pieced together whole Lepidodendron (Lepidodendrales, Lycophyta) and Psaronius (medullosan pteridosperm) trees from fossilized detached parts originally described as morphotype genera.
"Perhaps the greatest biofantasy of whole-plant reconstruction is the misconception that organ attachment is the only way to ascertain organ association in an acceptable manner."
This statement is quoted from page 147 of W. A. DiMichele and R. A. Gastaldo (2008), Plant paleoecology in deep time. Annals of the Missouri Botanical Garden 95: 144-198.
Therefore, it is possible that some detached, sterile, retuse Taeniopteris leaves; gigantopterid megaphylls; and Phasmatocycas-like megasporophylls catalogued and distilled in T. N. Taylor et al. (2009), might be fragments of the same seed plant species or represent whole new gymnosperm genera and taxonomic orders.
When whole seed plants are reconstructed from fossilized detached parts, the resulting species and character analyses may be critical in unraveling potential reticulations in several independent lines of anthophyte evolution that existed on Earth at the time of the angiosperm-gymnosperm split predicted by molecular clock studies (but also based on the timing of whole genome duplications [WGDs]), prior to the end-Permian extinction, 251 m.y.a.
Dating the angiosperm-gymnosperm split. What is interval in geologic time when angiosperms diverged from gymnosperms determined from molecular substitution data and timing of whole genome duplications?
Stephen A. Smith et al. (2010) report that the flowering plant crown group originated 217 million years ago, which is the Norian Age of the late Triassic Period.
Based on morphology alone, the explosion of diversity of phytophagous coleopterans and dipterans following the end-Permian extinction (EPE) predates the oldest known flowering plants by more than 100 million years. Yet, certain paleobotanists and plant ecologists assert that angiosperms originated suddenly during the Cretaceous Period.
This view contrasts with rbcL molecular clock estimates by Savard et al. (page 5166, 1994) who estimate the interval in geologic time when "extant seed plants shared their last common ancestor," between 275 and 295 million years ago (m.y.a.) during the early Permian or late Carboniferous periods.
Molecular phylogenetic studies of WUS-WOX homeotic transcription factors (Nardmann et al. 2009), numerous molecular clock studies of phytochrome (Mathews and Donoghue 1999), and estimates gleaned from phylogenetic analyses of chloroplast and ribosome genes (Sanderson and J. A. Doyle 2001, Wikström et al. 2001, Stephen A. Smith et al. 2010), suggest that seed plants split into the angiosperms and gymnosperms during the late Carboniferous or Permian periods of the Paleozoic Era about 300 m.y.a.
Assertions and generalizations on a Mesozoic origin of flowering plants by certain paleobotanists are unsupported by molecular phylogenetic studies of the Class III HD-Zip gene family, which points to a gene duplication event leading to the REVOLUTA and PHX (PHABULOSA/PHAVOLUTA-related) and C8 (CORONA/HB8-related) clade about 300 million years ago during the Carboniferous Period just prior to the angiosperm-gymnosperm split (Prigge and S. E. Clark 2006).
Revised relaxed-molecular clock estimates (Stephen A. Smith et al. 2010) and molecular phylogenetic studies of homeodomain proteins go along with revelations on the timing of WGDs in certain MIKC-Type MADS-box genes (S. Kim et al. 2004, Zahn et al. 2005). Specifically, the B-class MADS-box gene duplication in the APETALA3/PISTILLATA (AP3/PI) genetic lineage that paved the way for the evolutionary development of the first flower probably occurred 230 to 290 m.y.a. (S. Kim et al. 2004).
A possible paraphyletic Paleozoic origin of angiosperms contradicts proposals by Leebens-Mack et al. (2005) suggesting a monophyletic Mesozoic origin of a basal clade of flowering plants, which is based on estimates derived from cpDNA data.
Bell et al. (2005) age the crown group of angiosperms from 106.1 to 229.4 m.y.a.. Another estimate of the age of angiosperms determined by penalized-likelihood methodology is 251.77 m.y.a. (H. E. Schneider et al. 2004), based upon Bayesian relaxed-clock methods (J. L. Thorne and Kishino 2002).
Should the evolutionary clock and root of the seed plant clade that leads to flowering plants be calibrated to the time of the end-Permian extinction, 251.3 m.y.a?
Calibrating points of divergence in geologic time using molecular substitution data is statistically questionable (Graur and Martin 2004) and the subject of continuing debate.
"Despite the singular ecological significance and species diversity of angiosperms, they are not in a genealogical sense one of the major branches of land plants and did not originate with other major land plant clades (e.g. lycopsids, ferns, conifers, cycads, gingkos) during the middle or late Paleozoic."
The preceding quotation is from page 380 of S. L. Wing and L. D. Boucher (1998), Ecological aspects of the Cretaceous flowering plant radiation. Annual Review of Earth and Planetary Sciences 26: 379-421.
Despite wide disagreement on the timing of the angiosperm-gymnosperm split, there is growing consensus among some molecular systematists and paleobotanists on the potential existence of a ghost lineage of flowering plants.
Evo-devo of progymnosperm and seed plant SAMs: Paleozoic splits. I concluded in the first essay on the origin of angiosperms that cladogenesis of flowering plants may be traced back in geologic time to the end-Permian extinction and to surviving remnants of already divergent Permian seed plant lineages.
Clues gleaned from biochemical studies of vegetative growth and SAM organization in extant model lignophytes and monilophytes, when combined with insight gained from molecular phylogenetic analyses of class III homeodomain leucine zipper (Class III HD-Zip) genes and KNOTTED1 homeodomain proteins suggest that at least two (possibly three or more) evolutionary-developmental (evo-devo) programs potentially existed in early diverging Devonian lycopsid, progymnosperm, and seed plant populations.
If this hypothesis is correct then traditional ideas on the homology of leaf and sporophyll characters, which are gene expression products of transcription factors downstream of the SAM patterning developmental tool kit, should be reconsidered.
Joel Cracraft (page 349, 2005) brings several "issues" to the table regarding evo-devo and homology, which have a bearing on supposed homologies of seed plant reproductive structures and the present analysis (the comment in brackets [] is mine):
"If evo-devo is also about evolution [it certainly is, see reviews by Raff 2000, G. B. Müller 2007], how is the evolution of development among taxa to be studied?
What are the developmental characters (entities) being compared?
Does it mean that the observations ('characters') of evo-devo are merely to be mapped on a tree of choice?
Or do the comparative observations themselves have a contribution to make about relationships, alternative hypotheses of which affect interpretations of change?
In short, can evo-devo advance as an evolutionary discipline using biological homology, process homology, or other similar concepts?"
Several important reviews discuss from morphological, paleobotanical, and phylogenetic research perspectives, the evo-devo of innovative morphologies in divergent Devonian vascular plant lineages (Boyce and Knoll 2002, H. E. Schneider et al. 2002, Friedman et al. 2004, J. A. Doyle 2006, Hilton and Bateman 2006, Stein and Boyer 2006, Beerling and Fleming 2007, David-Schwartz and Sinha 2007, Sanders et al. 2007). All the preceding papers were overlooked in a general review of the evolution of developmental mechanisms in plants by Langdale (2008).
I agree with some of the philosophical arguments published by H. E. Schneider et al. (page 331-332, 2002), which were employed in their analyses of "new sources of data to answer long-standing questions about plant evolution," namely "integration of phylogenetic reconstruction, morphological studies, and developmental genetic data" through phylogenetic inference based on "a series of nested studies" (the phrase in brackets [] is mine):
(1) adoption of nucleotide sequence data gleaned from biochemical studies of coding and/or non-coding regions,
(2) making use of anatomical, biochemical, cytological, and morphological data sets to infer character evolution,
(3) comparison of these sets of data with gene trees of key developmental genes and evo-devo programs, and
(4) casting the results in light of a fundamental trend observed in vascular plant evolution that body plans simplify [by neoteny] over the course of deep time.
The state of evo-devo biology has progressed much since H. E. Schneider and colleagues published their book chapter in 2002. Some aspects of the traditional Hennigian concept of homology have been challenged by insight gained from studies of transcriptional regulation by homeotic cis-regulatory modules (Abouheif 1997, Abouheif et al. 1997, Janies and DeSalle 1999, Wray 1999, Brigandt 2003, Theißen 2005, Rosin and Kramer 2009, Theißen 2009).
Peter Endress (2003) suggests that evo-devo and paleobotanical studies might solve several problems with evolution of the perianth in extant basal angiosperms of the crown group.
Molecular dating of the angiosperm-gymnosperm divergence at about 300 m.y.a. remains an important piece of the puzzle which allows paleobotanists to focus the quest for the ancestor of angiosperms (or its detached pieces) to sedimentary beds, late Paleozoic in age.
Taking into account critical reevaluation and discussion of the homology concept when dealing with morphological characters determined by cis-regulatory modules, it may be necessary to modify and/or add to the "nested" approach used by H. E. Schneider et al. (page 331-332, 2002), restated in bulleted form, above. That is, the third and fourth bulleted lines of evidence might be considered before conducting the character analyses (item 2).
"It is first necessary to have phylogenetic evidence that observed differences in gene expression are meaningful in terms of morphological evolution."
This quote was transcribed from page 163 of L. Reiser, P. Sánchez-Baracaldo, and S. Hake (2000), Knots in the family tree: evolutionary relationships and functions of KNOX homeobox genes. Plant Molecular Biology 42(1): 151-166.
Molecular phylogenetic studies of WUS-WOX homeotic transcription factors (Nardmann et al. 2009), numerous molecular clock studies of phytochrome (Mathews and Donoghue 1999), chloroplast, and ribosome genes (reviewed by Sanderson and J. A. Doyle 2001, Wikström et al. 2001), and revelations on the timing of WGDs in certain MIKC-Type MADS-box genes (S. Kim et al. 2004, Zahn et al. 2005) point to a late Carboniferous or Permian divergence of seed plants into the angiosperms and gymnosperms.
Vásquez-Lobo et al. (2007), in an elegant study of FLO/LFY orthologs in Picea (Coniferales), Podocarpus (Podocarpales), and Taxus (Taxales), while employing ultrasensitive radioactive probes and histochemical methods, clearly demonstrate in situ expression of FLO/LFY gene transcripts in all the organs of female fertile SAMs. Shindo et al. (2001) suggest:
"Thus, it is likely that the induction of MADS-box genes by FLO/LFY homologues was established after ferns diverged from the seed plant lineage but before extant gymnosperms and angiosperms diverged."
The preceding passage is quoted from page 1205 of Shindo et al. (2001), Characterization of FLORICAULA/LEAFY homologue of Gnetum parvifolium and its implications for the evolution of reproductive organs in seed plants. International Journal of Plant Sciences 162(6): 1199-1209.
Certain homeodomain proteins (e.g. KNOTTED1), FLO/LFY proteins, phytohormones, and other small-sized (<60 kilodalton) transcription factors and their mRNAs move from cell to cell via plasmodesmata in the SAMs of some of the model eudicot angiosperms studied (Jackson 2002, J.-Y. Kim et al. 2003, Jackson 2005). Experimental data are needed to verify trafficking of homeodomain transcription factors and FLO/LFY proteins in fertile SAMs of extant flowering plants and gymnosperms.
A paper by Pham and Sinha (2003) on the evo-devo of a non-model gnetophyte does not address potential significance of KNOX homeodomain protein movements from the SAM to leaf primordia, thence to regions of indeterminate growth, when discussing ontogeny of accessory scaly bodies and reproductive organs in Welwitschia mirabilis.
The physiology of SAM macromolecular trafficking involving phytohormones, when extrapolated to seed plants in deep time is an important component of two prevailing models of cone organization and floral origins (Baum and Hileman 2006, Theißen and Melzer 2007).
Two or more traffic patterns of homeodomain protein movement and phytohormone flow might have emerged in meristems of aerial shoots of extinct progymnosperms and seed plants, but paleontologic evidence from studies of permineralized cells and tissues is lacking.
Clues on the possible existence of polar auxin flow in woody stems have been uncovered in comparative anatomical studies of a 375 million-year-old fossilized progymnosperm known as Archaeopteris and modern conifers (Rothwell and Lev-Yadun 2005). The fossil record reveals that auxin regulation of secondary growth was a conserved physiologic process (Rothwell et al. 2008).
Did KNOTTED1 and FLO/LFY proteins (among others) move from cell-to-cell in the SAMs and axillary SAMs of extinct Paleozoic vascular plants?
Baum and Hileman (page 16, 2006) propose that since the gibberellin signaling pathway positively regulates LEAFY (LFY) gene expression (and LFY integrates fertile meristem ontogenetic change from vegetative to reproductive), the bisexual cone axis in "... the common ancestor of angiosperms and gymnosperms would have gradually accumulated higher and higher levels of LFY protein during development."
When considering the evo-devo of SAM developmental programs in deep-time, conceptual models that incorporate movable homeodomain proteins and cis-acting promoters and enhancers, have potentially profound implications toward evolution of leaves, sporophylls, and lycopsid, progymnosperm, and seed plant reproductive axes.
Seed plant morphologies such as the arrangement of sporophylls on the mother plant and position of ovule attachment (terminal, abaxial, adaxial, or marginal) to stems and/or leaves potentially reflect diverse mechanisms of homeodomain protein trafficking in SAMS (at the organismal level), and evolution of modular units and cis-acting transcription factors (Class III HD-Zip, KNOX-ARP, YABBY) in paleopopulations.
Many of the supposed homologies of leaves, sporophylls, and ectopic structures including reproductive axes and ovular integuments could be incorrect based on the possibility that a fundamental dichotomy in SAM evo-devo emerged in Devonian vascular plant stock. Interestingly, lateral branches of the Devonian progymnosperm, Archaeopteris (Archaeopteridales) may arise from tissue fields resembling leaf primordia, and not axillary buds (page 481, T. N. Taylor et al. 2009).
Hickey and D. W. Taylor (1994) regard fertile axes of some seed plants as fertile trusses derived from telombs. There is no evo-devo evidence from the study of SAM maintenance and patterning genes and transcription factors supporting of Zimmermann's Telomb Theory. More research is needed.
Character analysis and homology assessment. A review by Rothwell et al. (2009) states that no known seed plant fossil transformational series supports past proposals on derivation of the angiosperm carpel and ovule integuments from Caytonia-like cupules, or the MMT. Professor Rothwell and company's quite correct insight from the rock record is incongruous with the idea floated by Specht and Bartlett (page 220, 2009) that extant gymnosperms "are sufficiently divergent from angiosperms to prevent reliable reconstruction of flower origins."
Inclusion of pteridosperms such as Caytonia or corystosperms (and their glossopterid ancestors) in a morphologically-based phylogenetic analysis of angiosperms might be fundamentally flawed on evo-devo grounds (bulleted list, above).
This assertion is explored from an experimental cladistic perspective to simplify data analysis and possibly ameliorate well-known vexing problems illuminated by J. A. Doyle and Endress (2009) in past, often conflicting seed plant phylogenies.
Homologies of the carpel and stamen. Reviews by D. W. Taylor and Kirchner (1996), Friedman et al. (2004), J. A. Doyle (2006), Scutt et al. (2006), Endress and J. A. Doyle (2009), Specht and Bartlett (2009), and Vialette-Guiraud and Scutt (2009) outline the great body of literature which has been devoted to evolutionary and experimental cladistic studies of carpel homologies.
A solitary carpel of the "primitive" magnoliid tree, Degeneria vitiensis (Degeneriaceae, Magnoliales), is depicted in the left-hand scanning electron micrograph, ×15. The right-hand image is a scanning electron micrograph of a Degeneria microsporophyll, ×5.
The Degeneria reproductive organs were field-fixed in 1986 and 1987 from dissected flowers collected in the canopies of tagged and vouchered trees growing in stands on the slopes of Mount Naitaradamu on Viti Levu Island, and from forested flanks of Mount Delaikoro, Vanua Levu Island, Fiji.
The scanning electron micrographs were prepared by Al Soeldner, Director, Oregon State University Electron Microscopy Laboratory. I thank the National Geographic Society for providing research funding for this study.
David Winship Taylor and Kirchner (page 119, 1996) provide a detailed discussion of the "phyllosporous origin" and "stachyosporous origin" hypotheses of carpel homologies.
The homology assessment of the angiosperm carpel, perianth, and flower by D. W. Taylor and Kirchner (Figure 6.3 page 124, 1996) would be confounded by the inclusion of a hypothetical Paleozoic gigantopteroid proanthostrobilus.
Phyllosporous origin hypotheses. The phyllosporous origin hypotheses is also coined the "megasporophyll-homology hypothesis" (page 119, D. W. Taylor and Kirchner 1996). Serge Mamay (Figure 7, page 33, 1976) was a 20th Century paleobotanist who proposed a conduplicately-folded megasporophyll hypothesis for a Paleozoic origin of Cycadales, but this fact apparently went unnoticed by D. W. Taylor and Kirchner twenty years later (1996).
David Winship Taylor and Kirchner's megasporophyll-homology hypothesis is based on morphological studies of primitive magnoliids by I. W. Bailey and Swamy (1951). The carpel of Degeneria vitiensis (Degeneriaceae, Magnoliidae) shown in the image on the left side of this page, was thought to be a conduplicately folded (inrolled) megasporophyll (I. W. Bailey and A. C. Smith 1942).
Evo-devo research on carpel, floral, and ovular transcriptional regulators in extant angiosperm model organisms (Scutt et al. 2006) do not preclude derivation of models of cone- and floral organization that explain curling, inrolling, and fusion in 300 million-year-old ovule-bearing gigantopteroid Phasmatocycas bridwellii megasporophylls to form carpels, ovaries, and pistils.
Two elegant models have been proposed to explain the origin of the flower from an ancient bisexual strobilus (Baum and Hileman 2006, Theißen and Melzer 2007), which cast doubt on earlier hypotheses from the standpoint of evo-devo of fertile SAMs.
Stachyosporous origin hypotheses. David Winship Taylor and Kirchner (page 128, 1996) assert that, "... open carpels are a teratology and are not homologous with ancestral conduplicate carpels." I disagree with this proposal as it conflicts with earlier ideas expressed by Mamay (1976) and the reanalysis of Phasmatocycas bridwellii megasporophylls by Axsmith et al. (2003).
A stachyosporous origin hypothesis better explains transformational series from ancestral glossopteridalean fertiligers to Mesozoic ovule-bearing cupules of Caytoniales, Corystospermales, Peltaspermales, and Petriellales, and not angiosperm carpels as argued by D. W. Taylor and Kirchner (1996).
Seed fern hypotheses. James Doyle's seed fern hypotheses (1978, 2006) that develop an earlier proposal by Hamshaw Thomas (1925) on evolution of the angiosperm carpel and ovule integuments from Caytonia cupules, and Sergei Meyen's gamoheterotophic hypothesis (Meyen 1986, 1988) are difficult to explain by modern evo-devo models of floral origin involving cis-acting homeotic transcription factors.
Carefully researched evo-devo proposals on the origin of the angiosperm outer integument from pteridosperm cupules involving transcriptional regulation of KANADI and YABBY, outlined by J. A. Doyle (2006) on pages 175 and 176 of his review, are not yet supported by experimental evidence gleaned from studies of basal angiosperms (e.g. Amborella or Nymphaea) or primitive magnoliids such as Degeneria, Drimys, and Magnolia.
Ovular and integumentary transcription factors such as KANADI and YABBY act downstream of cis-regulatory modules underpinned by shoot apical meristem (SAM) maintenance- and floral meristem organ identity and integrator genes. Further, it may be impossible to extrapolate to extinct seed plants, the gene duplication, isoform genesis, subfunctionalization, and/or neofunctionalization events that might have occurred in deep time within and between these gene families.
Frohlich and Parker's Mostly Male Theory (MMT) on ectopic ovules derived from male organs (2000) and later critiques of MMT based on paleontology of corystosperms and molecular evolution of cis-acting FLO/LFY homeotic selector genes (Frohlich 2002, 2003) are challenged by gene expression data of FLO/LFY homologs in female conifer cones and fertile SAMs (Vásquez-Lobo et al. 2007).
Developmental evolutionary studies of pteridosperms including Corystospermales are impossible to conduct since this line of evolution died-out during the Paleogene Period of the Tertiary Interval (McLoughlin et al. 2008).
Ophioglossoid origin hypothesis. Insight on alternative explanations for evolution of the carpel may be found in papers published by Kato (1990, 1991) on the anatomy and morphology of eusporangiate ferns.
Kato developed a proposal on a neotenous origin of the carpel from Paleozoic glossopterid fertiligers based upon morphological studies of extant Ophioglossales (Kato 1988), an early divergent order of monilophytes. Figure 1 on page 190 of Kato (1991) outlines evolutionary steps in the "ophioglossoid model for angiosperm carpel archetype."
Morphological-based proposals by Kato (1988, 1990, 1991) and Kato et al. (1988) namely, that Ophioglossales might represent living offshoots of early divergent homosporous, eusporangiate Devonian progymnosperms, are not supported by a recent, detailed anatomical study of the Botrychium shoot apical meristem (Rothwell and Karrfalt 2008).
When in geologic time do molecular systematists and pteridologists pin-point the monilophyte and seed plant split?
Pryer and coworkers (page 1586, 2004) estimate that "... initial divergence among monilophyte lineages ..." from seed plants occurred more than 360 m.y.a. during late Devonian time. Whisk fern and ophioglossoids, "... diverged from one another in the Late Carboniferous ...", more than 300 m.y.a.
Ferns and fern allies (monilophytes) potentially carry out development of fertile structures differently than extant seed plants based on biochemical and molecular phylogenetic studies of MIKC-type MADS-Box and FLO/LFY homeotic selector genes (Becker et al. 2000, Theißen et al. 2000, Hasebe et al. 2001, Himi et al. 2001).
Pryer et al. (2004) state:
"One of our remaining challenges will be to identify the Carboniferous ancestors of the whisk fern + ophioglossoid lineage."
The preceding passage is quoted from page 1594 of Pryer et al. (2004), Phylogeny and evolution of ferns (monilophytes) with a focus on the early leptosporangiate divergences. American Journal of Botany 91(10): 1582-1598.
Additional evo-devo research on ophioglossoid model monilophytes at the level of homeotic gene expression and homeodomain protein trafficking in meristematic tissues is needed. The leptosporangiate fern model organism Ceratopteris richardsonii is probably unsuitable for this purpose.
Fertiliger (glossopterid) origin hypotheses. Evolutionary-development (evo-devo) of pteridosperm cupules and fertiligers (Schopf 1976, Retallack and Dilcher 1981, Melville 1983) is open to alternative explanations (Pigg and Trivett 1994).
Ronald Melville's review on a Permian origin of angiosperms from glossopterids (1983) is a foundation for the homology assessment of fertiligers and carpels and a discussion of the gonophyll hypothesis. I will begin here at a later date.
Stamen homologies. An evaluation of the origin, evolution, and homologies of the angiosperm stamen and associated structures is published by Crepet and Nixon (1996), Endress (1996), Hufford (1996), and Ronse De Craene and Smets (2001). Additional resources include Endress and Hufford (1989) and Hickey and D. W. Taylor (1996).
There are two lines of reasoning on the homology of the stamen of flowering plants. Certain workers view the angiosperm stamen as fused tips of microsporangium-bearing branches of a telomb (see Stebbins 1974).
The prevailing view of Canright (1952) is that pollen bearing organs of Degeneriaceae are microsporophylls (see Takhtajan 1969). "The laminar, spirally arranged androecium of Degeneriaceae is probably the most plesiomorphic trait of any extant flowering plant" (page 223, J. M. Miller 1989).
A microsporophyll of Degeneria (Degeneriaceae, Magnoliidae) is imaged on the right side of the page, above. In my opinion, the microsporophyll of Degeneria is homologous with the laminar microsporophyll imaged in the earlier section on Phasmatocycas, and is not an anther as suggested by Endress and Hufford (1989) and Hufford (1996).
Elongate rice-shaped structures on the adaxial leaf surface of the fossil illustrated in the chapter on gigantopterids (see Phasmatocycas section) may be microsporangia. I found this Permian (Leonardian) fossil in the same bedding plane as leaves of Delnortea abbottiae in 1981.
The anatomy of the Phasmatocycas permineralization has not been studied. Its taxonomic affinities are unknown pending recovery of more material from the delnortea beds needed to demonstrate vascular connections with the mother plant.
Pollen homologies. Ultrastructural details of pollen permineralizations are of importance in phylogenetic analyses of seed plants and pin-pointing the time and place of flowering plant origins. Considerable importance is placed on the anatomy and homology of pollen ultrastructural characters (J. A. Doyle 2009, Tekleva and Krassilov 2009).
Angiospermous infratectal pollen structures (columellae and alveolae) and character variations of the nexine and tectum resembling both modern and fossil conifers and flowering plants, are known from permineralizations recovered from Permian rocks (Zavialova and Gomankov 2009). A discussion of these characters and their homologies is reserved for a later date.
Homologies of the angiosperm ovule. Ovules of flowering plants are composed of the embryo sac (megagametophyte), a nucellus (old megasporangium wall), and sterile integuments. Integuments are generally regarded as modified leaves. Many but not all angiosperms possess two integuments. Understanding the evolutionary-development of ovular integuments (Skinner et al. 2004) is a prerequisite to understanding certain character homologies (Friedman et al. 2004).
Considerable attention has been paid to the evolution of tissue layers (integuments) of angiosperm ovules, which is intertwined with the evo-devo of fertile leaves and shoots. Reviews by Friedman et al. (2004), J. A. Doyle (2006), and Kelley et al. (2009) are important starting points for this discussion.
Preintegumentary lobes and cupulate systems enclosed the seeds of primitive Devonian seed plants (Serbet and Rothwell 1992), and these structures were enveloped by dichotomously-lobed non-foliar cupulate systems (also termed "cupules") that probably functioned in protection of ovules and seeds (T. N. Taylor et al. 2009).
Thomas N. Taylor et al. (2009) do not regard Devonian seed plant cupulate systems as homologous with either megasporophylls of Paleozoic glossopterids or cupules of Mesozoic corystosperms and Caytoniales. I concur.
Frohlich (2003) offers additional discussion on the origin of ovule integumentation from Mesozoic pteridosperms, and evo-devo of leaves and ovules from the standpoint of YABBY signaling from shoot apical meristems (SAMs). But YABBY and KNOTTED1 homeodomain protein movement in the cells of SAMs might have been constrained by other factors such as distance of the fertile organ cell fields from the nuclear genes being activated.
James Doyle discusses YABBY within the context of ovule tissue layers and attachment point (abaxial or adaxial) to carpels and leaves in Amborella, a basalmost angiosperm (J. A. Doyle 2006).
Kelley et al. (2009) conclude that there are similarities between expression and transcriptional regulation of Class III HD-Zip, KANADI, and YABBY genes of integuments and leaves Arabidopsis (Brassicaceae, Brassicales, Rosanae). Based on biochemical studies of ovule polarity determinants, these workers state:
"Although these organs [integuments and leaves] have distinct evolutionary origins, a common set of polarity determinants appears to have been serially utilized in both sets of structures. The differences in the precise roles and interactions of the determinants of each structure may represent differences dating from their origin, or alternatively, differences arising from subsequent structural diversifications."
This quote is from page 1062 of D. K. Kelley, D. J. Skinner, and C. S. Gasser (2009), Roles of polarity determinants in ovule development. The Plant Journal 57(6): 1054-1064. The phrase in brackets [] is mine.
Evolutionary development of the angiosperm outer integument and its supposed homology with reproductive structures of corystosperms suggested by Frohlich (2003) and Caytoniales proposed by J. A. Doyle (2006), may be explained differently (Kelley et al. 2009).
For example, tissue layering of ovular integuments of flowering plants may be a consequence of splitting of cell fields, which are derived from the inner integument and not a subtending foliar organ (Skinner and Gasser 2009).
In conclusion, 300 million year old (Paleozoic) WGDs in the phylogeny of Class III HD-Zip genes (Floyd et al. 2006), and possibly YABBY, preclude derivation of the angiosperm carpel from any Mesozoic pteridosperm foliar organ.
Seed plant/progymnosperm phylogenetic analyses. I conducted two separate phylogenetic analyses based on the biological-process homology concept, Process PTDSP and Process ANTHO. Two distinct biological processes of SAM homeotic transcription factor and auxin/PIN protein trafficking are hypothesized in the last common ancestor of seed plants, prior to supposed WGDs that probably occurred during the late Devonian Period more than 350 million years ago.
I executed David L. Swofford's Phylogenetic Analysis Using Parsimony (PAUP) program, version 4.0b10 on a Windows machine, and a computed two separate phylogenies.
Anatomical and morphological data are from Bold et al. (1980), Cronquist (1981), Crane (1985), Trivett and Rothwell (1985), Serbet and Rothwell (1992), Crane (1996), Rothwell et al. (1996), Naugolnykh (2001), Axsmith et al. (2003), Friedman et al. (2004), Mundry and Stützel (2004), Z.-Q. Wang (2004), X.-Q. Liu et al. (2006), E. L. Taylor et al. (2006), J. A. Doyle (2009), Rothwell et al. (2009), T. N. Taylor et al. (2009), and Tekleva and Krassilov (2009).
Source references for specific characters are cited and discussed in each of the two analyses, below. Biomarker data are from D. W. Taylor et al. (2006).
Commands used in PAUP input blocks were simple and employed the most basic assumptions. Heuristic search methods were used to process the NEXUS files with maximum optimality parsimony settings. Characters were all the "unord" type with the "MultTrees" option selected. Bootstrap statistical methods were not used, which is in line with a critique of certain phylogenetic methods by Graur and Martin (2004).
Further, the number of possibly informative characters adopted in my preliminary cladistic exercises (fewer than 30) departs radically from published work of others (Nixon et al. 1994, Rothwell and Serbet 1994, J. A. Doyle 2006, Hilton and Bateman 2006, J. A. Doyle 2008, Rothwell et al. 2009) who analyzed more than twice the number of characters. It is just that there are too many gaps in the data on possibly critical seed plant groups (and too many question marks [?] in the matrixes) when a full complement of characters is used, including hypothetical process-level (evo-devo) characters.
The first phylogenetic analysis of core pteridosperms, Process PTDSP, is rooted with Aneurophyton (Aneurophytales) as the outgroup, and also includes Archaeopteris (Archaeopteridales), which are homosporous Devonian progymnosperms regarded by some paleobotanists as the Paleozoic ancestors of ferns and seed plants (T. N. Taylor et al. 2009).
Spermatophytes scored in the Process PTDSP study are Autunia (Peltaspermales), Callistophyton (Callistophytales), Cycadales, Lyginopteris (Lyginopteridales), Elkinsia (Lyginopteridales), medullosans (lumped together to follow J. A. Doyle 2006), Peltaspermum (Peltaspermales), Petriella (Petriellales), and the cupule- and fertiliger-generating pteridosperms classified in the orders Caytoniales, Corystospermales, and Glossopteridales.
A second analysis, Process ANTHO (including conifers) is rooted with Cycadales as the outgroup, and incorporates only those seed plant groups possessing simple and/or compound strobili (or in the case of female Ginkgo, which has no strobilus): angiosperms, Bennettitales, Coniferales, Czekanowskiales, gigantopteroids with protoflowers, Ginkgoales, Gnetales, Pentoxylales, Vojnovskyales, and conifers, among other key taxa (see character analysis Process ANTHO and Table 5).
For now, I leave out some seed plant groups from the phylogenetic analyses (e.g. Erdtmanithecales [including the taxa of charcoalified seeds, Friis et al. 2005 and 2006, Rothwell et al. 2009], Hermanophytales, Iraniales) but reserve the right to incorporate them into future revisions.
At a later date, I may also introduce a "temporal dimension" to the analyses (page 350, Cracraft 2005) by optimizing character transformations on trees, or by introducing minimum geologic age to nodes, or even by re-thinking the entire cladistic exercise.
Students and researchers alike are challenged to apply Cracraft's concept of historical-phylogenetic homology to evo-devo data (page 349, 2005) with added precision given to use of the concepts of "homocracy, homotopy, neofunctionalization, orthology, paralogy, and subfunctionalization" (page 211, Theißen 2005).
Character analyses are generally patterned along the lines of published work by Nixon et al. (1994), Rothwell and Serbet (1994), J. A. Doyle (1996), Hickey and D. W. Taylor (1996), J. A. Doyle (2006), Hilton and Bateman (2006), and J. A. Doyle (2008) while incorporating Devonian pteridosperms in the analyses, but differ in one critical respect: two distinct groups of Paleozoic seed plants (cupule- and fertiliger-generating pteridosperms and strobilus-generating anthophytes + gigantopteroids) are analyzed separately.
I will discuss the latest seed plant phylogenetic analysis by Rothwell et al. (2009) at a later date.
Further, not all characters used in the Process PTDSP analysis were appropriate to incorporate into the Process ANTHO study and vice versa. I employed a basic set of 23 to 27 characters for each of the two phylogenetic analyses for two reasons: one, absence of data (principally glaring gaps in data on reproductive characters) on poorly studied extinct taxa precluded use of potentially informative traits, and two, simplicity of the analysis is concordant with my lingering doubts on some character homologies in deep time.
Some of the ideas expressed herein might be justifiable on evo-devo grounds based on insight gained from molecular phylogenetic and biochemical studies of extant model vascular plant organisms and regulatory models of homeodomain proteins and cis-acting homeotic transcription factors (discussed in the previous section). Focused cladistic experiments may be included in future revisions of the analyses. More experimental data from evo-devo studies of a wider range of model extant lignophytes- and additional paleontological data are needed to shed light on anatomical and morphological homologies.
Based on the discussion in the preceding section, I conclude that two different evolutionary-developmental programs: Process PTDSP for core pteridosperms without anthophytes + gigantopteroids, and another termed Process ANTHO for anthophytes + gigantopteroids, might have existed in Paleozoic vascular plants.
Logically, vegetative and reproductive organs, e.g. leaves and sporophylls, of Process PTDSP pteridosperms and Process ANTHO seed plants might share no homology because these organs might be products of two different evo-devo programs downstream of SAM patterning genes and transcription factors.
It is possible and even likely that an evo-devo dichotomy triggered by WGDs developed in races of Devonian seed plants leading to at least two modes of SAM trafficking of homeodomain proteins, and divergence in the anatomy and morphology of reproductive organs.
More research is needed to pin-down the exact timing of these WGDs to calibrate seed plant divergences such as the supposed Process PTDSP and Process ANTHO split, and to support my novel proposal.
If the logic of this unorthodox approach is proven correct, pteridosperms might not constitute "the backbone of seed-plant phylogeny" (title, Hilton and Bateman 2006), but may instead constitute a loose, paraphyletic group derived from Devonian ancestral seed plant stock.
Core pteridosperms, progymnosperms, and cycads without anthophytes, conifers, and gigantopteroids: character analysis Process PTDSP. A numerical list of key seed plant characters with additional discussion appears below.
1A. General plant habit. Woody shrubs, vines, or trees possessing secondary growth = 0; herbs usually with little or no secondary growth = 1.
I scored this character the same way as J. A. Doyle (Appendix, Taxa and Characters, page S26, 1996) character 1; J. A. Doyle (Appendix 1, page 202, 2006) character 1; Hilton and Bateman (Appendix 2, page 160, 2006) character 2; and Nixon et al. (Appendix A, page 524, 1994) character 0. Arguments posed by Hickey and D. W. Taylor (1996) are inconsistent with evo-devo models of SAM organization in ferns.
2A. General plant architecture. Long shoots present, short (spur) shoots absent = 0; both long and short (spur) shoots present = 1.
Inclusion and polarity assignment of this character is in line with J. A. Doyle (Appendix, Taxa and Characters, page S26, 1996) character 4; Nixon et al. (Appendix A, page 525, 1994) character 2.
3A. Stelar anatomy. Protostelic = 0; eustelic = 1; polystelic = 2.
Nixon et al. adopted this character (Appendix A, page 525, 1994) character 4.
4A. Apical meristem layering. Tunica absent = 0; tunica present, consisting of a single layer = 1.
5A. Secondary tissues: wood. Manoxylic = 0; pycnoxylic = 1.
Nixon et al. incorporated the wood character in their analysis (Appendix A, page 526, 1994) character 9.
6A. Secondary tissues: rays. Uniseriate or biseriate = 0; multiseriate = 1.
This character was included by Hickey and D. W. Taylor [Table 8.3, page 219, 1996] character 15; Nixon et al. (Appendix A, page 526, 1994) character 11.
Polarity of the ray parenchyma character is reversed from J. A. Doyle (Appendix, Taxa and Characters, page S28, 1996) character 24.
7A. Leaf traces (nodal anatomy). One trace per leaf, unilacunar = 0; one trace per bundle = 1; trilacunar = 2.
Considerable discussion and several variations on the scoring of this complex trait have been published (J. A. Doyle [Appendix, Taxa and Characters, page S28, 1996] character 17; J. A. Doyle [Appendix 1, page 203, 2006) character 23; J. A. Doyle [Appendix, 2008] character 28; Hilton and Bateman [Appendix 2, page 161, 2006] character 20; Nixon et al. [Appendix A, page 525, 1994] character 5; Rothwell and Serbet [Appendix 1, page 473, 1994] character 17).
8A. Leaf venation (vegetative leaves). Open and dichotomous = 0; reticulate and anastomose = 1.
Scoring of the leaf venation character agrees with J. A. Doyle (Appendix, Taxa and Characters, page S27, 1996) character 8; J. A. Doyle (Appendix 1, page 203, 2006) character 31; J. A. Doyle (Appendix, 2008) character 36; Rothwell and Serbet (Appendix 1, page 473, 1994) character 9, and is consistent with my hypothesis that a fundamental dichotomy in the process of homeodomain transcription factor movement may exist in the evolutionary-development (evo-devo) of seed plant SAMs.
Adoption of multiple character states for vegetative leaf venation by Hickey and D. W. Taylor (Table 8.3, page 219, 1996) character 7, was considered for the Process PTDSP analysis but rejected.
9A. Leaves (vegetative). Pinnately compound = 0; simple and pinnate or pinnately-lobed = 1; simple and palmately-lobed or compound palmate = 2.
10A. Stomata. Haplocheilic = 0; dicyclic stomatal complexes = 1.
A new advanced stomatal character state is adopted for the Process PTDSP analysis based information from page 627 and page 629 of T. N. Taylor et al. (Figure 15.24, 2009), and knowledge that syndetocheilic stomata are not known from the Process PTDSP study taxa.
Further, leaf epidermal traits in taxa within the Process PTDSP line of pteridosperm evolution such as dicyclic stomatal complexes might share no homology with syndetocheilic stomata of angiophytes. This line of deductive reasoning follows my hypothesis that a fundamental dichotomy in transcription factor traffic patterns might exist in SAM development of Devonian seed plants possibly extending to downstream leaf epidermal patterning.
11A. Reproductive life cycle attributes Homosporous = 0; heterosporous = 1.
Choice and polarity assignment of aspects of the lignophyte reproductive life cycle is in line with Rothwell and Serbet (Appendix 1, page 473, 1994) character 1.
12A. Ovulate axes (or equivalent eusporangiate axes). Eusporangia clustered on terminal shoots, or ovules aggregated and sheathed by cupulate axes on leafless shoots = 0; ovules aggregated or solitary on ectopic shoots arising from laminar leaf midribs, or sessile or nearly so on laminar megasporophylls aggregated into simple cones or not = 1; ovules sessile or nearly so and enclosed by a cupular leaf (cupule) = 2; ovules sessile or nearly so and attached to a peltate leaf = 3.
13A. Microsporangiate axes (or equivalent eusporangiate axes). Eusporangial clusters, or microsporangia terminal on leafless shoots = 0; microsporangia clustered or fused on ectopic shoots attached to midribs of laminar microsporophylls aggregated into simple cones or not = 1; microsporangia clustered in synangia, sessile on pinnate leaves = 2.
14A. Microsporangia (or equivalent eusporangial clusters) attachment to axes or leaves. Terminal = 0; abaxial = 1; adaxial = 2.
James A. Doyle (Appendix 1, page 204, 2006) adopted this multistate character but added a fourth state (no. 49, lateral = 3), which I did not adopt.
15A. Pollen (or equivalent pre-pollen) shape. Spores or pre-pollen radial = 0; pollen bilateral = 1; pollen global = 2.
This character was employed by J. A. Doyle (Appendix 1, page 205, 2006) character 62; J. A. Doyle (Appendix, 2008) character 69; and Hilton and Bateman (Appendix 2, page 163, 2006) character 80, in their cladistic analyses of progymnosperms and spermatophytes.
16A. Megasporophylls (or equivalent): ovule number per cupule, laminar megasporophyll, or peltate leaf. One (rarely two) ovule = 0; three to many ovules (rarely two) = 1.
I scored ovule number along the lines of Nixon et al. (Appendix A, page 531, 1994) character 68.
Petriella (Petriellales) was difficult to score for this character. Most Petriella fossils contain three to six ovules per cupule, so a "1" was scored (pages 637-639, T. N. Taylor et al. 2009).
Uniovulate and strobilate Kannaskoppiaceae (pages 638-639, T. N. Taylor et al. 2009) are excluded from the analysis but may be included at a later date to better evaluate their character homologies and phylogenetic position.
Restudy of the Caytonia cupule by scanning electron microscopy (Nixon et al. 1994) reveals that this structure may be a megasporophyll homologue (Hilton and Bateman 2006).
This character is difficult to score for early Devonian seed plants since a lobed (dichotomous) non-foliar cupular system occurs in place of an ovule-bearing leaf (megasporophyll). Paleozoic progymnosperm seeds (e.g. Elkinsia), which are enclosed by a cupular system, were scored as "0."
Dichotomously-lobed, non-foliar cupulate systems have been termed "cupules" in the literature on Devonian seeds (page 511, T. N. Taylor et al. 2009), but these authors do not view cupulate systems as homologous with either megasporophylls of glossopterids or cupules of corystosperms and Caytoniales.
Callistophyton (Callistophytales) is one of the best known Paleozoic seed plants but details of ovule number per megasporophyll are problematic.
17A. Megasporophylls (or equivalent): position of ovule attachment. Terminal = 0; marginal or adaxial = 1; abaxial = 2.
Traditional ideas and supposed homologies of characters such as position of ovules on the megasporophyll, whether adaxial or abaxial (Figures 13 and 14, J. A. Doyle 2006), are called into question. I regard abaxial ovule attachment to the megasporophyll as the most derived character state in the Process PTDSP analysis, which is concordant with my suggestion that a wholly different evo-devo program of SAM transcriptional regulation and homeodomain protein trafficking occurred in this evolutionary line.
Position of ovule attachment on the Callistophyton (Callistophytales) megasporophyll is unclear but detached leaves of the morphotype genus Pseudomariopteris bearing Callospermarion-like ovules are known (page 595, T. N. Taylor et al. 2009). Based on this circumstantial evidence I scored a "?" for character 17A.
Several Paleozoic spermopteroid megasporophylls (see ANTHO analysis in next section) are known including Eophyllogonium cathayense (Mei et al. 1992), Phasmatocycas bridwellii (Axsmith et al. 2003), and Sobernheimia jonkeri (Kerp 1983; pages 709-711, T. N. Taylor et al. 2009). All of these possess stalked ovules attached to the abaxial surface of the megasporophylls, which is in contrast to marginal (to adaxial) ovule position on megasporophylls of true cycads (e.g. Crossozamia and Beania).
Caytoniales and Dicroidium odontopteroides (Corystospermales) both possessed megasporophylls with ovules attached to the lower leaf surfaces, which is in agreement with Retallack and Dilcher (1988) and T. N. Taylor et al. (2009).
This is a major departure from J. A. Doyle (page 194, 2006).
18A. Ovules (or equivalent eusporangium): position of attachment to axes or leaves. Erect = 0; inverted = 1.
Choice and polarity assignment of ovule orientation is in line with J. A. Doyle (Appendix, Taxa and Characters, page S30, 1996) character 32; and Rothwell and Serbet (Appendix 1, page 474, 1994) character 29.
Details of ovule attachment to the megasporophyll of Callistophyton (Callistophytales) are unclear. Based on this uncertainty I scored a "?" for character 18A.
19A. Ovule: integuments (or equivalent). Ovule (preovule) enclosed by preintegumentary lobes, or lobed integuments partially fused with the megasporangium = 0; ovule enclosed by one or more integuments = 1.
Preintegumentary lobes and cupulate systems enclosed the seeds of primitive Devonian seed plants (Serbet and Rothwell 1992), and these structures were enveloped by dichotomously-lobed non-foliar cupulate systems (also termed "cupules") that probably functioned in protection of ovules and seeds (T. N. Taylor et al. 2009).
20A. Megasporangium (nucellus, or equivalent). With a lagenostome = 0; with a pollen chamber = 1.
Choice and polarity assignment of this character is in line with J. A. Doyle (Appendix, Taxa and Characters, page S32, 1996) character 56; and Nixon et al. (Appendix A, page 532, 1994) character 75.
21A. Megagametophytes. Archegonia present = 0; archegonia absent = 1.
Many previous seed plant phylogenetic analyses (Hickey and D. W. Taylor [Table 8.3, page 221, 1996] character 49; Nixon et al. [Appendix A, page 533, 1994] character 86; Rothwell and Serbet [Appendix 1, page 476, 1994] character 60) include this organ (sometimes with multistate scoring e.g. J. A. Doyle [Appendix, Taxa and Characters, page S35, 1996] character 81), with the notable exception of Hilton and Bateman (2006).
22A. Sperm transfer. Zooidogamous = 0; siphonogamous = 1.
There is no paleobotanical evidence of sperm transfer in extinct Cycadales. Scoring of this character is based on living plants (Bold 1980). Unsubstantiated data for Hydrospermales (Elkinsia) and Lyginopteridales are included in the character matrix to follow rationale published by J. A. Doyle (2006), which is in line with a discussion of hydrospermalean reproduction by T. N. Taylor et al. (Chapter 13, 2009).
Inclusion and polarity assignment of this character is in line with J. A. Doyle (Appendix, Taxa and Characters, page S35, 1996) character 79; J. A. Doyle (Appendix 1, page 205, 2006) character 73; J. A. Doyle (Appendix, 2008) character 82; and Hilton and Bateman (Appendix 2, page 163, 2006) character 90.
Character data matrix Process PTDSP. A zero [0] in a cell denotes an ancestral (plesiomorphic) character state. Apomorphic (derived) character states are denoted in the table by the numeral one [1]. Multistate characters used in the analysis are: zero [0] (most plesiomorphic), numeral one [1] (neither plesiomorphic or apomorphic), two [2] (apomorphic), and three [3] (most apomorphic). The question mark [?] is equivalent to unassigned or missing data.
Several characters scored in the Process ANTHO study (see next section), e.g. companion cells, vessel elements, multilacunar nodal anatomy, modular female gametophytes, and double fertilization, were omitted from the Process PTDSP analysis as they do not occur in any known extinct progymnosperms and pteridosperms.
Further, at least one character used in previous attempts to resolve seed plant phylogeny using morphological data, secretory structs, was left-off both analyses. I could find no logical homology argument to justify inclusion of the character despite its inclusion in several seed plant phylogenetic analyses (J. A. Doyle [Appendix, Taxa and Characters, page S29, 1996] character 27; J. A. Doyle [Appendix 1, page 203, 2006] character 20; J. A. Doyle [Appendix, 2008] character 24; Hilton and Bateman [Appendix 2, page 161, 2006] character 32; Nixon et al. [Appendix A, page 526, 1994] character 17; and Rothwell and Serbet [Appendix 1, page 473, 1994] character 22).
Table 4 is a matrix of characters used in the PAUP input block for a preliminary phylogenetic reconstruction of progymnosperms, cycads, cupule- and fertiliger generating seed plants, and peltasperms (see below). The Devonian progymnosperm, Aneurophyton (Aneurophytales) is selected as the outgroup (ANE).
Additional lignophytes were also scored: Antarcticycas schopfii, including Delemaya spinulosa (Cycadales) = ANT, Archaeopteris (Archaeopteridales) = ARC, Autunia (Peltaspermales) = AUT, Lyginopteris (Lyginopteridales) = LYG, Callistophyton (Callistophytales) = CAL, Crossozamia (Cycadales) = CRO, Elkinsia (Hydrospermales), medullosans = MED, Peltaspermum (Peltaspermales) = PEL, and Petriella (Petriellales) = PET.
Key extinct groups in the PAUP "TAXA Block" that possess cupules or, alternatively fertiligers (Table 4) are: Caytonia (Caytoniales) = CAY, Dicroidium (Corystospermales) = DIC including detached branches of cupules described as Umkomasia, and Glossopteris (Glossopteridales) = GLO.
|
Paleobotany of Angiosperm Origins
Is there convincing fossil evidence of the first flowering plants or their antecedents? No, according to E. L. Taylor and T. N. Taylor (2009).
By knowing the interval in geologic deep time when molecular systematists predict divergence of reproductive patterning genes and regulatory proteins, paleobotanists might be able to better focus on candidate seed plant groups for detailed anatomical study of permineralized fossils (Crepet 2008).
With the exception of Armen Takhtajan's neglected hypothesis on a "neotenous origin of flowering plants" (page 209, 1976), none of these ideas when taken as a whole are neither compelling or plausible to many scientists, including the author.
