Evolution of Plants with Emphasis on Its Reproduction Form

Abstract

Chapter 1 of this book discusses the evolution of plants from the Pre-Cambrian era considering the starting time of alternation of generations. In the Paleozoic and Mesozoic eras, we will discuss the transitions from homospory to heterospory and the formation of the egg cell and the pollen grain as well as the evolution of seed plants. At the end of this chapter we will discuss the evolution of the angiosperms (flowering plants) and the sexuality transitions and the reproductive strategies. Also, the evolution of the unisexuality, self-incompatibility systems, selfing, and mixed mating systems will be discussed considering their evolutionary implications.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1.1 First Plants and the Alternation of Generations

The fossils of the first organisms that performed photosynthesis and produced oxygen were found in Warrawoona, Western Australia. These spherical and filamentous microorganisms with 3300 and 3500 million years (myr) were classified as photoautotrophic cyanobacteria (blue-green algae) (Schopf and Packer 1987). Further evidence of the evolution of plants from photosynthetic unicellular organisms that divide by mitosis was found in deposits in South Africa. These organisms, belonging to the Pre-Cambrian period (Table 1.1), were dated between 3200 and 3100 myr (Freeman and Herron 1998; Brown and Lomolino 2005). In more recent sediments of rocks of Southern Ontario, Canada, with approximately 2000 myr, specimens of blue-green algae were also found. These early ancestors of plants showed prokaryotic life because they did not have organized nuclei as in higher organisms. Thus, the oldest and the most reliable trace available from the emergence of eukaryotes has an age between 2000 myr (Mussa 2004) and 1400 myr (Zunino and Zullini 2003), when the first members of unicellular organisms called acritarchs, belonging to the Protist kingdom, appeared in the fossils record in China. It is widely accepted that the ancestors of land plants are in the Charales order of the Coleochaete genus (Graham et al. 2000). Protists were commonly found in aquatic environments, but some were able to live in terrestrial environments. Generally they had sexual reproduction, but some could reproduce asexually (Raven et al. 2007). Analyses of rock sediments with about 1000 myr revealed a wide variety of types of algae, the vast majority bluish-green. Apparently they had a true nucleus and mitotic divisions (Fig. 1.1), being among the first eucaryotic organisms that we know of so far (Schopf 1968; Banks 1970; Knoll 1992).

Table 1.1 Geological time scale with major evolutionary changes

Fig. 1.1

Schematic representation of the division by mitosis in unicellular algae (Karasawa et al. 2006)

In response to evolutionary changes, living organisms became increasingly diverse and complex in structure (Raven et al. 2007). These organisms have evolved for about another 500 myr until the first plants appeared with multicellular and erect growth. Geological evidence suggests the green algae Fritschiella as a probable ancestor of land plants, because metabolism features are similar to that of existing plants, which was not found in any other algae (Fig. 1.2). Fritschiella lived in fresh water, but could be found in any other terrestrial environment (Willis and McElwain 2002).

Fig. 1.2

Diagram of the probable ancestor of land plants (Drawn by MMG Karasawa based on Willis and McElwain 2002)

In the Cambrian and Ordovician periods (Table 1.1) tectonic activities were relatively intense promoting the reorganization of the continental plates deeply interfering in the sea levels, ocean currents, climate and geographical distribution of organisms affecting the life of all organisms of that time. Evidence suggests that in the high Cambrian all continents were distributed around the equator. The largest Gondwana moved to the South Pole while the two smaller coalesced into a continental landmass called Laurasia (Zunino and Zullini 2003). The reorganization of continental plates promoted the occurrence of glaciation (440 myr) that led to a dramatic reduction of sea level of approximately 70 m and around this period the first evidences of land colonization were found (Brown and Lomolino 2005; Lomolino et al. 2006). In parallel to changes in environmental conditions, changes were necessary in the structure, form and reproduction of plants to ensure their existence in the terrestrial environment (Willis and McElwain 2002). At the end of the Cambrian period green algae had evolved into highly complex shapes. But, the biochemical pathways, such as those that facilitated respiration and photosynthesis in the cyanobacteria, fundamental to the life of plants and algae who occupied aquatic environments, and the advent of meiosis, that promoted the emergence of more sophisticated forms of life, were established only in the Ordovician period (510~438 myr) (Bateman et al. 1998). Among the fossil specimens recorded, the best preserved is the Isochadites of Codiaceae family. This fossil shows gametocysts with reproductive structures, probably one of the earliest evidences of sexual reproduction (Banks 1970). The main algal groups from the Cambrian period were Dasycladaceae and Codiaceae, among the green algae, and Solenoporaceae, among the red algae. In the intermediate stage of the Ordovician period the Codiaceae presented segmented structures and an internal tubular structure found until today.

From the middle of the Ordovician period to the beginning of the Silurian (470~430 myr) (Table 1.1), fossils evidence were found from the development of specialized cells for water and nutrients transport, such as some other precautionary measures against desiccation, mechanical support and reproduction mode that decreased environmental water dependence (Willis and McElwain 2002). The aerial parts and the underground sporophytes of the first vascular plants differed little structurally one from another, but undoubtedly the primitive plants resulted in more specialized plants with differentiated body. These plants consisted of roots, that functioned in the fixation and absorption of water and minerals, and stems and leaves, which provided a well-adapted system to the necessities of life on earth (i.e., absorption of sun light, carbon dioxide from the atmosphere and water from the soil) (Raven et al. 1995, 2007). The Cooksonia Aglaophyton major, also known as Rhynia major (Edwards 1986), can be considered an intermediate stage in the evolution between primitive and vascular plants (Prototracheophyte) because they do not have tracheids. However, the presence of cells similar to moss hydroids was detected. During the transition, plants also underwent other changes that made possible their reproduction in terrestrial environments, with the production of resistant spores being one of the earliest stages for enduring drought (Raven et al. 2007). As an example, one can cite Cooksonias (Fig. 1.3) which were formed by sporangia containing spores inside (Mussa 2004). These sporangia could reach a maximum of 1.5 mm in diameter and 2.0 mm in length. The height of these plants ranged from 2.2 cm to 11 cm (Edwards et al. 2004).

Fig. 1.3

Diagram showing morphology of Cooksonia (Drawn by MMG Karasawa)

Plant evolution is associated with the occurrence of several changes in the gametophyte and the sporophyte (Fig. 1.4a). Currently, the most accepted hypothesis is that a spore producing ancestor, which lived in aquatic environments, has given rise to the first plants (Fig. 1.4a). This would have changed (by mutations) in gametophytic and sporophytic phases, some of which resulting in plants with an amplified gametophytic generation, which was nutritionally dependent (Fig. 1.4b) and presented the reproduction mode of most non-vascular plants (bryophytes). Some other changes resulted in plants with an amplified sporophytic generation, who produced nutritionally independent sporophytes (Fig. 1.4c), with a reproduction mode similar to that of vascular plants (tracheophytes) (Willis and McElwain 2002). The emergence of the gametophytic stage must have kept the dependence on water, which has become essential to transfer male and female gametes, as well as for the initial growth of the sporophytic embryo. Furthermore, plants that have developed an amplified sporophytic generation showed a decrease in water requirement as a consequence of a continuous drying of the environment, and were selected for a nutritionally independent sporophytic stage. Thereby, neither the spore production nor its spread needed water. Mutant individuals that arose in this period showed reduction in size and in gametophytes complexity compared to sporophytes (Fig. 1.4c) and gave rise to an independent sporophytic phase. This strategy increased the resistance to drought and dissection. Other mutants presented a larger gametophyte size and became physiologically dependent on moist environments to transfer their gametes (Fig. 1.4b) due to the amplified gametophytic generation (Drews and Ydegari 2002; Graham et al. 2000).

Fig. 1.4

Diagram showing the alternation of generations with (a) gametophytic and sporophytic phases in algae and (b) amplified gametophyte generation as in most bryophytes. Plants of simplified life cycle with (c) amplified sporophytic generation that occurs in all vascular plants (a) hypothetical algal precursor to terrestrial plants (b) bryophyte (c) tracheophyte. Drawn by Karasawa et al. 2006

Many algae groups are known to reproduce sexually and asexually, and sexual reproduction surrounds the alternation of two kinds of generations, viz., the diploid sporophytic and the haploid gametophytic phases. In the gametophytic phase, male and female gametes are released from the gametophyte and the male gamete swim to the female gamete. They then merge to produce a diploid zygote. The germination of the zygote form plants called sporophytes that, when mature, undergo meiotic division to form haploid spores that are released to form a new gametophyte.

The differentiation between the amplification of either the gametophytic or the sporophytic generation has persisted to the present. In the current vascular plants the vegetative sporophyte is the visible part of the plant (Figs. 1.4c and 1.5), whereas in the non-vascular group the vegetative gametophyte is the visible part of the plant (Figs. 1.4b and 1.5). This differentiation may be the explanation why, throughout the geological record, bryophytes have remained small and restricted to humid environments and mixed areas, while the evolution of tracheophytes not only turned them into the largest group of plants in the planet, but also allowed them to occupy the most diverse ecological niches.

Fig. 1.5

Sporophyte and gametophyte evolution (Drawn by MMG Karasawa)

It has been suggested that the appearance and gradual increase in the number of spores in the fossil record has not only been an indication of the sporophytic phase development, but also of the fact that they have become resistant to deterioration, because of the presence of sporopollenin in the wall. This chemical substance can be found in the wall of the pollen grain of non-vascular plants, vascular plants and also in some algae (Kenrick and Crane 1997). This complex polymer has the function of providing drought resistance, strength and protection against ultraviolet radiation. Evidence from fossil spores indicate that land plants were originated in the intermediate stage of the Ordovician period, while the greatest divergence between hepatics, ceratophyllum, mosses and vascular plants groups must have occurred at the end of the Ordovician and in the Silurian (Bateman et al. 1998). Some of the early visible spores found in the geological records belonged to the end of the Ordovician period (~450 myr). They had a tetrahedron arrangement while others from younger sediments (~430 myr) are found as isolated spores with a distinct trilete form (Gray and Shear 1992). This trilete morphology (Fig. 1.6) as well as the tetrahedron arrangement provides strong evidences for the meiotic division because a diploid cell, when dividing by meiosis, produces four haploid cells (in this case, spores). The significance of the fossil spores arranged with a tetrahedron form is that most of them, especially in the beginning of the Silurian period (~430 myr), probably represent the first evidence of development of the sporophytic phase in the life cycle of plants (Willis and McElwain 2002).

Fig. 1.6

Spores arranged in tetrahedron form (a); composition of tetrahedron of spores (b, c) and isolated spores with distinct trilete form (Drawn by MMG Karasawa based on Willis and McElwain 2002)

At the beginning of the Silurian period (~417 myr) fossil evidence was found of the earliest vascular plants. They presented globose sporangia with cutinized spores and isolated spores in trilete form. Multiple reproductive structures were also found. The growth habit of these plants was the determinate type (i.e., terminal reproductive structures), and they had asexual spreading rhizomes and/or sexual reproduction with spores (Banks 1970).

1.2 Homospory, Heterospory and the Evolution of the Egg and the Pollen Grain

Until the beginning of the Devonian period, the fossil records show only the presence of determinate growth habit. The first evidence of plant growth habit of indeterminate type comes afterwards. In the beginning of the Devonian, the Rhynie type is one of the most important evidences of growth habit and reproductive structures. Their fossil records show various forms of reproductive structures found in plants that may be isolated, multiple and even fused (Willis and McElwain 2002; Mussa 2004).

Evolution promoted the emergence of new types of plants and the extinction of old types in the Devonian period. The blue-green algae reached their peak in the beginning of this period, while the Characeae were found in more advanced stages of evolution. These algae inhabited fresh water, but occasionally were found in seawater, and can be recognized by the arrangement of its branches with a structure characterized by nodes and internodes, and their sexual organs, the oogonia, which was attached to the nodes. The oogonia was a single egg cell surrounded by coiled tubules. The most ancestral fossil of this group is represented only by its zygote. These individuals became highly specialized in the early Devonian period, and these structures have been conserved until today, differing only in some minor details (Banks 1970). Another marine algae group, also found in this period, was the Dasycladaceae and the Codiaceae, inhabiting environments that contained fresh water and also brackish water. It is believed that this group would have migrated from marine environments to fresh water, at the end of the Silurian period, because its zygote (oospore) was able to resist to desiccation, where the water supply could dry occasionally, that was an obvious adaptation to the new habitat. This type of spore resistance was not a common characteristic between seawater algae.

Land plants and Charophyceae algae may have moved to new environments at the same time in the past. According to the existing hypothesis, all plants have originated in marine environments wherein the migration to terrestrial environment would have led to new ecological niches. It has been reported, between the fossil samples of the Rhynie type, as found in the Devonian period, Zosterophyllum divaricatum with 400 myr, with sporangia attached laterally and at the apex of the stem (Gensel and Andrews 1987). Their spores were small, with a diameter of approximately 55–85 μm, ranging from circular to subtriangular, smooth and with a distinct trilete mark. This plant was approximately 30 cm in height and was composed of branches that grow from rhizomes. However, Psilophyton dawsonii (395 myr) and P. robustius had a central stem with undeterminate growth which grew to a height of 2–60 cm and had lateral branches with fertile apices, each apex consisting of approximately 32 sporangia (Willis and McElwain 2002; Mussa 2004).

The first representatives of vascular plants were the moss group, and among them three genera, Asteroxylon, Calpodexylon and Protolepidodendron, illustrate similar characteristics between the modern and the fossil type. In this phase (~400 myr) the plants produced globose and reniform sporangia with spores in the trilete form that, during this period, evolved from homospory (spores with the same size) (Fig. 1.7a) to heterospory (spores with different sizes) (Fig. 1.7b, c), in which the small spores are called microspores (3~5 μm) and the large spores are called megaspores (150~200 μm) (Mussa 2004). This is considered one of the most important courses of evolution for the emergence of seed plants (the gymnosperms) (Willis and McElwain 2002).

Fig. 1.7

Sporangium with homospores (a), sporangium with heterospores (b) and two sporangia on the same individual, one bearing only megaspores and the other bearing only microspores (c) (Karasawa et al. 2006)

It is postulated that the largest spores were the precursors of today’s megaspores, and the smallest were the precursors of pollen grains. The most accepted theory is that a mutation would have given rise to two sizes of spores (Fig. 1.8) (Thomas and Spiecer 1987).

Fig. 1.8

Diagram illustrating the transition from homospory to heterospory in vascular plants (Willis and McElwain 2002, modified by Karasawa et al. 2006)

These spores of different sizes would have been initially developed together in the same sporangia (Fig. 1.7b) and, subsequently, over the course of evolution, would have arisen in separated sporangia (Fig. 1.7c), the megaspore in the megasporangia and the microspores in the microsporangium (Andrews 1963; Banks 1970).

According to the fossil records, heterospory gradually evolved in the Devonian period. The evolution of megasporogenesis led eventually to the degeneration of three out of four products of female meiosis, which produces now only one viable megaspore per megasporocyte (Fig. 1.9a–c). The surviving megaspore can thus receive more nutrients from the mother plant (Fig. 1.9d) (Willis and McElwain 2002).

Fig. 1.9

Scheme illustrating the degeneration of megaspore indicating the probable evolution of the megasporogenesis in the megasporangium (Willis and McElwain 2002, modified by Karasawa et al. 2006)

But this megaspore was still subject to attack and desiccation, so that natural selection soon led to the evolution of fused sterile leaves located nearby for its protection, as shown in Fig. 1.10 (Thomas and Spiecer 1987).

Fig. 1.10

Stages in the evolution of megaspore protecting structures from sterile leaves existing nearby (Karasawa et al. 2006 based on Andrews 1963 and Willis and McElwain 2002)

It is believed that this development would have occurred around 370~354 myr with the evolution of the pro-gymnosperms (Willis and McElwain 2002). Support for this theory is provided by Genomosperma kindstoni, Genomosperma latens and Salpingostoma dasu (early Carboniferous), and Physostoma elegans, Eurystoma angulare and Stamnostoma huttense (late Carboniferous) (Andrews 1963).

The megaspore and the microspore evolved at the same time. The microspore has given rise to the pollen grain that could travel a long distance and was able to develop a pollen tube to reach the female gamete into the embryo sac, where the zygote is formed. Paleobotanical evidence shows that the pollen grain began its evolution around 364 myr, and only fossil spores are found before that. The first evidence of the pollen grain shown in the fossil record was termed pre-pollen and corresponds to an intermediate stage between the spores and the pollen grain, while it contained characteristics of spore (trilete form), but the evidence suggested that the germination took place on or near the opening of the megasporangium (Fig. 1.11) (Willis and McElwain 2002).

Fig. 1.11

Scheme illustrating the likely mode of pollination used by the pre-pollen to reach the pre-egg (Karasawa et al. 2006)

The pollen grain is distinguished from the spore in the structure and in the mode of gamete release. In heterosporic plants, the microspores release flagellated gametes at the spore distal end (i.e., the opening of the trilete), which swim to the archegonium for fertilization. The pollen grain, by comparison, produces a pollen tube at the distal end through which the gametes are transferred directly to the egg (Fig. 1.12).

Fig. 1.12

Diagram showing the difference in the release of the gametes from spores and from pollen grains (Willis and McElwain 2002, modified by Karasawa et al. 2006)

With the development of seed protection, it has become necessary to improve the mechanism of pollen reception by the gynoecium so that it could reach the ovule in the ovary to form the zygote by fertilization. In the first seed plants the egg protection was partial, and the sterile leaves were fused only at the megasporangium basis (complete fusion later would evolve producing the carpels), leaving the structure that received the pollen free (Fig. 1.13a). These structures, combined with the protection lobes, were highly effective in capturing the pollen grain carried by the wind. However, some of the first eggs had other mechanisms to receive pollen, such as the presence of inner and outer fringes in the integumentary lobes (Fig. 1.13b) and the droplet pollination (Fig. 1.13c). The droplet pollination mechanism used a substance consisting of an exudate, which adhered to the pollen after their deposition on the surface, until the dehydration and consequent volume reduction brought the pollen grain inside the micropyle allowing fertilization (Thomas and Spiecer 1987; Willis and McElwain 2002).

Fig. 1.13

Mechanisms of pollen reception present in the fossils. (a) Presence of lobes, (b) inner and outer fringes, and (c) droplet pollination (Willis and McElwain 2002, modified by Karasawa et al. 2006)

1.3 The Evolution of the Terrestrial Flora

From the end of the Devonian to the beginning of the Permian period (395~286 myr), the terrestrial flora evolved from small non-vascular and vascular plants to vegetation that included trees with 35 m or more. Due to adaptations to live in terrestrial environment, vascular plants have been ecologically successful, becoming numerous and diversified in the Devonian period (Willis and McElwain 2002; Raven et al. 2007). During this phase there was a significant change in the global environment, as the movement of continental plates promoted a dramatic change in the climate. Under these conditions major innovations occurred in the morphology of terrestrial vegetation, with the emergence of new plants (390~365 myr) and a relatively rapid increase in the number of species (Gensel and Andrews 1987; Lomolino et al. 2006). Everything indicates that the environmental selection pressure was probably the main factor in these events. The fossil records of that time suggest a steady increase in ecological complexity at all spatial scales. Ecosystems, from the beginning of the Devonian, where composed of plants with simple and dynamic interactions. Differences in local dynamic in landscape scales were small and difficult to differentiate. The communities consisted of groups of plants with clonal and opportunistic reproduction (Willis and McElwain 2002). Two are the groups of plants known to have evolved at this early period: Sciadophyton and Protobarinophyton (Banks 1970).

Typical plants with rhizoids and rudimentary roots co-evolved, supporting the turgor pressure and showing a history of evolution of homospory. The best known history is that of Rhynie, which provides crucial informations about the ecosystem. Yet sporophytic structure was simple; many sporophyte ecological strategies clearly co-existed, as the hability of Rhynie gwynne-vaughanii to disperse rapidly on the substrate by lateral and decidual branches. The diversification of the sporophyte was exchanged by a wide array in the morphology of the gametophyte which registers the many variations helping the syngamy. In the mid-Devonian period precursors of modern horsetail (Sphenopsida) were found. Two other extinct types, Cladoxylon and Aneurophyton, were also found. An empirical study of megafloras and deposition environments has shown the occurrence of early stages of landscape partition by a group of higher plants. Typical floras of wetlands were dominated by a genus with fern features (Rhacophyton), and the adjacent areas included the lycopods, which were distinguished from the interfluve floras, while the dry parts of the plains were dominated by the pro-gymnosperm Archaeopteris (Willis and McElwain 2002).

Also, at the end of this period (370~354 myr) the evolution of the ovule occurred giving rise to the seed, which was one of the most spectacular innovations that emerged during the evolution of vascular plants. The emergence of the seeds was one of the major factors responsible for the dominance of the current seed plants, which evolved over a period of several hundred million years. The main factor of this success was the fact that the seed gives the embryo stored food that becomes available in critical stages of germination and establishment, thus promoting a selective advantage in relation to related groups endowed with free spores and ancestral groups that liberated spores (Raven et al. 2007).

Apparently, the Pteridosperms, first seed plants, originated in humid landscape and then settled as opportunistic in disturbed landscapes with physical stress, including the relatively arid habitats. Fossil evidence showed that at the same time the main stem structure, that characterizes the most modern groups of gymnosperms, ferns, Sphenophyta and a series of lycopod groups were developed (Willis and McElwain 2002). When forming the structure of the main stem, the development of the conducting system of the central cylinder (the eustele) was a major novelty which allowed a greater growth in height and the effective transport of water and nutrients throughout the plant up to the canopy (Raven et al. 2007). Geological evidence suggests that the conductive system became progressively more complex over the evolution of vascular plants, apparently presenting, at the end of the Devonian period (~374 myr), three different types namely: protostele, siphonostele and eustele (Fig. 1.14a–c).

Fig. 1.14

Morphological differences between the fossils of the earliest eustele types. (a) protostele: vascular plants ~420 myr; (b) siphonostele: fossils with about 395 myr; (c) eustele: fossils with about 380 myr (Karasawa et al. 2006)

According to the fossil record, lycopods, sphenopsida, filicinaceae and pro-gymnosperms were the first trees producing spores. The ancestral lycopods were small plants, herbaceous and homosporous, as for example the Baragwanathia longifolia, with approximately 410 myr, while the first trees were found in fossils dating back to 370 myr, with the giant species Lepidodendron being one of the most commonest among the lycopods. It reached a height of 10~35 m and a diameter of 1 m. This species was heterosporous and the sporophylls were encountered in strobili containing microspores at the top and megaspores at the bottom (Willis and McElwain 2002; Mussa 2004).

The Sphenophyta currently comprises a group of 20 species, all having herbaceous growth habit and belonging to a single genus. Many fossils of this group have been assigned to the Carboniferous and Permian period (354~248 myr), including a series of arborescent forms, most of them belonging to Calamites, which grew to 18 ft of tall height or more. This plant had siphonostele stem with primary and secondary xylem and primary phloem. Another striking feature of the stem was the presence of nodes. Regarding the spore type they were homosporous. However, strong evidence point to the presence of heterospory in this group at the end of the Carboniferous (Willis and McElwain 2002).

The Filicinaceae can be found in the fossil record of about 360 myr, and many fossils in this group are quite similar to the present remaining forms. The Psaronius is an example of fossil of this group. It grew to a height of about 10 m, had long leaves and a protostele type of stem (Thomas and Spiecer 1987). Paleobotanical evidence suggests that in some species the root reached 1 m of diameter at the base of the stem. Most plants of this group were homosporous. In Psaronius the sporangium was big and had fused locus pairs (synangium). The bottom branches of the cones were fertile, suggesting that the arrangement evolved with the connecting to the megaphylls incorporating the sporangia at the bottom (Willis and McElwain 2002; Mussa 2004).

1.4 Evolution of Pro-gymnosperms and Gymnosperms

The second mass extinction occurred in the period between 395 and 290 myr. This event was marked by significant changes in the global environment accompanying the collision of the blocks that formed Gondwana and Laurasia during the Silurian, which formed the supercontinent Pangaea. The active movement of tectonic plates promoted a dramatic change in the climate, which changed from hot (24–32 °C—near the equator) to cold, with very low temperatures and dry climate (in the inner of the continent), and there were extensive glaciations in high regions in the southern hemisphere, with more intense effects in the inner part of the continent (Freeman and Herron 1998). During this period the sea level dropped 100–200 m. Moreover, land colonization and the consequent atmosphere CO₂ reduction contributed overwhelmingly to global cooling (Zunino and Zullini 2003; Lomolino et al. 2006).

The evolution of pro-gymnosperms occurred between the end of the Devonian period and the early Carboniferous period (~354 myr). The pro-gymnosperms comprise a group of plants representing the transition between ferns and gymnosperms. This group had some type of secondary xylem and phloem, the presence of bifacial vascular cambium and determined growth habit in some plants and indeterminate in others. In these, the most important advance, in relation to their ancestors, consisted in the fact that they presented a bifacial vascular cambium (i.e., a vascular cambium that produced secondary xylem and phloem). This type of vascular cambium is characteristic of seed plants and, apparently, first evolved in pro-gymnosperms. Among the pro-gymnosperms from the Devonian period, Aneurophyton (380~360 myr), a plant characterized by having tridimensional complex branching and protostele (i.e., a closed cylinder of vascular tissue), may be mentioned. Another important pro-gymnosperm was the Archaeopteris (370~340 myr) (Banks 1970). This plant, which also lived in the Devonian, had lateral branches with flat laminar structures considered leaves and presented stems of the eustele type (with single strands of vascular tissue arranged around the pith).

Regarding the reproductive system, most of the pro-gymnosperms were homosporous, but some species of Archaeopteris were heterosporous. Several groups of vascular plants without seeds thrived during the Devonian period, in which three of the most important were: Rhyniophyta, Zosterophyllophyta and Trimerophytophyta. These three phyla consisted of plants without seeds that presented a relatively simple structure, and all came to extinction at the end of the Devonian, approximately 360 myr. Only a fourth phylum of vascular plants without seeds, Progymnospermophyta, with intermediate characteristics between vascular trimerophyte without seeds and seed plants, did not become extinct. It is speculated that this phylum was the ancestor of the seed plants, the gymnosperms and angiosperms. Although these plants reproduced freely by the dispersal of spores, they produced a secondary xylem (wood) remarkably similar to extant conifers, being the only ones in the Devonian that produced a secondary phloem. The pro-gymnosperms and ferns probably originated from the older trimerophytes (Rhynia, Zosterophyllum and Trimerophyta), from which they differed primarily in having more elaborated and differentiated branch systems and more complex vascular systems than their ancestors (Raven et al. 2007).

With the decline of the seedless plant groups (spore-producing trees), only the Filicinaceae remained as today’s remnants of the Paleozoic era. The seed-producing plants appeared on the upper Devonian (~350 myr) and came to dominate the flora of the Mesozoic terrestrial landscape. The seed-producing plants were composed of five classes: pteridosperms (extinct), pteridophytes, cycads, ginkgoales and conifers. The first four had their peak at the end of the Paleozoic and early Mezozoic era. These differ considerably in the structure and form from their ancestors. The seeds of these plants were exposed in the same way as they are in the current pine strobilus (Andrews 1963; McAlester 1978).

The gymnosperms had great advantage over their ancestors because they were able to reproduce without external moisture. The male gametes did not need to swim to fertilize the female gametes, because they were able to form a pollen tube that carried the male gametes to the egg cell allowing the fertilization. The male gametophytic stage no longer required a liquid medium, because the pollen developed within a humid sporophytic plant tissue and the female gametophytic phase, reduced to the embryo sac, was embedded in the sporophyte. The pollen grain surrounded by a double impermeable membrane was highly effective in preventing water loss, and the thinner and more elastic inner membrane gave rise to the pollen tube. The small size and the large number of pollen grains allowed their transportation by wind to the stigma by issuing the pollen tube to reach the egg and promoted fertilization. The wind also made possible for plants that were far apart and from different individuals to become fertilized. After fertilization, the formed seed received all the nutrients (proteins, fats, starch, etc) which helped on the establishment of the embryo in the early stage of its development (Raven et al. 2007).

The pteridosperms was a group of seed plants that showed great development since the lower Carboniferous period. These comprised one of the principal groups of plants of coal-forming plants (McAlester 1978). On the other hand, Cycads and Ginkgoales dominated the landscape of the Triassic and Jurassic period, but declined rapidly during the Cretaceous, as the angiosperms developed, being relatively rare nowadays. It was in response to an increase in temperature and a decrease in moisture in the continent that the evolution of Cycads, Bennettitales and Ginkgoales occurred. These three groups still have living representatives, but in the Mesozoic era its global distribution was much more expressive (Willis and McElwain 2002).

Cycads belong to the group which currently comprises 10 genera and 100 species of plants, all dioecious (i.e., the population consists of plants with male strobili and plants with female strobili) (Mussa 2004), with no fossil record indicating the presence of monoecy (i.e., plants with separate sexes in cones of the same plant). Some of the earliest records are dated from approximately 280 myr (from Permian) and indicate that some species grew up to 15 m in height, although the first cycads were smaller, reaching about 3 m. The apices of these plants have been highly conserved through evolution, being very close to those found in the current cycads. Their reproductive organs were well documented in the fossil record. It is known that in these plants the female reproductive structure had eggs grouped into modified leaves called megasporophylls, and the male structure was located in modified leaves called microsporophylls, wherein each leave had small and compact pollen sacs attached on its surface. In groups of ancestral plants each pollen sac was able to produce a great number of pollen grains in the monolete form (Willis and McElwain 2002).

The Bennettitales, in turn, have fossil records dating from the beginning of the Triassic to the end of the Cretaceous (280~140 myr). This group had many morphological similarities to that of current cycads and also of the extinct ones. One of the genera of Bennettitales most commonly cited is Williamsonia (Banks 1970). This plant had reproductive structures that resemble the angiosperms, demonstrating that there is a close evolutionary connection between the members of this group and the first angiosperms, which were called pro-angiosperms. Numerous examples indicate, with few exceptions, that this group was at the beginning unisexual and then became bisexual. The female strobilus was composed by eggs surrounded by sterile leaves and a tubular-shaped integument comprising the mycropile. The male reproductive structure was composed of leaves and structures containing small pollen sacs composed of fused sporangia (sinangio). The pollen grain, whose shape was monolete, resembled that of the cycads. In this group, the occurrence of both wind and self-pollination has been suggested, and some evidence indicated that animal pollination may also have been used (Willis and McElwain 2002).

In the Ginkgoales group the records of the first fossils date from 280 myr, and only a single species is currently found, Ginkgo biloba. Fossil evidence suggests that this group was composed of at least 16 genera and contributed significantly to global vegetation. The great similarity between extinct species and Ginkgo biloba has led to its description as a living fossil (Thomas and Spiecer 1987). The main eustele stem had a great content of secondary xylem and show characteristics that are difficult to separate from certain conifers such as Pinus. Ginkgo eggs are born at the end of short lateral branches, in number of two or three, and are connected by a peduncle. The microsporangiums (pollen sacs), however, are born in the leaf axils of short lateral branches (Mussa 2004). The reproductive strategy used by Ginkgo biloba is a dioecious type, but fossil evidence indicates that there was much variation between the reproductive structures (Willis and McElwain 2002).

Another important group was the Glossopteridaceae. This group has also been suggested as a possible ancestor of the angiosperms, as these plants, which had deciduous and arborescent habit, also presented highly modified reproductive organs attached to the leaves. Fossil evidence suggests that these plants grew up about 10 ft of height and had an eustele stem type, indicating close similarity with the modern Araucaria. Its female reproductive structures were quite diverse, ranging from uni- to multiovulate structures (Willis and McElwain 2002).

Conifers appeared in the Carboniferous period, and most of the primitive forms are currently extinct. The earliest conifer, found in fossil records in Yorkshire, Swillingtonia denticulata, was dated as belonging to the Upper Carboniferous (~310 myr). This group showed an increase in the fossil record during the Permian period, yet the largest radiation occurred in the Triassic period (245~208 myr), with seven families (Podocarpaceae, Taxaceae, Araucariaceae, Cupressaceae, Taxodiaceae, Cephalotaxaceae and Pinaceae), which currently are widely dispersed. Among the conifers in evidence in the fossil records, Utrechtia is indicated as the most ancient plant. This plant, belonging to the Permian period, reached 5 m and possessed an eustele stem type with sap conducting vessels and tracheids. Its fossil specimen show morphological similarity to many existing conifers. The male and female reproductive structures of the plant were probably placed in different parts of the apex. Fossil evidence suggests that the pollen-producing male structure was very similar to current conifers (Thomas and Spiecer 1987). The female cones consisted of an axis of approximately 8 cm in length in which a reduced leave with the reproductive structure was found. This small fertile apex resembles an intermediate stage between the cordaites and the structure of modern conifers (Crane 1985; Doyle and Donoghue 1986). Thus, conifers, in general, can present a monoecious or dioecious reproductive strategy, containing male cones on the abaxial side of the leaf and egg cones in the bract surface (Mussa 2004).

1.5 Evolution of Angiosperms

As far as we know, the angiosperms diverged from a seed plant common ancestor between the late Jurassic and early Cretaceous between 130 and 90 myr (Crane et al. 1995), reaching dominance between 50 and 80 myr (Axelrod 1970), soon after the occurrence of the fifth mass extinction (Zunino and Zullini 2003). According to Stuessy (2004), the angiosperms originated from seed-producing fetuses (plants like ferns). These would have evolved in the Jurassic period, with the emergence of the carpel followed by the occurrence of double fertilization and only after these two evolutionary steps the mutations responsible for the appearance of the flower component parts (i.e., sepals and petals) should have occurred. This transition would have taken more than 100 myr to become complete.

The presence of extreme and inconstant weather would have restricted the location of these individuals to environments of higher altitudes and median latitudes, where there was a predominance of tropical dry weather until the early Cretaceous. However, with the fragmentation of Gondwana (~206 myr) the encroachment of the continents by ocean waters occurred (Lomolino et al. 2006), increasing the ocean surface around the continents. As a result, mild and homogeneous climate began for all continents, promoting the colonization of angiosperms in low altitude regions. On the other hand, the fragmentation and subsequent continental drift of Gondwana (originating South America, Africa, India and Australia) promoted the isolation of plants in the Cenomonian-Albian (Cretaceous intermediate phase) and significant climate change. The barriers to gene flow provided changes in allele frequencies and increased diversity due to pre-adaptive and adaptive functions (Axelrod 1970).

Morphological and molecular evidence and phylogeny studies have shown that the angiosperms had a monophyletic origin (Doyle and Donoghue 1986), with Amborella trichopoda being the most primitive angiosperm (Nymphaeales—aquatic lilies) and nearest sister group among them, from which the first diverging lineage would have appeared (Qiu et al. 1999; Soltis et al. 2000). On the other hand, two genera of Gnetales, Gnetum and Welwitschia, did not form a cluster with the angiosperms, showing a high level of consistency in the grouping with the conifers (Qiu et al. 2000). Fossil records confirm the possibility that all angiosperms have originated from a common ancestor derived from the gymnosperms and that this had no flowers, neither fused carpels, nor fruits. The oldest known fossil has been dated to 125 myr and was found by Dilcher et al. (2002), probably being the mother of all angiosperms. The fossilized material, belonging to the group of Liliaceae, has been found in China and was called Archaefructus sinensis. Archaefructus is considered a key fossil because it presents carpels; however, it does not have flowers. This absence of perianth parts and the presence of carpels and separate stamens, distributed along the axis in a vertical reproductive structure, posed questions about the possibility of the existence of unisexual flowers without perianth and more specialized forms in the base of the angiosperms (Friis et al. 2003). This information confirms the hypothesis that the angiosperms have started their evolution in the Lower Cretaceous (approximately 130 myr), reaching its dominance in the green world about 90 myr. Around 75 myr, many families existed already and some of the modern genera could also be found (Raven et al. 1995).

Currently, the angiosperms are worldwide dominant and have about 300–400 families and 240,000 to 300,000 thousand species, while the ferns have about 10,000 species and gymnosperms only about 750 species (Willis and McElwain 2002; Bernardes-de-Oliveira 2004). The appearance and rapid diversification of monocots and eudicots led these plants to an increasing dominance over the last 35 myr of the Upper Cretaceous (100~65 myr). At approximately 90 myr, several orders and families of existing flowering plants had already been established and the flowering plants had reached dominance throughout the Northern Hemisphere. In the following 10 myr they reached dominance in the Southern Hemisphere by having adaptive characteristics of drought and cold resistance such as smooth leaves of small size, presence of vessel elements with more efficient cells conducting sugar through the plant phloem and a resistant seed coat protecting the embryo against desiccation. The emergence of the deciduous habit also occurred early in the evolution of this group, allowing plants to become relatively inactive during periods of drought, extreme heat or cold, which probably contributed to the success recorded in the last 50 myr, when the climate in the world suffered frequent changes (Raven et al. 1995). One of the most important factors, for the angiosperms, has been the evolution of the reproductive system which allowed more precise pollination and a more specialized seed dispersal mechanism. Thus, individuals were able to occur widely dispersed in many different types of habitats such as desert, mountains and shallow waters, not developing only on open sea and polar regions. Other important factors of success were: a more developed autotrophic diploid phase, a reduced haploid stage, double fertilization and the development of carpels for a greater protection of the seeds (Paterniani 1974).

Thus, all angiosperms necessarily have flowers, fused carpels, double fertilization (responsible for the endosperm formation), microgametophytes with an extremely varied number of nuclei, stamens with two pairs of pollen sacs and the presence of sieve tubes and companion cells in the phloem (Friis et al. 1992; Bernardes-de-Oliveira 2004). So, the evolution of the angiosperms reported the presence of two new variants: the presence of floral whorls (sepals and petals), and the presence of both sexes in the same flower (hermaphroditism). Until this period (late Carboniferous—Early Cretaceous) flowers had no whorls, and each sex was located in a reproductive structure, on the same plant or separate plants. During evolution, the male and female strobiles, present in different locations and structures in gymnosperms, were found in a single arrangement. Additionally, the sepals and petals appeared, producing ornaments for the flowers which were beginning to evolve.

But how could this be possible?

Molecular studies have identified three factors (composed of one or more genes) that control the production of flower whorls, which were named factors A, B and C (Coen and Meyerowitz 1991). Currently, genes from gymnosperms that show a high similarity with floral initiation genes of flowering plants (transition from vegetative meristem to reproductive meristem) have also been isolated, suggesting the evolutionary conservation of the biological role of these genes (Lobo and Dornelas 2002). However, Kramer and Irish (2000) studying the level of conservation of these genes in lower eudicotyledons, magnoliids and monocotyledons, found that the ABC program is conserved only in some aspects while in others it showed a high degree of plasticity. Since genes are relatively well conserved in higher eudicotyledons, these may have been established only later in the evolution of angiosperms. In angiosperms these factors are responsible for the formation of sepals (Se), petals (Pe), stamens (St) and carpels (Ca) (Fig. 1.15), and in conifers (gymnosperms) its role is not well known yet. Apparently, factors B and C are the oldest existing in conifers. It is known that factor C alone determines the formation of carpels, but in association with factor B it is able to determine the differentiation of stamens. On the other hand, factor A alone determines the formation of sepals, while its association with factor B promotes the differentiation of the petals (Fonseca and Dornelas 2002). Thus, for the angiosperm flower to emerge it was required the presence of three factors (A, B and C) in association, and yet the evolution of a fourth factor, the transcription factor, called SUPERMAN (SUP), that should act on factor B to allow the expression of factor C to produce the carpels. Therefore, the evolution of angiosperms only became possible thanks to the presence of all these factors in combination (Fig. 1.15). For more details on this subject, see Chap. 2 of this book.

Fig. 1.15

Representation of the factors involved in the differentiation of vegetative meristems in the plant reproductive system (Karasawa et al. 2006)

The primitive angiosperms presented solitary flowers at the end of the branches or loosely arranged in axes as in many Paeonia species, and the branches above the internodes had reduced leaves and secondary flowers (Fig. 1.16a). The developmental pattern of the floral axis and the formation of stamens and carpels differed greatly and showed a long period without differentiation of meristems, with the occurrence of only an increase in size and, subsequently, the differentiation into three regions: central initial zone, peripheral and apical dome of the meristem. Initially, the development of the perianth and androecium in the most primitive angiosperms was poorly differentiated in sepals and petals. The tapetum was probably composed of bracts and modified leaves. In a number of genera, such as: Calycanthus, Paeonia and some species of Hibbertia, gradual transitions occurred in the leaves, which changed from bract structures (modified leaves) to sepals and typical petals (Stebbins 1974). Studies of floral structures and reproductive organs from the Cretaceous and Tertiary also show a general increase in morphological and organizational diversity of the reproductive organs of angiosperms throughout evolution (Fig. 1.16a–f). However, the fossil records of floral organs are incomplete, especially in the early stages of diversification of this group, and this is also consistent with the records of other organs such as leaves and pollen (Friis et al. 1992).

Fig. 1.16

Evolution of the flower types: (a) small with a few parts, (b) acyclic and hemicyclic, (c) monochlamydeous and unisexual, (d) cyclic, heterochlamydeous and actinomorphic, (e) epigynous and heterochlamydeous, (f) sympetalous, (g) epigynous and monochlamydeous, (h) zygomorphic, (i) form of a brush, (j) Papilionaceae family, (k) tube-shaped corolla (Friis et al. 1992, modified by Karasawa et al. 2006)

The phyllotaxis of the floral parts of the Albian Stage (Lower Cretaceous) is unclear, but a few forms show evidence of a spiral arrangement of the parts. At the beginning of the Cenomanian (early Upper Cretaceous) the two major types of phyllotaxis in angiosperms were already established, comprising acyclic flowers with the parts arranged in spirals and hemicyclic flowers (Fig. 1.16b), with the parts of the perianth arranged in partially spiral whorls (Basinger and Dilcher 1984). From the beginning to the intermediate stage of the Cenomanian, acyclic and hemicyclic flowers were widely scattered among the angiosperms, and its importance declined with the diversification of cyclic flowers (Fig. 1.16d), which predominated in the fossil floras of the Santonian-Campanian stage (Upper Cretaceous), but the fossil pollen suggests that probably these forms were already established at the end of the Cenomanian stage.

The information on the number of floral parts in the reproductive structures of the Albian stage is also scarce, whereas the number of carpels varied from 3–8 to over 100. The known number of stamens is three and five, but unfortunately this is based on only two floral structures. Polymeric (with numerous parts), acyclic and hemicyclic flowers, with an indefinite number of parts, were apparently prevalent in the Cenomanian stage. On the other hand, cyclic flowers (Fig. 1.16d) presented, mostly, five parts, but some evidence suggests the existence of flowers with four and six parts also in this period. Apparently, the first cyclic flowers were isomerous (with the same number of floral parts in each whorl). The heteromerous, in turn, have only been established in an intermediate stage of the Cenomonian period and dominated the Santonian-Campanian stage having, usually, the perianth and androecium in a number of five and the gynoecium with two to three carpels. On the other hand, the true trimerous flower types were established and were relatively common in the Maastrichtian stage (late Upper Cretaceous). The perianth of the beginning of the Cenomanian was already established with different types of calyx and corolla. In relation to the symmetry and fusion of floral parts, it appears that all were apparently actinomorphic with radial symmetry and with the perianth parts free. The bilateral symmetry in flowers has occurred approximately 60 myr after the origin of angiosperms and is found in many fossil records of the Paleocene and Eocene and Upper Cretaceous when it was associated with the presence of social insects, and its co-evolution occurred in a number of families at different stages (Dilcher 2000). Fossils of zygomorphic flowers (Fig. 1.16h) were found only in the Maastrichtian stage, but evidence indicates that zygomorphism may have been established at the beginning of the Campanian stage. As to the distinct differentiation of the flower parts, it is known that this was found in late Paleocene in Papilionoideae flowers, while the first sympetalous flowers were observed in fossils of the Santonian-Campanian stage and a number of them were found in the Maastrichtian (belonging to the final stages of Cretaceous). In the Cretaceous, sympetalous flowers generally show a shallow and wide open tube form (Fig. 1.16f), while the deep tube forms were established in the Paleocene and early Eocene (Friis et al. 1992).

Based on the information originated from fossils, it can be inferred that flowers of the first angiosperm presented individual carpels, unisexual or bisexual small flowers, and radial symmetry (Friis et al. 1992). Also Dilcher (2000) has verified the presence of only small and medium-sized flowers, among the oldest fossil record of angiosperms. This author believes that the flower sizes had a relationship with the size of the pollinator, and the subsequent variation in the size of these flowers suggests a wide range of pollinators. The author states further that water and wind also participated in the process of pollination.

As for the position of the ovary, the structure of the fossils of angiosperm flowers and fruits of the Albian and early Cenomanian was a hypogeous type (Fig. 1.17a). The epigynous flowers (Fig. 1.17c) were found well established in the beginning and middle of the Cenomanian stage, and their radiation apparently reached its peak in the Santonian-Campanian stage, comprising around two-thirds of all the floral structures of this stage, decreasing in the Tertiary period. Currently, the epigeous flowers are present in one quarter of all families (Grant 1950).

Fig. 1.17

Types of inflorescence according to the position of the ovary: (a) hypogeous; (b) perigynous; (c) epigeous (Karasawa et al. 2006)

With respect to the male reproductive structure, the first fossil records describe the existence of three stamens fused at the base, and also unisexual flowers with five stamens present in the Albian stage (intermediate stage of the Cretaceous) while in the Santonian-Campanian stage (Late Cretaceous) the stamens were well established in the fossil records. Fossil stamens with free filaments were found in the Cretaceous. The pollen sacs of all the anthers of the first known fossils have four sporangia, and only in the Santonian-Campanian stage the evidence of anthers with two sporangia appears. Dehiscence, initially, was longitudinal (Santonian-Campanian), while at the beginning of the Paleogene the first records of dehiscence by two or more valves were found, and in the Paleocene, the fossils presented the first anthers with apical dehiscence (Friis et al. 1992).

Pollen of the first angiosperms had a single opening, as found in monocots and in some other groups of angiosperms, as well as in the Cycadaceae, Ginkgoaceae and other groups. Currently, there are four known pollen types present in fossils found in the earliest angiosperms (Clavatipollenites, Pre-Afropollis, Spinatus and Liliacidites) and a fifth type (Tricoliptes) which can be found in the more recent angiosperms.

The female gametophyte has undergone major changes during the evolution of the angiosperms, with the evolution through modules (Fig. 1.18). It is believed that, in the beginning, the module was composed of four cells located in the micropyle region, comprising two synergids, one egg cell and one central cell, giving rise to diploid endosperm individuals. Williams and Friedman (2002) have shown that the presence of diploid endosperm was common in ancient lineages of angiosperms. Subsequently, the micropylar module would have suffered a duplication, which originated eight nuclei/seven cells, resulting in an embryo sac composed of the module located in the chalaza region (giving rise to three antipodes—which degenerate soon after fertilization), a central cell composed of two nuclei and the module located in the micropyle region (Friedman and Williams 2003). Thus, this is how individuals with triploid endosperm that presented 2:1 maternal/paternal ratio would have been originated (Williams and Friedman 2004). During evolution, the modules continued to be duplicated giving rise to endosperm with higher ploidy levels, with the presence of 1–14 nuclei in the polar region being sometimes detected (Friedman et al. 2008).

Fig. 1.18

Modular evolution of the female gametophyte (Drawn by Karasawa, MMG based on Williams and Friedman 2003, 2004)

The fossil record of the female reproductive structure of the first angiosperms have shown free carpels of an apocarpic type, this being the prevailing condition at the beginning of the Albian and the Cenomanian (Fig. 1.19). The syncarpy (fusion of carpels), in turn, was only established at the end of the Albian (intermediate phase of the Cretaceous), and was represented by a number of taxa in the early Cenomanian. Basinger and Dilcher (1984) described a fossil of about 94 myr. According to the authors, the fossil had pentamerous flowers with distinct sepals and petals, fused carpels and a floral receptacle. According to Friis et al. (1992), the syncarpic forms have become diversified at the end of the Cretaceous, being the most common reproductive structure in the Santonian-Campanian flowers.

Fig. 1.19

Evolution of the fruit types: (a) follicles and nuts of apocarpic ovaries; (b) capsules; (c) nuts; (d) drupes; (e) berries; (f) pods. (b–f: syncarpic ovaries) (Friis et al. 1992, modified by Karasawa et al. 2006)

The emergence of fused carpels was paramount in the evolution of angiosperms, this being the prevailing characteristic in their separation in relation to other seed plants. The fusion, almost always complete, has the task of protecting the unfertilized egg from the external environment. There are theories suggesting that the fusion of carpels occurred to promote protection against beetles and other herbivores. However, Dilcher (2000) suggests that this is more directly related to the evolution of the flowers bisexuality. With the flower evolution, the male and female organs were approximated, requiring, therefore, a protection against self-fertilization. To promote the necessary protection, mechanical (fusion of carpels) and chemical (self-incompatibility systems) barriers would have arisen, so that plants would prevent the growth of the pollen tube. Moreover, the addition or subtraction of sepals, petals and stamens was important to promote cross-pollination (outcrossing) and the emergence of nectaries was responsible for increasing the pollinization by insects.

Currently, 83 % of the taxa of existing angiosperms present syncarpy in the gynoecium (Endress 1982). Initially the syncarpic ovaries were, as it seems, partially separated and divided according to the number of loci corresponding to the number of carpels in the early Albian-Cenomanian stage. In the Santonian-Campanian, a number of different types with unilocular ovaries were developed, while the secondary divisions showed their first occurrence in the angiosperm fossils of the Maastrichtian stage (Friis et al. 1992). At this stage, “gynoecium” and apocarpic fruits (Fig. 1.19a) of the earliest known fossil record show no evidence of distinct “styles” and stigmatic area, as this type of characteristic shows inconsistency. The first syncarpic fruits (Fig. 1.19b–f) were found in the Albian-Cenomanian stage (lower third of the Cretaceous), where the fruits were apparently dry and without obvious modifications for dispersal. However, the evolution of syncarpy was relatively rapid and from the beginning to the end of the Cretaceous practically all kinds of syncarpic fruits were already established. The follicles and nuts were the fruits originated from apocarpic ovaries, while the separated capsules (Fig. 1.19b) were derived from syncarpic ovaries. The first fossils of fruits with pulp date from the middle third of the Cretaceous, while the first evidence of berries was found only in fossils belonging to the Maastrichtian stage (late Cretaceous) (Fig. 1.19e). The pulp fruits became relatively common during the early Paleogene period, increasing considerably their diversity in relation to size, indicating wide variation in dispersal mechanisms (Tiffney 1984).

It is believed that the first angiosperms experienced a variety of pollinators, being pollinated by water, wind or animals. However, it was the association with animals that provided their highest diversification (Bernardes-de-Oliveira 2004) throughout their evolutionary history.

1.5.1 Unisexuality and Reproductive Strategies

Today’s angiosperms exhibit a great diversity of reproductive strategies. The vast majority of angiosperms present individuals with hermaphrodite flowers (72 %), while only 11 % of the plants have unisexual flowers (Fig. 1.20), 7 % being monoecious and 4 % dioecious (Figs. 1.21 and 1.22), whereas the intermediate forms of sexual dimorphism (gynomonoecy and andromonoecy) represent 7 % and plants with both forms of unisexual and bisexual flowers comprise 10 % (Figs. 1.21 and 1.23) (Ainsworth 2000; Richards 1997).

Fig. 1.20

Types of flowers found in Angiosperms: female (left), hermaphrodite (center) and male (right) (Karasawa et al. 2006)

Fig. 1.21

Frequency of reproductive strategies present in the angiosperms (Karasawa 2005)

Fig. 1.22

Types of plants found in angiosperms: monoecious and dioecious (Karasawa et al. 2006)

Fig. 1.23

Reproductive strategies adopted by angiosperms (Ainsworth 2000, modified by Karasawa 2005)

The unisexual flowers can be found placed in different parts of a single plant (monoecious) or on different plants, forming dioecious populations (Fig. 1.22).

The monoecious populations may present plants with a gynomonoecious form (female and hermaphrodite flowers), andromonoecious (male and hermaphrodite flowers) or trimonoecious (male, female and hermaphrodite flowers). Likewise, dioecious populations may have gynodioecious forms (plants with female flowers and plants with hermaphrodite flowers), androdioecious (plants with male flowers and hermaphrodite plants), and also subdioecious forms (female flower plants, plants with male flowers and plants with hermaphrodite flowers) (Fig. 1.23).

But how could such a diversity of reproductive strategies in angiosperms have arisen?

1.5.2 Evolution of Unisexuality

Unisexuality, in angiosperms, evolved as an outcrossing (cross fertilization) promoter system whose primary function is to obtain reproductive success in many different habitats in which they are found. The different evolutionary forces—selection, mutation, migration and drift—acting on hermaphrodite individuals throughout its evolution would have promoted the emergence of monoecious and dioecious populations (Barrett 2002).

• Evolution of dioecy

Dioecy, most of the times, evolved from self-compatible species (which can self-fertilize) in response to a selective pressure to promote outcrossing (Bawa and Opler 1975) and represents the change in the pattern of resource allocation for the male and female roles. Moreover, sexual dimorphism changes the spatial distribution of resources for pollinators, seed dispersers and predators (Bawa 1980; Sato 2002). Although Lebel-Hardenack and Grant (1997) believe that this evolution could have occurred only to allow a better allocation of resources optimizing reproduction. Ainsworth (2000) emphasizes that dioecy is one of the most extreme mechanisms that, most of the time, arises due to the deleterious effects of inbreeding depression or stressful environmental conditions and consequent resource limitations that prevent the hermaphrodite plants to maintain sexual functions and may promote the emergence of individuals with separate sexes. According to Charlesworth (1991), the evolution of dioecy may occur in hermaphrodite, monoecious and in populations that present heterostyly, as shown in Figs. 1.24, 1.25 and 1.26, respectively.

Fig. 1.24

Evolution of dioecy in hermaphrodite plants (Drawn by MMG Karasawa)

Fig. 1.25

Evolution of dioecy in monoecious plant populations (Drawn by MMG Karasawa)

Fig. 1.26

Evolution of dioecy in plants with heterostyly (Karasawa et al. 2006)

In hermaphroditic populations, dioecy evolved as a result of at least two mutations, one causing male infertility which promotes the emergence of female plants, and another mutation causing female sterility making possible the appearance of male plants. Thus, one mutation would affect the production of pollen grains and the other the ovule production (Charlesworth and Charlesworth 1978, 1998; Charlesworth 1991). The authors believe it is unlikely that the two mutations would occur simultaneously for the establishment of dioecy, and that dioecy from hermaphroditism must have involved intermediate types in the population providing the presence of hermaphrodites, together with male and female sterile plants in the same population (subdioecious population) (Fig. 1.24). And that, in all cases in which the first mutation caused female sterility, with the emergence of androdioecious plants, the fall of dioecy would have been observed, because there is no registered case showing that this form have been able to evolve into established dioecious populations (Fig. 1.24). On the other hand, Sato (2002) reports based on mathematical models, have shown that those plants of separate sexes (dioecious) only become well established if there is a gradual reduction in male fertility or a reduction of the seeds production in hermaphrodite plants, providing the evolution of dioecy.

The evolution of dioecious plants, from monoecy, seems to involve only a single route, as male and female flowers already co-exist in a single plant, with the occurrence of mutations causing male and female sterility on different plants being sufficient. This would cause the separation of sexes in plants as shown in Fig. 1.25. Charlesworth and Charlesworth (1998) believe that for dioecy evolution to occur starting from monoecious populations, a series of mutations changing the proportions of male and female flowers in the plants are required until either sex is allocated on separate plants.

It is believed that distyly (a type of heterostyly) could originate dioecious plants due to the occurrence of mutations abolishing the male functions in some plants and female functions in others in order to give rise to plants of separate sexes (Lloyd 1979) (Fig. 1.26). One hypothesis is that the shift from distyly to dioecy is initiated by a change in pollination biology of these populations with the discontinuity of supplementary pollen between individuals which can occur in two ways: by promoting the flow between long stamens and pistils, and eliminating the usefulness and functionality of short stamens and pistils (Beach and Bawa 1980). However, it is worth highlighting that heterostyly may have had independent origins in plants pollinated by animals to increase the accuracy of pollination (Barrett et al. 2000).

• Evolution of monoecy

Monoecy and dioecy are quite different, as dioecy prevents selfing absolutely, while monoecy merely prevents pollination within the flower, but cannot prevent an individual to be self-fertilized.

Thus, just as plants have evolved as dioecious, monoecious plants may have originated from hermaphrodite plants by the suppression of male function in some flowers and the suppression of the female function in other flowers; however, this should have occurred in the same plant and have not been allocated to different plants as in the case of dioecy (Richards 1997).

Monoecious plants may also have been derived from dioecious plants, following the reverse path of dioecy (Fig. 1.25); however, this system must be considered with the onset of the female function in male plants, and vice versa for individuals in the same population, and in the end, these flowers had separate sexes in the same plant. However, it seems very unlikely that this evolutionary path has occurred.

The evolution of monoecious plants, from plants presenting heterostyly, seemed to be an easier and more likely mechanism, because this would involve the same steps discussed in the evolution of dioecy (Fig. 1.26); however, instead of unisexual flowers being placed in different individuals, these would be allocated in different parts of the same individual.

1.5.3 Evolution of Self-Incompatibility Systems

The origin and maintenance of the self-incompatibility systems are quite complex and there are still many questions surrounding its evolution that remain unanswered. It is believed that it arose several times during evolution, and to assure the establishment of the chemical system of self-incompatibility, the occurrence of strong inbreeding depression is necessary in individuals from self-compatible populations. Moreover, plants completely self-incompatible would not be established immediately in the population of self-compatible plants. It is believed that initially intermediate levels in the population would have been established, with compatible plants being among self-incompatible plants. Because the generated individuals did not show inbreeding depression, the self-incompatibility alleles would have a reproductive advantage by increasing its frequency in the population as to completely suppress the self-compatibility alleles. Another fact to consider is that alleles responsible for self-incompatibility do not increase in frequency if there is no inbreeding depression, because in the absence of depression both types of alleles would have the same reproductive advantage and, therefore, the individuals generated from self-compatible plants would not be eliminated (Clark and Kao 1994).

Thus, the reproductive advantage can be defined by the balance of two forces: one that controls the rejection of the pollen grain, prioritizing outcrossing, and that which acts in the opposite direction as to increase the frequency of progenies in environments where the presence of pollinators is low. So, the maintenance of self-incompatibility throughout generations will depend on the superiority of the progenies produced and the cost for reducing the number of individuals generated (Vallejo-Marín and Uyenoyama 2004).

1.5.4 Evolution of the Self-Fertilization System

The breakdown of self-incompatibility has occurred repeatedly during the evolution of angiosperms and provoked profound impacts on the genetic structure of populations (Stone 2002). The primary genetic cost of inbreeding is the deleterious effect of depression; however, it is not constant and varies depending on the level of self-fertilization. Allogamous populations that practice self-fertilization suffer the effect of depression generated by exposure of deleterious or partly deleterious recessive alleles until they are completely eliminated from the population. From this moment on, the population is prepared to continue to evolve, using the system of self-fertilization without having recurring detrimental effects.

Considering a self-incompatibility system associated with the presence of an ancestral clonality, it was found that the transition from self-incompatible clonal reproduction (SI C) to self-incompatible non-clonal reproduction (SI NC) rarely occurs, although the reverse is common (Fig. 1.27). However, the transition from SI C to self-compatible clonal reproduction (SC C) commonly occurs and is irreversible. Similarly, the SI NC system undergoes transition to autocompatible non-clonal reproduction (SC NC) is irreversibly. In contrast, the transition between SC C to SC NC occurs frequently, while the reverse varies according to environmental conditions (Vallejo-Marín and O’Brien 2007).

Fig. 1.27

Evolution of the self-incompatibility system from the self-compatible ancestral (Drawn by MMG Karasawa based on Vallejo-Marín and O’Brien 2007)

1.5.5 Evolution of the Mixed Mating System

The mixed mating system is common in higher plants (Ingvarsson 2002) and corresponds to the simultaneous occurrence of self-fertilization and cross-fertilization. Currently, there is strong evidence that it is generated mainly by high inbreeding depression (Goodwillie et al. 2005). The primary genetic cost of inbreeding is the deleterious effect of depression, however it is not constant and varies depending on the level of self-fertilization. In a condition where the inbreeding depression condition does not vary and its fluctuation occurs in a stochastic way between generations, with an average of approximately 0.5, self-fertilization is not necessarily selected. As a result of this fluctuation, inbreeding depression can be seen as an additional cost of self-fertilization that can be stabilized in a mixed mating system (Cheptou and Schoen 2002). The substantial frequency of species with intermediate crossing rates provides evidence that the mixed mating system can be a stable strategy; yet there is no measurement of this frequency, so that it cannot be stated that this mode of reproduction is stable or just a transition phase. For this, theoretical studies are needed with a larger number of taxa for greater precision in the conclusions (Goodwillie et al. 2005).

1.6 Evolutionary Implications

Along its evolution, plants have adapted to different forms of sexual and asexual reproduction. The knowledge of the mode of reproduction of the species is important because it has great effect on the colonization of different habitats and also in response to environmental changes.

To understand the evolutionary significance of the different mating systems, consider that in a population three different mutations arise that do not affect fertility or survival of the species, but the association of these may provide evolutionary advantage in the adaptation of the species. Assuming initially that the population presents asexual reproduction (every individual produced has the genotype identical to the mother), they would have much difficulty to gather the different mutations if they occur in different individuals, unless individuals already possessing a mutation would acquire new mutations. Therefore, it would require a long time until all individuals in the population would present all three mutations. Moreover, populations of sexual reproduction would quickly gather the different mutations that occur in different individuals by cross-fertilization, through the exchange of alleles between individuals within a short time, thus benefiting from faster selective advantage. On the other hand, we must consider that the population size also plays a fundamental role in the dynamics of dispersal of each new mutation, because if populations are too small mutations are not likely to be fixed and dispersed among the individuals of the population, and are usually lost by drift on both sexual and asexual forms (Crow and Kimura 1965; Hartl and Clark 1997).

The impact of the content and distribution of genetic variation within and among populations may represent an important role in the distribution of different characters, determining the extent and pattern of responses to natural selection. The reproductive system has shown that it plays a prominent role in this respect. Autogamous populations, due to the high level of homozygosity, have no potential variability within populations, because the mutations that arise are eliminated faster than in allogamous populations, which may limit their ability to respond to environmental changes. In general, it is expected that autogamous and asexual species have short lifespan (Holsinger 2000), because the offspring may not survive to reproduce, thus producing discontinuity in seed production (Herlihy and Eckert 2002). The size of the population also has an important effect on the diversity and on genetic risks of extinction, because small populations are more likely to lose alleles important for adaptation by drift, and more likely to cross between related individuals suffering the deleterious effects of inbreeding, increasing the risk of extinction as the levels of inbreeding increase (Blisma et al. 2000).

Moreover, it is important to consider that in nature rarely pure breeding systems are found (i.e., plants with a single reproductive system). A study evaluating the correlated evolution of self-incompatibility and clonal reproduction in Solanaceae performed by Vallejo-Marín and O’Brien (2007) found that there is a strong association between these forms of reproduction, and that all self-incompatible species of Solanum present clonal reproduction, supporting the hypothesis that clonal reproduction promotes reproductive success in the evolution of reproductive strategies in plants. According to the results clonality leads to benefits in colonizing species such as the Solanaceae, supporting the persistence of self-incompatible genotypes in the case of inhospitable environments where the presence of the pollinator is rare and/or the level of incompatibility is high reducing the sexual outcrossing progeny produced. On the other hand, clonality generates the aggregation of similar genotypes and this can lead to breakage of self-incompatibility system throughout evolution. The resolution of this paradox between clonality × evolution of self-incompatibility (Fig. 1.28) lies in the degree to which clonal propagation would be compensating or limiting seed reproduction, and the extent to which clonality would reduce pollen flow between established genotypes, which would be affected by clonal architecture, plant density and the type and presence of pollinator (Vallejo-Marín 2007).

Fig. 1.28

Paradox: clonality × self-incompatibility (Drawn by MMG Karasawa based on Vallejo-Marín 2007)

The effects of the pollinating agent have also proven to be effective in changing the distribution of sexual reproductive systems in nature. It has been found that anemophilous species (i.e., wind pollinated species) present a bimodal distribution, that is, either autogamous or allogamous modes of reproduction, with little or rare intermediate types. Moreover, animals pollinated species present a continuous distribution between the two types of reproductive system, ranging from autogamous forms to extreme allogamous forms, with all degrees of self-fertilization and outcrossing in the middle (Vogler and Kalisz 2001).

Finally, it has been found that modular evolution of cells present in the female gametophyte has led to an increased ploidy in the produced endosperm. And that this mechanism is evolutionarily beneficial and stable, presenting the following consequences: increased level of heterozygosity in the endosperm, reduced genomic conflict (increase in the maternal/paternal relation) and increase in the range of observed phenotypes. This increase in ploidy level leads to a higher heterozygosity level that, in turn, would have an effect on the nutrition and vigor of the formed embryo. Furthermore, embryos have been more vigorous when the polyploid endosperm was generated from allogamous crosses among unrelated individuals. Thus, it is believed that selection should favor the evolution of individuals presenting endosperm containing higher ploidy levels (Friedman et al. 2008).