Reproduction in Bryophytes

Abstract

Liverworts, mosses and hornworts are together called bryophytes, terrestrial plants with perennial/photosynthetically dominant gametophytes and ephemerous/dependent sporophytes. These plants are classified, according to their sexual system, as monoicous and dioicous, with a few lesser categories recognized by bryologists. Sexual systems in bryophytes are apparently related to breeding systems, with ‘out-crossing’ occurring in dioicous species and self-fertilization in monoicous ones. Dioicous species are associated with low frequency of fertilization and rarity of sporophytes, which can be caused by biased sex ratios, spatial separation of sexes, and absence of males or failure of males to express sex among populations. After fertilization and sporophyte development, bryophytes produce spores that give rise to new plants. However, other forms of reproduction are present among bryophytes, with asexual structures as gemmae, propagules and regeneration of fragments that are able to form new plants. In the three lineages of bryophytes evolutionary processes resulted in reduction of gametophytic and sporophytic traits, reverse transitions of sexual systems (i.e., monoicy to dioicy), “abnormal” ploidy shifts (through apospory and apogamy), precocious development and maintenance of juvenile traits in the adulthood (e.g., intracapsular germination and neoteny). Despite the impressive diversity of reproductive modes in bryophytes, these plants are still poorly known, and only recently have attracted more attention from botanists. Studies related to reproduction in bryophytes increased progressively from 1987, quickly reaching around 20 papers per year in 2001 and a maximum of 25 papers in 2011 in the “Web of Science” database. Many species and ecosystems remain unexplored regarding the bryophytes and their reproductive biology, especially in the tropics, claiming for natural history studies that will identify and characterize interesting systems for research. Moreover, future studies with focus on evolutionary biology, biogeography and functional ecology will promote a comparative framework of the reproductive patterns and processes among all plants and not only bryophytes.

Introduction

Liverworts (Marchantiophyta), mosses (Bryophyta) and hornworts (Anthocerotophyta), commonly known as bryophytes, are represented by approximately 18,700 species in the world. Bryophytes are terrestrial atracheophyte plants, whose life cycle has alternation of generations (Fig. 1A–F), with a green and perennial gametophyte and an ephemerous dependent sporophyte (Schofield 1985; Vanderpoorten and Goffi net 2009; Goffi net and Buck 2013).

Gametophytes are leafy (in all mosses—Fig. 2A–F, and the majority of liverworts—Fig. 2G–I) or thallose (in some liverworts—Fig. 2J–L, and all hornworts—Fig. 2M–O). Gametangia are composed by one or more sterile cell layers forming the wall, and inside contain gametes. Antheridia (male gametangia—Fig. 1B) have numerous antherozoids or sperm (motile male gametes), and archegonia (female gametangia—Fig. 1B) contain one single oosphere or egg (not motile female gamete). Fertilization (Fig. 1C) can occur when an antherozoid reaches, mediated by water, an oosphere. After that, a zygote develops into an embryo inside the archegonium. Embryo gives rise to a sporophyte with a foot, seta or stalk (except in hornworts) and capsule (sporangium, Fig. 1D).

Distinct from all terrestrial plants, bryophytes produces a single sporangium in each sporophyte (Singer 2010). Liverwort capsules generally open through four longitudinal slits or valves; hornworts by one or two longitudinal slits; and the majority of mosses have capsules opening through a transversal line resulting from the operculum detachment (Crandall-Stotler et al. 2009; Goffi net et al. 2009; Renzaglia et al. 2009; Goffi net and Buck 2013).

Inside a capsule, mother cells of spores (sporocytes) split out meiotically, generally originating tetrads of haploid spores. After maturation, spores are released from the capsule and dispersed usually by wind (Fig. 1E). In mosses, after the spore germination a fi lamentous phase called protonema is produced. The moss protonema differentiates into chloronema (cells with numerous chloroplasts and right transverse walls (Fig. 1F)), caulonema (cells with needle-shaped chloroplasts and oblique transverse walls) and rhizoids (brownish cells with no chloroplasts and oblique transverse walls). Leafy buds are generally formed on the caulonema and give rise to many genetically identical leafy gametophytes (Nishida 1978; Nehira 1983; Duckett et al. 1998). Among liverworts and hornworts, the protonema is much more ephemerous, being restricted to a short-fi lamentous, globose or cylindrical phase, which commonly develop into a single plant (Nehira 1983; Goffi net and Buck 2013).

In leafy species the gametophyte has rhizoids, caulid (stem), and phyllids (leaves). Rhizoids attach the gametophyte to the substrate; are unicellular and fi lamentous structures, generally hyaline in liverworts and hornworts; and multicellular, branched and brown-colored in mosses. Caulid is a vertically or horizontally growing axis, with an undifferentiated, very simple or complex anatomy. Respective of species, caulid may contain epidermis, cortex and central cylinder. In the central cylinder of some mosses, especially in the family Polytrichaceae, there are cells specialized in water and mineral conduction, the hydroids (tissue hydrom), and in photosynthates conduction, the leptoids (tissue leptom). In thalloid species, gametophyte has rhizoids attached to a flattened thallus with one or multiple cell layers, dichotomically branched or with a rosette-shape

(Crandall-Stotler et al. 2009; Goffi net et al. 2009; Renzaglia et al. 2009; Goffi net and Buck 2013).

Phyllids are attached to the caulid, and are generally green, small and one-cell layered. In mosses, phyllids are commonly helicoidally arranged, and have a thick central area (with more than one cell layer) similar to the central vein in leaves of other plants, called costa. Conversely, in liverworts, phyllids are arranged in three lines, or rarely in two or four, but do not have a true costa (Crandall-Stotler et al. 2009; Goffi net et al. 2009; Goffi net and Buck 2013).

Leafy bryophytes have gametangia at tips, axils or intercalary in branches or caulid. It is common to find modified phyllids covering gametangia and forming sexual branches. Terms used to classify the sexual branches among bryophytes can vary in the literature, and it is common to find perigonium (♂) and perichaetium (♀) for mosses, and androecium (♂) and gynoecium (♀) for leafy liverworts (Gradstein et al. 2001). Moreover, sterile structures, known as paraphyses, probably protect mosses’ gametangia against mechanical damages and dehydration, and also have a role in the dehiscence of the antheridia and secretion of substances to attract microarthropods that disperse sperm cells (Cronberg et al. 2006; Crandall-Stotler et al. 2009; Goffi net et al. 2009; Goffi net and Buck 2013). In hornworts gametangia are located inside the thallus; and complex thalloid liverworts have special receptacles or specialized branches, antheridiophores and archegoniophores, which contain the gametangia. In the thallose liverworts Ricciaceae, gametangia and sporophytes are formed and are totally embedded inside the thallus, capsules are cleistocarpous and release spores only after tissue decaying (Crandall-Stotler et al. 2009; Renzaglia et al. 2009).

In acrocarpous mosses, which comprise nearly half the species in the phylum, perichaetia are produced at the apex of the stem (terminal perichaetia), whereas in pleurocarpous species the perichaetia are located on lateral buds or short side branches, without vegetative leaves. There is a third condition where the perichaetia are terminal, although formed on lateral branches with vegetative leaves (cladocarpous mosses) (La Farge-England 1996). Traditionally, acrocarpy, pleurocarpy and cladocarpy were recognized as growth forms among the mosses. Nowadays, these terms are used to indicate perichaetium, and consequently, the sporophyte position in the gametophyte (La Farge-England 1996). Although the direction of growth in a moss is generally a cue for the growth form classification, it does not correspond to the perichaetial position, and some acrocarpous mosses grow horizontally and some pleurocarpous grow vertically (Glime 2007).

Retention and development of an embryo that is attached to mother plant is shared by bryophytes and all land plants, a reason for the clade name Embryophytes. A multicellular matrotrophic embryo depends on maternal tissues nutritionally and developmentally for at least some time during its early development (Graham 1996). Embryo nutrition occurs through transfer cells at the intersection between gametophyte and sporophyte, forming a tissue known as placenta. The embryo, and posteriorly the sporophyte, receive water, minerals and organic substances necessary for development from the gametophyte through the placenta (Goffi net et al. 2009; Vanderpoorten and Goffi net 2009). Moreover, in mosses a maternal calyptra (a remnant of the archegonial neck), with a multilayered cuticle, has a role of dehydration protection of the immature sporophyte and is able to increase the offspring fitness. The moss calyptra is the most ancient form of maternal protection by a cuticle in green plants (Budke et al. 2013).

Sexual Systems in Bryophytes and their Consequences

The sexual systems in bryophytes are classified into two main groups, monoicy and dioicy, each one with a few categories that are usually recognized by bryologists (Anderson 1980; Wyatt 1985; Mishler 1988; Table 1 and Fig. 3).

In bryophytes, unlike ferns and angiosperms, species having unisexual plants (the photosynthetically dominant phase) are overrepresented compared to bisexual ones. Among liverworts, about 70% of species are dioicous; in mosses dioicy reaches 55–60%; and in hornworts monoicy predominates among species (Wyatt 1982, 1994; Vanderpoorten and Goffi net 2009). On the other hand, ca. 6% of angiosperm species are dioecious (Renner and Ricklefs 1995).

Wyatt (1985) highlighted the importance for a unified terminology of sexual systems in plants, attributing to the lack of concordant terms in bryophytes and other embryophytes. However, there are good reasons to prefer terms like monoicous/dioicous over monoecious/dioecious among bryophytes. The sexual system in bryophytes describes the sexuality of the gametophyte while in other embryophytes (e.g., seed plants) it describes where and how unisexual gametophytes are borne on the sporophyte.

All seed plants have dioicous gametophytes. The terms monoecious and dioecious (sporophyte having one or two unisexual (dioicous) gametophytes) are meaningless for bryophytes since the sporophytes among these plants do not bear gametophytes (Allen and Magill 1987). Moreover, bryophytes are homosporous, whereas seed plants are heterosporous. In homosporous plants products of meiotic divisions (spores) give rise to egg-bearing and sperm-bearing gametophytes (dioicous) or gametophytes with both sex cells (monoicous). In heterosporous plants, after meiosis, gametophytes give rise to only one type of sexual cell (Zander 1984; Allen and Magill 1987). Therefore, the breeding systems in bryophytes and seed plants are differently influenced by the sexual systems, i.e., self-fertilization in bryophytes and in seed plants have different evolutionary consequences (Crawford et al. 2009). In the monoicous system, egg and sperm are originated from mitotic divisions in the same gametophyte, being genetically identical cells. Conversely, in the monoecious system egg and sperm cells are formed on different gametophytes (independent meiotic events in the sporophyte) and are genetically variable (Allen and Magill 1987).

Sexual and breeding systems in bryophytes are probably related, with ‘out-crossing’ in dioicous species and self-fertilization in monoicous ones (Gemmell 1950). Several records of self-compatibility are currently known among monoicous bryophytes, but recent studies suggest that self-incompatibility (intra-gametophytic level) also occurs in this group (Stark and Brinda 2013). According to Longton and Schuster (1983) the sub-divisions within the monoicy in bryophytes may reflect selection for out-crossing, from sinoicy (high selfing likelihood) to rhizautoicy (high crossing likelihood).

There are other mechanisms to prevent the self-fertilization among monoicous bryophytes, such as protandry (maturation of antheridia before archegonia on the same plant) and protogyny (maturation of archegonia before antheridia). Apparently, in most species of mosses, antheridia have earlier and longer maturation than the archegonia (Gemmell 1952; Longton and Schuster 1983; Stark 2002).

Sporophyte Production, Sexual Systems and Sex Ratios

Although the dioicy promotes out-crossing, this is balanced by a decrease in the sporophyte production (Mishler 1988; Crawford et al. 2009). Therefore, sporophyte production in bryophytes seems to be directly linked to the sexual system, since the rarity or absence of sporophytes has been mostly detected among dioicous species compared to monoicous ones (Longton 1976, 1992; Longton and Schuster 1983; Reese 1984; Laaka-Lindberg et al. 2000; Oliveira and Pôrto 1998, 2002; Pôrto and Oliveira 2002; Maciel-Silva et al. 2012a).

For instance, a study on British moss-flora (Gemmell 1950) showed a significant low frequency of sporophyte-bearing dioicous species, mostly explained by separation of the sexes and consequent mechanical difficulties in the transport of sperm to archegonia. The same study reported a higher frequency of successful sporophyte production among monoicous species, possibly due to self-fertilization, even though dioicous species were more widely distributed than monoicous ones. Still among British mosses, 87% of species with unknown sporophytes were dioicous and 83% of the monoicous species had sporophytes (Longton 1997). The other two studies on British moss (Longton 1992) and liverwort (Laaka-Lindberg et al. 2000) flora compared the life history traits (e.g., monoicy vs. dioicy; presence vs absence of sporophytes) of rare and common species and detected that 1) many rare species had no sporophytes compared to common species (consequences of sexual reproduction), 2) many monoicous species were rare compared to dioicous ones (probably adverse consequences of self-fertilization).

Among bryophytes in the Brazilian Atlantic rainforest, dioicous species that produced high numbers of sexual branches and gametangia per sexual branch (i.e., high reproductive performance) failed to produce sporophytes, probably due to the spatial separation of sexes (Maciel-Silva et al. 2012a). Spatial separation of colonies having different sexes is commonly verified among dioicous species of bryophytes, and occurs when a diaspore (spore, gemma or propagule) reaches a substrate, germinates and plants grow clonally forming a colony (Stark et al. 2005).

Dioicous bryophytes have obligatory cross-fertilization that results in high genetic variability compared to monoicous ones. However, this advantage is counter balanced with a low sporophyte production commonly recorded among dioicous species (Eppley et al. 2007). Main reasons for a low sporophyte production among dioicous species are related to spatial separation of female and male colonies (and populations), lower number of males than females, and absence of male plants (or failure to express sex among male plants) (Gemmell 1950; Longton and Schuster 1983; Bowker et al. 2000; Oliveira and Pôrto 2002; Stark et al. 2005, 2010).

Low number of male plants compared to female ones is commonly recorded in populations of dioicous bryophytes and results in a pattern recognized as female biased sex-ratios (Bowker et al. 2000; Bisang and Hedenäs 2005; Bisang et al. 2006; Stark et al. 2010). The real causes for this phenomenon remain unclear, but some hypotheses have been suggested to explain why the initial sex-ratio 1:1 verified among the spores (Stark et al. 2010) usually change during the gametophyte growth and maturation. Explanations are possibly related to sex-differential spore abortion, germination, protonema growth, and gametophore development and mortality in response to competition and/or abiotic effects (for a more detailed discussion see Stark et al. 2010, and Bisang and Hedenäs 2013).

Some studies on female-biased sex-ratio in bryophytes have begun to elucidate this biological puzzle. Constraints on one of the two sexes may begin during spore development, with spore abortion resulting in female and male biased sex-ratios. For instance sex-ratios of germinating spores of 4.1♀:1♂ in the moss Mnium undulatum and 0.89♀:1♂ in M. hornum (Newton 1972), and 1.5♀:1♂ in Ceratodon purpureus (Shaw and Gaughan 1993). In some species, biased sex-ratios begin after spore germination. Liverwort Sphaerocarpos texanus had high abundance of females in the field and growth chamber (cultures from 1:1 male to female spores), suggesting lower survival rate in males than in females (McLetchie 1992). In the laboratory, moss Bryum argenteum had unbiased (or slightly biased) sporeling sex-ratios, whilst a female-biased gametophyte sex-ratio predominated in the field (Stark et al. 2010).

Failure to express sex among male plants has been suggested to explain the female skewed ratio verified in bryophyte populations (Longton 1990). A hypothesis called “the shy male hypothesis” presumes that males and females have similar frequencies, but males produce sexual structures less often than females (Mishler and Oliver 1991; Stark et al. 1998, 2001, 2005), having a “reservoir of male plants in the non-expressing state” (Stark et al 2010). Male plants are generally rare (or do not express sex) in high-stress habitats like deserts (Bowker et al. 2000; Stark et al. 2010). However, recently the moss Drepanocladus lycopodioides was recorded having both genetically determined and phenotypically expressed sex-ratios being female-biased, which indicate that male plants compared to female ones do not necessarily fail to express sex. Sex-ratios were, indeed, female biased among sex-expressing and non-expressing plants (Bisang and Hedenäs 2013).

The spatial separation of males and females among dioicous bryophytes creates a potential obstacle for fertilization. Bryophyte sperm moves to the female archegonium through a water-fi lm, implying in a maximum distance for fertilization of no more than a few centimeters (Longton and Schuster 1983; Crum 2001). However, some species have developed strategies to increase the chances of fertilization over larger distances. Mosses in the genus Plagiomnium and in the family Polytrichaceae have splash cups with the antheridia exposed to rain action (Rohrer 1982; Andersson 2002); the liverwort Conocephalum conicum, after contact with water, produce airborne sperm that can travel long distances (Shimamura et al. 2008; Springer Videos 2010). Furthermore, microarthropods have been reported to mediate fertilization through sperm dispersal in the moss Bryum argenteum (Cronberg et al. 2006). Recently, Rosenstiel et al. (2012) demonstrated that sperm-dispersing microarthropods are guided by scents (volatile cue substances) emitted from fertile shoots of the moss Ceratodon purpureus, suggesting a fertilization syndrome somewhat comparable to insect pollination of flowering plants (Cronberg 2012; Rosenstiel et al. 2012).

Dioicous species seem to compensate their frequent failure to produce sporophytes through the asexual propagation for the species maintenance (Longton and Schuster 1983; Longton 1992, 2006). For instance, Une (1986) observed that 77.5% of Japanese dioicous species had asexual structures, which are found in only 9.9% of the monoicous species. Recently, Pôrto and Silva (2012) found 54% dioicous vs. 46% monoicous species in northeastern Brazilian Atlantic rainforest, where monoicous mosses produced sporophytes more frequently than dioicous ones (ca. 70% and 30%, respectively), and dioicous mosses and liverworts had large frequency of asexual structures compared to monoicous ones (ca. 35% and 15% for liverworts and 20% and 0.5% for mosses, respectively). Conversely, Crawford et al. (2009) analyzed life history data for 367 species of mosses and found no phylogenetic support for a positive correlation between asexual reproduction and dioicous condition. This relationship emerged only when species were treated as independent data points, indicating that asexual reproduction may be not an adaptation to dioicous condition, but a trait shared among relative species. Moreover, dioicous species were generally larger than monoicous ones, suggesting a compensation of the low production of sporophytes and spores by increased size and life-span (Crawford et al. 2009).

The Largest Diversity of Asexual Structures Among all the Plants

Asexual reproduction sensu lato is so common in bryophytes that it has been used as a diagnostic character for some taxa (e.g., genera within Lejeuneaceae, Bastos 2008). In no other plant group the asexual reproduction is as important as in bryophytes (Frey and Kürschner 2011), for instance, it is estimated that about 15% and 17% of North-American and British mosses, respectively, produce at least one type of specialized asexual structure (Longton and Schuster 1983; Glime 2007). If these asexual structures have small sizes, being dispersed over long distances, they may contribute considerably to the gene flow at both local and landscape scales (Pohjamo et al. 2006).

Among bryophytes the asexual reproduction s.l. is commonly divided in two basic types: (1) asexual reproduction sensu stricto and (2) clonal reproduction (cloning), including body fragmentation (Frey and Kürschner 2011). Therefore, asexual structures may be completely specialized or not. If they are able to detach from the parent shoot, being spatially or temporally dispersed, these asexual structures are recognized as asexual diaspores

(During 2001). Many bryophyte species, especially dioicous ones, reproduce asexually exclusively, by regeneration from more or less specialized caducous structures (leaves, leaf tips, shoots, branches, and bulbils) and by the production of specialized asexual structures (gemmae, protonemal gemmae, and rhizoidal gemmae; Frey and Kürschner 2011 and references therein). Based on different studies, the main types of asexual structures known in bryophytes are summarized below (Longton and Schuster 1983; Imura 1994; Glime 2007; Frey and Kürschner 2011). For a detailed account of the diversity of asexual structures among bryophytes, see a recent review by Frey and Kürschner (2011).

Propagules

Propagules have a differentiated apical cell, and sometimes show also leaves and rhizoids. They comprise caducous tips of branches, branches and leaves, flagella form branches and bulbils. Propagules differ from gemmae due to the apical cell, which originates a new shoot with no protonemal stage (Goffi net and Buck 2013).

Caducous shoot apices

These structures are little modified shoot tips, commonly deciduous along an abscission line. After the detachment from the mother plant, they may grow into a whole plant with rhizoidal development from the basal part, e.g., Campylopus spp., Bryum argenteum (Imura 1994; Frey and Kürshner 2011).

Caducous branches and branchlets (Cladia)

Caducous branches and branchlets are asexual propagules with normal size and reduced leaves, respectively. They are found in leaf axils at branch tips and have an abscission line. Cladia are generally recognized as small branches with reduced leaves, e.g., Dicranum flagellare, D. scoparium, Platygyrium repens, Pseudoeskeella nervosa, Cheilolejeunea oncophylla, Lejeunea laetevirens, L. cancellata, L. cardoti, and Microlejeunea epiphylla (Schuster 1983; Bastos 2008; Frey and Kürshner 2011).

Caducous leaves

Normal-size vegetative leaves and specialized diminutive leaves (brood leaves), which detach from the parent shoot and sometimes have rhizoids or young plants on borders. When present, brood leaves are frequently clustered on an axis, e.g., Hypopterygium didictyon, Pleurochaete squarrosa, Aulacomnium androgynum, Syntrichia laevipila, Campylopus fragilis, Dicranodontium longirostre, Dicranum montanum, Bazzania nudicaulis, Ceratolejeunea caducifolia, Cheilolejeunea adnata, C. decidua, Drepanolejeunea propagulífera, Lejeunea phyllobola, Rectolejeunea berteroana, R. emarginuliflora and Plagiochila corniculata (Giordano et al. 1996; Bastos 2008; Frey and

Kürshner 2011).

Fragments of leaves (and thallus)

Pieces of leaves, commonly tips and edges, where the leaf often breaks off, e.g., Campylopus fragilis, Dicranum viride, D. tauricum, Tortella fragilis, Acrobolbus ciliatus, Frullania microphylla, Lejeunea elliottii, and Plagiochila caduciloba. In the hornwort Nothoceros aenigmaticus (previously Megaceros aenigmaticus), thallus fragmentation supports the growth of geographically isolated male and female populations (Bastos 2008; Renzaglia et al. 2009; Vanderpoorten and Goffi net 2009; Frey and Kürshner 2011).

Bulbils

Small and multicellular, fi lamentous or spherical, caducous, budlike to thread-like branches, with reduced leaves, with short stalk, growing in leaf axils. They occur from one to several per leaf, e.g., Bryum spp., Leptobryum pyriforme, Pohlia spp. (Imura 1994; Frey and Kürshner 2011).

Gemmae

Gemmae, different from propagules, are asexual structures without an apical cell and germinate following a pattern that recapitulates the ontogeny of the whole plant. They are fi lamentous, thallose or globose structures, varying from one to usually many cells. Gemmae frequently have a stalk and are developed on different parts of the gametophyte (on leaf, thalli, on stem rhizoids, in leaf axils, endogenous, on protonema, and on specialized non-deciduous gemmiferous shoots, or thallose gemmae-bearing cups).

They are common in liverworts and mosses, e.g., Blasia pusilla, Anastrophyllum hellerianum, Cavicularia densa, Lunularia cruciata, Marchantia spp., Metzgeria spp., Calymperes spp., Syrrhopodon spp., Dicranum flagellare and Tetraphis pellucida (Cavers 1903; Kimmerer 1994; Bartholomew-Began and Jones 2005; Pohjamo et al. 2006; Jones and Bartholomew-Began 2007; Frey and Kürshner 2011; Goffi net and Buck 2013).

Endogenous gemmae

They are produced inside an initial cell on the basal lamina or on the ventral side of the costa in mosses, e.g., Grimmia torquata, G. trichophylla and Racomitrium vulcanicola. In liverworts, ovoid or ellipsoidal 2-celled endogenous gemmae occur at leaf tips or margins, e.g., Bazzania kokawana, Fossombroniaceae, Endogemma caespiticia and Riccardia spp. (Glime 2007; Frey and Kürshner 2011; Goffi net and Buck 2013).

Protonemal gemmae

One to a few (12)-celled gemmae on chloronemal filaments, with abscission mechanisms consisting in a thin-walled abscission (tmema-) cells that break easily, releasing the gemmae. Protonemal gemmae occur mostly in acrocarpous, but also in pleurocarpous mosses, e.g., Diphyscium foliosum, Dicranella heteromalla, Dicranoweisia cirrata, Dicranum montanum, D. tauricum, Tortula muralis, Orthodontium lineare, Schistostega pennata, Rhizomnium punctatum, Zygodon spp., Orthotrichum spp., Isopterygium elegans, Bryum spp., Mittenia spp., Orthotrichum obtusifolium, Oxyrrhynchium hians, Pseudotaxiphyllum elegans and Lepidopilum muelleri. Most records of protonemal gemmae among mosses are from cultivation in laboratory. The role of these gemmae is apparently to increase the initial establishment, especially when sporophytes are rare (Duckett and Ligrone 1992; Duckett and Matcham 1995; Maciel da Silva et al. 2006; Pressel et al. 2007; Frey and Kürshner 2011).

Tubers (and Rhizoidal gemmae)

Tubers are spherical to ellipsoidal or pyriform structures, thick-walled and with an apical cell. They are usually reddish to dark brown and subterranean, vary from 10 to more than one hundred cells and are attached to rhizoids. Tubers are resistant to drought and can store many substances like lipids and proteins. Dark color suggests anti-herbivory compounds or a filter that avoids germination under high light conditions (Imura 1994; Duckett and Ligrone 1992; Frey and Kürshner 2011).

In mosses, two germination modes are recorded from rhizoidal tubers: (1) tubers develop directly into a leafy shoot (without protonemal phase), when the apical cell is reactivated or (2) tubers produce secondary protonemata and form moss plants indirectly, when the apical cell is not reactivated. These asexual structures are common among acrocarpous mosses (e.g., Bryaceae, Dicranaceae, Ditrichaceae, Fissidentaceae, Pottiaceae, and Grimmia pulvinata), and are absent among pleurocarpous mosses (Risse 1987; Duckett and Ligrone 1992; Frey and Kürshner 2011).

Differences between rhizoidal tubers and rhizoidal gemmae are not clear in the bryological literature, and both terms are commonly used as synonyms. However, rhizoidal gemmae seem to lack an apical cell, and germinate into leafy shoots passing through a secondary protonema, e.g., Dicranella staphylina, D. schreberana and Leptobryum pyriforme (Risse 1987; Imura 1994; Duckett and Ligrone 1992). Both rhizoidal tubers and gemmae are common components of the asexual diaspore bank of bryophytes, surviving unfavorable conditions (During 2001).

In liverworts, tubers occur in Fossombronia spp., Petalophyllum spp. and Sewardiella spp. (Fossombroniales), Geothallus tuberosus spp. (Sphaerocarpales) and Riccia spp. (Ricciaceae), and are associated with adaptations to arid environments. In hornworts, they are common in Phaeoceros spp., and develop on apical parts of the thallus mostly during the dry season (Longton and Schuster 1983; Risse 1987; Imura 1994; Duckett and Ligrone 1992; Glime 2007; Frey and Kürshner 2011).

Some Reproductive Patterns and their Morphological and Evolutionary Consequences

In the three lineages of bryophytes several evolutionary processes resulted in reduction or simplification of gametophytic and sporophytic traits, reverse transitions of sexual systems, “abnormal” ploidy shifts and precocious development. Some of these processes may confer advantages during vegetative growth, reproduction and dispersal to short and long distances relative to habitats where species live and their life histories. Below, we present some of these patterns, which are directly or indirectly linked to reproduction in bryophytes.

Dioicy vs. Monoicy

The dioicy among bryophytes is considered a plesiomorphic character that is also present in the Charales and Coleochaetales algae, which are closely related to the embryophytes (Longton and Schuster 1983; Mishler 1988). However, this character has a complex history within the group, with at least 133 transitions between monoicy and dioicy among mosses (McDaniel et al. 2012). Moreover, reversals to dioicy are twice as frequent as the transitions to monoicy within this group. A possible explanation for the recurrent evolution of dioicy among mosses may be the sexual specialization. Sexual dimorphism is commonly found among dioicous species (Une 1984, 1985; Stark et al. 2001; Fuselier and McLetchie 2002; Hedenäs and Bisang 2011; Pichonet and Gradstein 2012) and indicates that males and females may be subject to conflicting selective pressures. There are different optimal phenotypes for males, which release sperm, and for females, which nurture embryo and sporophyte until maturation, spore dispersal and senescence

(McDaniel 2005; McDaniel et al. 2012). Regarding the liverworts, Devos et al. (2011) found that, at least in the Radula-clade (a group of leaf liverworts), the monoicy is a recent evolutionary acquisition associated to epiphytism, occurring six times independently, without reversions.

The evolution of the pollination in seed plants gave rise to selective pressures that favor genetic recombination (Bawa 1980; Bawa and Beach 1981), but in bryophytes the dependence of water for fertilization generates differences of fertilization likelihood vs. genetic recombination for monoicous and dioicous species (Longton and Schuster 1983; Longton 1992; 1997; Eppley et al. 2007). High fertilization rates and sporophyte production are generally associated with sexual reproduction. Among dioicous species this assumption is always true since out-crossing (i.e., crossing between two different gametophytes) is involved, but among monoicous species the fertilization commonly involves female and male genotypes from the same gametophyte, i.e., auto-fertilization or intragametophytic selfing (Crawford et al. 2009). Thus, in practical terms the sexual reproduction in monoicous species may, indeed, be a type of asexual reproduction. However, these species do not seem to be affected by inbreeding depression, since they rapidly purge recessive deleterious mutations through intragametophytic selfing (Eppley et al. 2007; Taylor et al. 2007). Conversely, in dioicous species, intragametophytic selfing can lead to sporophytic inbreeding depression (Taylor et al. 2007). There are evolutionary benefits and disadvantages involved in the reproduction of dioicous and monoicous bryophytes, and selective forces appear to maintain ratios of dioicy to monoicy in an almost unbiased way (compared to angiosperms) among the bryophyte species.

Neoteny

Neoteny is a type of paedomorphosis that leads to a retardation of somatic development relative to the normal onset of sexually mature features, resulting in the persistence of juvenile or pre-adult physical characteristics into adulthood. In bryophytes, the neoteny is generally expressed in the persistence of the protonema, which is more common in mosses than in liverworts. Protonemal neoteny is usually interpreted as an adaptation to growth in ephemeral or unstable substrate or habitats, such as twigs, living leaves and disturbed soil, where rapid maturation and completion of the life cycle are very important for survival (Schuster 1988; Gradstein et al. 2006). In mosses it is known in species of Buxbaumia, Discelium, Ephemeropsis and Pogonatum. In liverworts it occurs in Metzgeriopsis, Protocephalozia ephemeroides and Radula aguirrei (note that this is an invalid name: www.tropicos.org) and R. yanoella. Other forms of neoteny, with no maintenance of the protonemal stage, are present in the liverworts Myriocoleopsis, Aphanolejeunea, Cololejeunea, Chondriolejeunea, Metzgeriopsis, Aphanotropis, Calatholejeunea, Colura, Diplasiolejeunea, Macrocolura and Myriocolea (Gradstein et al. 2006).

Life-history traits of liverworts, like monoicy, lobule inflation, imbricate leaves, cell wall ornamentation, presence of asexual propagules and neotenic features are suggested to be adaptations to epiphytism s.l. (including epiphylly—habitat on leaves; Schuster 1988; Gradstein et al. 2006). A recent study of epiphyllous liverworts has shown that just the last four traits are correlated with the common epiphylly in a cross-species comparison. In addition, after a phylogenetic comparison only the presence of asexual propagules emerged as a true adaptation for the epiphylly (Kraichak 2012). The high correlation between neoteny and epiphylly, in a ‘cross species’ analyses, is due to the sharing of neotenic features among closely related liverwort lineages (e.g., Tuyamaella-Cololejeunea clade), favoring the colonization of leaves. Therefore, the neoteny appears to be an exaptation instead of an adaptation to epiphylly (Kraichak 2012). However, the hypothesis of its evolution in response to epiphytism should not be rejected and needs more investigation.

Apospory and Apogamy

Apospory (development of a gametophyte from sporophyte tissue without meiosis, i.e., without spore formation) and apogamy (development of a sporophyte, usually from the vegetative cells of the gametophyte, without union of gametes or fertilization) have been rarely observed in nature (El- Saadawi et al. 2012), but have been induced experimentally with bryophytes in laboratory (Fig. 4; Chopra 1988; Chopra and Kumra 1988; Bell 1992; Cvetić et al. 2005; Goffi net et al. 2009).

Apospory is usually induced in vitro by factors as suitable temperature and light, sufficient humidity and lack of sugar in the medium. On the other hand, apogamy is favored by the opposite conditions such as low light intensity, increased sugar concentration in the medium, or growth regulators (e.g., indol acetic acid) at low concentrations (Hughes 1969;

Chopra 1988; Chopra and Kumra 1988; Cvetić et al. 2005). Multiplication of a filamentous protonema (mosses) or a thallus (liverworts and hornworts) from sporophytic tissue is recorded mostly at controlled conditions in the laboratory (Lang 1901; Matzke and Raudzens 1968; Goffi net and Buck 2013).

Apospory is one of three way through which autopolyploidization—the doubling of the genome without hybridization—occurs in bryophytes. The two others are diplospory and syndiplospory (for details see Rensing et al. 2013). After an apospory event, natural polyploids may be formed from reprogramming of vegetative sporophytic cells. Aposporous gametophytes, different from their haploid progenitors, may have larger gametangia and sexual cells, which are likely due to the duplicated genome. An apospory event commonly leads to a bisexual gametophyte with functional gametes, i.e., a bisexual gametophyte (Bell 1992; Rensing et al. 2013). Although the apospory is recognized like an abnormal process in the life cycle of bryophytes, it may be responsible for some events of diocy-monoicy transition at least in mosses, where this phenomenon is more common than in liverworts (Bell 1992).

Image 5

Strong correlation between polyploidy and monoicous condition in mosses suggest that changes in chromosome number may alter the sexual system (Crawford et al. 2009). In the moss Atrichum undulatum, gametophytes with combined sexes were diploid or triploid, whereas gametophytes with separate sexes were haploid, diploid or triploid, suggesting that polyploidy has a role in the evolution of monoicy, but not necessarily always resulting in monoicy (Jesson et al. 2011).

Regarding the apogamous sporophytes, these generally have morphological modifications compared to “normal” sporophytes, including absence of stomata. Some abnormalities of apogamous sporophytes are caused during sporophyte development without the protection of a calyptra, resulting in decrease or no production of offspring. Viability and germination of spores vary among species. When a sporophyte develops from the haplophase, spores are usually unable to germinate. Conversely, if sporophytes grow directly from sporophytic tissue (diplophase), spores are usually viable (Cvetić et al. 2005 and references therein). After spore germination, protonemata grow originating leafy gametophytes and, consequently, gametangia and viable gametes. Leafy gametophytes have large morphological variation, mostly due to mutations during sporogenesis (Chopra 1988; Chopra and Kumra 1988; Cvetić et al. 2005).

Gametangia Number

Some authors have suggested a decreasing pattern of gametangia number among bryophytes, in species from temperate to tropical zones and from basal to derived lineages. Non-tropical liverworts generally produce gynoecia (female branches) with a large number of archegonia (12–25 in Cephalozia and 25–50 in Haplomitrium), whilst in tropical liverworts, especially in the families Radulaceae, Jubulaceae and Lejeuneaceae, there is a strong trend to gametangia reduction, e.g., Lejeuneaceae are generally monogynous. Similarly, the number of antheridia tend to decrease from non-tropical taxa (4–6(–16) per bract in Haplomitrium, Herbertus and Schistochilaceae) to tropical families (with (1)2–3 in Radulaceae, 1 in Porellaceae, 1–2 in Lejeuneaceae and Jubulacaeae, e 1(1–2) or rarely more in Plagiochilaceae; Schuster 1988; Gradstein 1991). In the hornworts there is an evolutionary trend to decrease the antheridia number per antheridial chamber, e.g., from the maximum of 30–80 per chamber in the basal Leiosporoceros to one per chamber in Dendrocerotaceae (the most derived taxon in the hornworts) (Renzaglia et al. 2009). Moreover, the family Dendrocerotaceae, Megaceros, Nothoceros and Dendroceros are tropical genera. In future, a phylogenetic comparison should be used to analyze the real role of decreasing number of gametangia among bryophytes, and elucidate if this trait, indeed, is an adaptation to life in tropics.

Furthermore, the number of gametangia in bryophytes may influence fertilization chances and sporophyte output in each species, and apparently is associated to the sexual system. In general, for bryophytes, the ratio of male to female gametangia per sexual branch is considerably higher in the dioicous taxa (Une and Tateishi 1996; Glime 2007). Based on a literature survey and unpublished data, Maciel-Silva et al. (2012a) recorded a ‘male-biased’ sex ratio of gametangia per sexual branch in dioicous bryophytes (♂ 77.79% and ♀ 22.21%; ♂:♀ ratio = 5.28), and in some degree in monoicous ones (♂ 54.75% and ♀ 42.25%; ♂:♀ = 1.77). The markedly biased sex ratio of gametangia in dioicous species compared to monoicous ones suggests that dioicous species have strategies to increase out-crossing (Glime 2007; Maciel-Silva et al. 2012a).

Green Spores and Precocious Germination

Green (or chlorophyllous) spores are common among bryophytes and some ferns. Since spore color is frequently related to spore longevity, green spores lose viability shortly compared to non-green spores (e.g., within a single day to a few months; Lloyd and Klekowski 1970; Pence 2000; Wiklund and Rydin 2004; Maciel da Silva et al. 2009a). Yellow and brown spores can last longer because they are protected against desiccation by thick walls and store nutrients in internal oil droplets (Mogensen 1981; Renzaglia et al. 2009).

Green spores are commonly associated to tropical taxa (e.g., liverworts: all Lejeuneaceae, Radula, Plagiochila; mosses: Neckeropsis disticha, N. undulata, Octoblepharum albidum, Pyhhrobryum spiniforme, and Thamniopsis incurva; hornworts: Megaceros, Nothoceros, and Dendroceros) (Schuster 1988; Renzaglia et al. 2009; Maciel da Silva et al. 2009 a, b; Maciel-Silva et al. 2014). In the hornworts, green spores are considered an apomorphic character, compared to yellow spores, occurring in three genera of the family Dendrocerotaceae (Renzaglia et al. 2009). Investigations with ultrastructural and phylogenetic approaches would clarify about the adaptive role of green spores among tropical species of bryophytes.

Green spores have fast germination, which in some species occurs inside the capsule (intracapsular germination) or still inside the spore wall (endosporic germination). Intracapsular germination is commonly recorded in tropical liverworts (especially epiphytes, e.g., Lejeuneaceae, Frullania, Plagiochila, Porella, and Radula; Schuster 1988; Thiers 1988), but appears rare among mosses. Kürschner (2004) recorded an example of intracapsular germination from the tropical moss Brachymenium leptophyllum, suggesting an achorous-strategy of life in which “protonema dwells out of the capsule, forming immediately buds and juvenile gametophytes that establish near the mother plant ... new plants remain on the same phorophyte, increase the chance of re-establishment of the population at these favoured sites and lower the risk of extinction by long-range dispersal”.

Endosporic germination occurs after the first cell division of the spore, without the rupture of spore wall (exospore), and gives rise to a globose sporeling covered by intact exospore. This precocious germination is present in liverworts, mosses and hornworts (Nishida 1978; Schofield 1981; Nehira 1983; Thiers 1988; Schuette and Renzaglia 2010). Tropical and epiphytic hornworts of the genus Dendroceros have green multicellular spores (i.e., spores with precocious germination inside the capsule), which are large when dispersed and may have around 100 cells. Similarly, spores in the liverwort Pellia have precocious germination inside the capsule, reaching 23–24 cells. Mosses of the genus Andreaea also have endosporic germination, but it occurs only after dispersal and at suitable conditions for germination (Nishida 1978; Nehira 1983; Schuette and Renzaglia 2010). Green spores associated to fast and precocious germination (intracapsular or endosporic) are suggested as a strategy adopted among different bryophyte taxa for desiccation tolerance and growth on epiphytic habitats, since the early stage of the gametophytes (protonema) remain protected by the capsule or exospore (spore wall).

Reproduction in Bryophytes: Now and Hereafter

Studies relative to reproduction sensu lato in bryophytes, including sexual and asexual cycles; reproductive phenology; diaspore development, dispersal and establishment; diaspore banks; reproductive effort and costs; evolution of sexual systems and trade-offs among life-history traits, have increased in number mostly during the last 30 years. To visualize publication trends on this topic, we carried out searches using the words’ reproduction AND bryophytes’ OR ‘reproduction AND liverworts’ OR ‘reproduction AND hornworts’ OR ‘reproduction AND mosses’ in the database of Web of Science (www.webofknowledge.com) from 1983 until 2013. We observed that studies directly or indirectly related to reproduction in bryophytes increased progressively from 1987, quickly reaching around 20 papers per year in 2001 and a maximum of 25 papers in 2011 (Fig. 5).

Assuming that all literature published is not available in the database of the Web of Science, and that there are studies written in languages other than English, we expect that these numbers are pretty higher. The increase of studies with bryophytes, especially linked to their reproduction, may be associated to improvements in microscopy technologies, informatics, geo-referencing, molecular biology, biochemistry and plant physiology. Students and researches worldwide have realized the importance of these plants to understand relevant issues in climatic changes, nitrogen deposition, evolutionary developmental biology (evo-devo), phylogenetics and life history theory of organisms.

Particularly in Brazil, studies on bryophyte reproduction started mostly in the last two decades (Oliveira and Pôrto 1998, 2001, 2002; Pôrto and Oliveira 2002; Bastos 2008; Alvarenga et al. 2009; Maciel-Silva and Válio 2011; Maciel-Silva et al. 2012a, b; Pôrto and Silva 2012; Maciel-Silva et al. 2013). Two explanations for this fact are: 1) the huge diversity of Brazilian ecosystems and consequently its large bryophyte species richness, which remain poorly known by scientists; 2) the significant increase in new bryologist training only in the last years. Brazilian ecosystems, which include tropical rainforests and dry forests, savannas, rock outcrops, mangrove, sandbanks, etc., have large potential to studies focusing on the reproductive biology of bryophytes.

Many species and ecosystems remain unexplored regarding the bryophytes and their reproductive biology, especially in the tropics, claiming for natural history studies that will identify and characterize interesting systems for research. However, it is also important that future researches walk hand in hand with other disciplines like evolutionary biology, biogeography and functional ecology, promoting a comparative framework of the reproductive patterns and processes among all plants and not only bryophytes.