Introduction

Bryophytes share much of their physiology with other green land plants, but there are also important differences; the similarities and differences do not necessarily fall in line with simple expectations. Because most bryophytes have “stems” and “leaves,” and tradition has regarded them as “lower” plants, it is too easy to think of them as underdeveloped miniatures of vascular plants – as organisms that have evolutionarily not yet “made the grade.” Raven (1977,1984) has emphasized the importance of supracellular transport systems in the evolution of land plants, and the physiological correlates that we must read alongside the anatomical structures of fossil plants. But the highly differentiated supracellular conducting systems exemplified by xylem and phloem are really only a prerequisite for large land plants. For simple physical reasons of scale, conduction of water and metabolites in bryophytes can be much more diffuse. Similarly, bryophyte and vascular-plant leaves and leafy canopies must be thought through and compared as photosynthetic systems from first principles, not by simple analogies between structures operating at radically different scales, which can be seriously misleading.

In adapting to the erratic subaerial supply of water, vascular land plants evolved xylem, bringing water from the soil to meet the needs of the above-ground shoots and leaves. Bryophytes in general adopted the alternative strategy of evolving desiccation tolerance, photosynthesizing and growing during moist periods and suspending metabolism during times of drought. These two patterns of adaptation are in many ways complementary.

Introduction

Bryophytes are familiar and attractive ingredients in many types of natural landscape. Their shaggy coverings on branches and boughs, crags and boulders, in waterfalls and on woodland banks, add distinction to the larger scene. Less appealingly, they grow occasionally on bizarre materials, like the leather of a discarded boot, or a rusty iron pipe. Even in modern cities where air pollution and the built environment may seem unrelenting, there are bryophytes able to colonize crevices in masonry, soil accumulations in gutters, and to soften the otherwise geometrical wildernesses of roof tiles with their rounded cushions. To the scientist all these situations provide taxing problems concerning the supply of necessary resources, the impact of the bryophytes on their habitat, and their responses to undesirable chemicals in the environment. This chapter describes the special problems that bryophytes encounter in obtaining essential mineral nutrients, and in dealing with non-essential elements and compounds. The substratumt on which a bryophyte grows can be a source of nutrients and of other chemicals that may cause stresses. Also, the periods of time for which different types of substrata are available for colonization by bryophytes vary enormously. Therefore, both the chemical properties and the wider ecological characteristics of different substrata are considered from the point of view of their suitability for bryophyte growth. Many examples of bryophytes behaving as specific indicators of particular chemical environments are given. Their uses in biomonitoring provide practical instances of this.

Bryophytes have a number of life history characteristics that make them interesting and tractable organisms for population studies. Their life cycle is unique among land plants in having a free-living (usually) perennial gametophyte and an annual sporophyte that remains attached to the gametophyte throughout its life. The sporophyte obtains a significant proportion of its nutrients from the maternal gametophyte to which it is attached, although in mosses it is green and photosynthetic when immature and premeiotic (Proctor 1977). A specialized tissue comprised of transfer cells occurs on either the sporophyte or gametophyte side, or both, of the junction of the two generations, and facilitates the movement of metabolites between the gametophyte and sporophyte (Ligrone et al. 1993). This physiological connection, in addition to the obvious genetic relationship between gametophyte and sporophyte, must contribute to complex evolutionary patterns into which bryophytes offer unique opportunities for investigation (Shaw & Beer 1997).

Although it is often said that it is the gametophyte rather than the sporophyte that is exposed to the external environment, there is no empirical evidence to suggest that the sporophyte is completely or even mostly shielded from the forces of natural selection. Indeed, the fact that “life history strategies” can be recognized and include both gametophytic and sporophytic traits (During 1979, Longton 1997) implies a history of selection on traits in both generations. Evolution of the whole bryophyte plant must involve interactions between each generation and the external environment, as well as interactions, genetic and physiological, between the sporophyte and gametophyte generations of a single plant.

Introduction

Bryophytes, in the broad sense, are the second largest phylum of land plants, after the angiosperms, and inhabit every continent. Among the bryophytes, mosses (the Bryophyta sensu stricto) are the most speciose group, comprising approximately 10000 or more species. They differ from liverworts and hornworts in a suite of characters (see chapters 1,2), including macroscopic features such as a gametophyte composed of stems with undivided and often costate leaves that are typically arranged all around the axis, and a sporophyte terminated by a capsule that is elevated by the elongation of a seta prior to maturity, and whose mouth is lined by teeth involved in regulating the dispersal of spores. Mosses along with liverworts and hornworts represent the oldest lineages among extant land plants (chapter 5 in this volume). Evidence is mounting that mosses are currently at the highest level of diversity in their evolutionary history (e.g., Kürschner & Parolly 1999). This rather recent diversification of mosses is likely correlated with the advent of angiospermous forests which provide a wide array of habitats. This trend is best exemplified in tropical rainforests, where most of the diversity in mosses is found.

Although mosses are rather small organisms, their morphology is relatively complex. Macroscopic characters are limited and most taxonomic concepts rely on features of the cells composing individual tissues. Throughout their evolutionary history, mosses have undergone repeated morphological reduction and simplification (Frey 1981), often as a result of colonizing specialized, and particularly xeric or ephemeral habitats (Vitt 1981).

Neoregelia, with about 100 species (Luther and Sieff 1996), belongs to subfamily Bromelioideae, and consists of two subgenera with largely nonoverlapping ranges: Neoregelia with about 90 species and subgenus Hylaeaicum with 10 species. Subgenus Neoregelia is confined to eastern Brazil except for one species each in northern Venezuela (N. cathcartii) and Amazonian Peru (N. johnsoniae). Subgenus Hylaeaicum is entirely Amazonian in parts of Colombia, Venezuela, Peru, Ecuador and Brazil.

Neoregelia is distinguished from the other bromelioid genera with nidular inflorescences (Canistrum, Nidularium and Wittrockia) by its asymmetric sepals and lack of petal appendages (Leme 1998a,b). However, more recent studies (Ramírez 1991, 1994) have determined that petal appendages occur in members of subgenus Hylaeaicum, and Leme (1997) reported these same organs in subgenus Neoregelia (N. carolinae), indicating need to re-evaluate the taxonomic utility of this character.

Taxonomic problems

Nidularium eleutheropetalum and N. myrmecophilum were successively assigned to different sections of Nidularium, and the genera Karatas and Aregelia until in 1890 Lindman placed them in genus Regelia, which he created by elevating the status of Nidularium subgenus Regelia Lemaire. In 1891 Kuntze had proposed the name Aregelia as a nomen novum for Nidularium, so that its typification must be identical to that of Nidularium. Therefore Mez's decision to use Aregelia for a genus segregated from typical Nidularium is invalid.

Genus Regelia was named after the German botanist A. von Regel, who served as superintendent of the Imperial Botanic Gardens in St Petersburg, Russia. Because the name Regelia had already been assigned to three species of Myrtaceae, Smith (1934b) created the name Neoregelia, considering Regelia Lindman and Aregelia Mez, 1896 non Kuntze, 1891, to be synonyms.

Although evolution is the theme that ties this volume together, until now adaptive aspects of phenotype such as CAM, the phytotelm shoot and the absorbing trichome have dominated discussions. Times, places and why these features emerged during Bromeliaceae radiation have received far less attention. At this point, we switch emphasis to beginnings and circumstances that influenced the adoption of those characteristics that define much of the family as exceptional for adaptive novelty and importance in communities. Fossils, ontogeny, phytogeography, paleoclimate, cytogenetics and the structure of the genome provide insight on the geologic history and phylogeny of Bromeliaceae and point out directions for additional inquiry.

Phylogenetic analysis informed by the molecular structure of key segments of DNA and a fuller understanding of the morphology and adaptive biology of representative species will eventually reveal the identities and dates of the major evolutionary events responsible for the distinctness of the more advanced Bromeliaceae among Magnoliophyta. Conditions in primordial habitat(s) head the list of enduring questions: were these sites dark and humid like the forest understory or exposed and dry? More fundamentally, did scarcities of mineral nutrients or drought play the more decisive roles in the evolution of the foliar indumentum and phytotelm shoot? Heterochrony has also figured prominently in speculations about bromeliad radiation, but without much thought given to the incentives (plant benefits) responsible for this process.

A framework that arrays extant lineages in evolutionary space and geologic time (cladogram) will also resolve long-standing disagreements about taxonomy.

All of the impressive functional and ecological variety expressed by some 2700 species of Bromeliaceae is grounded on a single body plan, or what Hallé et al. (1978) might consider one architectural model. Widespread occurrence of this same design among extant monocots and the paleoherb hypothesis (Taylor and Hickey 1992) suggest that early Magnoliophyta possessed much the same basic organization. Except for the occasional monocarp, a somewhat larger group of relatively caulescent species (Fig. 2.1), and another modest-sized assemblage of lateral-flowering taxa (Fig. 2.2B), the bromeliads share a distinctly modular bauplan characterized by sympodial branching that leads to series of attached, compact, terminally flowered ramets (Fig. 2.3). Roots, if present beyond the seedling stage, mostly emerge along the lower half of each module.

Vegetative form that favors life on arboreal and lithic substrates also imparts substantial horticultural value to many of the bromeliads. Moreover, some of these same features assure exceptional importance in ecosystems, including indispensability to extensive fauna with diverse needs (Chapter 8). Two plant characteristics warrant special note on all three counts: a generally compact, rosulate shoot (the ramet or module) that often impounds moisture and nutrient-rich solids (creates the phytotelma and consequently the phytotelmata; Fig. 2.4) and the usually peltate foliar trichome (Figs. 2.5–2.9). These attributes, combined with others involving roots and shoots, favor success, including occasional dominance in some of the most exacting kinds of ecospace colonized by vascular flora in tropical America (e.g., Figs. 1.2C, 7.1).

Beyond its numerous, valuable ornamentals, Bromeliaceae contains relatively few widely used species, pineapple and Spanish moss being the two notable exceptions. Pineapple, Ananas comosus, ranks among the most popular of the tropical fruits (Cobley 1976). Spanish moss (Tillandsia usneoides) was once an important source of low-grade fiber in the southeastern United States, with annual production of up to 5000 tons. This wideranging species also has important medicinal uses in several regions, and an undocumented amount of material continues to sell for floral arrangements in the United States.

Hortus Third (L. H. Bailey Hortorium 1976) describes nearly 250 selections distributed among 30 genera. Hybrids, some between members of different genera, substantially augment the hundreds of species in cultivation. Although flowers tend to be small and ephemeral, unusual vegetative forms, ornamented leaves (e.g., Figs. 2.17B, 2.18B) and brightly pigmented floral bracts assure horticultural interest. Red, orange or yellow inflorescences of the many bird-pollinated bromeliads often signal from impressive distances. Sizes ranging from diminutive Spanish moss to giant Puya raimondii further entice hobbyists, and shade-tolerance suits many taxa for indoor cultivation. Frequent capacity to grow on a variety of substrates, including drift wood, cork slabs and fern roots, further enhances the popularity of bromeliads.

Bromeliaceae figure prominently in several additional contexts including cameo appearances in Star Trek movies and Star Trek: The Next Generation. Tillandsia usneoides and other epiphytic bromeliads adorn the sets of Tarzan movies, belying the films' purported African setting. Recent appearances include Medicine Man,where Sean Connery's character finally identifies Tillandsia punctulata, or rather the ants nesting within, as the source of a cancer cure.

Tillandsioideae are mostly rosulate herbs characterized by entire leaf margins, radially organized peltate trichomes (Fig. 2.7), usually superior (or nearly so) ovaries, and three-parted capsules that contain plumoseappendaged seeds (Fig. 3.6J;Wittmack 1888; Baker 1889; Mez 1894, 1896, 1934–35; Harms 1930; Smith and Downs 1977; Rauh 1990). Stigma morphology varies more than in the other two subfamilies, with at least five different types present (Brown and Gilmartin 1989b; Gortan 1991; Figs. 3.1C, 12.1). Pollen morphology is similarly variable (Fig. 12.2). Grains are predominantly sulcate with a distal germination region, and represent the diffuse sulcus, insulae-type, operculum-type and Vriesea imperialis-type (Halbritter 1992). Catopsis is exceptional with its simple sulcus, while inaperturate pollen occurs in some Guzmania species. This subfamily comprises the genera Alcantarea (16 spp.), Catopsis (21 spp.), Glomeropitcairnia (2 spp.), Guzmania (176 spp.), Mezobromelia (9 spp.), Racinaea (56 spp.), Tillandsia (551 spp.), Vriesea (187 spp.) and Werauhia (73 spp.) (Smith and Downs 1977; Utley 1983; Till 1984, 1992b, 1995; Kiff 1991; Grant 1993a,b, 1994a,b, 1995a,b; Spencer and Smith 1993; Luther and Sieff 1996; Till et al. 1997).

A substantial literature dating back more than a century describes the bromeliad reproductive apparatus. Taxonomists working with dried specimens authored most of the early treatments. Interest continues, but specimen quality has improved allowing analyses to be more comprehensive. For example, Brown and Terry (1992) used liquid-preserved flowers and scanning electron microscopy to determine that the delicate petal scale that figures so prominently in the most recent monograph of the family (Smith and Downs 1974, 1977, 1979; Fig. 3.1) circumscribes some genera more convincingly than others. Wet material has also permitted determinations of when certain features appear during ontogeny, and accordingly, their utility for distinguishing taxa of low vs. higher rank.

Plant form underlying reproductive phenomena like pollination and seed dispersal and the genetic structure of populations are our primary concern for this review. Unfortunately, few of the hundreds of publications devoted to the reproductive apparatus of Bromeliaceae provide much insight on any of these subjects. Moreover, inquiry on flowers, fruits and seeds continues to be motivated primarily by interests in systematics. The exceptional report that does depart from tradition usually addresses the same question, namely who pollinates which bromeliad?

Today, molecular biology is augmenting the morphological data traditionally used to infer bromeliad history. However, cladograms based on nucleotide sequences must be more fully resolved than those illustrated in Chapter 9 to produce the phylogeny necessary to determine where, when and how often decisive features of the reproductive apparatus evolved.

The rapid and accelerating destruction of the world's tropical forests is widely known. Less appreciated are two additional facts: half of the estimated 300000 species of higher plants occur in these biomes, and epiphytes make up to half of the vascular floras of certain tropical forests (Benzing 1990). Moreover, bromeliads comprise most of the biomass of arboreal vegetation at many wet montane tropical American sites. Because members of this family significantly influence important forest processes and provide substrates and other resources for much canopy-based fauna, their preservation is vital to broader conservation efforts. Terrestrial Bromeliaceae, about half of the family, sometimes dominate communities where climates are harsh (e.g., cool or hot and dry) or substrates (e.g., rock) mandate unusual plant adaptations to obtain nutrients and water (Chapters 4 and 5). Too little information is available in the literature on the sizes and genetic structures of bromeliad populations to determine if more than just a handful of taxa are truly endangered (Chapter 6), and thus the following discussion draws heavily on unpublished observations and findings on other taxa.

Bromeliads tend to be locally abundant if not as diverse as co-occurring, ecologically similar flora such as Orchidaceae. Individual phorophytes or rock faces routinely support hundreds to millions of adults. One of the many populations of Tillandsia purpurea in the coastal desert of Peru (Pan-American highway, kilometer 348 north) covered approximately six square kilometers at a density of ∼50 ramets per square meter to total some 300 million ramets (Dimmitt 1989a).