Chapter 12

Pteridophytes of North America

Warren H. Wagner, Jr. Alan R. Smith

General Background

Pteridophytes, the ferns and so-called fern allies, comprise about 3% of the vascular plant species of continental North America north of Mexico. This assemblage of organisms includes very diverse elements such as aquatic quillworts (Isoëtes; Lycopodiophyta), desert cliff brakes (Pellaea; Polypodiophyta), nearly leafless whisk-ferns (Psilotum; Psilotophyta), and weedy horsetails (Equisetum; Equisetophyta). These groups are so divergent that they are placed in four different divisions (A.Cronquist et al. 1966).

What is the justification for grouping the pteridophytes together? Common features include: (1) the production of spores, the principal dispersal units; (2) the germination of spores and subsequent development to produce gametophytes that exist independently of the spore-producing plants; and (3) the fertilization of eggs by flagellate, swimming sperms. In addition, pteridophytes share a negative feature---the absence of seeds. The seedless condition is a primitive one that characterized all early land plants.

Most pteridophytes are basically herbs, some as small as 2--3 mm, and only a minority reach heights of more than 2--3 m. Only tree ferns, such as many Cyatheaceae, have tall trunks, and Cyatheaceae are essentially absent, except in cultivation, outside of the tropics. Even in the most tropical areas of the United States (e.g., southern Florida), there are no native tree ferns. Therefore, the tallest pteridophytes in North America north of Mexico are the climbing ferns (Lygodium); their height, however, is not accomplished by stems but rather by extremely long leaves that clamber up other plants. Species in the flora have twining leaves that may reach 3--7 m.

With the exception of certain grapeferns (Botrychium) and Isoëtes, no modern pteridophyte displays true secondary growth: the tissues are all primary, derived from a terminal meristem. Pteridophytes have no cambium, and they lack cork and secondary vascular tissues.

Pteridophytes are noted for retaining many primitive features of the earliest land plants. Some of the early pteridophytes were trees to perhaps 35 m tall with abundant secondary wood (e.g., the progymnosperms and certain lycopods of the Devonian [405--345 million years ago] and the Carboniferous [345--280 million years ago]). The club-moss (Lycopodiaceae), spike-moss (Selaginellaceae), and quillwort (Isoëtaceae) lines of evolution have been distinct from all other vascular plants since approximately the early Devonian. Whisk-ferns (Psilotum) and horsetails (Equisetum) are so distinctive that we are still in a quandary as to their relationships.

Evidence from the fossil record suggests that the earliest vascular land plants (those possessing xylem and phloem, the specialized water- and food-conducting tissues) had a life cycle similar to modern pteridophytes. These early examples are known today from fossil imprints and petrifactions, especially in Devonian deposits of eastern Canada and the northeastern United States, where some of the world's best preserved fossils of these ancient plants have been discovered. Remarkably detailed features are still preserved, including internal tissues, spore cases, and spores.

Some pteridophytes, such as Psilotum, seem nearly to have lost their ability to speciate; many others, such as Thelypteris, are apparently speciating actively today. Many of the modern groups show great diversity in form, life cycle, and habitat preference. We discuss some details of this diversity, especially as they pertain to an understanding of the taxonomic treatments in this flora, in the sections that follow.

Life Cycle

The pteridophyte life cycle is characterized by having two separate free-living plants, gametophytes and sporophytes, interconnected by stages of the sexual process. This phenomenon is referred to as alternation of generations. Ordinarily, the more conspicuous and dominant plant is the sporophyte (diploid and 2n), which is usually perennial and lives for an indefinite period. The gametophytes (haploid and n) tend to be inconspicuous (usually considerably less than 1 cm in the largest dimension) and short-lived. When fertilization occurs and the new embryonic sporophyte forms, the gametophyte dies. Gametophytes are occasionally the dominant or only generation present in a given species in certain temperate outliers of tropical species (see Modifications of the Life Cycle later in this chapter).

Gametophytes of closely related species and genera tend to be much alike and have been reported for only a small percentage of the species of North American pteridophytes. They are also difficult to detect because of their occurrence in small niches, and some grow underground. For these reasons, it is primarily the sporophytes that are described in this flora.

Organs---Roots, Stems, and Leaves

Sporophytes are made up usually of three organ systems: roots, stems, and leaves (fig. 12.1). Roots of pteridophytes have been poorly surveyed, and our knowledge of them is limited. They tend to arise along the stem, commonly near leaf bases. Most pteridophytes have very narrow (less than 0.5 mm thick) and wiry roots, but roots may be thicker (more than 1 mm) and fleshy. The roots may form a dense mass, as in Osmunda, that may be cut up and used as a substrate ("osmundine") for growing orchids or other epiphytes, or roots may be scattered and few, as in certain lycopods. Delicate root hairs may be abundant (Vittariaceae), sparse (most ferns), or totally absent (Ophioglossaceae). Most root hairs occur behind the root tips and die off, but they may be persistent in some groups.

The stems of pteridophytes are mainly true rhizomes, i.e., stems that run horizontally at or just beneath the surface of the ground. The growing tips, however, may be ascending or erect. Some ferns have upright stems like tiny tree trunks (caudices) and are unable to form colonies except for occasional upright branchings, as in Botrychium. Rarely a pteridophyte may have both upright caudices and horizontal rhizomes, these strongly differentiated as in Matteuccia. The rhizomatous habit of many pteridophytes results in extensive vegetative reproduction. Often a large continuous colony is actually a clone, and in terrestrial species, the individual leaves or plants that appear aboveground are from the rhizome system of a single original plant. In the jack pine (Pinus banksiana) forest of northern Michigan and Wisconsin, the rhizomatous brackens are the dominant understory, covering thousands of square kilometers. Some clones may be hundreds or even thousands of years old and cover a hectare or more, but because they lack secondary growth, their age is difficult to ascertain.

The clonal nature of growth is one of the main reasons why some sterile hybrid ferns and fern allies are capable of becoming persistent and sometimes common elements in the vegetation. Although most ferns are capable of propagating by rhizomes, others spread from leaves or roots (see below). The majority of individual pteridophyte plants encountered in forests, fields, fens, and along pond edges are vegetatively produced. In only a few terrestrial species, e.g., most moonworts and grapeferns, are individual plants formed only by sexual reproduction. Not surprisingly, the majority of plants that occur in narrow crevices of exposed rock cliffs also tend not to form clones. Even if they have the ability, the barren rock surfaces separating the crevices tend to keep the plants compact, small, and isolated.

Anatomically the stems of pteridophytes are simpler than in most seed plants because they are made up only of primary tissues. Complex patterns produced by secondary production of new xylem and phloem are absent. Even the primary tissues, however, may have elaborate arrangements, such as platelike systems of xylem (the plectosteles of certain Lycopodiaceae) or tubes of netted xylem-phloem bundles (the dictyosteles common in leptosporangiate ferns). Cross sections of stems reveal many patterns. These range from solid cores of xylem (protosteles) as in certain lycopods, to nearly closed tubes of xylem (siphonosteles) as in Dennstaedtia, to complex open tubular nets of xylem (dictyosteles) as in many ferns. Externally, the stems of most ferns tend to be covered with hairs or scales (see below), especially near the growing tips where indument protects the easily injured cell initials (meristem).

Leaves of pteridophytes are extremely diverse, ranging from unvascularized tiny projections (enations) as in Psilotum, to small, simple, needlelike or scalelike leaves with only a single central vein (microphylls) as in the lycopods, to rushlike microphylls up to 50 cm long in Isoëtes, and to the large complex leaves with an intricate pattern of many veins found in most ferns (megaphylls). In ferns the leaves are often referred to as fronds. These develop in a distinctive manner with the soft growing tip (meristem) rolled up in the center of the crozier (fiddlehead). The crozier is produced by a process of growth and unfolding, and the characteristic pattern thus formed is known as circinate vernation.

Leaf development that is initiated at the tip is called acropetal, and this development is characteristic of most modern ferns. Leaves of a few genera have another kind of development. For example, Ophioglossum leaves grow by intercalation and are conduplicate rather than circinate. Leaves of seed plants, in contrast, tend to develop by overall growth and expansion, i.e., by intercalary differentiation and enlargement.

The parts of a leaf include the petiole (stipe) and the blade (lamina). Petioles of ferns are diverse and offer many taxonomically valuable characteristics and character states. There may be a single round stele or vascular trace, or the trace may be U-shaped or omega-shaped, or there may be 2--10 or more separate traces. The patterns are characteristic of genera and sometimes families. For example, the petiole trace of most Aspleniaceae is X-shaped in cross section.

The main axis of the blade is called the midrib (rachis). If the blade of a fern is divided into leaflets, these are called pinnae, and if the leaflets are divided again into subleaflets, the subdivisions are called pinnules. To be defined as a pinna or pinnule, a leaf subdivision must be stalked or at least obviously narrowed at the base and not connected to adjacent pinnae or pinnules by greenish blade tissue; leaf subdivisions that are not stalked or narrowed are called segments. If a blade has no divisions it is termed simple; if it is once-divided only, it is regarded as 1-pinnate (fig. 12.2). If the pinnae are divided further into pinnules, the blade is 2-pinnate. A very finely dissected blade may be 3(or more)-pinnate (decompound).

Lobed blades are often mistakenly termed pinnate although, in fact, the segments are not narrowed at all at their bases. If the sinuses in a lobed leaf reach only part way to the rachis, the leaf is pinnatifid. If, however, the sinuses extend all the way to the rachis, the blade is considered pinnatisect. Terms referring to the dissection of the blade, along with terms pertaining to blade outline and relative petiole length, are used extensively in describing and identifying ferns.

Indument---Hairs, Scales, and Glands

In ferns, outgrowths of the epidermis of stems and leaves are common and varied, and these features are very important in description and identification. Simple outgrowths made up of a single cell or a chain of several cells are generally called hairs or trichomes and are regarded as the most primitive kind of outgrowth. Some primitive fern groups have only hairs (e.g., Ophioglossaceae, Schizaeaceae sensu stricto, Dennstaedtia and certain of its relatives). Trichomes that have two or three parallel rows of cells at the base and a single file of cells at the tip are called bristles (setae). More elaborate outgrowths that form flat plates of 3--20 or more rows of cells are scales (paleae). Scales may be attached basally or attached centrally on a small stalk; the latter are described as peltate. All transitional forms between hairs and scales may be found in some species, even on the same plant.

One of the more distinctive features of many hairs and scales is the presence of enlarged and rounded terminal secretory cells. Glandular hairs, sometimes called glands, may characterize particular species. Such glands are not to be confused with nectaries, which are also secretory structures. Nectaries are rare in pteridophytes and have been found in bracken and certain polypodies (S. Koptur et. al. 1982). The presence of glands in Dryopteris intermedia quickly distinguishes it from its look-alike, Dcarthusiana. Likewise, Gymnocarpium robertianum is glandular, and the similar Gdryopteris is without glands or nearly so. Glandular cells may occur at tips of hairs or scales. When the leaf of a given species or variety of fern is heavily glandular, the blade will actually stick to the paper when it is pressed (as with the rare glandular variety of Osmunda cinnamomea).

Some pteridophytes lack trichomes on stem and leaf surfaces. Most lycopods are hairless, but Palhinhaea has hairy stems. Hairlike structures are absent in most other fern allies, although lateral projections along the leaf margins, often called teeth or cilia, may resemble multiseriate hairs (LycopodiellaSelaginella). Lycopodium clavatum has terminal leaf hairs.

Gametophytes of most pteridophytes tend to have distinctive brownish or colorless hairs (rhizoids) that apparently serve to collect water and to anchor the plant. They may also have uniseriate hairs and glands on the prothallial margins and surfaces. Trichomes sometimes occur among or on the gametangia (as in Huperzia) or sporangia (as in Polypodium) and are then called paraphyses.

Veins---Free, Dichotomous, or Reticulate

The venation of ferns provides a rich source of information for identification (W.H. Wagner Jr. 1979). Vein patterns may best be seen by holding the blade up to the light, unless the tissues are very thick. Venation in dried, pressed specimens may be seen by clearing them temporarily with a drop of 95% ethanol.

Many ferns, especially primitive ones, have pinnately arranged free veins, with the major veins running out from the rachis laterally and the minor veins repeating the pattern. In leaves in which veins come together to form a network of loops or meshes (veins anastomosing; venation reticulate), the area within a loop is an areole (fig.12.3).

More advanced ferns often have various types of reticulate venation. Sometimes, as in Thelypteris, the anastomosed veins are found only below the sinuses between segments of the pinnae. In other cases, the basal veinlets connect to form a chain of areoles along the pinna midribs (as in Woodwardia). Vein connections are found throughout the blade in Asplenium rhizophyllum and Onoclea sensibilis. Rarely in North American ferns (but commonly in tropical ferns), the spaces within the areoles are occupied by single, free, unbranched or branched veinlets (fig.12.3) called included veinlets (as in Ophioglossum).

Rarely, the venation of a pinna or pinnule may be entirely dichotomous with the veinlets repeatedly forked from the base to the tip of the segments. This condition is illustrated by certain species of Adiantum and of Botrychium. One of the special features of truly dichotomous venation is that there is no midrib. The small, juvenile or first-formed leaves of many ferns have dichotomous venation, but distinct midribs are generally found in more adult, larger leaves.

Sporangia

The spore-producing organ of pteridophytes is the sporangium. With few exceptions, sporangia are borne on leaves or modified leaves (in Psilotum they are produced at the tips of short lateral branchlets). When the leaves that bear sporangia are like those that are only photosynthetic, the leaves are described as monomorphic. If, however, the sporangia are borne on leaves or leaflets that are strongly modified and different from the photosynthetic (or nutritive) leaves, the leaves are said to be dimorphic. In plants that show some divergence of fertile and sterile leaves, the fertile ones are often called sporophylls, and the sterile leaves are trophophylls (W.H. Wagner Jr. and F.S. Wagner 1977). Sporophylls tend to be taller or longer-petioled than trophophylls, and if they are photosynthetic at all, it is only in early stages of development. They usually have little or no laminar tissue. Also, sporophylls tend to be short-lived, lasting only long enough to produce and discharge the spores.

Sporangia of all fern allies and the primitive ferns are basically like those of the pollen-producing microsporangia of gymnosperms or the anthers of angiosperms in having thick walls with a number of cell layers. The sporangia open usually by a transverse split and produce hundreds or thousands of spores. This type of sporangium, which is most like the original vascular plant sporangia of the Lower Devonian, is referred to as a eusporangium, i.e., true or typical sporangium, the one found in practically all vascular plants.

Most ferns, however, have drastically modified and reduced sporangia that are so simple, they appear to be little more than elaborate trichomes bearing a tiny round spore case at the top. This distinctive type is called the leptosporangium (from the Greek leptos, slender or small). The spore case itself is few-celled, and the outer wall is only one cell thick at maturity.

The number of spores produced in leptosporangia is usually only 128, 64, or 32, most commonly 64. Most leptosporangia have a bow or annulus made up of strongly modified, thickened cells. The position and extent of the annulus are important characters that define a number of fern families, particularly the primitive leptosporangiate ones. The leptosporangia of Osmunda have patches of specialized, thickened cells. Those of Schizaeaceae, Lygodiaceae, and Anemiaceae have apical annuli. In Hymenophyllaceae, the annulus is oblique and not interrupted by the stalk. In most other leptosporangiate ferns, the annulus runs lengthwise from the base of the sporecase (where it abuts the tip of the stalk) over the distal end of the sporangium, thus dividing the sporangium into two equal halves. The stalk itself may be made up of 5--6 rows of cells, 2--3 rows (most ferns), or rarely only 1 row (as in Grammitidaceae and Aspleniaceae). The comparative morphology of leptosporangia deserves a great deal more attention than it has received, and it is a likely source of many additional taxonomic characters (K.A. Wilson 1959).

The sporangia are variously arranged, and the particular arrangement is a useful character in recognizing taxa. In Psilotum they are borne in fused clusters of three on short lateral branches; in all lycopods and quillworts, they are solitary and borne in or just above the axils of microphylls; and in the horsetails and scouring rushes, 5--7 pendent sporangia are borne on peltate sporophylls that are aggregated in cones. In ferns, sporangia are on leaves or leaf parts, in various arrangements, solitary to many, often in clusters (sori) that are linear, oval, or round; in some species the sporangia completely cover the abaxial surface (sporangia acrostichoid, as in Acrostichum). Sori are often protected by flaps formed by the leaf margin (false indusia) or specialized scales or cuplike coverings (true indusia) that are separately produced from the area on which the sorus is borne (receptacle). It is important, especially in ferns, to obtain fertile leaves that bear sporangia because this usually greatly simplifies identification, especially at the generic level.

Spores

Spores have been extensively employed in discriminating taxa of pteridophytes at all levels (A.F. Tryon and B.Lugardon 1991). Sometimes almost identical spore types appear in unrelated taxa, but generally there are good correlations of the spore type with taxonomy (A.F. Tryon 1986). The functions and adaptive significance (if any) of the different spore types are largely unknown. It is known that green spores (containing chloroplasts) are common in epiphytes growing in tropical rainforests (Hymenophyllaceae, Grammitidaceae) and in temperate genera that release their spores early in the spring (OnocleaMatteuccia, Osmundaceae).

Spores of most modern pteridophytes are all of one type, and the taxa are said to be homosporous. Individual spores are mostly 20--60 µm in length or diameter and are almost invisible without a microscope. En masse spores may be white, yellow, orange, green, brown, or nearly black. A number of pteridophyte orders have evolved the condition known as heterospory, in which two types of spores are produced by a given species: small spores mostly 20--30 µm long (called male) and large spores 200--700 µm in diameter (called female). Some of the life cycle modifications associated with heterospory will be discussed in a following section.

Spore shape and symmetry also vary. Spores may be tetrahedral or nearly globose, and the scar (laesura) on the inner (proximal) surface (toward the center of the original tetrad) may be triradiate; or the spores may be more or less reniform or bean-shaped and bilateral, with a single straight laesura. The type with the triradiate laesura is called trilete, and the one with the linear laesura, monolete.

Spores also differ greatly in the development of the wall layers. The basic layers are the endospore and the exospore, and they differ in their development and structure from the wall of pollen grains in seed plants (B.Lugardon 1978). An outer tapetal deposit, known as the perispore or perine, may also be evident under the light microscope as a massive buildup of ridges, wings, or warts. B.Lugardon (1971) determined that practically all ferns have at least a rudimentary perispore.

In nature, spores are distributed largely by air convection, and it is believed that long-distance dispersal by wind is far more common in pteridophytes than in seed plants. By and large, pteridophytes tend to have wider ranges than seed plants (A.R. Smith 1972). Most spores are easy to germinate on soil, flowerpot chips, and even tap water. Spores sometimes germinate after decades, but green spores remain viable only a week or so after their release from the sporangium. Spores of pteridophytes with underground gametophytes associated with fungi germinate only in the dark (D.Whittier 1972).

Germination of spores generally occurs on disturbed and exposed areas of moist soil, in rock crevices, or on rotting logs, often among bryophytes. The best places to look for young gametophytes are on shaded soil banks, at edges of streams, and on mossy bark at bases of trees. Under favorable conditions gametophytes appear by the thousands, especially when the spore parents are nearby. More common perhaps is the appearance of a single plant or a few isolated ones in a tiny microhabitat, but such isolated and sporadic specimens are rarely observed.

Gametophytes

Sexual fusion in pteridophytes occurs on inconspicuous, free-living plants known as gametophytes. These are often associated with similar-appearing but much more common mosses and liverworts. Most gametophytes are green and surficial (surface borne), but some are nongreen and subterranean. Surficial photosynthetic prothalli are generally flat and have clear-cut upper and lower surfaces. The upper surface in Equisetum and Lycopodiella is provided with projecting lobes or flattened processes. The under surface, i.e., the side next to the substrate, bears specialized hairs (rhizoids) that presumably function for anchorage and water uptake. Subterranean gametophytes tend to be much fleshier, either cylindric or thickly wafer-shaped, and yellowish to brownish. They are extremely difficult to find in nature.

Both surficial and subterranean gametophytes depend on free water through which the sperms must swim to reach the egg and achieve fertilization. The male gametes (sperms) are provided with special organelles equipped with cilia that propel the gametes by their motion. There may be two cilia per sperm (Lycopodiaceae, Selaginellaceae) or dozens (all other modern pteridophytes). Sperms are produced in specialized cases (antheridia) that are either sunken in the gametophyte or protrude from it. The antheridia release the sperms through a pore. Release depends on maturity of the sperms and the presence of water. The female gamete (egg) is located in a bottlelike organ, the archegonium. Sperms swim to the opening at the top of the neck of the archegonium. The neck provides a passageway to the enlarged base (venter), where the egg is located and where fertilization takes place.

Both antheridia and archegonia may be present on an individual gametophyte, especially if it is growing singly in a culture dish (E.J. Klekowski Jr. 1969). Thus it is possible to have intragametophytic selfing, the fertilization of an egg by a sperm from the same gametophyte. In nature, however, it is believed that the tendency, at least in ferns and Equisetum, is for gametophytes to pass sequentially through all-archegonial or all-antheridial stages. Female gametophytes release soluble substances (antheridiogens) that stimulate nearby gametophytes to develop antheridia only. This tends to promote cross-fertilization between gametophytes of the same or closely related species. It may help explain the well-known predilection for interspecific hybridization to occur, even in species with subterranean gametophytes (W.H. Wagner Jr. et al. 1985).

Embryos and Young Sporophytes

The embryogeny in a number of pteridophytic groups is still unknown or poorly known; what is known has been summarized mainly by D.W. Bierhorst (1971). The embryos of many pteridophytes develop their organs early, and it is often possible to distinguish the first leaf, the stem, and the root at a few-celled stage. Young sporophytes differ almost as much among the major groups as do the adult sporophytes. In some families, e.g., Equisetaceae, Lycopodiaceae, Psilotaceae, the early stages are simple and similar to greatly reduced adult stages. In many ferns, however, the juvenile leaves are often strikingly unlike the mature leaves; the "sporelings" have dichotomously constructed early fronds, the two halves corresponding to the two basal pinnae or vein trusses of the intermediate and mature leaves (W.H. Wagner Jr. 1952).

Chromosomes

Spore mother cells, located inside the developing sporangia, are often used to study the chromosomes of pteridophytes. The spectacular work of Irene Manton of the University of Leeds, whose book Problems of Cytology and Evolution in the Pteridophyta (1950) aroused pteridologists around the world, led to numerous profound changes in our taxonomic concepts. By making preparations of spore mother cells undergoing meiosis, she was able to determine the diploid number (2n) of chromosomes for individual species and also the base number (x) of chromosomes for each genus she studied. Additionally, she was able to observe the cytological effects of hybridization between species.

From the enormous number of studies that have been reported since 1950, we can make certain generalizations about pteridophyte chromosomes. Homosporous pteridophyte genera have high chromosome base numbers (i.e., the lowest haploid multiple) ranging from about x = 20 to x = 110. These are among the highest base numbers known in vascular plants. Genetic evidence suggests that homosporous ferns with high base chromosome numbers are diploid (C.H. Haufler and D.E. Soltis 1986), even though they may be derived from ancient polyploids (paleopolyploids). Heterosporous pteridophytes, however, have low chromosome base numbers like those of most seed plants, i.e., x = 7--11. This generalization applies across all divisions of pteridophytes.

Polyploidy may be superimposed on the base numbers and, at least among homosporous taxa, very high numbers are sometimes attained (in Ophioglossumnnumbers exceed 600). Of particular interest to taxonomists is the discovery that certain base numbers characterize major groups; adiantoid ferns have x = 29, 30; dryopteroid ferns, x = 40--42; osmundoid ferns, x = 22; huperzioid lycopods, x= 67, 68; equisetums, x = 108; and so on.

By observing meiotic chromosomes, we can usually determine whether a given plant is a hybrid or not. Meiosis in hybrids is generally very irregular, and the spores that result are malformed. Malformation of spores is generally attributed to the genetic imbalance resulting from uneven distribution of chromosomes. Sexual species usually have normal meiosis and spore production, whether they are diploid or polyploid. Even normal species may produce triploid individuals, however, by chance fusion of haploid and diploid gametes. Such individuals are invariably sterile (or apogamous, see below) because the odd number of chromosomes usually prevents a normal reduction division.

Hybridization is common in many groups of pteridophytes and plays an important role in the taxonomy of these plants. In a few cases, species may hybridize and the progeny retain the capacity for normal meiosis and spore formation. This is best illustrated in North America by two genera of club-mosses, Diphasiastrum and Lycopodiella, in which interspecific hybrids appear to be perfectly normal meiotically and produce well-formed, viable spores. With few exceptions, these club-moss hybrids are always in the minority in mixed populations, and their significance is poorly understood.

Modifications of the Life Cycle

From the life cycle described above there are many deviations. The evolution of heterospory extensively modified the sexual life cycle characteristic of homosporous pteridophytes. Heterosporic life cycles involve two strikingly different gametophytes, the microgametophyte (male) and the megagametophyte (female), both of which develop within the walls of their spores (endosporic development). The male gametophytes develop from small microspores, the female gametophytes from much larger megaspores, which are readily visible to the naked eye. They produce, respectively, the sperms and eggs. Fertilization takes place in water on the ground or in ponds. The heterosporous condition is known in pteridophytes as early as the Devonian and has evolved in a number of independent lines, including the ligulate lycopods (Isoëtaceae and Selaginellaceae), Marsileaceae, Salviniaceae, and Azollaceae. The heterosporous cycle may have been originally an adaptation to strongly seasonal or xeric habitats where moisture was available only for short periods during sporadic or vernal rains. Reproduction by this process is rapid because it is not necessary to grow a whole photosynthetic thallus, as in the homosporous cycle.

An asexual modification of the sexual homosporous life cycle is known as apogamy. The apogamous life cycle is common in leptosporangiate ferns and is especially widespread in the dry parts of western North America, among rock ferns like the cheilanthoids. Many well-known species, such as Phegopteris connectilisPellaea atropurpureaPteris cretica, and Asplenium resiliens, show this type of reproduction.

In the most common type of apogamy, there are two generations, but fertilization is not required for production of the sporophyte, which originates directly from the tissues of the gametophyte. The same chromosome number in sporophyte and gametophyte is maintained by doubling of chromosomes by endomitosis (nuclear division without cell division) in the formation of the spore mother cells. Therefore, only half as many spore mother cells develop, but each of these has double the number of chromosomes. These 4n spore mother cells undergo meiosis to produce normal-appearing 2n spores with the same chromosome number as the parent sporophyte. Generally, apogamy can be recognized because there are half as many spores as normal, and these are larger than for sexual members of the same genus. If the number of sexual spores is normally 64 per sporangium, the apogamous fern will have only 32.

Commonly an apogamous species will also have a sexual form, and the former may originate several times and in several different places from the latter (G.J. Gastony and L.D. Gottlieb 1985). Sexual species usually have an even number of genomes (2x, 4x, 6x) whereas apogamous ones have an odd number (3x, 5x), but exceptions are frequent (e.g., Pellaea glabella).

Apogamous taxa cannot hybridize with each other, but they are often capable of hybridizing with sexual taxa because they can produce antheridia and viable male gametes, even if they cannot form normal archegonia and eggs. Apogamy is inherited by the hybrid, as in Asplenium heteroresiliens (Aheterochroum [sexual] X A. resiliens [apogamous]).

In some pteridophytes one or the other generation is partially or entirely eliminated. Vegetative asexual reproduction is accomplished usually by creeping rhizomes. This is most notable in sterile hybrid ferns, such as Osmunda X ruggii (Oclaytoniana X regalis), which are capable of forming enormous colonies. One can only estimate the age of such large clones because the old rhizome branches die and rot away; they may be well over a thousand years old.

Specialized buds are also produced on one or another of the vegetative organs. Root buds are well known in certain rock-inhabiting Florida aspleniums. Such buds are also found in most species of Ophioglossum. Stems may produce minute brood bodies (gemmae), such as those found around the subterranean shoot apices of Psilotum. These play a role in colonization, especially in greenhouses, where plants are scattered from pot to pot by soil-borne gemmae. In certain species of prairie-inhabiting moonworts (Botrychium), masses of tiny spherical gemmae develop along the stem. These break off in the soil and become new plants.

Single leaf buds at the tips of long narrow fronds are familiar in a number of ferns, notably Asplenium palmeri and the walking fern Arhizophyllum. Buds are also formed along the rachises of a number of dark-petioled spleenworts, such as Amonanthes. The common eastern American Aplatyneuron produces a single bud in the center of the lowermost reduced pinna of large fronds, and a single plant is therefore able to produce clusters of numerous plants over a period of years.

In Actinostachys pennula, all new leaves are produced from the base of preexisting leaves, so that a tough, tangled mass of many dozens of petiole proliferations is produced. There is no main shoot apex at all; each leaf comes from a proliferation of a previous leaf. The most remarkable leaf proliferations in the North American fern flora are the highly specialized bulblets of Cystopteris bulbifera. Nearly spherical plantlets are abscised from the distal part of the rachis and pinna costae and fall or roll to positions where they germinate and form large colonies. The complex and somewhat samaralike gemmae of the typical fir-mosses (Huperzia) are notable for spreading widely not only the normal sexual phase of the species, but their sterile misshapen-spored hybrids as well.

Perhaps the most surprising examples of vegetative propagation are those found in primarily tropical genera that have apparently spread and remained in the temperate zones by gametophytic gemmae. Members of three families---Hymenophyllaceae, Grammitidaceae, and Vittariaceae---occur in the Appalachian region as far north as New England, but they often appear only as gametophytes that may produce extensive colonies from the minute, one- or few-celled propagules (D.R. Farrar 1967, 1985). The gametophytes themselves are simple algalike filaments or ribbons, easily confused with green algae or liverworts. These gametophytes have few or no gametangia, and they are apparently adapted to growing only in shaded, humid, bryophytic habitats, mainly in rock crevices and grottoes. They normally produce no sporophytes at all and are the only representatives of their respective families in the floras of a number of states. In this flora we key them out alongside typical sporophytic members of their genera.

Habitats

The optimal habitat for most North American pteridophytes is rich, moist, mesic forest. Ferns and fern allies reach their greatest abundance and diversity in such habitats. Many genera are common, e.g., HuperziaLycopodiumEquisetumOsmundaDryopterisPolystichumThelypteris, and Athyrium. The plants average 3--10 times as tall as those of rocky exposed habitats, and many of them are capable of producing extensive clones. It is common for more than one species in a genus to cooccur in favorable habitats, and it is in such instances that we can often gain great insight into taxonomic relationships (W.H. Wagner Jr. and F.S. Wagner 1983).

Because of extensive farming in the past, as well as fires and lumbering, much of the continent is spotted with areas of disturbance that are in the process of returning to a climactic and stable state. Such successional vegetational formations provide habitats for species dependent on disturbance. Probably the most conspicuous pteridophyte of this type is Pteridium aquilinum, which is widespread in fields and second-growth open forests. It forms expansive clones that may produce the dominant vegetation in, for example, Pinus banksiana--Populus tremuloides associations in the upper Great Lakes region.

Among other second-growth specialists are many club-mosses (Lycopodium and Diphasiastrum), Ophioglossum, and BotrychiumAsplenium platyneuron may form large colonies on old farm sites and in previously logged areas. Another widespread successional plant in eastern North America is Dennstaedtia punctilobula, found commonly in pastures and old fields. Roadside borders and banks provide habitats for such successional species, as do cultivated areas. The pteridophytes that occur in these habitats are not necessarily weedy. Indeed, one of our rarest southeastern pteridophytes, Botrychium lunarioides, occurs almost exclusively in weedy and grassy roadsides and banks, lawns, and cemeteries. The last is the normal habitat for a number of our minute tropical adder's-tongues in the southeastern United States. A typical cemetery in Louisiana might yield as many as four species of Ophioglossum and two species of Botrychium.

The naturalist Edgar T. Wherry (1885--1982) stimulated much interest in the importance of soil and ecology in the biology of pteridophytes, and his influence was profound. Wherry was first drawn to rock-inhabiting ferns because of his interest in geology and mineralogy. He pointed out that acidic rocks like granites, quartzites, and sandstones have their own suites of pteridophyte species (especially huperzias and certain aspleniums). Likewise, calcareous rocks like certain shales, limestones, and dolomites harbor sets of characteristic species, particularly other aspleniums, pellaeas, and notholaenas. It is interesting that ferns in calcareous habitats are commonly bluish or whitish in aspect. A few of the western North American rock ferns grow primarily on serpentine rocks (e.g., Polystichum lemmonii and Aspidotis densa).

Ferns of rocky habitats, such as cliffs, walls of sinkholes, boulders, and talus slopes, tend to be more or less xeromorphic and able to withstand considerable drying. They are usually small and tufted, and the rhizome is short, often appearing to be erect. The laminar tissue is mostly leathery. The apogamous life cycle occurs widely among rock ferns, and both sexual diploids and apogamous triploids may be found together on the same rocks. The major North American genera of rock-growing pteridophytes are HuperziaSelaginella (Srupestris group), PolypodiumCheilanthesArgyrochosmaNotholaenaPellaeaPolystichumAspleniumGymnocarpiumCystopteris, and Woodsia. One species of Dryopteris, a normally soil-growing genus, is Dfragans, which is restricted to various rock types in northern North America.

Wetland habitats for pteridophytes are varied and include such diverse types as lakes and ponds, fens, bogs, conifer swamps, and hardwood swamps. In the southern United States, Azolla and Salvinia are common in ponds and canals. In the South and Southwest, Marsilea and Pilularia are frequent along the banks of ponds and streams. Throughout North America various Isoëtes grow ephemerally in temporary ponds on prairies and rock outcrops or continually in permanent marshes, lakes, and slow-moving streams.

Marshes, fens, bogs, and swamps offer ideal sites for many genera. Open grassy marshes provide the habitat for the abundant Thelypteris palustris in eastern North America. Acid bogs in the Great Lakes area and along the eastern coastal plain are habitats for Lycopodiella and Pseudolycopodiella, which occur also in burned-over, damp, grassy and sandy areas, together with sundews (Drosera) and pitcher-plants (Sarracenia). At various places from Delaware to Newfoundland, the famous curly-grass fern (Schizaea pusilla) may be found in such acidic sites.

Deep hardwood and conifer swamps in eastern North America generally support species of EquisetumOsmundaDryopterisAthyriumWoodwardia, andThelypteris. The cooccurrence in swamps of black ash (Fraxinus nigra), red maple (Acer rubrum), and yellow birch (Betula alleghaniensis), as well as arbor vitae (Thuja occidentalis), is an excellent sign of potential pteridophytic diversity. Oddly, a number of normally wetland species may occasionally appear on rock cliffs, along ledges and near their bases. These include such unexpected occurrences on rock as Osmunda regalisAthyrium filix-feminaCystopteris bulbiferaWoodwardia areolata, and Thelypteris simulata.

Probably the best example of a truly xerophytic pteridophyte is Selaginella densa, which occurs as an inconspicuous, dull gray crust on the deserts and plains of an enormous section of the continent from Manitoba and British Columbia to California and Texas. In terms of total biomass, it may actually constitute our most abundant pteridophyte. In general, in desert regions pteridophytes grow in the protection of shaded fissures and ledges of rock cliffs and boulders. Compared to seed plants, pteridophytes are generally rare in xeric localities, probably a result of their need for water to achieve fertilization. As indicated above, some of these plants manage to exist in desert regions by remaining dormant, making appearances only in brief wet periods.

Epiphytic ferns and fern allies are sporadic, and there are only a few in the flora. The best known in western North America are Polypodium scouleri and Selaginella oregana, both occurring near the Pacific Coast. In the Southeast, Pleopeltis polypodioides forms huge masses on trunks, boughs, and in crotches of trees. Mosslike filmy ferns, especially Trichomanes petersii, produce carpets on damp trunks near the ground in shaded forests. In peninsular Florida, a number of epiphytes belonging to species common in the Antilles are found. The most unusual of these is Cheiroglossa palmata, which hangs from the leaf bases of palmettos (Sabal).

Some epiphytes occur also on rocks, as illustrated by the northern occurrences of Trichomanes petersii and Pleopeltis polypodioides on shaded rock faces and crevices. The reverse also occurs. Normally rock-inhabiting ferns such as Polypodium virginianum and Asplenium rhizophyllum may also be found occasionally as epiphytes in swamps. Both of the common terrestrial spinulose woodferns, Dryopteris carthusiana and Dintermedia, are occasionally found up to 4 m from the ground, growing in tree crotches.

Geography

The pteridophyte flora of North America north of Mexico is much larger (441 species) than that of Europe (about 175 species). But compared with that of eastern Asia (with an estimated 2000 species in China alone), it is very much smaller, perhaps only one-fourth or one-fifth as large. Eastern Asia is much more diverse topographically and climatically than North America, and it probably has the richest temperate-subtropical fern flora in the world. Most of North America is temperate or arctic, and much of the western part is arid.

The mountain systems of North America are simpler than those of China. Wet tropical or subtropical mountains harbor the bulk of the pteridophytes of the globe, and such mountains are absent in North America. One mountain in Borneo, Mt. Kinabalu, may have more species of pteridophytes than all of North America. The tiny Central American country of Costa Rica, with an area slightly less than half the size of Virginia, has nearly 1000 species of pteridophytes, more than double that of the entire flora area. Even the few tiny, mountainous, tropical islands of Hawaii boast a diversity that is 40% of that of the entirety of North America north of Mexico.

To discuss pteridophyte distribution, the continent may be subdivided into floristic provinces: Arctic and subarctic (Tundra and Northern Conifer Forest of H.A. Gleason and A.Cronquist 1964, as given in D.B. Lellinger 1985, fig. 3); eastern North America (Eastern Deciduous Forest, extended to include Newfoundland as a continuum); Coastal Plain (same as Gleason and Cronquist); Peninsular Florida (West Indian Province); Great Plains and Prairie (Grassland); Rocky Mountains--Cascade Mountains--Sierra Nevada (Cordilleran Forest); West Coast (Californian, plus the coastal area north to Alaska); and Sonoran (western Texas to Mojave Desert).

Most of the arctic species are circumboreal, and the number of species and their endemism is low (R.M. Tryon 1969). Eastern North America is very rich in species, especially in the Appalachian Mountain system. Some of these species, or at least closely related sister species, are found in Japan and China (M.Kato and K.Iwatsuki 1983; K.H. Shing 1988). Peninsular Florida is a very distinctive area of the continent floristically, but its pteridophytes are mostly at the edge of the range of widespread tropical American species (R.M. Tryon 1969). The Great Plains and Prairie have local areas of interest, usually small isolated elevated regions like the Cypress Hills (Alberta-Saskatchewan) and Black Hills (South Dakota), but otherwise the grasslands mostly are relatively uninteresting pteridologically.

In the Rocky Mountains, Cascade Range, and Sierra Nevada, pteridophytes are often rare and localized, and they occur at progressively higher elevations southward. The same species that occur nearly at sea level in northern Canada and Alaska may turn up at 3500 m at the southern ends of their ranges. Isolated mountain ranges in southern California and Arizona contain many primarily Mexican taxa. The West Coast region lies on the western slopes of the major ranges from California to Alaska and is rich in pteridophytes. Species seem to form overlapping south-north ranges, one species extending so far and overlapping another in the same genus. The diversity is greatest in California and decreases in Canada and Alaska where subarctic and arctic floras abut. Some of the most luxuriant pteridophyte growth in North America occurs as spectacular assemblages near the Pacific Coast, including prominent species of PolystichumDryopteris, and Athyrium, but the number of species is relatively low. It is sometimes difficult to draw a sharp line between the West Coast and the Cordilleran phytogeographic regions.

The demarcation of well-defined pteridophyte floristic areas is confounded by the presence of numerous disjunctions. Disjunctions of several hundred kilometers are common, and some disjunctions are much greater (W.H. Wagner Jr. 1972). A species that is common and dominant in one area may reappear as a highly localized outlier in another. A number of species have their centers in the west or southwest but have sporadic and rare disjunctions in the east; examples are Pellaea wrightiana (disjunct in North Carolina) and Astrolepis sinuata (disjunct in Georgia). Asplenium septentrionale is centered in the Rocky Mountains with outlying populations in West Virginia.

A spectacular series of west-east disjunctions is seen in the scattered isolated populations in the region from Lake Superior to northeastern Canada, including such predominantly western taxa as Botrychium hesperiumAspidotis densaCryptogramma acrostichoidesPolystichum scopulinum, and Thelypteris quelpaertensis. East to west disjunctions include such extreme examples as Cystopteris bulbifera and Asplenium platyneuron, both very widespread and frequent species in the east that reappear as isolated populations in Arizona.

Disjunctions also occur north to south in both the eastern and western mountains, where low elevation northern species climb higher and higher southward and end up very widely scattered on the topmost peaks. Since 1940, for example, plants of the three well-known northeastern moonworts, Botrychium lanceolatum subsp. angustisegmentumBmatricariifolium, and Bsimplex, have been reported farther south from Pennsylvania (where the species are frequent to common) down to southern North Carolina along the Blue Ridge Mountains (where the species are rare and localized at elevations above 1200 m). In the central Rockies, Bhesperium is known from numerous localities, but farther south, in Arizona, records for it exist only from Mount Baldy, at a very high elevation.

Although far fewer than in seed plants, introductions of exotic species of pteridophytes by human commerce have taken place. Naturalized alien taxa are found mainly in southeastern North America, but there are fewer than 30 species altogether. A few species, such as Marsilea quadrifolia, have spread in temperate areas of the East. Among the most prominent of the southern introductions are Thelypteris dentata, which is particularly common in and around greenhouses, and Macrothelypteris torresiana, now widespread in the Southeast. Cyrtomium falcatum is apparently spreading slowly in southern California and in the Gulf states. Of special interest is Pteris vittata, presumably from eastern Asia, which has established itself in Florida, where it has encountered the native analog, Pbahamensis, and crossed with it to form colonies of P. X delchampsii, a vigorous nothospecies, especially in the vicinity of Miami.

Between 20% and 25% of all of the reproductively competent, native pteridophyte species (including all sexual, apogamous, and gemmiferous taxa) are endemic to North America north of Mexico. The province of highest endemism is also the region of the greatest overall diversity, namely the Eastern Deciduous Forest. It contains approximately 50 endemics, twice the number of the West Coast Province and five times that of the southeastern coastal plain. As rich as the Florida hammocks are for ferns, most of the species there are nonendemic, and they are merely outliers of the Antillean flora.

Classification---The Higher Ranks

Classification and phylogeny of the approximately 10,000 species of pteridophytes are controversial because many of the differences are profound, and evolutionary lines have apparently been separate since the time of the earliest land vegetation. The problems of classification involve such questions as how much evolution has occurred, how extensive are the gaps in today's patterns of diversity, and where do connections occur. As to the last, intensive search has failed to reveal convincing connections between the lycopods, the horsetails, the whisk-ferns, and the true ferns. Highly speculative extrapolations have been made, but they are tenuous. Each of these assemblages possesses an ensemble of very distinctive characteristics. As a consequence, these groups have been accorded divisional status by many, including A.Cronquist et al. (1966; see table 12.1), whose classification we adopt. As a comparison, the approximately 250,000 species of flowering plants occupy a single division, Magnoliophyta.

The most evolutionarily advanced pteridophyte taxa are the most difficult to fix in the system because so many characters are involved, and apparently multiple parallelisms and convergences exist. During the past half century more attention has been given to the familial rank. The traditional family Polypodiaceae, which historically encompassed most of the leptosporangiate ferns, has been subdivided into many widely accepted families.

Among the genera that are still widely questioned as to family affinities is Ceratopteris. Is it a member of the Pteridaceae, perhaps a subfamily of it, or does it constitute a separate monotypic family, Parkeriaceae (as treated in this flora)? Some taxonomists (e.g., R.E.G. Pichi-Sermolli 1977) also segregate other genera in the flora of North America, e.g., Botrychium (Botrychiaceae), Cryptogramma (Cryptogrammaceae), Woodsia (Woodsiaceae), and Lomariopsis (Lomariopsidaceae) (R.E.G. Pichi-Sermolli 1977). Even Adiantum, treated here along with numerous other genera of cheilanthoid affinities, is recognized as being rather isolated and is sometimes regarded as the sole genus in Adiantaceae.

Much of the disagreement about family definition involves the group that is most numerous in the world today, the Dryopteridaceae sensu lato, ferns that have basically reniform or peltate indusia and well-developed perispores. Especially controversial are the Thelypteris group and the Asplenium group, both of which have been raised to family status on the basis of petiole anatomy, trichomes, chromosome number, and a general "coherence" in each, as well as an absence of genera linking them to other families. If we compare, on the basis of the actual number of differences, the elements formerly placed in the family Schizaeaceae sensu lato, namely the AnemiaLygodium, and Schizaea groups, we find that the differences among them are so numerous and striking that each should constitute a different family. The same is true also of the HuperziaLycopodium, and Lycopodiella groups of the traditional Lycopodiaceae.

Table 12.1 represents our classification of the pteridophytes, an attempt to bring together somewhat divergent classifications into a single, relatively consistent treatment. For other viewpoints, the reader should consult J.A. Crabbe et al. (1975), K.U. Kramer and P.S. Green (1990), R.E.G. Pichi-Sermolli (1977), and R.M. Tryon and A.F. Tryon (1982).

Classification---Genus to Variety

Historically, many fern genera have been established on the basis of characteristics (e.g., dimorphism, venation, and blade dissection) that we now regard as insufficient. A good example is in the spleenworts, where PhyllitisCamptosorus, and Ceterach have traditionally been separated from Asplenium but are now reunited with it. It has been demonstrated that these segregate "genera" are very similar to Asplenium anatomically and cytologically, and numerous hybrids among them have been found.

Other genera have been subdivided as we learn more about their morphology and tropical relatives. Such an example involves Thelypteris, in the past merged withDryopteris. In this case, anatomy and cytology, plus the total absence of hybrids, have led to recognition of separate genera. They differ in so many characters that most workers place them in separate families, as we do here.

In Flora of North America, we have attempted to maintain a more or less equal level of comparability among the genera. Some of the changes we have accepted (e.g., the generic separation of elements usually included within Athyrium s.l.) are still relatively novel, and in the future they may prove, at least in part, to be untenable. There have been numerous attempts to segregate genera in the past that have had to be reversed. Asplenium and its segregate genera, discussed above, is a good example, as is Woodwardia, from which Anchistea and Lorinseria once were separated. The family Thelypteridaceae has been variously treated in floras and monographs to comprise a single genus or 11 genera in North America. We have opted for an intermediate course, recognizing three genera in the flora. In some groups, e.g., the adiantoid and the dryopteroid ferns in particular, we have possibly accepted too many segregates, while in others, e.g., selaginellas and botrychiums, we have, perhaps, accepted too few. Only further research, plus quantitative evaluations of comparability, can be expected to resolve inconsistencies that may exist.

The most profound change in the taxonomy of North American pteridophytes involved the species concept and took place during the past 30 years, inspired in large part by the use of new evolutionarily informative characters and more objective approaches to taxonomy. This change can best be illustrated for eastern North America by comparing the treatments in the most recent edition of Gray's Manual of Botany (M.L. Fernald 1950) with the treatments given here. The former may be referred to as the "variety/form" school because of the emphasis on those ranks, and the latter the "species/hybrid" school. The former founded species primarily on relative similarity of morphology; the latter likewise employs morphology but also stresses anatomy, cytology (especially chromosomes), spore viability, breeding characteristics, and syntopic cooccurrence. The varietal category as used prior to 1950 included a wide conglomeration of elements that we now refer to, in our recent taxonomy, as either distinct species, hybrids, trivial forms, or geographical subspecies (W.H. Wagner Jr. 1960).

The category of variety was also used by authors of the variety/form school when they were in doubt as to whether the entity was a distinct species or not; its use tended to be a "trial balloon" without necessarily committing the author to recognizing the entity as a species. Dryopteris intermedia was made a variety of D. carthusiana (as D. spinulosa), to which it bore a certain resemblance. Diphasiastrum digitatum (as Lycopodium) was made a "variety" of D. complanatum. The sterile hybrid, Equisetum X ferrissii (E. laevigatum X hyemale) was construed as a "variety" of one of the parents, E. hyemale. The fertile hybrid, Dryopteris clintoniana (D. cristata X goldiana) was treated as a "variety" of D. cristataBotrychium oneidense was made a "form" of B. dissectum, and B. minganense a "form" of B. lunaria. Numerous other examples could be cited.

The Gray's Manual treatment in 1950 included 120 named species, 159 additional varieties and forms, and 12 named and recognized hybrids. Forty years later, those ratios have been drastically changed, as seen in the treatment herein. The number of species in eastern North America has been increased by approximately 40%, the number of varieties and forms has been reduced by 90%, and the number of named hybrids has been increased 400%, obviously a dramatic change in the systematics of pteridophytes. Pioneers in the presentation of the current taxonomy have been J.T. Mickel (1979) and D.B. Lellinger (1985), who brought together much of the modern monographic literature in their manuals of North American pteridophytes. More recently, W.J. Cody and D.M. Britton (1989) have provided a similar manual for Canadian pteridophytes. All of the problems have not been solved, however. The relationships within the Athyrium filix-feminacomplex are still not fully understood. Whether the mainly tropical Pteridium aquilinum var. caudatum is conspecific with the temperate northern hemisphere var. aquilinum is still debated.

It is especially gratifying to note that recent floristic accounts have, for the most part, been abundantly supported by new comparative characters such as chromosomes, spore wall structure, phenolic compounds, isozymes, and especially, closer observations of the behavior of populations in the wild. The role of hybridity in forming nothospecies (hybrid species) and secondary species is now better understood. Geographical segregates of species are treated as subspecies if the differences are numerous and as varieties if the differences are few, although there is, as would be expected, some arbitrariness in this regard.

Some authors occasionally use one or the other of these infraspecific categories to accommodate different cytological forms, such as diploid and tetraploid populations that differ hardly or not at all morphologically. These "subspecies" or "varieties" may coexist syntopically without successful interbreeding (e.g. Asplenium trichomanes subsp. trichomanes, 2n=72, and subsp. quadrivalens, 2n = 144). The category of form is usually applied to variants that differ in a single character (e.g., glandular form of Dryopteris expansa, dwarf form of Botrychium multifidum, forked form of Pellaea atropurpurea, or sterile-fertile frond intermediates of Onoclea sensibilis). It is now more clearly understood that there are few scientific reasons for recognizing such plants taxonomically. Horticultural or other reasons may exist for doing so.

To summarize the changes in classification at the generic rank and below, we provide the following figures. Approximately 75 species in the flora have undergone a name change as compared with the most recent comprehensive treatment of Canadian and United States pteridophytes by D.B. Lellinger (1985). This is, in part, a reflection of our changing generic concepts in ferns and allies, particularly in the lycopodioid and cheilanthoid assemblages. In addition, 58 taxa have been added to the flora since Lellinger's work. Of the additions, 29 are newly described taxa, mostly species. The total number of these recent changes approaches 30% of the pteridophyte flora!

The Flora of North America project itself has provided the impetus for many of these changes. Projects already underway when FNA began were accelerated so that the work could be incorporated into the flora. Recent studies have focused on taxonomically difficult genera in which allopolyploid and cryptic speciation (C.A. Paris et al. 1989) have occurred, e.g., IsoëtesBotrychiumPellaeaWoodsia, and Polypodium. To resolve problems in these genera and others, taxonomists have used cytotaxonomic and electrophoretic methods as well as more rigorous morphometric comparisons.

We view the changes described herein as a natural progression and growth in our understanding of complex relationships, changes made possible by new techniques and a more consistent application of older methods. As a result of these new studies, we believe that systematic pteridology is now a more viable and exciting area of endeavor than at any time in its history.

Interspecific Hybrids

There are approximately 100 species of hybrid origin in the flora, over 20% of the total. The role of interspecific hybridization in the diversity of North American ferns was grossly underestimated in the first half of the twentieth century. An early pioneer in documenting hybridization was Margaret Slosson (1902), who resolved the origin of Asplenium ebenoides by bringing together gametophytes of A. platyneuron and A. rhizophyllum, work later confirmed by K.S. Walter et al. (1982). Subsequently, R.C. Benedict, who was interested in fern genetics, especially in Nephrolepis, and E.T. Wherry, who was interested in fern ecology, especially in Asplenium, directed professional and amateur attention to reticulate fern evolution, culminating in major reinterpretations of a number of genera. The advent of chromosome studies (I.Manton 1950) refined the experimental methods, and isozyme analysis has substantiated hybrid origins in numerous plants. It is now widely accepted that the study of nothospeciation, i.e., the origin of taxa by hybridization, is an essential pursuit in pteridology.

Only closely related species can be expected to hybridize, normally only members of the same genus or subgenus. In North America, hybridization is frequent between members of Equisetum subg. HippochaeteHuperziaDiphasiastrumLycopodiellaBotrychium subg. BotrychiumPolypodiumCheilanthesPellaeaDryopterisPolystichum, and Asplenium, but it occurs in many other genera as well. Where extensive interspecific networks have been generated by hybridization in nature, we give diagrams summarizing the relationships in the generic treatments.

Some hybrid combinations are extremely rare and represented by only solitary or few individuals in a given locality. Examples are Dryopteris carthusiana X goldianaand Asplenium platyneuron X ruta-muraria (= A. X morganii). Other hybrids are common, e.g., Diphasiastrum sitchense X tristachyum (= D. X sabinifolium),Equisetum hyemale X laevigatum (= E. X ferrissii), and Dryopteris carthusiana X intermedia (= D. X triploidea). Common and/or conspicuous hybrids are often designated with a name in binomial form (rather than by the hybrid formula name).

Both sterile and fertile hybrids may display vigorous growth (and make excellent garden or greenhouse plants). Fertile hybrids produce normal spores and pass through a normal alternation of generations. They are capable of building up large populations either by vegetative or sexual means as discussed previously. If one wishes to distinguish sterile and fertile hybrids with the same parentage, the following convention (proposed by C.R. Werth and W.H. Wagner Jr. 1990 as an amendment of the International Code of Botanical Nomenclature) can be used. Hybrids that are sterile (sexually incompetent) are indicated by the X sign (Asplenium X bradleyi 2x form); if fertile, the times sign is placed in square brackets (A. [X] bradleyi, 4x sexual; A. [X] heteroresiliens, 5x, apogamous). Many hybrids, such as Dryopteris goldiana X ludoviciana (= D. [X] celsa or Asplenium montanum X rhizophyllum (= A. [X] pinnatifidum), were designated as normal species long before their hybrid origin was suspected.

Hybrid species (nothospecies) can be placed in two categories, depending on their reproductive system. Most often, homoploid hybrids (with the same chromosome number as the putative parents) are sterile. The malformed spores can be readily observed, even from herbarium sheets (W.H. Wagner Jr. et al. 1986). The spores are variable in size and shape, many of them obviously collapsed or otherwise distorted. Usually only a few out of thousands of such spores will germinate in culture.

Homoploid fertile hybrids are rare, and they are known in North America only in two genera of Lycopodiaceae, Diphasiastrum and Lycopodiella. In these hybrids, the chromosome behavior at meiosis appears to be entirely normal, as are the resultant spores. Why these homoploid fertile hybrids do not simply swamp the parental species is unknown, but the intermediates are distinctive, as a rule, and they are readily separated from their parents. Backcrossing has not been observed, and thus introgression has not been demonstrated, even though many of the hybrids, especially certain ones in Diphasiastrum, are frequent or even common.

In some predominantly sterile hybrids, large unreduced spores (with double the chromosome complement) are formed; these are theoretically capable of germinating to produce gametophytes. It is hypothesized that such gametophytes may be able to generate new sporophytes directly by apogamy (V.M. Morzenti 1962). This may explain why such "sterile" hybrids as Asplenium X trudellii (A. montanum X pinnatifidum) and Dryopteris X boottii (D. cristata X intermedia) are so common in nature.

The study of fern hybrids raises an unanswered question: Why have some sterile hybrids undergone chromosome doubling to restore fertility (allopolyploidy) while others have not? A number of sterile hybrids have never been known to double, even when they are very common, as are the numerous hybrids involving Dryopteris marginalis. Why has Asplenium X ebenoides (A. platyneuron X rhizophyllum) doubled only at one locality in Alabama, whereas A. [X] bradleyi (A. montanum X platyneuron) has practically always doubled? We still don't know.