Tom A. Ranker

A community-derived classification for extant lycophytes and ferns

Phylogeny has long informed pteridophyte classification. As our ability to infer evolutionary trees has improved, classifications aimed at recognizing natural groups have become increasingly predictive and stable. Here, we provide a modern, comprehensive classification for lycophytes and ferns, down to the genus level, utilizing a community‐based approach. We use monophyly as the primary criterion for the recognition of taxa, but also aim to preserve existing taxa and circumscriptions that are both widely accepted and consistent with our understanding of pteridophyte phylogeny. In total, this classification treats an estimated 11 916 species in 337 genera, 51 families, 14 orders, and two classes. This classification is not intended as the final word on lycophyte and fern taxonomy, but rather a summary statement of current hypotheses, derived from the best available data and shaped by those most familiar with the plants in question. We hope that it will serve as a resource for those wanting references to the recent literature on pteridophyte phylogeny and classification, a framework for guiding future investigations, and a stimulus to further discourse.

Phylogeny and evolution of grammitid ferns (Grammitidaceae): a case of rampant morphological homoplasy

We conducted phylogenetic analyses of the fern family Grammitidaceae using sequences from two cpDNA genes and from morphological characters. Data were obtained for 73 species from most recognized genera in the family. Thegenera Adenophorus, Ceradenia, Calymmodon, Cochlidium, Enterosora, and Melpomene were each strongly supported as being monophyletic. Other recognized genera that were not supported as monophyletic included Ctenopteris, Grammitis, Lellingeria, Micropolypodium, Prosaptia, and Terpsichore. Several previously unrecognized clades were identified, some of which are characterized by distinctive morphological features. Analyses of the distribution of morphological character states on our inferred phylogeny showed extremely high levels of homoplastic evolution for many different characters. Homoplasy for morphological characters was considerably greater than for molecular characters. Many of the characters that exhibited high levels of convergent or parallel evolution across the phylogeny are features that have been commonly used to circumscribe genera in this group (e.g., leaf blade dissection, various rhizome scale characters, and glandular paraphyses). Conversely, some of the characters that exhibited relatively low levels of homoplasy have either not been regarded as having taxonomic value or have been ignored (e.g., root insertion, rhizome scale sheen). Our data support a New World origin of Grammitidaceae, with Old World taxa generally being more evolutionarily derived. Several clades are either primarily Neotropical or primarily Paleotropical but also have a few members distributed in the opposite hemisphere. Thus, we postulate multiple, independent dispersal and colonization events in several lineages.

Genetic Evidence for Allopolyploidy in the Neotropical Fern Hemionitis pinnatifida (Adiantaceae) and the Reconstruction of an Ancestral Genome

Genetic evidence from enzyme electrophoresis and restriction site analyses of cpDNA was employed to explore the origin of the tetraploid Hemionitis pinnatifida by determining the role of diploid species of Hemionitis in contributing to the tetraploid genome. The results clearly supported the hypothesis that H. pinnatifida is an allotetraploid with H. palmata as one of its diploid progenitors. The other diploid parent was not identified among the existing species of Hemionitis and is either extinct or exceedingly rare. Genetic identity analyses compared the extant diploid genomes with the divergent diploid genomes combined in tetraploids. One of these divergent genomes was shown to be a good approximation of that of H. palmata, and the other is a reconstruction of that of the missing diploid parent. These analyses demonstrated that the missing ancestor was genetically more similar to H. palmata than either of these species is to the other diploid species surveyed. Such greater genetic similarity may have been important in allowing hybridization between the two diploid progenitors of H. pinnatifida. The data also suggest that tetraploid individuals may have been synthesized de novo at least five times among the three populations surveyed.

Evolution of high genetic variability in the rare Hawaiian fern Adenophorus periens and implications for conservation management

Abstract Adenophorus periens is a rain forest-dwelling epiphytic fern and is known only from one population on the island of Hawaii (Kahaualea) and from a few scattered individuals on the islands of Kauai and Molokai. A genetic survey was conducted of the Kahaualea population employing protein electrophoresis. The individuals sampled expressed extremely high levels of allozymic variability compared to two more widespread congeneric species ( A. tamariscinus and A. tripinnatifidus ). All individuals examined were produced via sexual recombination. Given the extremely young volcanic substrate upon which the Kahaualea forest has developed, the historical factors leading to the high genetic variability of A. periens are somewhat obscure. The outcrossed mating system of this species may play a key role in the maintenance of genetic diversity in combination with its perennial life cycle. Because the lack of neighboring populations precludes the influx of new genetic variability via gene flow, the population must rely solely on the appearance and incorporation of new mutations as sources of new variability. The data suggest that this genetically diverse population is not likely to suffer genetically from minor reductions in population size over the short term. The most obvious potential threats to A. periens are those affecting the demography of the population and include habitat destruction by lava flows, death of host trees through stand-level dieback, chance loss of individuals by a variety of mechanisms, and the adverse effects of gaseous emissions from volcanic activity and geothermal development. The State of Hawaii should undertake an artificial propagation and transplantation program to establish populations in less threatened, suitable habitats.

Antheridiogen and Natural Gametophyte Populations

could control the sex expression of fern gametophytes was made in the laboratory. In 1950, Dopp published a paper in which he showed that old media or aquatic extractions from media containing female gametophytes of bracken (Pteridium aquilinum) induced antheridia formation in young gametophytes of its own species or those of the male fern (Dryopteris filix-mas). This pioneer work was supplemented by two further publications in 1959 and 1962. In the United States, it was mainly Naf (for review see Nif, 1979) who started studies on antheridiogen in the late 1950s. Some of the most interesting questions that arose were: Does a selected fern taxon produce or react to antheridiogen? Are antheridiogen-systems widely distributed? What is the sex expression of isolated gametophytes? How much antheridiogen is required to induce antheridia? What happens to a culture of young sterile gametophytes when antheridiogen of a female or a meristic gametophyte is added? Do meristic gametophytes become insensitive to antheridiogen? Since D6pps initial work, a large number of publications have appeared. Nearly all of these have been based on laboratory investigations and our knowledge of antheridiogen has increased strikingly. It has been shown that at least three main classes of antheridiogens occur (Nif, 1979), antheridiogen A (bracken-antheridiogen), antheridiogen B (produced by members of the family Schizaeaceae) and antheridiogen C (produced by species of the genus Ceratopteris). Although the chemistry of only antheridiogen B has been studied in detail (Corey & Myers, 1985; Nakanishi et al., 1971; Yamane et al., 1979) many similarities among the various types of antheridiogens can be seen from the standpoint of physiology. Antheridiogens are all water soluble, are active in very low concentrations, and are formed by large, meristic gametophytes. At present, the term pheromone is preferred over the term hormone when speaking of antheridiogens (Nif, 1979). Although much has been accomplished, many questions may still be pursued through laboratory experiments. In contrast to the multitude of laboratory-oriented studies, only a few publications have focused on the occurrence and importance of antheridiogen in natural fern populations. The first publication giving evidence for the natural role of antheridiogen was that of Tryon & Vitale (1977). By mapping gametophytes on natural sites Tryon & Vitale (1977) could show that there was

Allopolyploid origin and population genetics of the rare orchid Spiranthes diluvialis.

The process of becoming and the attributes of being polyploid play a major role in the development and maintenance of genetic variation in allopolyploid species. A genetic survey employing protein electrophoresis on 12 populations of S. diluvialis, as well as on populations of eight congeneric species, was conducted to assess the putative allopolyploid origin of S. diluvialis and to determine the genetic variability within and among populations. Genetic identity values indicated S. diluvialis was more similar to S. magnicamporum (0.619) and S. romanzoffiana (0.727) than to any of the other congeneric species assayed. Similar to most allopolyploids, S. diluvialis showed high levels of fixed, or nearly fixed, heterozygosity and a high percentage of polymorphic loci (57.1-71.4%). The mean number of alleles per polymorphic locus in populations of S. diluvialis (2.6-3.3), however, was similar to mean values for both animal-pollinated, outcrossing, diploid species, and geographically restricted, diploid species (2.6 and 2.5, respectively). Genetic divergence among populations (mean Fst = 0.083) was low, leading to relatively high estimates of interpopulational gene flow (mean Nm = 5.41). Thus, each population harbors most of the genetic variability found within the species. The genetic variation observed within S. diluvialis supports the occurrence of at least two separate hybridization events giving rise to S. diluvialis.

CLONAL DIVERSITY IN ALPINE POPULATIONS OF POLYGONUM VIVIPARUM (POLYGONACEAE)

Asexual reproduction is extremely common among arctic and alpine species, and successful sexual reproduction may be rare. In asexual populations, the absence of segregation and recombination predicts reduced levels of genetic variation. Thus, genetic diversity within arctic and alpine plant populations may be quite low. Allozyme analysis of three alpine populations of Polygonum viviparum, a common herbaceous perennial with no observed sexual reproduction, revealed surprising levels of genetic diversity. There were 23 unique clones within a sample of 150 ramets. Few clones were large or distributed among all three sample populations; the majority of clones was rare (n ≤ 5) and unique to a single population. Genotypic diversity differed among the three sample populations. The number of clones and measures of diversity and evenness were much lower in the fell-field population than in wet meadow and dry meadow populations. Overall, genotypic diversity and structure of alpine P. viviparum are similar to other clonal species in which sexual reproduction is rare, and they are similar to the average for clonal species in general.

CONSERVATION GENETICS OF THE ENDANGERED ENDEMIC HAWAIIAN GENUS BRIGHAMIA (CAMPANULACEAE)

The endemic Hawaiian genus Brighamia (Campanulaceae) comprises two federally endangered, morphologically similar species, B. insignis from Kaua`i and Ni`ihau and B. rockii from Moloka`i. To assist the design of conservation management programs for these taxa, isozyme analyses were performed to assess the levels of genetic diversity at the population and species levels, including comparisons within and among seven natural populations and one ex situ collection each of B. insignis and B. rockii. Our sampling (N = 80) represents ~41% of all known individuals in the wild. Isozyme analyses revealed levels of genetic variation comparable to those reported for other Hawaiian flowering plant taxa but low levels of genetic variation at the population and species levels when compared to flowering plants in general. Ex situ individuals (N = 61) were genetically representative of natural populations and hence may appropriately serve as stock for population augmentations. The two morphologically similar Brighamia species were highly distinct genetically. The combination of morphological and ecological similarity with allozymic dissimilarity observed in Brighamia is unique among the Hawaiian taxa studied to date.

Systematic inferences from spore and stomate size in the ferns

Size of equivalent cells has traditionally been assumed to be constant within species and variable between species. Systematists have commonly measured cells with constant form, such as spores and stomates, as a means of distinguishing species and hybrids in polyploid complexes, since the best known factor determining cell size is ploidy level (Stebbins, 1950). Preliminary hypotheses about polyploid evolution can be generated directly from herbarium specimens based on spore and stomate measurement data. The first evidence of a relation between cell size and polyploidy in the ferns was Lawtons 1932 demonstration that prothallial cells, lower epidermal cells, and stomates in induced apogamous races of Dryopteris marginalis and Woodwardia virginica vary directly with ploidy level. Butters and Tryons 1948 work on Woodsia x abbeae included epidermal cell and annulus cell measurements that corroborated data from spore germination experiments to document an instance of somatic autopolyploidization. More recently a considerable body of evidence has been assembled supporting the contention that spore size is related to ploidy level in the ferns. Manton (1950) noted that Vancouver Island Cystopteris fragilis (n = 84) has smaller spores than Swiss Cystopteris alpina (n = 126). Hagenah (1961) postulated a polyploid series in species of Cystopteris based on spore-size measurements. In a monograph of Cystopteris, Blasdell (1963) concluded that it was possible to infer ploidy level from spore size in that genus. Blasdell assigned ploidy levels to cytologically unknown components of polyploid complexes based only on spore size (see Lovis, 1977 for a critique). Wagner (1966) demonstrated that there were two varieties of Gymnocarpium dryopteris: a larger-spored typical variety, which is tetraploid, and a smaller-spored variety disjunctum, which is diploid. Schneller (1974) recently provided an analysis of ploidy level and spore size in the Dryopteris filix-mas group of Europe. He concluded that ploidy and spore size are positively correlated for diploid to pentaploid cytotypes in a closely knit phylogenetic group. In an exhaustive analysis of spore features of northeastern North American Isoetes, Kott and Britton (1983) provided new insight into spore-size variability. Whereas megaspore variation within sporangia of a single plant, between sporangia of a single plant, and between plants of a population was found to be more or less equivalent and negligible, variation between populations of a species was demonstrated to be twice that between plants of a population. Microspores were found to be slightly less variable. Kott and Britton (1983) suggested that

Genetic Diversity in Alpine and Foothill Populations of Campanula rotundifolia (Campanulaceae)

Climatic constraints on insects in alpine environments have important consequences for the biology of their plant mutualists; in particular, reduced insect diversity and activity in alpine plant populations can result in pollinator‐limited seed set and, potentially, in low genetic diversity. However, highly effective pollination by bumblebees in alpine populations can compensate for low visitation rates. In this study we hypothesized that, because of highly effective pollination by bumblebees, alpine populations of Campanula rotundifolia would not experience more frequent cycles of pollinator limitation than low‐elevation populations and would therefore exhibit comparable levels of genetic variability and inbreeding to those found in foothill populations. Enzyme electrophoresis was used to assess genetic variability at nine putative loci in alpine and foothill populations of C. rotundifolia in Colorado. Genetic variability in C. rotundifolia was found to be comparable to that reported for other long‐lived herbaceous perennials. Measures of genetic variability and fixation indices did not differ between high‐ and low‐elevation populations and were consistent with Hardy‐Weinberg expectations. Nonsignificant FST values indicated no genetic differentiation among all populations.