Illinois Natural History Survey bulletin \ Monograph of the Neotropical Fern Genus Polybotrya (Dryopteridaceae) RobbinC. Moran Volume 34, Article 1 November 1987 Monograph of the Neotropical Fern Genus Polybotrya (Dryopteridaceae) RobbinC. Moran Illinois Natural History Survey Illinois Department of Energy and Natural Resources Department of Plant Biology University of Illinois at Urbana-Champaign Illinois Natural History Survey Bulletin Volume 34. Article 1 November 1987 This work is dedicated to my parents, Livia Ann Moran and John Howard Moran. Dr. Robbin C. Moran currently works at the Missouri Botanical Garden in St. Louis. Illinois Natural History Survey, Lorin I. Nevling, Chief A Division of the Illinois Department of Energy and Natural Resources Printed by Authority of the State of Illinois (62997—1,200—11-87) No charge is made for most publications of the Illinois Natural History Survey, and a list of those publications is available upon request. Single copies of most publications are available to anyone request- ing them. Requests for multiple copies should be made in writing and should explain the use to be made of the publications. Address correspondence to the Office of the Chief at the address below. Illinois Natural History Survey Natural Resources Building 607 East Peabody Drive Champaign, Illinois 61820 Citation: Moran, R.C. 1987. Monograph ofthe Neotropical Fern Genus /'o/)*o/r>'<'(Dryopteridaceae). Illinois Natural History Survey Bulletin 34 (I): U1.18. US ISSN 0073-4918 Contents Acknowledgments v Part One: Introduction and Discussion Materials and Methods 1 Taxonomic History of the Genus 3 Ecology 6 Geography 7 Morphology and Anatomy II Chromosome Numbers 27 Cladistic Analysis of the Species 27 Subdivision of the Genus 31 Relationships with Other Groups 34 Part Two: Taxonomic Treatment Notes on the Presentation of Data 37 Description of the Genus 38 Key to the Species of Polybotrya 39 Species Descriptions 43 Polybotrya subgenus Soromanes 1 . Polybotrya serratifolia (Fee) Klotzsch 43 2. Polybotrya polybotryoides (Baker) Christ 46 3. Polybotrya suberecta (Baker) C. Chr 50 4. Polybotrya andina C. Chr 50 Polybotrya subgenus Sorbifolia 5. Polybotrya sorbifolia Kuhn 53 6. Polybotrya fractiserialis (Baker) J. Smith 54 7. Polybotrya crassirhizoma Lellinger 58 8. Polybotrya espiritosantensis Brade 60 Polybotrya subgenus Polybotrya 9. Polybotrya caudata Kunze 60 10. Polybotrya goyazensis Brade 66 1 1 . Polybotrya pubens Martius 68 12. Polybotrya glandulosa Kuhn 71 13. Polybotrya lechleriana Mettenius 71 14. Polybotrya attenuata Moran 74 15. Polybotrya stolzei Moran 76 16. Polybotrya alfredii Brade 78 17. Polybotrya botryoides (Baker) C. Chr 80 18. Polybotrya lourteigiana Lellinger 82 19. Polybotrya pittieri Lellinger 84 20. Polybotrya cylindrica Kaulfiiss 86 2 1 . Polybotrya hickeyi Moran 88 22. Polybotrya puberulenta Moran 90 23. Polybotrya alata Moran 92 24. Polybotrya aequatoriana Moran 94 25. Polybotrya appressa Moran 94 26. Polybotrya altescandens C. Chr 97 27. Polybotrya gomezii Moran 99 28. Polybotrya osmundacea Willd 101 29. Polybotrya cyathifolia Fee 106 30. Polybotrya latisquamosa Moran 108 31. Polybotrya sessilisora Moran 108 32. Polybotrya canaliculata Klotzsch Ill 33 . Polybotrya semipinnata Fee 113 34. Polybotrya speciosa Schott 115 35. Polybotrya pilosa Brade 117 Names of Uncertain Application 119 Excluded Taxa 120 Literature Cited 122 Taxa and Distribution of Polybotrya 1 24 Distribution Maps 124 Index to Collectors' Numbers 132 Index to Taxonomic Names 137 Acknowledgments No one helped me more than the five pteri- dologists who carefully reviewed every page of this monograph: Dr. David B. Lellinger, Smithsonian Institution, Washington, D.C.; Dr. Alan R. Smith, University of California, Berkeley; Mr. Robert G. Stolze, Field Mu- seum of Natural History, Chicago, Illinois; Dr. Rolla M. Tryon, Harvard University, Cambridge, Massachusetts; and Dr. Warren H. Wagner, Jr., University of Michigan, Ann Arbor. Not only did they spend long hours reading my rough manuscripts, but they also patiently answered my numerous queries about the intricacies of fern taxonomy. My first field trip—to Costa Rica—was made on funds awarded by the Tinker Foun- dation. Doctoral Dissertation Improvement Grant # 83-06990 from the National Science Foundation provided most of my financial support, primarily for three additional col- lecting trips in Latin America and for a semes- ter of study at Harvard University. Two people were particularly helpful dur- ing my fieldwork in Latin America. First, Mr. Luis D. Gomez P. of the Museo Nacional de Costa Rica, San Jose, gave me lodging in his house and extended numerous other cour- tesies while I worked in Costa Rica. Second, Professor Francisco Ortega, UNELLEZ, Guanare, Venezuela, gave me the use of his herbarium and plant-drying facilities and took me on collecting trips through the Andes of western Venezuela. I thank them both for their hospitality and for sharing their knowl- edge of ferns. During my semester of study at Harvard University, Dr. Alice F. Tryon was extremely helpful in showing me how to obtain high quality scanning electron microscope photo- graphs of spores. Dr. Rolla M. Tryon helped with nomenclatural, systematic, and other matters. I learned much about pteridophyte biol- ogy from two summer field courses with Drs. Warren H. Wagner and Rorence S. Wagner, the first at Flathead Lake Biological Station, Montana, and the second at Mountain Lake Biological Field Station, Virginia. I thank them for these memorable summers. Dr. Kenneth R. Robertson of the Illinois Natural History Survey, Champaign, was my major advisor and provided much help and encouragement. Others who served on my doctoral committee were Dr. J. Leland Crane, Illinois Natural History Survey, Champaign; Drs. Johannes M. J. de Wet, Zane B. Carothers, and Malcolm L. Sargent, University of Illinois at Urbana-Champaign; Dr. Rolla M. Tryon, Harvard University, Cambridge; and Dr. Warren H. Wagner, Uni- versity of Michigan, Ann Arbor. Dr. Tom L. Phillips, University of Illinois at Urbana- Champaign, provided numerous helpful com- ments, esp)ecially on the Ecology and Geog- raphy sections. 1 completed most of the work for this monograph in the herbarium at the Illinois Natural History Survey, and I am greatly in- debted to that institution for its support and for the use of its facilities. Several biologists at the Survey helped me in various ways. Mr. John Taft and Ms. Mary Kay Solecki provid- ed useful discussion and patiently tested my keys. Dr. David Swofford was helpful with the cladistic analysis. Mr. Bill N. McKnight was a constant friend and critic throughout the study and assisted with the reproduction of the illustrations. 1 am also indebted to the editorial staff at the Survey, especially Patty Duzan and Eva Steger who typeset the man- uscript, Audrey Hodgins who edited it. and Molly Scott who helped with the production. I am grateful to the directors and curators of the following herbaria for the large quan- tities of valuable material they made available to me, often for a considerable period of time. The abbreviations used in the text follow the names of the herbaria; Arnold Ariwretum, Harvard University (A), Cambridge, Mas- sachusetts; Herbarium Jutlandicum, Univer- sity of Aarhus (AAU), Risskov, Denmark; Herbarium Amazonense, Universidad Na- cional de la Amazonia Peruana (AMAZ), Iquitos, Peru; Botanischer Garten und Botanisches Museum (B), Berlin, Federal Republic of Germany; British Museum of Natural History (BM), London, England; Herbier (CAY), Cayenne Cedex, French Guiana; Herbario Nacional Colombiano, Museo de Historia Natural, Universidad Na- cional (COL), Bogota, Colombia; Herbario Nacional de Costa Rica, Museo Nacional (CR), San Jose, Costa Rica; John G. Searle Herbarium, Field Museum of Natural History (F), Chicago, Illinois; Conservatoire etJardin botaniques de la Ville de Geneve (G), Swit- zerland; Gray Herbarium, Harvard Univer- sity (GH), Cambridge, Massachusetts; Rijksherbarium (L), Leiden, Netherlands; In- stituto Miguel Lillo de la Fundacion Miguel Lillo (LIL), Tucuman, Argentina; Facultad de Ciencias Naturales y Museo, Division Plantas Vasculares, Universidad Nacional de La Plata (LP), La Plata, Argentina; Herbario Nacional de Bolivia, Universidad Mayor de San Andres (LPB), La Paz, Bolivia; Her- barium of the University of Michigan (MICH), Ann Arbor, Michigan; Missouri Botanical Garden (MO), St. Louis, Missouri; New York Botanical Garden (NY), Bronx, New York; Museum National d'Histoire Naturelle (P), Paris, France; Academy of Natural Sciences of Philadelphia (PH), Philadelphia, Pennsylvania; Herbario Uni- versitario, Universidad Nacional Experimen- tal de los Llanos Occidentals "Ezequiel Zamora" (PORT), Portuguesa, Venezuela; Herbario del Instituto de Ciencias Naturales, Universidad Central (Q), (Juito. Ecuador; In- stituto de Ciencias, Pontificia Universidad Catolica del licuador (C?CA), Quito, Ecua- dor; Jardim Botanico do Rio de Janeiro (RB), Rio de Janeiro, Brazil; Department of Bot- any, University of California (UC), Berke- ley, California; United States National Her- barium, Smithsonian Institution (US), Wash- ington, D.C.; Herbario San Marcos, Museo de Historia Natural, Universidad Nacional Mayor de San Marcos de Lima (USM). Lima, Peru; Instituto Botanico (VEN), Caracas, Venezuela; Pringle Herbarium, University of Vermont (VT), Burlington, Vermont; Institut fiir systematische Botanik der Universitat Zurich (Z), Zurich, Switzerland. This work was originally part of a disser- tation submitted to the Graduate College of the University of Illinois at Urbana-Cham- paign in partial fulfillment of the require- ments for the degree of Doctor of Philosophy in plant biology. Monograph of the Neotropical Fern Genus Polybotrya (Dryopteridaceae) Robbin C. Moran Part One: Introduction and Discussion The need still is for more monographic work on tropical species, which represent the great majority of all ferns. R.E. Holttum(1982) The genus Polybotrya (Dryopteridaceae) in- cludes 35 species of neotropical ferns. It is distinguished by 1 ) strongly dimorphic leaves with fertile leaves that resemble skeletons of the sterile, photosynthetic ones; 2) usually high-climbing stems that are covered with scales; and 3) a unique stem anatomy with 5 to 12 circularly arranged meristeles, each sur- rounded by a black sclerenchymatous sheath, with numerous tiny leaf traces arching be- tween adjacent meristeles. The center of di- versity of the genus is the Andes, where 23 sjjecies occur, 12 of which are endemic. Tlie coastal mountains of southeastern Brazil are notable because they contain 5 species, all endemic. The range of Polybotrya is from Chiapas, Mexico, southward through Central America; the West Indies; northern South America southward along the Andes to Bolivia and Paraguay and eastward to the Guiana Highlands; the Amazon River basin and the Matto Grosso; and southeastern Brazil. Species of the genus typically inhabit wet, shaded, primary tropical forests from sea level to 2500 m, most often occurring at middle altitudes between 500 and 2000 m. Polybotrya is divided into three subgen- era: 1 ) Soromanes , leaves simply pinnate and veins anastomosing; 2) Sorhifolia, leaves simply or twice pinnate and veins free, close, and parallel; and 3) Polybotrya, leaves de- compound and veins free. Polybotrya cer- vina, a species usually included in Polybot- rya, is removed to the monotypic genus Ol- fersia (Moran 1986). Polybotrya is related to dryopteroid genera such as Arachniodes, Cyclodium, Maxonia. Olfersia, and Poly- stichopsis. Carl Christensen, the father of modem fern taxonomy, observed (1916) that Polybotrya may have arisen from Maxonia because both have high-climbing stems and strongly dimorphic leaves. The morphologi- cal and anatomical evidence presented here suggests that Polybotrya may have evolved instead from a Cyclodium-Wkt ancestor. I chose Polybotrya for study because two aspects of the genus immediately intrigued me: its strongly differentiated sterile and fer- tile leaves and its long, creeping hemiepiphy- tic stem (Fig. 1 ). Since these features evolved separately in unrelated fern genera, studying Polybotrya might well provide insight into broader questions of fern evolution. Polybot- rya was suited to monographic study because the number of species (35), all of which are neotropical, was manageable. Finally, no previous monographic work had been done on Polybotrya and many problems of nomenclature and identification remained to be solved. Materials and Methods This monograph is based on the study of about 25(X) herbarium sheets, which repre- sent approximately 8(X) individual collections from 30 herbaria (see acknowledgments). Unfortunately, Polybotrya at Kew could not be examined because their pyolicy precludes loaning specimens for use by graduate stu- dents (G.L. Lucas, in litt. 1983; pers. comm.). Kew, however, did send color slides of several critical type specimens. Illinois Na iural History Survky Vol.34. Art. 1 FiGliRE 1 . Habit sketches for two species of Polyhoina. Above. P. cr(issirhi:om(2. a climbing species; below. P. sorbijolia. a teirestnal species. Note that the fcnile leaves an; skeleionlikc and that iheir orientation is more erect than that of the spreading sterile leaves. November 1987 Monograph of Polybotra I spent seven months observing and col- lecting 1 8 of the 35 species of Polyhotrya in Costa Rica, Ecuador, Peru, and Venezuela and was able to make such important obser- vations about the biology of the species as presence of aerophores and mucilage on stems, duration and habit of sterile versus fertile leaves, and variation within and be- tween populations. During field work, her- barium, cytological, and anatomical mater- ials were also collected. Anatomical cross sections were prepared by freehand sectioning, and staining was rarely needed to determine cell types or tissue layers. Leaf segments were cleared by soak- ing them in 10 percent NaOH in a warm oven for 5 to 6 days. Several changes of clearing solution were usually needed to replace so- lution that had become darkly stained by phenolic substances. Since this procedure did not remove all the dark color from the leaf, the segments were placed in full-strength Clorox bleach from 10 to 60 minutes. This treatment usually rendered even the most stubborn leaves translucent for microscopic study. Taxonomic History of the Genus Polyhotrya was first described by Willdenow in his fourth edition of Linnaeus' s Species Plantarum (1810), as distinguished by "Cap- sules sessile, globose, aggregated in naked paniculate spikes. Nonindusiate" [my transla- tion]. Polyhotrya osmundacea, the sole species described in the new genus, received its specific epithet because of the fancied re- semblance of its fertile leaf with the fertile leaf apex oi Osmunda regalis. Since strongly contracted, nonindusiate fertile leaves were considered the principal feature of the new genus, other species with similar leaves were eventually placed in Polyhotrya. Since fully differentiated fertile leaves have arisen sepa- rately along many phyletic lines, this proce- dure resulted in a highly unnatural, poly- phyletic genus. Species placed in Polyhotrya at one time and then excluded are represented today in as many as 12 genera (see Excluded Taxa). Past Generic Concepts Each of the nineteenth-century pteridologists who wrote about Polyhotrya had a slightly different concept of the genus (Table I). Biume (1S28) placed many unrelated acros- tichoid ferns, including species of Loma- gramma. Stenosemia, and Bolhilis, in Poly- hotrya. SchotI (1834-1836) later restricted Pohhotrxa to the American species allied to P. osmunducea. The subsequent recognition of the Asian acrostichoids as distinct from the American genus Polyhotrya testifies to Schott's keen perception. Although Schott's classification was not immediately adopted by his colleagues, it was eventually revived by Smith (1875) and Christensen (1934). PresI (1836) relied heavily on venation patterns and anatomical characters in defining his genera and did not accept Schott's circum- scription of Polyhotrya. He merged species of Bolhitis sect. Egenolfianae with Polyhotrya because both had free veins, but he removed the anastomosing-veined species placed in Polyhotrya by Blume (1828). Smith (1841) largely followed this classification in his ar- rangement of fern genra. Fee (1845) was the first pteridologist to try to make sense out of the increasing number and diversity of acrostichoid ferns. He included the species related to P. osmim- dacea in his subgenus Eupolybotrya and created two subgenera of Polyhotrya (see Excluded Taxa) in which he put many species now placed in Bolhitis, Blechnum, Lomariop- .v/.v, and Teraiophyllum. Fee also erected the segregate genus Soromanes for species of Polyhotrya with anastomosing veins and 1- pinnate leaves. In my treatment. Soromanes is a subgenus of Polyhotrya. The German pteridologist Mettenius had a wide circumscription of Polyhotrya. He de- scribed several new species in the genus, some of which are here retained. The major- ity, however, have been placed in Bolhilis, Eluphoglossiiin , and Leptochilus. Hooker (1864; 194-195) and Baker in Hooker and Baker ( 1 874: 399ff . ) subsumed both Polyhotrya and Soromanes as subgenera in their large, eclectic genus Acrostichum. Illinois Natural History Survey Vol.34. An. 1 which consisted of any fern with acrostichoid sori. This decision was essentially a reversion to Swartz's (1806) concept of Acrostichum, and although both genera were recognized as subgenera, Polybotrya continued to include unrelated species that have since been placed in Atalopleris, Elaphoglossum, Psomio- carpa, and Teratophyllum. Hooker and Baker's classification was accepted by sev- eral pteridologists, especially by flora writ- ers, until the hegemony of Hooker's concepts was supplanted by the more natural views of Smith (1875). Smith accorded generic status to Polybotrya, restricting it to P. osmundacea and its immediate allies; however, he kept Soromanes as a distinct genus. Unfortunately, Smith's views were not immediately accepted by pteridologists. Christ (1897) and Diels (1899) classified many of the acrostichoids with free veins in Polybotrya, a decision that again resulted in an amalgamation of unrelated sfjecies. Chris- tensen ( 1905) in Index Filicum used Polybot- rya at the rank of genus, but it had essentially the same circumscription as it had had under Hooker. Unlike the writers cited above, he placed Olfersia cervina in Polybotrya. His classification was used by Schumann (1915) in her important work on the vascular supply in fertile leaves of acrostichoid ferns. Chris- tensen (1934) later removed the discordant species from Polybotrya and placed most of them in the genera where they are found today (Atalopteris , Egenolfia. Lomagramma. and Psomiocarpa) . Copeland's Genera Filicum (1947) adopted Christensen's earlier ( 1905) concept. Pichi-Sermolli (1977) treated Polybotrya. Table 1. Comparative treatment oi Polybotrya. Reference Polybotrya subgenus Soromanes* Polybotrya subgenus Sorbifolia* Polybotrya subgenus Pol\holr\a* Genera formerh included in Pol\botr\a WilidenowdSlO) Kault'uss{1824) Blume(1828) Schott( 1834-1836) Presl(l836) Smith (1841) Fee (1845) Hooker (1864) Smith (1875) Christcnscn(1905) Christcnscn(l934) Copeland(1947) Pichi-Scniiolli(1977) Tryon& rryon(1982) Soromanes Acro.^tichutn subgenus Soromanes Soromanes Pohholrxa Polybotrya Polybotrya Soromanes Polxbotr\a Acrostichum subgenus Polybotrya Polybotrya Pohbotrxa Polybotrya Polybotrya Polybotrya Pol\botr\a Polybotrya Polybotrya Polybotrya Polybotrya Polybotrya Polybotrya Polybotrya subgenus Eupolybotrya Acrostichum subgenus Polybotrya Polybotrya Pohbotrxa Polybotrya Polybotrya Polybotrya Pohbotrxa Olfersia Bolbitis. Lomma- gramma. Stenosemia Bolbitis Bolbitis Blcchnum. Bolbitis. Lomariopsis. Teratophyllum .Atalopleris. Psomiocarpa. Teratophyllum .Atalopteris. Bolbitis. Olfersia. Psomio- carpa. Teratophyllum Olfersia Olfersia * refers to the treatment of the cenus in this work. November 1987 Monograph of Poi.ybotra Soromanes, and Olfersia as distinct genera but placed Soromanes and Polyhotrya on separate branches of his phylogenetic dia- gram. Tryon and Tryon (1982) are the most recent authors to discuss Polybotrya and re- lated genera; their concept is like that used in the present work, except that they include Olfersia cervina in Polybotrya. In the present monograph, 1 restrict Polybotrya to the species allied closely to P. osmundacea, make Soromanes a subgenus of Polybotrya. and keep Olfersia cervina. which has been placed in Polybotrya by many recent pteridologists, in its own monotypic genus. 1 recognize 35 species of Polybotrya. all of which are neotropical and 10 of which are new. Work at the Species Level The species of Polybotrya have received less study over the years than those of most other fern genera. Fee (1845), Hooker (1864), and Hooker and Baker (1874) have been the only monographers of the genus. Although they attempted to identify all of the then-known species of Polybotrya. they did not always see types, overlooked several published names, put some species in synonymy with- out adequate study, and did not make detailed observations on the genus because they were studying hundreds of other ferns at the same time. Despite these shortcomings, the works of these three men have been the best source for identifying specimens of Polyhotrya in many regions of tropical America. The difficulty of identifying specimens of Polybotyra has been partially alleviated by local or regional floras, but these cover only a limited portion of the geographic range of the genus and are often incomplete. In many of these floras, types were rarely examined and names, therefore, were often applied incorrectly. Only Sodiro's (1897) treatment of the ferns of Ecuador is available for the Andean region; yet he was aware of only 6 of the 23 species that occur in that region. Vareschi's (1969) treatment for Ven- ezuela and Brade's (1971) for Brazil have helpful illustrations, but the names are often misapplied. The best treatment of Polybotrya for Mesoamerica is Stolze's ( 1981 ) excellent work for the Flora of Guatemala, but Guatemala lacks several species found in Mesoamerica. In brief, adequate keys and descriptions are lacking for identifying Polybotrya throughout most of its range and, as a result, many specimens have been mis- identified. In this century, most of the research on Polybotrya has focused on the specific level. Brade (1935, 1948) described two new species of Polybotrya in papers that included other ferns as well. Later, in a series of three papers published in the Brazilian journal Bradea ( 1969a, b,c), he described ten addi- tional new species from Brazil, Costa Rica, and Venezuela. These papers were based on field studies in the American tropics, and most of the types were plants he had collected by himself or with his brother Alfred. Brade (1971) published a synopsis providing keys and illustrations of the Brazilian species of Polybotrya. Unfortunately, he died before completing the descriptions, habitats, and distributional information. Only five of Brade's names are accepted in this mono- graph as representing valid species; the re- mainder are relegated to synonymy. More recently, Lellinger (1972, 1977) published two important papers on Polybot- rya. In the first, he described five new species from South America and offered interesting ecological notes on the genus. In the second, which deals with other ferns as well, he de- scribed two new species that are endemic to Colombia. Only three of these seven species are accepted here. About half of the species described by Brade and by Lellinger have, therefore, been relegated to synonymy in the present work, largely because I found older, obscure names during my research. In some cases, these names had not been u.sed since their original publication. The proliferation of names is one of the pitfalls faced by taxonomists when synthetic or revisionary works are unavailable for a group and in this case indicates how much Polybotrya has needed a monograph. Illinois Natural History Survey Vol.34. Art. 1 Ecology Polyholrya inhabits wet, shaded, tropical forests. These may be hot, humid, lowland rain forests or cool, cloud forests at high ele- vations. The genus rarely occurs in disturbed forests, and when it does, it is represented by only a few isolated plants. However, small scale disturbances within mature forest, such as light gaps created when a huge tree dies, appear to be important in establishing young plants (Lellinger 1972). During fieldwork in tropical America, I never observed young sporophytes or their associated gametophytes establishing themselves on roadbanks or on shaded, disturbed slopes adjacent to roads. Furthermore, all herbarium specimens that contained habitat information listed only pri- mary forest as the habitat. The restriction of Polyhotrya to wet, shaded, primary, tropical forest is reflected by its geographic distribu- tion (Fig. 2a). The genus is absent from such arid regions as central Mexico and the west- em coast of Peru and from such nonforested regions as the Llanos of Venezuela and Co- lombia (Figs. 2a & 3). The altitudinal range of Polyholrya is from sea level to 25(X) m. with most plants collected from 500-2000 m (Fig. 4). This range certainly corresponds to my field observations that Polyhotrya is most diverse, frequent, and abundant at middle elevations. Figure 2. Distribution of Polyholrya. a. composite of Maps 1-21; b. number of species (left bar), number of endemics (right bar). Countries or areas are outlined; those with only one bar have no endemics. See Table 3 for numbers of taxa in each countn, November 1987 Monograph of Polybotra Geography In the following discussion, the distribution of Pohhotrya is given according to the re- gional centers defined by Tryon (1972). These regional centers, all mountainous, were determined by their high endemism and species richness. The most important regional center in the diversification of Pohhotrya is the Andean, from Venezuela and Colombia south to Bolivia (Fig. 2b, Table 2). The Andean Re- gion contains 23 species, 12 of which are endemic, and I suspect that new species of Pohhotrya will be found there. Within this region. Colombia contains the richest Pohhotrya flora: 17 species, 6 of which are endemic (Table 3). The coastal mountains of northern Venezuela contain P. canaliculata and P. serratifolia, both occurring in cloud forests from 1200-2400 m. Pohhotrya canaliculata has a very limited distribution and is a true endemic to the coastal mountains (Map 19); it occurs outside of the Andean Region as defined by Tryon (1972). Polyhot- r\a serratifolia is considered a near- endemic, extending southwest into the Andes around Merida and eastward to Trinidad (Map 1). The Serra do Mar Mountains of the South- eastern Brazilian Region contain five species Figure 3. Distribution of Polybotrya species within regional centers of diversity and endemism for tropical American ferns defined by Tryon (1972) Primary centers arc indicated with dark shading: secondary centers, with light shading. The number on the left is the total number of species; the number in parentheses indicates how many of those species are endemic. Sec Table 4 for comparisons. Illinois Natural History Survey Vol. 34, Art. 1 of Polyhotrya, all of which are endemic (Fig. 3, Tables 2 & 3). I find it remarkable that this region has no species of Polybotrya in common with those of the other regional cen- ters (Table 4) or with intervening areas. This distinctiveness of the Polybotrya flora is ac- centuated by its lack of the simply pinnate Sf)ecies, that is, subgenera Soromanes and Sorhifolia. The biogeography of Polybotrya strongly supports data from many other sources that suggest that the isolated Serra do Mar Mountains are a distinctive biogeo- graphic center (Tryon 1972). Only one species, P. goyazensis, occurs in the Matto Grosso and intervening area be- tween the Andean and Brazilian regional cen- ters (Map 6). This species is closely related to the widespread P. caudata and is probably an example of the peripheral "budding" of a new species from the southern part of the range of P. caudata. The Amazon basin contains one endemic, P. glandulosa, which has been collected only three times (Map 7). Most species of Polybot- rya in the Amazon basin also grow in the Andean Region. In contrast, the coastal mountains of .southeastern Brazil have contributed no species to the Amazonian low- lands. My field experience in the Amazon basin of Ecuador and Peru showed that in forests on wet, sandy soils P. caudata was common; forests on lateritic soils, however, had P. crassirhizoma. P. caudata. P. osmun- dacea. and P. puhens. All five species were abundant in the western Amazon basin where I collected, and I suspect that they extend farther eastward into the Amazon basin than is shown on their range maps. The Guiana Highlands Region does not contain a distinctive Polybotrya flora. Only six species occur there and most of these are widespread elsewhere. The region, therefore, has a high fioristic affinity with the Andean and Central American regional centers (Table 4). Polybotrya sessilisora, which occurs in the Guiana Highlands and in the adjacent Amazon basin (Map 20), is probably more common in the Highlands and its range may be centered there. Polybotrya has two notable examples of Andean-Guianan range disjunc- 2500 - 2000 - 2 1500 Ul I- UJ 2 1000 500 - 2 3 31 33 I Figure 4. Altitudinal ranges of Polyholrya species. The numbers ab*ive the bars correspond to the species numbers assigned in the taxonomic treatment. November 1987 Monograph of Polybotra tions. TTie first example, P. lechleriana, is known in the Guianas from a single collection at Mount Roraima in Guyana. 1900 km from its nearest locality in Colombia (Map 7). This disjunction is probably best explained by long-distance spore dispersal from Andean populations. The second example, P. frac- tiserialis. occurs primarily in the foothills of the Andes and is disjunct in the southern Pakaraima Mountains of British Guiana, 2000 km from the nearest Andean population in Ecuador. Unlike P. lechleriana, P. frac- tiserialis is common and widespread in the Guianas (Map 3). The Central American Region contains seven species, including one endemic (Fig. 3, Table 2). All seven species occur in Costa Rica (Table 3). My fieldwork there showed that P. polyhotryoides, P. osmundacea. P. alfredii and P. gomezii (endemic) were par- ticularly common in the mountains and that P. caudata was frequent in the costal low- lands. Polyhotrya alfredii and P. gomezii were particularly abundant in cloud forests, often growing together. In the mountains of central Panama, just outside the Central American Region as defined by Tryon (1972), one endemic occurs, P. alata. The Greater Antilles Region, otherwise quite rich in fern species, has only one Polxhoirya species, the widespread P. os- mundacea (Figs. 2 & 3, Table 2). Puerto Rico lacks Polyhotrya but certainly has suit- able forest habitat. In the Lesser Antilles, P. osmundacea is found on Grenada. Martinique and Guadeloupe have the closely related en- demic P. cyuthifolia. The Caribbean islands have played a minor role in the diversification of Polyhotrya. Polyhotrya caudata and P. osmundacea clearly have the largest geographical and al- titudinal ranges of any species in the genus (Maps 5 & 18). With the exception of south- eastern Brazil, these two species extend nearly throughout the entire area covered by all of the other species in the genus. Polyhotrya occurs on only one oceanic island—Cocos Island, about 500 km (310 mi) southwest of Costa Rica in the Pacific Ocean. Three species occur there: P. Table 2. The geographic regions shown in Figure 3 and the species oi Polyhotrya that occur in them. 10 Imjnois Na riKAi History Survry Vol. 34, Art. 1 Table 3. Distribution of Polyhoina by country. Endemics are in boldface. Numbers in parentheses refer to the species numbers assigned in the taxonomic treatment. Mexico: 2 species, no endemics, polyholryoides (2), caudata (9). Belize: 2 species, no endemics, polybotryoides (2), caudata (9). GuATEMAt,A: 3 species, no endemics, polybotryoides (2), caudata (9), osmundacea (28). Honduras: 3 species, no endemics, polybotryoides (2), caudata (9), osmundacea (28). Nicaragua: 3 species, no endemics, caudata (9), alfredii (16), osmundacea (28). Co.STA Rica: 6 species, 1 endemic, polybotryoides (2), sorbifolia (5), caudata (9), alfredii (16), gomezii (27), osmundacea (28). Cocos Island: 3 species, no endemics, polybotryoides (2). caudata (9), osmundacea (28). Panama: 5 species, 1 endemic, polybotryoides (2), caudata (9), alfredii (16). alata (23), osmundacea (28). Cliba: 1 species, not endemic, osmundacea (28). Jamaica: 1 species, not endemic, osmundacea (28). Haiti: 1 species, not endemic, osmundacea (28). Guadeloupe & Martinique: 1 endemic, cyathifolia (29). Grenada: 1 species, not endemic, osmundacea (28). Trinidad: 3 species, no endemics, serratifolia (1), caudata (9), osmundacea (28). French Guiana: 3 species, no tr\Atm\cs. fractiserialis (6), caudata (9), osmundacea (28). Surinam: 2 species, no endemics. /r«(r/.?er/a//.v (6), caudata (9). Guyana: 4 species, no cnAetmcsfractiserialis (6), caudata (9), lechleriana (13). osmundacea (28). Venezuela: 6 species, 2 endemics, serratifolia (I), sorbifolia (5), caudata (9), glandulosa (12), osmundacea (28), canaliculata (32). Colombia: 17 species, 6 endemics. /7()/v/wrno((yf.?( 2 ).ii(/)<'rf(7a( 3 ),iorfe//o/ia (5), crojiir/i/roma (7), caudata (9), pubens (11), lechleriana (13), attenuata (14). stolzei (15). botryoides (17), lourteigiana (18), pittieri (19). hickcyi (21), altescandcns (26), osmundacea (28), latisquamosa (30), sessilisora (31). Ecuador: 14 species, 3 endemics, polybotryoides (2), suberecia (3), andina {4). fraciiserialis (6). crassirhizoma (7), caudata (9), pubens (11), lechleriana (13), alfredii (16), puberulenta (22), aequatoriana (24), appressa (25), altescandcns (26), osmundacea (28). Peru: 1 1 species, no endemics, polybotryoides (2). suherecta (i).fractiserialis (6). crassirhizoma (7), caudata (9), pubens (11), glandulosa (12). lechleriana (13), alfredii (16). altescandcns (26). osmundacea (28). Bolivia: 10 species, no endemics. /rac/i.vfn'u//.? (6), crassirhizoma (7), caudata (9), pubens (11), lechleriana (13), alfredii (16). hirkcyi (21). puberulenta (22). aequatoriana (24). osmundacea (28). PARACiUA"!': 1 species, not endemic, goyazensis (10). Brazil: 1 3 species, 5 endemics, sorhifrlia (5), crassirhizoma (7), cspiritosantensis (8). caudata (9), goyazcnsis (10), pubens (11), glandulosa (12). cylindrica (20), osmundacea (28), sessilisora (31). seniipinnata (33), speciosa (34), pilosa (35). November 1987 Monograph of Poi.ybotra caudata. P. osmundacea. and P. polyhot- ryoides. As noted in the previous paragraph. P . caudala and P . osmundacea have the most extensive range of any species in the genus. Although not as wide ranging, P. polyhot- ryoides has a long north-south distribution from Chiapas, Mexico, to Pasco, Peru, and is the most far-ranging species of the sub- genus Soromanes. This distribution supports Tryon's (1970) findings that the ferns of oceanic islands tend to be those that are wide- spread beyond the source area for the island (the source area for Cocos Island is defined by Tryon as Central America and Colombia). According to Tryon (1970), widely distrib- uted species have a broader ecological amplitude and therefore establish themselves more often on oceanic islands because of their ability to grow undera wide range of environ- ments. Morphology and Anatomy Information on the morphology and anatomy of Polyhotrya has been extremely useful for assessing affinities of related genera and for understanding functional aspects of the biol- ogy of the genus. Every phylogenetic argu- ment made here relies heavily on evidence from stem anatomy (along with other fea- tures, of course) in determining how Polyhot- rya relates to other fern genera and families. Anatomical studies have also answered other biological questions about the genus. For example, studies of the diplodesmic venation system showed how the sori are efficiently supplied with water and nutrients. Studies of sterile-fertile transitional leaves established that the diplodesmic veins are homologous with the veinlets of the sterile leaves. In short, anatomical studies of root, stem, and leaf have been valuable in understanding diverse questions about the biology of Polyhotrya. Roots A dense tangle of dark, tough, fibrous, adventitious roots ari.ses from the ventral side of climbing stems and from all sides of ter- restrial stems. No positional relationship exists between the leaf bases and the roots; roots are apparently borne at random. Roots arising directly from the stem measure about I mm in diameter, but their terminal branches may be only 0.3 mm wide. Roots branch at irregular intervals, increasing the number of tiny supportive rootlets. On climbing plants, these tenacious rootlets penetrate the outer layers of tree bark and firmly anchor the stem. This firm attachment hinders pulling the stem away from the tree. When a stem is forcibly pulled away from a trunk, the large roots of Polyhotrya remain attached to the stem, clinging to shredded pieces of bark, lichens, and mosses. I studied the root anatomy of six spe- cies— P. alfredii, P. caudata. P. gomezii, P. osmundacea, P. polybotryoides , and P. sorhifolia. All were similar. Figure 5a shows Table 4. Affinities offo/yiofrya floras between regional centers (.see Fig. 3). Floristic affinity is the percent species in common (C) of the total species in the two regions (A + B), i.e., 1(X)C/A + B. Regional Center (Tryon 1972) Total Species Species in Common Floristic Affinity (A + B) (C) (%) Andean &Guianan Central American & Andean Andean & Southeast Brazilian Central American & Guianan Guianan & Southeast Bra/.ilian Central American & Southeast Brazilian 29 30 28 13 11 5 12 Illinois Natural Hlstory Survey Vol.34. Art. I a root in cross section taken about 0.5 cm from the ventral surface of a scandent stem of P. osmundacea. The stele is weakly ellip- tic, protostelic, and comprises about one-fifth of the root's diameter. Two protoxylem poles occur; each is exarch and consists of four to seven protoxylem elements. The center of the stele contains three to five large metaxylem cells. Surrounding the xylem are phloem elements— minute, angular, thin- wailed cells about the same diameter as the protoxylem elements. They are not shown in Figure 5a because they were poorly pre- served. Surrounding the stele is the innermost layer of the cortex— the endodermis. It con- sists of a single, continuous, light-colored layer composed of thin-walled, rectangular Jg$k^ QA^ 0.1 mm 0.1 mm Figure 5. Root anatomy and various epidermal features of Polyhoirya. a.b. and c are the same size. a. root cross section of P. osmundacea: b. close-up of a stem meristele in cross section, showing root trace diverging about midway between the laterally diverging leaf traces: c. phloroglueinol-containing gland from the stem epidermis of P. cauduta; d. stomata on stem aerophore of P. caudata: e. fertile segment of P. stolzei, adaxial surface. Note different shapes of epidemial cells and the two kinds of hairs; the flaccid hair represents the type that intergrades with scales. EN endodermis; EP epidermis; LT leaf trace; PH phloem; RT root trace; SC sclerenchyma: SH sclerenchyma sheath surrounding a meristele; T tracheid; X xylem. November 1987 Monograph of Poi.ybotra 13 cells. Although the endodermis is narrow, it contrasts sharply with the dark, thicker- walled cells on either side. The anticlinal walls have casparian strips that are easily vis- ible even in unstained material. Outside the endodermis is a ring of dark, thick-walled fibers (Fig. 5a). This ring is the most conspicuous feature of Polyhotrya root anatomy. Unstained, the walls of the fibers appear orangy or reddish and may become so thick that they nearly occlude the entire lumen of the cell. The walls thicken evenly on all sides, unlike many angiosperms and some ferns where the deposition of cell walls occurs centripetaily (Bierhorst 1971; Wagner 1952), Outside the ring of dark fibers, the corti- cal cells take on a different appearance. In contrast to the cells of the inner ring of fibers, the outer cortical cells are parenchymatous with thin, irregular, dark brown cell walls. Cells of the epidermis are compact and quadrangular, in contrast to the shape of the cortical cells immediately below. The most conspicuously differentiated cells of the epidermis are the root hairs, generally 0.2- 1 .0 mm long and straight, narrow, and unicel- lular. Their color normally varies from brown to amber, but brilliant yellow hairs occur in several species. Unfortunately, the root anatomy of Polyhotrya cannot be compared with that of other genera of dryopteroid ferns because of inadequate information for those genera. Stem Habit. The stem is climbing in all species of Polyhotrya except for P . fractiserialis and P. sorbifolia, where it is terrestrial (Fig. I). Climbing stems grow horizontally about 0.5- 2.0 cm beneath the forest floor until they contact a tree. The stem then turns upward, anchoring itself to the tree by adventitious roots produced on the ventral surface. Stems usually climb to a height of 1 to 5 m, but the highest scandent stem I saw was just under 15 m. The longest underground stem that I saw was traced back 8 m from where it en- tered the soil at the base of its support tree. As a stem begins to climb, it grows wider. The diameter of a climbing stem may reach 3.5 cm; terrestrial stems seldom grow wider than I cm. The intemode length on climbing stems is generally 10-15 cm. Stems of the two terrestrial species of Polyhotrya differ in overall appearance from stems of the climbing species. Stems of ter- restrial species are about 1 5 -20 cm long with intemodes 1-3 cm long. These short inter- nodes impart a basketlike aspect to the ar- rangement of leaves instead of the linearly spaced aspect typical of the leaves of climb- ing species (Fig. I). The amount of scale cover is another difference. Stems of the ter- restrial species are less scaly than stems of climbing species due to the sloughing off or rotting away of scales, presumably because the scales are buried in moist, tropical soils where decay is rapid. Furthermore, the petiole bases hide the few scales that remain in terrestrial species; the best place to find scales is the stem apex, where there is less decay and fewer petiole bases. Despite these differences, both types of stems show affinity by sharing the distinctive anatomy described below. Branching is rare in stems oi Polyhotrya. I never saw dichotomy of the stem apex, and lateral branching was exceedingly rare in scandent stems, although occasional in ter- restrial portions. I did not investigate the re- lationship, if any, of branching to leaf posi- tion, but that examination might provide further evidence for comparisons with other genera of dryopteroid ferns. Anatomy. The stems of all dryopteroid fern genera have radially symmetric dictyo- steles; however, the stems of Polyhotrya are distinctive (Fig. 6). The most noticeable fea- ture in cross section is the vascular bundles — terete, circularly arranged, each bundle en- sheathed by a jet-black ringof sclerenchyma. The 5 to 12 vascular bundles are positioned around the central portion of the stem, their dark outlines contrasting vividly with the matrix of bright white ground parenchyma. Numerous tiny leaf traces emanate from the 14 III iNois Natural Hisiory Survey Vol.34. An. 1 PH -r^c,-^^ 1 Figure 6. Stem anatomy of Polyhoirya osmumlucca. typical of the genus. Lower left is a cross section of the stem with a mcristclc enlarged above. Note "xylary arm" in 2. AP aerophore: EN end^xlemiis; GP ground parenchyma: l.T leaf trace: MS meristele: P parenchyma: F^ pericycle: PH phloem: RT root trace: 1" trachcid: SH sclcrenchyma sheath consisting of fibers: X xylem: XP protoxylem. November 1987 Monograph of Polybotra 15 sides of the bundles, collectively forming a C-shaped line that connects adjacent bundles. This pattern is unmistakable (Fig. 6), and no other fern genus yet studied has this kind of stem anatomy. (For comparison of stem anatomy with closely related groups, see Re- lationships with Other Groups.) The epidermis of the stem is darkly sclerotized and one cell layer thick; it pro- duces numerous scales (described separately below). The cell walls of the epidermis are relatively straight compared to the sinuous ones of the leaf epidermis (cf. Figs. 5c & 9). The stems of Polyhotrya have minute glandular hairs on the epidermis. These short- stalked hairs have, in relation to their length, exceptionally large, round heads, 54-56 microns in diameter (Fig. 5c). Because these glands are tiny and few, they had been over- looked until found by Widen et al. (1983), who were also the first to report phloro- glucinols in Polyhotrya. The phloroglucinols are apparently produced and stored in these glands. In other dryopteroid ferns, such glands are known to secrete an oleoresin, along with various lipids, between the cuticle and the outer layer of the cell wall. The phloroglucinols are, presumably, located in this secretion (Widen et al. 1983). Using ex- tracts from the stem and petiole bases of P. caudata. Widen et al. found that aspidin BB was the principal compound, along with smaller amounts of albaspidin BB,desaspidin BB, and flavaspidic acid BB. Internal glandu- lar hairs, which probably also contain phloroglucinols, occur in the stems and petioles of other dryopteroid ferns (Mehra & Mittal 1961 ). I did not, however, find internal glands in Polyhotrya. Aerophores occur on the stems of most species of Polyhotrya but are apparently ab- sent in Sf)ecies having a mat of thick, woollike scales, such as in P. aequatoriana, P. alte- scandens, and P. crassirhizoma. On living stems, aerophores appear as yellowish white lines 1-3 mm wide that connect with aerophores on the petiole (for description of petiolar aerophores, see section below on petioles). Aerophores usually extend 2-6 cm down the stem from the base of the petiole. In cross section, they project about 0.5-1 mm above the surrounding surface and con- sist of thin-walled parenchyma cells that in- terrupt the otherwise continuous, thickened epidermis (Figs. 6 & I6g). Stomata abound and can be seen with a binocular microscope (Fig. 5d). Presumably, aerophores allow oxy- gen into the stem that is otherwise surrounded by compact, oxygen-impermeable, scleren- chyma fibers. Because aerophores shrink and darken upon drying, thereby matching the contour and color of the stem's scleren- chyma, they cannot be seen on herbarium material. Below the epidermis is a well-developed hypodermis, usually three to eight cells thick and darker than the epidermis that forms a conspicuous ring around the white ground parenchyma. The cells of the hypodermis are thick-walled sclerenchyma fibers that thwart cutting the stem, .sometimes even with a hefty machete, and I had problems cutting thin, even, cross sections for microscopic study. A ring of hypodermal sclerenchyma also en- circles the stem in closely related genera such as Arachniodes, Cyclodium, Maxonia, Olfer- sia, and Stigmatopteris (pers. obs.). Below the hypodermis lies the cortex, which consists of scattered clusters of brachysclereids (stone cells) in a matrix of whitish parenchyma. The parenchyma cells are generally rounded and contain abundant amyloplasts. In a freshly cut stem, the white color of these cells contrasts vividly with the darkly sclerotized hypodermal and meristelar sheaths. Sclereids never abound, as in Olfer- sia (Fig. 16i), but are scattered in clusters throughout the ground parenchyma. The dep- osition of secondary walls is extremely un- even; most of the wall is laid down on the side of the cells facing the inside of the sclereid cluster. A small, cuplike lumen can usually be seen on the side of the cell facing away from the center of the sclereid cluster. The thick, dark cell walls contain numerous simple pits and a lesser number of ramiform 16 M.LiNois Natural HrsxoRY Survey Vol. 34. An. I pits. Many Dryopteridaceae and Lomariop- sidaceae also have sclereid clusters in the ground parenchyma of their stems (pers. obs.). Inside the scierenchyma sheath that sur- rounds each meristele lies a band of paren- chyma two to four cells wide (Fig. 6). The cells of this layer are more compact and smaller than the cortical parenchyma, but like the cortical parenchyma, they contain amylo- plasts. This inner parenchymatous layer stops abruptly at the endodermis, which stands out as a single row of clear, narrow, rectangular cells interrupting the darker parenchyma on either side of it ( Fig . 6) . The next layer toward the center is the pericycle, which consists of compact, orangy cells that form a continuous band around the xylem and phloem. The phloem consists mostly of clear, an- gular, thin-walled sieve cells (Fig. 6.2 & 6.4). Phloem is broadest on the distal side of the stele but forms a thin band elsewhere. This band is broken in two locations by arms of the xylem that extend to the pericycle (Fig. 6.2). Scattered among the clear cells of the phloem are parenchymatous cells. Inside the ring of phloem is the xylem, which has large, conspicuous metaxylary tracheids (Fig. 6.3). These tracheids occur singly and in clusters within a matrix of dark brown (when unstained) parenchyma. The tracheids are generally rounded or somewhat angular; when adjacent to another tracheid, the walls are straight. Numerous pits occur between all the cells of the xylem. Pro- toxylem elements occur in the xylary "arms" (Fig. 6.2). Each meristele is, therefore, diarch and exarch. Root traces arise between the xylary arms of the meristele (Fig. 5b). The roots diverge through the cortex to the ventral surface of the stem. In contrast to the root traces, the leaf traces originate from the xylary arms. Four to six leaf traces are usually produced at each leaf gap from adjacent meristeles. During their passage from the cortex to the petiole, the leaf traces bifurcate and anas- tomose, forming a reticulum like that in Maxonia (Chandra 1975). Fach leaf trace is surrounded by a single layer of sclerotized cells (Fig. 5b). TTie walls of these cells are brownish and slightly thickened on the inner tangential surface. The distance from the in- ception of a leaf gap to the point where the leaf diverges from the stem is about 10-15 cm. Stem scales. Many spec\es of Polybotrya have distinctive stem scales. Some species, in fact, can be identified solely on the basis of their stem scales (thus the importance of always collecting part of the stem). Color is the easiest feature of the scales to use in iden- tifying the species of Polybotrya. Botanists should have no difficulty identifying P. alte- scandens, with its bright golden or yellow scales; P . crassirhizoma, with its dull orange or red scales; and P. latisquamosa. with its shiny, castaneous scales. Depending on the species, scales range from concolorous to slightly darker in the center to bicolorous with a dark central stripe. Scale habit is another helpful identifica- tion feature. Some species, such as P. alte- scandens and P . crassirhizoma. have narrow, densely tangled scales that impart a woolly appearance to the stem. Other species, such as P. appressa. P. caudata. and P. cylin- drica, have thick, dark, massive scales that are conspicuously appressed to the stem. Be- cause of their great width and spreading habit, the scales of P. serratifolia are distinct from those of other species of the subgenus Soromanes . Most species oi Polyhatrya have ascending and slightly spreading scales— a characteristic that is not particularly distinc- tive because of its prevalence in the genus. The shape of stem scales is usually not helpful in identification because most species have lanceolate to linear-lanceolate stem scales. Two species, however, P. alata and P. altcscandcns (Figs. 42 & 45). have ex- tremely long, attenuate scales that differ strik- ingly from those of the other species. Petiole scales are generally similar to those of the stem but tend to be shorter and wider. The petiolar scales of P. latisquamosa have be- come, as the specific epithet implies, greatly widened .so that this characteristic distin- November 1987 Monograph of Polybotra 17 guishes the sfiecies from all others in the genus (Fig. 50d). Despite variation in color, habit, and shape, stein scales are of two general types, here termed simply Type 1 and Type 2. Scales of Type 1 are thin and translucent and have easily visible cell walls (Fig. 7). Most of the scale is one cell layer thick, although the center may be thicker and darker. The mar- gins vary from denticulate to ero.se, with the teeth formed from the projecting ends of two adjacent cells (Fig. 7). Marginal teeth in 0.1 mm Figure 7. Stem scales of various Polybotrya species, a, b. i. P. crassirhizoma. P. altescandens; c. P. boiryoides; 18 Ii I iNois Natlrai. History Survey Vol.34. An. I Polyhotrya are never glandular, as in Bolhitis (Hennipman 1977) and Stigmatopteris (pers. obs.). Scales of Type I are attached basal I y at a single point or. more rarely, across the full length of the base. This point or line of attachment is always darkly sclerotized (Fig. 7a). Representative species with scales of Type 1 include P . crassirhizoma. P. goyazen- sis, P. puherulenta, and P. serratifoUa (Figs. 26, 29, 41, & 19). Scales classified as Type 2 are thick, opaque, and composed of many cell layers. The margins are entire or subentire. The base is greatly thickened, sometimes almost stipelike, and is curved and ascending. When the scale is removed, a circular or oval scar remains on the stem. Scales of this type are usually appressed, but they are squarrose in P. osmundacea. Representative species with scales of Type 2 include P. appressa, P. caudata, and P. cylindrica (Figs. 44, 28, & 39). In contrast to scales of Type I, which show tremendous variety in habit, color, and shape, those of Type 2 are relatively homogeneous. Evidence suggests that morphological transition occurs easily between the two types of scales. Species such as P. lechlehana. P. lourteigiana. P. osmundacea. and P. puhcns have both types of scales but on different individuals. Even on a single plant, scales belonging to Type 1 can occur on the climb- ing portion, with scales of Type 2 and/or intermediates on the terrestrial portion. If morphological transition is easily ac- complished, then evolutionary change in scale type could be expected to occur in closely related species pairs. This hypothesis is supported by two closely related species P. caudata and P. goyazensis (Figs. 28 & 29); the former has scales of Type 2 whereas the latter has scales of Type 1 . Clearly, the stem scales of Polyhotrya have been mor- phologically and evolutionarily plastic. The evolutionary advantage of scales in Polyhotrya is unknown. Their contribution to protecting the stem from mechanical dam- age is probably minimal since the stem is surrounded by a tough, resistant, scleren- chymalous sheath. Scales seem unnecessary on the older, hardened parts of the stem. Dur- ing fieldwork in Latin America. I found sev- eral plants of Polyhotrya that had lost all of their scales on the lower, older part of the climbing stem, yet the plants appeared vigor- ous. Perhaps the scales afford protection to the stem apex , where young tissues have not yet become sclerified. Another hypothesis is that the scales re- tain water by capillary action for later use by the roots. Retention would be advantageous to a climbing plant that has the leaves on the scandent portion of the stem separated by several meters from the roots in the soil. Stems collected in the field, however, never seemed to hold water between the scales and plants never appeared water stressed. The bark of the support tree, which is usually covered with water-retaining mosses, lichens, and organic debris, probably could provide most of the water needed by the plant. Further, the retention of capillar)' water be- tween stem scales might be harmful in a hot, humid, tropical environment that promotes decay. No satisfacton,' hypothesis, therefore, has been offered to explain the adaptive sig- nificance of scales in Polyhotrya. Even more difficult to explain is the adaptive advantages of the numerous minor modifications of scales. What could be the function of a dark, central stripe, of reddish color, of an appressed habit, or of a linear shape? How these and other minor modifica- tions of scales are significant in evolution, if indeed they are. is baffling. Leaves Petiole. The petiole base in some ferns is specialized to accumulate food and to per- sist as a storage organ long after the rest of the leaf to which it was attached has withered and decayed. These speciali7ed petiole bases, called "trophopods." are a new source of sys- tematic data in ferns (Wagner & Johnson 1983). Polyhotrya. however, lacks troph- opods. and the food-storing function of the trophopod is perfomied b)' the stem in the November 1987 Monograph of Polybotra 19 amyloplast-containing cells of its ground parenchyma. A cross section of the petiole of Polyhot- r\a reveals leaf traces that are arranged in a mushroomlike outline, with the base of the mushroom oriented adaxially (Fig. 8a). This pattern also occurs in Cyclodium trianae, a species closely related to Polyhotrya. and in some species of the similar family Lomariop- sidaceae (Hennipman 1977; Holttum 1978). Other closely related dryopteroid genera, however, such as Cyrtomium, Dryopteris, and Polystichum, display a C-shaped arrange- ment of leaf traces in the petiole (pers. obs.). These genera also have fewer leaf traces than Polyhotrya, probably as a consequence of their narrower petioles. Further study of petiole anatomy in dryopteroid genera may provide helpful information for assessing their relationships. In Polyhotrya. the two vascular bundles closest to the adaxial surface of the petiole differ from the lower vascular bundles in sev- eral respects. The most obvious difference is their shape: elongate and slightly curved in- stead of round (cf. Fig. 8b, d). Inside each of the two adaxial bundles, the xylem is hook- shaped in outline, in contrast to the lenticular shape seen in the lower traces. Only these uppermost vascular bundles, not the lower, yield the lateral traces that supply the pinnae. As they traverse the rhachis and petiole, all traces remain separate; they do not anas- tomose to form a reticulum within the petiole . Aerophores run down the sides of the petiole and join below with those on the stem. On living petioles, they appear narrow, linear, and light green. The surface of the aerophores bears stomata, and beneath them is parenchyma that contains intercellular spaces. This parenchyma, when seen in cross section (Fig. 8c), interrupts the otherwise continuous band of collenchyma that encir- cles the petiole. Upon drying, the aerophores darken and collapse, leaving a sulcus on either side of the central, adaxial sulcus of the petiole. Drying, therefore, makes the petiole trisulcate; in living petioles only a central sulcus is present. Petiole and stem mucilage. Mucilage has apparently never been reported in Polyhotrya or any other genus of Dryop- teridaceae. During fieldwork in Costa Rica and Ecuador, however, I found that stems and petiole bases of P. alfredii usually had a thick coat of translucent mucilage similar to that on the stems and/or petioles of certain Blcchnum and Thelypteris species. Upon dry- ing, the mucilage disappeared completely, although sometimes the scales exhibited a matted or flattened appearance, as if stuck together. Attempts to rehydrate the mucilage failed. Other species of Polyhotrya probably have mucilage. Several dried specimens of P. lechleriana and P. lourteigiana exhibited matted, flattened scales that suggested the earlier presence of mucilage. These two species are closely related to P. alfredii, an observation that suggests that mucilage is probably confined to the P. alfredii group (species nos. 13-20). I did not see matted, flattened scales in any other species of Polyhotrya. The mechanism of mucilage se- cretion and its adaptive significance, if any, is unknown. Nectaries. Koptur et al. (1982) were the first to report the existence of nectaries in Polyhotrya. They found them in P. osmun- dacea on the rhachis of unfurling leaves, near the yellowish, lateral, linear aerophores. Sev- eral times during fieldwork in Costa Rica, I noticed ants on young leaves, but I never saw them feeding on nectar. The nectar itself would certainly be nutritious because it con- tains sucrose, glucose, and fructose in con- centrations up to 35-75 percent by weight (Koptur et al. 1982). Amino acids also occur in the nectar, and Koptur et al. list those present. Although ants may benefit from the nectaries, additional fieldwork is needed to establish whether or not the plants benefit. Apparently, ants do not defend the plants— at least I was never attacked by ants while col- lecting Polyhotrya. Costa-costule architecture Costa-cos- tule architecture refers to the morphology of the juncture of these axes. Several kinds of 20 Illinois Na turai- His iory Survey Vol. 34, An. 1 0.1 mm FiouRE 8. Petiole anatomy of Potybotrya alfredii. a. cross section of petiole 2 cm above the stem .showing the arrangement of vascular bundles in the shape of a mushRxim: b. adaxial bundle showing hooked xylcm; c. aerophore; d. abaxial bundle. COL collenchyma; EN endodermis; EPI epidermis; T tracheid. November 1987 Monograph of Poi.ybotra 21 costa-costule architecture occur; the most thorough review of them is by Holttum (1959); more than any other pteridologist, he has shown the importance of this feature in the taxonomy of ferns. According to Holttum (1984), the dryopteroid genera have "midribs of ultimate leaflets grooved, the groove of the rachis bearing the leaflets being open to admit the leaflet-groove, the margin of the lamina of the leaflet being decurrent (but not prominent) down the side of the rachis; hairs of ctenitoid type lacking." This type of ar- chitecture is illustrated in Figure 47h. Holttum (1984) describes the tectarioid gen- era as having "midribs of ultimate leaflets more or less prominent (in Tectaria some- times slightly grooved) and bearing ctenitoid hairs, usually many." In Polybotrya. the basic type of architecture is dryopteroid, a finding that supports its classification among the dryopteroid ferns. Within Polybotrya. however, modifica- tions are found in the dryopteroid pattern of costa-costule architecture. Typical dryop- teroid costa-costule architecture is shown in the most primitive group in the genus— the group of species related to P . osmundacea. In contrast, the most advanced species in the genus — P. altemuita, P. lechleriana, and P. stolzei— always have the costular groove truncated by the ridges of the costa; therefore, the groove is not decurrent (Figs. 33b, 32b, & 34f). This truncated costa-costule architec- ture is considered to have been derived from the decurrent-grooved condition, as evi- denced from outgroup comparison with other dryopteroid ferns. Intermediates, such as P. aequatorianu and P. alfredii. have costular grooves that gradually fill up near the costa, so that the ridges of the costa are only slightly interrupted by the shallow, weakly decurrent groove of the costule (Fig. 43d). In P. alfredii and P. lourteigiana, these three types of costa-costule architecture can be found on different parts of the same leaf, although the intermediate type predominates. In summary, most species of Polybotrya have dryopteroid costa-costule architecture, but distinctive modifications of this basic type occur. Anatomy of the lamina. I studied leaf clearings and cross sections for 17 species of Polybotrya to determine differences between species or subgenera. Leaf clearings revealed that stomata were absent from the adax- ial epidermis but abundant on the abaxial sur- face between the veins. The long axes of the stomata ran almost parallel to the veins. The mean length of the stomata (for all species) was 56 microns, but means for the species ranged from 48 to 74 microns. Subsidiary cells surrounded the distal one-half to two- thirds of the guard cells. The shape of the subsidiary cells can be like the sinuous epi- dermal cells or slightly more rounded and circular (Fig. 9d,f)- In face view, the epidermal cell walls fit together like the pieces of a jigsaw puzzle (Fig. 9). This interlocking is best exemplified by the abaxial epidermal cells, which have a more highly irregular outline than the adax- ial ones. The cells above and below the veins, however, have straight walls. The slightly elongate axes of the cells are oriented parallel to the veins and point in the direction of the leaf margin. In cross section, the abaxial epidermis and adaxial epidermis form a single, compact layer of relatively thick-walled cells (Fig. 9g,h). Palisade parenchyma is absent from the mesophyll, which is composed mostly of large intercellular air spaces surrounded by long, cylindrical, parenchymatous cells. An arm of collenchyma extends from the abaxial epidermis to the vein, where it forms a sheath around the vein but does not continue to the adaxial side of the leaf. Leaf anatomy was similar in all species of Polybotrya. probably because the entire genus occupies a similar habitat. Hair types. Two fundamentally different types of hairs occur in Polybotrya. One type intergrades with scales, the other does not. The nonintergrading type is erect or spread- ing, usually cylindrical, and whitish or tawny. The size of hairs of this type varies tremendously, from unicellular hairs that are less than 0. 1 mm long and not visible to the naked eye to readily visible acicular hairs 22 Illinois Natural History Survey Vol. 34. An. 1 that have up to 15 cells and are up to 2 mm long (Fig. 10 a-i,k). The apical cell is point- ed, rarely rounded. All but five species of Polyhotrya have this kind of hair. Hair of the second type is fundamentally different from hair of the first type because each hair is a tiny, uniseriate scale and. there- fore, by definition, a hair. Every intermediate stage exists between these hairs and the scales, as evidenced by studying their change in form from the apex to the base of the costa (Figs. lOj & 36a,b). At the apex of the costa. ^cpZ9^r FiGLiRF. 9. Leaf anatomy of Polyhotrya. a-f. face views of the abaxial leaf epidermis The elongalcd cells with straight walls lie above the veins, g, h. cross sections of the leaf. a. Polyhotrya serratifolia: h. P. puhcns: c. P . fractiserialis: d. P. canalUulata: e, g. P. crassirhizoma: f. h. P. speciosa. November 1987 Monograph of Polybotra 23 only the tiny uniseriate scales, the "hairs," occur. These are readily distinguished from other hairs by their flattened cells, appressed habit, and darkened transverse walls. To- wards the base of the costa, the hairs become longer and are composed of more cells; soon are found "hairs" that are two cells wide near their base, that is, scales. These small scales grade imperceptibly into the larger scales on the rhachis. The scales on the rhachis and costae are, therefore, developmentally homologous with this second kind of hair. FiCiiiRF 10 Hair types in Polyhotrya taken from the abaxial surface of the eo.sta. a. P. puhens; b. P. semipinnata; c. P. altescandens: d. P. pillicri. e. P. alfrejii, f. P. ^kmdulosa: g. P. speciosa; h. P. osmundacea; i. P. caudata; j. P. lechleriana (reduced scale type); k. P. cylindrica. 24 iLi.iNors Natural History Survey Vol. 34. Art. 1 These hairs are called proscales since they are developmentally, and probably evolu- tionarily, precursors to scales (Moran 1986). Proscales occur on the lamina and axes of all species of Polybotrya and all genera of dryopteroid ferns. Proscales are, however, so minute that pteridologists often overlook them or do not bother to mention them. Pro- scales are smallest on the distal parts of the lamina (Fig. 49h) and at least 30x is needed to see them. Their flattened cells, often hav- ing reddish cross walls, occasionally twist upon drying, imparting the appearance of a Ctenitis-\\a\v. These are not true Ctenitis- hairs, however, because such hairs never in- tergrade with scales (Christensen 1913). This nonintergrading feature is rarely mentioned in the definition of Ctenitis-\\&\rs. Fertile Leaves Moran (1987) has presented a detailed study of the sterile and fertile leaves (trophophylls and sporophylls) of Polybotrya. Below is a synopsis of the subject. The fertile leaves oi Polybotrya resemble a skeleton of the sterile ones because their green lamina has been reduced to a narrow wing above the veins. This extreme dimor- phy, termed "holodimorphy," is distin- guished from other types of dimorphy by the nearly total loss of photosynthesis in the fer- tile leaf (Wagner & Wagner 1977). The re- duction of green tissue is the most conspicu- ous aspect of dimorphy, but differences in seasonality, duration, orientation, insertion on the stem, size, color, texture, and venation are also involved. Thus, sterile-fertile leaf dimorphy is an ensemble of characteristics and not merely a single feature of the leaf. Pteridologists previously described the sori of Polybotrya as acrostichoid, but three types of sori occur in the genus: botryoid, coenosoric, and a new type here termed "am- phiacrostichoid." Botryoid sori, which are round and discrete, are considered primitive on the basis of outgroup comparison with other dryopteroid ferns (Fig. 38b). The coenosoric type, which are oblong to linear, probably evolved from the botryoid type by basipetal fusion of the sori (Fig. 35h). The amphiacrostichoid type developed from a coenosoric ancestor by the expansion of the margins of the receptacle beyond the original adaxial surface; as a result, the receptacle assumed its own adaxial surface and sporan- gia appeared on both surfaces of the leaf (Fig. 30g). Coenosoric and amphiacrostichoid sori contain diplodesmic veins that are developi- mentally homologous with veins of the sterile leaf. Sporangia. The sporangial stalk of Polybotrya is three-celled at its apex, im- mediately below the capsule. Below the apex, however, only two rows of cells occur, each three to four (rarely 5) cells long (Fig. 1 1 ). Where the two rows of the base join the three cells at the apex, a paraphysis occurs; it is always uniseriate. unbranched, and multicel- lular. About three-fourths of the species of Polybotrya have paraphysate sporangial stalks, but paraphyses occur on only 40-80 percent of the stalks from a single sample. In dried or rehydrated material, the apical cell of the paraphysis is smaller and shriveled compared to the ceils below. In fresh mater- ial, the apical cell appears swollen and glan- dular. Sporangial stalks of P. fractiserialis are unique because they have lightbulb- shaped, glandular cells at the base of the paraphyses (Fig. I Ig). The functions of these glandular cells and paraphyses are unknown. Sporangial capsules of Polybotrya are typical of those found in related dryopteroid ferns. The mean number of annular cells for sf>ecies of Polybotrya ranges from 12 to 24, with most species having between 14 and 18. Both the epistomium and hypostomium con- sist of three or four thin-walled, transversely elongated cells. All species of Polybotrya have glabrous sporangial capsules, except P. imbcns. which has short, subulate hairs at the top of the capsule near the annulus (Fig. 30e,n. Usually two of these hairs occur on either side of the annulus. creating the impression that the capsule has "homs."" Spores. Spore sizes, measured by the longest axis, are given (when available) at the end of each species description. Numbers November 1987 Monograph of Poi.ybotra 2S are for spwres measured immediately after being placed in distilled water because this medium is readily available to taxonomists. Fifteen spores were measured per specimen. Spores of most Polybotrya species range from 45-65 microns in length, with extremes of 32 and 80 microns. According to Tryon and Tryon (1982). Polybotrya has relatively large spores compared to those of other gen- era in their tribe Dryopteroideae. The spores appyear dark brown when viewed with trans- mitted light under a compound microscope but deep orange when viewed with reflected light under a dissecting microscope. Spores of Polybotrya are fairly uniform within the genus compared to the variation encountered in other genera of dryopteroid ferns (e.g., Dryopieris and Polystu hum). Polybotrya spores are monolete, with the aperture linear and one-third to three-fourths the length of the long axis. The aperture is often obscured by the broad [lerispore folds and smaller spines. The exospore appears smooth (Fig. 12), as it does in spores of most genera of dryopteroid ferns. The perispore consists of two layers: the lower layer is thin and appressed to the exospore; the upper layer is thicker, with inflated folds and echinate to various degrees (Fig. 12). No constant differ- ences were found between the three subgen- era or smaller species groups. The principal differences between species appear in the prominence and density of peri- sporal folds and spines. Typically, the in- FiciURF. 1 1 . Sporangia of various Polybotrya species, a. P. allenuata: b. P. sorhifolia. showing both sides of the same sporangium; c. P. espirilosanlensis; d. P. speciosa: c. P. osmundacea; f,g. P.fracli- seriatis (note in g the globose, glandular cell al the base of the paraphysis). Illinois Natural History Survey Figure 12. Spovcs oU\>tyhotrva. a. P. semipinnaki-.b. P. osmumtacea-.c. P. scrralifolio.d. P.pinien. c. P. gomezii: f. P. alfredii: g. P. crassirhizoma: h. P. cylindrica: i. P. speciosa. a: Duane et al. 6W->-' (F) bC. Schiinkc .?S0 (GH). c: Femller 261 (GH). d: Lellingcr & de la Sola 251 (US), e: Moran 3241 (CR). f: M,mw 3I6S (CR). g; Plowman et al. 4025 (GH). h: Handro 222S (GH). i: Webh 26 (GH). c and h arc x 5(XK1, all others are x 1(XX). November 1987 Monograph of Polybotra 27 flated folds are well developed and the spines are so numerous that they impart a "fuzzy" appearance. However, the folds may be rela- tively low. as in P. goyazensis and P. sor- hifolia. or the spines may be reduced in height and/or density, as in P. speciosa (Fig. I2i). Species may also differ in spore size. This range is clearly shown by the two terrestrial species in the subgenus Sorbifoliu: P. sor- bifolia and P. fractiserialis differ greatly in spore size (40-47 and 52-56 microns long, respectively), a difference that probably re- flects different ploidy levels. Comparison of the spores shown in Figure 12 with the spores of other dryopteroid ferns illustrated by Tryon and Tryon (1982) shows that the spores of Polyhotrya are most like those of Maxonia and Stigmaiopteris (Tryon and Tryon include Cyclodium in Stigmatnp- leris). Cyclodium. which is most closely re- lated to Polyhotrya, has spores similar to those of Polyhotrya but less spiny (A.R. Smith 1986). This observation supports evi- dence from external morphology that these genera form a closely related group. Chromosome Numbers The only previous reports of chromosome numbers in Polyhotrya came from two sepa- rate counts of P. osmundacea. one from Jamaica and the other from Trinidad; both gave a chromosome number of n = 41 (Walker 1966; Smith & Mickel 1977). To add to the information about chromosome numbers in the genus. I collected meiotic material during fieldwork in Latin America. Young fertile segments were placed for 3-5 hours in distilled water saturated with paradichlorobenzene. The segments were then removed, blotted gently, and placed in a fixative of 3; 1 ethyl alcohol to glacial acetic acid. The material was stored in a freezer until it could be examined. Sporangia were squashed in a drop of aceto-carmine and then photographed . This work recorded new chromosome number counts for four species — P. alfredii. P. altescandens , P. polyhotryoides, and P. serratlfolia (Fig. 13). Each of the four counts had a chromosome number of h = 4 i . a find- ing that argues, along with morphological features, for classifying Polyhotrya among the dryopteroid ferns, all of which have n = 4\. Chromosome counts are still needed from the other species of Polyhotrya. espe- cially where polyploidy may be involved in the evolution of one species from another, as in P. fractiserialis and P. sorhifolia. Cladistic Analysis of the Species A cladogram of species relationships was constructed using the PAUP program, ver- sion 2.3 (Swofford 1985; PAUP is an ac- ronym for Phylogenetic Analysis Using Par- simony). Table 5 shows the input data used in the analysis, and Table 6 gives the charac- ters and character states used in the input data matrix. The character state trees (hy- pothesized evolutionary pathways) for the characters (Table 6) are shown in Figure 14. Further information on the characters and jus- tification of their postulated evolutionary pathways is given in the Morphology and Anatomy section. Because of homoplasy in the data set, PAUP found a large number of equally par- simonious cladograms. Therefore, a consen- sus cladogram (Fig. 15) was printed for the first 50 trees to determine the branching pat- terns they all had in common. The 50 trees agreed on the branching patterns for 12 species— about one-third of the genus. The groups that had congruent branching patterns were subgenera Soromanes and .Sorhifolia (Fig. 15, .serratlfolia— espiritosantcnsis) and the group of species related to P. caudata (Fig. 15, caudata—puhens). Homoplasy in the remaining species accounted for the numerous, equally parsimonious clado- grams. Nevertheless, distinct groups of species are shown on the consensus clado- gram, a result that supports the following subdivision of the genus. 28 Ii.i-iNOis Naiurai. History Survey Vol. 34. An. 1 "f Figure 13. Chromosome squashes of four sfwcies of Polybotrya. All squashes have n = 4l. My interpretations are at the right, a. P. altescimdem. Ecuador. Pichincha. Moron .^559 [CV\): b. P. polyhotryaides, Costa Rica, Cartago. Moran 2I7S (MO); c. P. serralijhlia, Venezuela. Trujillo. Moran 3709 (MO); d. P. alfredii, Costa Rica. Cartago. Moran 2442 (CR). November 1987 Monograph of Polybotra 29 Table 5. Data matrix for cladistic analysis of 35 character states and polaiily. Ancestor = hypothetica! ? = unknown character state. NA = not applicable species of Polybotrya. Sec text for discussion of ancestor possessing all primitive character states. 30 Illinois Nai ural History Survey Vol. 34, An. 1 Tabi F. 6. Characters and character states used in the cladistic analysis of 35 species of Polyhotrya. The numbers given to each character state correspond with those shown on one of the cladograms in Figure 14. 1 . Dissection of sterile leaves (Fig. 14a). = 4-pinnate; 1 = 3-pinnate-pinnatifid; 2 = 3-pinnate; 3 = 2- pinnate-pinnatifid; 4 = 2-pinnate; 5 = 1 -pinnate. 2. Venation (Fig. 14b). = close and long-parallel; 1 = obliquely ascending: 2 = anastomosing. 3. Type of fertile leaf (Fig. 14c). 1 = botryoid, but with lamina not completely reduced; 2 = fully botryoid; 3 = coenosoric; 4 = caudate. 4. Pinnule arrangement (Fig. 14d). 0= anadromic; 1 =catadromic. 5. Symmetry of pinnule base (Fig. 14b). = symmetrical and truncate; 1 = prolonged acroscopically and truncate; 2 = symmetrical and cuneate. 6. Submarginal connecting strand (Fig. 14e). = absent; 1 = several connections; 2 = present. 7. Stem scale base (Fig. 14d).0= attached by a single point; 1 = attached across the width of the base. 8. Hair type (Fig. 14b). O = long, acicular; 1 = small, jointed; 2= uncinate. 9. Pubescence of laminar surface (Fig. 14b). O = both surfaces pubescent; 1 = glabrous; 2 = abaxial surface pubescent. 10. Pubescence of laminar margin (Fig. I4d). = glabrous; 1 =ciliate. 1 1. Pubescence of costae (Fig. 14b). 0= uniform and dense pubescence; 1 = moderately pubescent or with scattered hairs; 2 = glabrous. 12. Lamina base (Fig. I4b). = deltate; 1 = reduced; 2 = cuneate. 13. Scale color (Fig. I4f). = golden or yellow; 1 = brown; 2 = reddish or bnghl castaneous; 3 = cream or whitish. 14. Receptacular hairs (Fig. 14d). = unbranched; 1= branched. 15. Size of lamina (Fig. 14d). 0= > 1 meter; 1 = < 1 meter. 16. Costal scale type (Fig. 140- = flaccid and ovate; 1 = linear to narrowly lanceolate; 2 = linear and tortuous; 3 = caducous. 17. Apex of .sterile leaf (Fig. !4d). = pinnatifid; 1 = subconform . 18. Stem habit (Fig. I4d). = hemiepiphytic; 1 = terrestrial . 19. Shape of the tertiary pinnules (Fig. 14b). = ligulate; 1 = oblong or ovate; 2 = obovate. 20. Spore size (Fig. I4d). = 40-50 microns long; I =52-56 microns long. \. 4 / November 1987 Monograph of Polybotra 31 Figure 15. Consensus tree for 50 equally parsimonious cladograms of species relationships mPolybot- rya. The number before the decimal refers to the character; the number after the decimal refers to the character state (Table 6). Subdivision of the Genus I subdivide Polybotrya into three easily rec- ognized subgenera: 1) Soromanes, sterile leaves I -pinnate, veins anastomosing; 2) Sor- bifolia. sterile leaves 1-2-pinnate, veins free, close, and parallel; 3) Polybotrya, sterile leaves decompound, veins free. Subgenus Soromanes ranges throughout Central America and the Andes, primarily in mountainous areas, but it is conspicuously ab.sent from the Amazon basin and from southeastern Brazil (Maps 1 & 2). This sub- genus consists of four well-defined species (species nos. 1-4) that occur in montane forests, primarily from 500-2000 m (Fig. 4). The only contemporary pteridologist who has maintained Soromanes at the generic level is Pichi-Sermolli (1977), who also placed Soromanes and Polybotrya on sep- arate branches of his phylogenetic diagram, associated with different generic groups. Other pteridologists, such as Christensen (1905), Copeland (1947), and Tryon and Tryon (1982) have subsumed Soromanes in Polybotrya. My research on both genera has shown that Soromanes is the closest genus to Polybotrya. Both genera share a unique stem anatomy: a circular grouping of meri- steles with each meristele surrounded by a dark, sclerenchymatous sheath (Figs. 6 & 16g). No other fern genus has this unmistak- 32 Illinois Natural History Survey Vol. 34, Art. I 1 cm Figure 16. Polybotrya and closely related genera, a. venation of sterile pinna. Cyclodium meniscioides; b. venation of sterile pinna, Polybotrya polyhotryoides: c. pinna of Cyclodium triamie var. trianae (compare to P. sessilisora and P. osmundacea): d. sterile pinna of Olfersia cenina. showing venation entirely unlike any Polybotrya; e-i. stem cross sections (dorsal surface is up) of e. Maxonia apiifolia, f. Lomuriopsis fendleri . g. Polybotrya caudata. h. Bolbilis lindigii. i. Olfersia cervina. AP aerophone; LT leaf trace; RT root trace; SH sclerenchyma sheath; SC sclereid or stone cell. November 1987 Monograph of Poi.ybotra 33 able stem anatomy, one of the hallmarks of Poh'hotryu. Other compelling similarities, such as holodimorphic leaves, long-creeping and densely scaly stems, petiole anatomy, and spore morphology, further demonstrate that Soromanes and Polybotrya are closer genealogically to each other than to any other genus; that is, they are sister groups. In my opinion, these similarities justify including Soromanes in Polybotrya. Subgenus Sorhifolia is primarily South American (Maps 3 & 4) and consists of four species (species nos. 5-8) that occur in low- elevation forests primarily from 0-I2(X) (1500) m (Fig. 4). Two species, P. sorhifolia and P . fractiserialis , are very closely related and contrast sharply with the rest of the genus because of their terrestrial stems. These two species also have distinctive venation: the veins are long, parallel, and sharply ascend- ing (Figs. 24 & 25). The third species, P. crassirhizoma, is dissimilar from the others and has dull orange stem scales and slightly more spreading veins (Fig. 26). It is an abun- dant fern in the western Amazon basin, from Colombia south to Bolivia. Polybotrya es- piritosanlensis is included in this subgenus, even though its leaves are 2-pinnate because its venation (Fig. 27b) is exactly like that of P. fractiserialis (Fig. 25e); this similarity suggests a close relationship. The species having decompound leaves, ail free veined (species nos. 9-35), make up the subgenus Polybotrya, which is further subdivided into four species groups. Since I see no purpose in giving these groups formal taxonomic names, I informally refer to them with the name of a typical, widespread species for each group, e.g., the "P. caudata group." The first three groups described below have catadromic pinnules; the fourth has anadromic pinnules. Pinnule arrange- ment, however, cannot be used to classify the species groups at a higher level because no other characters correlate with it. TheP. c«(«/«/« group (species nos. 9-12) is the most distinctive because it has peculiar, caudate fertile pinnules that are soriferous on both surfaces, that is, amphiacrostichoid. Ac- tually, what looks like the adaxial soriferous surface is the expanded, thickened margin of the receptacle; the true, phylogenetic, adaxial surface is reduced to a thin green line (see Morphology and Anatomy section). In addi- tion to this diagnostic sorus, the group is further united by the presence of whitish, septate, acicular hairs (Fig. IOa,f,i). Poly- botrya glandulosa is tentatively assigned to this group because of similarities in leaf cut- ting and pubescence; however, it has botryoid fertile leaves. This character is constant in P. pubens. but both P. caudata and P. goya- zensis can be glabrous. The P. caudata group is most frequent and abundant at low eleva- tions from 0-l(X)0 m, although P. caudata itself occasionally occurs up to 1900 m (Fig. 4). The second species assemblage, the P. alfredii group, consists of eight primarily An- dean species (species nos. 13-20). This group has a characteristic appearance, al- though it is difficult to describe because few features are constant. In general, the pinnae are short-stalked and crowded near the costa and rhachis. The pinnae bases are more or less equilateral and not acroscopically pro- longed as in the rest of the genus. The tertiary .segments are often oval or rounded, never narrow or strap-shaped. The grooves of the costules are either truncated by the ridges of the costa or very weakly admitted to the groove of the costa (see Morphology and Anatomy section). Most species in the P. alfredii group have botryoid fertile leaves. The third species group, the P. altescan- dens group, consists of seven species (species nos. 21-27). These species, with the excep- tion of one endemic to the mountains of Costa Rica, grow in the Andes at slightly higher altitudes than the rest of the genus (Fig. 4). This group has pinnule ba.ses slightly to strongly asymmetric, a characteristic that helps to distinguish it from the P. alfredii group. The fourth species group centers on the type of the genus, P. osmundacea. and differs from the rest of the genus by having anadrom- ically arranged pinnules. As shown in the illustrations, the species (species nos. 28-35) are very similar in dissection of the leaf. The 34 Illinois Natural History Survey Vol.34, Art. 1 pinnule base ascends obliquely on the basi- scopic side and prolongates conspicuously on the acroscopic side. This asymmetry imparts a distinctive appearance to the pinnule base (Fig. 48). Relationships with Other Groups Pteridologists have always classified Poly- hotrya with Dryoptehs and such closely re- lated genera as Arachniodes, Cyclodium. Cyrtomium, Maxonia, Olfersia, Polystichop- sis, Polystichiim. and Stigmaiopteris. These genera share the following features with Poly- botrya: base chromosome number .v = 41, dryopteroid costa-costule architecture, spores monolete with prominently inflated perispo- rial folds, more than three vascular bundles in the petiole, and petioles and petiolules not articulate at their bases. Familial Relationships The dryopteroid genera are most closely re- lated to the tectarioid ferns, such as Ctcnitis. Lastreopsis, and Tectaria. These genera con- stitute a natural group on the basis of their costa-costule architecture and special "C/fvi/V/.s-hairs." As noted in the Morphology and Anatomy section of this monograph, Polyhotrya lacks Crt'/;/7/.v-hairs, and their ab- sence argues strongly against a close relation- ship with the tectarioid genera. Furthermore, Polyhotrya has costa-costule architecture of the dryopterioid type rather than the tectarioid type. Polyhotrya might possibly be construed as a member of the Lomariopsidaceae. Three genera of Lomariopsidaceae — Lomagram- ma, Lomariopsis, and Teratophylltim— closely resemble Polyhotrya in overall habit because they have densely scaly, high- climbing stems, strongly differentiated sterile and fertile leaves, and nonindusiate, acros- tichoid sori. Furthermore, at the anatomical level, the three genera have diplodesmic ve- nation and dark, sclerenchymatous sheaths surrounding each meristele in the stem. Al- though no pteridologist has ever classified Polyhotrya with the Lomariopsidaceae, these similarities prompt me to consider the possi- bility. Stem anatomy is tremendously important in the taxonomy of Polyhotrya and the Lomariopsidaceae. The stem anatomy of Polyhotrya is unique and has been thoroughly described in the Morphology and Anatomy section. The stem anatomy of the Lomariop- sidaceae is also unique among ferns because it has a broad, strap-shaped, ventral meristele that differs from the remaining circular or oblong meristeles (Fig. 16f,h). \jkt Polyhot- rya. each meristele is surrounded by a dark, sclerenchymatous sheath. This elongated ventral meristele creates a dorsiventral dic- tyostele. Presumably, this ventral meristele was formed during phylogeny by the fusion of two, once-distinct meristeles like the upper ones (Holttum 1978). In fact, the two ventral meristeles have not completely fused in Lomariopsis, as evidenced by the shallow indentation in the ventral band (Fig. I6f). The ancestor of the Lomariopsidaceae, there- fore, had a radially symmetrical dictyostele of several to many, circularly arranged meri- steles with each meristele surrounded by a dark, sclerenchymatous sheath. In other words, the ancestral stem anatomy was like that of Polyhotrya. If one looks at Figure 16g. a cross section of a Polyhotrya stem, and imagines what it would look like if the two or three ventral meristeles were fused into a single meristele, the result would be a replica of the stem anatomy of the Lomariop- sidaceae. This compelling similarity in stem anatomy suggests that Polyhotrya and the Lomariopsidaceae arose from a similar an- cestral slock among the dryopteroid ferns. The possibility also exists, however, that parallel evolution accounts for these similarities in stems adapted for climbing. The leaf architecture of the two differs, hovve\er. and most species of Lomariop- sidaceae have siniph pinnate leaves. The leaves of Polyhotrya. like those of most drNopteroid ferns, are primitively decom- pound. All Lonianopsidaceae, except B()/- hitis. have articulate leaves and/or pinnae. In November 1987 Monograph of Poi.ybotra 35 contrast. Polyhutrya and other dryopteroid genera have nonarticulate leaves and/or pin- nae. In the climbing genera Lomuriopsis and Teratophyllum. leaves on the terrestrial por- tion of the stem (bathyphylls) differ greatly from those on the scandent portion (ac- rophylls). In Lomahopsis, the bathyphylls are less divided (usually simple and entire) than the acrophylls, and in Teratophyllum they are more dissected than the acrophylls. In Polybotrya . however, the bathyphylls and acrophylls are about the same size and shape. Venation is yet another difference. Sev- eral kinds of venation occur in the three gen- era of Lomariopsidaceae most similar to Polybotrya. Teratophyllum has simple or forked veins, as does Lomariopsis. except that the veins of the former unite with the cartilaginous margin (Holttum 1978). Polybotrya also has free veins, but the pattern is entirely different from that of these two genera, as can be seen by comparing the il- lustrations in this monograph with those .shown by Holttum (1978, Figs. 1-8). The veins ofLomagramma differ completely from those of these three genera; its veins form a network of three or more rows of oblique areoles without main veins. Venation, there- fore, does not support a relationship between Polybotrya and the Lomariopsidaceae. In summary, Polybotrya and some genera of Lomariopsidaceae share the following characteristics; climbing stems, dimorphic sterile and fertile leaves, diplodesmic veins, and dark sclerenchymatous sheaths surround- ing each meristele. The two groups, however, contrast sharply in such features of leaf ar- chitecture as the amount of dissection, pinnae articulation, acrophylls versus bathyphyll dif- ferences, venation patterns, and stem anatomy. This conflicting evidence is dif- ficult to assess. Certainly, the climbing habit, sterile-fertile leaf dimorphy, and diplodesmic veins have arisen many times in ferns, but the similarity in stem anatomy is less easily explained. I suspect, however, that the Lomariopsidaceae may have had a separate origin among the dryopteroid ferns, apart from Polybotrya. Strong evidence exists that Polybotrya was derived from a Cyclodium- like ancestor (see below). Relation to Similar Dryopteroid Genera Maxonia. This monotypic genus was first described by Christensen (1916), who ob- served that his new genus "must stand be- tween Polybotrya and certain species of Dryopteris grouped with D. ampli.s.sima [Arachniodes]." Copeland (1947) also held that Ma.xonia apiifolia (Swartz) C. Chr. rep- resented a phylogenetic intermediate between Arachniode.s and Polybotrya (Fig. 17), and most later pteridologists have agreed that Ma.xonia is intimately related to Polybotrya. Maxonia and Polybotrya appear very similar because both have highly differen- tiated sterile and fertile leaves and densely scaly, climbing stems. Nevertheless, a com- parison of stem anatomy shows some impor- tant differences (Fig. I6e,g). The most evi- dent dissimilarity is that each meristele of Ma.xonia is not surrounded by a dark scleren- chymatous sheath as in Polybotrya. Also, the dictyostele of Ma.xonia is dorsiventral (Chandra 1975; Walker 1972), unlike that of Polybotrya, which is radially symmetrical. Clearly, Mcixonia lacks the stem anatomy that characterizes Polybotrya, thus weakening the hypothesis that these genera are closely related. Strong evidence exists that Maxonia is most closely related to the American species of Arachniodes, in particular y4. macrostegia (Hooker) Tryon and Conant. Indeed, if Maxonia apiifolia had monomorphic leaves it would without doubt be placed in Arachniode.s. The similarities between the two genera are striking and are best seen in the sterile leaves. Both genera have pinnules arranged anadromically, basal pinnae elon- gated basiscopically, and laminae broadened notably at the base and of similar thickish texture. In addition, the pinnules and smaller segments have cuneate bases and acute apices, thereby imparting a characteristic "streamlined" appearance to the lamina that 36 Illinois Natural History Survey Vol. 34. Art. 1 contrasts sharply with the truncate or acro- scopically prolonged bases and rounded apices in Polyhotrya. Both Maxonia and Arachniodes have costa-costule architecture of the dryopteroid type (grooves decurrent into each other; Holttum 1984). Finally, both genera have remarkably similar brown, thick- ish, round-reniform indusia. The evolution of the climbing stem of Maxonia poses no problem \i Arachniodes is accepted as the ancestor. Since the American species oi Arachniodes have creeping stems, the potential to evolve a fully hemiepiphytic stem, as in Maxonia, was initially present and easily achieved. In short, Maxonia is similar to Polyhotrya because it arrived at a comparable evolution- ary grade or level, but it came from a different source (Fig. 17). Polyhotrya was probably derived from Cyclodium (see below), and Maxonia from Arachniodes, probably from an ancestor close to -4. macrostegia. Maxonia should not, however, be subsumed with Arachniodes because it is derived from it. It is convenient, and certainly in harmony with past taxonomic practice, to distinguish Maxonia generically on the basis of its leaf dimorphism and climbing stem. Olfersia. This genus consists of a single species, O. cervina (L.) Kunze, the place- ment of which has been controversial. Most recently, pteridologists have placed Olfersia in Polyhotrya because both have strongly di- morphic sterile and fertile leaves, nonin- dusiate sori, similar perispore morphology, and densely scaly, creeping stems. Yet Olfer- sia differs from Polyhotrya by its venation, conform apical pinna, and stem anatomy. Moran ( 1 986) studied Olfersia and concluded that it and Polyhotrya are sister taxa best maintained in separate genera. Cyclodium (sensu Smith 1986). Polyhot- rya was probably derived from an ancestor which, if it were alive today, would be placed in Cyclodium. This genus has two important prerequisites for the evolution of Polyhotrya. First, the fertile leaves of Cyclodium are slightly to strongly dimorphic. Second, the stem is creeping in all species of Cyclodium, and in several species it becomes fully hemiepiphytic. Cyclodium. therefore, has the genetic capacity to evolve two cardinal fea- tures of Polyhotrya: holodimorphic sterile and fertile leaves and a climbing stem. Moreover, Cyclodium contains species that look remarkably like certain sp)ecies of Polyhotrya. The 2-pinnate-pinnatifid leaf of C. trianae (Mett.) A.R. Smith var. trianae is strikingly like some species of Polyhotrya with respect to cutting and venation (cf. Fig. 16c to P. caudata. P. osmundacea. and P. sessilisora. Figs. 28, 47, & 51 . respectively). In fact, several times during fieldwork in Ecuador, I mistook sterile leaves of C. trianae for terrestrial leaves of P. caudata. Similarity is also seen in the simply pinnate lamina of C. meniscioides (Willd.) Presl, which has anastomosing venation notably like that found in Polyhotrya subgenus Soromanes and is also simply pinnate (Fig. 16a,b). This similarity of venation was noted long ago by Hooker and Baker (1874). I am not suggesting that these two species of Cy- clodium gave rise to Polyhotrya. but within both genera, species have evolved with simi- lar characteristics, and this homoloaous vari- Polytx)trya T Maxonia T Arachniodes Maxonia Polyhotrya T t Arachniodes Cyclodium T T Ancestor {Dryopteris?) FiciURi 17. Two phylogcnctic hypotheses concerning the relationships of Maxonia and Polyhotrya: left, the hypothesis proposed by Christensen (1916): right, the hypothesis proposed here. Sec text for explanation of characters involved. November 1987 Monograph of Poi.ybotra 37 ation implies a close genetic relationship. In conclusion, the close relationship between Cyclodium and Polybotrya is evidenced by their mutual possession of dimorphic leaves, creeping stems, similar evolutionary tenden- cies, and, at least in some species, remarka- bly similar leaf dissection and venation. Part Two: Taxonomic Treatment //; evohitionary biology almost all phenom- ena and processes are explained through in- ferences based on comparative studies. These, in turn, are made possible by very careful and detailed descriptive studies. It is sometimes overlooked how essential a com- ponent in the methodology of evolutionary biology the underlying descriptive work is. Ernst Mayr (1982) Notes on the Presentation of Data In most cases, the key can be used with speci- mens lacking fertile leaves. Fertile leaves, however, are often helpful in identification and are given as additional characters in many couplets. The key will generally not work when leaves are less than 45 cm long. To facilitate the comparison of similar species, 1 have arranged the species phylogenetically rather than alphabetically. The synonymy given for each species is complete. 1 have kept the descriptions short by not repeating characteristics that are constant for the genus. For example, all species of Polybotrya have a pinnatifid leaf apex; this characteristic is mentioned only in the generic description and is not repeated in each species description. The distribution maps (see appendix) were compiled from the specimens listed in the Specimens Examined section in each species treatment. In some cases, I could not find localities despite searching various atlases, indices, and gazetteers. The terms basiscopic and acroscopic are used frequently in the key and in descriptions to refer to the sides of pinnules (Fig. 18). The acroscopic side of a pinnule is the side directed toward the apex of the pinna bearing the pinnule. Conversely, the basiscopic side of a pinnule is that side directed toward the pinna base and rhachis. Two other frequently used terms, catadromic and anadromic , refer to pinnule arrangement. The pinnule arrange- ment is catadromic when the pinnule closest to the rhachis is basiscopic; when the pinnule closest to the rhachis is acroscopic, the ar- rangement is anadromic (Fig. 18). I have de- basiscopic side costule rhachis PINNULE acroscopic side tertiary segment costa catadromic anadromic PINNAE FiuuRK 18. Terms frequently used to describe the leaves of /'o/yfcofrya. See text for further cxplantions. 38 Illinois Natural History Survey Vol. 34. Art. 1 fined the specialized terms that refer to the fertile leaves, such terms as botryoid. coenosoric, and amphiacrostichoid, in the Morphology and Anatomy section and in the Description of the Genus section. Description of the Genus POLYBOTRYA WILLD Polybolrva Willd., Species Plantarum. ed. 4. 5;99. 1810. Type species; Polyhotrya osmun- dacea Willd. Soromanes Fee. Mem. Fam. Foug. 2 (Hist. Acrost.) 16. 1845.Type species; Soromanes serratifolium Fee = Polyhotrya serratifolia (Fee) Klotzsch. Bom'olhallus Klotzsch. Bot. Zeit. 4;104. 1846. nom. nudum. Type species; Botryothallus kunzei Klotzsch nom. nudum = Polyhotrya serratifolia (Fee) Klotzsch. Stem hemiepiphytic, long-creeping, or (two species) terrestrial and short-creeping, 1 .0-3.5 cm wide (excluding scales), the scandent portions un- branched. the terrestrial portions occasionally branched; dictyostelic. in cross section with a characteristic pattern of 4-10 circularly arranged meristeles. each meristele surrounded by a dark sclerenchyma sheath; aerophores present in most (all?) species, continuous and decurrent from the lateral aerophores of the petiole, darkening upon drying and not visible in herbarium specimens, in fresh material appearing yellowish, linear, 1-3 mm wide, slightly raised and bearing stomata; rootlets produced only on the ventral surface; scales numerous, densely covering the stem, spreading or appressed, primarily of two general types; 1) thick, dark brown, opaque, margms en- tire , attached across the length of the curved . thick- ened base, and 2) thin, variously colored (ranging from yellow, orange, red to brown), translucent, margins denticulate or highly erose. anached at a central, basal point. Sterile leaves up to 2.0 m long, reclining when mature, intcmodes generally 10-15 cm apart, bulbils lacking;/?*'//*'/*- with 8-16 vascular bundles arranged in a mushroomlike out- line, the base decurrent for a short distance on the stem, scaly, with scales shorter and wider than those on the stem, the aerophores present laterally as thin yellowish green lines, the adaxial surface slightly tlattcned, trisulcate in dried material; color stramineous, light green or darkened abaxially; rhachis trisulcate below becoming unisulcate above, the groove pubescent within; lamina papyraceous to subconaceous. 1 -pinnate (sub- genera Soromanes and Sorbifolia) or 1-pinnate- pinnatifid to 4-pinnate (subgenus Polyhotrya), usually lanceolate, the base rarely deltate or cuneate, the apex pinnatifid; pinnae not articulate to the rhachis. linear, lanceolate or deltate. sym- metrical or slightly prolonged acroscopically. a few species having the basal pair slightly pro- longed basiscopically; pinnules arranged caia- dromically or anadromically, symmetncal or asymmetrical at the base, if asymmetrical then with the basiscopic side reduced and oblique and the acroscopic side prolonged; grooves ofcosiules decurrent into the the grooves of the costa or trun- cated by the ridges of the costa and therefore not decurrent (intermediates exist); veins free or (in subgenus Soromanes) anastomosing, ending close to the margin, hydathodes absent: hairs of two types: 1 ) all species have tiny, reddish, appressed. jointed, flattened hairs on the lamina that grade into the scales (these actually represent reduced scales), and 2) most species have hairs that do not intergrade with the scales, these are whitish and cylindrical, ranging from unicellular and less than 0.1 mm long to 15 cells and 2.0 mm long, or uncinate. Fertile leaves appearing as a skel- etonized version of the sterile, more ephemeral than the sterile, produced only on the scandent portion of the stem in hemiepiphytic species: son nonindusiate and of three types; 1 ) round, discrete, usually at the tips of pinnately arranged stalks (botryoid type). 2) oblong or linear formed by the fusion of several different sori (coenosoric type), and 3) sori apparently occupying both surfaces of a caudate pinnule (amphiacrostichoid type); di- plodesmic veins present except in botryoid sori: sporangia with 64 spores per capsule, the capsule glabrous except in P. puhens which has setose hairs at the apex near the annulus: sporangial stalks with three rows of cells at the apex, two-rowed below, paraphvsate in most species by lateral hairs, these multicellular and unbranched (branched onlv in P. speciosa): annulus of 15-22 indurated cells: spores (32)45-65(80) microns long, dark brown when viewed with transmitted light under a compound microscope, deep orange w'hen viewed with retlected light under a dissect- ing microscope, monolete. aperaturc linear and i/,-y4 the length of the long axis, exospore smooth, pcrispore bilayered with inflated folds and cchinate to various degrees. .v = 41. November 1987 Moncxiraph of Polybotra 39 Key to the Species of Polybotrya 1. Sterile leaf 1-pinnate. 2. Veins of sterile leaf anastomosing (subgenus Saromanes). 3. Plants of Mcsoamerica; vein tips joined into a faint, continuous, submarginal connecting strand 2. P. potyhotryoides. p. 80. 3. Plants of South America and Trinidad; vein tips usually, but not always, free. 4. Lamina pubescent abaxially with uncinate hairs. Western Cordillera of Ecuador. 5. Hairs on abaxial surface 0.1-0.3 mm long, colorless, erect, 1-3 celled; free pinnae pairs 6-13 below the pinnatifid apex, mostly 15-21 x 4-5 cm 3. P. suberecia, p. 50. 5. Hairs on abaxial surface 0.5-1 .2 mm long, tawny, spreading, 5-12 celled; free pinnae pairs 4-7 below the pinnatifid apex, mostly 17-33 x 5.7-10 cm 4. P. andinii. p. 50. 4. Lamina glabrous abaxially. 6. Apex evenly pinnatifid, not similar to the lateral pinnae (Fig. 20a); vein tips free or in- completely anastomosing. 7. Stem scales linear, stiffish, mostly 0.3-1.0 mm wide, usually opaque and concol- orous; fertile pinnules more than 2.0 cm long. Colombia, Ecuador, and Peru. ... 3. P. suberecia, p. 50. 7. Stem scales lanceolate, flaccid, mostly 1-2 mm wide, often with a dark central stripe and lighter borders; fertile pinnules generally less than 2.0 cm long. Andes of Vene- zuela; Trinidad 1. P. serralifolia, p. 43. 6. Apex subconform, similar to lateral pinnae, or with one or a few small basal lobes (Fig. 20b, c); vein tips joined into a faint continuous submarginal connecting vein (see text for observation of this character). Southern Mexico to Peru 2. P. polybotryaides, p. 80. 2. Veins of sterile leaf free (subgenus Sorbifolia). 8. Stem terrestrial and short-creeping; scales brown, essentially concolorous; fertile pinnules com- monly round or oblong, usually less than 1 cm long. 9. Largest pinnae 7-10 times longer than broad, abaxial surface often with white or light brown, sessile, globose glands; apex of 3-7 pinnalike lobes (Fig. 24a), these with long decurrent bases; stem scales shiny brown, mostly transparent and denticulate; spores 40-47 microns long 5. P. sorbifolia, p. 53. 9. Largest pinnae 5-7 times longer than broad, abaxial surface without glands; apex of 2 or 3 lobes, their bases not long decurrent (Fig. 25b, c); stem scales dull brown, mostly opaque with entire or subentire margins; spores mostly 52-56 microns long 6. P. fractiserialis. p. 54. 8. Stem hcmicpiphytic and long-creeping; scales reddish brown or orange, with a prominent cen- tral stripe and lighter borders; fertile pinnules linear and usually more than 1 cm long 1. P. crassirhizoma, p. 58. 1. Sterile leaves 1-pinnatc-pinnatifid to 4-pinnatc (subgenus Polyholrya). 10. Pinnules of medial pinnae catadromic (Fig. 18). 11. Rhachis and costae pubescent, hairs 1.0-2.5 mm long. Amazonian lowland forests. 12. Lamina more than 15 cm wide at the base; petiole more than 15 cm long; largest pinna lobes with entire margins; stem scales usually denticulate, translucent, cream to casta- neous; fertile pinnules caudate. 3-15 mm wide, sori acrostichoid, apparently covering both surfaces; sporangial capsules setose (Fig. 30e,0 \\. P. puhens, p. 68. 12. Lamina up to 10 cm wide at the base; petiole up to 4 cm long; largest pinna lobes with ser- rate margins; stem scales entire, dark, and opaque; fertile pinnules botryoid, sori round, di.scretc, 1-3 mm wide; sporangial capsules glabrous 12. P. f>lundulosa . p. 71. 1 1 . Rhachis and costae glabrous, or if pubescent, hairs shorter than I mm long. Amazonia and elsewhere. 40 Illinois Natural HrsTORV Survey Vol. 34. Art. 1 1 3 . Ultimate segments or lobes of sterile leaf 0.5 - i . 5 mm wide and only one-nerved; lamina often pubes- cent on both surfaces. Andes of Colombia to Bolivia, Guyana 13. P. lechleriana. p. 71. 13. Ultimate segments or lobes of sterile leaf more than 1 mm wide and with several nerves; lamina rarely pubescent on both surfaces. 14. Stem scales bright golden or yellowish a/u/ the pinnatifid portions of the pinnae or pinnules with a single veinlet running towards the sinus arising directly from the costa or costule between the main lobes or vein groups (Fig. 45d). Andes of Colombia to Peru 26. P. aliescandens. p. 97. 14. Plants without the above combination of characters. 15. Laminar margins sparsely ciliate, hairs minute, less than 0.1 mm long (Fig. 28a) d/u^ fertile pinnules caudate, son amphiacrostichoid, covering both surfaces of the leaf. 16. Stem scales dull brown, opaque, appressed-ascending, margins subentire, base curved and thickened (Fig. 28h). Widespread 9. P. caudata, p. 60. 16. Stem scales shiny reddish to castaneous. membranous, spreading, margins dentic- ulate, base usually cordate at point of attachment (Fig. 29f). Paraguay and Brazil. 10. P. goyazensis. p. 66. 15. Laminar margins glabrous, or if sparsely ciliate, fertile pinnules not caudate and sori not am- phiacrostichoid. 17. Abaxial surface of sterile leaf pubescent, hairs fine, erect, whitish, less than 0. 1 mm long and costae scaly with numerous, golden brown, tortuous scales. Panama 23. P. alata. p. 92. 17. Plants without the above combination of characters. Panama and elsewhere. 18. Plants nearly glabrous throughout, even within the grooves and pinnule mar- gins cut less than Vi of the way to the costule. Costa Rica 27. P. gomezii, p. 99. 18. Plants usually pubescent on the major axes and within the grooves; pinnules cut more than % of the way to^e costule. 19. Lamina margins sparsely ciliate, hairs 0.1 mm long and stem scales opaque, appressed, the base curved and thickened. Southeastern Brazil. 20. P. cylindrka. p. 86. 19. Lamina margins glabrous; stem scales as above or thin, spreading, and trans- lucent. Plants not from southeastern Brazil. 20. Pinnule bases more or less symmetrical (Fig. 35); tertiary segments often inserted at about right angles to the costule. ovate to oblong, usually less than 2.5 (3.0) times longer than broad. 21 . Pinnules up to 2.5 X 1 .2 cm. with only 5-7 segments and;or lobes; lamina up to 55 x 26 cm. apex long-attenuate; major axes with nar- row, dark, tortuous scales. Colombia 14. P. attenuata. p. 74. 21. Pinnules larger than 2.5 x 1.2 cm, with more than 5 segments and/or lobes; lamina often larger than 55 x 26 cm, apex rarely long- attenuate; major axes rarely with narrow, dark, tortuous scales. 22. Sterile lamina slightly reduced at the base; basal pinnae less than 3.5 times longer than broad, usually elongated basiscopically; fertile leaves coenosone; sori not stipitatc but placed close to the midrib. 23. Sterile lamina pubescent adaxially. Colombia 15. P stohei. p. 76. 23. Sterile lamina glabrous adaxially. 24. Margins of tertiary segments entire to crcnate. serrate or lobcd; pinnules of medial pinnae cut more than half way to the costule for most of their length in pinnae 25 cm or longer, Mcsoamcrica. Ecuador, and Bolivia. 16. P. alfredii. p. 78. November 1987 Moncxiraph of Polybotra 41 24. Margins of tertiary segments entire, never crenate, serrate, orlobcd; pinnules of medial pinnae cut less than half way to the costulc for most of their length. Colombia. ... 18. P. lourleigiana, p. 82. 22. Sterile lamina broadest at the base; basal pinnae 3.5-4.0 times longer than broad, not strongly elongated basiscopically; fertile leaves botryoid; son stipitate, stalks 1-4 mm long. Colombia 19. P. pittieri, p. 84. 20. Pinnule bases acroscopically prolonged (Fig. 44); tertiary segments and lobes usually inserted obliquely to the costule, somewhat elongate and falcate, more than 2.5 times longer than broad. 25. Costal scales scattered, dark, ovate to lanceolate, flaccid (Fig. 52b); fertile leaves botryoid (Fig. 52c). Cloud forests of northern Venezuela 32. P. canaliculala. p. 111. 25. Costal scales absent, or if present, then linear, tortuous; fertile leaves rarely botryoid. 26. Sterile lamina 3-4-pinnate-pinnatisect, ultimate segments or lobes 1-2 mm wide; major axes on all sides with numerous, narrow (1-3 cells wide), reddish brown, tortu- ous, spreading scales; fertile leaves botryoid. Colombia 17, P. botnoides, p. 80. 26. Sterile lamina to only 3-pinnate-pinnatifid, ultimate segments or lobes more than 1-2 mm wide; major axes lacking numerous scales, or if scales present, then not as above. 27. Lamina puberulent on both surfaces; veins prominulous. Andes of Ecuador and Bolivia 22. P. puberulenia, p. 90. 27. Lamina always glabrous adaxially and usually so abaxially; veins not promin- ulous. 28. Stem scales dull brown, thick and opaque, attached across the length of the curved and thickened base, margins subentire. 29. Tertiary segments lobed, lamina usually 3-pinnate-pinnatifid. 30. Costae evenly pilosulous, hairs less than 0.1 mm long; costular grooves truncated by the ridgesof the costae and therefore notdecur- rent; grooves glabrous or nearly so within. Andes of Colombia and Bolivia 21. P. hickeyi, p. 88. 30. Costae glabrous or with only scattered hairs; costular grooves decur- rent into the costal grooves; grooves filled with reddish or brownish hairs. Widespread species 28. P. osmundacea, p. 101. 29. Tertiary segments entire or only minutely serrate at the apex; lamina mostly 2-pinnate-pinnatifid. Andes of Ecuador 25. P. appressa, p. 94. 28. Stem scales reddish brown or golden, thin and translucent, attached at a cen- tral point, base not curved or thickened, margins denticulate to erose. 31. Stem scales reddish brown; lamina with punctate, resinous glands; costal grooves packed with protruding hairs 0.3-0.8 mm long. Guadeloupe and Martinique 29. P. cyathifolia, p. 106. 31. Stem scales golden; lamina lacking punctate, resinous glands; costal grooves glabrous. Andes of Ecuador 24. P. aequatoriana. p. 94. 10. Pinnules of medial pinnae anadromic (Fig. 18). 32. Stem scales yellowish to golden; lamina usually with a single veinlet springing directly from the costa or costule between the main lobes or vein groups (Fig. 45d). Andes of Colombia, Ecuador, and Peru 26. P. altescandens. p. 97. 32. Stem scales brown, castaneous, or reddish; lamina lacking a veinlet as described above. 33. Sterile leaves 2-pinnate; pinnules entire, bases symmetrical andcuneate; fertile pinnules cy- lindrical and entire. Known only from the state of Espirito Santo, southeastern Brazil. 8. P. cspiritosantensis. p. 60. 33. Sterile leaves 2-4-pinnate; pinnules divided, lobcd, or serrate, the bases usually prolonged acroscopically; fertile pinnules lobed and/or divided. 34. Scales of the petiole base 5-6 mm wide, broadly ovate, castaneous. Colombia. ... 30. P. lalisi/uanwsa, p. 108. 42 Illinois Na I URAi. HisiOKY Survey Vol.34. Art. 1 34. Scales of the petiole base less than 4 mm wide, usually narrowly lanceolate, brown, castaneous, or reddish. 35. Abaxial surface of lamina evenly pubescent; stem scales reddish. Southeastern Brazil. 36. Costal hairs less than 1 mm long 34. P. speciosa, p. 115. 36. Costal hairs 1-2(2.5) mm long 35. P. pilosa, p. Ml. 35. Abaxial surfaccof lamina glabrous; stem scales variously colored. Southeastern Brazil and else- where. 37. Fertile leaves botryoid, i.e.. all ultimate soriferous segments round and discrete, not fusing to form an oblong or linear sorus (Fig. 52c). 38. Sterile lamina to 4-pinnate; stem scales dull to unaided eye. dark brown, with entire or subcntire margins; costal scales broadly ovate to lanceolate; costae glabrous or more rarely pubescent; hairs short, reddish, less than 0.1 mm long; sori short-stalked (Fig. 52c). Cloud forests, Andes of Venezuela 32. P. canaliculaia. p. 111. 38. Sterile lamina to 3-pinnate; stem scales shiny, dark castaneous, with denticulate-erose margins; costal scales narrowly lanceolate to linear; costae often pubescent, hairs 0.5- 1 .0 mm long, whitish; sori sessile (Fig. 51c). Lowland forests, northern Amazon basin and Guiana Highlands 3\. P. sessilisora. p. 108. 37. Fertile leaves coenosoric, i.e., some or most of the ultimate soriferous segments, especially those near the apex, fusing to form an oblong to linear sorus (Fig. 47g). 39. Abaxial surface of costulcs evenly hirsute, hairs less than 0. 1 mm long, or if glabrous, stem scales reddish. Southeastern Brazil. 40. Stem scales reddish; margins of tertiary segments or lobes crenate to lot>ed. .. 34. P. speciosa. p. 115. 40. Stem scales bright castaneous; margins of tertiary segments or lobes entire. ... 33. P. semipinnata. p. 113. 39. Abaxial surface of costulcs glabrous or variously pubescent by soft whitish hairs more than 0.1 mm long; stem scales rarely reddish. Plants not of southeastern Brazil. 41 . Costal grooves glabrous within; sterile lamina mostly 2-pinnate-pinnatifid. Costa Rica 27. P. gomezii. p. 99. 41 . Costal grooves pubescent within, hairs reddish to brownish; sterile lamina mostly 3-pinnate-pinnatifid. 42. Basal acroscopic segment of pinnules usually with a slight basal gibbosity on both margins (Fig. 49c); lamina membranaceous, almost always with reddish punctate glands abaxially. Guadeloupe and Martinique 29. P. cyathifolia. p. 106. 42. Basal acroscopic segment of pinnules lacking basal gibbosity, usually slight- ly reduced or oblique (Figs. 47 & 48); lamina thicker, papyraceous to charta- ceous, rarely with reddish punctate glands abaxially Widespread 28. P. osmundacea. p. 101. November 1987 Monograph of Poi.ybotra 43 Species Descriptions POLYBOTRYA subgenus SOROMANES (Fee) Moran comb. & stai. now Soromanes Fee, Mem. Fam. Foug. 2 (Hist. Acrost.) 16. 1845. Typf. sPF.crEs: Soromunei- serraiifolium Fee = Polybotrya serratifolia (Fee) Klotzsch. Polybotrya section Soromanes (Fee) Klotzsch, Linnaea 20:430. 1847. Polybotrya subgenus Soromanes (Fee), attri- buted incorrectly to Klotzsch by Fee, Genera Filicum 50, invalid. Acrostichum subgenus Soromanes Hooker, Species Filicum 5:256. 1864. Type species: Acrostichum caenopieris Hooker = Polybotrya serratifolia (Fee) Klotzsch. Sterile leaves I -pinnate; veins from adjacent costules anastomosing, 4-6 pairs, curved ascend- ing, the basal pair joining 'A- 'A the distance from the margin to the costa, then with an excurrent vein to the base of the above anastomosing pair, the distal vcinlcts strongly curved towards the apex, vein tips at the margin free or, in P. pohbot- ryoides. joined by a submarginal vein. Fertile pin- nae pectinate. 1. Polybotrya serratifolia (Fee) Klotzsch (Fig. 19, Map 1). Polybotrya serratifolia (Fee) Klotzsch, Lin- naea 20:430. 1847. Soromanes serratifolium Fee, Mem. Fam. Foug. 2 (Hist. Acrost.) 82, tab. 43. 1845. Type: P! (3 sheets), in Bory Herbarium, cited by Fee as "Polybotrya serrala, Galeotti, in herb. Bory, Habitat in Mcxicana rcpublica, Lagunetta (Galeotti). - V.S. in herb. Bory." but probably collected by Linden in Ven- ezuela and later distributed by Galeotti (sec text discussion). I have selected as the Icc- totypc that sheet which Fee illustrated for his tab. 43. Soromanes dentatum Fee, Mem. Fam. Foug. 2 (Hist. Acrost.) tab. 43. 1845. Fee's tab. 43 was erroneously labeled as "S. dentatum" in- stead of 5. serratifolium . His tab. 43 is a precise illustration of the lectotypc of P. ser- ratifolia. Soromanes integrifolium Fee, Mem. Fam. Foug. 2 (Hist. Acrost.) 82, tab. 42. 1845. Fertile leaf only, the sterile one is that of Cyctodium meniscioides. Type: Alexander Braun Herbarium (B). Botryothallus kunzei Klotzsch, Bot. 2^it. 104. 1846. nom. nudum, cited by Ettings- hauscn (1864). Polybotrya kunzei Ettingshausen, Denkschr. Akad. Wien. 22:66. fig. 2. 1864. Type: Only "in Colombia" was mentioned. Acrostichum caenopteris Hooker, Species Filicum. 5:256. 1864. nom. nov. for Soro- manes serratifolium Fee, non Kaulf. 1824, with same type. Polybotrya coenopteris (Hooker) Christ, Famkr. 44. 1897. Soromanes coenopteris (Hooker) Christ, Bull. Herb. Boissier, II. 3:613. 1903. Acrostichum hartii Baker, J. Bot. 371 . 1881 . Type: Trinidad, War/ 22S(holotypc:K, photo at US!; isotype: NY!). Polybotrya hartii (Baker) C. Chr., Index Filicum. 504. 1906. Polybotrya crassa Morton, Fieldiana Bot. 28:13. 1951. Type: Venezuela. Monagas: south-facing forested slopes above lime- stone bluffs, northeast of Guacharo, alt. 1300 -1400 m, 1 1 April 1945, Sleyermark 61991 (holotype:US!; isotypes: F!, VEN!). 5/emhemiepiphytic, 1 .5-3.0cmthick;scu/ej 0.8-1.5 mm long, 0.5-2 mm wide, narrowly lanceolate, reddish brown to light orange, with or without a prominent dark central stripe, membran- ous, lustrous, spreading, margins highly erose to merely denticulate at the apex. Sterile leaves up to 0.8 m long, dull green adaxially, pale green abaxially, chartaccous to subcoriaceous; petiole scaly at base, up to 30 cm long, trisulcate and flattened adaxially; lamina up to 50 x 27 cm, I- pinnate; pinnae 6-12 pairs, mostly (7)10- 20(23) X (2)3-4.5(5.5) cm wide, lanceolate, base rounded, cuncatc to truncate, margins entire to serrate, apex acuminate; veins in pinnated groups 3-6 mm apart, the tips arcuate, free; axes with a few scattered scales, rarely pubescent, the hairs tiny, less than 0. 1 mm long, unicellular, whitish; grooves nearly glabrous within. Fertile leaves 2- pinnate, often about the same length as the sterile 44 Illinois Natural His iory Survky Vol. 34. Art. 1 I cm F.GURF 19 PoWhcn'a serranfolia (Fde) Klo.^sch. a. habit; b. stem scales (note ca.se margins and d°k central stnpes,; c. petiole scale; d. s.enlc pmna; e-g. fertUe p.nnae show.ng vanat.on .n cutung^ oarK central sinptM, y^ j ,^ w -rff ^ WiiH'flfUl UW {MO) d: Aristei;uieU2 .^963 h. sporangium with paraphysis. a-c; van der Wcrjf & Ww^JwUi -^•^- " *'^'y' (VEN). c Fendler 261 (MO), f.h: Broadway 9947 (GH). g: type, probably Luuicn (P). November 1987 Monograph of Poi ybotra 45 leaves; pinnae linear to narrowly triangular, evenly long-tapered to apex, up to 14x2.5 cm; pinnules entire to occasionally slightly lobate, linear, oblong or sometimes clavate . 4 - 1 5( 20) mm long, 2-3 mm wide, the adaxial margins folding together at maturity, giving the pinnules a cylindri- cal appearance; sori coenosoric. continuously cov- ering the pinnule, occasionally lobcd at the base or botryoid; sporangial stalks paraphysate; spores (44)50-60(63) microns long. n = A\. Other illustrations: Fee's tab. 43 is an excel- lent, precise illustration of the lectotype; Vareschi. Fl. Venez.. Helechos. vol. 1, tab. 71. 1969 (as P. crassa). Polybotrya serratifolia grows in undisturbed, wet, premontane and cloud forests from 1200- 2400 m elevation. It is known only from the moun- tains of northern Venezeula and Tnnidad (Map I). Van der Werff and Smith (1980) report this species (as Polybotrya sp., aff. crassa) from the state of Falcon, Venezuela, where it grows in wet premontane forest with another Venezuelan en- demic, P . canaliculata. The distinctive, wide, flaccid, crose scales of P. serratifolia (Fig. 19b,c) distinguish it from the other species in subgenus Soromanes . Polybotrya serratifolia differs from the similar P. polybot- ryoides by its submarginal connecting vein. This vein, however, is not easily seen because the con- necting vein of P. polybotryoides is very faint and the vein tips of P. serratifolia. though free, are arcuate and simulate a submarginal connecting vein. Polybotrya hartii and P. crassa are placed in synonymy, although at first sight they look differ- ent. Both were originally distinguished from P. serratifolia by their more dissected (botryoid) fer- tile leaves. But these botryoid leaves represent a break-up of the coenosorus— a phenomenon that I interpret as atavistic. These highly divided fertile leaves arc not uncommon; even the type at Paris contains a sheet with a 3-pinnatc fertile leaf (Fig. 19g). As might be expected, intermediates exist between coenosoric and botryoid fertile leaves (Fig. 190- Still, the most common type of fertile leaf in P. serratifolia is 2-pinnate (coenosoric) with relatively short pinnules (Fig. 19a, c). In addition to its finely divided fertile leaf, Polybotrya crassa was distinguished by having an acroscopic auricle and several lobes at the pinna base. This condition represents nothing more than a part-fertile, part-sterile leaf; such transitional forms arc commonly found in other species of subgenus Soromanes and frequently cause taxonomic confusion. Accordingly. P. crassa is here placed in synonymy. Fee's citation of the type locality and collector ("Mexicana republica. Lagunctta (Galeotti). -V.S. in herb. Bory")seemstobe the result of confusion. One of the labels on the type specimen reads (my translation): "Lagunetta, sent from Galeotti, Oc- tober 1845." The word Lagunetta appears without mention of Mexico. Fee probably thought that Galeotti had collected the specimen in Mexico, the principal American country in which Galeotti collected (Morton 1971:63). Galeotti returned to Europe in 1840, where he sold many duplicates of his own and other collections. Morton (1971) observed: "Galeotti collected only in Mexico and a few specimens in Cuba but is often cited as the collector of plants from Brazil, Venezuela, and Colombia, but these plants were actually collected by Linden, and Galeotti was merely the distributor of the Linden plants." Presumably, Jean Jules Lin- den collected the type of P. serratifolia, which was later distributed by Galeotti. Although Linden collected in Cuba, Mexico, and Guatemala, his last and most profitable trip was to Venezuela and Colombia (1841-1844) where, based on other col- lections, this species is known to occur. Polybot- rya serratifolia has not been collected in Mexico. I cannot find a town called "Lagunetta" in Mexico, but a town with that name is located in the state of Lara, Venezuela, and I suspect that is where Linden collected the type. Specimens examined: Trinmjaij. Presioe 1491 (MO). 1492 (MO); heights of Aripo. Broadway 9947 (GH. US). 9949 (F, NY. US). Fay H59 (BM), Venezuki.a. Aragua: Cordillera Inierior. Cerro El Pauji. TopoF.I Pauji.al surdcElConsejo.iVcvt'rmart cS Sloddari IIS051 (GH. VEN);ccrcaTcjerias.V«r«eW 7764 (YEN); Parque Nacional "H. Pillicr." bosque dc Rancho Grande. Tschudi 167 (YEN); Colonia Tovar. Fendler 235 (BM). 261 (GH. K. MO. NY. PH. US); Colonia Tovar. Moriiz 277 (BM). Falcon: Curimagua (Core), van Cotthem 1327 (UC); Sierra de San Luis. Montana de Paraguariba. van der Werff & Winnfield 3430 (MO) Distrito P'ederal: between El Junquito and Colonia Tovar. Sleyermark 91756 (GH. YEN); El Jun- quito. Schnee 615 (YEN); off road Caracas-Colonia Tovar. in Ibresl below Club Jundolandia. Rirry 948 (YEN). Lara: Lagunetta. (jtdeolli \.ii. (P); Dito Moran, 8.7 mi SE of Sanare. Parquc Nacional Yacambu. A.R. Smith 1259 (PORT. UC); Dito. Moran. Rivero el at. I60fi (PORT). Dlto. Iribarrcn, Parquc Tercpaina, Iriarie 52 (PORT); Dtio Andres Elroy Blanco. 7 km de Sanare. 46 Illinois Natural History Survey Vol.34, Art. I Rivera 5I3A (PORT); sclva arriba de Sanare, Aris- leguiela 396J (US, VEN); 22(X) m sobre los bafios de Ri'o Claro, A.R. Smith 4055. Merida: verlientes del Rio Capaz, arriba de La Azulila, Sleyermark & Ruhe 97133 (VEN); rich forest above Hacienda Agua Blanca, above La Azulita, Sleyermark 56114 (F, US) Monoagas: south-facing forested slopes above limestone bluffs, northeast of Guacharo, all. 1300-1400 m. II Apnl 1945, Sleyermark 61991 (F, US. VEN). Portuguesa: 15 km E of Chabasquen, 67 km NNW of Guanare, Sleyermark et al . 726675 (PORT). /266W) (UC, VEN). Trujillo: arriba dc Escuquc, entre Estuque y La Mesa de San Pedro, Sleyermark 104717 (MO, VEN); Dtto. Bocono, ca. 10 mi SW of Batalal on road to Bocono, Laguna dc Aguas Negras, A.R. Smilh etal. 922 (PORT, UC); 2 km NW of Caserio Cerros de Guaramarcal, 42 km SE of Bocono, Moran 3709 (F, MO, PORT, VEN). Yaracuy: El Amparo hacia Candelaria, a 7-10 km al Norte dc Salom, Sleyermark el al. 106758 (NY, MO. PORT); Distnto Bruzual, Serram'a de Aroa, 1 1-15 km NNE of Urachichc, 3 km NE of Caserio Buenos Aires, Sleyermark et al. 124749 (PORT, UC, VEN); Dtto. Bolivar, entre las Parchitas, Tierra Fria y Ojo de Agua, Onega <$ Smilh 2491 (PORT), 2510 (PORT). 2. Polybotrya polybotryoides (Baker) Christ (Figs. 20& 21, Map 2). Polybotrya polybotryoides (Baker) Christ, BulL Herb. Boissier, U. 1:70. 1901. Acrostichum polybotryoides Baker, J. Bot. 207. 1881. Type: Colombia. Norte de San- tander: Ocaiia, on trees in the forest, 70(X) ft., Kalbreyer 1254 (holotypc: K, color slide at MO!; photo GH!, MO!). Acrostichum juglandifolium Baker, J. Bol. 207. 1881. nom. illeg.. turn Kaulfuss, 1824. Type: Colombia. Antioquia: Kalbreyer I79S (holotypc: K, color .slide at MO!; photo GH!, MO!). Polybotrya juglandifolia Christ, Bull. Herb. Boissier, II. 4:96.S. 1904. nom. now for Acrostichum juglandifolium Baker, non Kaul- fuss, with same type. Polyholrya juglandifolia Christ var. lobata Christ, Bull. Herb. Boissier, II. 6:168. 1906. Type: Costa Rica. Cartago: Rio Navarro, Coll. Inst. Costaricensis no. 16769. Werckle s.n. (P?). Polybotrya kalbreyeri C. Chr., Index Filicum. 504. 1906. nom. superfl. for Polybotrya juglandifolia Christ; with the same type. Polybotrya aucuparia Christ, Bull. Herb. Boissier, II. 6:166. 1906. Type: Costa Rica. Cartago: Valley of the Ri'o Navarro, 1400 m, Werckle 16770 (P!; photos F!, NY!, UC!). S/em 0.5 -2.0cm thick, hemiepiphytic; 5ca/« linear, mostly 9-15 mm long, 0.5-1 .0 mm wide, ascending with spreading tips, lustrous, golden yellow when fresh, turning yellow brown to purple brown upon drying, margins denticulate. Sterile leaves up to 1 .45 m long; petioles up to 45 cm long; lamina up to 1 .0 m. 1-pinnate. but occasion- ally becoming pinnate-pinnatifid to 2-pinnate in transitional sterile-fertile leaves, subcoriaceous, dark green adaxially, pale green abaxially. nearly glabrous on both surfaces: pinnae up to 12 pairs, mostly 12-20(30) X 3-6(8) cm, lanceolate, the base round, cuneate or subtruncate, with the acro- scopic side slightly prolonged, the margins entire, crenate or serrate, the apex acuminate; veins in pinnate groups 5-12 mm apart, the side branches ascending and anastomosing at acute angles with the adjacent ones (rarely, the veins fail to anas- tomose locally), the vein tips connected by a faint intramarginal connecting strand; axes nearly gla- brous or with scattered hairs, these unicellular and colorless. Fertile leaves 2-pinnate, 12-28 cm broad; pinnules oblong to linear, 0.9-2(2.5) x 1- 2 mm; sori coenosoric, continuously covering the abaxial surface of the pinnules; sporangial stalks paraphysate; spores mostly (45)50-65(68) mi- crons long. ;? = 41 . Other illustrations: Ettingshausen, Famkr. fig. 2. 1864 (as P. kunzei, portion of pinnae show- ing vein); Hooker. Icones Plant. 1877. pi. 1690 (is Acrostichum polybotryoides) x\A p\. 1691 (as A. juglandifolium): Rovirosa. Pterid. Sur. Mex. pi. 60. 1909 (as A. coenopteris): Smith. Flora of Chiapas, part II. fig. 71. 1981 (as P. aucuparia): Stolz-c. Fieldiana Bot. II. fig. 60d. I98I (portion of pinna with venation, as P. aucuparia). Polyholrya polybotryoides grows in wet. shaded, tropical forests from sea level to 1400 (1850)ni. It has the largest range of any species in subgenus Soromanes, even occurring on Cocos Island (Map 2). Gome? (1976) reported this species from Nicaragua, but I have not seen a specimen from that country . The shape of the leaf apex varies clinally from southern Mexico to Peru (Fig. 21 ). From southern Mexico to Costa Rica, the apex is pinnatifid, hav- November 1987 Monograph of Polybotra 47 FiciURE 20. Polybotrya polybotryoides (Baker) Christ, a. northern apex form. Guatemala; b. sterile leaf with intermediate apex form, Panama; c. southern apex form, Ecuador; d. fertile leaf; c. sterile pinna; f. stem scales, a: Sleyermark 37320 (F). b; Croal 12143 (MO), c.d; Moran 3569 (F). c,f; Moran 2I7H (CR). 48 Illinois Natural History Survey Vol. 34. Art. 1 ing two to four basal lobes; this apex does not resemble the lateral pinnae. South of Costa Rica, the apex resembles the lateral pinnae because it has only one or two basal lobes with the remainder of the apex relatively narrow and the margins en- tire to crenatc. Since apex shape varies clinally and no other characters correlate with it, I interpret the two extreme shapes as geographic variations of the same species and do not believe that these should be named. Plants having an intermediate apex shape do not appear to be hybrids, since they do not have aborted spores. In South America, the nearly conform apex of P. polybotryoides is taxonomically useful because the three other species of subgenus Soromanes , which might pos- sibly be confused with this species, have strongly and evenly pinnatifid apices. The subconform apex in southern populations of P. polybotryoides is derived within the genus, as evidenced by com- parison to all other species of Polybotrya and to all other genera of dryopteroid ferns. Polybotrya polybotryoides is most like P. sub- erecta. Another morphological character separat- ing these two, besides apex shape, is the submar- ginal connecting vein present in P. polybotryoides (Fig. 20e)butonly partially formed in P. suberecta (Fig. 22d). Unfortunately, the connecting vein is faint and difficult to see because of the thick lamina and, in some specimens, margins that have be- come revolute upon drying and thereby hide the region just inside the margin. The connecting strand is best seen with magnification on the abax- ial surface of the pinna or with the unaided eye and the pinna held up to strong light. Specimens examined: Mbxico Chiapas: Tum- bala, Rtnirosa 972 (GH. PH); Finca Mexiquito, Piirpus 6761 (BM, F, GH, MO, NY, UC, US), 7245 (BM, GH, US); 18-20 km N of Ocozocoautla de Espinoza, along road to Mai Paso, Breedlove & Smith 21886 (MICH, NY, UC). Bhlizk. Cayo: Hwy 28.5 mi S of Belmopan, Croat 24566 (CR. MO); Toledo district, Maya Mts , between Rio Caraval and Union Camp, Boutin & ScMosser 5902 (MO). Gliatkmala. Alta Verapaz: 7 mi up ihc road to Oxec along road off Hwy 7F. bclwecn Tucuru and El Estor ca. 6 km NE of Pan/os, Croat 4I6J7 (CR, MO). Izabal: along Rio Bonita, Cerro San Gil, Steyermark 41698a (F, US); Cerro San Gil, Steyermark 41870 (F). Quezaltenango: near Calahuache. Standley 67135 (F); Finca Pirencos. below Santa Maria de Jesus, Standley 6820J (F). San Marcos: above Finca F.l Porvenir, Vol- can Tajumuico, up I.onia Bandera Shac, Steyermark 37320 (F). Honduras. Atlantida: Lancelilla Valley, near Tela, Standley 53955 (F. US); Montana U .Manga, 30 km SE de La Ceiba, Nelson el al. 3296 (MO); Cordillera Nombrc de Dios, Gome: 7027 (CR). Comayagua: Quebrada Potrero, Cerro Azul dc Meambar, Cdme:69l4 fCR). Cortes: mountains on N side of Lake Yojoa, Morton 7629 (US). Costa rica. Cartago: Valley of the Rio Navarro. 1400 m. Werckle 16770 (P. photos F. NY. UC); along Camino Rai? de Hule. SE of Plantanillo, Croat 36800 (CR. MO); Raiz de Hule. Moravia de Chirripo. Ocampo 727 (CR); Chitaria. forest near old jailhouse. Moran 2170. 2171. 2173. 2176. 2178 (CR. F. MO. NY): Chitaria. Valeria 329 (US), 33133 (CR ); Finca Navarro. Maxon 639 (NY). Cocos Island: Wafer Bay. Gomei 3324 (CR, F, US); Wafer Valley, Pittier 16232 (CR, US); Wafer Bay river valley. Holdridge 5153 (GH). Klawe 1545 (US); trail between Chatham and Wafer bays. Gome: 18064 (CR); Chalham Bay. Jimene: M. 3209. 3210 (CR. F. GH. MICH); Chatham Bay. Four- nier 357 (NY). Puntarenas: Osa Peninsula, on ridge 9,.S km W of Rincon de Osa. Mickel 2742 (NY); above San Vito at Finca Wilson. NE of home. Evans & Bowers 3152 (MO); vicinity of biological field station at Finca Wilson, 5 km S of San Vito de Java. Mickel3l80 (NY). San Jose: Carrillo. Brade 372 (UC). Panama . Cocle: Continental divide on road to Coc- lesito. Hammel 3503 (CR) Darien: El Llano-Carti road. Churchill <4 de Nevers 4993 (MO); Serrania de Pirre, along ascent of Serrania de Pirrc above Cana Gold Mine between Rio Cana and Rio Escucha Ruido. Croat 37757 (MO). 37794 (MO) Panama: El Llano-Carti road, 13.7 km N of Pan-American Highway. Folsom 3590 (MO); Cerro Campana. ca. 10 km SW of Capira, trail to summit, Mori & Kallunki 3574 (MO. NY); I mi upstream from Frizzel's Finca Indio. on slopes of Cerro Jefe, Foster & Kennedy I8I4{V. MO); Cerro Jefe, Web- ster et al. 16467 (UC); along road to Cerro Campana. Croat 14687 (F. MO. NY); trail to Cerro Campana. Kirkhride & Hayden 274 (MO. NY); Cen-o Campana. FSU Field Station. Kennedy et al. 2074 (MO); cloud forest on Cerro Campana above Su Lin Motel. Crixj/ 14742 (MO); Cerro Campana. above Su Lin Motel. Croat 4266 (MO); Cerro Campana. near FSU building. Croat 121 14 (F. MO). 12143 (MO. US). 14786 (MO); 8-12 km N of El Llano, along new El Llano-Carti road. Nee et al. 8768 (MO); Cerro Campana. lower slopes abtne FSU cabin, Crra;i 3565 (F. GH. Q. QCA); Tinalandia Resort. N side of Rfo Toachi. about 25 km E of Sto. Domingo. Moran 3561 (F. GH. Q. QCA); ca. 2 km N of Mindo. Hacienda San Vicente. Fo.sler 85-37 (UC). Pfrl'. Junin: Pro\ Tamia. Chanchamayo. Esposlo 10928 (USM). 4. Polybolrya andina C Chr (Fig. 23. Map 2). Polybolrya andina C. Chr.. Index Filicum 7. 1905. noin. nov. for Acroslichum insigne Baker, mm Fee 1872-73. with same type. .Acroslichum insigne Baker. J. Bot. 167. 1877. nom. illeg.. mm Fee 1872-73. Typf: Ecuador. Pichincha: "Andes of Quito." So- diro (holotypc: K, photo GH!, US!: isotypes: AAU!, CW.. Q!. UC!). November 1987 Monograph of Polybotra SI rrrrf E u 5 cm FrouRF 22. Polyhotrya suherecta (Baker) C. Chr. a. uncinate hairs from abaxial leaf surface; b. stem scales; c. sterile leaf; d. sterile pinna; e.f. fertile pinnae, a: Ollgaard ct al. 37H23 (AAU). b-d: Moran 3546.5 (F). e: Smith 983 (NY), f: Dodson 7380 (US). 52 Illinois Natural History Survey Vol.34, Art. 1 FiGURF. 23. Polyholryti iimUtui C. Chr. a. fertile pinna; b. sterile pinna; c. sterile leaf: d. multicellular uncinate hairs from aba.\ial surface of the leaf; e. stem scales, a: Sodiros.n. (AAU). b-e; Moran3563 (F). November 1987 Monograph of Polybotra 53 Stem 1-1.5 cm thick, hemiepiphytic; scales mostly 8-13 mm long, dark castancous. shiny or dull, opaque, margins denticulate. Sterile leaf up to 1 .0 m long; petiole up to 30 cm long, Vi to as long as the lamina, scaly at base, becoming less so upwards; lamina mostly 60-80 x 30-50 cm, deltatc or broadly ovate-lanceolate, subcoria- ceous, glabrous above, below densely pubescent with spreading, tawny, uncinate, 4-to-IO-cellcd hairs, these 0.5-1.2 mm long; pinnae 17-33 x 5.7-10 cm, mostly less than 3.5 times as long as broad, 4-7 pairs below the pinnatitld apex, mar- gins entire, base cuneate-rounded, petiolulate, gradually becoming sessile distally; veins anas- tomosing, the tips uniting discontinuously; axes pubescent, the hairs like those on the lamina. Fer- tile leaves slightly smaller, 40-80 x 20-32 cm, pubescent with hairs like those of the sterile lamina; pinnae pectinate; pinnules proximally lobulate, becoming entire apically, 3-6 mm wide; sori coenosoric, covering most of the abaxial sur- face of the pinnule; spore length unknown. Sodiro (1897) noted that this species "grows in the tropical and subtropical region up to 1200 m in the forests around Sto. Domingo de los Co- lorados and in the Nanegal and Mindo valleys" (translation mine). This species is very rare and restricted in range and has been collected only twice from the western cordillera of Ecuador west of Quito (Map 2). I saw only three plants at the Tinalandia site. Polyhotryu siiherecta and P. poly- hotryoides also occurred at this locality and were more abundant. This species is most closely related to P. suh- erecta. from which it differs by its longer and wider pinnae, fewer (4-7) free pinnae pairs below the apex, and spreading, multicellular, tawny hairs on the abaxial surface (Fig. 23d). Pubescent plants of P. suherecta occur in the region of Ecuador where P. andina grows. Within this region, the hairs of both species are uncinate, but in P . suh- erecta they are much shorter (0.1-0.3 mm), 1- or 2-celled, erect, and easily overlooked by the naked eye (Fig. 22a). Only these two species of Polyhotryu have uncinate hairs. Because the fertile leaf of ^. andina. known only from Sodiro's type collection, had very few spores, I could not make an adequate measurement of spore length. Specimens examined: Ecuador Pichincha: Hotel Tinalandia, casi 25 km al cste dc Slo Domingo lie los Colorados, bosque primario arriba la monlana al lado norte del Rio Toachi, HKK) m, Mnrim .i563 (F. Q. QCA); "Andes of Quito." Sodiro s.n. (K, photo GH, US; AAU. GH, Q, UC). POLYBOTRYA subgenus SORBIFOLIA Moran, suhiien. now Type spixiks: P. sorhifolia Kuhn, l.innaca 34:64. 1869. Folia l-pinnata. pinnae lineares vel lanceo- latae, 6-10 sexies vel decies longiores quam lutiores. apice longiacuminato : venae sihi paral- lelae pinnatae. 3-5 in quoque lurmae, liherae. Stem terrestrial and short-creeping (2 spp.) or hemiepiphytic and long-creeping (I sp.). Sterile leaves 1-pinnate; pinnae lincarto lanceolate, 6-10 times as long as broad, apices long-acuminate; veins in pinnate groups of 3-5, strongly ascend- ing, free all the way to the margin. Fertile leaves pectinate or more rarely, moniliform. 5. Polybotrya sorhifolia Kuhn (Fig. 24, Map 3). Polyholrya sorhifolia Kuhn, Linnaca 36:64. 1869. Based on var. salicifolium Hooker and with same type. Acrostichum caenopteris Hooker var. salici- folium Hooker, Species Filicum 5:257. 1864. Typr: Brazil. Pcrnambuco; Scrrado Araripe, Gardner 1901 (holotypc: K, photo GH!; frag- ment NY!). Polyholrya salicifolia Leilinger, Amcr. Fern J. 62:54. 1972. Type: Colombia. Santander: vicinity of Puerto Berrio, between Carare and Magdalena River, 100-700 m, 8 June 1935, Haui;ht 1757 (ho\otype: US!; isotypc: BM!). Stem terrestrial, 1-2.5 cm thick; stem scales shiny brown, linear, 0.4-1 x 10-20 mm. mem- branaceous, spreading, the margins denticulate. Sterile leaves up to 1.3 m; petiole 15-50 cm long; lamina lanceolate to oblanceolate, 0.3-1.5 x 0.15-0.50 m, chartaccous, with 12-18 pinnae pairs, the apex abruptly acuminate, its lowermost lobes decurrcnt; pinnae lanceolate-acuminate, (7.3)10-22(25.5) X (0.7)1-3(3.5) cm, the base truncate on its acroscopic side, cuncate on the basiscopic side, the margins entire, crenate, usu- ally scirate at the apex, the abaxial surface gla- brous or glandular, the glands scattered, round, whitish to reddish, sessile glands; veins slender, sharply ascending, long parallel, usually oblique to the costa, the tips free; (uo with a few scattered, apprcssed .scales. Fertile leaves lanceolate, 0.3- 1 .3 X 15-40 m, pinnatc-pinnalifid;/)m/i(;(' linear, 4-15x0.7-1.0 cm, monilltorm because of the round to oblong pinnules, these 1 3 mm wide; 54 Illinois Natural Hlstory Survey Vol. 34. Art. I sori coenosoric, completely covering the abaxial surface of the pinnules; sporangial stalks paraphy- sate; spores mostly 40-47 microns long. Other illustrations: Murillo. Cat. Illustrado de las Plantas de Cundinamarca, 2; 10.^. 1966 (as P. scrrotifolici): Varcschi, Flora Vcncz., Helcchos, vol. 1, lab. 7.V 1968 (as P. serratifolia): Brade, Bradea 1(9):62, fig. 3. 1971 (as P. ser- ratifolia); Lellinger, Amer. Fern J. 62, figs. 3. 4. 10. 1972 (as P. salicifolia). The distribution of P. sorbifolia is spotty but extensive (Map 3). This species looks like P.frac- liserialis; see that species for comparison. Unlike all other species in the genus, both P. sorbifolia and P. fractiserialis are terrestrial — never climb- ing—and it is doubtful that they could climb be- cause their stems are only short-creeping. Both species grow in shaded, rocky habitats along streambanks, from 0-1200 m. 1 collected P. sor- bifolia in Costa Rica at the El Rodeo site. It is common at this locality, growing on talus along a stream in the bottom of a humid, limestone ravine that is surrounded by dry uplands. Unfortunately, the specific epithet "satici- fiilia" cannot be used for this species even though its pinnae bear a strong resemblance to leaves of certain Salix species, most notably 5. nigra and S. amygdaloides. This resemblance was noted by Hooker (1864) and Lellinger (1972). Specimens examined: CosrA Rica. San Jose: Alajuelila. Alfaro S07.i (US); El Rodeo. Moran 3145 (CR. F. GH. MO. PORT). Knighl v n (US), Gome: 7122 (CR). Hunnewelt 16514 (MICH) Vhnl/uki.a. Aragua: Parquc Nacional "H. Pit- tier," Rancho Grande, Tscluuli 162 (VEN). Steyermark elaL 95827(1 (US) Portuguese: Dtto Araure, FilaSan Jose, al oeste de Sla, Lucia, Onega cS Grimann 2707 (PORT). Yaracuy: "La Enjalma" al sur de Chivacoa. Vareschi & Pannier 26M (US. VEN). Co(.()Mi)iA, Boyaca: vallc del Rio Cusiana, entre Pajarilo y Gua/ul. Murill,> 1457 (COL). 1491 (COL). Cundinamarca: entre Nilo y la qucbrada de Agua de Diosilo. Miirillo el al. 2H9 (COL. US) Magdalena: Santa Maria, near Jiracasaca. H.H Smilh 1052 (1-. GH. L. MICH. MO, NY, PH, US). Meta: along Cono Rosa Blanca, a small stream outside of Villavicencio, Kirkbride 404 (MO. NY). Santander: vicinity of Puerto Berrio, between Carare and Magdalena River. Haught /7.57(BM. US). Bra/m Goias: .Serra Dourada. 17 km (straight line) S of Goias Velho, h km NE of Mossaniedes, An- derson 10152 (NY). Para: Serra dos Carajas. Serra Norte, near waterfall near AMAZ Exploration Camp. Silvu et al. BG 526 (AAU. F. GH. MICH. NY. UC). Pernambuco: Serra de Araripe. Gardner 1901 (NY. plioio of K specimen at GH). Roraima: Indian trail from Surucucu to Uaica, Maita Mts. Prance etal. 10466 (NY). 6. Polybotrya fractiserialis (Baker) J. Smilh (Fig. 25. Map 3). Polybotrya fractiserialis (Baker) J. Smilh. Hist. Filicum 133. 1875. Acrostichum fractiseriale Baker. Synopsis Filicum 414. 1868. Type: Peru. San Mani'n; "in sylvis monlis campana, terrestre."" Dec. 1855, Spruce 4337 (holotype: K. color slide at MO!; photo GHI. USD. Acrostichum plumbicaule Baker. Synopsis Filicum 413. 1868. Type: Peru. San Martin: Terapoto, Ad rupes secus rivularum. Aug. 1855, Spruce 4090 acaoiype. K. photo GH!. photo and fragment P!. US!). Polybotrya plumbicaulis (Baker) J. Smith. Hist. Filicum 133. 1875. Stem terrestrial. 1-1 .5 cm thick, short-creeping with inlemodes 1-3 cm long, apex scaly, behind the apex usually naked or only sparsely scaly; scales dull brown, opaque, 0.3-0.9 x 8-17 mm, ascending, margins entire to more rarely denticu- late. Sterile leaves to 1.4 m tall; petiole about equaling the lamina, stramineous to lead-gray; lamina slightly reduced or broadest at the base, chartaceous to subcoriaceous, apex with one or two major basal lobes, merely crenate-lohulate above; pinnae mostly 9-15 pairs, linear to oblong. (1 1)15-25(28) X (2.1)2.5-4.0(4.5) cm, margins serrate, especially towards the acuminate apex, base rounded to cunealc. w ith the acroscopic mar- gin usually more oblique; veins numerous, fine. and in closely parallel pinnate groups, occasion- ally with a simple vein springing directly from the cosla; axes usually glabrous or with fine, subulate, 0.1-0.2 mm hairs Fertile leaves creel. 2-pinnatc. commonly taller than the sterile, to 1.5 m long; pinnae linear, mostly 7-12(17) x 5-10(15) cm; pinnules round, oblong, or linear. 1-3 mm wide; .son coenosoric. completely covering the abaxial surface of the pinnule; sporangial stalks paraphy- sale. often with a bulbous glandular cell al base of the paraphysis; spores (48)52-56(60) microns long. November 1987 Monograph of Polybotra 55 Figure 24. Polyhoirya sorhifolia Kuhn. a. sterile leaf; b. stem; c. ferlllc leaf; d. sterile pinna; e. stem scales, a-e: Moran 3145 (F). 56 Ii.i.iNois Natural History Survey Vol. 34. Art. 1 Fic-.URE 25. Polyhotrya fracliserialis (Baker) J. Smith, a. fertile leaf: b. apex of sterile leaf; c. sterile leaf; d. stem scales; e. sterile pinna; f. stem, a: C. Schunke 666 (F). b.d.c: Moran 3536 (F). c: Gonggrijp & Stohel 3300 (MO), f: Tn.on & Tryon 5221 (GH). November 1987 Monograph of Polybotra 57 Other illustnitions: Brade. Bradca 1(9):62. fig. 4. 1971 (as P. polyhotryoides). This species, like P. sorhifoUa. is terrestrial and grows in wooded talus slopes, usually along rocky strcamsides. The altitudinal range of P. frac- lixerialis is from 200-1500 m, but about 80 per- cent of the collections come from the 400- 8(X) m range. Although principally a forest species, 1 have seen vigorous fertile plants growing along sunny streamsides in Ecuador, an observation which suggests that this species has more tolerance to higher light conditions than its cogenerics. The range of P. fracliserialis occupies two discontinuous regions: I ) the Andes from Bolivia to Ecuador, and 2) the Guianas (Map 3). A dis- tance of 900 km separates the northernmost popu- lation in Ecuador from the westernmost population in British Guiana. All populations of this species occur on the eastern side of the Andes. Their ab- sence from the western side is probably due to the extreme dryness that characterizes that side of the Andes south of the equator. Polyholrya fracliserialis and P. sorbifolia look very much alike and can be easily confused. Several characters, however, distinguish the two. The easiest feature to recognize is the shape of the medial pinnae; those of P. fracliserialis are relatively shorter and broader than those of P. sorbifolia (see key and descriptions); this stoutness is accentuated just below the apex (Fig. 25). The apex of P. sorbifolia is another difference; it has more numerous, decurrent, narrower pinnae (Fig. 24). Usually, the veins of P. sorbifolia are at a more oblique angle to the costa than those of P. fracliserialis. Another difference, albeit more rec- ondite, is found in the sporangial stalks of P. fracliserialis. which have a lightbulb-shaped gland at the base of the paraphysis (Fig. 1 If.g); P. sor- bifolia lacks such a gland (Fig. I lb). In fact, no other species of Polybolrya has a glandular cell on the paraphysis or elsewhere on the sporangium. Polyholrya sorbifolia, however, has round, sessile glands on the lamina; such glands are lacking on P. fracliserialis. Spore size is yet another diffencc; P. frac- liserialis has much larger spores than P. sorbifolia. This difference suggests that P . fracliserialis may be a p Ecuador Napo: 27 km SE of Coca, petroleum well Auca 4, Moran 3616 (F. GH. NY. Q. (JCA. US); 20 km NE of Coca, 5 km N of Joya de las Sachas, Moran 3615 (F, GH, NY. PORT. Q. (JCA, US); about 10 km SE of Tena. 3 km E of Puerto Nuevo on road to Puerto Mishahuallf. Moran 3535 (F. GH. NY. Q. (JCA. US); Bimbino, on the Rfo Pacuno, lOhrW of confluence with Rfo Napo, Whilmore 752 (BMl; Rfo Cuyabeno. about 2 km upstream from Puerto Bolfvar. Brandbyge el al. 33684 (AAU); San Pablo de los Sccoyas. Brandbyge & Asanza C 32881 (AAU); San Pablo de los Sccoyas. Rfo Wai si aya. Brandbyge ei al. 32617 (AAU); San Pablo de los Secoyas. on path to Shushufindi . Brandbyge el al. 32544 (AAU); .SO km NE of Coca. Lugo S. 3293 (GH); Canton Putumayo. Rfo Aguanco, town of Dureno, Plowman el al. 4025 (GH); Anagua. Parquc Nacional Yasuni. SEF forest project area. Ollgaard el al. 38S.t6 (AAU). 39084 (AAU) Pas- taza: Curaray. Vallc dc la Mucrte, Holm-Sielsen el al. 22484 [AAV. Q. QC.\): basin otRio Pasta/a. Pacayacu- Sarayacu region. Gill 47 (NY); village of Rio Chico. 8 km from Puyo. Shemluck 280 (F); Rfo Bobonaza, be- tween oil exploration camp ChichiaMa and Dcstacamcnio Caho Po70. Ollgaard el al. 34855 (AAU. Q. QCA); Rio Bobonaza. hclween Cachilama and the outlet of Rio Bulco. Ollgaard el al .U748 l..\.\V): Rio Pastaza. near the Peruvian border, around Destacamento Ishpingo, November 1987 Monograph of Poi.ybotra 59 Figure 26. Polyhotrya crassirhizoma Ullingcr. a. fertile leaf; b. sterile leaf; c. fertile pinna; d. sterile pinna; e. stem scales, a: 0llgaard el al. 34H55 (AAU). b-e: Moran 3615 (F). 60 Illinois Natural History Survey Vol.34. Art. 1 0llgaard el al. 34970 (AAU); 2 km W of Yuralpa, S border of Ri'o Napo. Holm-Nielsen & Jeppsen 9H7 (AAU). Peru. Huanuco: Prov Pachilae, Ri'o Pozouso, Foster 9284 (F, MO, USM). Junin: Chanchamayo Val- ley, C. Schunke 157 (F), I5H (F), 661 (F), SI2 (F); Schunke Hacienda, above San Ramon. KilUp & Smith 24605 (NY, US); E of Quimiri Bridge, near La Merced, Killip & Smith 23876 (NY, US); Hacienda Schunke. La Merced, Macbnde 5602 (F, US); La Merced, Chan- chamayo, Soukup 1065 (F). Loreto: Prov. Maynas, Quebrada Yanomono, Explorama tourist camp, Rio Amazonas above mouth of Rio Napo, Transect 6, Gentry et al. 27880 (MO), Moran 3640. 3641 (F, Q, QCA); Mishuyacu, near Iquitos, Klug 1386 (NY, US); Gamitanacocha, Rio Mazan, J Schunke 268 (F. GH. NY, UC, US), 14285 (US); Veradcra de Mazan, Croat 20763 (MO); Rio Napo near Entrada de Isla Inayuga, Croat 20551 (MO). Bolivia. Cochabamba: Prov. Carrasco, conflu- ence of Rio Leehe with Rio Isarsama, Beck 1635 (LPB). Brazil. Acre: Cruzeiro do Sul, vicinity of Serra da Moa, Prance et al. 12180 (NY). 8. Polybotrya espiritosantensis Brade (Fig. 27, Map 21). Polybotiya espiritosantensis Brade, Rodri- guesia 10:28, tab. 3. 1948. Type: Brazil. Es- pi'iito Santo: Municipality of Itaguagu. Jati- boca, virgin forest, 700-8(X) m, A. C. Brade 18224 (holotype: RB!). Stem 1-2 cm wide, hemicpiphytic; scales mostly 10-15 x 1.0-2.5 cm, bright reddish brown, spreading, membranous, the center often darkened to varying degrees, margins highly erose-dcnticu- late, occasionally fimbriate. Sterile leaves up to 1.2 m long; petiole to 30 cm long, about half as long as the lamina, scaly at base; lamina ovate, 2-pinnate proximally, becoming 1 -pinnate dis- tally, 60-90 x 50-70 cm, subcoriaceous, pale green, nearly glabrous on both surfaces, the apex subconform and pinnalike but with one or two basal lobes; pinnae alternate, 6-8 pairs, the apex conform, like the lateral pinnules, the upper pinnae simple and resembling the pinnules of the lower; pinnules anadromic throughout, 3-4 pairs, ellip- tic, 10-13 X 2-3 cm, the margins entire, the apex acuminate, the base cuneatc, the proximal pin- nules with stalks 2-4 mm long, the distal pinnules sessile: veins slender, none conspicuously thicker than the others, long, parallel. 1-2 branched, the tips free; axes with a few scattered, appressed, narrow .scales, otherwise glabrous. Fertile leaves 2-pinnate, amphiacrostichoid; pinnules caudate, entire, appearing cylindrical when malure: sporan- gial stalks paraphysate; spores (52)54-60(62) microns long. Other illustrations: Brade's original descrip- tion has an excellent photograph of the tyf)e: Brade, Bradea 1:67. tab. 6. fig. 5. 1971 (stem scales only). Polybotrya espiritosantensis is endemic to the state of Espirito Santo in southeastern Brazil (Map 21 ). 1 expect this species will eventually be found in other parts of mountainous coastal Brazil. Polybotrya espiritosantensis has the most distinc- tive laminar cutting of any species in the genus. No other species has the simple, entire pinnules that evenly taper at the base and apex (Fig 27b). Unlike all other species of Polybotrya that have pinnatifid apices, the leaf and pinnae apices of this species are conform or nearly so (Fig. 27a). The numerous, close, long, parallel veins (Fig. 27b) are like those of the 1 -pinnate species P. fractiserialis and P. sorbifolia and probably indi- cate a close relationship. Specimens examined: Brazil. Espirito San- to: Santa Thereza, 900 m, Foster & Foster 854 (GH, US); Municipality of Itagua^u. Jatiboca. virgin forest. 700-800 m. A. C. Brade 18224 (RB). POLYBOTRYA subgenus POLYBOTRY.\ Type species: P. osmundacea Willd., Sp. PI. ed. 4. 5:99. 1810. Polvhotrya subgenus Eupolybotrya Fee, Mem. Fam. Foug. (Hist. Acrost.) 2:16. 1845. Acrostichum section Polybotrya Hooker, Species Filicum 5:244. 1864. Stem hemicpiphytic; sterile leaves 1-pinnate- pinnatifid to 4-pinnate; veins free. 9. Polybotrya caudataKunze (Fig. 28. Map 5). Polybotrya caudala Kunzc, Linnaea 9:23. 1834. T>pe: Peru. Huanuco: Pampayaco, July 1829. Poeppig s.n. (B!). Polypodium adianloides Aublet. Hist. PI. Guiane Franijoisc 2:962. 1775. nomen itieg.. non Bumi. (1768). Type: Guyana. Aublet s.n. (holotype: BM!, Morton photo 6626 at B!. F!. GH!). November 1987 Monograph of Polybotra 61 E o FiGURF, 27. Polyhotrya espiritosantensis Bradc. a. sterile leaf; b. sterile pinnule; c. stem scales d fertile pinna, ad; Brade 18224 (RB). 62 Illinois Naturai History Survey Vol.34. Art. 1 Olfersia caudala (Kunze) Kunze, Linnaca 21:206. 1848. Psomiocarpa caudala (Kunzc) Presl, Epim. Bot. 162. 1849. Acroslichum cuudulum (Kunze) Hooker, Species Filicum 5:244. 1864. nomen illeg. mm Hooker (1840). Polyhotrya acuminata Kaulfuss var. v'dlosa Christ, Prim. Fl. Costar. 3( 1 ):9. 1901 . Type: Costa Rica. Limon: Llanuras de Santa Clara, 500 m, Biolky 10688 (lectotype: US!; isotype: CR!). Polyhotrya villosula Christ, Bull. Herb. Bois- sier, II. 6:168. 1906. Type: Costa Rica. Santa Clara, Las Delicias. 1897, 500 m, Pittier 10688 (P?). Polyhotrya costaricensis Brade, Bradea 1:11, tab. 1 , fig. 1 . 1969. Type: CcstaRica. Limon: Hundrisser Ranch, Atlantic shore, Sept. 1909, Brade & Brade 374 (holotype: HB; isotypes: NY!, UC!). Stem 0.5-2.5 cm thick, hemiepiphytic; scales linear-lanceolate to narrowly triangular, 8-20 x 1-2 mm, dull brown, opaque, entire to denticulate, the base elevated, thickened, curved, and ap- pressed. Sterile leaves up to 2 x 1 m; petioles 30-70 cm long, about 'A the length of the lam- ina; lamina mostly 2-pinnatc-pinnatirid, rarely 3- pinnate at the base, firm-chartaccous. glabrous to pilose, the hairs 0.2-1.5 mm long, acicular. the margins sparsely ciliate. the hairs minute, less than 0.2 mm long; pinnae up to 20-45(60) x 7- 20(30) cm. subdcltate. acuminate, subcquilateral. slightly more developed on the acroscopic side; pinnules slightly prolonged acroscopically, the ba.sc truncate to slightly cordate, catadromically arranged in the medial pinnae; veins free, some- times with a single vein springing from the costa between the pinnate groups; axes nearly glabrous or pubescent to various degrees, the hairs acicular, whitish; grooves usually pubescent within, often densely so at the junctures. Fertile leaves similar in size to the sterile, mostly 2-pinnale. but with some of the larger pinnules lobed at base; pinnae caudate. 4-8(1 1) X 0.5-1. 5 cm, apparently sorif- erous on both surfaces; spores (43)46-50(53) mi- crons long. ;i = 41 . Other illustrations: Fee, Mem. Fani. Foug (Hist. Acrost), Atlas, tab. 34, 1845; Vareschi. Fl. Vcncz., vol. 1. tab. 73. 1968; Brade. Bradea 1(9):63. fig. I; p. 67. fig. 6.; Croat. Fl. Barro Colorado Is., figs. 28 & 29. 1978 (as P. villosula); Stolze, Fieldiana Bot. n.s. 6. fig. 60a,b. 1981; Tryon & Tryon. Ferns & Allied Plants, figs. 80.8, 80.9, 80.22, 80.23. 1982. Polyhotrya caudata is one of the most widely distributed species in the genus (Map 5). It grows primarily in lowland forests from sea level to 800 m, but specimens from the Andes have occasion- ally been collected as high as 1900 m. I found this species at many of the sites where I collected in Costa Rica, Ecuador, and Peru, but not in Ven- ezuela. Polyhotrya caudata is one of only three species in the genus that occur on an oceanic is- land—Cocos Island, about 500 km (310 mi) southwest of Costa Rica. Polyhotrya villosula was said to differ from P. caudata only by its long and villous pubes- cence, but specimens are usually glabrous or fully pubescent, with few intermediate forms. Since pubescence does not correlate with any other character and I found no difference in geographical range between the two forms, I regard P. villosula as a synonym of P. caudata. The juvenile leaves in glabrous plants of this species may be difficult to distinguish from P. osmundacea. A good character to separate the two species is the minutely and sparsely ciliate leaf margin oi P . caudata (Fig. 28a); that off. osmun- dacea is always glabrous (Fig. 47). The thick, dull brown, curved, and appressed stem scales (Fig. 28b, h) also help distinguish P. caudata from many other Polyhotrya species. Specimens examined: Mexico. Chiapas: easi- em highlands near Guatemalan border. Breedlove 34101 (CAS). Guatemala. Alta Verapaz: near Rio Icvolay. near Hacienda Yaxcbnal. 5 mi NW ofCubilquitz. 250- .^()0 m, Steyermark 44675 (F); CubilquiU. 350 m, Tuercklu-im I4S (US). SUS (P. Z). 7SI2 (US). 8040B (US), S04I (GH, NY). Izabal: between Bananera and "La Prcsa" in Montana del Mico. Steyermark 3S229 (F), .i.S27l (F). .<9/97 (F): Rio Chacon. 30 m. Johnson 1221 (US); between Los Amales and Izabal. Sierra del Mico, Kellerman 7J54 (F. Wl); along Rio Fno. 65 m, Sleyer- mark .0 (CAY. Z); Saul Tampoc. Granville -fH-f I (CAY); Haule Riv. Mana: Saut Grous Tigrc. Cremers 7552 (CAY, Zl; Haute Crique Baboune. affluent de la Riv Mana, Cremers 7369 (CAY, Z); Haute Crique Waamahpann. au depart du chemin indien, Granville 971 (CAY); Frontierc Guyanc- Surinam, Tuma Humac, Granville 99 1 (CAY, Z); Crique Gabaret. 25 km de rembouchurc. layon N-W, Oldeman 1933 (CAY, NY); Cnque "Roche Fende" (aftleunl de la Comte) a 1 km environ de son embouchure, Granville B.47II (CAY, Z); W of Saul on trail to Monts Galbao. Boom & Mori 1856 (CAY, NY); Piste allant de Cilron vers le Massif du Decou Decou, Billiel & Jailin I6S3 (CAY); Region de Paul Massff du Decou Decou, Crem- ers 7951 (CAY, Z). SuRtNAM. West Rivier, 2-5 km SW of Juliana Top, 275-300 m, Imin elal. 54896 (NY); no locality, Hostmann s.n. (NY); 45 km S of Paramaribo, new road to Hanover, N of Zanderij, Tryon & Kramer 561 1 (GH, MICH, NY); about 25 km S of Paramaribo, l.imieman 4570 (GH), Kramer 1954 (MO); 3 km S of Juliana Top, 12 km N of Lucie River, 300-325 m. /mm 55/6/ (MO, NY); Para Dist . Limleman & Teimissen 15291 (Z); Suriname River. Plantage "Accaribo." d'An^renwnd s.n. (Z). Guyana. Kamuni Ck.. Groete Ck., Essequibo River, Maguire <4 Fan.thawe 22855 (GH, NY); Es- sequibo River. Persatid 372 (F); Rockstone. Gleason 582 (GH. NY); Demerara. Jenman s.n. (NY); Barima River. Jenman s.n. (NY); 3 mi S of Chodikas, Guppy 462 (BM); Essequibo River, Moraballi Ck., near Bar- tica, Richards 803 (BM); lower Cuyuni River, Sandwith 1561 (BM). Venezukla. Bolivar: Dtto. Heres, Campamenio Gun, Fernandez 951 (PORT, UC); Roraima, Schom- hurgks.n. (B, NY); without locality, anno l843,Sf/iom- burgk 1659 (B); F of Cerro El Picacho, N of l.as Nieves and Las Chicharras, 45 km N of Tumeremo, vicinity of Beborah. Altiplanicic de Nuria, 6(X)- 650 m, S/cvcrmorA: 89120 (NY, VEN); a lo largo de la Quebrada Acarabisi, limiles del Eslado Bolivar con la zona dereclamacion, Aymard el al. 952 (PORT, UC) Portuguesa: Depio Paez, Pozo Blanco, enire Acarigua y Payara, 190 m, Onega 636 {PORT. UC, VEN). Sucre: selvadcl Rcctan- gulo N del Lago dc Tuanoco, Lasser & Vareschi 3926 (VEN); vicinity of Cristobal Colon, Avicagua, Broad- way 560 (GH. NY, US) Territorio Federal Delta Amacuro: Depto. Antonio Diaz, 9°I5'N, 6()"57'W, upper reaches of riverine forest of Cano Atoiba tributary of Boca Araguao. Sleyermark el al. 115033 (VEN); Depto Antonio Diaz, vicinity of Caho Jotajana ( = ticrra alta). tributary of Caiio Guiniquina. NW of Epana near boundary with Depto. Tucupita, 9°I5'N. ai'lO'W, 50 m. Sleyermark elal. 1 15021 (MO. UC. VEN); Depto. Pcdemalcs (boundary with Depto Tucupita): Caiio Simoina. west of Isia Coucuina. S of Barra de Coucuina. .50 m, Sleyermark el al. II433I (MO, VEN). I I43.Q (MO, UC, VEN). Colombia. Amazonas: Trapecio Amazonico, Loretoyacu River, 100 m, Schulles <$ Black 8467 (GH, US). Cauca: Cali, Lehmann 2998 (BM); en la orilla opuesla Puerto Lim6n-Rio Caqueta, Mora 4430 (COL); Gorgona Island, off Nariiio. Taylor 1223 (MICH, NY); Rio Timbiqui, B.T. 443 (GH). Choc6: slopes and ridge of Loma del Cuchillo. ca. 15 km WSW of Chigorod6, 150-400 m. Leilinger & de la Sola 643 (LLP. US); upper Rio Truando. 2 km SSW of the confluence of Rio Nercua near the MADUREX Camp. 100 m. Ul- linger & de la Sola 589 (COL, LP, US); trail to Miniquia E of Puerto Mulis (Bahia Solano), 20-1 20 m. Leilinger & de la Sola 26 (CR. COL. LP. US); Rio San Juan, 3.5 km SW of Andagoya. just NE of the mouth of the Rio Suruco. 60 m. Leilinger & de la Sola 496 (COL, LP); Municipio de Rio Sucio, Parque Nacional Los Catios. alrededor del Campamento de Tilupo. 250-320 m. Forero & Jaramillo 1745 (COL, MO). Cun- dinamarca: Cordillera Oriental, MesaNegra.Gazuguan Valley. 6 km NW of Medina. 580 m. Grant 10436 (F, US). El Valle: La Cumbre, Cordillera Occidental, 1700-2200 m, Killip 11343 (GH, NY, US); Finca La Pradera ca. 6 km SW of El Cairo on trail to Rio Blanco, between EI Brillanle and Boquer6n, Leilinger & de la Sola 790 (COL, US). Narino: Pambana, between Rio Pimbi and Rio Cuembi, on Rio Telembi, above Bar- bacoas, 50 m, Ewan 16846 (BM, GH, UC). EcuAtxjR. Los Rios: Rio Palenque biological sta- tion, km 56 Quevedo-Sto. Domingo road, Evoy 104 (NY). Morona-Santiago: Pachicutza, at "Escuela is- comisional [sic] Cardinal Dofner," km 140 on road Loja Gualaquiza, 900- 1 000 m, Holm-Nielsen el al. 4489 (AAU, Q, QCA), 4495 (AAU, Q, (JCA); Sucua, Swingle el al. 70-02-05-2 (UC, US). Napo: ca.si 10 km SE de Tena, 3 km este de Puerto Nuevo por camino a Puerto Misahualli, 300 m, Moran 3534 (F, Q, QCA); 10 km al sur de Tena a Puerto Misahualli, Moran 3588 (Q, QCA); Anangu, Parque Nacional Yasuni, SEE proj- ect, 260-350 m, 0llgaard el al . 38932 (AAU, Q, QCA); 27 km SE of CiKa, Moran 3617 (MO, C, QCA). Pas- taza: Rio Bufeo, northern tributary of Rio Bobonaza, 300 m, Ollgaard el al. 43798 (AAU, Q, QCA); Rio Bobonaza, around houses between Huagraeachi and Cachitama, below Montalvo, 300 m, Ollgaard el al. 34634 (AAU, Q, QCA); oil exploration camp Chi- chirota, on the Rio Bobonaza, 300 m, 0llgaard el al. 35290 (AAU, Q, QCA) Sur de SanUnder: vicinity of Barran ca Bermeja, Haughl 1325 (MO) PlRU. Ama/.onas: Prov dc Bagua, left bank of Rio Maraiion opposite Quebrada Mirana (opposite km 277 of Maraiion road above Cascadas de Mayasi), 425- 450 m, Wurdack 2011 (GH, US, USM) Cuzco: Prov, La Convcnciiin, 73°40'W, 12°30'S, at Camp Zx;ro, l\Om. Dudley //5/.? (GH, US); Camp I, 910-940 m, in J. Knox's quadrate, Dudley 10168 (GH); Prov. Paucariambo, Mautainiza (?], 80()-9(X) m. Vargas 17800 (GH) Hu^nuco: Tingo Maria, 615-1100 m, Allard 22593 (US); Cuchcro, Poeppig s.n. (BM); hills 66 Illinois Natural History Survey Vol.34. An. 1 E of Tingo Mari'a, Croat 21191 (MO); 10 km S of Tingo Maria, Slork & Norton 9509 (F. UC. US); Pampayaco, Poeppig 201 (B). Loreto: Prov. Mayna,s. Peler Jensen's Explorama Lodge, 50 mi downriver from Iquitos at Yanamono Ck., Moran 3663 (AMAZ, USM); Prov. Maynas. about 10 km SW of Iquitos at zoological park, Moran 3671 (AMAZ, USM); Pebas on the Amazon River, Williams 1739 (F); Caballo-Cocha on the Ama- zon River, Williams 2137 (F); Ri'o Mazan, 100-125 m, C. Schunke 380 (GH, NY); Prov. Maynas, Rio Itaya, 10 km S of Iquitos, Tryon (S Tryon 5204 (GH); lower Rio Huallaga, 155-210 m, Williams 3999 (F, US); Quebrada Nawampa, Croat 17620 (MO); Puerto Arturo, lower Rio Huallaga below Yurimaguas, 135 m, Killip & Smith 30690 (NY, US); Mishuyacu, near Iquitos, 100 m, Klug 1166 (F, NY, US); Quebrada Tahuayo above Tamishiyaco, Croat 197/7 (MO); Rio Itaya above Iquitos, Croat 19220 (MO); pnmary forest 17 km SW of Iquitos on road to Puerto Almendara, Croat 18388 (MO); 1 2 km SW of Iquitos, Croat 18219 (MO). Madre de Dios: 12°49'S, 69°17' W, Prov. Tambopala, Tam- bopata Nature Reserve, ca. 30 air or 70-80 river km SSW of Puerto Maldonado at effluence Rio Torre/Rio Tambopata, 260 m, Barbour 4764. 4790 (F, MO); Par- que Nacional de Manu, Cocha Casha Biological Station, Foster P-84-42 (F). Pasco: Puerto Bermudez, 375 m, Killip & Smith 26637 (NY, US). Puno: San Gavan, Lechler 2.?2 /, 2i29 (B ). San Martin: San Roque , 1 350 - 1500 m, Williams 7681 (F, US); Cerro de Campana, Spruce 4634 (P). Bolivia. La Paz: Prov. Sud Yungas, Limite de los Deptos. La Paz/Beni, Ri'o Quiquibey, 320 m. Beck 8037 (LPB); San Carlos, Mapiri, Mapiri River, Tate 422 (LPB, NY); Mapiri, Buchtien 35 (B), 290 (NY, UC). Santa Cruz: Prov. Sara, Ri'o Yapaicani, 400 m, Steinbach 7499 (B, F, GH, MO, Z); bosque del Rio Surutii, 400 m, Steinbach 3032 (US). Brazil. Acre: Cruziero do Sul vicinity, Rio Moa between Igarape Ipiranga and Aquidaba. Prance et al. 12069 (F, LP, NY, UC). Amazonas: prope Sao Gabriel da Cachoeira, ad Rio Negro, Spruce 21 16 (B. BM, P. RB); Larges, on the Amazon River 1 km below mouth of Rio Negro, Conant 940 (F, GH, NY). Para: Serra dos Carajas, AMAZ camp AZUL, Sperling 5914 (GH, NY); Bclem, Huber 2599 (BM), 7235 (BM). 10. Polybotrya goyazensis Brade (Fig. 29, Map 6). Polybottya goyazensis Brade, Bradea 1:24. tab. 1, fig. 1. 1969. Type: Brazil. Goias: Goiania, piimai^ forest, December 1936. A. C Brade 15373 (holotype: RB!; isotype: NY!). Polybotrya macedoi Brade, Bradea 1:24. 1969. TvpF,: Brazil. Goias: IVlunicipio Jatai. Localidad Balsans. 1 November 1950, Macedo 2682 (holotype: HB; isotypes: HB, MO!. SP. US!; paratypcs; Brazil. Goias: Fazenda Queixado, 8 December 1948, Macedo 1447 (RB. SP); Fazenda Balsamo. 15 December 1948, Macedo 1521 (SP). Stem 1-2.5 cm thick, hemiepiphytic; scales spreading, membranaceous, generally 6-12(16) mm long, dark castaneous or dark reddish with lighter borders, margins strongly denticulate to erose, the base cordate to various degrees, often black and sclerified at the point of attachment. Sterile leaves up to 1 .45 m long; petiole 'A to V2 as long as the lamina; lamina mostly 2-pinnate- pinnatifid, lanceolate to ovate, 1.2 x 0.9 m, the abaxial surface nearly glabrous to densely pubes- cent, the hairs whitish, acicular, up to 1 mm long, glands occasionally present, these reddish, sessile, resinous; pinnae up to 45 x 23 cm, lanceolate to narrowly deltate, the suprabasal ones soon becom- ing pinnatifid; pinnules narrowly deltate, mostly falcate, catadromically arranged above the base. 4-8( 10) X 1-2(3) cm, the acroscopic side slightly prolonged, the margins entire to crenate. ciliate, the hairs minute, 4-10 celled; axes nearly glabrous abaxially or densely pubescent, by tiny, unicellu- lar hairs, scales very small or lacking, adaxially pubescent in the central groove with reddish hairs, 0.3-0.8 mm long; grooves usually ciliate on the ridges. Fertile leaves 2-pinnate, amphiacrosti- choid: pinnules caudate, often lobed or undulate at the base; sporangial stalks paraphy sate; spores (44)48-62(70) microns long. Other illustrations: See original description of P. goyazensis: Sehnem, Fl. llust. Catarinense, pi. 18, 1979. Polybotrya goyazensis occurs in Paraguay — the only species of Polybotrya known from that country— and in the southern half of Brazil but not in the coastal mountains (Map 6). This range is unusual because it comprises regions apparently not occupied by other species of Polybotrya. I suspect that P. govazcnsis evolved from south- ernmost, isolated populations of P. caudata. Polybotrya goyazensis has stem scales that are spreading, shiny, membranaceous, translu- cent, darkened in the center with lighter borders, margins strongly denticulate to crose, and the base cordate around the darkened p camp, Brandbyge & Asanza C. }IH70 (AAU) Napo: Reseria Faunistica Cuyabeno. al lado nortc de Laguna Grande. Balslev 4787 (AAU, CR. Q, QCA). 4802 (AAU, Q, QCA). Laegaard 51136 (AAU.Q,QCA),5//50(AAU,Q.QCA),5/224(AAU, Q, QCA); Parque Nacional Yasunf, in the area of the SEP project. Ollgaard el al. 38969 (AAU). 39039 (AAU). 39040 (AAU); 6 km along Ri'o Pano. Hotm- Nielsen & Jeppsen 663 (AAU). Pastaza: Rfo Bolxjnaza. oil exploration camp Chichirota and Destacamento Cabo Pozo, Ollgaard el at. 34894 (AAU. Q. (JCA); Curaray. SE of the airstrip, Holm-Nielsen el al. 22121 (.AAU); oil exploration camp Chichirota. on the Rio Bobonaza. 0llf;aard el at. 35350 (AAU); Ri'o Bobonaza. between Cachitama and the outlet of Ri'o Bufeo. Otlgaard el at. 34703 (AAU). Santiago-Zamora: Taisha. Cazalet & Pennington 7720 (BM. F. K. NY, US). Peru. Amazonas: ridge crest of Quebrada Chuivi (above km 278 of Mararion road), valley of Rio Maraiion near Cascadas dc Mayasi. W'urdack 1933 (USi Junin: E of Quimiri Bridge, near La Merced. Kittip hurumbo. 7-8 km al esle del Puente de Palo Blanco. Mariscal Cacercs. Tivachc Nucvo. J. Schunke 5789 (NY. US); San Roque. Wil- liams 7620 (F). Bolivia. La Paz: Prov, Larecaja. Consata 7 km linvards Mapiri, Beck 4924 (F. LPB); region of Mapiri. San Amonio. fi/«/im';i.«5(US), //2.*(US), II24{K\): region of Mapin. San Carlos, Buchticn 260 (VC). 299 (NY); Mapin. Rushy 442 (NY). Br \/ii .Amazonas: Rio Negro. \S\9.Martiuss.n. (photo ofM specimen aiBM) Para: Martius s.n. (photo ot M specimen al BM) November 1987 Monograph of Polybotra 71 12. PolybotryaglandulosaKuhn(Fig. 31, Map?). Polyhotrya glandiilosa Kuhn, Linnaea 36:65. 1869. Type: Brazil. Amazonas: San Gabriel, no date. Spruce s.n., not 2116 as indicated on Morion negative, (holotype: B!), Polyhotrya suhelliplica Lcllinger, Amcr. Fern J. 62:56. 1972. Type: Peru. Loreto: Mis- huyacu, near Iquitos, 100 ni, Kliif; 1390 (holotype: US!; isotypes: F!, NY). Stem 1-2 cm thick, hemiepiphytic; scales dull brown, thick, opaque, concolorous to bicolorous, curved-appres,sed, 8-15 x 1.5-2.5 mm, the mar- gins entire, the base thickened and expanded, slightly elevated, the dorsal surface sometimes with a medial groove. Sterile leaves up to 75 cm long; petiole very short, up to 2.5 cm long; lamina 2-pinnate-pinnatifid, subelliptic, 65-73 x 22-28 cm, the base cuneate, about 8 cm wide, the apex long-attenuate, the abaxial surface with or without sessile, red, resinous, round glands, the margins ciliate with hairs similar to those of the veins and axes; pimiae ca. 35 pairs, the medial ones 14-16 X 3-3.5(4) cm, sessile or short-stalked, oblong, the base truncate; pinnules catadromic, oblong, falcate at the apex, those of the medial pinnae 15-20 X 5-7 mm, the margin entire, crenate or lobed, generally with 6-10 veins, the basal basi- scopic margin decurrent; axes and veins densely pilose on both surfaces, the hairs tawny, pluricel- lular, acicular, 1.0-2.5 mm long, the scales few or absent. fer///e /eavM narrowly elliptic, 45 x 15 cm, 2-pinnatc-pinnatirid, botryoid; medial pinnae 6-8 X 1.5-2 cm; axes sparsely pilose, with scat- tered, linear, appressed, dark scales; sporangial stalks paraphysate. Pohbotrxa glandulosa, which has been col- lected only three times, is one of the rarest species of Polyhotrya. It is the only species in the genus that is endemic to the Amazon basin (Map 7); however, 1 suspect it will be found in the adjacent Guiana Highlands as that region becomes better explored. I failed to find this species during my fieldwork in the Iquitos area, in part because the location of "Mishuyacu" is unknown. The eleva- tional range for this speties is 100-140 m. This species is unlikely to be confused with any other Polyhotrya since, unlike other species in the genus, the lamina tapers gradually to an extremely short petiole (Fig. 31c). The long, acicular hairs on the axes and veins further distin- guish this species from most other congenerics. Polyhotrya puhens. however, has long acicular hairs and leaf cutting similar to P. glandulosa; since it grows in Amazonian forests, the two species could be confused. Polyhotrya glan- dulosa, however, can be distinguished from P. puhens because the former has a botryoid, instead of an amphiacrostichoid, fertile leaf and a short petiole. The specific epithet refers to the red, .sessile, round, resinous glands on the abaxial suface of the type specimen (Fig. 31d). The two other col- lections, however, lack such glands. This variabil- ity is not unusual, since glands are variably present in all gland-bearing species of Polyhotrya. Specimens examined: Venezuei.a. Territorial Federal Amazonas: Dept. Ri'o Negro, 0-3 km N of Cerro dc Neblina Base Camp, on the Rfo Mawarinuma, 140 m, Ue.mer 16293 (MO, UC). Peru. I.oreto: Mishuyacu, near Iquitos, 100 m, Klug 1390 (F, NY. US). Brazil. Amazonas: Sao Gabriel, no date. Spruce s.n. (B). 13. Polyhotrya lechleriana Mettenius (Fig. 32, Map 7). Polyhotrya lechleriana Mettenius, Filices Uchler. 1:4, tab. 1, figs. 1-5. 1854. Type: Peru. Puno: San Gavan, Lechler 2156 (lec- totype: B!; isotype: L!; fragments F!, US!; photo GH! of K). Acrostichum lechleriamim (Mettenius) Hooker, Species Filicum 5:246. 1864, nom. illeg., non Mettenius 1856. Stem to 1.5 cm thick, hemiepiphytic, mucilaginous?; .scales lanceolate, up to 15x3 mm, cream to dull brown, thin, appressed, the margins denticulate to entire. Sterile /ca/ pubes- cent throughout, the hairs 3-12 celled, 0.3-1.2 mm long, lax, spreading, colorless; /jc/iV)/? 1/1-1/2 as long as the lamina, stramineous; lamina fine- ly divided, (4)3-pinnate-pinnatisect, lanceolate, both surfaces pubescent, especially along the axes and veins, eglandular, the apex acute, not long- attenuate; /^mmyc lanceolate to ovate, (7)10-15 x (2)3-7(12) cm, the base truncate, short-stalked, crowding the rhachis;/?(>if!M/e.v catadromic, oblong with subparallcl sides, the base truncate, nearly sessile, crowding the costa, the apex acute to rounded; ultimate segments single veined, talcatc, 0.5-1 mm wide; (i.vc.v pubescent abaxially, usually with a single scale at the juncture of the costa and 72 Illinois Natural History Survey Vol.34. Art. 1 Fir.URF. 31. Polybotrya i>landiilo.ia Kuhn. a. stem scales, b. fertile leaf; c. stcnle leaf: d. aba.xial surface of sterile pinna showing sessile reddish glands (as dots) and acicular hairs; e, medial pinna. a-c,e: Klug 1390 (US, F). d: Spruce s.n. (B). November 1987 Monograph of Poi.ybotra 73 1 cm Figure 32. Polyhotrya lechteriana Mcltcnius. a. sterile leaf; b. rhachis-costa juncture of fertile leaf; c. costal hairs; d. rhachis-costa juncture of sterile leaf; e. stem scales. a,c; Spruce 4744 (P). b,d: Stiihel 914 (B). e: Dudley 10325 (GH). 74 Illinois Naturai, History Survey Vol.34, Art. 1 costulc, the scale ovate, apprcssed. thin, brown; grooves glabrous or puberulent within, truncated by the ridges of the next lower order, usually becoming shallower near the juncture. Fertile leaf 3-4-pinnate, botryoid. pubescent with hairs like those on the sterile leaf; sori distinct, round. 0.5- 0.8 mm wide, on short stalks, these 1-2 mm long, the receptacle moundlike; sporangial slalks paraphysate; spores {50)54-60(65) microns long. Other illiistralions: Hooker. Second Century of Ferns, lab. 97. 1861; Mettenius's original dc- .scription has an excellent illustration of the type at Berlin. Polybolrya lechleriana occurs in rich mon- tane forests from (100)1000-1500 m. This species is primarily Andean but has a notable range dis- junction in the Guiana Highlands at Mt. Roraima in Guyana. 1900 km from the nearest known popu- lation in Colombia (Map 7). This disjunction is best explained as an example of long distance spore dispersal. Another significant Andean- Guianan range disjunc'ion occurs in P. frac- tisericitis (Map 3). This species is easily recognized by its finely divided sterile lamina with very narrow, single- veined segments and lobes (Fig, 32). The finely cut leaves, so unlike others in the genus, arc prob- ably what prompted Copeland (1947) to remark that this species "looks out of place here |in Polyhotrya]" But P. lechleriana has all the fea- tures of a typical Polybolrya including the unique stem anatomy of the genus. Some forms of this plant resemble Polybolrya stolzei an endemic to the Andes of Colombia, but P. lechleriana differs by the narrower width of its ultimate segments or lobes, each of which is one-veined (see key). Polybotrya lechleriana probably evolved from a less dissected ancestor in the P. alfredii group by cessation of the growth of the marginal meristem to produce a more finely divided sterile lamina. This species is named in honor of WilibaldLechler( 1814-1856). a Gemian pharmacist, botanist, and explorer, who sent many of his collections to Mettenius. Specimens examined: Guyana. Mi Roraini.i. Waruina Trail. Pcrsaiul 114 (NY). CoLOMHiA C'auca: forests of Rio Timbiqui. l.chmann H T4I.< ((iH) Choco: alredcdorcs dc San Jose del Palmar. Ccrro S dc la pohlaciiin. Forcrti pe: Colombia. Antioquia: Guatape, forests on slopes above river, 1/77, McAlpin <$ Kuhn 77-12 (holotype: NY!). The plant from which the type specimen was taken is cultivated in the fern greenhouse. New York Botanical Garden (accession no. 332/78). Planta hemiepiphytica: caulis ca. I cm diam. : squamulis membranaceis. linearibus. acutis. 9- 12 mm longis. 0.3-0.7 mm latis. in margine den- ticulatis vel integris: laminae steriles lanceolatae vel anguste dellalae, 3-pinnatae. apice longiat- tenuatae, usque ad 55 cm longae. 26 cm latae. glabrae vel punctatae . glandulis resinaceis. rotun- dis: pinnulae catadromicae . 2.0-2.5 cm longae. 0.8-1 .2 cm latae. apice rotundatac: segmenta ter- tiaria obovata. ad basim cuneata: costae et rhachides sulcalae puhescenies in sulco. squamis angustis, fuscatis. denticulatis: folia fertilia bot- ryoidca: sori rotundi. ca. I mm tongi. pedicellati . Stem about I cm wide, hemiepiphytic; scales light to dark brown, mosth concolorous or rarely with a narrow hyaline border, thin, linear, 9-12 >^ 0.3-0.7 mm, the margins subcntire to denticulate. Sterile leaf up to 8 m long; petiole V\-Va as long as the lamina; lamina to 55 x 26 cm, 3-pinnatc, narrowly lanceolate or triangular, the apex long-attenuate, the tissue with or without November 1987 Monograph of Polybotra 75 FiciiiRF-; 33. Pnlyhoirya altcniiala Moran. a. sterile leaf; b. pinnules and eosla-rhachis juncture (note the abundance of narrow, dark scales on the axes); c. stem scales; d. two pinnules of fertile leaf; e. fertile leaf, a-c: McAlpin & Kuhn 77-12 (NY). 76 Illinois Natural History Survey Vol. 34, Art. 1 punctate, resinous, spherical, sessile glands; pin- nae 10-13 X 3.5-4.0 cm long, lanceolate to ob- long, the base truncate, short-stalked, crowding the rhachh; pinnules catadromic, 2.0-2.5 x 0.8- 1.2 cm, with only 5-7 lobes or segments, the apex acute or rounded, the base truncate, sessile or with a short stalk less than 1 mm long; tertiary segments obovate, entire to slightly lobed, con- taining several vein branches, the base cuneate; axes pubescent within grooves and especially ab- axially, the hairs 4-12 celled, 0.2-0.3 mm long jointed, with reddish cross-walls; costal scales numerous, narrow, dark, denticulate, tortuous, ap- pressed; grooves pubescent within, the hairs red- dish, multicellular. Fertile leaves botryoid, 3-pinnate, densely scaly, the scales grading into jointed hairs (the hairs are actually reduced scales); sori stipitate, the stalks 1-3 mm long; sporangial stalks paraphysate. Polybotrya altenuata, named for its distinc- tive, long-attenuate leaf apices, is endemic to the Andes of Colombia (Map 7). This species differs from P. lechleriana. a similar species, by its long- attenuate apex, oval tertiary segments, resinous glands on the lamina, and the darker, narrower scales on the axes. The pinnules, furthermore, are shorter and stubbier, having only 5-7 lobes and/or segments, and these have a cuneate base (Fig. 33a,b). The presence of glands on the lamina is not constant. The living plant from which the holotype was made has always produced leaves having round, sessile, resinous glands on the undersur- face. The Lehmann specimen, however, lacks such glands. Similar glands are also variably pres- ent in other species, such as P. osmundacea and P. glandulosa. Specimens examined: Colombia. Antioquia: Guatape, McAlpin & Kuhn 77-12 (NY). Nariho: along river Pipulquer. west slopes of the western Andes, Lehmann 500B (US). 15. Polybotrya stolzei Moran, sp. nov. (Fig. 34, Map 9). Polybotrya stolzei Moran, sp. nov. Type: Colombia. El Valle: Santa Rosa, Dagua Val- ley, forest along Rio Cabellcte, 200-300 m, 22 September 1922, Killip 11549 (holotype: US!; isotypes: GH!, NY!, PH!). Planta hemiepiphytica: caulis 0.6-1.5 cm diam., Imucilaginus?): squamae fuscatae, opacae, lineares, integrae, usque ad 12 mm lon- gae. 0.4 mm latae: lamina sterilis 3-pinnata. plerumque pubescens in superftciebus ambabus: rhachis et costae pubescentes. squamis ovatis tenuibus aliquot; pinnulae catadromicae , pro parte maxima sessiles, usque ad 33 "X- 9 cm; seg- menta tertiaria ovalia vel oblonga Integra, (5)6- 8(10) X (2)3-4(5) mm. Folia fenilia ignota. Stem 0.6-1.5 cm thick, hemiepiphytic, mucilaginous (?); scales dull brown, appressed, opaque, linear, 0.4-12.1 x 0.3-1. mm, margins subentire. Sterile leaves up to I m long; petiole 'A to 1/2 as long as the lamina; lamina lanceolate orsubdeltate, 30-65(80) x 19-32(60) cm, 3-pin- nate, usually pubescent on both surfaces, always so along the veins beneath, the hairs whitish to tawny, articulated, 0.1-0.4 mm long, spreading; pinnae to 33 x 9 cm, narrowly triangular to lan- ceolate, equilateral or, in the basal pair, with the basiscopic side slightly prolonged, the base trun- cate, short-stalked, crowding the rhachis; p/nnu/ej catadromic, oblong, the base truncate, nearly ses- sile, the apex rounded to acute; tertiary segments ovate to oblong, (5)6-8(10) x (2)4(5) mm. with several vein branches, the margins entire to slightly lobed; axes pubescent abaxially, with a few thin, ovate, brown scales, especially at the pinnule junctures; grooves nearly glabrous to mod- erately puberulent within, never packed with long- protruding hairs, truncated by the ridges of the next lower order. Fertile /fa/ unknown. This new species is named for Robert G. Stol/e, pteridologist at the Field Museum of Nat- ural History, Chicago. Illinois. He onginally suggested this genus to mc and has given much help and encouragement dunng the preparation of this monograph. Polybotrya stolzei is endemic to the western Andes of Colombia (Map 9). occurring in wet. shaded forests from 2(X)-1750 m. The laminar cutting of P. stolzei is intermediate between P. lechleriana and P. alfredii. Extremeh large pin- nules of P. stolzei look like those of P. alfredii but are distinguished by the pubescence on both surfaces of the lamina and by the blunter apices of the medial pinnules. Smaller, more delicate forms of P. stolzei may look like P. lechleriana but are distinguished by their oblong tertian, seg- ments with more than one vein f>er segment. Un- November 1987 Monograph of Poi ybotra 77 FiciURF. 34. Polybotrya stolzei Moran. a. costal scales from the pinnule-costa junctures; b. abaxial view of medial pinnules of "c"; c. medial pinna; d. stem scales; e. basal pinna of large leaf; f. adaxial view of rhachis-costa juncture, a-d: Killip 1 1549 (GH). e,f: Leilinger & de la Sola 2H4 (COL). 78 III iNois Natural Hisidry Survky Vol.34. Art. 1 fortunately, none of the collections contain fertile leaves and this structure, therefore, cannot be com- pared to the distinctive botryoid fertile leaf of P. lechlerianii. Specimens examined: Colombia. Choco: NW side i)f Alto del Buey, LeUinger & de la Sola 2fiO (COL, US), 2H4 (COL, US); principal ridge and slopes 2 km E of San Jose del Palmar, 15.5()-I6()() m, LelUnger & de la Sola 747 (COL, LP, US), 74« (COL, LP, US). El Valle: Santa Rosa, Dagua Valley, forest along Rio Caballele, 200-3(X)m, /fi7/(p //549 (GH, NY, PH, US). 16. Polybotrya alfredii Brade (Fig. 35, Map 8). Polybotryci cdfredii Bradc, Bradca 1:12, tab. 1, fig. 2. 1969. Type; Costa Rica. San Jose; Tablazo, 1900 m, July 1908, (Brade's origi- nal description states August, apparently an error.) Brade & Brade 98 (holotypc; HB; isotypc; NY!). Polybotrya alfredii Brade f. carpinterae Brade, Bradea 1:13. 1969. Type: Costa Rica. San Jose: Carpintera, terrestrial, 1500 m, 25 November 1908, Brade & Brade s.n. (holotype: HB). Polybotrya gracilis Brade, Bradca, 1:14, tab. 1, fig. 3. 1969. Type: Costa Rica. Cartago: Tablazo, 1900 m, 28 August 1908, Brade 554 (holotypc: HB). Stem 1-2 cm wide, hemiepiphytic, the apex in living plants usually covered by a thick mucilage that disappears upon drying; scales dull, dark to light brown, opaque to translucent, 8-16(22) x 1- 1.5(2.0) mm wide, narrowly lanceolate, spread- ing, attached across the width of the slightly thick- ened base, the margin denticulate or more com- monly entire. Sterile leaves up to 1 .4 m x 0.7 ni; petiole up to 35 cm long; lamina ovate, up to 3-pinnate-pinnatifid, subchartaceous, often drying greyish green, the proximal margins of the pinnule bases often ciliatc, the hairs less than 0.1 mm long; pinnae up to 45 X 22 cm, apex acuminate: pinnules up to 12 X 5 cm, ovate to narrowly trian- gular, the base truncate and symmetrical, those of the basal pinnae anadron)ic or subcqual, those of the medial pinnae catadromous; ullimute seg- ments ovate to oblong, entire, crenulate, serrate or lobed, the apex obtuse or acute; axes evenly pilosulous abaxially, the hairs 0.1-0.2 mm long. usually tawny; grooves pubescent within, very shallow or even disappearing within about 1 cm of the rhachis so that the costa is almost terete where it joins the rhachis. Fertile leaves up to 0,8 X 0.35 m, up to 3-pinnate-pinnatirid, ovate, coenosoric; axes pubescent, the hairs short, 0.1- 0.2 mm long, usually with ovate to lanceolate, appressed, thin, denticulate to fimbiiate scales; sporungial stalks paraphysate; spores (50)55- 62(66) microns long. Other illustrations: See the original descrip- tions cited above. The altitudinal range of Polybotrya alfredii is 7(X)-190O m. In Ecuador, this species grows in rich montane forests in the eastern Andes. In Costa Rica, this species occurs primarily in wet. shaded cloud forests, usually growing alongside P. gomezii. Polybotrya alfredii produces numer- ous terrestrial leaves that often predominate the forest floor. Twice in Costa Rica and once in Ecuador, 1 saw fertile leaves produced from the terrestrial portion of the stem. These are the only instances in which I saw fertile leaves produced tertestrially by a scandent species. Polybotrya alfredii looks like P. lourteigiana but can be distinguished by its more highly cut lamina. Since leaves lower on the stem tend to be smaller and less divided than those higher on the stem, cutting should be compared using pinnae over 25 cm long (sec couplet 24 in the key). The two species also differ in the shape of the ultimate segments or lobes; P. alfredii has relatively more ovate and shorter segments than P. lourteigiana (compare Figs. 35 & 37). Unlike all other species of Polybotrya. P. alfredii has the stem apex and the lowermost 5 to 1 5 cm of the petiole covered w ith a thick, translu- cent mucilage. Mucilage is totally lost upon drying and therefore not visible on herbarium specimens. The function of this mucilage is unknown. More fieldwork is needed to determine if this unusual mucilage occurs in two closely related species. P. lechleriana and P. lourteigiana. Nectaries are another unusual feature of the morphology of P. alfredii During tleldwork in Costa Rica, I saw nectaries on the rhachis of P. alfredii like those described by Koptur et al. (1982) for P. osmun- dacea. Sec the Morphology and .Anatomy section of this monograph for further information concern- ing nectaries. November 1987 MONCWRAPH of POI.YBOTRA 79 15 cm FiciURF. 35. Polyhoina alfredii Bradc. a. sterile leaf; b. slcm scales, the middle shown sideways; c-g. pinnules, acroscopic side to the right; h. fertile pinnule, adaxial view; i. adaxial surface of costa and pinnule base, a.b.f.h.i: Moran 2442 (CR). c; Moran 3532 (F). d: Buchtien 5164 (US), e: Moran 3214 (CR). g: Dwyer 8337 (MO). 80 iLLiNors Natural History Survry Vol. 34. An. 1 I put P. gracilis in synonymy with P. alfredii, although the former appears distinct because the leaf is smaller, only 2-pinnate, and has obtuse pinnule apices (Fig. 35e). Polyhotrya gracilis re- sembles P. alfredii, however, in such characteris- tics as scales, pubescence, stem mucilage, habitat, and range. Furthermore, at several localities in Costa Rica, I collected plants that exhibited com- plete intergradation between the typical large leaves of P. alfredii and the smaller ones of P. gracilis. Polybotrya gracilis, therefore, represents a small-leaf form of P. alfredii that has well- rounded pinnule apices. I chose the name P. al- fredii over P. gracilis because an isotype has been seen. This species is named in honor of Alexander Brade's brother and field companion, Alfred Brade. Specimens examined: Nicaragua. Rivas: Isla de Ometepe, NW slope of Volcin Maderas, Stevens i\). no iiK-ality. Dii.fcn UbSfi (GH); Porto Dom Pedro II, Dusen 44:.hBS\): November 1987 Monograph of Polybotra 87 FiouRF. 39. Polyhotrya cylindrica Kaulfuss. a. sterile pinna; b-d. sterile pinnules showing variation in amount of dissection; e. fertile pinnules; f. stem scales, top view; g. stem scale, side view. a,e: Goedas 217 (NY), b: Mosen 3050 (B). c.f,g: Schmah I5S (UC). d: Du.sen 6926 (F). 88 Illinois Natural History Survey Vol. 34, Art. I Vdlla Grande, Dusen 6984 (BM); Porto de Cima, Du.sen 6926 (F, GH, LIL, MO, NY, PH, UC, Z); Tacarehy. Dusen 15351 (BM, F, PH), 15353 (GH, MO, PH, UC. Z) Rio de Janeiro: without locality, Hunnewell 1851 1 (MICH); Distiito Federal, Guanabara, Represa Camori, Brade 12576 (RB); Rezende, 80 mi WNW of Rio, on Paraiba River, Rio Palmital, Castellanns 25710 (F); Serra de Itatiaia, Brade J 2614 (NY, RB); Brene. Glaziou 955(RB); vicinity of Meio da Serra, Smith & Brade 2286 (GH); Estrada Velha de Barrira a Teresopolis, Duarie & Pereira 65323 (F, LP); Corcovado, Miers s.n. (NY); Teresopolis, Brade 9365 (NY, UC); Serra da Carioca, Brade 13982 (RB). Rio Grande do Sui: S. Leopoldo. Lehnun 2562 (LIL); Porto Alegre, Fargens 251 (UC). Santa Catarina: Municip Ibirama, Horto Florestal. Smith & Klein 7546 (US); Horto Florestal, I.N. P., Ibirama, Rett: & Klein 1 101 (US), 1102 (US). 2665 (US), 3466 (US); Blumenau, Wamow, Gnedas 217 (NY, UC, US); Itajai. Reit: 151 (LIL); Municipality Brusque, Azambruja, Smith and Reitz 6135 (COL, GH, MO, US); ReseiA'a Florestal, dos Piloes, Duarie 3179 (LP, NY); Mina Velha, Garuva, Sao Francisco do Sul, Reil: & Klein 4635 (US), 4682 (US): Sabia, Vidal Ramos, Reit: & Klein 5108 (US); Passo Mansa, Haer- chen 124 (NY); Joinville, Schmal: 78 (MO), 158 (NY, UC); Tres Barras, Garuva, Sao Francisco do Sul, Reit: (4 Klein 5549 (US); Sta. Catherine 1834, Caudichaud 73 (P); Azambuja-BrtJsgui, Reit: 2825 (LIL, P); Blumenau, Mu//fr60.?9 (MICH). Sao Paulo: Sao Bcnlo. Luetzelhurg 254 (MICH, US); Sen-a da Bocaina. Brade 20931 (LP); Morro das Pedras, Iguape, Brade s.n. (US); Paranapicaba, Biological Station, Handro 1229 (GH, US); Ague funda, Handro 2228 (GH, US); Serra do Mar, Wackel 21556 (GH); Campo Grande to Alto da Serra, 40 km SE of Sao Paulo, Tryon & Tryon 6591 (GH); Santos, 1874, Mosen 3050 (B, P). State un- known: Gla:iou 2428 (photo of K specimen at GH); no locality, Claussen s.n. (P); Micken s.n. (NY); Capao do Fradie, Schneus 3412 (GH); Morro do Aniao. Sehnem 3092 (GH). 21. Polybotrya hickeyi Moran, sp. nov. (Fig. 40, Map 13). Polybotrya hickeyi Moran, sp. nov. Type; Bolivia. Cochabamba: Prov. Chapare, road from Cochabamba to Villa Tunari, in thick rain forest, very heavily shaded, on trunk of Ficus 1700 m, 23 March 1980. Hickey 801 (with Eshbaugh) (holotype: GH!; isotype: MU). Caulis hemiepiphyticus. 1-2 cm diam.: squa- mae alrohrunneae, vulgo opacae. leviter nilidae, concolores vel marginihus tenuiorihus el pal- lidioribus ad apicem: lamina usque ad 3-pinnata- pinnatifida. deltaia, usque ad 1 m longa. leviter pubens inferne. pilis cylindricis. appressis, usque ad 0. 1 mm longis: pinnae usque ad 46 ^ 25 cm. deltalae. paribus proxirrmlibus acroscopice pro- longatis; pinnulae catadromicae supra basin, del- tatae. ohlongae vel lanceolatae: costae sine sulcis prominentibus: venae prominulae superne. apice leviter incrassatae. Sori discreti. rotundi. oppositi vel suboppositi. Stem 1-2 cm thick, hemiepiphytic; scales dark chocolate brown, mostly opaque, concolor- ous, or the apex with thinner, lighter borders, the margins denticulate to entire, the base thickened, attached across its full width. Sterile leaves up to 1,5 m (?); petiole with thin, ovate to lanceolate scales; lamirui to 3-pinnate-pinnatifid, deltate. to about I m long, the abaxial surface slightly pu- berulent, the hairs cylindrical, appressed, up to 0.1 mm long; pinnae up to 46 x 25 cm, deltate. the proximal pair basiscopically prolonged, the distal ones gradually becoming acroscopically pro- longed; pinnules catadromic above the base, del- tate to oblong-lanceolate, the basiscopic margins thickened and decurrent on the costa, the base slightly prolonged acroscopically, the apices acuminate; tertiary segments also with a thick- ened, decurrent, light-colored basiscopic margin; veins prominulous adaxially. the tips thickened and ending just behind the margin; costules bor- dered by a raised, erect flap or wing of tissue; axes often not prominently grooved, pubescent by ca. 0. 1 mm long, colorless, erect hairs, the scales appressed, denticulate, flexuose. Fertile leaves botryoid; sori with few fusions, arranged mostly oppositely or suboppositely. Polybotrya hickeyi grows in high elevation forests in the Andes of Colombia and Bolivia; it has not yet been collected from Ecuador or Peru (Map 13). This species looks most like P. puberulenta— see that species for comparison. This plant is named for one of the collectors. Dr. R. James Hickey. who has made impc>nant con- tnbutions to the study oflsoetes and Lycopodium. The Colombian specimens look like the Bo- livian type specimen except for the stem scales. The Bolivian type has dark, chocolate brown scales with denticulate margins and lighter borders apically (Fig. 40g). The Colombian specimens have merely dull brown, concolorous scales with November 1987 Monograph of Poi.ybotra 89 Figure 40. Polyhotrya hickeyi Moran. a. fertile pinnules; b,c,d. sterile pinnules, acroscopic side to the right; c. medial pinna; f. basal pinna (note that the two proximal basal pinnules arc fertile); g. stem scales (Bolivia); h. stem scale (Colombia); i. abaxial surface of distal pinnules, a.d: Ludwif; 262 (NY). b.h; Grant 10283(F). c: Killip & Smith 202/0 (US), c.f.g.i: holotypc. HickeyHOI .wnh Eshbaugh (GH). 90 Illinois Natural History Survey Vol.34, Art. I entire margins and a prominently curved and thick- ened base; therefore, they appear appressed (Fig. 40h). This thickened and curved scale base appears somewhat elevated, like those of P. caudata and P. cylindrica . The lamina of P. hickeyi turns a peculiar yellowish green upon drying. Specimens examined: Colombia Cauca: W slope of W Cordillera, above Carpinteria. Alston S278 (BM) Cundinamarca: Cordillera Oriental, Toquiza. Gazaunta Valley, Cordillera de Helicona, 15 km NW of Medina, 2200 m. Grant 10283 (F, US); Fusagasuga, Ludwig 262 (BM, MO, NY) Norte de Santander: Pica- Pica Valley, above Tapata (N of Toledo) 2100-2400 m, KUlip <* Smith 20210 (GH, NY, US). Bolivia. Cochabamba: Prov. Chapare, road from Cochabamba to Villa Tunari, 1700 m, Hickey 801 with Eshbaugh (GH). 22. Polybotrya puberulenta Moran, sp. now (Fig. 41, Map 13). Polybotrya puberulenta Moran, sp. nov. Type: Ecuador. Napo; camino Baeza-Tena, 5 km al sur de Cosanga. Cordillera Oriental, bosque perturbado, 1 100 m, 4 de enero 1984, Moran 3528 (holotype; F!; isotypes: COL!, GH!, LPB!, MO!, NY!, Q!, QCA!, UC!, US!, VEN!). Caulis hemiepiphyticus: squamae aurantiacae sordidae vel luteae, memhranaeae. late patenies. centroleviterfuscatae. marginibus denticulatis vel erosis: lamina 2-pinnata-pinnatisecta perlate ovata, crassa.flavovirenles paltide, pubents ulrin- que, pilispatentihus.flexuosis, ca. O.l mm longis: pinnae usque ad 34 x 18 cm, infimus hasiscopice prolongatae: pinnulae catadromicae oblongae. in margine incrassatae et decurrentes: lohi tantum crenali. apice truncati vel rotundati: venae pro- minulae superne. apice leviter incrassatae. Folia fertilia 4-pinnata. tuibus dense pubescentibus. pilis 3-5 mm longis et squamis linearibus. appres- sis vel leviter patentibus: segmenta opposita vel subopposita. receptaculo pubescenii , pilis 1-2 mm longis. articulatis. Stems 1-2 cm thick, hemiepiphytic; .scales yellow to sordid orange, darker in the center with light yellow borders, membranous, spreading, 1- .3 mm wide, the margins crosc to denticulate, at- tached across the width of the narrowed base, in fresh material somewhat rugose transversely. Sterile leaves up to 1 .2 m long; petiole up to 27 cm long; lamina to 2-pinnate-pinnatisect, broadly ovate, thick and somewhat fleshy in living mater- ial, color light yellowish green, puberulent on both surfaces, the hairs spreading, flexuous, ca. 0.1 mm long, intergrading with the scales on the lower suri'ace; pinnae up to 34 x 18 cm, the proximal pairs basiscopically prolonged, the distal ones grad- ually becoming equilateral or slightly acroscopi- cally prolonged; pinnules catadromic throughout, oblong, the basiscopic margin thickened, decur- rent on the costa. the apices obtuse to acute, base cuneatc and slightly prolonged acroscopically; coslules bordered by a raised flap of lighter colored tissue; lobes of the largest pinnules merely crenate. the margins thickened, the apices obtuse or trun- cate; veins prominulous adaxially, the tips slightly thickened and ending just behind the margin; axes evenly pubescent abaxially , the hairs 0. 1 mm long, tawny, erect, scales caducous, usually lacking; grooves pubescent within, becoming shallow and flat at the junctures, not admitted to the groove of the next lower order. Fertile leaves 4-pinnate. the axes densely pubescent, the hairs jointed. 3-5 mm long, the scales 5-10 mm long, linear, ap- pressed to slightly spreading; sort opposite or sub- opposite; sporangial stalks paraphysate; recepta- cle pubescent by jointed hairs, 1-2 mm long, that protrude beyond the sporangia; spore size un- known. Polybotrya puberulenta grows at high eleva- tions in the Andes of Ecuador, Bolivia, and prob- ably Peru (Map 13). The specific epithet refers to the sterile lamina, which is puberulent on both surfaces. Another distinctive feature of this species is the stem scales, which are relatively wide (1-3 mm), thin, spreading and yellowish when viewed with transmitted light. The lamina is relatively thick and fleshy; the mesophyll cells apparently collapse when dry and the veins become promin- ulous. The lamina, which is always broadest at the base, has a peculiar light greyish green color when dry. This species differs from P. hickeyi in several characteristics of the stem scales, pinnule arrange- ment, lamina dissection, pubescence, paraphysis length, and scalyness of the major axes. Both species, however, have thickened, decurrent. basiscopic margins. The medial, less cut pinnules of P. hickcvi may approach in outline those of P. puberulenta. but the largest pinnules of the basal November 1987 Monograph of Polybotra 91 Figure 41. Polybolrya puberulenta Moran. a. pinnules; b. stem scales; c. sterile leaf; d. abaxial surface of costa and pinnules, a-d: Moran 3528 (F). 92 Ii.r.iNois Natural History Survey Vol. 34. Art. 1 pinnae are much more highly divided in P. hickeyi (compare Figs. 40f and 41a). Unfortunately, Rusby's Bolivian collection consists only of the fertile leaf, but I believe it belongs to this species because of its cutting, in- dument type, and yellowish stem scales. The long, protruding paraphyses are easily seen with a hand lens and offer an excellent character to distinguish this species from P. hickeyi. Specimens examined: Ecuador. Napo: camino Baeza-Tena, 5 km al sur de Cosanga, Cordillera Orien- tal, bosque perlurbado, 1 100 m, Moran 3528 (COL, F, GH, LPB, MO, NY, Q, QCA, UC, US, VEN). Bolivia. La Paz: Yungas, 6000 ft, "only the fertile frond collected," Rusby 443 (NY). 23. Polybotrya alata Moran, sp. nov. (Fig. 42, Map 14), Polybotrya alata Moran, sp. nov.. Type: Panama. Panama; Campo Tres, 3 mi NE of Altos de Pacora, primary forest, alt 500-800 m, epiphytic, the caudex closely appressed, sterile, 10 March 1973, Croat 22706 (holotype: MO!; isotype: L!). Caulis hemiepiphyticus usque ad 2.5 cm diam.i squamae lineares. 15-25 x 0.8-1 .2 mm, chrysohrunneae, concolores vel centris leviterfus- catis. appressis, ascendentibus, marginihus den- ticulatis. Lamina 2-pinnata-pinnatifida, superne glabra, inferne pubens, pilis circa 0. 1 mm longis, niveis, suhulatis: pinnae usque ad 30 x 8 cm, an- guste triangulares, apicibus attenuatis, zonis mediis alatis secus costam: pinnulae catad- romicae, usque ad 6x2 cm, lobis catadromicis, crenatis: rhachis el costae pubescentes et squamatae. Folia fertilia ignota. Stem 2.5 cm thick, hemiepiphytic; scales linear, 15-25x0.8-1.2 mm, membranaceous, golden brown, concolorous or with a slightly dark- ened central stripe, appressed-ascending, margins denticulate. Sterile leaves up to 1.3 m long; petiole to 45 cm long, scaly, pubescent with 0.2 mm long, unicellular, colorless hairs; lamina 2-pinnate- pinnatifid, the adaxial surface glabrous, the ab- axial surface lightly to moderately pubescent, the hairs unicellular, less than 0. 1 mm long, whitish, subulate; pinnae up to 30 x 8 cm, narrowly trian- gular with a long tapering apex, 3-4 times as long as broad, the lowest becoming pinnatifid in the apical one-third, the transition zone between pin- nate and pinnatifid marked by a narrow, straight, decurrent, green wing parallel to the costa: pin- nules catadromically arranged in the medial pin- nae, lobed almost to the costule, inequilateral at base, up to 6 x 2 cm, the proximal acroscopic ones longer than the proximal basiscopic; pinnule lobes catadromically arranged, the margins cre- nate, the apices rounded; axes pubescent ab- axially, the hairs 0. 1-0.2 mm long, subulate, col- oriess, often difficult to see on the lamina, inter- mixed on the costae with long, scraggly. subclath- rate, denticulate scales; groove usually glabrous, abruptly truncated by the raised ridge of the costal groove, the junctures short-pubescent. Fertile leaves unknown. Polybotrya alata has been collected only twice, both times in the mountains of Panama (Map 14), but 1 suspect it also occurs in the moun- tains of southern Costa Rica. Distinctive features of this species are the narrowly triangular pinnae with long attenuate apices and the pinnules that are lobed almost to the costule. thereby making the lamina 2-pinnate-pinnatifid (Fig. 42a). The transition zone from the free pinnules to the pin- natifid apex differs from all other species of Polybotrya by the thin, basally decurrent. green wing, which imparts an alate appearance to the costa (Fig. 42c). The stem scales are also distinc- tive by their long length (15-25 x 0.8-1.2 mm) and shiny, golden brown color. The short, even pubescence on the abaxial surface of the leaf separates P. alata from all other Central American Polybotrya. except P. caudata. which can also have pubescent leaves. The laminar hairs of P. caudata. however, are longer ( l-3mm) than those of P. alata and are 4-10 celled. The two species also differ sharply by their stem scales. The only other Central American species with which P. alata could be confused is P. osmun- dacea. but that species is easily distinguished by its anadromic pinnule arrangement and nonalate costae. Furthermore, the costular groove of P. alata is truncated by the ridges of the costa; in contrast. P. osmundacca has uninterrupted, decur- rent grooves (Fig. 47h). Specimens examined: Panama Panami: Cam- po Tres. .1 mi NE of Alios dc Pacora. Croat 2270ft (MO. I.) Veraguas: ^ mi W of Santa Fc on ro-id past Escucia Agncola Alto Picdra on Pacific side of divide, 800-1200 m. Croat 23011 (MO). November 1987 Monograph of Polybotra 93 E u ^ 1 Ficn^RR 42. Polyho.ry^a alota Moran. a. sterile pinna; b. margin of stem scale; c abaxial surface ofdistal p,nnules (note pubescence and decurrent p.nnule bases); d. stem scale, a-d: CroaTlTm (MO) 94 Illinois Natural Hisiory Slrvky Vol.34. An. 1 24. Polybotrya aequatoriana Moran, sp. nov. (Fig. 43, Map 15). Polybotrya aequatoriana Moran, sp. nov. Typr: Ecuador. Napo: Cordillera Oriental, camino Baeza-Tena, 34 km a! sur de Baeza, 2100 m, 30 de diciembre 1983, Moran 3512 (holotypc: F!; isotypcs: AAU!, COL!, GH!, LPB!, MO!, NY!, Q!, QCA!. UCi). Caulis hemiepiphyticus. 1 .5-3.4(4.0) cm diam.; squamis sordide aurantiacn-luteis. nitidis, memhranaceis, imple.xis. lanatis. in marline mi- nute denticulatis, 1-2(3) X 12-20(25) mm: lamina lanceolata, usque ad 1.4x0.7 m, 2-pinnata- pinnatifida, modice vel sparse puberula subter, pilis appressis, 0. 1-0.3 m longis; pinnae longidel- tatae. 25-35 x 13-18(20) cm: pinnulae catad- romicae, acroscopice prolongatae, basiscopice valde obliquae: lobi integri vel vade crenati: rhachis et costae pubescentes, pilis rigidis albidis. sutcis profundis pubescentibus. Pinnulae fertile s lineares, vulgo 4-6 x 0.5-1 .0 cm, segmentis op- positis vel suboppositis circularihus vel oblongis, 3-5 X 1-2 mm. Stem 1.5-3.5(4) cm thick, hemiepiphytic; scales dingy orange yellow, mostly darkening upon drying, shiny, membranous, tangled and woollike, the margins minutely denticulate, 1- 2(3) X 12-20(25) mm. Sterile leaves up to 1.6 m long; lamina lanceolate, to 1 .4 x 0.7 m, 2-pinnate- pinnatifid, moderately to lightly pubescent below, the hairs fine, appressed, 0. 1-0.3 mm long, grad- ing into scales along the axes, often with a few scattered resinous, circular glands (use at least 30 x ), the margins somewhat thickened and lighter colored, often becoming revolute upon dry- ing; pinnae long-triangular, 25-35 x 13-18(20) cm; pinnules arranged catadromically above the base, prolonged acroscopically, reduced and very oblique basiscopically; lobes entire or rarely shal- lowly crenate; axes pubescent below with un- evenly scattered, whitish, erect hairs 0. 1-0.2 mm long, scaly by linear, flexuous, darkened, often appressed scales, these most abundant at the junctures; grooves deep and di.slinct, filled with hairs, these sometimes clavatc, ca. 0.1 mm long, reddish, the ridges well detlned. stramineous, con- tinuous with those of the costules but truncated by the ridges of the rhachis. Fertile leaves bot- ryoid; pinnules linear, mostly 4- 6 x 0.5-1 .0 cm; sort arranged oppositely or suboppositely, round to oblong; sporangial stalks paraphysate; spore size unknown. Polybotrya aequatoriana occurs in the cloud forests of Ecuador and Bolivia (Map 15), from 2100-2410 m, the highest known altitudinal range of any species of Polybotrya. Polybotrya aequa- toriana is named for Ecuador, where I collected the type. Polybotrya aequatoriana can be easily sepa- rated from P. hickeyi and P. appressa by its stem scales, which are thin, yellowish orange, and membranaceous (Fig. 43b); see P. appressa for further comparison with that species. Polybotrya aequatoriana has highly reduced basiscopic lobes on the sterile pinnules (Fig. 43c), and the fertile pinnules are much narrower than those of other congenerics. Specimens examined: EcuAtxsR Napo: Cor- dillera Oriental, camino Baeza-Tena. 34 km a! sur de Baeza. 2100 m, Moran 3512 (AAU, COL, F, GH, LPB, MO, NY. Q, QCA. UC), 3585 (F. Q. QCA); road Baeza-Lago Agrio. ca 114 km from Lago Agrio. 1750 m. OUgaard et at. 35776 (AAU. QCA). Bolivia La Paz: Prov Sud Yungas. Huancane, 7 5 km hacia el sud sobre el camino nuevo. 2410 m. Beck 3108 (LPB); same locality, but at 6 5 km. 2280 m. Beck 3060 (LPB, NY). 25. Polybotrya appressa Moran. sp. nov. (Fig. 44, Map 17). Polybotrya appressa Moran, sp. nov. Type: Ecuador. Napo: Cordillera Oriental, camino Baeza-Tena. 34 km al sur de Baeza. bosque virgen, 2100 m, 24 de enero 1984. Moran 3586 (holotypc: F!; isotypcs: GH!, MO!, NY!. Q!. QCA!. US!). Caulis hemiepiphyticus. 1-2 cm diam.: squamae concotores, obscure brunneae. integrae. appressae, vulgo 1-2 x 0. 1-0.2 cm; lamina ovata vel late deltata. 2-pinnaia-pinnatifida. superne glabra, inferne pilosa, pilis dispersis. ca. 0.1 mm longis: pinnae usque ad 37 x 18 cm: pinnulae pro- longatae acroscopice. latis basiscopicis reductae: lobis rotundatis vel plusminu.'ive truncatis: rhachis et costae acqualiier pubescentes, pilis usque ad 0. 1 mm longis. sulcis glabris. Foliafertilia ignota. Stem 1-2 cm thick, hemiepiphytic; scales concolorous, dull brown, thick, opaque, entire. * November 1987 Monograph of Polybotra 95 FicjURF. 43. Polvboirya cwquatoriana Moran. a. sterile leaf; b. stem scales; c. costa and pinnule bases, abaxial view (note only scattered hairs); d. rhachis, costa, pinnule bases, adaxial view, groove minutely pubescent, thin, a-d; Moran 3512 (F). 96 Ii.i.iNois Natural History Survey Vol.34. An. I »^»lj«'v Figure 44. Polyhotrya appressa Moran. a. sterile leaf; b. stem scales: c. eosta and pinnule bases, abaxial view; d. costa and pinnule bases, adaxial view, a-d: Moran 3586 (F). I November 1987 Monograph of Polybotra 97 appressed, mostly 1-2x0,1-0.2 cm. Sterile leaves up to 1.5 m long; lamina ovate to broadly deltate, dark green above, slightly paler below, 2-pinnate-pinnatirid, up to 1.3x0.74 m, gla- brous adaxially , abaxially with a few scattered, ap- pressed hairs, these less than 0. 1 mm long, inter- grading with scales on the costae; pinnae to 37 x 18 cm, widest at the base, the apex long-tapering, the acroscopic basal pinule usually prolonged beyond the rest, pinnules arranged catadromically, prolonged acroscopically and reduced basiscopi- cally at the base, the proximal ones with acuminate apices, the distal ones with falcate and truncate apices: lobes rounded to somewhat squarish, entire or rarely minutely serrate at the apex; axes evenly pubescent abaxially, the hairs short, less than 0.1 mm long, the costal scales present or caducous and absent, linear to lanceolate, appressed, light brown, thin; grooves sparsely pubescent within, not very deep or prominent, bordered by broad, shallow ridges, not admitted to the groove of the next lower order. Fertile leaf unknown. Polyhotrya appressa is known only from the eastern Andes of Ecuador (Map 17), where it grows in cloud forests at high elevations. This new species is named for its distinctive, appressed stem scales (Fig. 44b), which help separate it from similar species, many of which have spreading, membranous stem scales. Polyhotrya caudata has similiar dull, thick, opaque stem scales, but that species does not grow at high elevations. Polyhotrya appressa differs from P. aequa- toriana. a closely allied Ecuadorian species, by its stem scales, groove architecture and pubes- cence, and lamina shape. Polybotyra hickeyi is also very closely related, especially by its short, even, abaxial, costal pubescence, and the indis- tinct adaxial grooves that are glabrous within; but it differs pnmarily by its more finely cut lamina. In addition, the lamina of P. appressa is dark green, in contrast to the pale yellowish green laminae of P. hickeyi and P. aequatoriana. Specimens examined: Ecuador. Napo: Cor- dillera Orienlal, camino Bacza-Tena. .14 km a! sur dc Baeza, Bosque virgcn, 2100m. 24deenero \9M.Morun i5«6 (F, GH, MO, NY. Q, QCA, US); Cordillera Orien- tal, 5 km al sur de Cosanga. camino Bacza-Tena, 1(X)0 m. 4 de enero 1984. Moran iSll (Q, QCA); Cerro Huacamayos, on road Baeza-Tena, ca. 34 km from Baeza, 0llgaard ei al. 53824 (AAU, QCA). 26. Polyhotrya altescandens C. Chr. (Fig. 45, Map 121. Polyhotrya alte.uandens C. Chr.. Index Filicum 7. 1905. nam. nov. for Acroslichum chrysolepis Sodiro, non Fee 1869. Acrostichum chrysolepis Sodiro, [Anal. Univ. Quito (Xl)77:56l. 1894.|Crypt. Vase. Quit. 485. 1893. mm. illeg., non Fee 1869. Typr: Ecuador. Pichincha: selva tropical, de Pilaton-Toachi, scptiembre 1892. Sodiro s.n. (holotypc; PI). Stem (1)2-3(4) cm thick, hemiepiphytic; scales golden to yellowish, mostly concolorous, shiny, tangled and forming a dense "woor" around the stem, linear. 0.6-1.3 x 15-25(32) mm, base cordate and darkened at the point of attachment. Sterile leaves up to 2 m long; lamina to 1 .65 m long, lanceolate, to 2-pinnate-pinnatifid or rarely 3-pinnatc, usually moderately pubescent, the hairs fine, appressed, more or less cylindrical, multicel- lular, 0.2-0.4 mm long, inlcrgrading with scales on the costa; pinnae evenly tapering to a truncate base, the acroscopic side slightly prolonged, espe- cially in the distal pinnae. 14-25(35) x 3-10(17) cm. soon becoming pinnatifid in the distal por- tions; pinnules anadromous or subequal proxi- mally. catadromous distally, entire tocoarsely ser- rate, the basal acroscopic one almost always con- spicuously prolonged beyond the others; axes pubescent abaxially, the hairs subulate, colorless, 0.1-0.2 mm long, unicellular, the scales few, scattered, denticulate; grooves slightly pubescent within, those of the costulc dceurrent into those of the costa, costal groove gradually becoming shallow and expanded before the juncture with the rhachidial groove; veins in pinnate groups, in large pinnae and pinnules the lowermost basiscopic veinlet often springing directly from the costa or coslulc. appearing as a single isolated veinlet be- tween the main pinnate groups. Fertile leaves 3- pinnate, pubescent by jointed hairs 0.5-1.0 mm long, these grading into linear, tortuous scales; sori obovatc or oblong; sporangiul stalks para- physate; spores (54)59-69(72) microns long. « = 4 1 . This species has an altitudinal range of (8(K)) 1200-2500 m. Most of the collections come from the western cordillcra of Ecuador, but two apparently disjunct locations occur in Colombia 98 Illinois Natural History Slrvey Vol. 34. Art. 1 FiuuRi- 45. Polybolrya allescimdens C. Chr. a. stem scales; b. fertile leaf; c. fenile pinnae; d. sterile pinnule (note isolated veinlet between the main pinnate groups); e. fertile pinnule (note obovate seg- ments), a.b: Moran 3559 (CR). c.e: Rimbach 91 (F). d: Steyermark 52S12 (F). November 1987 Monograph of Polybotra 99 and Peru (Map 12). I suspect that this species occurs in many other valleys of the Colombian and Peruvian Andes, and that lack of collecting accounts for the apparent disjunctions. My field experience in Ecuador showed this species to be quite common in the western cordillera but absent from the eastern cordillera. Polyhotrya allescandens can be immediately recognized by its bright golden to yellowish scales that thickly cover the stem. These scales are longer and narrower than those of any other species of Pohbotrya (Fig. 45a). The lamina is less cut than that of similar Andean species, and the pinnae soon become pinnatifid in their distal portions, a characteristic that further accentuates this less- divided look. The sori are usually clavate or short- oblong, in contrast to the more elongated sori of other species (Fig. 45e). An unusual tendency of the venation further distinguishes this species: in large pinnae and pinnules the lowermost basi- scopic vein migrates onto the costa or costule and therefore appears as a single isolated veinlct be- tween the main pinnate groups (Fig. 45d). Specimens examined: Colombia Antjoquia: bosquc bajii la cumbrc cerca dc Boqucron. camino cntre Medellfn y Palmilas. 2.100-:.'i(X) m. HoJi:cfi\. US); Bath, Gilbert s.n. (GH); upper soulhem slopes and summit of Mac- casucker Bump, Maxon 9522 (GH. NY. PH); SE slopes of Stone Hole Bump, Maxon 8978 (GH. NY, PH); Com Puss Gap and vicinity, trail W over Blue Mis.. Wilson & Murray 564 (BM. GH. MICH); Cora Puss Gap, Proc- tor 3969 (PH. US). HAm. Massif du Nord. Chavar>, Ekman 4734 (US); Camp Pcrrin. Ekman 5214 (US) Grenada. Grand Eiang, Beard 1252 (UC. US); no localitv , Fraser .v. n . ( P); in Mirabeau Mts. . Broadway November 1987 MONCXjRAPU of POIYBOIRA 105 2520 (Z); St. Georges, Azimas. 1896, Broadway s.n. (NY); without locality, Sherring 13 (BM). TRINIDAI3. Pass, Arima Valley. Fleming i 74 (MICH); carretera al mar cerca de Villa Arleaga, Gutierrez & Barkley 1 70109 (GH); Municip. Anori, Pro- videncia, Soejarto2805 (COL) Boyaci: Muzo, Lindsay 262 (BM). Cauca: Agua Clara, along hwy from Buena- venturea to Cali, Killip ecific epithet refers to the raised flap af tissue that borders the costules and accentuates the depth of the central groove. This character is not diagnostic, since other decompound Polyhot- rya species also have raised costular ridges, al- though in P. canaliculata these ridges tend to be relatively taller and more well developed. When available, the botryoid fertile leaves [Fig. 52a,c) distinguish this species from P. os- mimdacea and its allies. Distinguishing vegetative -haracters arc the dull brown, concolorous stem scales, and the major axes with prominent, ovate to lanceolate, flaccid scales (Fig. 52b). The pin- fiule arrangement of P. canaliculata may be either anadromic or catadromic. This variation is pecu- liar because the pinnule arrangement is usually jonslant within most species of Polyhotrya. A similarly variable species in this respect is P. ^omezii, an endemic from Costa Rica. I interpret P. canaliculata as a primitive species of Polyhot- n,a because of its decompound lamina and bot- ryoid fertile leaf. I have chosen the Moritz collection as the lectotype because of its wide distribution in her- baria and the excellent quality of the specimens. Specimens examined: Vknk/uki.a. Aragua: Colonia Tovar, 1846, Moriiz 278 (B, BM. F-, GH, L, NY. P. US); Colonia Tovar. 1846. Karsten 13 (B. BM. L; photo of L specimen. NY): Parque Nacional Henri Pittier. bosque de Rancho Grande. Tschudi 166 (VEN); Colonia Tovar, 1854-5, Fendler 262 (GH, MO, NY, P, PH. US) Falc6n: Sierra de San Luis, arriba de Sta. Maria, alt. I2(X) m. 5 June 1979, van der Werff 3486 (MO. UC); Sierra de San Luis, selva nublada. entre La Chapa y Una, Steyermark 99185 (VEN). Yaracuy: Dito. Bolivar, entre las Parchitas. Tierra Fria y Ojo de Agua. Onega & Smith 2498 (PORT). 2511 (PORT). State unknown: Andes of Venezuela, 1889, Coebel s.n. (P). 33. Polybotrya semipinnata Fee (Fig. 53, Map 18). Polybotrya semipinnata Fee, Crypt. Vase. Bresil. 1:16. 1869. Type: Brazil. Rio de Janeiro: Yacuacanga, 15 June 1869, Glaziou 2427 (K, P!, RB!, US!; photo of K specimen at US!). Aspidium scandens Raddi, Plant. Brasil. 1:34, tab. 49. 1825. Type: Brazil. Raddi s.n. (FI; isotype: K, photo GH!). Polyhotrya scandens (Raddi) Christ, Bull. Herb. Boissier, IF 4:965. 1904. nom. illegil., non F6e 1852. Stem 1-1.5 cm thick, hemiepiphytic; scales membranous, spreading, mostly 9-12 x 0.5-1.2 mm, bright castaneous, concolorous or with a dark central stripe and lighter borders, the margins denticulate to strongly erose. Sterile leaves up to 1 m long; petiole 'A to 'A the length of the lamina; lamina ovate to lanceolate to 3-pinnate but mostly 2-pinnate-pinnatifid throughout, generally 45- 60(75) X 30-55(64) cm, coriaceous, glabrous ex- cept on major axes; pinnae broadly triangular to ovate , mostly l5-30x Il-I7cm, divided at base and soon becoming pinnatifid distally; pinnules 5-10(13) X 2-3 cm, lanceolate, anadromic throughout, the bases mostly cuneate, unequal, the basiscopic side more narrowly cuneate and the acroscopic side slightly prolonged, proximal ones with a 3-5 mm long stalk, rarely with the lower- most acroscopic segment cut to the costule; axes pubescent abaxially with hairs less than 0.1 mm long, stiff, colorless, subulate, the scales few or absent; grtmves glabrous or nearly so within, decur- rent on those of the next lower order. Fertile leaves smaller than the sterile, 3-pinnate (-pinnatifid), cocnosoric; tertiary segments oblong, generally 3-7 mm long; sporangial stalk paraphysate; spores (48)50-62(66) microns long. Iii.iNOis Na ruRAi. History Survey Vol.34. Art. I Figure 53. Polyholrya scmipiiiiuiui Fee. a. sterile pinnules; b. sterile leaf; c middle pinnae of fertile leaf; d. stem scales, a-c: BraJc S()6I (PH). d: Diuirlc ct al. 65322 {¥). November 1987 Monograph of Polybotra lis Other illustrations: Raddi, Plant. Brasil. tab. 49. 1825; Mettenius, Filices Hon. Lips., 23, tab. 2, figs. 1-6. 1856 (as P. acuminata); Brade, Bradea 1:64, fig. 1; 67, fig. 9. 1971 (as P. scan- dens). Polybotrya semipinnata is one of five Polybotrya species that are endemic to the coastal mountains of southeastern Brazil (Map 18, Table 2), a distribution that emphasizes the biogeo- graphic distinctness of the Serra do Mar Moun- tains. The altitudinal range of P. semipinnata is from 800 to 1000 m. This plant, easily recognized by its lamina cutting, is not readily confused with other species of Polybotrya. The pinnule bases are distinctive because they are stalked and more narrowly cuneate on the basiscopic side. The pinnae are stouter and much less divided compared to those of other decompound Polybotrya species; the dis- tal portions soon become pinnatifid, a characteris- tic that accentuates this less-cut appearance. An earlier published name, P. acuminata Link, has often been applied to this species. Since the type specimen cannot be located and I cannot discern from Link's description the species he had in mind, 1 treat P. acuminata as a name of uncer- tain application. Specimens examined: Brazil Rio De Janeiro: Yacuacanga, Glaziou 2427 (P, RB, US; photo of K specimen at US); Estrada Veiha da Bocaina, Teresopolis, Duarle el al. 65322 (F, LP); Serra dos Orgaos, Teres6polis can. Quebrafrasco, 1000 m, Brade 16456 (RB). Sao Paulo: Piruhyba. Loefyren & Duvall 36217 (RB); propc Rio Grande ad Sao Paulo Railway, 800 m, Wett.':lein & Schiffner .<^.n. (P); Alto da Serra, Lueder- waldl s.n. (BM, NY), 21547 (NY); Iguape. morro das Pedras, Semnha Peroupara, Brade 8061 (NY, PH, UC, US); Pilar, Gerdes 102 (NY, UC). SU. Catarina: with- out locality, Schwacke s.n. (P). 34. Polybotrya speciosa Schott (Fig. 54, Map 21). Polybotrya speciosa Schott, Genera Filicum tab. 7. 1834. NEOTYlfE: (here chosen) Schott, Genera Filicum tab. 7, based on material from "Brasiliae provincia Scbastianopolitana." Polybotrya tomentosa Brade, Arq. Inst. Biol. Veg. Rio de Janeiro 1:224, fig. 2, plates 2 and 3. 1935. Type: Brazil. Minas Gerais: Serra do Itatiaia Maromba, 25 June 1930, Brade 1035 1 (holotypc: RB!), Polybotrya osmundacea Willd. var. cris- popaleacea Rosenst., FeddesRepert. 21:349. 1925. Type: Brazil. Sao Paulo: Alto da Serra, 11 February 1925, Brade 5838 (S; isotype: UC!). Polybotrya litoralis Brade, Bradea 1:26, tab. 1 , fig. 2. 1969. Type: Brazil. Rio de Janeiro: Angra dos Reis, Serra do Mar, 29 June 1935, Brade 14943 (holotype: RB!). Polybotrya rosenstockiana Brade, Bradea 1:27, tab. 1, fig. 3. 1969. Type: Brazil. Rio de Janeiro: Serra dos Orgaos, Corrego Beija- flor, Brade 16579 (holotype: RB!; isotype: LP!). Stem 1-3 cm thick; scales commonly brick red, rarely dull brown, concolorous or with a dark central stripe, generally 8-20(27) x 0.5-1 .5(2.0) mm, spreading, membranous, the margins den- ticulate to strongly erose. Sterile leaves up to 1.4 m long; petiole 'Ao- Vi the length of the lamina; lamina to 1.2x0.8 m, lanceolate to ovate, to 3-pinnate but mostly 2-pinnate-pinnatifid through- out, tomentose to glabrous, the margins sparsely ciliate to glabrous; pinnae to 40 x 15 cm, free pinnules usually 5-9; pinnules acroscopic, short to long triangular, mostly 4.5-8.0(10.0) x 1.5- 3.5 cm, the base stalked, the stalk 2-4 mm long, the acroscopic side prolonged, the basiscopic side oblique; tertiary segments generally oblong, the margins entire to crenulate or dentate; a.xes tomen- tose to glabrous or subglabrous, the hairs usually 0.1-0.5 mm long, the scales few, appressed, tor- tuous, narrow; grooves pubescent within, the hairs reddish. Fertile leaves coet)osor\c , 3-pinnate; jrpo- rangial stalks paraphysate; receptacle glabrous or with multicellular, branched paraphyses, these as long as or slightly longer than the sporangia; spores (56)60-75(82) microns long. Other illustrations: See original descriptions cited above; Brade, Bradea, tab. 3, fig. 4 (as P. littoralis); tab. 4, fig. 1 (as P. rosenstockiana); tab. 4, fig. 3; tab. 6, figs. 12, 13 & 15. 1971. Polybotrya speciosa is endemic to the Serra do Mar Mountains along the coast of southeastern Brazil (Map 21). It differs from the four other species of Polybotrya there by its combination of strongly denticulate, red stem scales, finely cut lamina, and multicellular, branched paraphyses (Fig. 54). The pubescence of the abaxial surface varies from densely tomentose to nearly glabrous. 116 Illinois Natural History Survey Vol 34. Art. 1 1 Figure 54 Pohbotna speciosa Schott. a. stenle and fertile leaves; b. branehed paraphyscs; c. pin- nules; d.e. stem scales, a.b.e; Brade 16579 (RB). c.d; BroJe 1035 1 (RB). November 1987 Monograph of Polyboira 117 Five of the specimens examined had the adaxial surface slightly pubescent whereas the remaining were completely glabrous. No other species of Polybotrya, except P. pilosa, has branched para- physes. The spore size varies greatly, perhaps reflect- ing different ploidy levels. The averages from eight collections, with 15 spores measured from each, are as follows (in microns): 56. 57, 57, 58, 69, 69, 79, 82. This problem needs further study of additional collections andcytological samples. I have been unable to locate Schott's type, but I feel quite certain that Schott's excellent plate represents this species. Schott's specimens be- came part of the Cardinal Hynald herbarium now located in Budapest, Hungary (BP). The curator of the fern collection at Budapest, Mr. Tibor Szer- dahelyi, informed me (in litt.) that much of Schott's type material had been destroyed during World War II and that he could not find the type. My placement of the three species described by Brade in synonymy with P. speciosa requires comment. I find no differences between the types of P. tomentosa and P. liloralis— the two might well have been collected from the same individual. Given this likeness, I find it odd that Brade ( 1 969c) did not mention P. tomentosa in his discussion after the description of P. litoralis. I also place P. rosenstockiana and P. osmundacea var. cris- popaleacea in synonymy, although they differ slightly from most specimens of P. speciosa by their somewhat broader stem scales (Fig. 54, com- pare d & e). They also differ by their leaves, which are less pubescent, primarily so along the axes and veins. Both these characters, however, intergrade and do not correlate with any others. Specimens examined: Brazil. Rio de Janeiro: Serra dos Orgaos, Corrego Beijaflor, Brade 16579 (LP, RB); Serra do Taquaral, Brade 17464 (MO. NY);There- zopolis. Brade 977.? (BM. UC). 9M3 (NY); Parque Nacional de Serra dos Orgaos. de la Sola 234i (LP); Serra dos Orgaos. Morro Assu. Luelzelhurg 6H5H (US); Organ Mts . Rose & Russell 20790 (US); Corcavado, collected by the U.S. South Pacific Exploring Expedi- tion. 1838-42 (NY. US); Mt. Tijuca. Cuyler4796{\}S): Angra dos Reis. Serra do Mar. Brade 14943 (RB); Guanabara. Estrada do Sumaro, Pahsl el al 6772 (LP). Minas (ierais: Serra do Itatiaia. Maromba. Brade 1035 1 (RB); Itatiaia. Maromba. Brade 20214 (F. LP. MO. NY); same locality. Brade 12616 (BM); Tijuca. Alston ,S99/(BMl Sao Paulo: Alto da Serra. flralia (Ret/.) Mettenius, Fil. Lcchler. 2:12. 1859. = Leplochilus zeylandicus (Houtt.) C. Chr. (fide Christensen, 1905). Polybotrya rhizophylla (Kaulf.) Presl, Tent. Pterid. 231. 1836, = Bolbilis rhizophylla (Kaulf.) Hennipman (fide Hennipman 1977). Polybotrya serrulala Fee, Mem. Fam. Foug. (Hist. Acrost.) 76, pi. .39, fig. ii. 1845. = Bol- hilis rhizophylla (Kaulf.) Hennipman (Jlde Hen- nipman 1977). Polybotrya sinensis (Baker) C. Chr., Index Fil. 57. 1913. = Bolbilis sinensis (Baker) Iwatsuki (fide Hennipman 1977). Polybotrya sorbifolia (L.) Keyserling, Pol. Cyath. Herb. Bung. 32. 1873. nomen illegil. = Lomariopsis sorbifolia (L.) Fee (fide Proctor 1977). Polybotrya stenosemioides (Baker) Copel., Polypod. Philipp. 40. 1905. = Helerogonium stenosemioides (Baker) C. Chr. (Jlde Holttum 1975). Polybotrya subqiiinqitefida (Fee) Mettenius, Fil. Lechler. 2:12. 1859. =lLeptochilus laiifolium (Meyen) C. Chr. (fide Christensen 1905). Polybotrya laccaefolia (J. Smith) Mettenius, Fil. Leehler. 2:12. 1859. = Leplochilus latifolius (Meyen) C. Chr. (Jlde Christensen 1905). Polybotrya lenuifolia (Desv.) Kuhn, Fil. Afr. 52. 1868. = Slenochlaena lenuifolia (Desv.) Moore (fide Christensen 1905). Polybotrya leysmanniana (Baker) Posthumus, Ree. Trav. Bot. Neeri. 33:872. 1930. = Stenosemia leysmanniana (Baker) Diels (fide Christensen 1905). Polybotrya trilohaia (J. Smith) Mettenius, Fil. Lips. 24. 1856. = Leptochiltts latifolius (Mcycn) C. Chr. (Jlde Christensen 1905). Polybotrya tripartita (Hooker & Grev.) J. Smith, Jour. Bot. 4:150. 1841. - Peltapteris tripartita (Hooker & Grev.) Morton (fide Morton 1955). Polybotrya vivipara Hooker, Exotic Flora, 2: pi. 107. 1825. = Bolhilis appendiculala (Willd.) Iwatsuki subsp. vivipara (Hooker) Hennipman (fide Hennipman 1977). Polybotrya wilkesiana Braekenridge, U.S. Expl. Exped. 16:80, tab. 10. 1854. =Teralophyllum wilkesianaum (Braekenridge) Holttum (fide Holttum 1978). 122 iLLfNois Natural History Survey Vol 34. Art. 1 Literature Cited Bierhorst, D.W. 1971. Morphology of vascular plants. Macmillan, New York. Bl.UME, C.L. 1828. Enumeratio plantarum Javae et in- sularum adjacenlium Fasc. II Filices. Leiden. Brade. A.C. 1935. Contribu^ao para a flora do Ilatiaia. Filices novae Brasilianae. III. Arquivosdo Institute de Biologia Vegetal 1:223-230. Brade, A.C. 1948. Contribugao para o conhecimento da flora do estado do Espirito Santo (I. Pteri- dophyta). Rodriguesia 10(21 ):25-56. Brade, A.C. 1969a Algumas especies novas de filicineas de Costa Rica da cole?ao Alfred & Alex- ander Curt Brade Bradea 1:11-17, lab. 1-5. Brade, AC. 1969b. Duas especies novas do genero Polybotrya (Polypodiaceae) da Venezuela. Bradea 1:19-21, tab. 1-2. Brade, AC. 1969c. Algumas especies novas do genero Polybotrya da flora do Brazil. Bradea 1:23-28, tab. 1. Brade, AC. 1971. O genero Polybotrya no Brazil. Bradea 1:57-67. Chandra, S. 1975. Some morphological aspects of the rhizome of Maxonia C. Chr. (Dennstaedtiaceae). Brenesia 6:1-7. Christ, H. 1897. Die Famkrauter der Erde. Fischer, Jena. Christensen, C 1905. Index Filicum. H. Hagerup, Copenhagen. Christensen, C. 1913. A monograph of the genus Dr/opteris. Part 1 . Del kongelgelige danske vi- denskabemes selskabs naturvidenskabelige og mathematiske afhanlinger, VII. 10:55-282 Christensen, C. 1916. Maxonia, anew genus of trop- ical American ferns. Smithsonian Miscellaneous Collections. 66(9): 1-4. Christensen, C. 1934. Index Filicum. Supplementum lertium pro annis 1917-1933. H. Hagerup, Copenhagen Copeland, E.B. 1947. Genera Filicum. Chronica Botanica, Waltham, Mass. DIEIS, L. 1899. Polypodiaceae. In A. Engler and K. PrantI, Die Natiirlichen Pflanzenfamilien, 1, 4: 1 39- 339. FfiE, A.LA. 1845. Deuxieme Memoire: Histoire des Acrostichees. Veuve Bcrger-Lcvraull, Stra.sbourg. G6MFJ'., L.D. 1976 Contribucioncs a la plcridologi'a centroamericana, I. Enumeratio Filicum Nicara- guensium. Brenesia 8:41-57 Hennipman, E. 1977. A monograph of the fern genus Bolbitis (Lomariopsidaccae) Leiden Botanical Series No. 2, Leiden University Press Hcn.TTUM, RE. 1959. Vegetative characters distin- guishing the various groups of ferns included in Drvopteris of Chrislenscn's Index Filicum and other ferns of similar habil and son Gardens' Bulletin (Singapore) 17:.36l-367. HoLTTUM, R.E. 1963. Cyatheaceae. Flora Malesiana, II, 1:65-176. HoLTTUM, RE. 1975. The fern genus Heterogonium Presl. Kalikasan 4:205-231. HoLTTUM, RE. 1978. Lomariopsis group. Flora Malesiana, II, 1:255-330. HoLTTUM, RE 1982. The continuing need for more monographic studies of ferns. Fern Gazette 12:185- 190. HOLTTUM, RE. 1984. Studies of fern genera allied to Tectaria, I. A commentary on recent schemes of classification. Fern Gazette 12:313-319. Hooker, W.J. 1864 Species Filicum, vol. 5. London. Hooker, W.J., and J G. Baker. 1874. Synopsis Filicum, 2nd ed. London. Kaulfuss, G.F. 1824. Enumeratio Filicum. Cnobloch, Leipzig. KoPTUR, S., A.R. Smith, and I. Baker. 1982. Nec- taries in some neotropical species of Polypodium (Polypodiaceae): preliminary observations and analyses. Biotropica 14:108-113 Lellinoer, D.B 1972 Five new species of South American Polybotrya. American Fern Journal 62:49-56. Lellinoer, D.B 1977. Nomenclatural and taxonomic notes on the pteridophytes of Costa Rica, Panama, and Colombia, I. Proceedings of the Biological So- ciety of Washington 89:703-732. Mayr, E 1982 The growth of biological thought. The Belknap Press of Harvard University Press, Cam- bridge, Mass. Mehra, P.N.. and T.C. Mpttal 1961 Significance of internal glands in relation to filicin. Planta Medica 9:189-199. Mettenius, G. 1856. Filices lechlerianae Leipzig Mickel, J.T. 1980. Relationships of the dissected elaphoglossoid ferns. Brittonia 32:109-117. Moran, R C. 1986. The neotropical fern genus Olfer- sia. American Fern Journal 76:161-178. Moran. R.C 1987. Slenlc-ferlilc leaf dimorphy and the evolution of soral types in Polybotrya (Dryop- teridaceae). Systematic Botany 12:617-628. Morton, C.V 1955 Notes on Elaphoglossum. MI. The publication of Elaphoglossum and Rhipidop- teris. American Fern Journal 45:1 1-14. Morton, C V. 1971. The fern collections in some European herbaria, VII. American Fern Journal 61:59-75. Pichi-Sermoi 11, REG. 1977. Teniamen ptendiv phytorum genera in taxonomicum ordinem redi- gendi Webbia 31:313-512. Posthl'mus, O 1928 The ferns of Surinam and French and British Guiana Published by the author. Malang, Java. PRE.SL, KB. 18.36. Tentamen Ptcridographiae Haasc, Prague November 1987 Monograph of Poi.ybotra 123 Proctor, G.R. 1977. Pteridophyta Pages 1-414 in R. A. Howard, ed. Flora of ihe Lesser Antilles, vol. 2. Arnold Arboretum of Harvard University. Jamaica Plain, Mass. SCHOTT, H. 1834-1836. Genera filicum. Vienna. Schumann, E. 1915. Die Acroslicheen und ihre stel- lung im system der fame. Flora 108:201-260. Smith, A.R. 1986. Revision of the neotropical fern genus Cyclodium. American Fern Journal 76:56- 98. Smith, A.R., and J.T. Mickel. 1977. Chromosome counts for Mexican ferns. Brittonia 29:391-398. Smith, J. 1841. An arrangement and definition of the genera of ferns. Journal of Botany 4:147-198. Smpth, J. 1875. Historia filicum MacMillan, Lx)ndon. SODiRO, A. 1897. Cryptogamae Vasculares Quitensis. Quito. Stafleu, FA. 1967. Taxonomic literature. Regnum Vegetabile, vol. 52. Stafleu, F.A., and R.S. Cowan. 1981. Taxonomic literature, vol. Ill: Lh-O. Regnum Vegetabile, vol. 105. Stol^e. R.G. 1981. Fems and fern allies of Guatemala. Part 2: Polypodiaceae. Fieldiana: Botany, new series, 6:1-522. SwARTZ, O. 1806. Synopsis Filicum. Bibliopoli Novi Academici. Kiliae. SwoFFORD. D.L. 1985. PAUP user's guide, version 2.3. Illinois Natural History Survey, Champaign. Tryon.R.M. 1970. Development and evolution of fern floras of oceanic islands. Biotropica 2:76-84 Tryon, R.M. 1972. Endemic areas and geo- graphic speciation in tropical American fems. Bio- tropica 4: 1 2 1 - 1 3 1 . Tryon, R.M.,and A.F. Tryon. 1982. Fems and allied plants, with special reference to tropical America. Springer-Verlag, New York. VAN DER Werff, H., and A.R. Smhh. 1980. Pteridophytes of the State of Falcon, Venezuela. Opera Botanica 56:1-34. Varhschi. V. 1969 Helechos. Rora de Venezuela, 1(1), Edicion Especial del Instiluto Botanico, Caracas. Wagner, W H , Jr. 1952. The fern genus Diellia, its structures, affinities, and taxonomy. University of Califomia Publications in Botany 26:1-21 1. Wagner, W.H.. Jr., and DM. Johnson. 1983. Trophopod, a commonly overlooked storage struc- ture of potential systematic value in fems. Taxon 32:268-269. Wagner. W.H., Jr., and F.S.Wagner. 1977. Fertile- sterile leaf dimorphy in fems. Gardens' Bulletin (Singapore) 30:251-267. Walker, T.G. 1966. A cytotaxonomic survey of the pteridophytes of Jamaica. Transactions of the Royal Society of Edinburgh 66:169-237. Walker, T.G. 1972. The anatomy of Maxonia apiifolia: a climbing fern. British Fern Gazette 10:241-250. Widen, C-J, J. Sarvela, and DM. BRirroN. 1983. On the location and distribution of phloroglucinols (filicin) in fems. Annales Botanici Fennici 20:407- 417. Willdenow, C.L. 1810. Caroli a Linne species plan- tamm, vol. 5, 4lh ed. Nauk. Berlin. Windisch. P.G. 1982. Specimens from Fee's pterido- logical collection at the Botanical Garden of Rio de Janeiro American Fern Journal 72:56-60. 124 Illinois Natural History Survey Vol.34. Art. I Taxa and Distribution of Polybotrya Numbers correspond to the species numbers as- signed in the taxonomic treatment. 1 . Polybotrya serratifolia (F6e) Klotzsch: Trini- dad, Venezuela. 2. Polybotrya polybotryoides (Baker) Christ: Mexico, Belize, Guatemala, Honduras, Costa Rica, Panama, Colombia, Ecuador, Peru. 3. Polybotrya suberecta (Baker) C. Chr.: Co- lombia, Ecuador, Peru. 4. Polybotrya andina C. Chr.; Ecuador. 5 . Polybotrya sorbifolia Kuhn; Costa Rica, Ven- ezuela, Colombia, Brazil. 6. Polybotrya fractiserialis (Baker) J. Smith: French Guiana, Surinam, Guyana, Ecuador, Peru, Bolivia. 7. Polybotrya crassirhizoma Lellinger: Colom- bia, Ecuador, Peru, Bolivia, Brazil. 8. Polybotrya espiritosantensis Brade: Brazil. 9. Polybotrya caudata Kunze: Mexico, Guate- mala, Belize, Honduras, Nicaragua, Costa Rica, Panama, Trinidad, French Guiana, Surinam, Guyana, Venezuela, Colombia, Ecuador, Peru, Bolivia, Brazil. 10. Polybotrya goyazensis Brade: Brazil, Para- guay. 1 1 . Polybotrya pubens Martins: Colombia, Ecua- dor, Peru, Bolivia, Brazil. 12. Polybotrya glandulosa Kuhn: Venezuela, Peru, Brazil. 13. Polybotrya lechleriana Mettenius: Guyana, Colombia, Ecuador, Peru, Bolivia. 14. Polybotrya attenuata Moran: Colombia. 15. Polybotrya stohei Moran: Colombia. 16. Polybotrya alfredii Brade: Nicaragua, Costa Rica, Panama, Ecuador, Peru, Bolivia. 17. Polybotrya botryoides (Baker) C. Chr.: Co- lombia. 18. Polybotrya lourteigiana Lellinger: Colombia. 19. Polybotrya pittieri Lellinger: Colombia. 20. Polybotrya cylindrica Kaulfuss: Brazil. 21. Polybotrya hickeyi Moran: Colombia. Bo- livia. 22. Polybotrya puberulenta Moran: Ecuador, Bolivia. 23. Polybotrya alata Moran: Panama. 24. Polybotrya aequatoriana Moran: Ecuador, Bolivia. 25. Polybotrya appressa Moran: Ecuador. 26. Polybotrya altescandens C. Chr.: Colombia, Ecuador, Peru. 27. Polybotrya gomezii Moran: Costa Rica. 28. Polybotrya osmundacea Willd.: Guatemala. Honduras, Nicaragua, Costa Rica. Panama, Cuba, Jamaica. Haiti. Grenada, Trinidad, French Guiana, Guyana, Venezuela. Colom- bia, Ecuador. Peru. Bolivia. Brazil. 29. Polybotrya cyathifolia Fee: Guadeloupe, Martinique. 30. Polybotrya latisquamosa Moran: Colombia. 31. Polybotrya sessilisora Moran: Colombia, Brazil. 32. Polybotrya canaliculata Klotzsch: Vene- zuela. 33. Polybotrya semipinnata ¥€t: Brazil. 34. Polybotrya speciosa Schott: Brazil. 35. Polybotrya pilosa Brade: Brazil. Distribution Maps The small open dots on the twenty-one distribution maps that follow indicate towns. Other symbols are explained in the legends. November 1987 Monograph of Polybotra 125 Map 1. Distribution of Polyhotrya suherecta (Baker) C. Chr. (diamond) and P. serratifolia (Fde) Klotzsch (circle). Map 2. Distribution of Polybolrya polyhoinoides (Baker) Christ (circle) and P andina C. Chr. (square). 126 Illinois Natural History Survey Vol. 34. Art. 1 Map 3. Distribution of Potybolna sorhifolia Kuhn (diamond) and P. fractiseriaUs (Baker) J. Smith (circle). Map 4. Distribution of Polybolrya crassirhizoma Lellinger. November 1987 Monograph of Polybotra 127 Map 5. Distribution of Polybotrya caudata Kunze. Map 6. Distribution of Polyboirya pubens Mat- tius (circle) and P. goyazensis Brade (diamond). 128 Illinois Natural History Survey Vol.34. Art. 1 •) - -^ ' Map 7. Distribution of Potybotrya lechleriana Mettcnius (circle), P. attenuata Moran (triangle), and P. glandulosa Kuhn (square). Map 8. Distribution of Polybotna atfredii Brade (circle) and P. boiryoides (Baker) C. Chr. (triangle). Map 9. Distnbution of Polyholrya siohei Moran. Map 10. Distribution of Polybotrya lourteigiana Lellinger. Map 1 1 . Distribution of Potybotrya pittieri Lel- linger. November 1987 Monograph of Polybotra 129 Map 12. Dislnbution of Polybotryaaltescandens C.Chr. Map 13. DhlribMion of Polyholrya hickeyi Moran (circle) and P. puberulenta Moran (triangle). Map 14. Distribution of Polyhotrya gomezii Moran (circle) and P alata Moran (triangle). Map 15. Distribution of Polyhotrya aequatoriana Moran. Map 16. Distribution of /'o/>'b<>/o'ac>'/iminca Kaulfuss. Map 17. Distribution of Po/>*o/- rya appressa Moran. 1 30 iLLtNois Natural History Survey Vol. 34, Art. 1 18 Map 18. Distribution of Polyhotrya osmumlacea Willd. (circle) and P. semipinnaUi Fdc (triangle). Map 19. Distribution of P. latisquamosa Moran (solid square). P. cimaUcuhua Klotz.sch (triangle), and P. lyathifolia Fee (open square). November 1987 MONCXJRAPH of POLYBOTRA 131 Map 20. Distribution of Polyhotrya sessilisora Moran (circle) and P. pilosa Bradc (triangle). Map 21 . Distribution of Polybotrya speciosa Schott (circle) and P. espiriiosaniensis Bradc (triangle). 132 Illinois Natural History Survey Vol. 34. An. 1 Index to Collectors' Numbers Numbers in parentheses refer to the species num- bers assigned in the taxonomic treatment. Acosta Solis 6358 (2). Acuna 17531 (28). Aguilar 947 (6). Alfaro 8073 (5). Allard 20661 (6); 21609 (28); 21997 (28); 22305 (6); 22526 (6); 22593 (9). Alston 7341 (28); 7641 (28); 8278 (21); 8991 (34). Alverson et al. 342 (2). Anderson 10152 (5). Antonio 4019 (2); 5133 (28). Argent & Richards 6651(10). Argent eta! . 6336 (10). Aristeguieta 1780 (28); 3963 (1). Armond 298 (2). Atwood 3711 (28). Aymard et al. 952 (9). B.T. 443 (9). Bailey & Bailey 541 (9). Balslev 4787 (11); 4802 (11). Balslev & Madsen 10649 (28). Barbour 4764 (9); 4790 (9). Barclay 2199 (9). Barkley & Gutierrez V. 1897(28). Barnard et al. 411 (28). Beard 1252 (28). Beck 1635 (7); 3060 (24); 3108 (24); 4924 (11); 8037 (9). Bennett 23 (28). Benoist 1271 (9). Berget al. P18138 (28). Berry 948 (1). Billict & Jadin 1683 (9). Biolley 73 (16); 10688 (9). Boom & Mori 1856 (9). Boutin & Schlosser 5902 (2). Brade 372 (2); 554 (16); 5838 (.34); 8061 (33);9365 (20); 9773 (34); 9787 (35); 9843 (34); 10351 (34); 12576 (20); 12614 (20); 12616 (34); 13982 (20); 14943 (34); 15373 (10); 16456 (33); 16579 (.34); 16824 (16); 17464 (34); 18224 (8); 19972 (20); 20214 (34); 20733 (35); 20931 (20). Bradc & Brade 67 (16); 98 (16); 374 (9). Brandbygc & Asanza C. 3 1 870 ( 1 1 ); 32881 (7). Brandbygc ct al. 32544 (7); 32617 (7); 33684 (7). Breedlove 34101 (9). Breedlove & Smith 21886 (2). Brenes 11676 (16); 14248 (16); 21980 (16). Britton et al. 1267 (28); 1935 (9); 2144 (9); 2276 (28). Broadway 560 (9); 2520 (28); 5358 (9); 5589 (28); 5717 (9); 6459 (28); 6902 (28); 9207 (9); 9947 (1); 9948 (28); 9949 (1);9950 (28); 9951 (28). Buchtien 11 (6); 13 (6);35 (9&ll);260 (11); 290 (9); 298 (28); 299 (11); 1015 (6); 1066 (28); 1123 (11); 1124 (11); 2138 (6); 5164 (16); 5308 (13). Bues 1743 (13). Bunting 9516 (28). Burch 4414 (16); 4505 (16); 4613 (28). Burger & Antonio 11263(9). Burger & Stoize 5865 (9). Cain 74 (28). Camp 1298 (28); 1359 (28). Cardenas 1249 (6). Castellanos 25710 (20). Cazalet & Pennington 7720 (11). Chrysler 4836 (9). Chrysler & Roever 5095(16). Churchill 5776 (28). Churchill & de Nevers 4993 (2). Clement 725 (28). Clute 259 (28). Conant 940 (9); 1016 (31): 1080 (31); 1482 (31). Cooper 10240 (16). Comman 544 (9); 927 (16); 1178 (16); 1233 (16). Crcmers 4460 (9); 6374 (9); 6523 (9); 7369 (9); 7552 (9); 7951 (9). Croat 690 (28); 738 (28); 4266 (2); 5114 (9); 6850 (9); 7362 (9); 8025 (9); 8153 (9); 9000 (9); 9004 (9); 9103 (9); 10804 (9); 11544 (9); 12114 (2); 12143 (2): 13431 (16); 14687 (2); 14742 (2); 14786 (2); 15081 (9); 15256 (9); 17161 (2); 17367 (9); 17620 (9); 17689 (11); 18219 (9); 18388 (9); 18476 (28); 18508 (11); 19220 (9); 19717 (9); 19771 (11); 20287 (28); 20543 (28); 20551 (7); 20763 (7): 20786 (28); 20851 (28); 21014 (6): 21152 (6); 21191 (9); 21987 (28); 22260 (28); 22682 (28); 22706 (23); 22789 (2); 23011 (23): 23244 (2); November 1987 Monograph of Polybotra 133 24541 (9): 24566 (2); 27648 (2): 33452 (9); 35176 (9); 35608 (16); 36007 (16); 36669 (16); 36761 (16); 36800 (2); 37757 (2); 37794 (2); 38047 (28); 41637 (2); 49718 (13); 51154 (6); 55844 (26); 56510 (3). Croat & Folsom 34051 (2). Croat & Grayum 60022 (28). Croat & Porter 15342 (9); 15436 (9). Crosby 76 (28). Crueger 139 (9). Cuatreeasas 14246 (28); 15527 (18); 22143 (18). Cuyler 4796 (34). Davidse & Gonzalez 19444 (28). DcWolf 385 (16). Dodson 7380 (3). Dodson et al. 8679 (3). Donnell Smith 6939 (16). Duarle 3179 (20). Duarte & Pereira 65323 (20). Duarte et. al. 65322 (33). Dudley 10168 (9); 10325 (13); 11513 (9); 13005 (28); 13258 (16); 13290D (2); I8265A (16). Dumont et al. 7439 (28). Dunn & LeDoux 22005 (9). Duque-Jaramillo 1868 (18). Dusen 4423 (20); 6926 (20); 6984 (20); 13658 (20); 15351 (20); 15353 (20). Duss 1503 (29); 3897 (29); 4719 (29). Dwyer 8337 (16). Dyer A225 (9). Eggers 5324 (28). Ekman 3769 (28); 4734 (28); 5214 (28); I42I0 (28). Englesing 291 (9). Esposto 10928 (3). Evans & Bowers 2792 (28); 2944 (16); 3152 (2). Evoy 104 (9). Ewan 16846 (9). Fargens 251 (20). Fault 12583 (28). Fay 345 (28); 373 (28); 472 (9); 859 (1). Fendlcr 69 (28); 105 (9); 235 (1); 261 (1); 262 (32). Fernandez 951 (9). Fiedler* Koptur 51 (27). Fisher 132 (28). Fleming & Fleming 52 (28). Folsom 3590 (2). Forero & Jaramillo 1745 (9); 2469 (13). Foster P-84-42 (9); P-84-91 (6); 85-37 (3); 85-162 (2); 4020 (6); 7455 (6); 7858 (11); 7937 (6); 9284 (7). Foster & Foster 854 (8). Foster & Kennedy 1814 (2). Foumier 357 (2). Gardner 1901 (5). Gastony 43 (28). Gaudichaud 73 (20). Gentle 2947 (9). Gentry & Juncosa 41021(28). Gentry et al. 27365 (6); 27880 (7). Gerdes 102 (33). Gerrera 1624 (13). Gill 47 (7). Glaziou 955 (20); 2427 (33); 2428 (20); 14456 (10). Gleason 582 (9). Goedas 217 (20). Gomez 3324 (2); 3354 (9); 4528 (9); 6914 (2); 7027 (2); 7122 (5); 18064 (2); 18906 (16); 19489 (9 & 28). Gomez &Cittar 6409 (9); 6731 (9). Gomez et al. 20423 (9). Gonggrijp & Stahcl 3300 (6). Grant 10283 (21); 10436 (9); 10556 (3). Granville 971 (9); 991 (9); 1143 (9); 1165 (9); 3674 (6); 3865 (28); 4012 (9); B.4711 (9); 4841 (9); 5121 (6); 5219 (6). Grayum & Sleeper 3448 (16). Grijalva 300 (28). Guppy 462 (9). Gutierrez & Barkley 170109(28). Haerchen 124 (20). Hammel 3503 (2). Handro 1229 (20); 2228 (20). Harley et al. 10922 (10). Marling et al. 7535 (28). Hart 53 (9); 228 (1); 229 (28). Hassler 11618 (10). Hatch 90 (28). Hatschbac-h 7424 (20); 8141 (20); 10749 (20); 25118 (10). Haught 1325 (9); 1757 (5); 5498 (13). Hayes 8 (9). Henri-Stanislas 1709 (2); 1710 (28). Herrcra 12(X) (6). Hickey 801 (21). Hioram 2495 (28). Hioram & Clement 6424 (28). Hodge 6592 (26); 6780 (2). Holdridgc 5153 (2). 134 Illinois Natural History Survey Vol.34. Art. I Holm & litis 200 (9 & 28). Holm-Nielsen & Jeppsen 663 (11); 987 (7). Holm-Nielsen et al. 4489 (9); 4495 (9); 22121 (11); 22484 (7). Huber 2599 (9); 7235 (9). Humboldt 459a, b (28). Hunnewell 16514 (5); 18511 (20). Idrobo & Schultcs 1106 (30). Idroboct al. 10116 (28). Inarte 52 (1). Irwin 55161 (9). Irwin et al. 54669 (6); 54784 (6); 54896 (9). Jameson 33 (26). Jaramillo et al. 30783 (28). Jenny 2137 (9); 2368 (28); 2846 (28); 3122 (9); 10826 (9); 11195 (28). Jimenez M. 803 (16); 3200 (9); 3209 (2); 3210 (2); 3299 (28); 3304 (28). Johnson 1221 (9). Johnston 452 (9). Jones & Facey 3246 (9). Juncosa 1369 (17); 1467 (3). Kalbreyer 1254 (2); 1798 (2); 1873 (17); 1877 (3). Karsten 13 (32). Kellerman 7354 (9). Kennedy et al. 2074 (2). Kilhp 2544 (9); 2845 (9); 2929 (9); 5163 (16); 5202 (16); 5386 (16); 5429 (16); 11343 (9); 11549 (15). Killip & Cuatrecasas 38902 (28). Killip & Lasscr 37756 (28). Killip & Smith 15341 (28); 20210 (21); 23876 (7); 23916 (11); 23991 (28); 24605 (7); 25452 (28); 25921 (13); 25972 (11); 26194 (11); 26543 (II); 26637 (9); 26955 (11); 28760 (11); 30690 (9). Kirkbridc 404 (5). Kirkbride & Hayden 274 (2). Klawe 1474 (28); 1.504 (28); 1545 (2). King 1166 (9); 1386 (7); 1390 (12); 3208 (28). Knapp 4966 (9). Knapp & Dressier 5461 (2). Knapp & Mallet 5133 (9); 5402 (9). Knapp etal. 4479 (9); 4542 (2). Kramer 1954 (9). l-acgaard 511.36 (I1);5I150 (11);5I224 (II). Lankester 653 (16). Lasser & Vareschi 3926 (9). Lechler 2156 (13); 2176 (13); 2321 (9); 2329 (9). Lehmann B.T413 (13);500B (14); 2998 (9). Lehnun 2562 (20). Lellinger&delaSota 26 (9);39 (2); 190 (2); 213 (28|; 248 (3); 250 (28); 251 (18); 280 (15); 281 (13); 284 (15); 387 (18); 496 (9); 553 (28); 589 (9); 643 (9); 747 (15); 748 (15); 768a (18); 790 (9); 843 (18); 894 (3); 899 (18); 948 (18). Lellinger & White 1244 (16); 1361 (16); 1436 (28). Lent 299 (28); 1084 (27); 2642 (16); 3537 (27). Leon 226 (6). Liesner 820 (2); I6I33 (28); 16293 (12). Liesner & Judziewicz 14855 (16). Liesner et al. 8359 (28). Lindeman 4570 (9). Lindeman & Teunissen 15291 (9). Lindsay 262 (28). Loefgren & Duvall 36217 (33). Ludwig 262 (21). Luederwaldt 21547 (33). Luetzelburg 254 (20); 6858 (34). Lugo S. 3293 (7). Lundell 6416 (9). Macbride 5602 (7). Macedo 1447 (10); 1521 (10); 2682 (10). Madison etal. 3356 (2); 4851 (26). Maguire & Fanshawe 22855 (9). Maguire et al. 46068-A (6); 54384 (6). Maurel & Maurel 3816 (28). Maurice 699 (16). Maxon 412 (16); 639 (2t; 2293 (28); 2470 (28); 4641 (9); 6895 (9); 8978 (28); 9366 (28); 9464 (28); 9522 (28). Maxon & Har\ey 8242 (16). Maxon & Killip 148 (28); 756 (28). Maxon et al. 6817 (9). McAlpin & Kuhn 77-12 (14). Mendez 75 (9). Mexia 6246a (6). Mickel 1944 (9); 2003 (16); 2624 (28); 2742 (2); 2803 (28); 2817 (28); 2958 (16); 3126 (16); 3180 (2): 3368 (9); 3426 (16); 3570 (9); 9472 (9). Molina R. 1907 (9). Mora 4160 (18); 4287 (13); 4430 (9). Moran 2167 (28); 2170 (2); 2171 (2); 2173 (2); 2176 (2); 2178 (2); 2182 (9); 2185 (9); 2186 (9); 2237 (16); 2241 (16); 2325 (16); 2441 (27); 2442 (16); 3145 (5); 3160 (27);3168 (16);3214 (16);3241 (27); November 1987 Monograph of Polybotra 135 3271 {27);3338 (27);35I2 (24);3527 (25); 3528 (22); 3530 (2); 3532 (16); 3534 (9); 3535 (7); 3536 (6); 3544 (26); 3546.5 (3); 3547 (28); 3559 (26); 3561 (3); 3562 (2); 3563 (4); 3564 (26); 3565 (3); 3569 (2); 3570 (16); 3585 (24); 3586 (25); 3588 (9); 3590 (2); 3592 (28); 3593 (2); 3600 (28); 3612 (28); 3615 (7); 3616 (7); 3617 (9); 3618 (28); 3640 (7); 3641 (7); 3642 (28); 3663 (9); 3671 (9); 3688 (26); 3709 (1); 3718 (28). Morcly & Whitefoord 655 (28); 675 (28); 695 (28). Mori & Kallunki 3574 (2). Moritz 277 (1);278 (32). Morton 7629 (2). Mosen 3050 (20). Mullcr 6039 (20). Murillo 1457 (5); 1491 (5); 2580 (28). Murilloet al. 289 (5). Nee & Stockwell 11610 (2). Nee et al. 8768 (2). Nelson et al. 3296 (2). Nieil 2610 (9). Niemcyer 44 (28). Ocampo 727 (2); 877 (16). Oldcman 1933 (9). 011gaard et al. 34634 (9); 34703 (11); 34748 (7); 34855 (7); 34894 (11); 34970 (7); 35124 (6); 35290 (9); 35350 (11); 35776 (24); 37702 (26); 37820 (26); 37823 (3); 37859 (3); 37867 (26); 38836 (7); 38845 (28); 38932 (9); 38894 (28); 38969 (11); 39039 (11); 39040 (11); 39084 (7); 39086 (28); 43798 (9); 53824 (25). Onega 636 (9). Ortega & Grimann 2707 (5). Ortega & Smith 2387 (28); 2491 (1); 2498 (32); 2510 (1); 2511 (32). Pabst el al. 6772 (34). Perkins 1163 (28). Pcrsaud 114 (13); 372 (9). Pipoly 3542 (9); 38 17 (9); 3824 (9); 3880 (9); 3901 (9); 5146 (28); 5299 (9). Pittier 587 (19); 4492 (9); 9016 (28); 9076 (28); 10688 (9); 12416 (28); 16232 (2). Plowman & Davis 4542(13). Plowman & Thomas 13678(28). Plowman et al. 4025 (7). Pocppig 201 (9). Porter et al. 4763 (9). Prance & Silva 59677 (10). Pranceetal. 10466 (5); 12069 (9); 12180 (7); 15332 (31). Prestoe 1491 (1); 1492 (1). Proctor 3969 (28); 16506 (28); 22269 (28). Proctor et al. 27091 (9). Purpus 6761 (2); 7245 (2). Ratter et al. 2047 (10). Rcitz 151 (20); 2825 (20). Reitz& Klein 1101 (20); 1102 (20); 2665 (20); 3466 (20); 4635 (20); 4682 (20); 5108 (20); 5549 (20). Richards 803 (9). Richardson 2037 (28). Riedel 81 (35). Rimbach 31 (13); 91 (26); 312 (26). Rivero 513A (1). Riveroet al. 1608 (1). Rojas 10182 (10); 10431 (10); 10804 (10). Rose & Russell 20790 (34). Rossbach 3624 (9); 3625 (9); 3628 (28); 3710 (9). Rovirosa 972 (2). Rusby 442 (II); 443 (22). Sandwith 1561 (9). Scamman 5984 (28); 5985 (16); 5986 (16); 5987 (28);7152 (28);7153 (28);7155 (16); 7156 (9); 7482 (9); 7687 (28). Scamman & Holdridge 7998 (9). Schipp 273 (9); 8101 (28); 8108 (28). Schmalz 78 (20); 158 (20). Schnee 615 (1). Schneus 3412 (20). Schomburgk 1659 (9). Schultes & Black 46-266 (7); 8467 (9). Schultes & Cabrera 13963 (31); 16053 (11). Schunke, C. 157 (7); 158 (7); 164 (28); 165 (6); 380 (9); A214 (6); 661 (7); 666 (6); 705 (28); 812 (7); 1341 (28); 1395 (28); 1396 (6); 1451 (28). Schunke, J. 268 (7); 275 (28); 380 (28); 5785 (28); 5789 (11); 10173 (6); 10200 (26). Seavems 56 (9). Sehnem 3092 (20). Seifriz 24 (3). Shafer 4458 (28); 8889 (28). Shemluck 280 (7); 304 (16). Shcrring 13 (28). Silva et al. BG526 (5). Skogetal. 5036 (2). Skutch 3018 (28); 3236 (16); 4637 (28). 136 Illinois Natural History Survey Vol. 34. Art. I Smith, A.C. 48/293 (16); 778 (16); 2984 (31); 61713 (13). Smith, A.R. 1259 (1);4055 (1). Smith, A.R. et al. 922 (1); 1347 (28); 1388 (28). Smith, H.H. 983 (3); 1050 (28); 1052 (5). Smith & Brade 2286 (20). Smith & Klein 7546 (20). Smith & Reitz 6135 (20). Sodiro 81 (28). Soejarto 2805 (28). Sola, de la 2343 (34); 5181 (28); 5252 (16). Soukup 1065 (7). Sperling 5914 (9). Sperling & Bleiweiss 5034 (26). Sperling et al. 5915 (28). Spruce 2116 (9); 3880 (11); 4090 (6); 4337 (6); 4634 (9); 4744 (13); 4740 (II); 5685 (28). Standley 9047 (9); 24195 (9 & 28); 25085 (9); 27597 (9); 33648 (16); 37096 (28); 53955 (2); 53983 (9); 67135 (2); 68203 (2). Standley & Valerio 47127 (28); 48622 (9). Starry 91 (9). Sleeves & Ray 504 (28). Steinbach 3032 (9); 7499 (9). Stemetal. 523 (9); 1034 (16). Stevens & Krukoff 6518 (16); 8827 (9); 12056 (9); 12666 (9); 12741 (9); 13005 (9). Stewart 241 (28). Steyermark 37320 (2); 38229 (9); 38271 (9); 39197 (9); 39895 (9); 41698a (2); 41870 (2); 44675 (9); 52812 (26); 56114 (1); 61991 (1); 62015 (28); 89120 (9); 91756 (1); 94952 (28); 99185 (32); 104717 (1); 107148 (28). Steyermark & Davidse 116499 (28). Steyermark & Liesner 120634 (28). Steyermark & Nevling 95930 (28). Steyermark & Rabe 71756 (28); 96145 (28); 97133 (1). Steyermark & Stoddart 1 1 805 1 ( 1 ) . Steyermark et al. 95827a (5); 101552 (28); 106758 (1); 114331 (9); 114332 (9); 115021 (9); 115033 (9); 122814 (28); 124749 (I); 125039 (28); 126675 (1); 126680 (1). Stolze 1488 (16). Stork & Horton 9509 (9). Stubel 875 (28); 913a (13): 914 (13). Sugden 1151 (28). Swmgle et al. 70-02-05-2 (9). Sydow 339 (26). Tate 422 (9); 502 (6); 513 (6). Taylor 447 (28); 1223 (9). Tonduz 9006 (28); 9016 (28); 13337 (28); 14568 (28); 18879 (28). Tryon & Kramer 5611 (9). Tryon&Tryon 5204 (9); 5221 (6); 6591 (20). Tschudi 162 (5); 166 (32); 167 (I). Tuerckheim 148 (9): 358 (9); 7812 (9); 8040B (9); 8041 (9). Tyson 2109 (9). Underwood 1246 (28); 1250 (28); 1323 (28): 1612 (28). Valerio 63 (16): 328 (28): 329 (2); 333 (28); 2356 (16); 33133 (2). van Cotthem 1327 (1). van der Werff 3486 (32). van der Werff & Gonzalez 5202 (28). van der Werff & Wingfieid 3430 (1 ). Vareschi 3147 (28); 7764 (I). Vareschi & Gessner 1875(28). Vareschi & Pannier 1686 (28); 2636 (5): 2660 (28): 2715 (28). Vargas 11280 (6); 17743 (28); 17800 (9). von Sneidem 1590 (19). Wacket 21556 (20). Walker T10995 (28). Watt 160 (28): 7276 (28). Webb 26 (34). Webster et al. 16467 (2). Weddell 952 (35). Wcrckic 559 (16); 16770 (2). Wetmore & Woodworth 130(9). Whitmorc 752 (7). Wiggins 11073 (26). Williams 1188 (6): 17.39 (9): 2137 (9); 3136 (11); 3999 (9); 4797 (11); 7620 (ID; 7681 (9). Wilson & Murray 564 (28). Wilson & Webster 549 (28). Wingficld 6914 (28). Wright 786 (28). Wurdack 1854 (6): 1933 (11): 2011 (9). November 1987 Monograph of Polybotra 137 Index to Taxonomic Names Accepted names are set in roman type; new taxa and combinations appear in bold type; ail other names are italicized. A number after a name refers to the number of the accepted taxon . Abbreviations refer to sections of this monograph: ex (excluded names), unc (names of uncertain application), sub (subdivision of the genus). Acrostichum botryoides 1 7 caenopteris 1 var. salicifolium 5 canaliculatum Til caudatum 9 var. pubens 1 1 chrysolepis 26 cylindricum 20 fracliseriale 6 hackelianum 3 hanii I incisum 20 insigne 4 juglandifoUum 2 lechlerianum 1 3 osmundaceum 28 plumbicaule 6 polybotryoides 2 pubens I 1 suhereclum 3 Aspidium scanJens 33 Botryothallus see Polybotrya kunzei 1 Dryopteris guentheri 1 1 Olfersia caudata 9 cervina ex Polybotrya sect. Arthrobotrya ex sect. Egenolfia ex sect. Lomagramma ex sect. Teraiophyllum ex subg. Ectoneura ex subg. Egenolfia ex subg. Polybotrya sub subg. Sorbifolia sub subg. Soromanes sub acroslichoides ex acuminata unc. var. villosa 9 aequatoriana 24 alaU 23 alfredii 16 forma carpinlerae altescandens 26 andina 4 apiifolia ex appendiculaia ex appressa 25 arfakensis ex aristeguielae 28 arliculala ex aspidioides ex asplenifolia ex attenuata 14 aucuparia 2 aurita ex bifurcata ex blumeana ex botryoides 1 7 canaliculata 32 caudata 9 cervina ex cicutaria ex coenopteris I costaricensis 9 crassa 1 crassirhizoma 7 crespiana 1 1 cyathifolia 29 cylindrica 20 decorata 1 1 duplicato-serrata ex espiritosantensis 8 exallaia ex filiculifolia ex flabellala ex fractiserialis 6 fraxinifolia ex frondosa 20 fulvostrigosa unc furcata ex gaudichaudiana ex glandulosa 12 gomezii 27 goyazensis 10 gracilis 1 6 hamiltoniana ex /lar/// I helferiana ex hickeyi 2 1 incisa 20 138 Iij iNois Natural History Survey Vol. 34. Art. 1 intermedia ex juglundifolia 2 var. lobata 2 kalhreyeri 2 kunzei I latisquamosa 30 lechlcriana 13 litoralis 34 lomarioides (Blume) Kuhn ex lomarioides Mettenius unc lourteigiana 18 macbridei 7 macedoi 9 marattioides ex marginata ex monlana ex nana ex neglecia ex nieuwenhuisenii ex nodiflora ex nutans unc orientalis ex osmundacea 28 var. crispopaleacea 34 var. cylindrica 20 var. frondosa 20 var. incisa 20 peltata ex pilosa 35 pittieri 1 9 plumbicaulis 6 plumieri ex polybotryoides 2 polyphylla ex prolifera ex pteroides ex pubcns 1 1 puberulenta 22 quercifolia ex rhizophytla ex rosenstockiana 34 salicifolia 5 scandens Fee unc scandens (Raddi) Christ semipinnata 33 serratifolia 1 serrulata ex sessilisora 3 1 i/neni/j ex sorbifolia Kuhn 5 sorbifolia (L.) Keys, e; speciosa 34 stenosemioides ex stolzei 1 5 subelliptica 1 2 suberecta 3 subquinquefida ex taccaefoliii ex tenuifolia ex teysmanniana ex tomentosii 34 trapezoides unc trilobata ex ex 28 9 ex ex 33 tripartita vareschii villosula vivipara wilkesiana Polypodium adiantoides 9 Psoniiocarpa acuminata unc caudata 9 Soromanes see Polybotrya coenopteris I dentatum 1 integrifolium 1 serratifolium 1 Manuscripts of high quality dealing with any aspect of natural histon* will be considered for publication in one of the series of the Illinois Natural History Survey: Bulletin. Biological Motes. Circular, and Special Publication. The recommendations of two or more outside referees chosen by the Publications Committee of the Survey arc required before a manuscript is accepted as a Bulletin or Biological Note. Authors who arc not employees of the Survey arc required to pay pnnting costs. Manuscripts should follow the recommendations of the third edition of the Council of Biology Editors Style Manual except that the names of journals in literature cited arc to be sf)cllcd in full. The Survey expects to publish only one or two manuscripts by non-Survey authors each year. Send three copies of manuscripts to be considered for publication to Office of the Chief. Illinois Natural History Survey, 607 Kast Peabody Drive. Champaign. Illinois 61820. linois Natural History Survey 07 East Peabody :hampaign, Illinois 61820 i Division of the Illinois Department of Energy and Natural Resources