sign in sign up
<<
Home , Adenophorus, Adenophorus tamariscinus, Cibotium, Cibotium menziesii, Ctenitis, Ctenitis squamigera

Pacific Science (1995), vol. 49, no. 1: 31-41 © 1995 by University of Hawai'i Press. All rights reserved

Evolution of Hawaiian and Allies in Relation to Their Conservation Status1

WARREN HERB WAGNER, JR. 2

ABSTRACT: Evolutionary and conservational differences between Hawaiian and angiosperms involve differences in histories (free-living generations, fertilization, and dispersal). Very high base chromosome numbers characterize the homosporous pteridophytes. Long-distance spore dispersal took place mainly from Old WorId and pantropical ancestors, ac­ counting for some 80% of the taxa. The ratio of native to angio­ taxa in Hawai'i averages roughly 1: 6, much higher than in continental with 1: 14. Two hundred twenty-four pteridophyte taxa, including hy­ brids and naturalizations, are known in Hawai'i. The 170 native orthospecies inC'lude endemics (highly variable taxa with polymorphies involving one or more characters, monophyletic swarms, and solitary endemics) as well as nonendemics. nothospecies compose an important additional com­ ponent, as do naturalized orthospecies. Most of the hybrids are sterile inter­ mediates that propagate by vegetative means; sexual hybrids are rare. The per­ centage ofnaturalized species is only one-fourth that of angiosperms. Hawaiian pteridophytes have evolved much more slowly than the angiosperms, as shown by lower endemism (75% versus 91 % overall and relatively fewer one- or two­ island endemics) and much smaller species swarms (average 1.5 versus 16.0 de­ scendants from each introduction in the 20 most species-rich genera, re-' spectively). Anticipated listing of Hawaiian rare and endangered fern species will probably comprise ca. 17% of the natives, including four believed to be extinct. Naturalized species compose only one-fourth of the percentage in an­ giosperms, and very few are pests. by humans and feral mammals is the major conservation problem. Although artificial spore banks and whole- culture may help save some rare pteridophytes, the most promising procedure is establishment of natural preserves.

COMPARING THE PATTERNS of Hawaiian pter­ In this essay I have tried to establish a broad idophytes with their counter­ view as a framework for more detailed in­ parts reveals numerous differences, especially vestigations. Most of the observations re­ in their relative numbers, their endemism ported here are based on intensive studies by (both overall and island-by-island), the num­ the author with the help of Florence S. ber and size of monophyletic species swarms, Wagner over the past 8 yr, together with a incidence of , and extent of hy­ number of other individuals. We have also bridization, all of which bear upon our un­ relied heavily on substantive contributions in derstanding of their evolutionary processes. particular groups, such as Elaphoglossum (Anderson and Crosby 1966), Cystopteris (Blasdell 1963), Ophioglossum (Clausen 1 Project supported by NSFB5R90, Survey and In­ 1954), Thelypteridaceae (Holttum 1977), ventory for Hawaiian Pteridophyte . Manuscript accepted 27 April 1994. (Johnson 1985), (Kato 2 Department of , The University of Michigan, 1984), Pteris (W. H. Wagner 1949, 1968), Ann Arbor, Michigan 48109-1048. and Diellia (W. H. Wagner 1952, 1953) My 31 32 PACIFIC SCIENCE, Volume 49, January 1995

Hawaiian research was actually begun in common in the Islands, viz. ar­ 1947, with the help of Lincoln Constance, buscula (Kaulf.) Spring; this is also the only Annie M. Alexander, and Harold St. John. one that is polymorphic with many varia­ The ferns and fern allies differ radically tions, some of which have been named. In from in their life history biology. North America, the percentage of hetero­ It is not surprising, therefore, that there are sporous pteridophytes is more than eight strong differences in their evolutionary pat­ times as great. Because of the low incidence terns, and this is well illustrated by their dif­ ofheterospory, the following discussion deals ferences in on oceanic islands, such only with homosporous pteridophytes, and as those of Hawai'i. Pteridophytes have a the major basis for comparison will be the distinctive alternation of independent free­ indigenous species. living generations: the familiar spore-pro­ A poorly understood feature of all homo­ ducing plants that are dominant and long­ sporous pteridophytes is the number of chro­ lived, and the tiny -producing plants mosomes (Wagner and Wagner 1980). The that are inconspicuous and short-lived. In base numbers range mainly from 20 to 70, flowering plants, the gamete-producing gen­ much higher than those in flowering plants. eration has become parasitic on the spor­ Some ofthe largest genera in Hawai'i display ophyte and only the diploid stage is free. such numbers (e.g., Asplenium with x = 36, Ferns and fern allies reproduce simply by and Dryopteris with x = 41). Polyploid spe­ wind-blown , but flowering plants re­ cies are frequent, as in the dryopteroid ferns, quire transfer by grains to the ovaries with n numbers of 41, 82, 123, and 164. The where the embryonic seed or occur importance of these high numbers is still not and where fertilization takes place on the fully understood. It has been suggested that plant. In pteridophytes, fertilization requires the polyploid base numbers arose in antiq­ merely layers of water on the substrate in or uity and -silencing effectively diploidized through which the ciliated sperm can swim the , so that the paleopolyploids are from the male organs to the female or­ genetically no more heterozygous than typi­ gans. In flowering plants the process is much cal flowering plants (Haufler 1987). Neo­ more complicated and specialized: the sperm polyploids may occur within species or be must be transferred by pollen tubes. The derived from sterile hybrids between species. whole process of in flowering The latter situation involving hybridization plants is mediated mainly by the behavior of between two known parental species produc­ volant , especially (primarily ing allopolyploids seems to be rare among and Hymenoptera) and , Hawaiian pteridophytes, unlike those of the latter especially in the Tropics. Thus there North America and temperate Asia, where are profound differences in the life cycles of such allopolyploids are frequent to common, pteridophytes and angiosperms, and these no at least in certain genera. doubt bear upon the differences in their is­ The immigration of pteridophytes to Ha­ land evolution. wai'i has been primarily due to the long-dis­ Most pteridophytes are homosporous, but tance dispersal of spores. The majority of several groups are heterosporous and have native Hawaiian pteridophytes, estimated to male and female spores, which store food be over half of the taxa, appear to have their implanted in them by the parent plants. The closest affinities in the Old World. Of the re­ native heterosporous orders represented in mainder, some 60% are pantropical, and less the Hawaiian flora are Selaginellales (spike­ than 20% are New World in their relation­ , two species), (quillworts, ships. Even though some spores could be one species), and Marsileales (water , carried among the feathers of migrating one species). Why these heterosporous pter­ birds, the overwhelming pattern has surely idophytes are so few and have evolved so lit­ been by wind because spores are readily air­ tle in Hawai'i is unexplained. Only one ofthe borne. Presumably, a majority of spores have heterosporous species is widespread and the ability to produce that can Hawaiian Ferns: Evolution and Conservation-WAGNER 33

become under certain conditions bisexual, derived from single original introductions thus making it possible to have intra­ that diverged into two or more monophyletic gametophytic mating. Many spores germi­ species; solitaries are single isolated species nate and form female gametophytes, which, presumably resulting by direct descent from if not fertilized for a long period, may pro­ original ancestors or surviving branches of duce male organs and male capable ancient swarms. The native nothospecies in­ of fertilizing eggs on the same gametophytes. clude taxa that originated by local hybrid­ Thus a single spore from thousands of kilo­ ization involving native and (or) recently in­ meters away could start a whole colony. The troduced species. Only one nothospecies is ages of the current high islands (i.e., those thus far demonstrated to have resulted from with well-developed rain ) range from hybridization that took place outside of Ha­ ca. 5.7 to 0.5 million yr old. Most native wai'i. Naturalized describes those species species occur on all the major islands, with known to have been introduced into the ar­ only a minority of approximately one in six chipelago and established mainly in the last species one- or two-island endemics. The lat­ century by Western commerce. Numbers of ter more localized species are probably either each of these groups are shown in Table 1. relicts of once widespread species, species of Patterns of in Hawaiian pter­ highly specialized habitats, or species of re­ idophytes in many respects suggest various cent arrival. underlying evolutionary processes. These The mode of long-distance dispersal by patterns assume several different forms: (a) spores together with the ability for a single Within endemic species variation-mono­ spore to become established and form a pop­ types to polytypes, taxa with a high level of ulation, plus the nondependency of pter­ genetically controlled variation in one or idophytes on animals for fertilization, prob­ more characters. (b) Within endemic ably explain in large part the high percentage -the presence of two or more re­ of ferns versus flowering plants in floras of lated sister-species obviously derived from a oceanic islands. Hawai'i is no exception; 17% single progenitor. (c) Solitary endemics­ of the species in the indigenous vascular flora isolated taxa with no near Hawaiian rela­ are pteridophytes, approximately one-sixth. tives, probably mostly evolved locally with­ According to preliminary data on oceanic out species splitting. (d) Indigenous non­ floras, provided to me by Robbin Moran of endemics, essentially lacking evolutionary the Missouri Botanical Garden, the average change-local taxa that appear to be identi­ percentage of pteridophytes of nine oceanic calor nearly identical to counterparts in island groups is 23 (range, 14-34) %. For six other geographical areas. (e) Nothospecies­ continental floras, preliminary data indicate sexually sterile or fertile derivatives of two an average of only 7 (range, 3-9) %. The data on the composition of the pter­ idophyte component of the Hawaiian flora TABLE 1 are still somewhat in a state offlux as a result TYPES AND NUMBERS OF PrERIDOPHY'I'E SPECIES of new findings in connection with an in­ INHAWAI'I tensive research program that started in 1987. Any of the numbers given below are HAWAIIAN FLORA NO. OF TAXA still subject to modification as the investiga­ tion proceeds. A distinction is made between I. Native 197 orthospecies (species that according to our A. Orthospecies 170 1. Edemics 127 current evidence are the result of cladogeny a. Swanns 48 or divergent evolution), and nothospecies b. Solitaries 79 (those resulting from hybridization or re­ 2. Nonendemics 43 ticulation). All of the orthospecies listed be­ B. Nothospecies 27 II. Nonnative naturalized 27 low are separated into endemic and non­ Total taxa 224 endemic. Species swarms are presumed to be 34 PACIFIC SCIENCE, Volume 49, January 1995

separate evolutionary lines that reticulate. (f) vergent speciation, but they could represent Recent naturalizations with no known local rejoining segregates just as well. Much of the evolution-usually more or less weedy culti­ evolution of Hawaiian pteridophytes may vars that escape and spread spontaneously, have involved divergent disjunction, local either locally or widely. differentiation, and subsequent rejunction, Hawaiian species that display extensive and both processes are occurring simulta­ polymorphy in single characters are listed in neously in different taxa and on different is­ Table 2. All of these have been treated in the lands. The polymorphies may involve in­ past as two or more species. Only by com­ cipient new speciation events caused by parative field studies of their -has mutation, selection, and/or drift involving it been possible to show that they compose one or more characters of a single island intergrading populations. The extreme forms population. Occasional island-to-island spore have differences that equal or exceed those of dispersal could allow the new genotype to many normal pairs or suites of sister-species. infiltrate the old genotypes and thus produce Perhaps some of these represent stages in di- mixtures and multiple variations. One of the most striking examples of this phenomenon involves the O'ahu plant TABLE 2 known as hillebrandii (Hook.) K. POLYMORPHIES IN SINGLE CHARACfERS APPARENTLY NOT Wilson, morphologically a very distinct ACCOMPANIED BY SPECIATION species in its , which seems to constitute one extreme and forms a series of in­ 1. Sori at laminar level to stalked: Dryopteris glabra trogressants with the other extreme, A. tri­ (Brack.) Kuntze pinnatifidus (Kaulf.) Hook. & Grev. (Figure 2. Sori dorsal to submarginal to marginal: Deparia I). We regard them as separate species. prolifera (Kaulf.) Hook. & Grev., Athyrium micro­ phyllum (Sm.) Alstor However, so common are the intermediates 3. Sori coenosoral to schizosoral: Pteris lidgatii that Bishop (1974) placed them all, including (HiUebr.) Christ, Diellia erecta Brack. the A. hillebrandii extreme itself, in a single 4. Sori indusiate to exindusiate: honolulensis polymorphous species. He stated that "rel­ (Hook.) Copel. 5. blades glabrous to pubescent, or glandular: atively uniform local populations may differ Dicranopteris linearis (N. L. Burm.) Underw., strikingly from one another." Asplenium contiguum Kaulf. The most typical A. hillebrandii extremes 6. Leaf blades pinnate to tripinnate: Asplenium macraei are on O'ahu. The islands of Kaua'i and Hook. & Grev., Diellia erecta Hawai'i show the least hillebrandii influences. 7. False veins absent to present: Polypodium pellucidum Kaulf. The differences between the extremes are given in the key below.

Stipe and lower rachis 0.8-1.0 mm thick, blade outline lanceolate, blade base little to much reduced; blade 3-pinnate, pinnule bases stalked; pinna shape lanceolate to ovate, median pinna length 3.0-4.5 cm; median pinna shape lanceolate, pinnae approx­ imate to overlapping, ultimate segments 0.7-1.1 by 0.2-0.4 mm; lamina chartaceous, green to dark green (alive). All Islands A. tripinnatifidus Stipe and lower rachis 1.0-1.3 mm thick, blade outline oblong, blade base not reduced to little reduced; blade division 2-pinnate; pinnule bases broadly adnate, basal pinna shape linear, median pinna length 4-6 mm, pinnae remote, separated by more than a pinna width; ultimate segments 1.2-2.0 by 0.4-0.7 mm; lamina coriaceous, light green to yellowish (alive). Mainly O'ahu A. hillebrandii

Forty-eight endemic Hawaiian species be­ Several of the swarms, namely Adenophorus, long to monophyletic swarms of 12 genera Oligadenus, Diellia, and , are gen­ and two to six species each (Tables 3, 4). erally considered to be endemic genera: the

us' ,1;611 f 1 Midii 88¢i@ MH'lWUJ1if1idlldib#@$$tii9#MM'M¥!WMN&M!!ijuwilfPtl'!ijffi!diSibl!Jif!Hi Hawaiian Ferns: Evolution and Conservation-WAGNER 35

FIGURE I. Single frond on left: Adenophorus tripinnatifidus, Kahokuamanui, Kaua'i, Rock 5572 (MICH). Cluster of fronds on right: A. hillebrandii, Kawailoa, O'ahu, Topping 3785 (MICH). first two related to , the third to each colonist, and the latter only 1.5 (±0.1). Asplenium, and the fourth to Blechnum. Ta­ Nearly 80 species are solitary endemics, all of ble 4 compares the Hawaiian species-rich them belonging to genera known outside of clusters of the angiosperms and the pter­ Hawai'i. These may be relicts of species now idophytes. According to these data, the for­ extinct in their original homes or derivatives mer show an average increase of 16 (±2) for ofpast introductions that have undergone all 36 PACIFIC SCIENCE, Volume 49, January 1995

TABLE 3 or most of their evolution in the Islands. It is particularly interesting to compare the in­ ENDEMIC ORTHOSPECIES SWARMS IN HAWAIIAN FERNS (CLOSELY RELATED CLUSTERS OF SPECIES HAVING cidence of polyploidy in the divergent species A SINGLE ANCESTOR) of monophyletic swarms with the solitary endemics. Only 2% of the species of swarms SPECIES SWARMS NO. OF SPECIES are polyploid, but 58% of the solitary species so far counted are polyploid (Table 5). Thus, Doryopteris decipiens (Sm.) Hook. 3 (Kaulf.) Hook. there appears to be good evidence that the & Grev. 6 presence of polyploidy has had a dampening. Oligadenus periens (L. E. Bishop) W. H. effect on the evolution of pteridophytes. Wagnerined. 3 Many of those taxa that are polyploid ap­ menziesii Hook. 4 parently lack the ability to form local deriv­ Thelypteris cyatheoides (Kaulf.) Fosberg 2 Kaulf. 5 atives. Doodia kunthiana Gaud. 2 A survey of island polyploidy that in­ Diellia falcata Brack. 5 cludes the Canaries, Madeira, Trinidad, New Deparia prolifera (Kaulf.) Hook. & Grev. 4 Zealand, Ceylon [Sri Lanka], , and Dryopteris glabra (Brack.) Kuntze 3 Tristan da Cunha gives a range from 25 to Dryopteris unidentata (Hook. & Am.) Copel. 3 65% (Walker 1984). On the basis of current Elaphoglossum alatum Gaud. 4 knowledge, Hawai'i appears to be squarely in the middle with 45%.

TABLE 4

TwENTY MOST SPECIES-RICH GENERA AND PRESUMED NUMBER OF ORIGINAL COLONISTS

FLOWERING PLANTS (W. L. Wagner et al. 1990) PTERIDOPHYTES

PRESUMED NO. OF PRESUMED NO. OF GENUS NO. OF SPECIES COLONISTS GENUS NO. OF SPECIES COLONISTS

Cyrtandra 53 4-6 Asplenium 19 19 52 I Dryopteris 15 8-II Pelea 47 I Thelypteris 13 7-II 27 I Elaphoglossum 9 9 25 3-4 Adenophorus 9 I Clermontia 22 Sadleria 6 I 22 I Diellia 6 I 21 I 5 5 20 2 (polyphyletic) Cibotium 5 I 20 Deparia 5 I 20 1-2 4 4 20 1-2 Pteris 4 4 19 I Cheilanthes 4 I 19 I Oligadenus 4 I Chamaesyce 15 1-2 Vandenboschia 4 3-4 15 I Grammitis 3 2-3 14 I Ophioglossum 3 3 13 2 Microlepia 3 2 Labelia 13 2 3 3 12 I Polystichum 3 3 Totals 469 26-32 125 79-88

l::lAlJ:iI4iii%i~~eiUm! 'iM' '#k5 ;C'jl"iiftHM MfIi'AiWti-W@FpWlgIM'!'i:a&l&S&l&Ll&&E1iiiBWW4A iM Hawaiian Ferns: Evolution and Conservation-WAGNER 37

TABLE 5 zornes that grow and branch extensively.

ENDEMIC FERN ORTHOSPECIES Among the -reproduced sterile hybrids (WITH KNOWN CHROMOSOME NUMBERS) are Pityrogramma x mckenneyi W. H. Wagner (P. austroamericana Domin x P. No. of species in the 12 species swarms 44 calomelanos [L.] Link), and the stem-re­ No. that are diploid 43 98% produced, Microlepia x adulterinum W. H. No. that are polyploid 1 2% Wagner (M speluncae [L.] T. Moore x M No. of solitary species 43 strigosa [Thunb.] Presl), Thelypteris x pal­ No. that are diploid 18 42% meri W. H. Wagner (T cyatheoides x T No. that are polyploid 25 58% Of the 25 polyploid species: dentata [Forsk.] E. St. John, Asplenium x 17 (68%) are tetraploid kokeensis W. H. Wagner (A. cookii Copel. x 4 (16%) are hexaploid A. aethiopicum [N. L. Burm.] Becherer), and 4 (16%) are octoploid Nephrolepis x medlerae W. H. Wagner (N exaltata x N multiflora [Roxb.] Jarrett ex Morton). In contrast, some hybrids appar­ ently are unable to reproduce in any way, Some 43 of the indigenous Hawaiian taxa such as x Lindsaeosoria flynnii W. H. are nonendemics, known from outside Ha­ Wagner (Lindsaea ensifolia Sw. x Odonto­ wai'i. Most likely these are more recent ar­ soria chinensis [L.] J. Sm.), and hybrids rivals than the endemics and have had little among the ferns Cibotium and Sadleria. time to evolve locally. To be sure, it will Each individual results from a separate ferti­ probably be shown that at least some ofthese lization. nonendemic natives have minute technical In addition to sexually sterile hybrids, differences from their counterparts elsewhere, there are a few allohomoploids (e.g., Diellia but in terms of ordinary taxonomic practice, falcata x unisora W. H. Wagner), allopoly­ all of these can be interpreted as conspecific ploids (e.g., Asplenium adiantum-nigrum L.), with taxa known elsewhere. A small number and apogamous x sexual hybrids (e.g., Pte­ of them are New World species, some of ris cretica L. x P. irregularis Kaulf.). In which tend to occur in drier, higher altitude terms of the provenance of the parents, the situations (e.g., East and the volcano largest number ofnothospecies is endemic x regions of the island of Hawai'i), such as endemic. But there are also other combina­ Pellaea ternifolia (Car.) Link, Asplenium tions (Table 6). Most noteworthy is the black L., Tectaria cicutaria (L.) Copel., spleenwort, Asplenium adiantum-nigrum, and Nephrolepis exaltata (L.) Schott. Marsi­ which arose as an ancient allopolyploid of lea villosa and Asplenium rhomboideum two European species and has now become Brack. are but little differentiated from M. widespread in the Northern Hemisphere, a vestita Hook. & Grev. and A. fragile Presl, remarkable example of a nothospecies the respectively. However, as noted above, the majority of nonendemic indigens appear to be Old World in affinities. TABLE 6 Among the taxa of hybrid origin (i.e., no­ thospecies), roughly 75% are sterile primary PROVENANCE OF NOTHOSPECIES PARENTS crosses. However, many of these plants, al­ though sexually sterile, are capable of form­ Endemic x Endemic 12 Endemic x Indigenous 5 ing vast populations. Some, such as sterile Indigenous x Indigenous 5 hybrids of Huperzia and , can re­ Endemic x Naturalized 1 produce by specialized gemmae, on either Indigenous x Naturalized 1 aerial or subterranean parts. A few hybrids Naturalized x Naturalized 2 Foreign x Foreign 1 apparently propagate by root proliferation, Total 27 but more commonly they reproduce by rhi- 38 PACIFIC SCIENCE, Volume 49, January 1995 parents of which now occur thousands of single-island and two-island endemics. There kilometers away. (On the basis of isozyme is a general correlation between endemism electrophoresis, Ranker et al. [1992] sug­ and the age of the island, as shown in Table gested that this species was introduced in 7. This must be taken into consideration in Hawai'i at least several different times. An conservation efforts. Unquestionably, Kaua'i alternative explanation is that the descen­ is the most vulnerable island because of its dants of a single founder have undergone numerous single-island endemics. isozyme mutations.) The paucity of known A preliminary summary of presumed ex­ allopolyploids in Hawai'i is unusual in com­ tinct, endangered, and threatened to very parison with Asia and North America. Cur­ rare Hawaiian pteridophytes is given in Ta­ rent numbers of the cytogenetic categories of ble 8. It will be noted that surprisingly few hybrid taxa now known in Hawai'i are as follows: sexually sterile, 19; allohomoploid, 4; allopolyploid, 3; sexual x apogamous, 1, TABLE 8 for a total of 27. HAWAIIAN PrERIDOPHYTE SPECIES OF SPECIAL CONSERVATION CONCERN

PROBLEMS IN CONSERVATION Extinct Botrychium subbifoliatum Brack. The conservation status of Hawaiian pter­ Deparia kaalaana (Copel.) M. Kato idophytes is considerably more favorable Diellia leucostegioides (Baker) W. H. Wagner ined. than that of the angiosperms, in terms of Diellia mannii (D. C. Eaton) W. Robinson survival of species. It will be relatively easier, Endangered therefore, to save pteridophyte species if the Doryopteris takeuchii W. H. Wagner (Hook. & Am.) Copel. proper precautions are adopted. Only about Cystopteris douglasii Hook. 12% of Hawaiian ferns and fern allies of all Diellia pallida W. H. Wagner categories have been naturalized successfully W. Robinson in the Islands, compared with ca. 45% of an­ Dryopteris podosora W. H. Wagner & Flynn giosperms (W. L. Wagner et al. 1990). The D. tenebrosa W. H. Wagner Huperzia haleakalae (Brack.) Holub overall rate of endemism among native an­ hawaiiensis W. C. Taylor & W. H. Wagner giosperms is considerably higher than among Marsilea villosa Kaulf. pteridophytes, 91% versus 75%, respectively. (HiIIebr.) W. H. Wagner ined. The majority of angiosperms are -pol­ P. nutans (Brack.) W. H. Wagner ined. Pteris (Schizostege) lidgatii (Hillebr.) Christ linated and have, therefore, greater vulner­ Thelypteris boydiae (D. C. Eaton) Iwats. ability to habitat changes. Unlike the angio­ Threatened to very rare , the pteridophytes have relatively few Adenophorus abietinus (D. C. Eaton) K. A. Wilson Adiantum capillus-veneris L. Asplenium hobdyi W. H. Wagner TABLE 7 A. rhomboideum Brack. A. schizophyllum C. Chr. ONE- AND TwO-ISLAND ENDEMICS (FIVE MAJOR ISLANDS) Cystopteris sandwicensis Brack. Diellia erecta Brack. Single island: D. unisora W. H. Wagner Kaua'i 15 (5.7)" Doodia lyonii Degener O'ahu 3 (3.6-2.6) Dryopteris parvula W. Robinson Moloka'i I (1.8) D. tetrapinnata W. H. Wagner & Hobdy Maui 7 (1.6-0.8) Gonocormus prolifer (Blume) Prantl Hawai'i 0(0.5-0.6) Lindsaea repens (Bory) Thwaites var. macraeana (Hook. & Am.) Mett. ex Kuhn Two islands: Microlepia mauiensis W. H. Wagner Kaua'i-O'ahu 4 (5.7-2.6) Oligadenus periens (L. E. Bishop) W. H. Wagner Maui-Hawai'i 6 (1.6-0.5) ined. Sadleria unisora (Baker) W. Robinson a Parentheses: estimated ages of islands in millions of .

• ;m~...... u::=w::oz:o:::t...... :li£CJ~&:::e:m::::::a:s::::s:a:tOO!W:2Z!M'l:tlUftttPhit''''Kdfi'!§+b (6 ICAN iAA~.!£iMP?!feli'&U!E!iHS!1EUi!i!!D Hawaiian Ferns: Evolution and Conservation-WAGNER 39 taxa are listed as extinct. It is conceivable vasive pteridophytes and even fewer that are that one or more of these will be redis­ serious pests in Hawaiian ecosystems. Ferns covered, in view of the recent upswing of in­ and other pteridophytes are less popular than terest in and exploration of Hawaiian flora. flowering plants for , and there (Two of the taxa previously listed as "ex­ are far fewer of them transferred to Hawai'i. tinct" were rediscovered in 1992-1993.) Pteridophytes are nonwoody and only a few Among those listed as endangered, most are of them reach any stature, most being less known from only one or two locations; oth­ than 1 m tall, except for the vining and tree­ ers occur in habitats especially liable to de­ fern species. The most abundant naturalized struction. ferns currently are Adiantum raddianum Presl. Many of the Hawaiian ferns and fern al­ (rocky, shaded, moist banks), Pityrogramma lies have probably always been more or less austroamericana (dry open fields and exposed rare and local, especially those confined to lava surfaces), Phlebodium aureum (L.) J. Sm. distinctive habitats such as rocky streambeds; ( in second-growth ), Phymato­ steep, dark, mossy banks; very wet rain for­ sorus scolopendria (N. L. Burm.) Pichi. Serm. est; and bogs. Although some species (e.g., of (abundant ground cover and climber in many Diellia) occurred primarily in (now mainly disturbed habitats), Thelypteris dentata and destroyed) dry forest, the number is not T parasitica (L.) Fosberg (abundant along large. The bulk of Hawaiian rare pter­ roadsides and paths), Blechnum occidentale idophyte species tends to occur in the middle­ L. (with the foregoing), Deparia petersenii and high-elevation rain forests at 450-1830 (Kunze) M. Kato (with Thelypteris), and m, and most of them are apparently intoler­ Nephrolepis multiflora (many habitats, espe­ of very warm ambient temperatures and cially abundant as a ruderal, also forms drought. Some seem to require nearly in­ enormous populations on exposed lava cessant mist and rain. Many are very delicate fields). Currently, we are witnessing a dra­ and sensitive, such as some ofthe grammitids matic spread of the Cyathea cooperi and filmy ferns. There is preliminary evidence (Hook. ex F. Muell.) Domin on several is­ that the montane pteridophytes have larger lands. In truly pristine forests, except along spores than those of low altitudes and dry pathways and waterways, there still remain a forests (Carlquist 1966). number of localities where practically all of The disappearance of Hawaiian pter­ the pteridophytes are natives. idophytes has been caused mainly by habitat Some of the Hawaiian endemic pter­ destruction, at first by Polynesians and later idophytes can no doubt be maintained either and much more extensively by Western as viable spores or as cultivated mature world commerce. of large plants under suitable conditions, with a view herbivorous mammals, especially feral sheep, to reintroduction as a part of habitat rec­ cattle, goats, and pigs, are demonstrably lamation. However, many, if not the major­ among the primary factors that have led to ity of pteridophytes, are difficult to maintain reduction of the pteridophyte populations. in botanical gardens. Some, such as grammi­ Also, invasive plants, such as , , tids and filmy ferns, have short-lived green guava, , Christmas , various photosynthetic spores. Others have highly members of the Melastomaceae, and many specialized habitats and associations. Many other flowering plantspecies, enter disturbed of these cannot tolerate dry conditions or areas and often take over completely. high greenhouse temperatures, and so must The first analysis of naturalized ferns in be placed in expensive controlled growth Hawai'i was published over 40 yr ago chambers. Very few botanical gardens have (Wagner 1950). Since then, many introduced either the facilities, the personnel, or the species reported there have expanded their funds to support such cultures in any major ranges and numbers considerably, and other way in the framework of present-day eco­ new arrivals have appeared. There are vari­ nonucs. ous reasons why there are relatively few in- My own conclusion, after years of study of 40 PACIFIC SCIENCE, Volume 49, January 1995 tropical ferns and of culturing ferns in green­ BLASDELL, R. F. 1963. A monographic study houses, is that the surest way of saving the of the fern genus Cystopteris. Mem. Tor­ nearly 130 Hawaiian endemics is to find ex­ rey Bot. Club 21: 1-102. actly where they live in their natural envir­ CARLQUIST, C. 1966. The biota of long-dis­ onment and to make definitive efforts to tance dispersal. III. Loss of dispersibility maintain .those areas as strictly protected in the Hawaiian flora. Brittonia 18(4): preserves. A number of such places still exist 310-335. in Hawai'i and can be preserved starting now CLAUSEN, R. T. 1954. ofthe if the public and the politicians are willing to Hawaiian Islands. Am. J. Bot. 41 :493­ cooperate with the conservationists and the 500. scientists. I am extremely impressed with the HAUFLER, C. H. 1987. Electrophoresis is following general areas: the Koke'e Alaka'i­ modifying our concepts of evolution in Wai'ale'ale region in Kaua'i, parts of the homosporous pteridophytes. Am. J. Bot. still-remaining dry forest in the Wai'anae 74: 953-966. range of O'ahu, the Pu'u Kukui and Pu'u HOLTTUM, R. E. 1977. The Thelypter­ 'Eke slopes ofWest Maui, the Waikamoi and idaceae in the Pacific and Australasia. Al­ Palikii regions of Haleakala of East Maui, lertonia 1(3): 169-234. and certain parts of the forests from KIlauea JOHNSON, D. M. 1985. Systematics of the to Glenwood on Hawai'i. Taken together, New World species of Marsilea (Marsi­ these and other still-surviving regions can leaceae). Syst. Bot. Monogr. 11: 1-:-87. probably continue to support all or most of KATO, M. 1984. A taxonomic study of the the Hawaiian endemic pteridophytes in the athyrioid fern genus Deparia with main natural living state. reference to the Pacific species. J. Fac. Sci. Univ. Tokyo 13: 375-430. RANKER, T. R., S. K. FLOYD, and P. G. TRAPP. 1992. Genetic evidence for multi­ ACKNOWLEDGMENTS ple colonizations of Asplenium adiantum­ Among many contributors to this project I nigrum onto the Hawaiian Archipelago. am especially indebted to Kenneth Kane­ Am. J. Bot. 79 (no. 6, abstracts): 119. shiro of the Hawaiian Evolutionary Biology WAGNER, W. H., JR. 1949. A reinterpretation Program, University of ; Robert of Schizostege lidgatii (Baker) Hillebr. Hobdy, Wildlife and , State of Ha­ Bull. Torrey Bot. Club 76: 444-461. waii; Flynn, National Tropical ---. 1950. Ferns naturalized in Hawaii. Botanical Garden; Joel Lau, The Occas. Pap. Bernice P. Bishop Mus. 20(8): Conservancy of Hawaii; and Daniel D. Pal­ 95-121. mer, Bishop Museum. Gerald Carr and Clif­ ---. 1952. The fern genus Diellia, its ford Smith of the University of Hawaii, and structure, affinities, and . Univ. Jane Medler and Fiona Norris of the Bishop Calif. Publ. Bot. 26: 1-212. Museum have provided institutional facilities ---.1953. An Asplenium prototype ofthe and assistance. Florence S. Wagner has par­ genus Diellia. Bull. Torrey Bot. Club ticipated in all stages of the study. 80:76-94. ---. 1968. Hybridization, taxonomy, and evolution. Pages 113-138 in V. H. Hey­ , ed. Modem methods in plant tax­ LITERATURE CITED onomy. Academic Press, New York. ANDERSON, W. R., and M. R. CROSBY. 1966. WAGNER, W. H., JR., and F. S. WAGNER. A revision of the Hawaiian species of 1980. Polyploidy in pteridophytes. Pages Elaphoglossum. Brittonia 18: 380-397. 199-213 in W. H. Lewis, ed. Polyploidy: BISHOP, L. E. 1974. Revision of the genus Biological relevance. Plenum, New York. Adenophorus (Grammitidaceae). Brittonia WAGNER, W. L., D. H. HERBST, and S. H. 26: 217-240. SOHMER. 1990. Manual of the flowering

CIJiRtIf'lW!2f2i' I i 2+' Wi *eeeu @i$$/d waRE'" Hawaiian Ferns: Evolution and Conservation-WAGNER 41

plants of Hawai'i. University of Hawaii 141 in A. K. Sharma and A. Sharma, eds. Press and Bishop Museum Press, Hono­ Chromosomes in evolution of eukaryotic lulu. 2 vols. groups. Vol. 2. CRC Press, Boca Raton, WALKER, T. G. 1984. Chromosomes and . evolution in pteridophytes:. Pages 103-