Page  5 ï~~2002 THE MICHIGAN BOTANIST BRYOPHYTES ASSOCIATED WITH HART'S TONGUE FERN (ASPLENIUM SCOLOPENDRIUM VAR. AMERICANUM) ON DOLOMITE ROCKS IN MACKINAC COUNTY, MICHIGAN Janice M. Glime Department of Biological Sciences Michigan Technological University Houghton, MI 49931 ABSTRACT Nineteen bryophyte taxa were found within the circle of influence of the hart's tongue fern. These formed zonation patterns on individual rocks, with pleurocarpous mosses such as Thuidium recognitum, Anomodon attenuatus, Brachythecium spp., and Plagiomnium affine predominating. One- and two-year-old fern sporophytes were found in disturbed patches where the thick moss mats were missing and only a thin layer of young mosses, fern and moss rhizoids, and detritus covered the rock. Older sporophytes occurred predominantly on the sides of rocks where they were supported by thick mats of mosses. INTRODUCTION Asplenium scolopendrium Linnaeus var. americanum (Fernald) Kartesz & Gandhi is federally listed as threatened in the United States, with over 90% of the individuals occurring in only two counties of central New York (Cinquemani Kuehn & Leopold 1992). Known also as Phyllitis scolopendrium (Linnaeus) Newman var. americanum Fernald, the hart's tongue fern, it grows in small populations on dolomite rocks of the Engadine series in Mackinac County, Michigan, where it is protected by U. S. Forest Service ownership. The dolomite series has both Mg and Ca, providing an essential blend of nutrients that permit plants to grow (U.S. Fish and Wildlife Service, Division of Endangered Species, 1990), whereas many other kinds of limestone rocks are lacking in some essential nutrient. The hart's tongue-bearing rocks of this series are covered with bryophytes, whereas in New York Cinquemani Kuehn and Leopold (1993) found that the mature plants more typically were on bryophyte-poor rocks. Because little is known about the ecology of the Michigan hart's tongue populations, it is desirable to understand their association with these bryophytes. It is possible that the bryophytes provide some necessary aspect to the microhabitat of these endangered ferns. And because shaded dolomite boulders are themselves an uncommon occurrence in Michigan, it is also likely that these boulders provide a unique environment for bryophytes. It is the purpose of this study to determine the bryophytes associated with the hart's tongue fern and to examine possible re

Page  6 ï~~THE MICHIGAN BOTANIST Vol. 41 lationships they might have with the fern as an indication of topics needing further study in the quest to protect these endangered plants. METHODS In July 1993, we visited two hart's tongue fern sites (Taylor Creek and East Lake, Mackinac County, MI), guided by Dorothy Evans of the U. S. Department of Agriculture, Forest Service, Escanaba, MI, and compiled a species list for the bryophytes. Our primary objective was to determine what comprised the typical communities and to examine the rocks for unusual bryophytes. At the same time, we looked for clues as to any possible role the bryophytes might play in the success of the fern, particularly in the gametophyte stage. We sampled 21 plots at East Lake and 39 plots at Taylor Creek. Whenever possible, we re-located the plots previously marked and numbered for the fern study by staff of the Forest Service, Escanaba, MI. Plots consisted of a radius of influence of ca. 15 cm. In addition, we listed any species seen but not within the circle of influence. Since the ferns were almost exclusively on the sides of boulders, mosses occurring on the tops of the boulders were outside the circle of influence, but all of these appeared at some location within the circle of influence around a hart's tongue fern. Nomenclature follows Anderson et al. (1990) for mosses, Crum (1991) for liverworts, and Flora of North America Editorial Committee (1993) for ferns. Voucher specimens of bryophytes are deposited in the cryptogamic herbarium of Michigan Technological University (MCTC) and with the U.S. Department of Agriculture, Forest Service, Escanaba, MI. RESULTS AND DISCUSSION Bryophyte associations Nineteen species of bryophytes were found in association with the hart's tongue fern on the boulders. Seventeen were present at East Lake; eleven were present at Taylor Creek (Table 1). Seven species that occurred on only one or two rocks in association with ferns at East Lake were absent from Taylor Creek despite the larger number of samples there; only two species from Taylor Creek were present in only one or two associations and absent from East Lake. Thus, nine abundant species were common to both sites. Most of the rocks were completely covered by bryophytes, with only small areas of recent disturbance that provided areas of exposed rock. Several species dominated the rocks: Anomodon attenuatus, Brachythecium rutabulum, Brachythecium sp., Plagiomnium affine, Thuidium recognitum. Only Anomodon attenuatus (12% of NY samples collected) was in common with the dominant bryophytes in the New York study (Cinquemani Kuehn & Leonard 1993), which together with Brachythecium oxycladon (Brid.) Jaeg. & Sauerb. (42%), Mnium cuspidatum Hedw. (24%), and Eurhynchium pulchellum (Hedw.) Jenn. (12%) comprised the most common NY taxa. Since all of these taxa are common in this geographic area, the overlap of only one common species suggests that conditions most likely are different. Plagiothecium laetum was common in a narrow zone at the bases of rocks, especially at Taylor Creek. Examination under the microscope revealed several taxa that were not seen in the field: Platydictya confervoides, Homalia trichomanoides, Lophozia barbata. The development of a species list was compli

Page  7 ï~~2002 THE MICHIGAN BOTANIST TABLE 1. Percent frequency of bryophyte species from circle of influence plots at East Lake (21 plots) and Taylor Creek (39 plots). East Lake Taylor Creek Anomodon attenuatus (Hedw.) Hiib. 33 33 Anomodon rostratus (Hedw.) Schimp. 5 Barbilophozia barbata (Schreb.) Loeske 10 Brachythecium shiny 71 59 Brachythecium rutabulum (Hedw.) Schimp. in B.S.G. 33 31 Cephalozia sp. 5 Cephaloziella rubella (Nees) Warnst. 5 Dicranum scoparium Hedw. 5 Fissidens adianthoides Hedw. 19 13 Homalia trichomanoides (Hedw.) Schimp in B.S.G. 5 Mnium marginatum (With.) Brid. ex P. Beauv. 10 21 Plagiochila porelloides (Torr.) Lindenb. 10 Plagiomnium affine (Bland. ex Funck) T. Kop. sensu lato 14 74 Plagiothecium laetum Schimp. in B.S.G. 19 44 Platydictya confervoides (Brid.) Crum 5 Porella platyphylla (L.) Pfeiff. 3 Rhodobryum ontariense (Kindb.) Par. in Kindb. 14 10 Thuidium recognitum (Hedw.) Lindb. 57 46 Tortella tortuosa (Hedw.) Limpr. 5 cated by the low light at the sites and the restrictions on collecting that were necessary in order to protect the fern sites for future expansion. It is likely that other small taxa, especially leafy liverworts, were not located. The bryophytes seem to exhibit a zonation pattern that is worthy of further study. The tops of the rocks and a considerable distance down the sides are covered by Brachythecium rutabulum and a second undetermined species that appears to be a different, smaller, shiny Brachythecium. On some rocks, the top is covered instead by Thuidium recognitum. Below that is a zone of Anomodon attenuatus. Near the bases of many of the rocks there is a narrow band of Plagiothecium laetum. The hart's tongue fern seems to be concentrated in the middle and bottom zones. Patchiness Fissidens adianthoides and Mnium marginatum tend to occur as invaders in disturbed patches where large clumps of the dominant pleurocarpous mosses such as Brachythecium, Thuidium, Plagiomnium, and Anomodon have been torn off. Kimmerer & Allen (1982), studying disturbed patches of sandstone cliffs, observed similar patterns for Fissidens obtusifolius where large, intertwined bryophytes had been torn off. As on the boulders, the dominant cliff species occurred in distinct vertical zones, whereas all species had a patchy distribution on open spaces that had resulted from disturbance. Fissidens obtusifolius had the greatest cover in the low, most disturbed cliff zone, whereas Conocephalum conicum was virtually absent there. Kimmerer & Allen concluded that its absence was due to severe disturbance, where large, intertwined, weakly attached patches

Page  8 ï~~THE MICHIGAN BOTANIST Vol. 41 were removed as a group. The smaller, upright Fissidens, on the other hand, had individual plants strongly attached by rhizoids and was not intermingled, therefore not losing large masses at one time. In undisturbed areas, F obtusifolius was easily overgrown by C. conicum. In more stable areas, Gymnostomum aeruginosum and Mnium marginatum occurred in the same zones as the Conocephalum, but they were not dominants. On the dolomite boulders of the present study, as with C. conicum, when any part of the larger pleurocarpous mosses is detached, it is likely to take with it a large portion of the adjoining clone. Establishment of F obtusifolius in areas of disturbance can occur within 30 days, but the percent cover drops to less than 10% within one year (Kimmerer & Allen 1982). At the dolomite boulders, it is likewise a species of Fissidens, F adianthoides, that is the primary invader of disturbed patches. Likewise, in some of these, Mnium marginatum is a common invader, occurring only where the thick moss mats had detached. The one large mat-forming species that enters early in this post-disturbance succession is Thuidium recognitum, but in its early stages of development it has a growth form not very different from that of the Fissidens or Mnium. On the sandstone cliffs, the Fissidens was not replaced with Fissidens, but rather by a group of intermediate successional species: Brachythecium rutabulum increased sixfold and Gymnostomum aeruginosum increased threefold (Kimmerer & Allen 1982). It was not clear which of the pleurocarpous mosses invaded next on the limestone boulders, but it appeared that the most likely event was invasion by the existing mosses at the periphery of the disturbance patch. Interaction of fern and bryophytes Both the young sporophytes and the mature sporophytes of the ferns were attached to the sides of the bryophyte-covered boulders. On one or two occasions we found mature sporophytes on the ground in a clump of mosses that apparently had fallen from the adjacent rock. The mosses were typical of the mosses on the rock and were generally not abundant on the soil. It does not appear that the species of bryophyte within our study sites are important to the ferns, but the stage of succession may be. We were able to find several colonies of young sporophyte fern leaves that appeared to be first-year plants, although we were unable to see any gametophytes to verify this (Fig. 1). These were concentrated very close to the bottom of the rocks, near and below mature plants. It appeared that, without exception, these had close contact with the rock surface, although in all cases some sort of primary cover existed. Often this cover was a thin (ca. 1 mm) growth of very young Thuidium recognitum, Mnium marginatum, Plagiomnium affine, and Fissidens adianthoides. In some cases, it seemed to be a thin felt of fern roots, moss rhizoids, or other brown material including detritus. When we tried to examine these young sporophytes with a hand lens, there was no way to avoid blocking our own light, thus making it nearly impossible to determine if any gametophyte tissue remained. By late July, it is likely that most of the gametophytes would already be dry and withered away. The important observation is that these young sporophytes occurred where the bulky bryophyte cover had been removed by disturbance and

Page  9 ï~~2002 THE MICHIGAN BOTANIST 2002 THE MICHIGAN BOTANIST FIGURE 1. Asplenium scolopendrium first-year sporophyte (left) and another several years old (right) with invading mosses on a felt of rhizoids and detritus left behind by a disturbed moss mat. Note the thicker moss mat at right, which is able to support the developing fern rhizome. there was access to the rock substrate. The moss mat was too thick for either gametophytes or sporophytes to succeed on the rock at the base of the mat. Although we looked on every rock that had mature ferns, we were unable to locate any young fern sporophytes or gametophytes on the thick bryophyte mats. It is possible that the hart's tongue, like Hymenophyllum in the southern Appalachians (Hicks 1979), has a life cycle in which the gametophyte has a different optimal habitat from that of the sporophyte. It appears that the hart's tongue requires disturbance that removes the thick moss mat in order for the gametophyte to become established, but that the mature sporophyte may require the support of the mosses to remain positioned on the sides of rocks. Hence, the bare patch must be of the right size for its surrounding bryophytes to grow in concert with the fern sporophyte and offer support at just the right time when the sporophyte rhizome is too heavy to cling to the rock without that support. In this remote Michigan forest site, there are few sources of disturbance. The location of disturbed patches near the bottom and on the sides of the rocks may be the result wildlife activity or of patches of heavy snow or ice that fall as a chunk, carrying with them the bryophytes frozen into the mass. The frequency of disturbed patches, however, is small, and the patches likewise are small, making it possible for the large mats to invade easily and support the heavy rhizomes as they mature. Futyma (1980) found that A. scolopendrium mature plants were typically as

Page  10 ï~~10 THE MICHIGAN BOTANIST Vol. 41 sociated with bryophytes in his Michigan plots. On the other hand, Cinquemani Kuehn and Leopold (1993) had quite different results in New York. They found that cover of mature A. scolopendrium was negatively correlated (r=-0.206, P<0.05) with bryophyte cover but that immature A. scolopendrium was positively correlated (r=0.425, P<0.05). Sporelings were predominantly on bryophyte-covered rocks (79.9%) but only 46.0% of immature and 8.8% of mature A. scolopendrium were associated with bryophytes. One possible interpretation is that where mature A. scolopendrium is present, the bryophyte cover is reduced by competition with the ferns for substrate, but Silvertown (1983) contends that competitive exclusion and niche separation are not important in limestone pavement habitats because of the occurrence of the plants "in discrete compartments which give clear limits to possible niches." Furthermore, competition would not account for the remainder of the rock being bare, as in the New York study. Rather, in the New York study it appears that crevices with organic matter were most important for mature plants, perhaps relating to nutrient value of the rocks. The differences in dominant bryophyte species between New York and Michigan may indicate the distinctions here-could it be that the New York bryophyte taxa were less conducive to fern growth, or are they merely indicative of differences in substrate chemical and physical properties that likewise affect the preference of the ferns for bare rock? From the New York study, we know that bryophytes, slope position, canopy openings, herbaceous plant cover, shrub cover, deciduous vs conifer cover, and rock crevices all are important determinants of the success of A. scolopendrium (Cinquemani Kuehn & Leopold 1993). If at the New York site the canopy cover was less and moisture conditions less than in the Michigan sites, bryophyte cover may have been needed to protect the young plants from desiccation and/or excess light, whereas the mature plants required access to crevices for both nutrients and water. Bodziarczyk (1992) found that the European variety of A. scolopendrium has a negative correlation with light intensity and suggested that forest floor herbs could play a significant role in protecting the young plants from high light intensity. By contrast, Cinquemani Kuehn and Leopold (1993) found that sporelings at their New York site had a positive correlation with light intensity, but they did suggest that during drought the sporelings with higher light intensities might be injured by the concomitant higher temperatures and evaporation. Temperature for germination, on the other hand, seems less critical, assuming that the spores germinate before the onset of summer heat. Pangua et al. (1994) determined experimentally that spores germinated equally successfully at 10, 15, 20, and 250C. Measurements of these and other microhabitat parameters for Michigan populations are clearly needed. As in any study with threatened taxa, it is difficult to experiment to find the best or necessary growing conditions without risking a negative impact on the plants you are trying to protect. The Michigan populations are too small to withstand such risk. Although several of the New York populations are sufficiently large, the conditions there are clearly different, and it is likely that the Michigan populations will prove to be genetically distinct. Based on the findings of this preliminary study, there are several observations in addition to microclimate

Page  11 ï~~2002 THE MICHIGAN BOTANIST 11 monitoring that need to be developed as testable hypotheses and that could conceivably be studied without endangering the populations: The bryophytes may exhibit a vertical zonation on the rocks that is related to light and moisture availability. New (for the Upper Peninsula of Michigan) or rare bryophyte taxa for the state of Michigan may occur on the dolomite boulders in other locations in the region. Disturbance of the bryophyte cover is likely to play a major role in the spread by spores of Asplenium scolopendrium. Bryophytes may provide the support necessary to maintain the position of the mature fern on the side of a boulder. Taxa of bryophytes associated with the fern gametophyte may be different from those associated with the sporophyte. ACKNOWLEDGMENTS Amy Linn, Janet Marr, and Tim Barbour helped with the field collections of bryophytes. The U. S. Department of Agriculture, Forest Service, Escanaba, MI supported the study. LITERATURE CITED Anderson, L. E., H. A. Crum, & W. R. Buck. 1990. List of the mosses of North America north of Mexico. Bryologist 93: 448-499. Bodziarczyk, J. 1992. The structure of selected hart's tongue, Phyllitis scolopendrium (L.) Newm. populations, as related to ecological factors. Ekologia Polska 40: 439-460. Cinquemani Kuehn, D. M. and D. J. Leopold. 1992. Long-term demography of Phyllitis scolopendrium (L.) Newm. var. americana Fern. in central New York. Bulletin of the Torrey Botanical Club 119: 65-76. Cinquemani Kuehn, D. M. and D. J. Leopold. 1993. Habitat characteristics associated with Phyllitis scolopendrium (L.) Newm. var. americana Fern. (Aspleniaceae) in central New York. Bulletin of the Torrey Botanical Club 120: 310-318. Crum, H. A. 1991. Liverworts and Hornworts of Southern Michigan. The University of Michigan Herbarium, Ann Arbor. 233 pp. Flora of North America Editorial Committee. 1993. Flora of North America North of Mexico. Vol. 2. Pteridophytes and Gymnosperms. Oxford University Press, New York. Futyma, R. P. 1980. The distribution and ecology of Phyllitis scolopendrium in Michigan. American Fern Journal 70: 81-87. Hicks, M. L. 1979. Quadrat studies of Hymenophyllaceae populations and their bryophyte associates in the southern Appalachians. ASB Bulletin 26: 80. Kimmerer, R. W. and T. F. H. Allen. 1982. The role of disturbance in the pattern of a riparian bryophyte community. American Midland Naturalist 107: 370-383. Pangua, E., S. Lindsay, and A. Dyer. 1994. Spore germination and gametophyte development in three species of Asplenium. Annals of Botany 73: 587-593. Silvertown, J. W. 1983. The distribution of plants in limestone pavement: Tests of species interaction and niche separation against null hypotheses. Journal of Ecology 71: 819-828. U.S. Fish and Wildlife Service Division of Endangered Species, Species Accounts. American hart'stongue fern. Phyllitis scolopendrium (L.) Newman variety americana Fernald. Source: Endangered and Threatened Species of the Southeastern United States (The Red Book) FWS Region 4 -As of 8/90.