‘Dovova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 94 January 1992 No. 877 Che New England Botanical Club, Inc. 22 Divinity Avenue, Cambridge, Massachusetts 02138 RAGDORA NORTON H. NICKERSON, Editor-in-Chief Associate Editors DAVID S. BARRINGTON RICHARD A. FRALICK A. LINN BOGLE GERALD J. GASTONY DAVID E. BOUFFORD C. BARRE HELLQUIST CHRISTOPHER S. CAMPBELL MICHAEL W. LEFOR WILLIAM D. COUNTRYMAN ROBERT T. WILCE GARRETT E. CROW RHODORA (ISSN 0035-4902). Published four times a year (January April, July, and October) by The New England Botanical Club, > 23 Divinity Ave., Cambridge, MA 02138 and printed by Allen Press, Inc., 1041 New Hampshire St., Lawrence, KS 66044. Second class postage = at — MA and at additional ouing offices. and cytotaxonomy in their broader implications, will be consid- eed. A Articles are subjected to peer review. RHODORA assesses page charges SUBSCRIPTIONS: $45.00 per calendar year, net, postpaid, in funds payable at par in the United States currency at Boston. Remittances payable to RHODORA. Send to Treasurer, 22 Divinity Ave., Cam- bridge, MA 02138. ee Regular $35; Family $45. Application form printed BACK VOLUMES AND SINGLE COPIES: Some available; infor- mation and prices will be furnished upon request to the Treasurer. ADDRESS CHANGES: In order to receive the next number of RHO- DORA, changes must be received by en Treasurer prior to the first day of January, April, July, or Octobe POSTMASTER: Send address Py to RHODORA, 22 Divinity Ave., Cambridge, MA 02138. Beye ge FOR CONTRIBUTORS: Inside back cover, January MANUSCRIPTS: Send to: Joan Y. Nickerson Managing Editor, RHODORA Phippen-LaCroix Herbarium Dept. of sore S ae University Medford, MA 0 Cov Rhexia virginica L., meadow Pere: is found from Nova Scotia to Georgia, but is rare at the northern limits of its range. The < northern outlier of the : sale tRbhodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 94 January 1992 No. 877 RHODORA, Vol. 94, No. 877, pp. 1-14, 1992 A NEW SPECIES OF BLEPHILIA (LAMIACEAE) FROM NORTHERN ALABAMA RICHARD W. SIMMERS AND ROBERT KRAL ABSTRACT Blephilia subnuda Simmers & Kral is described from the Cumberland Plateau province of northern Alabama. This species is compared with other species and forms of Blephilia, from all of ony it differs i in its puberulent calyx tube. — B. ciliata it differs primarily also glabrate below the middle, leaves aan calyx mostly shorter, and corollas ni in ground color (white to pale purple). From glabrate forms of B. hirsuta it differs primarily in stem vesture, leaves with fewer trichomes, calyx with longer, puber- ulent tube and longer teeth of lower lip, and darker nutlet color. Key Words: Blephilia, new species, northern Alabama INTRODUCTION Blephilia Rafinesque has been considered by all recent authors to comprise two somewhat variable species, namely B. ciliata (L.) Benth. (Monarda ciliata L., Sp. Pl. 23, 1753) and B. hirsuta (Pursh) Benth., both confined to eastern North America. The former has its morphs most often in open areas, open woods and glades. In strong contrast is the other previously described species, B. hir- suta, which is primarily a mesic-woodland plant, taller (usually over 1 m), with softer, stouter stems, larger, thinner, smoother and sharper-tipped leaves, paler and softer inflorescences, shorter calyces, smaller and paler corollas and larger resin dots. The two have a broad sympatric range, namely from Vermont south to northern Georgia, west to southern Ontario, lowa and Missouri. Blephilia ciliata ranges further south, into the Coastal Plain of Alabama, Mississippi, Arkansas and Oklahoma, while B. hirsuta 1 MISSOURI BOTANICAL APR 06 1992 @annen LIRRARY Z Rhodora [Vol. 94 extends further north, into Minnesota and southwestern Quebec, as well as occupying higher elevations in the southern Appala- chians. Thus we note that these two blephilias share much geographic range but overlap little in habitat; however, the two species do at times share some habitat. We have identified putative hybrids, ostensibly allowing genetic exchange by backcrossing. But even working within such a context and even when such intermediates have been identified, we have come across plants of still another and distinctive morphology that cannot be explained in terms of character states either shared or distinctive to one or the other. This taxon, discovered by the senior author in 1972 and located in the rugged, calcareous, wooded terrain of the Cumberland Pla- teau of northeastern Alabama, we are proposing as a new species. Blephilia subnuda Simmers & Kral, sp. nov. Haec species habitu cum Blephilia ciliata (L.) Benth. optime congruens sed differt stolonibus per florescentiam nullis, caulibus inferne glabratis, superne t parte maxima in duo- bus lateribus sulcatis supra petiolos foliorum praeditis, laminis subtus glabratis, tubis calycum externe puberulis (trichomatibus usque ad 0.6 mm longis, pro parte maxima 0.03-0.10 mm longis). Strict perennial herb (25-) 30-60 cm high, perennating by de- cumbent offshoots to 25 cm long, these produced only after an- thesis. Sterile and fertile stems mixed in lax clusters, the stems at anthesis ca. 1-3 mm thick, the upper internodes above petioles sulcate, the branching (if present) ascending; stem surfaces com- monly greenish to deep purple, glabrous or sparsely and minutely puberulent to glandular-puberulent, also with frequent tiny resin- dots from base to midstem, then from ca. midstem to tip of inflorescence, sparsely to densely mixed-trichomiferous, the tri- chomes spreading to recurved, even appressed, unicellular to mul- ticellular, the longest on nodes and stem angles. Leaves spreading or slightly ascending, in the inflorescence sometimes reflexed (mostly spreading), the petioles (0.25-) 0.5-2.0 (—2.5) cm long, ciliate, puberulent; blades broadly ovate to narrowly oblong or lanciform, broadest in outline at stem base, largest at ca. midstem, progressively smaller, narrower, subsessile or sessile upstem into the inflorescence, the largest ones (2.6—) 3.6-10 (-11.9) cm long, 1992] Simmers and Kral—Blephilia 3 (0.9-) 1.2—4.0 (-5.0) cm wide; apices acute to acuminate; margins serrate to crenate-serrulate or subentire; bases round to broadly cuneate, subtruncate or attenuate, often oblique; adaxial surfaces deep green, glabrate to sparsely scabridulous; abaxial surfaces paler, glabrate or sparsely unicellular-puberulent on larger nerves; both sides moderately dotted with small, flattened, sessile glands ca. 0.08 mm wide, and puberulent to glabrate along edges. Flower clusters compact, 1-4, globose to hemispheric, 20-1 20-flowered, involucrate, 0.9-2.5 cm wide (at calyx level), the lower ones most distant and largest; involucral leaves green, sometimes purplish; involucral bracts numerous, loosely imbricate, the outer ones ovate to lanceolate, 7-16 mm long, 1.4-7.5 mm wide, acuminate, hirsute-ciliate, spreading in anthesis, pale green to carmine-pur- ple, the inner ones grading to linear; pedicels 0.4-2.5 (-3.5) mm long, glabrate; calyx in anthesis slightly excurved, 5.0-8.6 mm long, the tube cylindric, slightly campanulate or flaring, (2.5-) 3.2-4.8 (-5.3) mm long, distally with a transverse hispid annulus within, externally sparsely puberulent with trichomes mostly 0.03- 0.10 mm long, a few up to 0.6 mm long, sparsely glandular- punctate; upper lip 0.7-1.7 mm long, the subulate teeth 1.2-3.2 mm long, distally setose, the setae (3-) 5-10, 0.6-2.0 (-2.5) mm long, the lower lip shorter, the teeth similar but 1.3-2.8 mm long, sparsely setose; corolla externally villous, (7.5) 8.8-12.0 mm long, the upper lip linear, 1.4-1.6 mm long, the lower lip broader, longer, 3.2-5.8 mm long, 2.44.2 mm wide (both densely pus- tulate), externally white or light lilac to mauve, internally irreg- ularly purple-maculate on the lower lip. Fertile stamens 2, the filaments 5.5-8.4 mm long, the anthers versatile, slightly exserted, rose-purple and white to light yellow, ca. 0.8-1.2 mm long, the pollen white, hexacolpate, strongly suggesting a miniature canta- loupe in form and sculpture, ca. 0.040 mm long and 0.034 mm wide. Nutlets 4, commonly ovoid, 0.7-0.9 mm long, 0.5-0.7 mm thick, uniformly black to brown-spotted or brownish, shallowly alveolate. Chromosome number unknown. Type: United States. Alabama. Madison Co.: ca. 8 km NE of New Market; T1S, R3E, Sect. 18 SW'4, elev. 900-1100 ft., frequent on rich N-facing slopes over limestone, cove hardwoods S of Mountain Fork Creek, 1 June 1983, R. W. Simmers 3423 (HOLOTYPE: GH; ISOTYPES: AUA, BH, BM, F, GH, K, MO, NCU, NY, PH, RSA, SMU, TENN, UNA, US, VDB). 4 Rhodora [Vol. 94 ADDITIONAL SPECIMENS EXAMINED: U.S.A. Alabama. Jackson Co.: 6 mi. NW of Swain on AL HW 65, SW-facing limestone slope, cove hardwoods, 4 Jun 1975, Massey & Whetstone 4601 (vps); just W of dirt road along Hurricane Creek ca. 2 mi. S of Walls of Jericho, Hytop Quad., cove Sasi ira over limestone, 19 May 1979, Simmers 3256 (BH, GH, K, SMU, UNA, US, VDB); above Paint Rock Creek ca. 5 mi. N of Princeton, sndrock-oulder-stewn ‘topes over limestone, wooded bluffs, 7 Jun 1982, Kral 68691 (GH, MO, NY, TENN, UNA, US, vpB); along AL 65 on NW side of Round Mt., recently tes steep, rich woods over limestone, occasional, T1S, R4E, Sect 17 NE, 1 Jun 1983, Simmers 3424 (vps). Madison .. Northern end of Logan Point, north of Monte Sano, alt. 430 m, 31 May 1972, Clausen 72-16 (GH); ca. 8 km NE of New Market, T1S, R3E, Sect 18 SW'4, el. 900-1100 ft., frequent on rich N-facing slopes over limestone, cove hardwoods S of Mountain Fork Creek, 28 Jul 1972, Simmers 2650 (GH, vpB), 17-18 May 1976, Simmers 2963 (vps); ca. 8 km NE of New Market, T1S, R3E, Sect 18 SW'4, frequent on alluvial terrace on S side of Mountain Fork Creek, 27 May 1978, Simmers 3181 (AUA, BH, F, GH, K, MO, NY, PH, RSA, SMU, US, VDB); W face Huntsville Mtn., at end of Deborah St., E. side Huntsville, limerocky, hardwood-forested Cotinus site, 17 Jun 1983, Kral 70174 (GH, Mo, US, vpB); Green Mountain, ca. 1300 ft. a.s.1., without date, Cindy Drake, s.n. (vDB) DISCUSSION Blephilia subnuda is compared with the other two blephilias in 1. It stands nearest to B. ciliata in features of height, branch- ing, calyx limb (relative length of teeth), bract ciliae, nutlet color and shape, and phenology. However, B. subnuda has longer pet- ioles, thinner, smoother, sharper-tipped leaf blades, and smaller, paler corollas, and thus approaches B. hirsuta in these characters. As is true of many genera of Lamiaceae, the character and amount of vesture is given considerable weight, as are pigments. In B. subnuda, the most striking characters are in the smooth or vir- tually smooth lower and middle internodes, the smooth or nearly smooth leaf surfaces, the puberulent calyx-tube (trichomes over 0.6 mm long are lacking) and the smooth, decumbent offshoots produced after anthesis. It is in degree and character of the trichomes of the stems, leaves and calyx, and to a lesser degree pigmentation, that the three species are best distinguished. Our observations are as fol- lows: In Blephilia ciliata, the most common and most variable spe- cies, the slender stems rarely range higher than 8 dm. In the sun forms (the most common situation), stems are densely pale-re- curved-puberulent proximally, on most plants densely so on su- prapetiolar faces. This puberulence (Figure 1) may be with or 1992] Simmers and Kral—Blephilia 5 Table 1. Character states in Blephilia taxa. B. hirsuta Character B. ciliata B. subnuda (typical) Offshoots Stolons by anthe- | Decumbent Erect to lax i shoots after shoots by au- anthesis tumn eight 28-88 cm 23-56 cm 50-135 cm Branching Rarely, short Rarely, short Usually several Stem pubescence Petiole length Blade shape Apex of upper blades Base of upper lades Pubescence, ab- axial side blades Calyx-tube length Calyx-tube pu- bescence Length of upper calyx teeth Length of lower calyx teeth Corolla ground color Time of anthesis ? Nutlet color (usually under cm) Dense, short, usually retrose 1-12 mm Oblong to elliptic to ovate Blunt Often cuneate Usually + dense- may some pilosity also 4.0-8.4 mm Long trichomes numerous, mostly slender, m 0.4-3.3 mm 1.2-3.5 mm Shades of purple or blue May-June Black when present Glabrate below middle, + re- trorse above 4-22 mm Variable, elliptic + acute Variable, usually obtuse Glabrate, a few tiny unicellular trichomes on midvein 2.5—5.3 mm Puberulent, tri- chomes under 6mm 1.2-3.2 mm 1.3-2.8 mm White or pale purple (lilac) Mid-May-Mid June Black, black with -brown ar- eas, or brown branches over 15cm Hirsute, spread- ing trichomes often 1-2 mm 9-42 mm Ovate (to lanceo- late) Acuminate to Usually obtuse Pilose on larger nerves; tiny fine trichomes sional 2.2-3.6 mm Long trichomes frequent distal- ly, + twi when dry 1.0-2.8 mm 0.5-1.3 mm White July—October (variants earli- Tan to reddish- brown [Vol. 94 IR Nae y: ih "i i" a Figures 1-4. Stems (middle to upper internodes) of Blephilia. 1-2. B. ciliata from large population near Demopolis, Marengo Co., Alabama (Simmers 4 2952); pilose extreme, plant e; usual pubescence, plant a. 3. B. subnuda from type locality NE of New Market, Madison Co., Alabama (Simmers 3181). 4. B. hirsuta from Fletcher’s Hollow, Marshall Co., Alabama (Simmers 3313). Scale bars = | mm 1992] Simmers and Kral—Blephilia 7 Figures 5-8. Undersides of cauline leaf-blades of Blephilia. 5. B. ciliata from Highland Co., Virginia (Simmers 3189-e). 6. B. ciliata from Tuscaloosa Co., Alabama (Simmers 2953-e). Figures 7-8. B. subnuda from type locality, Madison Co., Alabama (Simmers 3197). Figure 7. Typical leaf, 8. Maximum density of trichomes on midvein. Scale bars = 0.1 mm. 8 Rhodora [Vol. 94 without a mingling of spreading long trichomes, particularly along stem angles, less often with such long trichomes predominant. Toward midstem a uniform puberulence may persist (Figure 2), or the vesture may consist solely of scattered or dense pilosity. Upstem, typically the pubescence becomes denser, of interme- diate length, is recurved and often tomentose. The adaxial leaf blade surface is typically smoothish with scattered resin dots, the abaxial surface with short, usually soft, erect or curved puberu- lence, sometimes in addition with scattered pilose trichomes on the larger nerves (Figure 5). Both stems and leaves in these sun forms are generally more purple-pigmented than in the other two species. In shade forms there are extremes in which stem pubes- cence may be almost uniformly scattered-pilose and the thinner leaves scattered-pilose on both surfaces, rarely nearly smooth save for puberulence on the nerves beneath (Figure 6). In typical Blephilia hirsuta, a strong plant usually exceeds | meter; the stems are thicker and softer. Stem pubescence is more uniformly long, spreading, and often villous (Figure 4); it ranges over the lower and median internodes from sparse to copious, particularly along angles and at nodes. Upstem pubescence be- comes thicker and is often tomentose. In addition, the stems have frequent recurved to appressed puberulence and occasional glands. The principal leaf-blades of B. hirsuta, the largest in the genus, often are totally smooth between veins with scattered, often crisped pilosity confined to larger veins above (Figure 9), the sessile glands scattered, wider, more depressed, and paler. There are also oc- casional fine, possibly glandular trichomes present on the larger veins (Figures 9, 11), as well as puberulence. The calyx (Figures 14, 17) and involucral bracts have longer, softer, more villous pubescence than in other blephilias. Blephilia subnuda, as noted in Table 1 and in the description, is the smoothest blephilia; the lower and often midstem inter- nodes are totally smooth or with scattered recurved puberulence and have tiny resin-dots only. Upstem the pubescence becomes — Figures 9-11. Undersides of cauline leaf blades of B. hirsuta. 9. from Varna, Tompkins Co., New York (Simmers 2992-5). 10. from Little River trail near Elkmont, Great Smoky Mountains National Park, Sevier Co., Tennessee (Patrick & Simmers 3186-b). 11. from Fletcher’s Hollow, Marshall Co., Alabama (Simmers 3225). Scale bars = 0.1 mm. 1992] Simmers and Kral—Blephilia 10 Rhodora [Vol. 94 moderately dense on angles and faces above the petioles (Figure 3); the longer trichomes (0.5—1.5 mm long) are concentrated on the angles, the shorter ones are more numerous in the grooves. Adaxial leaf surfaces are often totally glabrous other than for glands or a few scattered trichomes on larger veins (Figure 8). The calyx tubes (Figures 13, 16) are puberulent with mainly uni- cellular trichomes 0.03-0.10 mm long with a few longer tri- chomes, especially at the orifice. The teeth of the upper calyx lip often have more numerous (5-10) setae at or close to their tips than are found in other Blephilia. Occasional populations of Blephilia, especially in the Great Smoky Mountains of Tennessee (e.g., Patrick & Simmers 3186), which have stems glabrate proximally and have glabrate leaves nonetheless differ significantly from B. subnuda, particularly in ways that align them more with B. hirsuta. Such plants in summer or autumn produce strongly pubescent offshoot stems (field- checked by the senior author in November, 1979), and have a more uniform pubescence on upper parts of flowering/fruiting stems. They have relatively short petioles (3-15 mm) and the blades of upper cauline leaves are ovate to lanceolate, are acu- minate with the larger nerves (midvein at least) abaxially rather densely puberulent (Figure 10), much more so than in B. subnuda (compare with Figure 8). Their nutlets are fuscous (dark brown), much darker than is true for typical B. hirsuta, in which they are tan to reddish-brown. These Smoky Mountain plants are consid- ered here to be closest to B. hirsuta; possibly they have been introgressed or mixed with genes from B. ciliata, as is strongly Suggested by the character states of petiole length, nutlet color, and relatively early phenology (often being in anthesis in June). Fernald (see Day, 1899, p. 221, fn.) described Blephilia hirsuta var. glabrata from specimens collected by M. A. Day from a population on Mt. Equinox in Vermont (GH! NEBC!). Unfor- tunately, Fernald, who cited Day 140 & 141, did not indicate which of these he considered typical! The lot is variable, suggesting — Figures 12-14. Calyces of Blephilia. 12. B. ciliata from Tuscaloosa Co., Ala- bama (Simmers 2953-e). 13. B. subnuda from type locality NE of New Market, Madison Co., Alabama (Simmers 3181). 14. B. hirsuta from Fletcher’s Hollow, Grant, Marshall Co., Alabama (Simmers 3225-g). Scale bars = 1 mm. Simmers and Kral—Blephilia Rhodora [Vol. 94 VR) dovetl 1992] Simmers and Kral—Blephilia 13 a hybrid swarm, but none of them is much like the Alabama plants here named B. subnuda. One of the Day collections (Day 141) is nearly identical to the above-described Smoky Mountains plants, particularly as to the strongly puberulent larger nerves on its cauline leaves. To summarize, we have a species that combines several of the salient characters of each of the two previously described Blephilia but which adds some distinctive features of its own, particularly a unique vesture (especially of the calyx). We consider that it is most likely that B. subnuda represents a stabilized hybrid. Fea- tures of vesture are an indication that selection of these characters has occurred since the initial hybridization event. A recent col- lection (Simmers 1990-7) from a population in DeKalb County, Tennessee is intermediate between B. ciliata and B. hirsuta but is unlike B. subnuda primarily in being densely pilose on stem and (usually) leaf: its calyx-tubes have longer trichomes than do those of B. subnuda. The corolla ground color of these DeKalb County plants is fairly uniformly pale violet. A few specimens close to typical B. ciliata have been observed and collected (Sim- mers 1990-8) on the fringe of this population, but are outside our concept of B. subnuda. Blephilia subnuda has been found thus far only in two counties of northeastern Alabama north of the Tennessee River on the Mississippian limestones of the dissected Cumberland Plateau. In this region it often associates with rarities such as Neviusia alabamensis A. Gray and Viburnum bracteatum Rehder. Within its known range it is the most abundant Blephilia; it is locally frequent to fairly abundant, usually on shaded, moist outcrops of limestone or on slightly disturbed, fairly shaded sites including alluvial terraces. We have not found B. ciliata within the range of B. subnuda, but B. ciliata has been collected just outside it in Huntsville, Madison Co. (Jada Leo, s.n., vDB). Populations of Blephilia closest to B. hirsuta but having reduced pubescence occur within the range of B. subnuda in Jackson Co., Alabama (Kral 47572, 47574, vps; Simmers 3470 & 3741, vbB, etc.); these Figures 15-17. Calyces of Blephilia. 15. B. ciliata from near Demopolis, Ma- rengo Co., Alabama (Simmers 3255-n). 16. B. subnuda from type locality NE of New Market, Madison Co., Alabama (Simmers 3181). 17. B. hirsuta from Teal Hollow, Lincoln Co., Tennessee (Simmers 3283-b). Scale bars = 0.1 mm. 14 Rhodora [Vol. 94 populations were not observed within 1 km of known populations of B. subnuda. The upland region where this endemic grows has a higher rainfall than do the surrounding areas. Therefore it is possible that the greatly reduced pubescence of B. subnuda is an adaptation to local environmental conditions. ACKNOWLEDGMENTS We thank Thomas S. Patrick for his time spent on the Etec Autoscan SEM, and for his help in collecting in the Great Smoky Mountains National Park. Thanks are also extended to Gordon Hunter and various staff at Tennessee Technological University, and to B. Eugene Wofford and others at TENN for their help with both examination and storage of study material. We are grateful to Carroll Wood, Michael Canoso, and others at Harvard Uni- versity Herbaria for their assistance and encouragement. Con- structive comments of the anonymous reviewers and the editors are sincerely appreciated. R.W.S. 8187 MACEDONIA RD. COOKEVILLE, TN 38501 HERBARIUM, BOX 1705, STA. B VANDERBILT UNIVERSITY NASHVILLE, TN 37235 RHODORA, Vol. 94, No. 877, pp. 15-37, 1992 CONTRIBUTIONS TO THE ALPINE FLORA OF THE NORTHEASTERN UNITED STATES PETER F. ZIKA! ABSTRACT Alpine and sub-alpine cliff floras have changed since botanists first explored the mountaintops of the northeastern United States; results of recent extensive 634 3 ele + be Bee ck - 1 datak + 4 ++h = 11010) Sur y v New distribution and population data are presented for 27 rare northern taxa in Vermont, New York, New Hampshire and Maine. Paronychia argyrocoma iS reported for the first time from Vermont. Eleven historical records for rare species are dismissed as errors. Most newly reported rare plant locations appear to be previously unexplored areas and do not represent recent dispersal. Population declines are indicated for Arenaria rubella in Vermont and Luzula spicata in New York, although the populations are in remote and pristine sites. Local extirpations are presumed in Vermont for Arnica lanceolata, Castilleja septentrionalis, Em- petrum eamesii ssp. atropurpureum, Geocaulon lividum, and Solidago cutleri, and in New York for Cassiope hypnoides. Key Words: Alpine flora, endangered species, extirpation, Maine, New Hamp- shire, Vermont, New York Over the last three decades, recreational use of mountaintops in the northeast has reached all-time highs (Waterman and Wa- terman, 1989). Widespread soil compaction and trail erosion threatens rare members of the alpine flora. Air pollution and climatic warming are also potential threats. As a result, many alpine species are on state lists of rare, threatened or endangered plants (Countryman, 1978; Thompson, 1989: Storks and Crow, 1978; Mitchell et al., 1980; Zika, 1990a; Dibble et al., 1989, 1990). Mountain species in these references are the subject of this in- vestigation. e study was initiated when Draba cana Rydb. and Carex atratiformis Britt. were re-located in Smugglers Notch, after a 100-year hiatus in their collecting record. Had other high-ele- vation species been overlooked by recent workers? Countryman (1978) suspected the “rare” grasses of Vermont were undercol- lected, and this suspicion was confirmed for many species (Zika, 1990b). In this paper, for alpine and subalpine species I compare ! Current address: Oregon Natural Heritage Program, 1205 NW 25th, Portland, OR 97210. 15 16 Rhodora [Vol. 94 some historical and contemporary data, provide evidence sug- gesting i or declines in several taxa, reject eleven historical records, discuss one new state record, and record relevant pop- ulation and ecological data for new stations and re-locations of some state rare, threatened or endangered species. Distributions of most mountain species in the northeastern United States were well-known by the turn of the century (Pringle, 1876, 1897; Peck, 1899, 1900; Brainerd et al., 1900; Fernald, 1901, 1907; Kennedy, 1904; Pease, 1964). Collecting activity in alpine areas peaked between the 1870’s and World War I, a period of intense professional and amateur collecting for herbaria. The results were thousands of pressed specimens and ample literature. METHODS Literature and herbarium searches provided a historical frame- work and a list of sites for intensive field surveys. Current pop- ulation data were generated by repeatedly visiting classic collect- ing localities at Smugglers Notch (Lamoille Co., VT), Mt. Mansfield (Chittenden and Lamoille Cos., VT), Camels Hump (Chittenden and Washington Cos., VT), Mt. Pisgah (Orleans Co., VT), the Presidential Range including Mt. Washington (Coos Co., NH), and in New York’s Essex Co.: Mt. Marcy, Panther Gorge, Indian Pass and Wallface Mountain. Other peaks were investigated, but less frequently. Field work was conducted between 1979 and 1990. All species reported as locally extirpated in New York or Ver- mont were observed by the writer on Mt. Katahdin, Mt. Wash- ington, or elsewhere, to learn the proper “search image,” microsite requirements and associates. RESULTS Species are listed alphabetically; vernacular names follow Har- ris et al. (1964). Agrostis mertensii Vermont’s fourth station for Agrostis mertensii Trin., boreal bentgrass, was recorded in 1989, on the ledgy schist summit of Mt. Hunger, Worcester, Washington Co., at 3490 feet (Zika & Dann 10662 vt). About 20 plants were growing in turf dominated 1992] Zika— Alpine Flora 17 by Deschampsia flexuosa (L.) Trin., with Carex brunnescens (Pers.) Poir. ex Lam. The colony was limited to a few square meters, and no plant was found on the adjacent exposed summit of White Rocks Mountain. Arenaria rubella Arenaria rubella (Wahlenb.) Sm., marble-sandwort, is known from only two locations in the eastern United States. Historical data for the Vermont population, in Smugglers Notch, include a west-side 1894 collection (NEBC, vT) labeled “rare,” and an 1894 collection (vt) from north-facing cliffs in another west-side ravine, labeled “in abundance.” East-side collections end in 1932 (MAss). A plant was collected on the roadside in 1926 (Mass). Arenaria rubella can no longer be described as common or abundant in Smugglers Notch; its distribution and abundance have declined. Only two ravines currently support the species. In the last decade the population was restricted to two habitats: steep gravelly slopes, and inaccessible shelves on the tall cliffs above. The population fluctuated from 5-50. When plants were scarce most seedling establishment was observed on steep unconsoli- dated substrates, in mats of Saxifraga oppositifolia L. Following landslides, populations increase with recruitment on newly ex- posed gravels. The high slopes of Smugglers Notch landslides are seldom vis- ited by humans and are protected as part of the Mt. Mansfield State Forest. Portions are closed during the Falco peregrinus nest- ing season. Arnica lanceolata Arnica lanceolata Nutt., hairy arnica, is a showy subalpine and alpine species. Schweinfurth and St. John (St. John, 1987) col- lected it once in Vermont, in 1911, on the northern spur of Mt. Mansfield above Smugglers Notch. It has not been seen again, despite repeated searches, and is presumed extirpated in Vermont. The forested historical locality is remote and undisturbed. In New York A. lanceolata shows an interesting historical dis- tribution along the northern of the two Indian Pass Brooks. It was collected in five sites between 1899 and 1989, along 2.5 km of the stream. The population appears to episodically colonize 18 Rhodora [Vol. 94 anorthosite shores when conditions are favorable for seed dis- persal and seedling establishment. Presumably seeds are dispersed by the brook, by wind or by animals traveling along the riparian corridor. The low elevation colonies apparently did not persist. Peck (1899) could not re-locate his original streamside station. Re- peated searches in 1988 and 1989 by Steve Clemants, Rose Paul and the writer determined that the uppermost colony was the only one extant. This location was a wet exposed NE-facing cliff above the brook, and may have been a stable Arnica habitat and hence a reliable seed source for the lower sites, which are all forested at present. Arnica lanceolata (Zika 10707 nys) was associated with Houstonia caerulea L., Thalictrum pubescens Pursh, Aster um- bellatus Mill., Carex debilis Michx., Spiraea latifolia (Ait.) Borkh., Vaccinium uliginosum L. and Lycopodium selago L. One wonders what led to the loss of Arnica sites on Indian Pass Brook. Logging dams used 1913-1915 might have altered stream- bank habitats by changing the seasonality and volume of flow. Arnica requires openings; an important factor may have been declining light levels concomitant with shoreline forest regrowth following logging in 1913-1915. Beaver activity or siltation fol- lowing logging are possible explanations; Arnica clones may be short-lived. Combinations of factors are possible as well. Asplenium viride The station of Asplenium viride Huds. credited to Camels Hump, Vermont “/egit Pringle,’ (Brainerd et al., 1900; Eggleston, 1905; Eggleston et al., 1915) must be discounted. Pringle apparently never collected the species there, nor mentioned it in his publi- cations. Pringle was a fern enthusiast, eagerly collecting the un- usual species he discovered in the 1870’s in Vermont (Anony- mous, 1877; Davis, 1936). Surely he would have collected and published such a find from Camels Hump; he did so for his 1876 discovery of Asplenium viride at Smugglers Notch and Mt. Mans- field, as an addition to the United States flora (Pringle, 1876, 1897). There he collected considerable quantities for distribution. Pringle collected on Camels Hump only in the years 1874, 1875, and 1876 (Pringle, 1897), never returning ‘“‘because its subalpine area is limited, and consequently the number of rare plants to be found there is small.” No Asplenium was found in the course of intensive field work on Camels Hump either by the writer or by 1992] Zika—Alpine Flora 19 many others, including members of the American Fern Society in 1926 (Chisolm, 1926). No Camels Hump voucher could be found at GH, HNH, MO, NEBC, NY, SJFM, TUFT, VINS or VT. There was no reference to a Washington Co. Pringle collection in either the manuscript notes for Harry Ahles’ unpublished flora of New England, nor in the Vermont checklist (Atwood et al., 1973). Those works were based on the holdings of more than 30 public and private herbaria in the region. The Asplenium viride citation for Camels Hump is believed to be a confusion with Pringle’s record of Woodsia gla- bella on Camels Hump (Pringle 1897; Chisholm, 1926; Zika, 1982), which was not mentioned by Brainerd et al. (1900), Eg- gleston (1905) or Eggleston et al. (1915). Calamagrostis pickeringii Greene (1987) did not credit Vermont with any records of Pickering’s reedgrass. Although he verified the identification of an 1877 collection from Vermont (Pringle s.n. us), Greene’s an- notation suggested the sheet was a mislabeled collection from New Hampshire, since no other collections are known from Ver- mont. Pringle’s specimen bears no locality data besides “Ver- mont.” It is the basis of reports in Brainerd et al. (1900), Eggleston et al. (1915), Fernald (1907, 1950) and Hitchcock and Chase (1971). I disagree with Greene’s interpretation of Pringle’s Vermont record, and accept the Vermont record, as did Pringle’s contem- poraries. Pringle was a careful collector and there is no indication that he mislabeled other rare species he found in Vermont, the vast majority of which were confirmed by later collectors. The distribution of Calamagrostis pickeringti Gray in northern New York and northern New Hampshire suggests it should also be in Vermont’s mountains. Its habitats and associated flora in New York (Zika and Jenkins, unpubl. data) suggest Vermont’s station could have been a boggy sphagnum area on the ridgeline of Mt. Mansfield. Pringle collected heavily there in 1877. Scrutiny of the damp mossy habitats on Mt. Mansfield’s ridge- line has shown that a number of rare species have disappeared this century, including Geocaulon lividum (Rich.) Fern., Goodyera repens (L.) R. Br., Listera cordata (L.) R. Br., and apparently _ pickeringii. All the historical collection sites, in or near bogs, have been degraded by human disturbance. 20 Rhodora [Vol. 94 Carex atratiformis The only historical station for Carex atratiformis Britt., black- ish sedge, in Vermont was in Smugglers Notch, where it was last collected in 1879. This population was rediscovered in 1979 (Zika 1028 mass, 1740 vt, 10892 vt), on both the east and west slopes of the notch, between 2200 and 3400 feet, growing with Agrostis scabra Willd., Asplenium viride Huds., Carex leptalea Wahl., C. scirpoidea Michx., Conioselinum chinense (L.) BSP., Listera con- vallarioides (Sw.) Nutt. ex Ell. and Rubus pubescens Raf. Cassiope hypnoides Moss plant, a circumboreal species, is at the southern limit of its range in northern New England and New York. Contrary to Mitchell (1986), Miller (1989) doubted that Cassiope hypnoides (L.) D. Don was ever a contemporary member of the New York flora. However, there are compelling arguments the species was present in the 1800’s, suggesting it is now extirpated in New York. Two herbarium specimens (Haberer 542 nys, PA) and correspon- dence from the collector (NY State Museum botany files) are robust evidence the species was on Panther Mountain, town of Morehouse, Hamilton Co., in 1879. House (1924) confused the locality with a Panther Mountain in Essex Co., but Haberer wrote to House and emphasized Cassiope was in Hamilton Co. Steve Clemants of the New York Natural Heritage Program could not find Cassiope in 1987 at Panther Mountain, Morehouse. As Miller (1989) pointed out, the habitat is apparently no longer suitable. The mountain is currently dominated by hemlock-hard- wood forest, presumably established after logging eliminated a more boreal evergreen forest type in the 1800’s. In New England Cassiope is an alpine snowbank species. Apparently a peripheral low-elevation habitat once existed on Panther Mountain. Partial shade from a pre-logging spruce-fir canopy may have sheltered accumulated winter snowdrifts and avalanches at the base of one of the small cliffs, forming a marginal ice cave talus community (Reschke, 1990, p. 51). Cassiope habitat could have been altered irreparably with logging, when the resulting second-growth was predominantly deciduous, and the crucial coniferous shade re- quired to retard snowmelt was lost. A second New York site for Cassiope was credited to C. C. 1992] Zika— Alpine Flora 21 Parry in the 1840’s (Peck 1891, 1900) from the summit of Mt. Marcy. Suitable habitat on Marcy’s summit exists. Parry was a careful botanist who often worked in alpine areas (Britton, 1890; Parry, 1889-1897), so his report was not likely to be based on a misidentification of this unique member of the northeastern al- pine flora. Miller (1989) and Adams et al. (1920) rejected the Mt. Marcy record because the voucher specimen could not be located. Certainly the missing collection creates some doubt, but fewer than 10% of the pre-1850 alpine vouchers from Vermont have survived to the present (Zika, unpubl. data); thus the Mt. Marcy Cassiope record cannot be summarily dismissed. Indeed, collec- tions of the species at a lower elevation in Hamilton Co. suggest it potentially was in the High Peaks alpine zone in Essex Co. in the 1800’s. Field work by Ketchledge (1984), DiNunzio (1972, unpubl. M. S. thesis, State U. of NY, Syracuse), the writer and others has failed to relocate C. hypnoides on Mt. Marcy. By 1880, upon failing to re-locate Cassiope, Peck (1891, 1900) suspected the species had already been extirpated on Mt. Marcy. I concur, and suggest the species has been extirpated in New York State. The alpine community of Mt. Marcy was severely degraded this century by visiting hikers and campers, although the peak is currently a designated wilderness. Trampling may have affected the historical Cassiope location. Remote Panther Mt. remained undisturbed following logging. Castilleja septentrionalis Oakes (1842) reported the first Vermont collection of Castilleja septentrionalis Lindl. (as Bartsia pallida L.) by Tuckerman and Macrae in 1839, “‘on the north side of Mansfield Mountain, near the summit.” It is not known whether this site was on the ridgetop alpine zone, or on the sub-alpine cliffs or streams in Smugglers Notch where the species is extant today, more than one-half mile northeast of (and 1500 feet below) the Mt. Mansfield ridgeline. No data are available from the label of the voucher collection; it has not been found, like most of Tuckerman’s and Macrae’s col- lections from this period. Support for an alpine rather than a sub- alpine station comes from Tuckerman’s journal entries for 1839, at the Amherst College Archives. On 3 July, Bartsia pallida was seen on the “northeast side [of the] Chin,” the summit of the alpine ridge. Further evidence is on page 88 of the same journal a2 Rhodora [Vol. 94 for 1839. There is a list of plants found “by Edward Tuckerman junior in the alpine regions of high mountains in N. E.” and it includes Bartsia pallida from ‘“‘Mansfield mtn. Vt. 1839,” and from the White Mountains in 1838. Castilleja septentrionalis has not been found on the Chin of Mt. Mansfield since 1839. Recent intensive field searches by the writer and by Green Mountain Club rangers and caretakers (e.g., Bristow et al., 1977) indicate that pale painted cup has disappeared from this site, despite its current status as a protected natural area. Draba cana Draba cana Rydb. (D. lanceolata of northeastern authors, not Royle) was discovered in Smugglers Notch in 1878 (Pringle s.n. vt). Eggleston (1895) quoted Pringle’s description of the popu- lation as, “‘a patch I could have covered with my hat.” Eggleston (1895) “found a similar patch in 1893” and labeled his collections “very rare” (Eggleston s.n. HNH, NEBC, NY). A tiny station was documented in 1979 (Zika 1051 Mass) at the base of an exposed xeric cliff on the west side of the notch, at 2700 feet, under a spineless Rosa blanda Ait., rather distant from the landslide gul- lies that are usually explored. The writer and David Barrington found another tiny colony in 1990 at 3000 feet, on a rotten north- facing cliff on the west side of Smugglers Notch (Zika 10909 vt). Draba cana is a locally common plant in Vermont only on the southwest face of Mt. Pisgah, where hundreds are extant. It was apparently first collected there in 1862 by Mann, Tuckerman, et al. (Kennedy, 1904; Fernald and Knowlton, 1905). Empetrum eamesii ssp. atropurpureum Empetrum eamesii spp. atropurpureum (Fern. & Wieg.) D. Love, purple crowberry, was collected on Mt. Mansfield between 1851 (Russell s.n. A) and 1908 (Flynn s.n. NEBC, vT). It has not been seen more recently, despite intensive field searches by the author and many others, and is presumed extirpated in Vermont. Empetrum nigrum Black crowberry, E. nigrum L., was discovered at a fourth station in Vermont by Harry T. Peet, Jr., in 1982. About 20 plants were located on a large schist outcrop at 3250 feet, on the south 1992] Zika— Alpine Flora 23 flank of Bolton Mountain, Bolton, Chittenden Co. (Zika, Peet & Sulek 6325 vt). It was growing with Amelanchier bartramiana (Tausch) Roem., Betula papyrifera Marsh. var. cordifolia (Regel) Fern., Carex brunnescens (Pers.) Poir. ex Lam., C. debilis Michx., Cystopteris fragilis (L.) Bernh., Ledum groenlandicum Oeder, Nemopanthus mucronatus (L.) Loesener, Picea mariana (Mill.) BSP., Sorbus americana Marsh., Vaccinium boreale Hall & Aal- ders, V. myrtilloides Michx., V. uliginosum L. and a collection tentatively identified as V. boreale x uliginosum. Festuca brachyphylla Festuca brachyphylla Schult. & Schult., alpine fescue, was not credited to Vermont or New York State by Frederiksen (1982) in his treatment of this arctic and western alpine species. Apparently reports of F. saximontana Rydb. from the northeastern United States can be referred to F. brachyphylla. Gleason and Cronquist (1963, p. 55) incorrectly state the habitat is “alpine summits of N. Y. and New England.” Festuca brachyphylla is not found on summits or in alpine areas in the northeastern United States. The two historical stations in the region are Smugglers Notch and Wallface Mountain. Small extant populations were confirmed at both sites in 1990. These subalpine habitats are similar, being at ca. 3000 feet, on north-facing steep rocky turf, and partially shad- ed by tall cliffs. Associated species include Campanula rotundi- folia L., Carex scirpoidea Michx., Draba arabisans Michx., Hous- tonia caerulea L. and Saxifraga paniculata Mill. The bluish foliage of alpine fescue helps distinguish it from similar capillary-leaved cespitose plants on the cliffs, such as sterile Deschampsia flexuosa (L.) Trin. and Carex eburnea Boott. Some New York reports (Smith, 1965; Mitchell, 1986; Cle- mants, 1989) are based on a 1964 collection from Whiteface Mountain, Wilmington, Essex Co. (Smith et al. 37420 nys). The Whiteface specimen is Festuca tenuifolia Sibth. Geocaulon lividum Northern comandra is rare and threatened in New Hampshire (Pub. Law 93-205, Res-N 301.02), although Stern (1979) found it was widespread in the Mahoosuc Range on the Maine-New Hampshire border. Most Geocaulon populations in the northeast are quite small, and the distinctive rhizome is collected all too 24 Rhodora [Vol. 94 often. Geocaulon lividum is a cryptic species, easily overlooked unless in fruit. In nature, G. lividum resembles a solitary un- branched shoot of Comandra umbellata, and is colored like Vac- cinium uliginosum L., a common associate. Geocaulon lividum was re-located in 1984 at Eggleston’s 1901 station on Mt. Clinton in the Presidential Range, Bean Grant, Coos Co., N. H. (Pease, 1964). The solitary plant was observed on the edge of the Appalachian Trail at 4100 feet elev., in Abies scrub. To help protect the population, no voucher was collected. In Vermont, northern comandra was known from a single site, a tiny bog on Mt. Mansfield, where it has not been found since 1901, despite intensive field searches, and is now presumed ex- tirpated. The bogs on Mt. Mansfield have been degraded by hu- man disturbance despite the current protected status of the alpine ridge. Geum peckii Only one report from Vermont was found for Geum peckii Pursh, mountain avens (Atwood et al., 1973). The citation was based on a collection (Carpenter s.n. vt) labelled: “Mt. Willough- by slope, 21 July 1926.” Mt. Pisgah, on the shore of Lake Wil- loughby, Westmore, Orleans Co., Vermont, was often called Mt. Willoughby by collectors. There is evidence Carpenter was care- less with his label data, and his Vermont record is rejected for several reasons. No recent workers have found this endemic of New Hampshire and Nova Scotia in the Lake Willoughby area. Furthermore, Car- penter (1927) did not include the species in his article on recent additions to the Vermont flora. The date of collection also suggests a locality error. On 20 July 1926, Carpenter was collecting Scirpus cespitosus L. “near [the] summit, Mt. Washington, N.H.” (Car- penter s.n. vt). It would be difficult even with modern transpor- tation to be collecting on Mt. Pisgah one day after hiking on Mt. Washington. Why do this and then return to Mt. Washington? Carpenter was camping in the White Mountains a week later; he and Frank Dobbin saw Geum peckii on Mt. Washington on 29 July (Dobbin, 1927). Surely Vermont’s G. peckii voucher is from Mt. Washington, N. H. Incongruous Carpenter labels for speci- mens of Scirpus cespitosus and Phyllodoce caerulea are also dis- cussed below. 1992] Zika— Alpine Flora 25 Luzula spicata Spiked woodrush, Luzula spicata (L.) DC., was first found in New York by Peck in 1898, who described the station (Peck, 1900, p. 645) from the: “top of Wallface Mountain, Essex County ... It is found in considerable abundance along the brow of the precipice that forms the western wall of Indian Pass.” Spiked woodrush is no longer common on the brink of Wallface. In 1989 only 50 plants were found (Zika 10746 Nys), in one tiny area, growing in the thin band of alpine vegetation dominated by Des- champsia flexuosa (L.) Trin., Juncus trifidus L., Potentilla triden- tata Soland ex Ait., Solidago spathulata spp. randii (Porter) Cronq. and Vaccinium uliginosum. This population is threatened by trampling when rock climbers are belaying or completing their climb. A second Luzula population on gravelly talus directly be- low suggests a seed rain from the clifftop. The lower population of Luzula spicata is growing in a grami- noid-dominated turf among Agropyron trachycaulon (Link) Malte, Agrostis scabra, Bromus ciliatus L., Carex debilis Michx., C. echinata Murr., C. houghtonii Torr., C. scirpoidea Michx., and Poa nemoralis L. Luzula spicata is rare and very local at the cliff base, and is threatened with trampling by rock climbers. The few plants observed in 1989 were along the approach to one of more than 20 technical rock climbs on the ramparts of Wallface (Mellor, 1988). In Smugglers Notch, Vermont, Luzula spicata was recorded between 1879 (Brainerd s.n. vt) and 1908 (Kirk s.n. NEBC). A small population was re-located (Zika 10903 vr) while exploring with Cathy Paris and David Barrington in 1990, on a steep damp rocky open north-facing turf partially shaded by the steep wall of a gully on the west side of the notch. Paronychia argyrocoma Paronychia argyrocoma (Michx.) Nutt. (including var. albi- montana Fern.), whitlow-wort, was not included in the flora of Vermont (Atwood et al., 1973; Seymour, 1969). Ray Angelo and the late Harry Ahles brought to my attention two Vermont spec- imens. One sheet has collections from several states, and several labels, one reading: “Willoughby Mt., Stowe Vt., Aug. 1865” (Ex herb. F. J. Bumstead, M. D. s.n. MASS). The site data on this label 26 Rhodora [Vol. 94 are puzzling. There are several plants mounted near the label so it is not clear if the collections came from both Stowe, Lamoille Co. and Willoughby Mountain [Mt. Pisgah], Westmore, Orleans Co. Or perhaps the collections are from only one of the two label sites, and the labeler was confused or in error. The Willoughby area seems more likely. Potential P. argyrocoma habitat near Lake Willoughby should be explored, including a granitic cliff with Potentilla tridentata Soland. ex Ait. on Wheeler Mountain. This site closely resembles good Paronychia habitat on several New Hampshire peaks, including Mt. Willard, Crawford’s Notch. The second Vermont Paronychia argyrocoma collection is la- beled: “rare, Vermont,” with no date (Ridler 339 BEDF). This label data is typical for a C. E. Ridler collection. His herbarium, with specimens from many states, was described by Huntington (1881) as “extensive.” Ridler’s vouchers include a number of rare species from cliffs by Lake Willoughby. Ridler (1884) described botaniz- ing the cliffs above Lake Willoughby, but did not mention finding Paronychia. If he found the species in the area, it presumably would have been on a return trip after 1884. There are no recent records for Paronychia argyrocoma from Vermont, but more field work in the Lake Willoughby region is needed to determine its current status. Phleum alpinum Phleum alpinum L., alpine timothy, reported from “alpine regions, Vermont, F. H. Horsford” (Dole, 1937), is not supported by an herbarium collection. The report is rejected. It is believed to be a confusion with P. pratense L., a weed in Vermont’s alpine regions. Phyllodoce caerulea Vermont’s only report of Phyllodoce caerulea (L.) Bab., moun- tain heath (Atwood et al., 1973) rests on a specimen labelled: “Alpine Garden, Mt. Willoughby, 5000-6000 feet, 20 July 1926” (Carpenter s.n. vt). Mt. Willoughby is now Mt. Pisgah (2750 feet), a peak with a “flower garden” (Kennedy, 1904). Mt. Washington, N.H. (6288 feet), has an “alpine garden.” Dobbin (1927) de- scribed finding on Mt. Washington “the mountain heath, Phyl- lodoce coerulea [sic], which we had never before seen.” Dobbin 1992] Zika—Alpine Flora ar was with Carpenter, and the date was nine days after the appar- ently mis-dated Vermont collection was made. The altitude data on the “Vermont” specimen also implies it was collected in New Hampshire. As with Geum peckii, P. caerulea is rejected from the flora of Vermont. Prenanthes boottii Prenanthes boottii (DC.) Gray, a rare northeastern endemic, is often thought to be restricted to alpine habitats (e.g., Crow, 1982). On Mt. Mansfield it was found (Zika 4712, 4728 vt) on three schist ledges with an eastern exposure, between 3500 and 3900 feet, well below treeline, growing with Abies balsamea, Asplenium viride Huds., Juncus trifidus L., and Lycopodium lucidulum x selago. It may be in similar situations on other major peaks. Peck (1900) discovered Boott’s rattlesnake-root on Mt. Marcy in New York in 1898. He noted it was “very rare” but collected at least five plants. It is possible he overcollected and extirpated the population, as there are no subsequent records despite the intensive searches by Adams et al. (1920), DiNunzio (1972, un- publ. M. S. thesis, State Univ. of New York, Syracuse) and many others, including the writer. P. boottii is extant on two other Adirondack peaks. Mt. Marcy is in a designated wilderness area, but the summit has been degraded by recreational use this century. Poa fernaldiana Poa fernaldiana Nannf., wavy bluegrass, was last seen in Ver- mont in 1897. A tiny population was re-located in 1990, while botanizing with Everett J. Marshall, on the Nose of Mt. Mansfield. It was associated with Agrostis mertensii, Carex brunnescens, Hierochloe alpina (Sw. ex Willd.), Lycopodium selago and mosses, on a steep northeast slope at 3900 feet. Human disturbance in the vicinity dates back to a road and hotel constructed in 1858 at the base of the Nose (3850 feet). Communication facilities were installed on the Nose’s summit (4062 feet) in 1954 (Hagerman, 1975). One unintended result was the establishment of weedy populations of Poa compressa L. and P. pratensis L. They have abrogated some alpine habitat previously available to P. fernal- diana, Hierochloe alpina and Carex bigelowii on the summit of the Nose. 28 Rhodora [Vol. 94 Pyrola minor Mountain pyrola, Pyrola minor L., is inconspicuous and cryp- tic. Mountain forms of the more common P. elliptica Nutt. often produce large mats of the reduced, rotund foliage typical of P. minor. At its historical site in Vermont, on the floor of Smugglers Notch near Big Spring (Eggleston, 1895), P. minor has not been seen since 1896, despite numerous searches, and is presumed extirpated. A new Vermont station for Pyrola minor (Zika 6211 vt) was located at 3500 feet along the Long Trail on Camels Hump, Hun- tington, Chittenden Co., growing with Abies balsamea, Chelone glabra L., Rubus pubescens Raf., Solidago macrophylla Pursh, Thalictrum pubescens Pursh, and Veratrum viride Ait. The pop- ulation consisted of 73 plants in 1982. The colony, along a wet, eroding area on the footpath, is threatened by routine trail work such as ditching and waterbar installation or maintenance. A historical population in Wilmington Notch, Essex Co., New York, was not re-located after several days of searching in 1989. More field work is required to determine its status there. However, S. J. Smith believed the colony was destroyed by altering the highway alignment in the notch in the 1970’s (Alvin Breisch, pers. comm). Salix herbacea Salix herbacea L., dwarf willow, was reported from Camels Hump, Vermont by Carpenter (1927), Dole (1937) and Bean et al. (1956). Carpenter’s original report was founded on a misiden- tified sheet of bearberry willow (S. uva-ursi) at the Pringle Her- barium (Kirk, 1950). Salix herbacea is rejected from the Vermont flora. Salix planifolia Tea-leaved willow, Salix planifolia Pursh, is an alpine and sub- alpine species in the northeastern United States. Seymour’s (1969, 1982) habitat summary: “meadows, swamps” is incorrect for our area, implying low elevation habitats. The statement appears to based on records from other parts of the species’ range, or on misidentified specimens of Salix discolor Muhl. from low ele- 1992] Zika— Alpine Flora 29 vation sites. Four such records are discarded: Vermont collections from Burlington, Fairfax, Stratton and Westford, cited in 1973 by Atwood et al. Vermont’s only station for S. planifolia is on Mt. Mansfield, and it is extant where Pringle discovered it in 1S7i. Scirpus cespitosus Random extinctions are predicted on functional islands of al- pine habitat by island biogeographic theory (MacArthur and Wil- son, 1967). This assumption probably explains how deer’s hair can be locally dominant in the alpine zone in New Hampshire and New York, yet absent in appropriate habitat in the alpine zones of Mt. Mansfield and Camels Hump. Scirpus cespitosus L. is present on the west slope of Mt. Mansfield on sub-alpine ledges at 3000 feet (Zika 724 vt). It is a dominant on the northeast side of Mt. Mansfield, in Smugglers Notch, where it occurs with many of the species found in the Willoughby flora. The current distri- bution of S. cespitosus suggests it may have been present on the summit of Mt. Mansfield, and became locally extinct prior to the first botanical investigations. Examination of macrofossils from the bogs on Mt. Mansfield could answer this question. Scirpus cespitosus is absent on the sub-alpine cliffs of five peaks around Lake Willoughby in Westmore, Orleans Co., Vermont. A solitary Mt. Pisgah record (Carpenter s.n. vt) from 21 July 1926 appears to be one of several mis-labeled alpine Mt. Washington records (see Geum peckii and Phyllodoce caerulea), and is dis- missed. Solidago cutleri Solidago cutleri Fern., alpine goldenrod, is endemic to the mountains of northern New York, Vermont, New Hampshire and Maine. On some ranges it is a common component of the alpine community. In Vermont the population in Smugglers Notch was last documented in 1891. A tiny population was re-located in 1990 on a schist shelf in a ravine on the west side of the notch, at an elevation of 3000 feet. It was growing with S. spathulata ssp. randii, which is common in the area. The presence of sus- pected hybrids and the extreme rarity of S. cutleri suggests that genetic swamping may be a problem for this population. 30 Rhodora [Vol. 94 The population of alpine goldenrod on the Chin of Mt. Mans- field has not been documented after 1908 (Flynn s.n. vr), and is presumed extirpated. It seems unlikely this bright-flowered spe- cies could be overlooked during the intensive searches conducted by the author, Bristow et al. (1977) and others. The Chin has been heavily trampled along hiking trails, but is otherwise little changed from the turn of the century. Botanical collecting may have been important in the decline of this population. Streptopus xoreopolus Streptopus xoreopolus Fern., mountain twisted stalk, was con- sidered “possibly extirpated” in Maine (Dibble et al., 1989) and was not seen in searches in 1988 (Dibble et al., 1990). A small population was located in 1990 on Hamlin Peak, T3, R9, Pis- cataquis Co. Five flowering plants were in a snowbank community at the base of an an alpine cliff with northeast exposure at 4650 feet elevation, growing with S. roseus Michx. and S. amplexifolius (L.) DC. The hybrids displayed the strongly clasping leaves and habit of S. amplexifolius, combined with the hispid herbage and pink flowers typical in S. roseus. Associates included Cassiope hypnoides, Gaultheria hispidula, Loiseleuria procumbens (L.) Desv., Luzula parviflora, Maianthemum canadense, Phyllodoce caerulea, Poa fernaldiana, Rubus pubescens, and Vaccinium ces- pitosum. To help preserve the colony, no voucher was collected. Vaccinium cespitosum Vaccinium cespitosum Michx., dwarf bilberry, is unique among the mountain Vaccinium species, with its serrated obovate leaves and solitary axillary flowers. It is probably overlooked because it is low and inconspicuous, and may be more common in river gorges in the northeast than the limited number of current records indicate; Jenkins and Zika (1987) found it on the Missisquoi, Wells and West River drainages in Vermont. The species is scarce but extant at several locations along the ridgeline of Mt. Mans- field, from near the Octagon to the Adams Apple, between 3650 and 4370 feet. In New York a new station was observed along the trail between Lake Tear of the Clouds (4300 feet) and the Panther Gorge shelter (3250 feet), southeast of Mt. Marcy, Keene, Essex Co. Another new population was observed on wet groun 1992] Zika—Alpine Flora 31 along the trail on the southeast shoulder of Mt. Haystack, Keene, at 3750 feet. About 700 plants were seen on the northeast slope of Mt. Marcy, where it is common between 4800 and 5200 feet in sheltered alpine areas, near treeline, in the general area where Peck (1900) reported it. Viburnum edule Squashberry, Viburnum edule (Michx.) Raf., often occurs in small populations of 1-6 shrubs in the northern Appalachians. A number of colonies are sure to remain undocumented until there is more off-trail exploration of the headwaters of small mountain brooks, a favored habitat for the species. Viburnum edule is extant at four locations on the ridgeline of Mt. Mansfield, Vermont, between 3500 and 3900 feet, growing with Salix planifolia and other shrubs (Zika 4714, 4715, 4719 vt). The Mansfield popu- lation extends into the towns of Cambridge and Stowe, Lamoille Co., as well as into Underhill, Chittenden Co. A new Vermont locality is on the trail to the abandoned fire tower on Mt. Monad- nock’s east slope, at 3000 feet, in Lemington, Essex Co. (Zika 9217 vt). Another new station is on Camels Hump, at 3500 feet in Huntington, Chittenden Co. Nine shrubs were found in wet thickets (Zika 6212 vT). In New York, new locations of small populations were sighted in 1989 on Santanoni Brook (Newcomb), and Algonquin Peak (Keene, Zika 10685 nys). One colony was seen on Phelps Brook (North Elba) in 1981 (Zika 5077 vr). Extensive colonies (10-100 shrubs) were seen along the northern Marcy Brook at Indian Falls (Keene, Zika 1981 and 1989 observations), on Opalescent River below Uphill Leanto (Keene), on two small brooks draining north- east and southwest of Lake Arnold (Keene), and on the southern Marcy Brook that drains Panther George (Keene). All these sites are in Essex Co., in the Adirondack Mountains. DISCUSSION How recently established are these newly reported montane stations? Why were the localities not documented in the herbaria of the numerous, and apparently quite thorough, collectors of the last century? Has there been a recent widespread dispersal of these rare taxa? This last possibility is doubtful. Consider six rare moun- Ee Rhodora [Vol. 94 tain taxa in close proximity in the Adirondacks: Epilobium horne- mannii Reichenb., Festuca brachyphylla, Loiseleuria procumbens, Luzula spicata, Poa fernaldiana, and Salix herbacea. Historically, each is documented from a single site. If they suddenly appeared on a second peak it would suggest recent dispersal, which simply has not happened. A more likely, though speculative, explanation is that “new” populations of Prenanthes boottii, Agrostis mertensii, Empetrum nigrum etc. are apparently relictual early Holocene populations overlooked or inaccessible to previous botanists. Three Vermont examples, from Mt. Mansfield, Bolton Mtn., and Camels Hump, are given as circumstantial evidence to illustrate the point. Bear Pond, which is not visible from the alpine ridge, was not mentioned on any of the ca. 1200 herbarium labels I examined for Mt. Mansfield. Bear Pond populations of Asplenium viride, Carex atratiformis, Prenanthes boottii, and Viburnum edule were apparently ignored by earlier botanists who collected these species from more accessible areas. Harold St. John (pers. comm.) bush- whacked through formidable krummholz in this area in July 1911 because it was still trailless; he and earlier botanists could easily have missed the ledges with rare species in such difficult terrain. A modern trail now provides access to the vicinity. Empetrum nigrum and its alpine associates on a Bolton Moun- tain cliff were discovered in a remote trailless area only when scouting for a new route for the Long Trail in 1982 (Peet, pers. On Camels Hump, the dismantling of overnight huts in the hut clearing (3800 feet) and the construction of Gorham Lodge 0.4 miles to the north (3400 feet) was not done until 1950 (Peet, 1977). Thus only recently has there been convenient lodging north of the hut clearing. Naturally the botanical work of the last century focused on the alpine regions south of the hut clearing, and not on the low diversity boreal forest to the north, where a trail was not built until 1910-1911 (Waterman and Waterman, 1989, P. 358). Day hiking trails used by Robbins, Macrae and Tuckerman (Oakes, 1842) were from the east or west, as today. Longer (over- night) hikes from the north apparently were avoided by botanists, e.g., Chisolm (1926) reported a botanical field trip climbing from the east, not the north. ““New” sites for Pyrola minor and Vibur- num edule in this paper are on the north trail. Viburnum edule, and to a lesser degree Vaccinium cespitosum, 1992] Zika— Alpine Flora 33 are the only species studied with substantial numbers of newly reported stations. These sub-alpine (to low elevation) species were largely found in areas accessible to modern botanists due to re- cently constructed hiking trails and roads. More sites should be found with the off-trail exploration of mountain headwater brooks (Viburnum edule) and boating in rapids and river gorges (Vac- cinium cespitosum). It is becoming clear that the floras on our alpine peaks are changing; the data show several species declined or became ex- tirpated from alpine areas this century. The reasons for this change are not clear, but several possibilities can be offered. Effects of human disturbance, generally associated with recreation and bo- tanical collecting, may have been detrimental in the past. The Little Ice Age ended in the early 1800’s (Dansgaard et al., 1975); Hamburg and Cogbill (1988) noted an average summer temper- ature increase of more than 2°C at stations in Massachusetts and New Hampshire since 1830. Climatic amelioration following the Little Ice Age could have a negative impact on disjunct species in alpine areas. If so, then current predictions of increased global warming may have serious consequences in the plant commu- nities in northeastern alpine areas. Chronic pollution may also be a potential threat in the future. Further studies of historical extinctions of alpine species are needed in alpine areas across the northeast (e.g., Dibble et al., 1990). Demographic studies are needed for the apparently de- clining populations of Arnica lanceolata, Arenaria rubella, and Luzula spicata in order to determine the proximate and ultimate causes of population shifts; all are in remote, protected areas, and the latter two are in essentially pristine habitats. They may be bellwether species for the boreal flora of the northeast. ACKNOWLEDGMENTS For their assistance in the field, I am indebted to D. S. Bar- rington, W. D. Countryman, J. C. Jenkins, C. Leunig, C. A. Paris, R. Paul, H. T. Peet, K. Regan, C. A. Savonen, J. Sulek, and A. Van Sweringen. Invaluable assistance was provided by the late Harry Ahles, Ray Angelo, Charles V. Cogbill, J. K. Dean, Craig Greene, Janice Hall, Deborah Lewis, Harold St. John, and Mary Walker. I thank the curators of A, BDI, BEDF, BUF, GH, HNH, MASS, MO, NEBC, NY, NYS, PA, SJFM, SYRF, TUFT, US, 34 Rhodora [Vol. 94 and VT for loans or access to their collections. Funding came from Vermont Fish and Wildlife Department’s Nongame and Natural Heritage Program, The Nature Conservancy, the Pringle Herbarium, the Vermont Bird and Botanical Club, the New York Natural Heritage Program and the New York Department of En- vironmental Conservation. LITERATURE CITED ApaMs, C. C., G. P. Burns, T. L. HANKINSON, B. Moore AND N. TAYLor. 1920. Plants and animals of Mount Marcy, New York. Ecology 1: 71-94, 204-233, ANONYMOUS. 1877. Pringle’s plants. Bull. Torrey Bot. Club 6: 199. Atwoop, J. T., W. D. COUNTRYMAN, R. A. Jervis, D. H. MILLER, F. C. SEYM AND M. L. SmitH. 1973. Check List of Vermont Plants. Vermont Bot. pee Bird Club. Bean, R. C., C. H OWLTON AND A. F. Hitt. 1956. Eleventh report of the committee on er distribution. Rhodora 58: 125-134. BRAINERD, E., W. W. EGGLESTON AND L. R. Jones. 1900. Flora of Vermont. Free — a Burlington, VT. Bristow, P., M. ROHMAN, P. ROHMAN AND J. PETERSEN. 1977. Some vascular plants af the alpine zone of Mt. Mansfield, Vermont. Report to the Green Mountain Club, Montpelier, VT. Britton, N. L. 1890. Charles Christopher Parry. Bull. Torrey Bot. Club 17: 74-75. ARPENTER, D. S. 1927. Additions to the flora of Vermont. Vermont Bot. and Bird Clubs Joint Bull. 12: 25. CHISHOLM, M. L. 1926. A field meeting with Dr. W. S. Monroe. Amer. Fern J. COUNTRYMAN, W. D. 1978. Rare and Endangered Vascular Plant 9 me me me me me me EN me . cordifolius x A. pilosus var. pilosus. 2n = 32. CANADA. Ontario. Waterloo Co.: Wellesley Twp., along farm rd., S & Ch 9099. . dumosus L. 2n = 32. U.S.A. South Carolina. Florence Co.: US-51 N of Pam- plico, S & Ch 6116. . elliotti Torr. & Gray. 2n = 16. U.S.A. Florida. Lake Co.: FL-46 at Wekiwa R.., S et al. 5342 me nN eee 1992] Semple et al.—Astereae 51 Table 1. Continued. A. foliaceus Lindl. 2n = 64. U.S.A. Utah. San Pete Co.: E of Ephraim, rd. to top of Wasatch Plateau, S & Ch 8891. Summit Co.: Wasatch N.F., For. Rd. just E of UT-150 just S of N.F. border, S, Su & Ah 9213. Wyoming. Carbon Co.: S of Arlington, Rocky Creek Trail, ca. 2 km S of I-80, S, Su & Ah 9205. A. laevis L. var. laevis. 2n = 48. CANADA. British Columbia. N of Windermere, Lake Drive by Wenderemere L., disturbed woods, S & Bt 436] (count by L. Brouillet). U.S.A. Montana. Silver Bow Co.: SE of Butte, near Continental Dn . 2 mi. E of access to I 90, s & Bt 4422 (count by L. Brouillet). A. lanceolatus Willd. ssp. / le & Chmielewski. 2n = 64. U.S.A ma. Yavapai Co.: US- 89 (MP309) SE of Prescott, by creek, S & Ch 8998. A. lanceolatus ns lanceolatus var. hirsuticaulis Semple & Chmielewski. 2n = 32. s. sconsin. Bayfield Co.: WI-27 21.3 km S of Douglas Co. Rd. A, S$ of i. S 9084. Douglas Co.: Co. Rd.-U S of Amnicon Falls State Park, A. lanceolatus ssp. lanceolatus var. lanceolatus. 2n = 64. U.S.A. Wisconsin. Doug- las Co.: Co. Rd.-U S of Amnicon Falls State Park, S 9071. A. lanceolatus ssp. lanceolatus x A. pilosus var. pilosus. 2n = 48. CANADA. Ontario. Waterloo Co.: Wellesley Twp., by RR, S & Ch 9096. A. lateriflorus (L.) Britt. 2n = 16. U.S.A. North Carolina. Wilson Co.: NE of Eureka, NC-222 2 mi. N of county line, S & Ch 6009. 2n = 32. Pennsylvania. Centre Co.: Nittany Mts., PA-144 just W of Pleasant Gap, S & R 7602. A. aff. lentus Greene. 2n = 64. U.S.A. California. Napa Co.: — junction of Fairground Rd. and American Cr., SE of Napa, S & Ch 891 : —— Lam. 2n = 80. U.S.A. Wisconsin. Douglas Co.: a of Barnes, pond near WI-27, S 9078. : linariifolius L. 2n = 18. U'S.A. Florida. Okaloosa Co.: US-90, E of Crestview, 2 mi W of Shoal R., S. Bt & Canne 3893. Mississippi. Simpson Co.: MS-13 0.6 mi N of MS-28, S of Mendalhall, S, Bt, & Canne 3815. A. macrophyllus L. 2n = 72. CANADA. Nova Scotia. Yarmouth Co.: NS-340 just N of Carleton, S & Keir 4858. U.S.A. Pennsylvania. Elk Co.: US-219 S of Ridgeway, N of Hwy.-948, S & R 7594. Wisconsin. Sheboygan Co.: Kettle Moraine State For., Kettle Hole Geological Marker Site, slopes of kettle hole, S 9091. A. modestus Lindl. 2n = 18. CANADA. Alberta. Jasper Nat’l. Park: N of Alta- 16, W of Clairvaux Creek (trail site), S & Bt 4323. British Columbia. Cassiar Hwy. 33 km S of Cassiar, Mighty Moes Campground, KP580.2, edge of Cotton L., Ch et al. CC4726. = 16. U.S.A. California. Plumas Co.: CA-89 SW of Crescent Mills, s et ait 5719. oolentangiensis Riddell. 2n = 32. U.S.A. Wisconsin. Douglas Co.: WNW of Barnes, margins of kettle hole pond by WI-27, 2.1 km N of Co. Rd.-A, S nN nN pa S ie = Z e s tN 3 ~ 9082. A. aff. parviceps (Burgess) Mackenzie & Bush. 2n = 32. U.S.A. Missouri. Wash- ington Co.; MO-8 6.6 km W of Potosi, S, Su & Ah 9393. A. pilosus Willd. var. pilosus. 2n = 32 U.S.A. Arkansas. Dallas Co.: N of Princeton, AR-9 just N of AR-48, S & Ch 6402. aa Williamson Co.: TN-96 E 52 Rhodora [Vol. 94 Table 1. Continued. ay mh A. a A, A. A, of Triune, S & Ch 6289. West Virginia. Hampshire Co.: WV-28 4.5 mi. S of Springfield, S & Ch 5883. 2n = 48. U.S.A. Kentucky. Hopkins Co.: KY- 109 1 km N of Charleston, S & Su 9431. . pilosus Willd. var. pringlei (Gray) Blake. 2n = 48. U.S.A. Massachusetts. Berkshire Co.: S of Otis, MA-8 0.9 mi. S of MA-23, S & Bt 3628. Virginia. Northumberland Co.: Resdville. S & Ch 5988. . prenanthoides Muhl. 2n = 32. U.S.A. New York. Hamilton Co.: NY-28 by Eighth L., S & Bt 3665. Lewis Co.: Lowville, NY-26 0.1 mi S of NY-12 & 26, S & Bt 3650. priceae Britt. 2n = 64. U.S.A. Tennessee. Trousdale Co., US-231 3.3 km N of Cumberland R., S & Ch 9129. . puniceus L. 2n = 16. CANADA. Ontario. Prescott Co.: Hwy-17 between Alfred and Plantagenet, picnic area, S 2413. U.S.A. Michigan. Mackinac Co.: US-2 10.6 mi. E of Naubinway, S & Ch 5007. 2n = 32. Minnesota. Mille Lacs Co.: US-169 between Rum River Rest Area and rd. to Onamia, S 9064. radulinus Gray. 2n = 27 (triploid). U.S.A. California. Mendocino Co.: E of Ft. Bragg, CA-20 at Chamberlain Creek For. Demonstration Area picnic ground, S & Hd 8552. (This is the same population sampled as S et al. 5677 and previously reported as triploid by Semple et al., 1983) . reticulatus Pursh. 2n = 18. U.S.A. Florida. Liberty Co.: NW of Sumatra, near branch of Kennedy Creek by FI-379, Godfrey s.n. (fruit collection only). schreberi Nees. 2n = 54. CANADA. Ontario. Halton Reg. Mun.: S of Waterdon, Grindstone Creek Valley, S, Bradley & Axon 9047. York Co.: Richmond Hill, golf course, Ch & Cameron 2337. . Shortii Lindl. in Hook. 2n = 16. U.S.A. Tennessee. Marshall Co.: Henry Horton State Park, Wilhoite Mill Hiking Trail, S & Ch 9116. sibiricus L. 2n = 18. CANADA. Alberta. Ghost River Valley, Chinnappa & Ch CC4906. Jasper Nat’l. Park: Sunwapta Pass, S & Bt 4347. British Colum- bia. Cassiar Hwy., 53 km N of Good Hope, 52 km S of Alaska Hwy., Ch et al. CC4713; Chilkat Pass, base of Nadahini Mt., Haines Hwy., KmP147.7, Ch et al. CC3959, Teska River Flats, Alaska Hwy., KmP600, Ch et al. CC3775. Yukon. Carcress Desert N of Carcress, Klondike Loop 2, S of White- horse, Ch et al. CC3882; Dempster Hwy., KmP107-108, beside Blackstone River, Ch et al. CC4512; NE of Stewart Crossing, Siver Trail Hwy., KmP19, Ch et al. CC4564; 5 km S of Pelly Crossing, Klondike Loop, KmP460,Ch et al. CC4601; South Canol Rd., KmP165, Lapie L. access rd., Ch et al. CC 4676, Chet al. CC4677; junction of Sheep Creek and Rose River, Ch et al. CC4691. A. . S of Deadhorse, Dalton Hwy., MP407, Prudhoe Bay, Ch et al. CC4387; Edgerton Hwy., MP26-25, 7 mi. N of Chitina, Ch et al. CC4069, Edgerton Hwy. near Kennicott River, Ch et al. CC4080; pe NW of Kiskulana Bridge, MP14, Ch et al. CC4075; Richardson Hwy., 7.7 of McCullen Creek, 35 km N of Paxson, Ch et al. CC4087; Glenn oi 6 km E of Little Nelchina R. Bridge, near Nelchina State Hwy. maintenance station, Ch et al. CC4111; Pond E of KmP137.3, Long L., Ch et al. CC4134; NE of Fairbanks, Chena Hot Springs Rd., MP39.5, Granite Tours Trailhead, N Fork of Chena River, Ch et al. CC4233; Elliot Hwy., just SW of Baker 1992] Semple et al.—Astereae 53 Table 1. Continued. Creek, MP138, Ch et al. CC4281; Dalton Hwy., 12.2 mi. N of Coldfoot, MP187.2, Ch et al. CC4291. A. simmondsii Small. 2n = ca. 64. U.S.A. Florida. Indian River Co.: E of Fellsmere, o. Rd.-512 just E of I-95, S 7525. A. geingniy L. 2n = 16. CANADA. Nova Scotia. Yarmouth Co.: NW of Carle- NS-340 S of Richfield, edge of lake, S & Keir 4860. U.S.A. Vermont. nate Co.: Rockingham, S 6884. A. umbellatus Mill. var. pubens Gray. 2n = 18. U.S.A. Wisconsin. Douglas Co.: US-2 E of Maple, S 9703. A. undulatus L. 2n = 32. U.S.A. Alabama. Cleburne Co.: N of Chulafinnee, US- 431 5 mi. S of I-20, S & Ch 6310. A. urophyllus Lindl. 2n = 16. U.S.A. New York. Cattaraugus Co.: 7.6 km N of Ellicottville, S & R 7570. Minnesota. Crow Wing Co.: Nisswa, S & Hd 8790. Missouri. Washington Co.: MO-8 6.6 km W of Potosi, S, Su & Ah 9391]. isconsin. Bayfield Co.: WI-27 21.3 km S of Co. Rd.-A, S of Barnes, S 9085. Citssonss godfreyi Semple f. viridis Semple. 2n = 10. U.S.A. Florida. Bay Co.: Tyndall .B., NCO Beach East, progeny of A. Johnson 8016a. G. mariana a ) Ell. ton = 8. U.S.A. Florida. Liberty Co.: FL-20 opposite Hosford t, S & Godfrey 3101. Erigeron aff | onal Hall. 2n = 18. U.S.A. California. Tulare Co.: For. Rd. J-41 E of Johnsondale, 11.5 km W of Sherman Pass Summit, S & Hd 8652. E. acris L. ssp. politus (E. Fries) Schinz & Keller. 2n = 18. CANADA. Alberta. Jasper Nat’l. Park: ca. 3 mi. W of Jasper, S & Bt 4328. E. divergens Torr. & Gray. 2n = 18. U.S.A. California. Mono Co.: N of Lee Vining, shore of Mono L., S & Hd 8719. 2n = 36. New Mexico. Taos Co.: NW of Holman, NM-3, MP53, Sipapu Ski Area, S & Hd 8069. E. eatonii Gray. 2n = 9,. U.S.A. Wyoming. Albany Co.: Medicine Bow Nat'l. For., S of Lincoln Monument (I-80 at Happy Jack Rd.), S 8797 E. foliosus Nutt. var. hartwegii (Greene) Jeps. 2n = 18. U.S.A. California. Del Norte Co.: US-199 W of Gasquet, S & Hd 8520. E. — Greene var. formosissimus. 2n = 18. U.S.A. — Coconino of Flagstaff, SnoBowl Rd., below ski area, S & Ch 900 E. formosissimus Greene var. viscidus (Rydb. ) Cronq. 2n = 18. US. "7 New Mex- ico. Taos Co.: N of Vadito, NM-3 3.5 km N of NM-75, S & Hd 8065. E. inornatus (Gray) Gray. 2n = 18. U.S.A. California. Siskiyou Co.: SW of Etna, S of Etna Summit, S & Hd 8485. E. simplex Greene. 2n = 18. U.S.A. Colorado. — Co.: W of Poncha Springs, US-50 4.4 km E of Monarch Pass, S & Hd 7 E. speciosus (Lindl.) DC. 2n = 18. U.S.A. rors — Co.: N of Flagstaff, Schultz Pass Rd., N of US-180, S & Ch 9006. E. subtrinervis Rydb. 2 = 18. U.S.A. New Mexico. Lincoln Co.: W of Alto, NM- 532, Sierra Blanca Ski Area, S & Hd 8094. E. ee 4 Ponte (Salisb.) Nutt. 27 = 18. U.S.A. Delaware. Sussex Co.: US- of Georgetown, 3 km W of DL-30, S & R 7643. Wisconsin. Douglas = fe) Barnes, pond near WI-27, 2.1 km N of Co. Rd.-A, S 9077. Grindelia camporum Greene. 2n = 12. U.S.A. California. Glenn Co.: NW of Alder 54 Rhodora [Vol. 94 Table 1. Continued. Springs, For. Hwy.-7, S & Hd 8564. Stanislaus Co.: W of Patterson, Co. Rd. J-17 2.3 km E of Minnear Campground, Frank Raines Park, S & Hd 8584. Haplopappus gracilis Nutt. 2n = 4. U.S.A. New Mexico. Catron Co.: US-180 W of L .5 mi. E of state line, S & Hd 8029. Hz. Saaarvosus Hook. & Arn. 2” = 10. U.S.A. California. Monterey Co.: Rd. G-14 S of King City, just N of Liggett Military Base, S & Ch 8939, S of Big Sur, E of Pfeiffer Beach, S & Ch 8930. Ventura Co.: N of Ventura, Santa Ana Rd. W of CA-33, S. & Ch — Hoeternthera nners var. glandulissimum Semple. 2n = 36. U.S.A. Kentucky. Warren Co: US213 10.9 km SE of I-65, S of seta Green, S & Ch 9 H. psammophila Se cenieith. 2n = 18. U.S.A. Arizona. La Paz Co.: US-60 just W of Wendon, S & Ch 8994. Isocoma veneta (H.B.K.) Greene. 2n = 24. U.S.A. California. Santa Barbara Co.: -1 ca. 1 mi. S of CA-135, MP30.1, S & Ch 8964. San Diego Co.: Del Mar, near beach, S & Ch 8992. Leucelene ericoides (Torr.) Greene. 2n = 32. U.S.A. Arizona. Coconino Co.: SW of Winslow, AZ-87 6.6 km SW of county line, S & Hd 7913. NMamannuthene eee H.B.K. 2n = 4,. U.S.A. Arizona. Apache Co.: cCarrell Rd by I-40, ca 3 mi. W of Chamber, S, Su & Ah 9367 — ger (Pursh) Nutt. 2n = 18. U.S.A. Massachusetts. Barnstable Co: of MA-137 and Long Pond Rd., S 3366. ieee hes (Nutt.) Nutt. 2n = 12. U.S.A. Texas. Howard Co.: S of Big Spring, FR-700, S & Hd 8210 pr hens apargioides (Gray) Greene, 2n = 12. U.S.A. California. Mono Co.: CA-108 a few km W of US-395, picnic area W of Walker R., S & Hd 8722. e. ge (Hook.) Greene. 2n = 24. U.S.A. Utah. Emery Co.: UT-29 15.1 km of UT-57, W of Orangeville, S & Hd 7802. joe altissima L. var. altissima. 2n = 54. U.S.A. Nebraska. Jefferson Co.: US-136 3 mi. N of county line, S & Bt 7349. S. altissima L. var. gilvocanescens (Rydb.) Semple. 2n = 36. U.S.A. Minnesota. Olmsted Co.: US-14 (MP206), SW of Rochester, S 9056. S. arguta Ait. 2n = 18. U.S.A. North Carolina. Transylvania Co.: US-276 near Connestee Falls, S & Ch 6174. S. caesia L. 2n = 18. U.S.A. Delaware. New Castle Co.: DL-82 S of Yorkin, ¢4. 4 km W of DL-52, S & R 7620. Kentucky. Adair Co.: NW of Toria, beech woods near Cumberland Parkway (MP38.1), S & Su 9445. Christian Co.: KY-109 SE of Dawson Spring, N end of Pennyrile Forest St. Pk, S & Su 9432. Maryland. Washington Co.: US-40 N of Greenbrier State Park, S & R a i Ee ee _ Co.: PA-372, below W side of bridge at Sus- aR.,S&R7 S. californica Nutt. 2n = “s % S.A. California. San Benito Co.: W of Hernandez, ga Rd., Lorenzo Vasques Cyn., S, Su & Ah 9346. S. Pi Ss L. aff. var. canadensis. 2n = 36. U.S.A. Tennessee. Davidson Co.: junction of TN-171 and US-41, S & Ch 9124. S. canadensis L. ssp. salebrosa (Piper) Keck. 2n = 18. U.S.A. Wyoming. Teton a 1992] Semple et al.— Astereae 55 Table 1. Continued. Co.: US-26 0.3 mi. E of Teton N.F. boundary, S & Bt 7219. 2n = 36. Colorado. Gunnison Co.: N of Gunnison, CO-135 7.3 km S of Almont, S & Hd 7752. 2n = 54. Montana. Musselshell Co.: US-12 4.2 mi. W of Musselshell, S & Bt 6993. S. confinis Gray. 2n = 18. U.S.A. California. Kern Co.: E of Bakersfield, Kern R. yon, Democrat Flat springs, S, Su & Ah 9362. Ventura Co.: N end of Fillmore, S & Ch oui 0. along Clear Ce, f Clear Creek Rd. 4.65 ker NE of Coulieas Rd., S, Su & Ah 9349, 7.15 km NE of Coalinga Rd., S, Su, & Ah 9352. 2n = 18. California. San Benito Co.: SW of New Idria, Clear Creek Rd. 7.3 km NE of Coalinga Rd., along Clear Creek, S, Su & Ah 9354, S, Su & Ah 9355. S. hispida Muhl. var. tonsa Fern. 2n = 18. CANADA. Ontario. Lambton Co.: Pinery Prov. Park, Bakows S. juncea Ait. 2n = 18. U.S.A. Minnesota. Sherburne Co.: US-169 near Mississippi R., S of Elk River, S 9060. S. multiradiata Ait. 2n = 18. CANADA. British Columbia. S of Chilkat Pass, Haines Hwy., Jarvis Glacier area, Mt. McDonnell, S of Seltat Creek, C h et al. CC3920. S. nana Gray. 2n = 18. U.S.A. Colorado. Gunnison Co.: Kebler Pass Rd. 14.8 km E of CO-133, E of Somerset, S & Hd 7779. (This is a correction of a previous report for S. sparsiflora in Semple and Chmielewski 1987; see text for discussion. S. parryi (Gray) Greene. 2n = 18. U.S.A. Utah. Sevier Co.: ca. 18 km S of Gooseberry, near Gooseberry Rd., S & Ch 8895 S. petiolaris Ait. var. petiolaris. 2n = 36. U.S.A. North Carolina. Richmond Co.: US-1, 4.6 mi. S of NC-220, S, Brammall & Hart 3044. S. ptarmicoides (Nees) Boivin. 2n = 18. U.S.A. Wisconsin. Douglas Co.: WNW of S. rigida L. ssp. humilis (Porter) Heard & Semple. 2n = 18. U.S.A. Minnesota. Sherburne Co.: US-169 near Mississippi R., S of Elk River, S 9057. S. rigida L. ssp. rigida. 2n = 36. U.S.A. Illinois. Livingston Co.: E of Forrest, between US-24 and RR, S & Bt 7386. Stephenson Co.: US-20 E of Freeport, S 9050. Indiana. Benton Co.: US-52 just NW of Fowler, S & Hd 8335. Minnesota. Sherburne Co.: Olmstead Co.: US-14 (MP206), SW of Rochester, S 9054. S. rugosa Mill. 2n = 18. CANADA. Ontario. Wellington Co.: Arkell, S 7397. 2n = 36. U.S.A. Delaware. Sussex Co.: US-9 3 km W of DL-30, E of Georgetown, S & R 7641. Maryland. Washington Co.: US-40 N of Greenbrier State Park, S & R 7655. 2n = 54. New York. Chattaraugus Co.: N of Ellicottville, S & St. S. sempervirens L. 2n = 18. U.S.A. Delaware. Sussex Co.: near DL-1 , by Delaware Sea Shore State Park, S & R 7648. Maryland. US-301 1 km E of Drawbnidge, N of Grasonville, S & R 7633. S. — Kunth ssp. simplex var. simplex. (syn: S. glutinosa var. glutinosa). 2n 18. CANADA. British Columbia. Alaska Hwy., KP600, Teska River flats, 56 Rhodora [Vol. 94 Table 1. Continued. just N of Teska service station, Ch et al. CC3773, Alaska Hwy., 21 km SE of Fireside, Ch et al., CC3783; Cassiar Hwy., 6 km N of Boya L. Provincial Park turnoff, Ch et al., CC4720; Cassiar Hwy., KP729.8, at Yukon border, Ch et al., CC4709; 12 km N of Iskutm just N of Tsasbye Creek, Ch et al., CC4752; Telegraph Creek Rd., 1 km NE of Telegraph Creek, Ch et al., CC4738; Telegraph Creek Rd., 40 km S of Dease L., , Ch et al., CC4735; Telegraph Creek Rd., 5 km S of Dease L., Chet al., CC4731. Yukon. Campbell Hwy., KP570, 21 km SE of Carmacks Junction, Ch et al., CC 4606; Campbell Hwy., KP560, just N of turnoff to Frenchman L. Gov’t. Campground, Ch et al., CC4608; Campbell Hwy., KP539, N of turnoff to Frenchman L. Gov't. Campground, Ch et al., CC4610; Campbell Hwy., KP410, 10 km SE of Faro Junction, Ch et al., CC4614; S of Whitehorse, Klondike Loop 2, KP108.3, Ch et al., CC3883; Klondike Loop 2, 2 km N of MacGregor Creek, KP406- 407, Ch et al., CC4604; 5 km S of Pelly Crossing, Klondike Loop, KP460, Chet al., CC4600. U.S.A. Alaska. N of Dot L., Alaska Hwy., KP1367.5, Ch et al., CC4470; Dempster Hwy., KP12.2, Ch et al., CC4561; Glenn Hwy., 1 km E of Tolsona Lake Rd., W of Tolsona Creek, Ch et al., CC4102; Siver Trail Hwy., NE of Stewart Crossing, KP19, Ch et al., CC4563. S. ngs Gray. 2n = 18. U.S.A. Texas. Culberson Co.: Guadalupe Mountain 1. Park, Bowl Trail, 7700’, S & Hd 8180. 2n = 36. Arizona. Coconino a .. N of Flagstaff, SnoBowl Rd., below ski area, S & Ch 9003. S. spectabilis (D.C. Eaton) Gray. 2n = 18. U.S.A. California. Modoc Co.: CA- 5 km W of Alturas, N of Rattlesnake Butte, S, Su & Ah 9299. Shasta Co.: CA-299, wi Creek Park, Shasta N.F., near Pit R., bank of Hat Creek, S, Su & Ah 9 Ss ere Nutt. = 18. U.S.A. Minnesota. Aitkin Co.: SW of McGrath, bog -18 4km E of MN-65, S 9067. Wisconsin. Portage Co.: NE of Stevens ee WI-10 E of WI-34, marsh, S 9086. S. wrightii Gray. 2n = 9,,. U.S.A. Arizona. Pima Co.: Mt. Lemmon, near summit, S, Hd & Love 7929, 2n = 18. Arizona. Yavapai Co.: just E of Prescott, US- 89 (MP317), S & Ch 9000. New Mexico. Grant Co.: NM-15 39.2 km N of US-180, N of Silver City, S & Ch 9 Virgulaster ascendens (Lindl.) Semple. 2n = Be: U.S.A. Arizona. Coconino Co.: Spring Valley, 6 mi N of Parks, S et al. 5563. 2n = 26. California. Sierra Co.: CA-49 W of Yuba Pass, campground, S & Hd 8407. 2n = 52. Utah. Salt Lake Co.: E of Midvale, UT-190, lower end of Big Cottonwood Cyn, S, Su & Ah 9239. Virgulus x amethystinus (Nutt.) Reveal & Keener (V. ericoides x V. novae-an- gliae). 2n = 10. CANADA. Ontario. Waterloo Co.: Wellesley Twp., by RR, S & Ch 9093. ¥ —_ a Reveal & Keener. 2n = 10. CANADA. British Columbia. ie km N of Edgemont, 12 km N of Radium, Ch CC4899. U.S.A. a. Mono Co.: town of Mammoth Lakes, S & Hd 8685. V. rerio L) Reveal 2 a 2n = 8. U.S.A. Florida. Leon Co.: W of Tal- assee, Godfrey 82228. V. ae. (L.) Reveal ne Keener var. ericoides. 2n = 10. CANADA. Ontario. 1992] Semple et al.— Astereae 57 Table 1. Continued. Waterloo Reg. Mun.: Wellesley Twp, S & Ch 9094. 2n = 20. U.S.A. Illinois. Winnebago Co.: Kilbuck Bluffs Forest Reserve, SW of Rockford, S 9048. V. ericoides var. pansus (Blake) Reveal & Keener. 2n = 10. U.S.A. Arizona. Yavapai Co.: US-89 NE of Peoples Valley, S & Ch 8995. Utah. Uintah Co.: US-191 2.2 km N of Vernal, S & Ch 8874. V. falcatus (Lindl.) Reveal & Keener. 2n = 30. U.S.A. Colorado. Boulder Co.: E of Lyons, S et al. 5815. V. patens (Ait.) Reveal & Keener var. patens. 2n = 20. U.S.A. Georgia. Gordon Co.; S bound rest area, I-75 N of Calhoun, S et al. 7403. V. patens var. patentissimus (Torr. & Gray) Reveal & Keener, 2n = 20. U.S.A. Ar . Yell Co.: E of Danville, AR-10 13.9 km W of AR-7, S & Hd 8288. Xylorhiza tortifolia (Torr. & Gray) Greene. 2n = 12. U.S.A. Utah. Garfield Co.: UT-95 at View Point Rd., W of L. Powell and Hite Marina, S 88/8. et al., 1983), Semple et al. (1983) noted that diploids occurred throughout the species’ range. The few known tetraploids came from its western margin near the prairies in central Canada and the midwestern United States. Authenticity of the one tetraploid report from northeastern Iowa (Semple et al., 1983) has come into question, and thus it should be accepted only with caution. We have discovered that the plant in the greenhouse at Waterloo confirmed to be tetraploid A. puniceus had the same collection number as a second plant, a tetraploid A. ontarionis; thus the lowa provenance of the A. puni- ceus plant is not certain. Since 1983, 75 additional sites have been sampled by this lab- oratory or others, and the distribution of diploids and tetraploids conforms to the pattern illustrated by Semple et al. (1983). Dip- loids were found at 74 sites in Nova Scotia and Ontario in Canada, and Maine, Michigan, Minnesota, and Virginia in the United States (Table 1; Hill, 1983; Semple, 1985; Chmielewski, 1987). An additional tetraploid (with the /ucidulus morphotype) was found in eastern Minnesota (Table 1) near the edge of the range. Thus, although one of the four previous tetraploid reports 1s ques- tioned, the new report replaces it to maintain the pattern as first presented. Aster sibiricus and A. meritus In North America, Aster sibiricus L. was found to be diploid 2n = 18 at 43 sites in Alaska, Alberta, British Columbia, North- 58 Rhodora [Vol. 94 west Territories, and the Yukon (Table 1; Mulligan et al., 1972: Semple and Brouillet, 1980; Chinnappa and Chmielewski, 198 Tk Aster meritus A. Nels. has been reported to be tetraploid at four sites in Montana and northwestern Wyoming (Semple et al., 1983; Semple, 1985). The habitats and ranges of the two species do not appear to overlap. Aster sibiricus is an alpine/tundra taxon of northern Eurasia, Alaska, and western Canada. Aster meritus is a montane to subalpine taxon of the northcentral Rocky Moun- tains in the United States; it has been treated as A. richardsonii Spreng. var. meritus (A. Nels.) Raup and A. sibiricus L. var. meri- tus (A. Nels.) Raup. Although the two taxa are closely related, we feel species status is more appropriate for the more southern member of the complex. Additional study, however, is recom- mended to clarify the status of A. meritus and its relationship to A. radulinus. Solidago guiradonis, S. confinis and S. spectabilis Solidago guiradonis, S. confinis, and S. spectabilis clearly form a closely related species complex that is the far western element of the S. missouriensis/S. juncea group of goldenrods. Guirado’s goldenrod in the strict sense is an endemic restricted to ultramafic arroyo and creek margins in southern San Benito County, Cali- fornia and adjacent Fresno County. It is morphologically close to and perhaps conspecific with Solidago confinis, which occurs in usually boggy or marshy ground and creek margins from San Benito County southward in the coastal ranges and northern Inyo County southward in the Sierras to northern Baja California. Solbrig et al. (1964), Beaudry (1969) and Semple et al. (1989) reported S. confinis to be diploid in southern California. The two species differ primarily in basal leaf width and in branching pat- tern in the club-shaped capitulescence; S. guiradonis is narrower in both traits. The counts of 2n = 9, and 2n = 18 (Table 1) are the first reports for S. guiradonis in the strictest sense. However, a report by Raven et al. (1960) for S. confinis from San Benito County (Solbrig 2816 ps!, cas!) may represent the first report for S. guiradonis, if a broader interpretation of the latter species is adopted. Based on field observations of a number of small an large populations of both taxa, there appears to be only a limited morphological basis on which to separate the two. Character states for diagnostic traits of Solidago guiradonis appear only to be extremes at the lower limits of much greater ranges of variation 1992] Semple et al.— Astereae 59 included in S. confinis. It would not be unreasonable to merge the two species under the older name S. guiradonis and treat the two at some infraspecific rank. The significance of environmental factors on morphology needs further study; S. guiradonis s.s. is endemic to special ultramafic soils and S. confinis is not. For clarity we have presented our reports under two names. Individuals of Solidago confinis can be difficult to separate from S. spectabilis (D.C. Eaton) Gray, which is native to moist habitats in the Great Basin area and adjacent Sierra Nevadas. The ranges of the two overlap in Inyo and Kern Counties in California. Broad phyllary tips with flat margins are generally diagnostic features of S. spectabilis; both S. confinis and S. guiradonis have narrow phyllary apices with inrolled margins. The number of ray florets does not appear to be diagnostic for any of these three taxa. A multivariate analysis of morphological variation should be un- dertaken to determine which traits best distinguish the three taxa. Solidago sparsiflora Solidago sparsiflora is one of several closely related goldenrod taxa native to the southwestern United States and Mexico; the other two are S. californica Nutt. and S. velutina DC. Taylor and Taylor (1984) treated S. sparsiflora as a subspecies of S. velutina with the combination S. velutina ssp. nevadensis (A. Gray) Taylor and Taylor. Nesom (1989b) noted that the complex included additional Mexican species, and Semple et al. (1990) noted the need for additional study of the complex, which is similar to S. nemoralis of the prairies and woodlands of eastern North Amer- ica. Each member of the complex has a distinct geographic range, and thus the cytogeography of each can be examined separately with caution. Regardless of the epithet applied, a sufficient portion of the range has been sampled to permit some generalizations regarding the cytogeography of S. sparsiflora. In total, counts have been reported for 38 locations from the United States (Table 1; Raven et al., 1960; Anderson et al., 1974; Ward and Spellenberg, 1986, as S. velutina ssp. nevadensis; Semple et al., 1984; Semple, 1985; Semple and Chmielewski, 1987; Semple et al., 1989). Dip- loids (2n = 18) are known from Arizona (12 sites throughout the state), Colorado (six sites generally in the central part of the state), Nevada (one site west of Las Vegas), Texas (Guadalupe Mt.), and Utah (four sites). Tetraploids (27 = 36) have been less frequently collected: Arizona (one site near Flagstaff), New Mexico (four 60 Rhodora [Vol. 94 sites), Nevada (one site near Utah), Utah (two sites in the Wasatch Plateau), and Wyoming (one site in the Big Horn Mts.). The single report for a hexaploid (2 = 54) is from a site west of Boulder, Colorado. Polyploids occur throughout the range in the United States, but represent a greater portion of the sample in the north- ern part of the range (five out of the 15 sites north of the 37th parallel) than in the southern part of the range of the species (four out of 24 sites south of the 37th parallel). In addition, the species has been reported to be diploid at three locations in central Mexico (Ward and Spellenberg, 1986). Any study meant to clarify the most suitable rank to assign S. sparsiflora will have to take into consideration the fact that the taxon occurs at three ploidy levels. Brammall and Semple (1990) and Semple et al. (1990) determined that in the related S. nemoralis complex ploidy level was signif- icant in distinguishing subspecies but was not diagnostic. In the course of examining vouchers for this report, the voucher for one diploid report for S. sparsiflora (Semple and Chmielewski, 1987) was determined to be S. nana (location given in Table 1). Semple and Heard 7799 (wat!) has the more corymbiform capitu- lescence and the short dense pubescence on its stems and leaves generally characteristic of S. nana. The voucher was collected well after blooming; the involucres at flowering would probably have been larger than normal for S. sparsiflora. The voucher is part of the same taxon as the voucher for the only other report for S. nana (Semple et al., 1989), which was also diploid. Solidago simplex Ringius and Semple (1987) discussed the cytogeography of this species throughout its range under the synonym Solidago gluti- nosa. Nesom (1989a) noted that the types of S. simplex and S. glutinosa were conspecific and that the former name was older and had nomenclatural priority. Ringius and Semple (1987) de- termined that all reports for ssp. simplex were diploid. The 19 reports for diploids from Alaska and western Canada (Table 1) further confirm that ssp. simplex is diploid. ACKNOWLEDGMENTS This work was supported by Natural Sciences and Engineering Research Council of Canada Operating Grants to J.C.S. The fol- 1992] Semple et al.—Astereae 61 lowing people are thanked for their assistance in the field: T. Ahmed, B. Axon, W. Bessey, D. Bradley, R. Brammall, L. Brouil- let, D. Cameron, J. Canne, C. C. Chinnappa, D. Farris, C. Hart, S. Heard, B. Johnson, R. Keir, T. Love, G. Ringius, B. Semple, B. Smith, and B. A. Suripto. Viable achenes and rootstocks pro- vided by W. Bakowsky, J. K. Morton, R. K. Godfrey, and A. Johnson were much appreciated. Collections made in Jasper Na- tional Park and Banff National Park in Alberta and Guadalupe Mountain National Park in Texas were done so with permission of Parks Canada and the United States National Park Service. Guy Nesom and R. Hartman are thanked for assistance in iden- tifying collections of some taxa. Luc Brouillet is thanked for pro- viding several unpublished counts of taxa in Aster. LITERATURE CITED ANDERSON, L. C., D. W. KyHos, T. Mosquin, A. M. PowELL AND P. H. RAVEN 1974. Chromosome numbers in ————— IX. Haplopappus and other Astereae. Amer. J. Bot. 61: 665-671 BEAuUpDRY,J.R. 1969. Etudes sur les Solidago IX. Une Troisieme liste de nombres chromosomiques des taxons du genre Solidago et de certains genres voisins. Naturaliste Canad. 97: 431-445. BRAMMALL, R. A. AND J. C. SEMPLE. 1990. The cytotaxonomy of Solidago nemoralis (Compositae: Astereae). Canad. J. Bot. 68: 2065-2069. CHINNAPPA, C. C. AND J. G. CHMIELEWSKI. 1987. Documented plant chromo- some sniaibers - 1. Miscellaneous counts from western North America Sida 12: 409-4 CHMIELEwsKI, J. G. "te Cytogeographic studies on North American asters. III. Asters of southern Ontario. Rhodora 89: 41-45. Hitt, L. M. 1983. Chromosome numbers of twelve species of Aster (Asteraceae) om Virginia. Castanea 48: 212-217. Huziwara, Y. 1958. Karyotype analysis in some genera of Compositae. V. The chromosomes of American Aster species. Jap. J. Genet. 33: 129-137. sie = bad baie Data on Ehremmasnne en in Aster (Asteraceae), the st ps of certain North American specie Beinn 32: 240-261. Léve, A. AND D. Léve. 1964. Chromosome number of plants. I. Taxon 13: 99- 110. MULLIGAN, G. A., W. J. Copy AND N. GRAINGER. 1972. IOPB chromosome number nek XXXVII. Taxon 21: 498-499. Nesom, G. L. 1989a. Solidago simplex (Compositae: Astereae), the correct name for S. glutinosa. Phytologia 67(2): 142-14 989b. Taxonomy of Solidago velutina (Asteraceae: Astereae) with a new related ae from Mexico. Phytologia 67: 297- Raven, P., O. SoLBriG, D. KyHos AND R. SNow. 1960. Chromosome numbers in eee: L Astereae. Amer. J. Bot. 47: 124-132. 62 Rhodora [Vol. 94 Rinatus, G. S. AND J. C. SEMPLE. 1987. Cytogeography of the Solidago spathu- lata-glutinosa complex (Compositae: Astereae). Canad. J. Bot. 65: 2458- 2462. SEMPLE, : i pate Chromosome number reports LXXIl. Taxon 30: 703-704. Fam. Compositae, Tribe —~ ar ie 87: 517-527. ROUILLET. 1980. Chromosome numbers and satellite chromo- some eeephalony in Aster and Lasallea. Amer. J. Bot. 67: 1027-1039. [Lasal- lea = Virgulus] AND J. G. CHMIELEWSKI. 1987. Chromosome numbers in Fam. Com- positae, Tribe Astereae. II. Additional Counts. Rhodora 89: 319-325. , R. A. BRAMMALL AND J. G. CHMIELEWSKI. 1981. Chromosome numbers of goldenrods, Euthamia and Solidago (Compositae-Astereae). Canad. J. Bot. 59: 1167-1173. , J. G. CHMIELEWSKI AND R. A. BRAMMALL. 1990. A multivariate study of Solidago nemoralis (Compositae: Astereae) and comparison with S. cal- ifornica and S. sparsiflora. Canad. J. Bot. 68: 2070-2082. HINNAPPA. 1983. Chromosome number determi- nations in Aster L. (Ganpostae) with comments on cytogeography, phylog- eny and chromosome morphology. Amer. J. Bot. 70: 1432-1443. AND Lane. 1989. Chromosome numbers in Fam. Com- positae, Tribe Astereae. III. Additional counts and comments on generic limits and ancestral base numbers. Rhodora 91: 296-314 —., 2G. Runes, C. LEEDER AND G. Morton. 1984. Chromosome numbers 1 Solidago (Compositae-Astereae). II. Additional counts with comments on cytogeography. Brittonia 36: 280-292. Socsric, O., L. ANDERSON, D. KyHos, P. RAVEN AND L. RUDENBERG. 1964. oe numbers in Compositae V. Astereae II. Amer. J. Bot. 51: 513- es Cc. E.S. AND R. J. Taytor. 1984. Solidago (Asteraceae) in Oklahoma and Texas. Sida 10: 223-251. VAN —_ P. 1963. Cytotaxonomic studies in Michigan asters. Michigan Bot. 7-27. ——— AND F. F. Sterx. 1973. Chromosome numbers in Aster. Rhodora 75: 26-33. Warp, D. E. anp R. W. SPELLENBERG. 1986. Chromosome counts of angio- sperms of western North America. Phytologia 61: 119-125. i. C. S. AND C. X. DEPARTMENT OF BIOLOGY UNIVERSITY OF WATERLOO WATERLOO, ONTARIO CANADA N2L 3G1 IG. € BIOLOGY DEPARTMENT SLIPPERY ROCK UNIVERSITY SLIPPERY ROCK, PA 16057 U.S.A. RHODORA, Vol. 94, No. 877, pp. 63-97, 1992 A FLORISTIC AND VEGETATION ANALYSIS OF A FRESHWATER TIDAL MARSH ON THE MERRIMACK RIVER, WEST NEWBURY, MASSACHUSETTS! FREDRICKA ANN CALDWELL AND GARRETT E. CROW ABSTRACT Plant community structure and selected physical parameters influencing vege- tational patterns were studied in a Merrimack River freshwater tidal marsh. A flora of 88 vascular plant species was documented. Special consideration was given to three species, Scirpus fluviatilis, Bidens eatonii and E riocaulon parkeri, listed as rare for the Commonwealth of Massachusetts. Species abundance data Wwera At be a“ os ** 1 ‘“ : +h ter } y wa + y the i 7 program TWINSPAN. Key Words: Freshwater tidal marsh, plant community, plant classification, TWINSPAN, rare plants, northeastern Massachusetts INTRODUCTION Freshwater tidal marshes develop in coastal rivers flooded daily by the incoming tides. The advancing salt front produces a salinity gradient from the ocean upstream to that portion of the river no longer under tidal influence. Coastal wetlands along this gradient are Classified on a salinity-based system which defines a freshwater tidal marsh as one with an average annual salinity of less than 5%o (parts per thousand) (Odum et al., 1984). These freshwater or near-freshwater conditions occur because the tidal influence exceeds the advance of the underlying salt wedge, causing a back- flow of the overlying freshwater into the marshes. Freshwater tidal marshes occur where the salt front reaches its upstream limit in late summer when the river discharge is low. . The fluctuating physical conditions induced by changing tides make a freshwater tidal marsh a unique environment for vege- tation. Several studies have been conducted describing the dis- tribution and abundance of intertidal vascular plants from these marshes (Nichols, 1920; Fassett, 1928; Ferren and Schuyler, 1980, Metzler and Rosza, 1982; Odum etal., 1984; Odum, 1988). Fresh- ' Scientific contribution No. 1719 from the New Hampshire Agricultural Ex- periment Station. 63 64 Rhodora [Vol. 94 water tidal marshes have been described by several authors as transition zones between fresh and salt water habitats (Philipp and Brown, 1965; Anderson et al., 1968; Ferren, 1976; Garofalo, 1980; Haramis and Carter, 1983), and species diversity across this transition zone typically decreases as salinity increases (Odum, 1988). A number of unusual species restricted to estuaries, as well as rare plants, have been documented from this type of marsh (Fernald, 1903; Fassett, 1925a, 1925b, 1928; Mathieson and Fra- lick, 1973; Ferren and Schuyler, 1980; Crow, 1982; Sorrie, 1987; Barrett, 1989 MS thesis, Univ. of Conn., Storrs, CT). The forces of strong tidal and river currents and ice floes have a strong influence on the morphology and vegetation of a fresh- water tidal marsh. Marsh soil is directly affected by sediment accretion and erosion as well as by other factors (Ahnert, 1960; Dionne, 1968, 1969, 1974; Garofalo, 1980; Serodes and Troude, 1984). Zonal patterns of vegetation may develop in response to en- vironmental gradients in freshwater tidal marshes. Some studies have correlated plant community structure and productivity with certain environmental parameters such as elevation of the marsh, tidal submergence, salinity, soil texture, soil organic material and soil redox potential (Disraeli and Fonda, 1979; Hutchinson, 1982; Ewing, 1983, 1986). Freshwater tidal marshes occur on both coasts in North Amer- ica, but the Pacific coast marshes are not as extensive as those along the Atlantic (Disraeli and Fonda, 1979). In Massachusetts, two river systems support sizable freshwater tidal marshes, the North River (Plymouth Co.) and the Merrimack River (Essex Co.). The Merrimack River arises in the White Mountains of New Hampshire and empties into the ocean near Newburyport, Mas- sachusetts, where the Merrimack forms a wide estuary at its con- fluence with the Parker River, north of Plum Island. This study of the Merrimack River freshwater tidal marsh was initiated with the following objectives: 1) to conduct a complete floristic survey of the area: 2) to collect species abundance data and classify the vegetation into cover types; 3) to document pop- ulations of plants listed as rare for the Commonwealth of Mas- sachusetts; 4) to describe the marsh habitat, including plant cover types, tidal fluctuation, free water salinity, and soil organic matter content; and 5) to assess the relationship between plant com- munity structure and selected environmental characteristics. 1992] Caldwell and Crow—Freshwater Tidal Marsh 65 Site Description The study site occurs along a segment of the Merrimack River in the Town of West Newbury, Essex County, Massachusetts, approximately nine miles from the ocean (42°49’N Lat., 71°57'W Long.). The study area begins at the Indian River and extends east for 114 miles to the Artichoke River. Three marsh areas occur between these two tributaries, separated by sections of rocky shoreline (Figure 1). The average area of each marsh is approx- imately 400 x 150 meters. Because the high tide zone borders a sloping forested riverbank, the Merrimack River marsh contains only two small areas where high marsh habitat has developed. These areas occur as slightly elevated islands near the mouths of the Artichoke and Indian Rivers, and appear to be sections of land that have become iso- lated by a backwash of these tributaries from tidal flooding. MATERIALS AND METHODS Environmental Measurements Tidal Amplitude Tidal amplitude was recorded during September and October of 1989 using a Stevens Model F Recorder. Since the Merrimack River is used extensively for boating, it was not advisable to place the recorder in the river channel at the low tide level. A reference location was chosen for the recorder platform in Area #3 (see Figure 1) at the edge of the vegetation nearest the river channel at low tide. Therefore, data do not include full low tide range of the river, but they do measure tidal amplitude and duration of flooding for all vegetation zones in the marsh. Surface elevation of this marsh area relative to the recorder platform was measured using a line level and meter rod. The longest transect in the center of Area 3 was chosen to represent the average elevation of sampled areas. Salinity Free-water salinity measurements were taken in the spring and fall of 1989 using a portable Y-S-I model 33 salinometer. Mea- 66 Rhodora [Vol. 94 AMESBURY MERRIMACK RIVER WEST NEWBURY Figure 1. Map of study site. surements were recorded during the high tide cycle at several sampling stations in the marsh and at various depths within the water column. Temperature and conductivity were also recorded. Soil Organic Matter Content Soil samples were collected in June 1989 from Area #1 for organic matter content determination. Three replicates were col- lected at 5 cm and 10 cm depths at four intervals along a transect corresponding with different plant cover types. Samples were oven- dried at 80°C for 24 hours and ashed in a muffle furnace at 475°C for 24 hours. Percentages of water and organic matter content of the soil were calculated from these weights. Vegetation Analysis Vegetation data were collected in August and September of 1988 and 1989. These months were chosen for vegetation sam- pling because combined patterns of all species produce a pea community biomass in August (Doumlele, 1981). Sampling was done using the stratified random sampling method (Muellet- Dombois and Ellenberg, 1974). Three transects were established 1992] Caldwell and Crow—Freshwater Tidal Marsh 67 in each of the three marsh areas, and 23 transects were established along the rocky shore areas at 50-meter intervals for a total of 32 transects. Each transect ran perpendicular to the edge of the river and extended from the beginning of the rooted vegetation exposed at low tide to the upper edge of the high tide zone. A quadrat size of 2 x % meter was used in a total of 323 sample plots. One quadrat was randomly located within each five meter segment of marsh transect and three meter segment of rocky shore transect. A shorter random sampling segment was chosen for the rocky shore transects because the length of each of these transects was relatively short, and elevational gradient was steeper. Percent cover was recorded to estimate the abundance of 30 vascular plant species occurring in the quadrats. Cover was de- fined as the projection of the crown or shoot area of a species to the ground surface expressed as a percent of the quadrat area (Mueller-Dombois and Ellenberg, 1974). Data were analyzed using TWINSPAN (Two-Way Indicator Species Analysis), a Fortran program designed to construct a clas- sification of samples that are then used to classify species ac- cording to their ecological preference (Hill, 1979). The program groups quadrat samples and species by repeated dichotomies. Three ordinations are used in the dichotomy determination: |) primary ordination, which uses a reciprocal averaging method; 2) refined ordination, which uses differential species; and 3) in- dicator ordination, which uses indicator species (Hill, 1979). Floristics An inventory of the vascular plants of the Merrimack River freshwater tidal marsh was undertaken during the 1988 and 1989 field seasons. Voucher specimens were collected and deposited at NHA. Documentation of three rare plants, Bidens eatonil, Er- iocaulon parkeri and Scirpus fluviatilis and their habitat descrip- tion were also included in this study. Identifications were based on Aquatic and Wetland Plants of Northeastern North America (Crow and Hellquist, in press), A Manual of Vascular Plants of Northeastern United States and Adjacent Canada (Gleason and Cronquist, 1963) and Gray's Man- ual of Botany (Fernald, 1950). Additional references included: New Britton and Brown Illustrated Flora (Gleason, 1952) and 68 Rhodora [Vol. 94 Aquatic Vascular Plants of New England Parts 1-8 (Hellquist and Crow, 1980, 1981, 1982, 1984; Crow and Hellquist, 1981, 1982, 1983, 1985). RESULTS AND DISCUSSION Environmental Measurements Tidal Amplitude The total range of tidal amplitude measured for 76 cycles was 0-150 cm, with high tides ranging from 82-150 cm; mean high tide calculated from all recorded tides was 115 cm. The elevation of the marsh surface was 40 cm, as measured from the base of the tide gauge to approximately mean high tide along a 147 meter transect. A rise in elevation of 15 cm occurred along the first 25 m nearest the river; the remaining 25 cm of rise in elevation occurred over the last 122 m of the marsh. This gradual slope across the marsh platform is typical of freshwater tidal marshes, and has been described by several authors (Metzler and Rosza, 1982; Ewing, 1983; Odum et al., 1984; Mitsch and Gosselink, 1986). Many sections of the high tide zone end in an erosional step that abuts a sloping river bank. This step accounts for much of the difference in high tide water levels and the measured slope of the marsh surface. A similar geomorphology was reported by Hutch- inson (1982) in the Fraser River tidal marshes of British Colum- bia. Other sections end in a forested slope which was beyond the scope of this study. Duration of inundation in the Merrimack River marsh at the elevation of the tide gauge was approximately 16 hours per day during two tide cycles. Mean high tide data (115 cm) were use for this determination. Water level plotted against duration of inundation at the plant/soil interface shows submergence time for different elevations across the marsh platform (Figure 2). Cov- er types such as Scirpus pungens and Spartina alterniflora that occur at the lower edge of the vegetation zone are partially sub- merged for 16 hours a day and fully submerged for a portion of that time, depending on their height and elevation. Decreased levels of light and gas exchange, characteristic of a submerged 1992] Caldwell and Crow— Freshwater Tidal Marsh 69 WATER LEVEL vs DURATION OF INUNDATION——\ 120 MEAN HIGH TIDE 115 CM WATER LEVELS ( CM ) DURATION OF INUNDATION ( HR ) Figure 2. Water level versus duration of inundation at the plant/soil interface for two tide cycles taken from 115 cm mean high tide data. habitat, make this environment suitable for only a select number of emergent plants. The mid-marsh and backmarsh portions of the surface below 40 cm elevation are flooded for at least 11 hours per day at the plant/soil interface (Figure 2). Cover types occurring in this area include Scirpus tabernaemontanii-type, Acorus calamus-type and Zizania aquatica-type. Although tide gradient front-to-back in the marsh seems relatively small, it is probably a significant factor in controlling distribution of these cover types. In a study based on a linear regression analysis of selected physical parameters in the Nooksack River delta in Bellingham Bay, Washington, Disraeli and Fonda (1979) reported that the most important environmental factor affecting plant distribution was elevation above mean low water. They further added that this factor controlled frequency and duration of submergence and indirectly affected all other factors. Using analysis of variance to assess which physical factors governed plant species distribution, Hutchinson (1982) reported that “‘species distribution is clearly controlled by the elevation of the marsh platform and the asso- ciated tidal regime.” Several other authors report similar conclu- sions (Johnson and York, 1915; Nichols, 1920; Philipp and Brown, 1965: Metzler and Rosza, 1982; Ewing, 1983; Keddy, 1983; Cha- breck, 1988). 70 Rhodora [Vol. 94 Table 1. Water analysis data recorded from eight stations at high tide in the Merrimack River marsh study site. Average Average Salinity Conduc- Gee Salinity Range tivity Date 1989 (%o) (%o) (umho) CC) Cr May 20 0.0 ~ — 20 68 June 18 0.0 ~ 106 20 68 Sept 16 1.4 0.8-4.8 1742 21 70 Salinity Except for one sampling period in the fall, all samples showed 0%o salinity. The only detectable salinity occurred in mid-Sep- tember, when it averaged 1.4%o in the water column with a range of 0.8 to 4.8% (Table 1). These readings classify the study site as a freshwater tidal marsh, since by definition, this type of marsh must be tidal and have an average annual salinity of less than 5%o (Odum et al., 1984). The seasonal salinity pattern reported here supports previous data recorded from this area of the Merrimack River. Jerome et al. (1965) reported that during an eight month survey only one salinity measurement of 5.0% was recorded in October from the Artichoke River Station in the Merrimack River. In another study conducted on the Merrimack River, Miller et al. (1971) indicated that the limit of the salt intrusion at high tide varied from 4.3 to 10.9 miles from the mouth depending upon the season. They found gross fluctuations in salinity both daily and seasonally in this section of the river. Disraeli and Fonda (1979) also found a similar salinity pattern in the Nooksack River in Bellingham Bay, Washington, with a maximum salinity of 5.19%e in October. Soil Organic Matter Content Soil organic matter and water content measurements may help to determine the nature of the marsh soil, which varies consid- erably among river systems and marsh types (Chabreck, 1988). Soil samples for these det were collected; organic mat- ter analysis showed a gradient from 1.49 percent in the low tide zone to 19.81 percent in the backmarsh (Table 2). This type of 1992] Caldwell and Crow—Freshwater Tidal Marsh 71 Table 2. Soil analysis data taken at four intervals and at two depths along a transect in Area 1. These data represent an average of three samples. % Organic Content % Water Content Distance from of Soil of Soil Vegetation Edge (Low Tide Zone) = Depth (Meters) 5 cm 10 cm 5 cm 10 cm 5 1.64 1.49 26.73 20.95 a5 2.46 eat 30.89 25.78 75 6.85 4.29 43.66 o.17 125 19.81 17.86 68.09 64.41 gradient is predictable since the action of the tides sweeps away the organic debris from the low- and mid-marsh areas and de- posits it in the backmarsh. Whigham and Simpson (1975) also found that soil organic concentration increased on a gradient from actively flooded stream banks to less actively flooded areas in the Hamilton Marshes in New Jersey. The soil in the Merrimack River foremarsh is mainly gray clay mixed with some silt and sand. Since the organic content of this soil is less than 15 percent, it may be classified as a mineral soil (Dachnowski-Stokes, 1940). By contrast, the backmarsh substrate can be classified as muck since it contains an organic content in the 15 to 50 percent range. Large deposits of silt are also an important component of the mid- and backmarsh soils. 7 Although a great deal of between- and within-site variability exists, average organic content taken from a series of samples from the east coast of North America ranged from 20 to 70 percent, with a mean of 35 percent (Odum, 1988). Along the Delaware River in New Jersey, Whigham and Simpson (1975) Showed a range of 14 to 40 percent organic matter. Relative to these findings, the Merrimack River marsh has a low organic content. These results may be due to the absence of “high marsh’ areas relative to other freshwater tidal marshes. High marsh areas are elevated sections of the marsh that are less exposed to tidal currents. They are usually flooded only during seasonal high tides and, for a variety of reasons, poorly drained. These conditions facilitate the build-up of organic matter in the soil. The low mea- surement may also be a result of insufficient sampling. Water content in the Merrimack River marsh tends to parallel organic content. A front-to-back marsh gradient exists with re- pe: Rhodora [Vol. 94 spect to soil water content, with a range of 27 percent to 68 percent respectively (Table 2). Both Disraeli and Fonda (1979) and Hutch- inson (1982) found that soil water content was highest in the high- and mid-marsh areas at low tide. They attribute these findings to differences in soil texture and drainage patterns from front to backmarsh. There is usually a higher proportion of fine sand in low marsh substrate, resulting in a more rapid drainage of these soils during ebb tide. In contrast, the backmarsh contains a more organic matter and drainage patterns are poorer. In the Merrimack River marsh, backmarsh is typically waterlogged due to gradual surface slope and dense network of rhizomes of Acorus calamus in the mid- marsh section. Differences in percent organic and water content between 5 cm and 10 cm depths are shown in Table 2. In all cases, percent Organic matter and water content decreased with depth. The dy- namics of an intertidal marsh prevent long-term accumulation of organic matter, and consequently less is found at greater depths. Plant Cover Types The vegetation was classified using TWINSPAN into eight plant cover types (Figure 3). Four hierarchical divisions were used to cluster the 323 quadrats and 30 plant species into these eight groups. In the first dichotomy, TWINSPAN separated the 66 quadrats corresponding to the rocky shore sites from the 257 marsh quad- rats. Five plant cover types were further delineated from the anal- ysis of marsh vegetation data and were named according to their dominant vascular plant species. The Spartina alterniflora covet type is the first to be defined on the second divisional level. The Scirpus tabernaemontanii and Acorus calamus cover types are split on the third divisional level, but they are not completely defined until they are separated from the Sagittaria graminea and Zizania aquatica cover types on the fourth divisional level. Di- visions with less than five quadrats are not considered sufficiently distinct to designate as a cover type. In the rocky shore areas, three plant cover types are recognized at three divisional levels, Spartina pectinata cover type at level two and Scirpus pungens and Amaranthus cannabinus cover types at level three (Figure 3). Fewer quadrats were sampled in the 1992] Caldwell and Crow—Freshwater Tidal Marsh a2 323 257 66 Marsh Cover Types Rocky Shore Cover Types 11 246 58 8 Spartina Spartina alterniflora pectinata 87 159 28 30 Amaranthus Scirpus cannabinus pungens 20 iy CF WT -_ 22 Sagittaria Scirpus Acorus Zizania graminea __tabernaemontanii calamus aquatica Figure 3. TWINSPAN analysis showing the eight plant cover types classified at four hierarchical levels. rocky shore sites since these areas comprised a considerably small- er portion of the study site than the three large marshes. Because of the comparatively small number of samples and sparseness of vegetation, it is not meaningful to describe cover types beyond the third divisional level in these areas. . The two habitats identified by the first dichotomy exhibit dif- ferences in topography, substrate and vegetational patterns. The three marsh areas are broad with sections of dense vegetation, a relatively flat surface and a silty-clay organic substrate. Borders of these marshes adjacent to the river channel consist of extensive mud flats that are exposed at low tide and are devoid of vegetation throughout the year. The overriding impression of the vegeta- tional pattern in these marsh areas is one of zonation due to broad bands of dominant species, although the borders may be highly irregular (Figure 4). In contrast, rocky shoreline areas which occur between marshes are characterized by more steeply sloping surfaces, stony-sandy substrates and mosaic patterns of sparse vegetation (Figure 5). 74 Rhodora [Vol. 94 Area 3 1SO RE MARSH COVER TYPES ‘inane HHH Hes) Mid 335 8 8 Gs Bs aS Bs is HHH Mi 5s 55 ss | S58 GSN ER #8 #2 #8 $8 Sa 8 a Ba Sy PSbobebebsbenipebsies SBS SES SESS ESS BSS ES SS SES SS SS NN NE ES 5B BB BT NB I Mt Be See eter —-LI_N TTL HM} Spartina alterniflora 4 Sagittaria graminea Scirpus tabernaemontanii Acorus calamus Zizanié aquatica Figure 4. The distribution of marsh cover types in three marsh areas. Each square represents a 2 x 2 meter quadrat within a 5-meter segment of transect. The transects are depicted vertically and are separated horizontally by distances ranging from 50-150 meters. These shore areas are more exposed to erosive currents, ice scour- ing, and wind than are the marshes, and this high degree of dis- turbance may be a contributing factor to the patchy distribution of vegetation. While each cover type throughout the marsh is characterized by one or more dominant plants, some species such as Polygonum punctatum and Sagittaria latifolia do not have 4 strong affinity for any particular area within the intertidal zone, and are found in several different cover types. 1992] Caldwell and Crow—Freshwater Tidal Marsh ie Area 3 150 ROCKY SHORE COVER TYPES Area | 125 ai Area 2 1ool_|_| # M E nme ea 7; a me — E ra il R zl T € 5 ee) 50 ne x asi_| | RIVER a Amaranthus cannabinus Scirpus pungens Spartina pectinata Figure 5. The distribution of rocky shore cover types. Each square represents a ¥2 x '% meter quadrat within a 3-meter segment of transect. The transects are depicted vertically and are separated horizontally by distances ranging from 50- 150 meters. MARSH PLANT COVER TYPES Spartina alterniflora Cover Type The Spartina alterniflora cover type is one of the less common of the eight plant associations described. The community can be characterized by three species, but Spartina alterniflora domi- nates, with a mean cover of 26 percent (Table 3). _ . Spartina alterniflora is typically found in more saline habitats, but since freshwater tidal marshes are transitional to salt marshes, 76 Rhodora [Vol. 94 able 3. Mean percent cover for 25 species in 8 plant cover types. 1 = Spartina alterniflora. 2 = Sagittaria graminea. 3 = Scirpus tabernaemontanii. 4 = Acorus calamus. 5 = Zizania aquatica. 6 = Spartina pectinata. 7 = Amaranthus can- nabinus. 8 = Scirpus pungens. (Species with <1% cover are omitted.) Cover Types 1 g 3 4 3 6 ii 8 Number of Quadrats lt 20 6 Taf 22 8 28 30 Sagittaria graminea — 38 2 2 4 —- 1 2 Ludwigia palustris — _ -_ _ 1 - - Elodea nuttallii 1 — 3 1 — = - = Zizania aquatica _ - _ —- 42 _ 1 _ Spartina alterniflora 26 — Scirpus tabernaemontanii 9 9 latine americana _ — Sagittaria latifolia — 2 y (4) m 1 5 Ww N — | | | | | — w On —_ | UANNWAWAS! WwW | - | — sh italy punctatum _ 1 = nS § _ Ww WN Ww Sium suave — _ 1 Lythru rum salicaria = = = ms lw < g S23 = g 8 S = ‘ Ss = ~ & | | | | | | ice | | W Om OW + | | tS s Le cy = & 8 x 8 a § "A 8 8 4 at 4 | | | ee: | | | el © | wr | Scirpus pungens = J _ = a this saltwater cordgrass may occasionally be found upstream un- der nearly freshwater conditions (Ferren, 1976). In the Merrimack River marsh, Spartina alterniflora i is most abundant at the sea- ward edge of the study site in the low tide zone. A dense network of rhizomes and roots makes it very resistant to erosive forces, allowing it to colonize unstable areas (Garofalo, 1980). Although Scirpus tabernaemontanii grows with these clumps of cordgrass, it is usually found in the midmarsh areas where it forms nearly homogenous zones. The other components of this cover type; such as Elodea nuttallii and Elatine americana, are submerged aquatics that are tolerant of the alternately flooded and exposed mud flats in the low tide zone. 1992] Caldwell and Crow—Freshwater Tidal Marsh ivi Sagittaria graminea Cover Type The Sagittaria graminea cover type (Table 3) extends from exposed mud flats in the low tide zone to fringes of the backmarsh. The dominant species, Sagittaria graminea, is typically associated with Scirpus pungens and Scirpus tabernaemontanii in the low- to mid-marsh areas, and with Zizania aquatica, Sagittaria lati- folia, Bidens cernua and Polygonum punctatum in the backmarsh. This distribution occurs primarily because Sagittaria graminea has adapted to a wide range of environmental conditions. When Sagittaria graminea occurs in the low tide zone where it is submerged throughout most of the day, the plants form colonies of sterile rosettes with linear-lanceolate phyllodia. In the backmarsh, however, the plants are only occasionally submerged, and it is under these conditions that they develop larger, broader leaf blades and occasionally flowers. Ferren and Schuyler (1980) reported that plants of the upper intertidal zone had several ver- ticils on their inflorescence axes, while those growing in the lower intertidal zone had one to rarely three verticils of flowers. Flow- ering specimens of S. graminea in the Merrimack River marsh usually had one or two verticils. Sagittaria graminea rarely pro- duces achenes in New England (Hellquist and Crow, 1981), and none was observed in this study. Scirpus tabernaemontanii Cover Type This cover type is characterized by the soft-stemmed bulrush, Scirpus tabernaemontanii (Table 3), which forms a broad band from the lower edge of the vegetation zone to the midmarsh area. Other rhizomatous perennials such as Acorus calamus and Sag- ittaria latifolia are included in this association, but they rarely inhabit the low tide zone. The submerged aquatics, Elodea nut- tallii and Elatine americana, grow intermixed with these emer- gents and form small patches on the surface of the mud in fre- quently inundated areas. This cover type is found mainly at sites with a silty clay substrate, and its borders are abrupt when the substrate changes to sand and gravel near the rocky shore areas. The dominant species in this cover type, Scirpus tabernae- montanii, is a rapidly growing perennial which reaches heights of three meters or more. Stored nutrients in the rhizomes may allow 78 Rhodora [Vol. 94 the early and rapid development of plants of this species that obtain their peak biomass in early June along with Acorus calamus (Whigham et al., 1978). The height, abundance and early emer- gence of S. tabernaemontanii make it one of the predominant species of the entire freshwater tidal marsh. Acorus calamus Cover Type The Acorus calamus cover type spans by far the largest area of the Merrimack River marsh, extending from midmarsh to the border of the backmarsh in some sections. The TWINSPAN anal- ysis clustered 137 quadrats and 11 species in separating out this community. Because this cover type encompasses such a large area and tidal range, its species diversity is relatively high, despite the overriding dominance of A. calamus (Table 3). Dense rhi- zomatous colonies of this plant dominate the midmarsh section and may extend up to sixty meters along certain transects. Ap- parently this species reproduces almost exclusively by rhizomes since it is a common plant in these marshes and has not been documented in several seed bank surveys (Leck and Graveline, 1979; Leck and Simpson, 1987; Simpson et al., 1983). The backmarsh elements of this cover type are characterized by the broad-leaved emergents Sagittaria latifolia, Pontederia cor- data, and occasionally Peltandra virginica. These species are most abundant in wet organic mucky substrates. Although Sagittaria latifolia also occurs in the midmarsh, plants are larger and flower more frequently in the backmarsh. Scirpus fluviatilis is another important member of this backmarsh cover type, and is discussed in the rare plant section. The annual plant component of this cover type is found most often intermixed with Acorus calamus. The dense and partially exposed system of rhizomes of 4. calamus gives the marsh an irregular surface which provides a microhabitat for seeds of an- nuals such as Sium suave, Bidens cernua, B. connata, and B. eatonii. In winter, clumps of these rhizomes are uplifted along with ice blocks and deposited downstream along the shoreline (Hardwick-Witman, 1984 MS thesis, Univ. of New Hampshire, Durham, NH). Observations over the two year-study period, however, revealed that erosive forces prevented establishment of these clumps in the low tide zone in the Merrimack River marsh. 1992] Caldwell and Crow—Freshwater Tidal Marsh 79 Zizania aquatica Cover Type This cover type consists of several perennial backmarsh emer- gents, but it is dominated by the late-summer annuals Zizania aquatica and Bidens spp. Zizania aquatica requires soft mud and slowly circulating water typical of a backmarsh habitat (Odum et al., 1984), but it may also be found in small depressions through- out the midmarsh (Figure 4). Other associated species include Sagittaria graminea, Pontederia cordata, Polygonum punctatum, Sium suave, Ludwigia palustris and Eleocharis smallii (Table 3). The late maturity of Zizania aquatica and Bidens makes this cover type the last to appear in the seasonal succession of marsh vegetation. Root growth predominates during early development of Z. aquatica, but later in the season, shoot growth may be up to 6.5 cm per day (Good and Good, 1975); this shallow-rooted grass can reach heights of 4 m. Lodging of large stands of this plant has been observed in the study site due to waves from storms and heavy boat activity on the river. ROCKY SHORE COVER TYPES Spartina pectinata Cover Type This freshwater cordgrass cover type represents a small portion of the rocky shore vegetation. Its definition by TWINSPAN in- volved only eight quadrats. Spartina pectinata, the dominant spe- cies in this cover type, occurs most often along back borders of rocky shore areas in sandy substrate with other species such as, Aster novi-belgii and Lythrum salicaria. It also grows mixed with annuals that are common there (Table 3). The two major peren- nials in this cover type, Spartina pectinata and Aster novi-belgii, have rhizomatous root systems that enable them to withstand the adverse conditions characteristic of this section of the Merrimack River marsh. This band of shoreline is one of the most highly disturbed areas since it is often filled with debris that is shifted back and forth along the shoreline with each changing tide. Amaranthus cannabinus Cover Type The Amaranthus cannabinus cover type is the most diffuse of the eight described. It occurs most often in the rocky shore areas, 80 Rhodora [Vol. 94 but may be found occasionally in backmarsh areas where the substrate is firm and gravelly (Figure 5). In rocky shore areas, this cover type extends from low-to-high tidal elevations, without an apparent zonal pattern. Since the Amaranthus cannabinus cover type is comprised mainly of annuals (Table 3), abundance of the different species may change dramatically from year to year. Leck and Graveline (1979) found that annuals comprised seven of the ten most nu- merous species encountered in seed bank experiments conducted in the Hamilton freshwater-tidal marshes on the Delaware River near Trenton, New Jersey. They further reported that the river- bank, which had been scoured of all vegetation by ice and waves during the winter, produced dense growths of annuals during the summer. In the Merrimack River marsh, these rocky shore areas occur around two points that project into the river and are highly exposed to wind and waves (Figure 1). These extremely disturbed sites provide an open habitat for fast-growing annual species, with the rocky substrate affording the seeds protection from river and tidal currents. Backmarsh occurrence of the Amaranthus can- nabinus cover type may be due to similar stony substrate in these areas. Scirpus pungens Cover Type Scirpus pungens dominates this cover type, intermixing only occasionally with patches of Sagittaria graminea (Table 3). Scir- pus pungens forms a nearly monospecific zone in the low tide mud flats or occurs as scattered populations throughout the rocky Shore areas (Figure 5). The mean cover value of other species associated with this cover type is less than | percent. When this small-stemmed perennial colonizes the lower limit of the low tide zone, it is submerged throughout most of its life. In the spring, the cold temperature of the mud and overlying river water delay shoot emergence until mid-May, a time when Acorus calamus is already beginning to flower. These plants remain somewhat stunt- ed throughout most of the growing season and rarely produce flowers. Deschenes and Serodes (1985) found that Scirpus pungens (cited as Scirpus americanus) can withstand nearly 100 percent submersion under freshwater conditions, but its population size declines with increasing salinity. 1992] Caldwell and Crow— Freshwater Tidal Marsh 81 ch + lass ‘Dus pung ely rapidly on the more elevated sections of rocky shore areas, however, and flowers in early June. This is one of the few rhizomatous plants that colonizes sloping surfaces and stony substrates of the Merrimack River shore areas. Hutchinson (1982) and Odum (1988) considered this species to be a pioneer, capable of quickly colonizing disturbed or bare areas by seed dispersal. RARE PLANT DOCUMENTATION AND HABITAT DESCRIPTION Three plant species listed by Sorrie (1987, 1990) as rare for the Commonwealth of Massachusetts occur in the Merrimack River marsh, including Scirpus fluviatilis (special concern), Eriocaulon parkeri (endangered) and Bidens eatonii (threatened). Scirpus flu- viatilis has a wide distribution in fresh and tidal rivers across the United States. Eriocaulon parkeri and Bidens eatonii are confined to estuaries along the east coast (Crow and Hellquist, in press; Hellquist and Crow, 1982). Six current sites have been documented for Scirpus fluviatilis in Massachusetts, including the Merrimack River freshwater tidal marsh (Sorrie, pers. comm.). These plants occur most often in the middle and backmarsh areas of the study site, rooted in a muddy, silty substrate and growing on the slightly elevated bor- ders of islands in the marsh. Although large stands develop, most plants lack inflorescences, and reproduction is almost exclusively vegetative by means of an extensive rhizome system. Ferren and Schuyler (1980) reported similar observations of S. fluviatilis pop- ulations along the Delaware River. Leck et al. (1988) also reported that although this species occurred in their study area, 1t was not represented in the seed bank survey nor in the field as seedlings; they concluded that it rarely reproduces by seeds. . The second rare plant, Eriocaulon parkeri, is reported in only four locations in Massachusetts, including the Merrimack River marsh (Hellquist and Crow, 1982; Sorrie, pers. comm.). One small population was documented near the west end of the study site, and a second population, reported further down river, could not be located. Three boat docks have been built out from the riv- erbank, and a large section of shoreline clearing occurred during the two years of this study; this activity may account for disap- pearance of the second population. 82 Rhodora [Vol. 94 The estuarine pipewort, Eriocaulon parkeri, grows in a gravelly, sandy substrate on the upper slope of the rocky shore in open sunlight. These plants are submerged for ca. 11 hours a day at high tide; the substrate in this area, although sandy, appears to remain continuously wet even at low tide. Associated vegetation is sparse, consisting of species such as Scirpus smithii, Cyperus bipartitus, Isoetes echinospora, Equisetum fluviatile, Ludwigia palustris and Lindernia dubia. Although the range of Eriocaulon parkeri extends from Maine to South Carolina in estuaries along the coast, it is listed as rare and endangered for New England, Maine, Massachusetts and Connecticut (Crow et al., 1981; Hellquist and Crow, 1982). It is also rare in Delaware, New Jersey and New York (Tucker et al; 1979: Mitchell et al., 1980; Snyder and Vivian, 1981). Efforts should be made to protect the remaining populations of this rare plant and to preserve its habitat. The Merrimack River site is especially vulnerable due to recent and potential marsh destruc- tion. The third rare plant species, Bidens eatonii, was discovered by Alvah A. Eaton in 1902 from the “brackish shores of the Mer- rimack River in Newburyport, Massachusetts” (Fernald, 1903), just downstream from the study area. Since then, eight weakly defined varieties have been described, four of which have been documented from the Merrimack River estuary, var. eatonii Fern., var. fallax Fern., var. illicita Blake and var. kennebecensis Fern. (Fassett, 1925a; Sherff, 1937). In addition to these varieties, three other species of Bidens also occur in the study site, B. cernua, B. frondosa and B. connata. Since Bidens cernua has simple sessile leaves and B. frondosa has compound petiolate leaves, both are easily recognized in the field. However, Bidens eatonii and B. connata are very similar and have simple petiolate leaves that frequently have basal lobes, and flowers that typically lack ray florets. In many cases it was impossible to distinguish between these two species in the field, especially since they were in the vegetative stage at the time of vegetation sampling. Therefore, these two species were grouped together in the abundance data for vegetation analysis. Characters that have been considered important in distinguish- ing these two species include number, size and shape of terminal heads, as well as the surface features of achenes. Bidens connata has large (30-60 flowers) rounded heads, and B. eatonii has small- 1992] Caldwell and Crow— Freshwater Tidal Marsh 83 er (8-30 flowers) campanulate heads (Fassett, 1925a, 1925b). However, in examining many petiolate, simple-leaved specimens of Bidens in the Merrimack River marsh, a range of terminal head sizes and shapes was observed. Sorrie (1987) also found similar intermediate head characters in plants from a Threemile River population in Bristol Co., Massachusetts. Achene surface features such as striations and tubercles also proved to be unsatisfactory as distinguishing characters. Both species have striate achenes (Fassett, 1925b) and the presence of tubercles, often used to distinguish B. connata from B. eatonii, was reported by Fernald (1908) to be an inconstant character. Taxonomic identity could be further complicated in that all four species commonly grow together on tidal shores and in brack- ish marshes, and hybrids have been reported to occur between some of them. Fassett (1925a) described B. x multiceps, a hybrid of B. connata and B. eatonii from the Taunton River estuary of Massachusetts. Also, in a study of Bidens connata in Ontario, Crowe and Parker (1981) suggested that the Thunder Bay pop- ulation, thought to be B. connata, may be an agamospermously- reproducing hybrid between B. cernua and B. frondosa. Both species occur in the same area and the supposed hybrid is inter- mediate in gross morphology between them. They further con- clude that certain other taxa of Bidens which are poorly differ- entiated from B. connata, such as B. eatonii, “may likely be elements of a widespread agamic complex.” As a result of studies in the field and the examination of nu- merous herbarium specimens, there is good reason to question the taxonomy of Bidens eatonii. Whether B. eatonii is truly a distinct species, or whether it is a hybrid, or represents an ecotype or ecophene, remains to be seen. Clearly, additional studies are needed to resolve the taxonomic status of Bidens eatonii in order to determine accurately the taxonomic identity of these plants. FLORA OF THE MERRIMACK RIVER FRESHWATER-TIDAL MARSH The vascular flora of the Merrimack River freshwater tidal marsh consists of 88 species distributed among 66 genera in 38 families. Forty-five species are dicots and forty species are mono- cots. The best represented families are the Asteraceae, Cyperaceae and Poaceae. Rhodora [Vol. 94 Pteridophytes ISOETACEAE Isoetes echinospora Durieu Quillwort. Occasional; at the east edge of Area 1 in the gravelly sub- strate in the high tide zone. Caldwell 273, 488. EQUISETACEAE Equisetum arvense L. Common Horsetail. Occasional; throughout the marsh in the high tide zone. Caldwell 407. Equisetum fluviatile L. Water Horsetail. ommon; in gravelly substrate in the high tide zone. Cald- well 258, 426. Equisetum hyemale L. Scouring Rush. Occasional; midmarsh with Acorus calamus at east end of study site. Angiosperms Dicotyledons CABOMBACEAE Cabomba caroliniana Gray Fanwort. Occasional; in the Artichoke River. Caldwell 449. CERATOPHYLLACEAE Ceratophyllum demersum L. Coontail. Occasional; in the Artichoke River. Caldwell 448. RANUNCULACEAE Caltha palustris L. Marsh-marigold. ommon; in the high tide zone throughout the marsh. Caldwell 210. Ranunculus repens L. Creeping Buttercup. ommon; in gravelly areas between deep marshes in high tide zone. Caldwell 203, 235. Thalictrum pubescens Pursh Tall Meadow-rue. Common; in the backmarsh near the woods. Caldwell 212. 1992] Caldwell and Crow— Freshwater Tidal Marsh 85 URTICACEAE Pilea pumila (L.) Gray Clearweed. Common; in the backmarsh near the high tide zone. Cald- well 305, 380, 501. AMARANTHACEAE Amaranthus cannabinus (L.) Sauer (= Acnida cannabina L.) Water-hemp. Abundant; throughout the marsh, especially in the gravelly areas between the deep marshes. Caldwell 244, 283, 284, 484. CARYOPHYLLACEAE Sagina procumbens L. Pearlwort. Occasional; in wet depressions in the high tide zone. Cald- well 229. POLYGONACEAE Polygonum arifolium L. Tearthumb. Common; in the backmarsh. Caldwell 302. Polygonum cuspidatum Sieb. & Zucc. Japanese Knotweed. Common: in the backmarsh near edge of the woods. Cald- well 268. Polygonum punctatum Ell. Water Smartweed. Abundant; throughout the marsh. Caldwell 255, 482. Rumex crispus L. Yellow Dock. . Common; at the edge of the marsh in the high tide zone. Caldwell 230, 251. ELATINACEAE . Elatine americana (Pursh) Arn. (= E. triandra var. americana (Pursh) Fassett) Waterwort. Abundant; growing on the mud throughout marsh. Cald- well 291, 476, 487, 504. BRASSICACEAE a Cardamine pensylvanica Muh. ex Willd. Pennsylvania Bitter Cress. Occasional; at the edge of marsh, especially in gravelly areas between the deep marshes. Caldwell 404, 423. 86 Rhodora [Vol. 94 PRIMULACEAE Lysimachia ciliata L Occasional; at the edge of the marsh near woods. Caldwell 468. Lysimachia lanceolata Walt. Occasional; in the backmarsh. Caldwell 248, 261. Lysimachia terrestris (L.) BSP. Swamp Candles. Abundant; throughout the marsh near high tide zone. Cald- well 454, 471. ROSACEAE Potentilla norvegica L. Uncommon; at the edge of the marsh near woods. Caldwell 452. Sanguisorba canadensis L. American Burnet. Uncommon; in open areas between the deep marshes near the high tide zone. Caldwell 276, 511. FABACEAE Amorpha fruticosa L. False Indigo. Occasional; in the high tide zone between Area 2 and Area 3. Caldwell 417, 520. LYTHRACEAE Lythrum salicaria L. Purple Loosestrife. Occasional; in the backmarsh. Caldwell 239, 445. ONAGRACEAE Ludwigia palustris (L.) Ell. Water-purslane. Common; throughout the backmarsh in muddy substrate. Caldwell 310, 381, 475. BALSAMINACEAE Impatiens capensis Meerb. Spotted Touch-me-not. Abundant; throughout the elevated areas and the back- marsh. Caldwell 306. APIACEAE Cicuta bulbifera L. Bulb-bearing Water-hemloc Occasional; in the backmarsh in Area 1. ae Ne 345, 522. 1992] Caldwell and Crow—Freshwater Tidal Marsh 87 Cicuta maculata L. Water-hemlock. Common; in the backmarsh. Caldwell 451, 436. Sium suave Walt. Water-parsnip. Abundant; throughout the marsh, especially in the stony areas between the deep marshes. Caldwell 240, 322, 474. APOCYNACEAE Apocynum cannabinum L. Indian-hemp. Occasional; in the high tide zone between Area 2 and Area 3. Caldwell 440, 505. LAMIACEAE Lycopus americanus Muhl. Water-horehound. Common: in the backmarsh. Caldwell 242, 323, 379, 495. Mentha arvensis L. Water Mint. Common: throughout the marsh in the high tide zone. Caldwell 242, 462. Physostegia leptophylla Small (= P. arboriginorum Fern.) Occasional; in the shade near the high tide zone between Area 2 and Area 3. Caldwell 512. Scutellaria lateriflora L. Skullcap. ccasional; in the backmarsh near the edge of the woods. Caldwell 378, 499. CALLITRICHACEAE Callitriche verna L. (= C. palustris L.) Water-starwort. Occasional; along the edge of the Indian River and the west side of the island near the Artichoke River. Caldwell 228, SCROPHULARIACEAE Chelone glabra L. Turtlehead. Common: in the backmarsh and the edge of the island near the Artichoke River. Caldwell 329, 336, 509. Lindernia dubia (L.) Pennell False Pimpernel. Common; in the high tide zone at the edges of the deep marshes. Caldwell 262, 480. Mimulus ringens L. Square-stem Monkey-flower. — Common; throughout the marsh in the high tide zone. Caldwell 245, 441, 469, 521. 88 Rhodora [Vol. 94 RUBIACEAE Galium palustre L. Bedstraw. Occasional; in the muddy areas of the backmarsh in the high tide zone. Caldwell 217, 224, 500. ASTERACEAE Aster novi-belgii L. New York Aster. Abundant; throughout the marsh in the high tide zone at the edge of the woods. Caldwell 338, 516, 519. Bidens cernua L. Beggar’s-ticks. Abundant; throughout the midmarsh but more frequently in the backmarsh. Caldwell 281, 303, 309, 328, 481. Bidens connata Muhl. ex Willd. Common; throughout the marsh. Caldwell 355, 358. Bidens eatonii Fern. Common; in the gravelly and elevated areas of the marsh. Caldwell 506, 507, 510, 525. Bidens frondosa L. Occasional; in the backmarsh in the high tide zone. Cald- well 335, 340, 524. Eupatorium dubium Willd. ex Poir. Joe-pye Weed. Common; at the edge of the marsh in the high tide zone. Caldwell 274, 496. Eupatorium perfoliatum L. Boneset. Abundant; in the elevated areas and the backmarsh in the high tide zone. Caldwell 437, 465, 497. Solidago sempervirens L. Seaside Goldenrod. Common; between the deep marshes in the high tide zone. Caldwell 325, 517, 518. Tussilago farfara L. Coltsfoot. Uncommon; between Area 2 and Area 3 in the high tide zone. Caldwell 405. Monocotyledons ALISMATACEAE Alisma subcordatum Raf. (= A. plantago-aquatica var. parvl- florum (Pursh) Torr.) Water-plantain. Uncommon; in the backmarsh of Area 1. Caldwell 271, 478. 1992] Caldwell and Crow—Freshwater Tidal Marsh 89 Alisma triviale Pursh (= A. plantago-aquatica var. americana Schultes & Schultes) Water-plantain. Uncommon; in the backmarsh of Area 3. Caldwell 494. Sagittaria graminea Michx. var. graminea (S. eatoni J. G. Sm.) Abundant; throughout the marsh in muddy silty substrate. Occurring as sterile rosettes in the low tide zone. Caldwell 232, 241, 254, 261, 279, 280, 429, 439. Sagittaria latifolia Willd. Arrowhead. Abundant; throughout the middle and backmarsh in mud- dy substrate. Caldwell 264, 278, 331, 332, 477, 493. HYDROCHARITACEAE Elodea nuttallii (Planch.) St. John Waterweed. Abundant; throughout the marsh in muddy substrate or in large patches in open areas in the low tide zone. Caldwell 222, 257, 294, 433. Vallisneria americana Michx. Tape-grass. Common; washed up on shore throughout the marsh. Caldwell 277, 498. POTAMOGETONACEAE Potamogeton crispus L. Curly-leaved Pondweed. Occasional; rooted in the mud in the low tide zone or washed up on shore. Caldwell 221, 479. Potamogeton nodosus Poir. Occasional; in the muddy substrate of the high tide zone; plants vegetative only. Caldwell 218, $33, $12,327. Potamogeton perfoliatus L. Clasping-leaved Pondweed. Occasional; rooted in the mud in the low tide zone and at the edges of the Artichoke River. Caldwell 219, 420, 450. Potamogeton robbinsii Oakes Robbins’ Pondweed. Uncommon; in the Artichoke River. Caldwell 447. NN CCS RN ARACEAE Acorus calamus L. Sweet-flag. Abundant: occurs as a large zone mid-marsh with scattered populations near the edges. Caldwell 204, 402, 411. Peltandra virginica (L.) Schott & Endl. Arrow-arum. Common; in the deep mud of backmarsh. Caldwell 424, 434. 90 Rhodora [Vol. 94 LEMNACEAE Wolffia columbiana Karst. Water-meal. Uncommon; washed up on shore. Caldwell 313. COMMELINACEAE Commelina communis L. Asiatic Dayflower. Occasional; in the high tide zone between Area 2 and Area 3. Caldwell 459. ERIOCAULACEAE Eriocaulon parkeri Robins. Pipewort. Occasional; at the east edge of Area | in gravelly substrate in the high tide zone. Caldwell 259, 272, 319, 492. JUNCACEAE Juncus acuminatus Michx. Occasional; in the gravelly substrate between the deep marshes. Caldwell 234, 253, 446, 455. CYPERACEAE Carex hormathodes Fern. Uncommon; in the high tide zone at the east edge of Area 1. Caldwell 427. Carex paleacea Wahlenb. Occasional; in the gravelly substrate between the deep marshes and elevated backmarsh. Caldwell 205, 413. Carex stipata Muhl. ex Willd. Uncommon; in the high tide zone at east edge of Area |. Caldwell 421. Cyperus bipartitus Torr. (= C. rivularis Kunth) Occasional; in the gravelly substrate in the high tide zone. Caldwell 491. Cyperus esculentus L. Yellow Nut-grass. Occasional; in the gravelly substrate in the open areas. Caldwell 292. Dulichium arundinaceum (L.) Britt. Three-way Sedge. Frequent; in large clumps in the gravelly areas and in the backmarsh. Caldwell 456, 508. Eleocharis smallii Britt. Spike Rush. Frequent; in the gravelly areas and the elevated borders of the islands. Caldwell 211, 247, 287, 408, 425, 457. 1992] Caldwell and Crow— Freshwater Tidal Marsh 91 Scirpus fluviatilis (Torr.) Gray River Bulrush. Frequent; in the midmarsh and elevated borders, partic- ularly around the islands. Caldwell 285, 312, 400, 401, 430, 443, 503. Scirpus pungens Vahl. (S. americanus of American authors, misapplied to this taxon) Three-square Bulrush. Abundant; throughout the marsh in the low tide zone and the gravelly areas. Caldwell 214, 233, 249, 270, 286, 442. Scirpus smithii Gray Uncommon; in the gravelly substrate with Eriocaulon par- keri at the east end of Area | near the high tide zone. Caldwell 490. Scirpus tabernaemontanii K. C. Gmel. (= S. validus Vahl.) Great Soft-stem Bulrush. Abundant; midmarsh in the silty muddy substrate. Cald- well 213, 256, 431. POACEAE Agrostis stolonifera L. var. stolonifera (= A. stolonifera var. major (Gaud.) Farw.; A. alba var. stolonifera (L.) Sm.) Redtop. Occasional; at the edge of the marsh in the high tide zone. Caldwell 470. Calamagrostis canadensis (Michx.) Beauv. Bluejoint. Frequent; at the edge of the marsh in the high tide zone. Caldwell 227, 428. Dactylis glomerata L. Orchard Grass. Uncommon; at the edge of marsh in the high tide zone near the woods. Caldwell 206. Panicum dichotomiflorum Michx. var. geniculatum (Wood.) ern. Uncommon: in a gravelly area of the river bank that was cleared by landowner. Caldwell 344. Phalaris arundinacea L. Reed Canary Grass. Occasional; in the high tide zone throughout the marsh. Caldwell 208, 414. Spartina alterniflora Loisel. Saltwater Cord Grass. Frequent; in large clumps in the silty substrate near the low tide zone. Caldwell 485, 506. Spartina pectinata Link Freshwater Cord Grass. Frequent; in the gravelly substrate in the high tide zone. Caldwell 236, 250, 444. 92 Rhodora [Vol. 94 Zizania aquatica L. var. aquatica Southern Wild-rice. Abundant; scattered throughout the marsh, but especially abundant in the deep mud of the backmarsh. Caldwell 260, 267, 486. SPARGANIACEAE Sparganium eurycarpum Engelm. Large-fruited Bur-reed. Occasional; in the backmarsh and bordering the Indian River. Caldwell 269, 304, 418. TYPHACEAE Typha angustifolia L. Narrowleaf Cattail. Occasional; in the elevated areas bordering the island near the Indian River. Caldwell 223. PONTEDERIACEAE Pontederia cordata L. Pickerel-weed. Abundant; in the deep mud of the backmarsh and occa- sionally the midmarsh. Caldwell 238, 467. IRIDACEAE Tris pseudacorus L. Yellow Iris. Occasional; in the high tide zone and the elevated areas bordering the islands. Caldwell 201, 226. Tris versicolor L. Blue Flag. Occasional; in the high tide zone between the deep marsh- es. Caldwell 202. CONCLUSION Several biotic and abiotic factors influence the composition and distribution of vegetation in a freshwater tidal marsh. Three en- vironmental parameters that may be related to plant community structure were explored, including salinity, organic content of the soil, and water level fluctuation. Although salinity is a hydrological component of a freshwater tidal marsh, levels of less than 5% measured at the study site were probably too low to be a major influence on the structure of plant cover types. Also, the fact that salinity occurred near the 1992] Caldwell and Crow—Freshwater Tidal Marsh 93 end of the growing season seemed to further negate its influence on community structure. Soil organic content measurements indicate that a large per- centage of the marsh, encompassing many cover types, contains mineral soil with a low level of organic matter. Although species diversity is the highest in backmarsh cover types where soil or- ganic content is highest, more sampling needs to be done to sub- stantiate any conclusions. On the other hand, duration of inundation at the plant/soil interface appears to be an important factor in structuring marsh cover types. Species diversity is lowest in cover types such as those dominated by Spartina alterniflora and Scirpus pungens, where duration of flooding is the longest (16 hr./day). Few emer- gent species can tolerate this extensive period of inundation. Mid- to backmarsh cover types such as those dominated by Acorus calamus and Zizania aquatica have more diversity. However, this analysis does not hold true for rocky shore areas where pre- sumably the high degree of disturbance is an overriding influence on vegetation patterns. Two other important factors contributing to plant community structure in the Merrimack River marsh that should be mentioned here are plant growth forms and physical disturbance by ice floes. The most successful plants in this fluctuating environment are either annuals or strongly rhizomatous perennials. A proliferation of rhizomes allows these plants to take hold in areas of constant sediment deposition and erosion, while storing nutrients for early emergence and rapid growth. The seeds of annuals, in contrast, find protection in the microrelief of the marsh surface and rocky shoreline. The Merrimack River is a relatively large and rapidly moving river that partially freezes during the winter. Therefore, ice floes also have a major impact on the vegetation of the marsh and rocky shoreline. Plants along the banks are sheared by ice scour- ing. Additionally, sediments containing seeds and rhizomes are entrapped and transported by ice chunks drifting up and down- stream on tidal and river currents, often being deposited along the river’s edge or carried out to sea. Freshwater tidal marshes are the least studied wetlands in the U.S. (Odum et al., 1984). This environment is dynamic, with many distinctive plant species. Research in many more freshwater tidal rivers is surely needed to better understand this unique hab- Itat. 94 Rhodora [Vol. 94 ACKNOWLEDGMENTS We thank A. Linn Bogle, T. Lee and A. Mathieson for their helpful comments on the manuscript. This research was funded in part by the the Massachusetts Natural Heritage and Endangered Species Program and the University of New Hampshire Central University Research Fund (#3012). LITERATURE CITED AHNERT, F. 1960. Estuarine meanders in the Chesapeake Bay area. Geogr. Rev. (New York) 50: 390-401 ANDERSON, R. R., R. G. Necciia AND R. D. RAppLEYE. 1968. Water quality and plant a along the upper Patuxent River, Maryland. Chesapeake Sei. 9: 6. CHABRECK, 2 1988. Coastal Marshes: Ecology and Wildlife Management. University of Minnesota Press, Minneapolis, Crow, G. E. 1982. New England’ s Rare, Threatened, and Endangered Plants. ‘S. Gov. Print. Off., Washington, DC. ——., W. D. Courrevman, G. L. Cuurcu, L. M. Eastman, L. J. MEHRHOFF ane I. M. Storks. 1981. Rare and endangered vascular plant species in New England. Rhodora 83: 259-299. —— anp C. B. HELiquist. 1981. Aquatic vascular plants of New England: Part 2. ‘Tolar and Sparganiacaeae. New Hampshire Agric. Exp. Sta. Bull. 517. AND 82. Aquatic vascular plants of New England: Part 4. Juncaginacaeae, Scheuchzeriaceae, Butomaceae, Hydrocharitacaeae. New Hampshire Agric. Exp. Sta. Bull. 520. = AND 1 Aquatic vascular plants of New England: Part 6. Trapaceae, Hideripaceae Hippuridaceae. New Hampshire Agric. Exp. Sta. Bull. 524. AND ———. 1985. Aquatic vascular plants of New England: Part 8. Lentibulariaceae. New Hampshire Agric. Exp. Sta. Bull. 528. ——— AND - Aquatic and Wetland Plants of Northeastern North America. University of eee Press. Madison, WI (in press). Crowe, D. R. AND W. H. PARKER. 1981. iinidiention and Agamospermy of Bidens in stent Ontario. Taxon 30: 749-760. DACHNOwSKI-STokgs, A. P. 1940. Structural characteristics of peats and mucks. 9-400. Soil Sci. 50: 389-4 DESCHENES, J. AND J. B. SERODES, 1985. The influence of salinity on Scirpus americanus tidal marshes in the St. Lawrence River Estuary, Quebec. Canad. J. Bot. 63: 920-927. Dionne, J. 1968. Sch By hst f the St. Lawrence Estuary. Amer. J. Sci. 266: 380-388, ———.. 1969. Tidal flat erosion by ice at La Pocatiere, St. Lawrence Estuary. J. Sediment. Petrol. 39: 1174-1181. 1992] Caldwell and Crow—Freshwater Tidal Marsh 95 . 1974. How ice drift shapes the St. Lawrence. Canad. Geogr. J. 88: 4-9. Disraci, D. J. AND R. W. Fonpa. 1979. Gradient analysis of the vegetation in a brackish marsh in Bellingham Bay, Washington. Canad. J. Bot. 57: 465- 475 Doumtete, D. G. 1981. Primary production and seasonal aspects of emergent plants in a tidal freshwater marsh. Estuaries 4: 139-142. Ewina, K. 1983. Environmental controls in Pacific northwest intertidal marsh plant ina ces Js Bot. 61: 1105-1 116. 1986. dients in a Pacific northwest brackish intertidal marsh. Estuaries 9: 49-62. Fassett, N.C. 1925a. Bidens eatonii and its varieties. Rhodora 27: 142-146. . 1925b. A key to the northeastern American species of Bidens. Rhodora 27: 184-185 . 1928. Vegetation of the estuaries of northeastern North America. Proc. Boston Soc. Nat. Hist. 39: 73-130. FERNALD, M. L. 1903. A new Bidens from the Merrimac Valley. Rhodora 5: 93 —. 1908. Bidens connata and some of its American allies. Rhodora 10: 197-203. . 1950. Gray’s Manual of Botany, 8th ed. American Book Co., New York. FERREN, W. R., Jr. 1976. Aspects of intertidal zones, vegetation, and flora of the Maurice River system, New Jersey. Barton1 —67. . E. Scuuyier. 1980. Intertidal canal plants of river systems near Philadelphia. Proc. Acad. Nat. Sci. Philadelphia 132: 86-120 GaRoFALo, D. 1980. The influence of wetland vegetation on tidal stream channel migration and morphology. Estuaries 3: 258-270. Gueason, H. A. 1952. The New Britton and Brown Illustrated Flora of the Northeastern United States and Adjacent Canada. 3 vols. New York Botan- ical Garden, New York. GLEAson, H. A. AND A. CRonquist. 1963. Manual of Vascular Plants of North- eastern United States and Adjacent Canada. D. Van Nostrand Co., New York. Goon, R. E. AND N. F. Goop. 1975. Vegetation and production of Woodbury Creek-Hessian Run freshwater tidal marshes. Bartonia 43: 38-4 S, AND . 1983. Distribution of adhaneseedl aquatic acrophytes i in the tidal Potomac River. Aquatic Bot. 15: 65-79. tieceareee C.B. anpG. E. Crow. 1980. Aquatic vascular plants of New En- gland: Part 1. Zosteraceae, Potamogetonaceae, Zannichelliaceae, Najadaceae. New Hampshire Agric. Exp. Sta. Bull. 515. AND 1981. pane ee plants of New England: Part 3. Alismataceae. om Hampshire Agric. Exp. Sta . Bull. 518. ~~ SAD Aquatic ann plants of New England rt 5. Saber see ge Araceae, Lemnaceae, Xyridaceae, Eriocaulaceae, and Hampshire Agric. Exp. Sta. Bull. 523. AND 1984. Aquatic vascular plants of New England: Part 7 Cabombaceae, Nymphaeaceae, Nelumbonaceae, and Ceratophyllaceae. Nie Hampshire Agric. Exp. Sta. Bull. 527. Hit, M.O. 1979. TWINSPAN: A fortran program for arranging multivariate data in an ordered two-way table by searticss of the individual and 96 Rhodora [Vol. 94 attributes. Department of Ecology and Systematics, Cornell University, Ith- ca, NY aca, : HutTcuinson, I. 1982. Vegetation-environment relations in a brackish marsh, Lulu Island, Richmond, B.C. Canad. J. Bot. 60: 452-462. JEROME, W. C., A. P. CHEsmore, C. O. ANDERSON, JR. AND F. Grice. 1965. A study of the marine resources of the Merrimack River Estuary. Monograph Ser. No. 1, Div. of Mar. Fish., Dept. of Natl. Res., The Commonwealth of Massachusetts, Boston. JoOHNson, D. A. AND H. H. York. 1915. The relation of plants to tide levels. Publ. Carnegie Inst. Wash. 206, Keppy, P. A. 1983. Shoreline vegetation in Axe Lake, Ontario: effects of ex- posure on zonation patterns. Ecology 64: 331-344, Leck, M. A. AND K. J. GRAVELINE. 1979. The seed bank of a freshwater tidal marsh. Amer. J. Bot. 66: 1006-1015. AND R. L. Simpson. 1987. Seed bank of a freshwater-tidal wetland: turnover and relationship to vegetation change. Amer. J. Bot. 74: 360-370. , R. Simpson, D. WHIGHAM AND C. F. Leck. 1988. Plants of the Hamilton marshes: a Delaware River freshwater-tidal wetland. Bartonia 54: 1-17 Matuieson, A. C. AND R. A. FRALICK. 1973. Benthic algae and vascular plants of the lower Merrimack River and adjacent shoreline. Rhodora 75: 52-64. METZLER, K. ANDR. Rosza. 1982. Vegetation of fresh and brackish tidal marshes of Connecticut. Newsletter Conn. Bot. Soc. 10: 1-3. MILter, B., D. NoRMANDEAU, G. PIEHLER, P. HALL, A. MATHIESON, R. FRALICK, D. TuRGEON, P. MAHONEY AND W. Owen. 1971. Ecological Study, Mer- rimack River Estuary, Massachusetts. Report for the U.S. Army, Corps of Engineers by Normandeau Associates, Inc. and Vast, Inc. MitTcueLt, R. S., C. J. SHE EAN. 1980. Rare and endangered vascular plant species in New York state. N.Y. State Museum in cooperation with the U.S. Fish and Wildlife Service, Albany, NY Mitscn, W. AND J. GosseLink. 1986. Wetlands: Tidal-Freshwater Marshes. Van Nostrand Reinhold Co., Inc., New York. MUELLER-Domgpors, D. AND H. ELLENBERG. 1974. Aims and Methods of Veg- etation Ecology. John Wiley, New York. Nicuots, G. E. 1920. The vegetation of Connecticut: VII. The associations of depositing areas along the seacoast. Bull. Torrey Bot. Club 47: 511-548. Opum, W. E. 1988. Comparative ecology of tidal freshwater and salt marshes. Annual Rev. Ecol. Syst. 19: 147-176. , T. J. Smrtu III, J. K. Hoover aNp C. C. MclIvor. 1984. The ecology of tidal freshwater marshes of the United States east coast: a community profile. U.S. Fish and Wildlife Service FWS/OBS-83/17, Washington, DC. Puiuipp, C.C. AND R. G. BRown. 1965. Ecological studies of the transition-zone vascular plants in South River, Maryland. Chesapeake Sci. 6: 73-81. Seropes, J. B. AND J. P. TRoupe. 1984. Sedimentation cycle of a freshwater tidal flat in the St. Lawrence estuary. Estuaries 7: 119-127. SHERFF, E.E. 1937. The genus Bidens. Field Mus. Nat. Hist., Bot. Ser. 16, Parts I and II. Simpson, R. L., R. E.Goop, M. A. LEcK AND D. F. WHIGHAM. 1983. The ecology of freshwater tidal wetlands. Bioscience 33: 255-259. 1992] Caldwell and Crow— Freshwater Tidal Marsh 97 SNYDER, D. B. AND V. E. VIVIAN. — ss and endangered " vascular plant species in New Jersey. The Conser Center, Inc. in cooperation with the U. S. Fish and Wildlife Service, [Newton Corner, MA\]. Sorrig, B. A. 1987. Notes on the rare flora of Massachusetts. Rhodora 89: 113- 6. 1990. Rare Native Plants of Massachusetts. Massachusetts Division of Cishecies and Wildlife, Boston. Tucker, A. O., N. H. Dit, C. R. BRoome_, C. E. PHILLips, M. AND J. MACIARELLO. 1979. Rare and Endangered Vascular Plant Species i in Delaware. The Society of Natural History of D with the U.S. Fish and Wildlife Service, [Newton Corner, MA]. WuicHaM, D. F. AND R. L. Simpson. 1975. Ecological studies of the Hamilton Marshes progress report for the period June 1974-January 1975. Rider Col- lege, Biology Dept., Lawrenceville, NJ. —— J. McCormick, R. E. Goop anp R. L. Simpson. 1978. Biomass and primary production in freshwater tidal wetlands of the middle Atlantic coast. In: R. E. Good, D. F. Whigham and R. L. Simpson, Eds., Freshwater Wet- lands: Ecological Processes and Management Potential. Acad ic Press, New York. DEPARTMENT OF PLANT BIOLOGY UNIVERSITY OF NEW HAMPSHIRE DURHAM, NH 03824 RHODORA, Vol. 94, No. 877, pp. 98-99, 1992 NEW ENGLAND NOTE AN INDIGENOUS POPULATION OF CLINTONIA BOREALIS (LILIACEAE) ON CAPE COD SHIRLEY G. Cross On August 29, 1990 in East Sandwich, MA I was collecting botanical specimens for the Green Briar Nature Center of the Thornton W. Burgess Society, and was startled to come upon a colony of approximately 50 plants of Clintonia borealis (Ait.) Raf. The colony was growing on a wooded bank above a swampy area traversed by spring-fed streams running down into the adjoining saltmarsh. Only a few plants appeared to have flowered. There were no fruiting scapes. I returned to the site in early May of 1991. Three budded scapes were present. On this occasion an additional colony was discov- ered which had no flowering scapes, possibly due to the close shading of Smilax rotundifolia L. The entire area, swamp and banks, is in the category described by Svenson and Pyle (1979) under the heading of “The fresh borders of saltmarshes.” It has the black soil they speak of and the rich flora of Symplocarpus foetidus (L.) Nutt., Dryopteris si- mulata Davenp., Woodwardia areolata (L.) Moore, Arisaema tri- phyllum (L.) Schott, Viola cucullata Ait., and Chrysosplenium americanum Schwein. overshadowed by old red maples (Acer rubrum L.), tupelo (Nyssa sylvatica Marsh.), beech (Fagus gran- difolia Ehrh.) and witch hazel (Hamamelis virginiana L.). I enlisted the help of Dr. Walter H. Hodge and returned to the site with him on May 14th, 1991. On seeing the site he agreed with me in thinking it to be an old undisturbed area. He photo- graphed the Clintonia colony and the blossoming scape. I after- wards pressed the scape and basal leaves and am depositing the specimen with copies of the photographs in NEBC, under the Green Briar Nature Center’s label. (The Green Briar Nature Cen- ter is the teaching arm of the Thorton W. Burgess Society.) Copies of the photographs are also deposited at SPWH and at the Herbarium of the Cape Cod Museum of Natural History in Brewster, Massachusetts. 98 1992] New England Note 99 LITERATURE CITED Svenson, H. K. AND E. Pye. 1979. The Flora of Cape Cod Museum of Natural History, W. Brewster, MA. THORTON W. BURGESS SOCIETY 6 DISCOVERY HILL ROAD E. SANDWICH, MA 02537 NEBC AND NEW ENGLAND WILDFLOWER SOCIETY SYMPOSIUM SCIENTIFIC AND MANAGEMENT ISSUES IN RARE PLANT CONSERVATION The New England Botanical Club and the New England Wild- flower Society will sponsor a symposium on the scientific and management issues in rare plant conservation. Symposium topics will include taxonomic issues in plant conservation, setting pri- orities for protection, re-introduction and transplantation, and habitat management. The symposium will be held on March 21, 1992 at Bentley College in Waltham, Massachusetts. Additional details and registration materials will be mailed in January, 1992. For information, write to: Symposium Committee The New England Botanical Club Harvard University Herbaria 22 Divinity Avenue Cambridge, MA 02138 RHODORA, Vol. 94, No. 877, pp. 100-101, 1992 BOOK REVIEW Alice F. Tryon and Bernard Lugardon. 1991. Spores of the Pter- idophyta: Surface, Wall Structure, and Diversity Based on Electron Microscope Studies. 648 pp. Springer-Verlag, 175 Fifth Ave, New York, NY 10010. ISBN 0-387-97218-8. (Price: $98.00, hardcover.) The spore-wall features of spore-dispersed plants such as ferns and clubmosses have long been known for their beauty and vari- ation. In a new book by Alice Tryon and Bernard Lugardon, these features have been exhaustively portrayed in an unparalleled com- pendium of data on spore-wall ultrastructure. This abundantly illustrated volume focuses both on surface features as seen under the scanning electron microscope and the interior-wall features visible in transmission electron micrographs. It will prove a land- mark volume for students of plant evolution and ecology as well as sedimentary petrologists and commercial groups interested in identifying productive sedimentary strata. This work is the product of a remarkable international coop- eration; Bernard Lugardon developed the transmission electron micrography in France and Alice Tryon pursued the scanning electron microscope work in the United States. Their years of work together have paid off in a highly integrated volume with a single message supported by two enormous data sets relevant to the same biological issue, the significance of spore design to sys- tematics and evolutionary studies in the spore-dispersed plants. The book includes two sections, a general introduction and a survey of the genera. The introduction provides a valuable over- view of spore structure and development; it is rich in detail and at the same time accessible. The long-standing confusion about the chemical and structural correspondence of seed-plant and fern spore coats is resolved in a lucid and well-illustrated essay: we can now Say that there is little if any homology between the two wall structures. At the same time we gain an understanding of the deposition process resulting in the complex, attractive struc- ture of the mature spore. The heart of the book is the profusely illustrated survey of genera. Worldwide in scope, the Survey includes a broad repre- sentation of all the major groups of ferns, lycopsids, horsetails, and psilophytes. The text for each genus addresses the circum- 100 1992] Book Reviews 101 scription and size of the genus, its range, and details of spore size and morphology, as well as a longer section outlining the rela- tionships of the group based on spore morphology and insights into variation within the group. A literature summary ends each generic treatment. The illustrations for each genus include an array of Scanning Electron Microscope images of the surface fea- tures of the spores, and Transmission Electron Micrographs of interior features are often included as well. The images are tech- nically of the highest quality, and the standard magnification for all whole-spore images allows easy visual comparison of spore size. The utility of the volume is further increased with a glossary of the technical terms relevant to characterizing spore morphol- ogy. This work of Tryon and Lugardon is a milestone in the un- derstanding of spore wall structure and development. The ex- haustive review of spore diversity coupled with the well developed analytic section in the introduction supports the continued de- velopment of spore architecture as a character system for system- atic studies in the ferns and other spore-dispersed plants. DAVID S. BARRINGTON PRINGLE HERBARIUM DEPARTMENT OF BOTANY UNIVERSITY OF VERMONT BURLINGTON, VT 05405-0086 RHODORA, Vol. 94, No. 877, p. 102, 1992 BOOK REVIEW Day, Robin and Paul M. Catling 1991. The Rare Plants of Prince Edward Island. Syllogeus Series No. 67. Canadian Museum of Nature. Prepaid orders available from the Canadian Mu- seum of Nature, Direct Mail Section, P.O. Box 3343, Station D., Ottawa, Ontario, Canada K1P 6P4. (Price: $4.95.) This booklet is a continuation of the rare plants of Canada series. It lists the rare vascular plants of Prince Edward island and includes a range map in Prince Edward Island with gener- alized locations, references, the ranges in North America, habitat, and status in other provinces. The rare vascular plants of Prince Edward Island is a must for the person interested in the rare species of Canada and North America. The style is similar to those published for the other provinces. The only plant listed that probably does not belong is Phragmites australis, an introduced weedy species throughout North America. C. BARRE HELLQUIST DEPARTMENT OF BIOLOGY NORTH ADAMS STATE COLLEGE NORTH ADAMS, MA 01247 102 RHODORA, Vol. 94, No. 877, pp. 103-105, 1992 BOOK REVIEW A. Borhidi. 1991. Phytogeography and Vegetation Ecology of Cuba. 858 pp. (including 380 figs. as photographs, graphs, maps, diagrams, plus 16 colored plates, and a packet with ‘Map of the Natural Potential Vegetation of Cuba,” in color, 1:1,250,000). Akadémiai Kiad6 (publishing house of the Hungarian Academy of Sciences), Prielle Kornélia u. 19-35, H-1117 Budapest, Hungary. ISBN 963 05 5292 7. (Price: $74, cloth.) Phytosociological studies and classification of the vegetation of European countries are well known. The school producing such studies in North America has been small and is represented most recently by Pierre Dansereau (1957). In the Caribbean area, Stehlé published such treatments for Guadeloupe (Stehlé, 1935), as did Ciferri for Hispaniola (Ciferri, 1936) and Dansereau for Puerto Rico (Dansereau, 1966). In this volume Borhidi applies such techniques to the vegetation of Cuba. Following the“victory of the Popular Revolution” in Cuba in 1959, botanical research was re-established in the sixties with extensive programs to recollect the vegetation for a new Flora of the Republic of Cuba (Capote, 1988). Along with the building of collections, special studies had as preliminary tasks: “typification of the vegetation units; analysis of the structure and composition of the vegetation units; comprehensive bioclimatological studies of whole country; exploration of the inner and outer phytogeo- graphic relationships in Cuba in order to prepare a new ‘phyto- geographic regionalization’ of Cuba to link all in geobotanic work which has been prepared for most countries in Europe but which 1S available only in a few tropical countries” (Borhidi, 1991). The author of this book is A. Borhidi of Hungary, the principal con- tributor to these studies, but the work also includes the studies of H. Manitz and J. Bisse of the German Democratic Republic, V. Samek of Czechoslovakia, and many Cuban botanists who were their students and protegés. Borhidi’s volume contains a most welcome concise study of the geography of Cuba with a map showing current divisional boundaries. There is a chapter devoted to updating our knowledge of the various relationships between soil and vegetation, which historically have been important to Cuba. Borhidi also analyzes 103 104 Rhodora [Vol. 94 the vegetation on the basis of life-forms and offers a phytogeo- graphic characterization of the flora of Cuba, on which a vege- tational map is based. A life-form analysis is given for the 6375 species of spermatophytes to be included in the new Flora of the Republic of Cuba. Woody plants comprise 53.3% of these species; 40 different woody formations are recognized by Borhidi. These formations are comparable to the descriptive units given by Ca- pote et al. (1988), but are described here with pseudo-Latin phrase names, which even the author occasionally does not repeat cor- rectly. The map of the natural potential vegetation by Borhidi and Mufiiz is at the scale of 1:1,250,000, dated 1969-1970 (in this 1991 volume), and is approximately 12” by 36”. Thirty-three vegetational units are given in six basic colors, some also overlaid with lines and dots. Three of the colors are intended to be sub- divided, but the divisions are not readily distinguishable in my copy. Over half of the book is devoted to a systematic survey of the plant communities which are classified in a phytosociological hierarchy of 27 classes, 55 orders, 86 alliances, and 186 associ- ations. Each is named, again with a pseudo-Latin epithet, dated and described, with an author’s name cited. The implication of priority is indicated in abbreviations such as nom. nud., nova, hoc locus, syn., or ined. Some comparable categories are recog- nized by Ciferri or Stehlé, but are renamed or emended in this volume. A long appendix, as an out-of-place Table 25, lists the taxa found in relevés in the 40 communities of woody plants. Both life-forms and geographic elements are indicated. Taxa are listed alphabetically. There is an index to subject matter, including figure and table references, and an index to Latin plant names as used outside of the phytosociological tables. A good bibliography is given which shows how much of the available literature on Cuba has unfortunately not reached our local libraries. A personal visit to Cuba is not possible for most botanists at the present time. However, an understanding of the complexities of the vegetational units in the flora of Cuba is easily obtained from this comprehensive work by Dr. Borhidi, who makes clear his contention that ecological research should be geobotanical, chorographical, and plant geographical. The anticipated Flora of the Republic of Cuba will be a welcome companion volume to this study. 1992] Book Reviews 105 LITERATURE CITED Capote, R. R., R. BERAZAiN, AND A. LeIvA. 1988. Cuba. Jn: D. G. Campbell and D. Hammond, ie i inventory of Tropical Countries. New York DANSEREAU, P. 1957. Biogeography: An Ecological Perspective. Ronald Press, 1966. Studies on the Vegetation of Puerto Rico: 1. Description and Integration of the Plant Communities. Inst. Carib. Sci. Spec. Pub. No. 1. STEHLE,H. 1935. Essai d’écologie et de géograr que. Basse-Terre, Gua- deloupe. RICHARD A. HOWARD HARVARD UNIVERSITY HERBARIA CAMBRIDGE, MA 02138 NEBC AWARD FOR THE SUPPORT OF BOTANICAL RESEARCH The New England Botanical Club will again offer an award of $1000 in support of botanical research to be conducted in relation to the New England Flora during 1992. This award is made to stimulate and encourage botanical research on plant species or communities which occur in New England, and to make possible visits to the New England region by those who would not oth- erwise be able to do so. The award will be given to the graduate student submitting the best research proposal, and is not limited to graduate students at New England institutions nor to members of the New England Botanical Club. Preference will be given to research dealing with field studies in systematic botany, biosys- tematics, plant ecology, or plant conservation biology. Applicants Should submit a proposal of no more than three double-spaced pages, a budget (the budget will not affect the amount of the award), and a curriculum vitae. Two letters, one from the stu- dent’s major professor, in support of the proposed research are also required. Proposals and supporting letters should be sent before 28 February to: Awards Committee The New England Botanical Club 22 Divinity Ave Cambridge, MA 02138 106 MEETING ANNOUNCEMENT AND CALL FOR ABSTRACTS NEW ENGLAND BOTANY GRADUATE STUDENTS The seventh New England Botany Graduate Student Meeting will be hosted by the Department of Plant Biology at the Uni- versity of New Hampshire, Durham, NH on Saturday April 11, 1992. Attendance is open to all. Paper presentations will be largely restricted to graduate student research (completed or in progress) representing all areas of botany (systematics, ecology, reproduc- tive biology, anatomy, physiology, etc.). Time slots for paper presentations are limited and prior registration is required. Ab- stracts are due by Monday, March 13, 1992. For information and abstract forms, contact: Linda Fahey or Lenny Lord Department of Plant Biology University of New Hampshire Durham, NH 03824 Tel.:(603) 862-3860 RONALD L. STUCKEY INITIATES ENDOWMENT FUND FOR THE OHIO STATE UNIVERSITY HERBARIUM Ronald L. Stuckey, Professor of Botany at The Ohio State Uni- versity, recently presented a gift of $30,000 to the University’s Foundation to initiate an endowment for the support of the Uni- versity Herbarium. The presentation was made as a final surprise announcement at Professor Stuckey’s retirement party celebrating his 26 years of teaching at the University. Designated as the Ronald L. Stuckey Herbarium Fund, the endowment will aid in the studies of the flora of Ohio, a particular concern of the donor. The establishment of the endowment fund for the University Herbarium not only marks the occasion of Dr. Stuckey’s retirement from teaching, but also commemorates the 100th anniversary of The Ohio State University Herbarium, which was founded in 1891 by the University’s first Professor of botany, Dr. William A. Kellerman. Professor Stuckey served as curator of the herbarium from 1967 through 1976. 107 THE 1991 JESSE M. GREENMAN AWARD The 1991 Jesse M. Greenman Award has been won by Scott Zona for his publication ““A monograph of Sabal (Arecaceae: Coryphoideae),”” published in Aliso 12: 583-666, 1990. This monographic study is part ofa Ph.D. dissertation from Claremont Graduate School, Claremont, California, under the direction of Dr. Sherwin Carlquist. The Greenman Award, a certificate and a cash prize of $500, is presented each year by the Missouri Botanical Garden. It rec- ognizes the paper judged best in vascular plant or bryophyte sys- tematics based on a doctoral dissertation published during the previous year. Papers published during 1991 are now being ac- cepted for the 24th annual award, which will be presented in the summer of 1992. Reprints of such papers should be sent to: Dr. P. Mick Richardson Greenman Award Committee Missouri Botanical Garden P.O. Box 299 St. Louis, MO 63166-0299, U.S.A. In order to be considered for the 1992 award, reprints must be received by June 1, 1992. 108 THE NEW ENGLAND BOTANICAL CLUB 22 Divinity Avenue Cambridge, MA 02138 The New England Botanical Club is a non-profit organization that promotes the study of plants of North America, especially the flora of New England and adjacent areas. The Club holds regular meetings, has a large herbarium of New England plants, and a library. It publishes a quarterly journal, RHODORA, which is now in its 94th year and contains about 400 pages a volume. Membership is open to all persons interested in systematics and field botany. Annual dues are $35.00, including a subscription to RHODORA. Members living within about 200 miles of Boston receive notices of the Club meetings. To join, please fill out this membership application and send with enclosed dues to the above address. Regular Member $35.00 Family Rate $45.00 For this calendar year For the next calendar year — Name Address City & State Zip Special interests (optional): Vol. 93, No. 876, including pages 307-412, was issued November 14, 1991. 109 Ae ei ites i ii i i leila anatase alti neem. | ine Sietaeheenaiinen INFORMATION FOR CONTRIBUTORS TO RHODORA Submission of a manuscript implies it is not being considered for publication simultaneously elsewhere, either in whole or in Manuscripts should be submitted in triplicate (an original and two xerographic copies) and must be double-spaced (at least *”) throughout including tables, figure legends, and literature cita- tions. The list of legends for figures and maps should be provided on a separate page. Footnotes should be used sparingly. Do not indicate the style of type through the use of capitals or under- scoring, particularly in the citation of specimens. Names of genera and species may be underlined to indicate italics in discussions. Specimen citations should be selected critically, especially for common species of broad distribution. Systematic revisions and similar papers should be prepared in the format of “A Monograph of the Genus Malvastrum,” S. R. Hill, Rhodora 84:1-83, 159- 264, 317-409, 1982, particularly with reference to indentation of keys and synonyms. Designation of a new taxon should carry a Latin diagnosis (rather than a full Latin description), which sets forth succinctly just how the new taxon is distinguished from its congeners. Papers of a floristic nature should follow, as far as possible, the format of “Annotated List of the Ferns and Fern Allies of Arkansas,” W. Carl Taylor and Delzie Demaree, Rho- dora 81: 503-548, 1979. For bibliographic citations, refer to the Botanico-Periodicum-Huntianum (B-P-H, 1968) which provides standardized abbreviations for journals originating before 1966. All abbreviations in the text should be followed by a period, except those for standard units of measure and direction (compass points) For standard abbreviations and for guidance in other matters of biological writing style, consult the CBE Style Manual, 5th ed. (original title: Style Manual for Biological Journals). In preparing figures (maps, charts, drawings, photos, etc.) please remember that the printed plate will be 4 x 6 inches; be sure that illustrations are proportioned to reduce correctly, and indicate by blue pencil the intended limits of the figures. (Some “turn-page” figures with brief legends will be 32 x 6 in.) Magnification/reduction values given in text or figure legends should be calculated to reflect the actual printed size. An Abstract and a list of Key Words should be supplied at the beginning of each paper submitted, except for a very short article or note. All pages should be numbered in the upper right-hand corner. Brevity is urged for all submissions. RHODORA January 1992 Vol. 94, No. 877 CONTENTS A new species of Blephilia (Lamiaceae) from Northern Alabama Richard W. Simmers and Robert Kral Contributions to the alpine flora of the northeastern United States Peter F. Zika A new form of Triphora trianthophora (Swartz) Rydberg, and part 3 of observations on the ecology of Triphora trianthophora (Orchidaceae) in New Hampshir. Tr SO a a Nomenclatural notes on North American Hypoxis (Hypoxidaceae) wives sveobea void FR eT Oe ee Ge ee Chromosome member Sotermainntions i in Fam. Compositae, Tribe Aste cad ew} oo a alu Status ui Se ee me species of Aster and Soli, an C. Semple, Jerry G. Chmielewski and Chunsheng Xiang ...... ristic and vegetation sis of a freshwater tidal marsh on the Mer- assachusetts rimac = bury, M: Fredricka Ann Caldwell and Garrett E. Crow ENG: NOTE An indigenous population of Clintonia borealis (Liliaceae) on Cape Cod ONO Te RA re REVIEWS Spores of the Pteridophyta: Surface, Wall Structure, and Diversity Based on Electron Microsco pe Studies Dans & Raion The Rare Plants of Prince Edward Island C. Barre Hellquist Phytogeography and — Ecology of Cuba Richard A. Howar ANNOUNCEMENTS Rooora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB TC EE a Vol. 94 April 1992 No. 878 Che New England Botanical Club, Inc. 22 Divinity Avenue, Cambridge, Massachusetts 02138 RAGDORA NORTON H. NICKERSON, Editor-in-Chief Associate Editors DAVID S. BARRINGTON RICHARD A. FRALICK A. LINN BOGLE GERALD J. GASTONY DAVID E. BOUFFORD C. BARRE HELLQUIST CHRISTOPHER S. CAMPBELL MICHAEL W. LEFOR WILLIAM D. COUNTRYMAN ROBERT T. WILCE GARRETT E. CROW RHODORA (ISSN 0035-4902). Published four times a year (Jan April, July, and October) by The New England Botanical Club, 33 Divinity Ave., Cambridge, MA 02138 and printed by Allen Press, Inc., 1041 New w Hampshire St., Lawrence, KS 66044. Second class postage paid at Boston, MA and at additional mailing offices. RHODORA is a journal of botany devoted primarily to the flora of North America. Scientific papers and notes relating to this area and floristically related areas, and articles concerned with systematic bot- any and cytotaxonomy in their broader implications, will be consid- ered. Articles are subjected to peer review. RHODORA assesses page charges. SUBSCRIPTIONS: $45.00 per calendar year, net, postpaid, in funds payable at par in the United States currency at Boston. Remittances payable to RHODORA. Send to Treasurer, 22 Divinity Ave., Cam- bridge, MA 02138. MEMBERSHIPS: Regular $35; Family $45. Application form printed sone BACK VOLUMES AND SINGLE COPIES: Some available; infor- mation and prices will be furnished upon request to the Treasurer. ADDRESS CHANGES: In order to receive the next number of RHO- DO ges must be received by the Treasurer prior to the first day of January, April, July, or October. POSTMASTER: Send ede changes to RHODORA, 22 Divinity Ave., Cambridge, MA 02138. INFORMATION FOR ohio Inside back cover, January and April. MANUSCRIPTS: Send to: Joan Y. Nickerso Managing Editor, "RHODORA Phippen-LaCroix Herbarium Dept. of Biology, nig University Medford, MA 0215 Cover Illustration Rhexia virginica L.,, meadow beauty, is found from Nova Scotia to Genie but is rare at the northern limits of its range. The only northern outlier of tropical family relent ria it has distinctive 3-veined leaves stamens. Original artwork by Josephine Ewing. NI, en me mmm cunt, TRbodora JOURNAL OF NEW ENGLAND BOTANICAL CLUB Vol. 94 April 1992 No. 878 RHODORA, Vol. 94, No. 878, pp. 111-134, 1992 FLORAL VARIATION AND TAXONOMY OF LIMNOBIUM L. C. RICHARD (HYDROCHARITACEAE) RICHARD M. LOWDEN ABSTRACT Floral and geographic evidence for infraspecific classification of the monotypic genus Limnobium (incl. si paella native ~ the Americas are presented. A continuum in floral form between populations disclose the presence of just one latitudinally diverging spe- cies, Limnobium spongia (Bosc) Steudel, in which two subspecies, L. spongia Lowden, comb. nov., are recognized. Numerical analysis of floral variation in living populations and their evolutionary trends are treated for these two sub- species. Special field collections of this freshwater, non-submerged, monoecious continuous variation was found in the form and number of petals, stamens, stam- Inoides, stigmas, carpels and locules. Key Words: Limnobium, monospecific, Limnobium spongia subspecies, infra- ee variation, flower variation, population diversity, aquatic plant, Ameri INTRODUCTION The first documented field observations of floral variation ina population of this New World native aquatic plant were made by José Celestino Mutis, director of the Royal Botanical Expedition of the New Kingdom of Granada at Santa Fe de Bogota (Bogota), Colombia. On two occasions (April 29, 1783 and January 26, 1784) he wrote in his diary (Instituto Colombiano de Cultura 111 MISSOUR! BOTANICAL JUL 21 1992 | 112 Rhodora [Vol. 94 Hispanica, 1958) concerning the problems encountered to draw and classify the inconsistencies found in flowers of this interesting plant which grew then, and still abounds today, in lagoons and stagnant water about Bogota. Beyond his interest to conserve flowering material for drawing purposes, Mutis had collected a flower sample that enabled him to determine some of the variations that characterize the unisex- ual flowers of this plant for this particular locality in south Amer- ica. In male flowers, Mutis found three sepals, three petals, and usually seven stamens but sometimes nine [‘‘Por lo regular hallé siete machos y pocos de nueve. . . . casi invisibles.’’]. At the Ma- drid Botanical Garden (Spain), Mutis Plates 259, 260 & 260A (Diaz-Piedrahita, 1985, Laminas XXXV-XXXVII) show male flowers with conspicuous bilobed petals and seven stamens. Even though Mutis did not refer in his diary to these petals as being bilobed, these plates are proof that they were observed as such Over two centuries ago. It is somewhat surprising to note that this obvious character has been overlooked in the herbarium and in nature until now. In the same manner, Mutis found considerable variation in female flowers [“‘Muchos pistilos con inconstancia”], having three sepals, apparently no petals, nine to 15 stigmatic lobes (Plate 260; shows a flower with 11 stigmatic lobes), ovary uniloculed, and fruit elongate (Plates 260 and 260A). A keen sense of observation in the field led Mutis to consider he did not have at hand ordinary flowers with a constant number of parts, but instead a highly variable plant whose classification would require further field observations. Two decades later, flower variants of a similar plant population were observed by M. Bosc (1807) and L. C. Richard (1814; de- scribed as Limnobium) from a marsh in South Carolina (Basse- Caroline), United States. The male flowers were described as hav- ing three sepals, three petals, eight to 12 stamens, sometimes more [Bosc, 1807: “huit a douze étamines, et quelquefois plus, . . . 2] and two or three staminodes [Richard, 1814: “‘La colonne stam1- nifére est terminée 4 son sommet par deux ou trois petits filamens tres-courts et aizus.”; Planche no. 8 (B-7) shows nine stamens and at least one staminode]. Female flowers demonstrated three sepals, three petals, three staminodes [Planche (E-2) shows two of these staminodes], 12 stigmatic lobes [Bosc, 1807; . Six styles profundément bifurqués . . .”], ovary six loculed, fruit glo- ie SES SS 1992] Lowden—Limnobium 13 bose. Richard (1814) was not surprised to find male flowers with different numbers of stamens considering the number of parts found in other members of the Hydrocharitaceae. It is only in recent years during the present decade that com- parable knowledge of another population has been presented by Armando T. Hunziker (1982) from Cabalango (Dept. Punilla), Province of Cordoba, Argentina. In male flowers he counted three sepals, three petals, six stamens only and no staminodes; while, in female flowers there were three sepals, zero to three petals (sometimes one or two, rarely three), four or five staminodes (usually five in pairs or solitary opposite the sepals 2+2+1, if four, then 2+1+1 or 2+2), stigmas usually six but sometimes five (ten or 12 stigmatic lobes), ovary uniloculed with four or five inconspicuous placental partitions, and fruits oblong to ovoid. Despite his proposal to maintain the Middle and South American plant as a distinct genus (Hydromystria), Hunziker rightfully con- sidered our present knowledge of this plant to be incomplete, hence a better classification depends on new data. As such, up until now, these exact but only too few isolated field observations explain neither the extent nor the frequency of flower variants in populations that would permit the taxonomic treatment of this plant throughout its known geographic distri- bution in the Americas. In like manner, the same applies to her- barium studies (Diaz-Miranda et al., 1981; Cook and Urmi-Ko- nig, 1983) that make assertations concerning classification of this plant based on the meager existence of dried specimens, whose exact identity based on the number of flower parts is for the most part unknown. These treatments have not recognized the occur- rence of flower variants in living populations nor on dried her- barium material. Consequently, due to insufficient field studies, not only the taxonomy but the developmental biology has been misinterpreted for this plant. At different geographical locations in the Americas, I have col- lected and observed the extent of variation in unisexual flowers along with their frequency of expression. The population at each locality was recognized as a dynamic and natural unit demon- strating observable biogenetic tendencies which relate to the num- ber and kind of flower structures found in flower variants. Based on these new findings, I present a corresponding infraspecific Classification. 114 Rhodora [Vol. 94 Table 1. Traits characterizing subspecies of Limnobium spongia. Primitive Derived . Spongia L. spongia Character subsp. spongia subsp. /aevigatum Male: Petals essentially unlobed 1-2 and more lobes Stamens and many (form 34 trimerous few (form 1-3 trimerous staminodes whorls) whorls) Female: Petals 3 conspicuous, essentially apetaly common or petals unlobed less than 3, with 1-2 and more lobes Staminodes more numerous fewer Stigmas many deeply-bifid (fill 2nd fewer, somewhat less-bifid and 3rd trimerous stylar (fill Ist and 2nd trimer- whorls) ous stylar whorls) Carpels almost closed, globose open, elliptic ovary one lo- ovary with 6-9 parietal culed with 3-6 intruding placental divisions parietal placentae (Fig. 2f-g) (pseudolocules, Fig. 3f-) SUBSPECIES CHARACTERIZED: PRIMITIVE VERSUS DERIVED TRAITS Critical observations of floral structures disclose a consistent and continuous integration in the number and form of stamens, stigmas, staminodes, petals, locules and carpels that negates the existence in the Americas of two distinct monotypic genera, Hy- dromystria and Limnobium (Ascherson and Giirke, 1889; Ryd- berg, 1909; Diaz-Miranda et al., 1981: Hunziker, 1982), and of even the single genus Limnobium having two or more species (Dandy, 1959; Cook et al., 1974: Cook, 1982; Cook and Urmi- Konig, 1983). These new field results based on knowledge of living populations elucidate a single genus with just one very dynamic species, Limnobium spongia (Bosc) Steudel, for which a subspe- cies distinction was determined based on the inseparable nature of two diverging developmental patterns that demonstrate lati- tudinally phylogenetic trends for the establishment of primitive and derived floral traits. These traits which characterize the two recognized subspecies of Limnobium spongia are summarized in able 1. 1992] Lowden—Limnobium 115 30° 0°} | 30° i 10°} : | ! —10° or 40° + 4 — § ed | 3 10° | \ ! ; \ \ oo cc ee eee i Ben a od cal wie 90° 70° 50° 30 essere... Figure 1. Distribution of the monotypic genus Limnobium in the Americas, showing L. spongia subsp. spongia (@) in the United States and L. spongia subsp. laevigatum (©) in the West Indies (Greater and Lesser Antilles), Mesoamerica and South America. Throughout its known geographic distribution in the Americas (Figure 1), Limnobium spongia is a species of great vegetative and reproductive plasticity. Its transitional nature is reflecte by the terrestrial to aquatic habitats it occupies and by its variable imperfect floral states, which demonstrate a latitudinal continuum In specialization from one subspecies to the other, Limnobium spongia subsp. spongia is more northern in latitude (limited to the eastern United States, Figure 1). In its northern- 116 Rhodora [Vol. 94 IX VY i =e an NIA : Figure 2. Staminate and pistillate plants of Limnobium spongia subsp. lae- vigatum from the Dominican Republic (Lowden 3601): a. floating ae segment with fleshy aerenchymous leaf blades and pedicel; b. ventral view of male flower having 3 sepals, 4 oetia and 6 "ati stamens; c. single longitudinally dehiscing mucronate-tipped stamen; d. fem ale —— a 1992] Lowden—Limnobium 117 most range (Horseshoe Lake, Illinois), this subspecies appears to have a greater number of floral parts with fewer structural mod- ifications. Plants are generally larger, having expanded leaf blades with prominent nerves. In both male and female flowers, three unlobed petals is a relatively constant trait. On the other hand, Limnobium spongia subsp. laevigatum oc- curs in more southern latitudes (south of the United States, Figure 1). Here more specialization takes place in latitudes further south (from the Dominican Republic to Argentina), where there is ev- idence of greater reductions in the number of floral parts. Plants are usually smaller and leaves inconspicuously nerved. More bi- lobed and irregularly-lobed petals are found in flowers of both sexes of this subspecies. In female flowers, apetaly and connation between concordant stigmatic lobes are considered to be highly specialized conditions, just as is reduction in fused carpel walls of fruits. FLORAL STRUCTURES: UNISEXUAL FLOWERS This comparative study of the gross floral morphology treats flowers collected at anthesis, which occurs between noon and late afternoon before the sun loses its thermal intensity. At this time of day in all populations, flower stalks are completely extended and a spectacular array of both male and female flowers can be observed above the water surface. Plants collected otherwise might be partly responsible for the misconception that plants are “rare” or “uncommon,” stamens are “sessile” (Hauman, 1915; Hun- ziker, 1981), male flowers are “‘less common” and female inflo- a are “much more commoner” (Cook and Urmi-K6nig, Female and male flowers of Limnobium have an outer reflexed perianth whorl (Figures 2b and 2e) consisting of three, rarely four (in female flowers), purplish-tipped enrolled sepals. The petals in the alternating inner perianth whorl may be absent (Figure 2e), reduced to short rudiments (Figure 3e), or exceed by 1-2 mm the ‘i aving 3 sepals, 0 petals, 3 awl- 2 hairy stigmatic lobes (6 deeply pendent fruit without visible CMM. Plant segment at anthesis; e. pistillate flower h shaped staminodes (1-1-1 opposite sepals) and | bifid stigmas); and f. lateral view of a mature elliptic Partitions in cross section, g. Cited specimen is deposited at U 118 Rhodora [Vol. 94 Figure 3. Limnobium spongia subsp. /aevigatum (a)+e): a.—c. from the Do- minican Republic (Lowden 3601): leafy seg f ious plant showing emergent rooted habit with staminate flower at anthesis and mature fruits of pistillate inflorescence borne on connective stolon; b. ventral and c. lateral views of floating juvenile plant propagators showing first true leaf and root; d. female | 1992] Lowden—Limnobium 119 length of the sepals (Figure 4). Three erect whitish-mauve petals (Figure 4) characterize male flowers, whereas in female flowers there is an obvious reduction in the number and size of petals. In Limnobium spongia subsp. spongia, female flowers have three well-developed petals (Figure 3h), while in L. spongia subsp. /ae- vigatum they show a conspicuous apetalous trend (Figures 2e and 3d-e). Lobing of petals is also a significant trait in male and female flowers. The presence of unlobed petals predominates in both sexes of Limnobium spongia subsp. spongia (Figures 3h and 4a), whereas there are more bilobed petals (Figures 3d-e and 4b) and irregular petals with 3-4 minute lobes (Figure 4d) in L. spongia subsp. /aevigatum. In this latter subspecies, bilobed petals appear to be more prevalent in male flowers, since fewer petals exist in female flowers due to apetaly. Contrary to former reports (Kaul, 1968; Diaz-Miranda et al., 1981; Hunziker, 1982), my observations on male flowers reveal neither a constant number of stamens nor the absence of stami- nodes. A highly variable number of mucronate-tipped stamens (Figures 2c and 4) and awl-shaped staminodes (Figures 4a and 4c-d) make up the staminal column consisting of alternating trim- erous whorls of connate staminal bases. These staminal parts (Figure 4) develop acropetally in the androecium where stamens or staminodes belonging to the first, third and fifth trimerous whorls are found opposite the sepals, while those in the second, fourth and sixth trimerous whorls are found opposite the petals. Stamens and staminodes become progressively smaller the higher up One moves on the staminal column, where eventually stamens may be reduced to staminodes that appear to be only anther- Stalks (Figures 4a and 4c-d). In Limnobium spongia subsp. spon- = sepals) and 6 stigmatic lobes (3 bifid stigmas). Limnobium spongia subsp. spongia (f}-(h): f. two pendent globose fruits from Florida (Lowden 3937) showing in cross section, &- visible partitions of 9 pseu- dolocules; and h. female flower from Florida (Lowden 393 9) having 3 sepals, 3 Petals (2 shown), 7 staminodes (6 shown of 3-2-2 opposite sepals) and 17 stigmatic lobes (8 stigmas: 7 bifid and 1 trifid). Cited specimens are deposited at UCMM. 120 Rhodora [Vol. 94 Figure 4. Floral parts in male flowers of Limnobium spongia subsp. spongia (a) and L. spongia subsp. laevigatum (b)(f): a. 3 sepals, 3 unlobed petals and 13 3 bilobed petals and 9 fertile stamens from the Dominican Republic (Lowden 394 7&3 sepals, 3 unlobed petals and 7 fertile stamens (7th just a half anther) with | staminode from Argentina (Lowden 3952); d. 3 sepals, 3 petals (varying 1992] Lowden— Limnobium 121 gia, the androecial column consists of 3-4 trimerous whorls of fertile stamens (Figure 4a), while in L. spongia subsp. /aevigatum there are fewer fertile stamens (Figures 4b-f); just the first three trimerous whorls become filled. In northern latitudes the general tendency is toward twelve staminal parts in Limnobium spongia subsp. spongia, while in southern latitudes the trend is toward Six staminal parts in L. spongia subsp. laevigatum. No real evidence was observed in fresh male flowers of pistillate parts along the central floral axis. In female flowers no true stamens are found, only rudiments existing in one whorl as subulate antesepalous staminodes (Figure 2e). In both subspecies, these staminodes are frequently found in pairs. On the other hand, staminodes increase in numbers with a corresponding increase in the number of stigmatic lobes in Limnobium spongia subsp. spongia (Figure 3h), whereas female flowers with fewer stigmas have solitary staminodes opposite se- pals in L. spongia subsp. laevigatum (Figures 2e and 3e). Even a true pistillate flower, without staminodes (Mosquera, Colombia; Lowden 3929), was observed in this latter subspecies. The superior portion of the gynoecium in these imperfect fe- male flowers consists of complete or incomplete alternating trim- erous whorls having usually bifid stigmas (Figure 3e). Stigmas having one or three lobes occur infrequently. The three bifid stig- mas (six stigmatic lobes) in the first (Figure 3e) and third whorls are found opposite the petals (if absent, then alternate with the sepals), while stigmas in the second trimerous whorl are found Opposite the sepals and staminodes. In general, female flowers of Limnobium spongia subsp. spongia have more deeply bifid stig- mas filling the second to the third trimerous stylar whorls, while fewer but somewhat less bifid stigmas fill just the first to the second trimerous stylar whorls in L. spongia subsp. laevigatum. The inferior compound ovary bears few to many ovules borne On intruding parietal placental partitions formed by the connation of adjoining ventral carpel margins which show no signs of com- plete closure. The number of stigmas and placental projections 1s a from 2-4 lobed) and 6 fertile stamens with | staminode from Colombia (Lowden 3933). e. 3 sepals, 3 unlobed petals and 6 fertile stamens from the Dominican Republic (Lowden 3601); and f. 3 sepals, 3 petals (2 slightly bilobed) and 5 fertile stamens from Argentina (Lowden 3950). Cited specimens are deposited at UCMM. iz Rhodora [Vol. 94 indicative of the number of fused carpels. The almost-closed car- pels of the globose fruits (Figures 3f-g) in Limnobium spongia subsp. spongia have six to nine parietal placental divisions (pseu- dolocules), while the open carpels of elliptic fruits (Figures 2f-g) in L. spongia subsp. laevigatum have just one locule with usually three to six intruding parietal placentae. An overall comparison of sexes in this monoecious plant (Lim- nobium spongia) reveals that male flowers more closely represent the imperfect state (unisexual), as evidenced by absence of pis- tillate parts, whereas female flowers present with great consistency an ancestral whorl of staminal rudiments opposite sepals, thus calling attention to a previous bisexual origin. Staminodes are well-developed in both sexes, yet frequently overlooked. DIVERSITY IN POLLINATION MECHANISMS Contrary to former belief (Cook, 1982; Cook and Urmi-K6nig, 1983), the pollination mechanism in the genus Limnobium is not constant as in other genera of the Hydrocharitaceae. In this family, where the primitive nature of Limnobium is well known, it is not surprising to find a variety of pollination mechanisms (Hauman, 1915), depending on which floral modifications are present. As such, the sexual structures evolving in these imperfect flowers present a specialization in pollination mechanisms. My field experience indicates that wind and insects appear to be the principal agents of pollination in both subspecies of Lim- nobium spongia. The slightest movement of the extended male pedicels (Figure 2a) brings down showers of sulfur-colored pollen onto the plumose stigmatic lobes of the shorter-stalked female flowers (Figure 2d). Powdery pollen masses reaching the water surface even float to stigmas of partially submerged female flow- ers. In nature, large numbers of minute aphid nymphs (Aphididae) and their predators, the Ladybird beetles (Coccinellidae), are fre- quent visitors that settle on the floral parts at anthesis. Pollination by wind (anemophily) is most likely more effective in Limnobium spongia subsp. spongia where fertile stamens and stigmas are more abundant. On the other hand, in the more spe- cialized flowers of L. spongia subsp. /aevigatum insect (ento- mophily) and even surface pollinations might be more effective due to the reduction in the amount of pollen and stigmatic surfaces in these flowers having fewer sexual parts. 1992] Lowden— Limnobium 123 FLORAL VARIATION AND CHROMOSOME COUNTS Cook and Urmi-KG6nig (1983) pointed out that in the Hydro- charitaceae, “‘one may find a greater degree of variation some- times in a single species,” such as in Ottelia alismoides (L.) Per- soon which has “3-12 stamens, 3-10 carpels and may or may not develop dissepiments.” They considered that this range of vari- ation exceeded that found in Limnobium. My field study reveals that Limnobium is just as, or even more, variable than Ottelia alismoides. Limnobium spongia has 3-14 stamens and 3-10 carpels with or without developed dissepi- ments. Considerable aneuploidy occurs in both of these species. As such, variation is even more accentuated in Limnobium where chromosome counts can be correlated with morphology and ge- ography, while in Ottelia alismoides attempts to find correlations of this nature have failed (Cook and Urmi-KG6nig, 1984). In Limnobium there exists a definitive relationship between character specialization, geographic latitude and chromosome number (Table 2). Significant differences in frequencies of floral variants for given population localities permit this overall for- mulation of male and female flower types in relation to known chromosome reports. Chromosome counts (Table 2) elucidate an aneuploid series (2n = 26, 27, 28, 29 and 30) in Limnobium spongia. In the northern range of the species, the base number of x = 6 has been reported from Paynes Prairie, Florida (Bernardello and Moscone, 1986). Here, the more primitive L. spongia subsp. spongia shows a rel- ative constant diploid count of 2” = 24, which corresponds with more stamens and staminodes in male flowers, while female flow- ers have more petals, stigmas and staminodes. On the other hand, chromosome counts from southern latitudes (Guadeloupe and Argentina) reveal plants with diploid complements of 2” = 26- 30. Frequencies for these counts (in parentheses) indicate a pro- gressive increase in the diploid number as one goes further south where flower specialization is evidenced with fewer floral parts in plants of both sexes of L. spongia subsp. /aevigatum. Supposing this to be the case, one would expect to find more counts of 2n = 30 in plants at the southernmost extreme of the distribution of this subspecies (Figure 1; San Miguel Del Monte, Argentina). Further studies in the genus Limnobium should take into ac- count the existence of these floral types in nature. In this manner, exact relationships might be established between cytotypes and 124 Rhodora [Vol. 94 Table 2. Relationship between character specialization, geographic latitude and chromosome number in Limnobium spongia. Subspecies Overall Floral Type Locality & Diploid Counts Subsp. spongia Subsp. /aevigatum male: 12 stamens, pet- Is 1-lobed female: 3 unlobed pet- als, 8-9 stigmas, staminodes 2-2-2 to 3-3-2 male: 6 stamens, petals 1- 2-lobed female: 0-2 petals, 1-2- lobed; 5-6 stigmas; staminodes 1-1-1 to 2-1-1 male: 6 stamens, petals 1-2-lobed female: 0-1 petals, un- lobed; 5-6 stigmas; staminodes 2-1-1 to 2-2-1 male: 6 stamens, un- lobed petals, rarely 2-lobed female: 0-2 petals, 1-lobed, rarely 2-lobed; 5-6 stigmas; staminodes |-1-1 to 2-1-1 U.S.A.: Florida Paynes Prairie; 2n = 24 (96 cells from 68 seed- lings) (Bernardello & Moscone, 1986) Guadeloupe: West Is- land, Maire-Galante; 28(4), 29(2), 30(1) _ (Cook & Urmi-K6nig, 1983) Argentina: Cordoba— Cabalango & Embalse Rio II; 2n = 28 (75 metaphase plates from 51 plants) (Moscone & Bernardello, 1985) Argentina: Buenos Aires, Riachuelo; 2n = 26(4), 27(3), 28(5), 29(5) (Cook & Urmi-K6nig, 1983) their corresponding morphological floral countertypes. Field ex- perimentation between both subspecies of Limnobium spongia will most likely disclose the genetic pathways existing between floral traits and chromosomal germplasm. A better understanding of these cytological and morphological variation patterns in this primitive genus would undoubtedly extend our knowledge of phy- logenetic trends in the Hydrocharitaceae. TAXONOMY AND GEOGRAPHY Limnobium L. C. Richard, Mém. Classe Sci. Math. & Phys. Inst. France (Paris), pp. 66-67, fig. 8. 1811 (pt. 2, publ. 1814). 1992] Lowden—Limnobium 125 Hydromystria G. F. W. Meyer, Prim. Fl. Esseq. 152-153. 1818. Jalambicea Cervantes in De La Llave & Lexarza, Nov. Veget. Descr. 2: 12. 1825. Rhizakenia Rafinesque, Autikon Botanikon, p. 188. 1840 Trianea Karsten, Linnaea 28: 424. 1857. Hydrocharella Spruce ex Bentham in Bentham & Hooker, Genera Plantarum 3(2): 1883 Plants monoecious (Fig. 3a), perennial aquatic herbs, stems trailing and rooted at nodes, leaves floating (aerenchymous) or aerial, blades rounded, cordate to spatulate, principal nerves con- verging (Figs. 2a & 2d). Flowers imperfect (Figs. 2b & 2e), in- conspicuous, true petals white-mauve, in both sexes slightly lon- ger and rarely as broad as the green-purplish sepals (Fig. 2b), apices narrowly rounded to irregularly lobed (Figs. 3d-€ & 4), or petals reduced to rudiments or absent only in female flowers (Fig. 2e). STAMINATE FLOWERS (Fig. 4) with stamens and staminodes glabrous, united in one staminal column arising acropetally in alternating antesepalous or antepetalous trimerous whorls (Fig. 4b), anthers much longer than broad, mucronate-tipped (Fig. 2c). PISTILLATE FLOWERS (Figs. 2e, 3d-e & 3h) with staminodes gla- brous, arranged in only one antesepalous whorl (Fig. 3h), stigmas bifid nearly to the base, exceeding the length of the perianth. In the New World, the genus Limnobium might be confused in North America with the European frogbit, 7 ydrocharis morsus- ranae L., which has been introduced in Ontario and Quebec, Canada (Louis-Marie, 1958; Catling and Dore, 1982). Only re- cently, the European frogbit has spread from these Canadian prov- inces to New York (Roberts et al., 1981) where the American Limnobium spongia (Bosc) Steudel reaches its northern limits in the eastern United States (Figure 1). Despite their vegetative sim- ilarity, these two species are morphologically distinct, based on flower characters. The European frogbit is a dioecious plant whose showy white- yellowish flowers have broadly-rounded petals that are much lon- ger and broader than the sepals. In male flowers, the anthers of stamens are more or less oval, nearly as long as wide, bases of filaments are united in pairs, staminodes are hairy (at times so Is the filament of the innermost stamen). In female flowers the an- tepetalous staminodes are glandular, while the antesepalous ones are elongate and hairy; stigmas are bifid for one-half their length and much shorter than the petals. 126 Rhodora [Vol. 94 Limnobium is a monotypic genus, widely distributed in the temperate and tropical parts of the Americas. Type species: Limnobium spongia (Bosc) Steudel 1. Limnobium spongia (Bosc) Steudel, Nom. Bot. ed. 2, 45. 1841. Type basionym: Hydrocharis spongia Bosc, Ann. Mus. Hist. Nat. Paris 9: 396-398, pl. 30. 1807. U.S.A., “‘Basse Caro- line,” Bosc (G, P). Aquatic plants, floating to emergent, forming dense floating mats or inhabiting mudflats, roots fibrous, stems stoloniferous, nodes form leafy rosettes; leaves stipulate, rounded cordate or spatulate, green-purplish, entire, with 3-7 longitudinal nerves, cross-nerves less conspicuous, floating blades usually aerenchy- mous with short petioles, aerial blades with longer petioles and lacking a central spongy layer. Inflorescences cyme-like, sessile or short-pedunculate, subtended as flowers by floral (spathe) bracts, lanceolate to oblanceolate, few- to many-flowered, opening usu- ally one at a time. Flowers unisexual, pedicels extended at an- thesis, reaching approx. 11.0 cm in both sexes, sepals 3—4 (usually 3), boat-shaped, wider than petals, petals 0-4 (usually 3, except for the obvious apetalous trend in female flowers south of the United States), linear to lanceolate, slightly longer than sepals, unlobed or irregularly 24 lobed (2-lobed petals are not uncom- mon). STAMINATE FLOWERS with 3-14 fertile stamens (6 or !2 common), anthers approx. 4.0 mm long with mucronate tips, longitudinally dehiscing (Fig. 2c), 0-6 awl-shaped staminodes (if present, usually one), approx. 0.5—-4.0 mm long. PISTILLATE FLOW- ERS with 0-4 antisepalous staminodes (usually 1-2), awl-shaped, reaching 4.0 mm long, 3-10 bifid hairy stigmas (6-19 stigmatic lobes), however, 5-6 are common (10-12 stigmatic lobes). Fruits pendent, beaked at apex, elliptic to globose (Figs. 2f & 3f), many seeded, berry-like, unilocular, placentation parietal, 3-9 intruding placental partitions, mature seeds beaked and covered with tr- chomes, approx. 0.1-0.3 mm long, seedling propagators (Figs. 3b-c) floating and inconspicuous. In the Americas, Limnobium spongia occupies a variety of ecological habitats ranging from ponds, lagoons, swamps and lakes along hilly coastal plains at sealevel to mountain plateaus reaching 3000 m (Bogota, Colombia). However, there is apparently n0 1992] Lowden—Limnobium 127 variation in plants because of the change in altitude. Other aquatic plants commonly associated with Limnobium spongia are Eich- hornia crassipes (Mart.) Solms, E. azurea (Sw.) Kunth, Pistia stratiotes L., Lemna spp., Wolffiella sp., Nelumbo lutea (Willd.) Pers., Nymphaea spp., Azolla sp., Salvinia sp. and Ricciocar- pus sp. KEY TO THE SUBSPECIES OF LIMNOBIUM SPONGIA la. Leaves cordate-based (especially in juvenile and floating blades) to rounded; fruits globose, 6-9 placental divisions; female flowers usually with 3 petals, number of bifid stig- mas usually 6-9 (i.e., 12-18 stigmatic lobes); male flowers frequently have 12 fertile stamens (9-11 stamens not un- ut: | ———— aes L. spongia subsp. spongia lb. Leaves spatulate (rarely cordate based in floating blades); fruits elliptic, 3-6 intruding parietal placentae; female flow- ers without petals or if present usually less than 3, number of bifid stigmas commonly 4-6 (i.e., 8-12 stigmatic lobes); male flowers frequently have 6 fertile stamens (7-9 stamens not uncommon) ......... L. spongia subsp. laevigatum In general, plants are larger in Limnobium spongia subsp. spon- gia as compared with a decrease in size and increase in number of plants per unit area in L. spongia subsp. laevigatum. Big aerial plants of the American frogbit, L. spongia subsp. spongla, reach the height of mature water hyacinth (Eichhornia crassipes) in rivers and lakes of Florida in the southeastern United States. Here, from the “St. John’s River runs and lakes to Lake Okeechobee bays” (Bodle, 1986), these large plants of L. spongia subsp. spon- gia inhibit water flow or navigation and have been the target of an aquatic plant control program. On the other hand, in C olombia the smaller (less conspicuous) forms of L. spongia subsp. laevi- Satum flourish in drainage ditches where they recycle nutrients from wastewaters on the Bogota Plateau. In both subspecies there are exceptionally small flowering plants that inhabit mudflats in juxtaposition with larger aerial-flowering Plants which normally float in entangled mats. The scarcity of these smaller flowering plants at northern latitudes made it 1m- Possible to determine further local relationships between plant 128 Rhodora [Vol. 94 size and number of flower parts beyond those already established for the overall distribution of each subspecies. la. Limnobium spongia (Bosc) Steudel, subsp. spongia, Nom. Bot. ed. 2, 45. 1841. Type basionym: Hydrocharis spongia Bosc. ee spongia Bose. Ann. Mus. Hist. Nat. Paris 9: 396-398, pl. 30. 1807. Basse-Caroline Limnobium bosci L. C. Richard, Mém. Classe Sci. Math & Phys. Inst. France is), pp. 32-34, fig. 8. 1814 (pt. 2, 1811 ie. noiielia Nuttall, Gen. North Am. ‘PL 2: 241. 1818. me Plants taller (reaching approx. 5 dm), leaves with 5—7 principal nerves, secondary nerves usually visible, floating and aerial leaves cordate-based or rounded. Flowers of both sexes (Figs. 3h & 4a) usually with 3 well-developed unlobed petals (female flowers rare- ly with 2 or 4 petals; bilobed petals infrequent in both sexes). STAMINATE FLOWERS (Fig. 4a) with 8-14 (12 usually) fertile sta- mens and 0-5 staminodes (frequently one or even more present) in the upper (third through the sixth) trimerous whorls of an extended staminal column (Fig. 4a). PistrLLATE FLOWERS (Fig. 3h) with more (up to 4) staminodes in groups opposite sepals, stam- inodes robust, stigmatic lobes numerous (commonly 12-18). Fruits globose (Figs. 3f-g; approx. 1.0-1.6 cm wide) with 6-9 parietal placental divisions or pseudolocules, at maturity the outer ovary walls decay leaving partit Throughout most of its distribution, this subspecies is a fre- quent inhabitant of swamps of bald cypress (Taxodium distichum (L.) Richard). Of particular interest was an herbarium specimen collected by Rogers (Rogers, 8773-A) from Rankin County, Mis- Sissippi. This dried specimen of a male flower has an unusual number of staminal parts: ten stamens and nine staminodes. The innermost structure extending from the staminal column appears to be a bilobed hairy stigma. Fresh material of male and female flowers from this particular locality might make it possible to ascertain the somewhat doubtful nature of this dried flower. Up until now, this male flower is the only one which I have seen with signs of a bisexual state. a (Figure 1.) This subspecies is endemic to tem- p regions in the United States where it mainly —_—— - 1992] Lowden— Limnobium 129 occurs along the Gulf of Mexico and the Atlantic coastal lowlands from Louisiana-Florida-Delaware in the south, reaching a north- ernmost limit in the state of New York (Braddock Bay, Lake Ontario) in the northeast, extending to the southwest from the Mississippi River delta into Oklahoma and Texas, and in the Mississippi valley north to southern Illinois and southeastern Missouri in the midwest. The unsubstantiated reports (Catling and Dore, 1982) from Monmouth County, New Jersey (Mackenzie, 1922; Fassett, 1957) and Lake County, Indiana (Deam, 1940) have not been included in this study, as well as specimens not seen by this author but cited by others for the states of Oklahoma (Langdon, 1959) and Kentucky (Catling and Dore, 1982). lb. Limnobium spongia (Bosc) Steudel, subsp. laevigatum (Hum- boldt & Bonpland ex Willdenow) Lowden, comb. nov. Type basionym: Salvinia laevigata. Salvinia laevigata Humb. & Bonpl. ¢ ex Willd., Sp. Pl. 5: 537. 1810. [“St. Fe de Bogota”] (HOLOTYPE: B; ISOTYPE gta stones F.W. Meyer, Prim. Fl. Esseq., 152-153. 1818 [“plan- amburg,” Guyana, Rio Essequibo] (HOLOTYPE: GOE 7): Jalan Aig Cervantes in De La Llave & Lexarza, Nov. Veget. Descr. 2: 5. [““Frequentissima in defossis Mexicanis”]. Pata sinclairii Bentham, Bot. H.MLS. ae 175. 1844. [“Guayaquil”’] OLOTYPE AND ISOTYPE: K). Trianea bs Karsten, Linnaea 28: 424-425. 1857. [“in planitiei Bogota- nae”’} (Iso US). Limnobium Walenitenive (G. F. W. Meyer) Grisebach, Fl. Brit. West Ind. Isl. 506. 1862. (Hab. Trinidad!, Cr.; [Guiana!]’’) ee echinospora Spruce ex Bentham in Bentham & Hooker, Gen. PI. 3(2): 452. 1883. Brasil, [“flum. Amazonum. ad ostium flum. Solimoes”], cai 593 (HOLOTYPE: BM; ISOTYPES: G, K). Hydrocharis se ate (G. F. W. Meyer) O. Kuntze, Rev. Gen. Pl. 3(3): 297. oncepcion de Paraguay”’]. Hydromystria sinclairi (Bentham) Huaman, Anales Mus. Argent. Ci. Nat. “Ber- nardino Rivadavia” (Buenos Aires) 27: 326-327. 1915. Limnobium iabiesce (Karsten) Delay, Bull. Soc. Bot. France 88: 481. 1941. gor, laevigatum (Humb. & Bonpl. ex Willd.) Heine, Adansonia 8: 315. 968. laude rig agi (Humb. & Bonpl. ex Willd.) Morton, Contr. U.S. Nat. Herb. 38: 270. 1 ae aeviga (Humb. & Bonpl. ex Willd.) A. T. Hunziker, Lorentzia 5-8. Myaromysia —— (Humb. & Bonpl. ex Willd.) Diaz- Miranda et al., Bot. J. n. Soc. (London) 83(4): 318. 1981. 130 Rhodora [Vol. 94 Plants shorter (Figs. 2a, 2d, & 3a; reaching approx. 4.5 dm), leaves (Fig. 3d) with 3-5 principal nerves, secondary nerves in- conspicuous, floating and aerial leaves (Figs. 2a & 2d) spatulate, rarely cordate-based. Only male flowers (Fig. 2b) usually with 3 petals (2 or 4 petals occur), while in female flowers petals reduced in size and number to the point where apetaly is common (Fig. 2e). In both sexes bilobed petals (Figs. 3d & 4b) are generally just as pronounced as unlobed petals (with an increase to 3-lobed and 4-lobed petals in female and male flowers, respectively. STAMI- NATE FLOWERS (Figs. 2b & 4b—f) with 3-10 (6 usually) fertile stamens and 0-3 staminodes (frequently absent but when present, One is common) in the upper (second through the fourth) trim- erous whorls of the shortened staminal column (Fig. 4e). PIsTIL- LATE FLOWERS (Figs. 2e & 3d-e) with zero (rare) or few (up to 3) staminodes opposite sepals, staminodes weak, stigmatic lobes fewer (Commonly 8-12). Fruits elliptic (Figs. 2f-g; approx. 1.0- 1.5 cm long to 0.4-3.5 mm wide) with 3-6 intruding parietal placentae, unilocular. On the Caribbean islands of Puerto Rico and Martinique, I did not find any populations. In Puerto Rico, urban developments have destroyed habitats at Loiza Aldea; sugar cane plantations now extend from Arroyo to Patillas, just as rice fields and cattle pastures predominate from Vega Alta to Vega Baja and Florida. In Martinique, the ponds amongst the coastal hills have dried up and the Caravelle Peninsula today supports only a very dry bush vegetation affected by the encroaching sugar cane fields and live- stock. Deserving special mention are those herbarium materials that helped to delineate the geographic limit of Limnobium spongia subsp. /aevigatum in Middle America. The drawing of the Mex- ican plant (Sesse & Mocina, F) shows clearly a male flower with 6 stamens and a female flower with 12 stigmatic lobes. Male specimens from Nicaragua (Nichols 1058, Mo), El Salvador (Roh- weder, GH) and Costa Rica (Crow & Charpentier 5933, Mo) have one, two or three bilobed petals, respectively, and 6 fertile sta- mens. Only the specimen from El Salvador has in addition one staminode. In the Greater Antilles, a Cuban specimen (Leon 9032, NY) disclosed diagnostic data for both sexes. The female flower has zero petals, 4 staminodes, 6 bifid stigmas and an ovary without dissepiments, while the male flower has 7 fertile stamens. Spec- 1992] Lowden— Limnobium 131 imens from Puerto Rico show 6 stamens (Britton & Britton 7949, ny) and elliptic fruits (Sintenis 5779, Mo). DISTRIBUTION: (Figure 1.) This tropical to temperate (in south- ern latitudes and altitudes) subspecies is found throughout the West Indies on the Caribbean islands of Cuba, Hispaniola (Do- minican Republic) and Puerto Rico in the Greater Antilles, and Antigua, Montserrat, Guadeloupe, Martinique, St. Lucia and Trinidad in the Lesser Antilles; in Mesoamerica it mainly occurs along the Pacific lowlands of Mexico, Guatemala, El Salvador, Nicaragua, Costa Rica and Panama; and in South America it occurs in Colombia, Venezuela, Guyana, Surinam, French Gui- ana, Ecuador, Peru, Brazil, Paraguay, Uruguay and Argentina. Worthy of mention are references to specimens from localities not personally examined but cited by other authors for Montserrat (Howard, 1979) and Trinidad (Simmonds, 1967). The existence of Limnobium spongia subsp. laevigatum in Guyana is based on the type specimen of Hydromystria stoloniferum (“‘Essequibo.”’) in the Grisebach Herbarium at Goettingen. Of further interest are reports of localities in southwestern Brazil (Hoehne, 1948: Co- rumba and Caceres, State of Mato Grosso), which lead one to expect further plant collecting in Brazil (Figure 1) will reveal a more than infrequent occurrence of this plant. CONCLUSIONS Plant collectors should recognize flower variants in nature and determine the number of flower parts along with their frequency of expression in populations. This relevant collection data should accompany locality information on herbarium labels. These kinds of field observations will be most helpful in understanding the real extent of variation in populations as compared to dried her- barium specimens which demonstrate none, or at most one, of these flower variants. The dissemination of Limnobium spongia in the Americas ap- Pears to be through natural means instead of by artificial intro- ductions. Perhaps this natural dissemination has contributed to the expression of a continuum in characters which may be ob- served as two diverging developmental patterns between North and South American latitudes. The overall similarity of flower 132 Rhodora [Vol. 94 variants indicates that these developmental patterns have evolved from a common gene pool. Inbreeding studies would be of great interest for experimentally determining the exact relationships between cytotypes and their corresponding morphological floral counterparts. Populations of- fer the diversity in flower variants, pollination mechanisms and cytotypes that make Limnobium spongia an attractive laboratory organism for a better understanding of species specialization and its phylogenetical implications in the Hydrocharitaceae. ACKNOWLEDGMENTS This research is a contribution to the Aquatic Vascular Flora Project of Hispaniola directed by the author at the Moscoso Her- barium (UCMM), Pontificia Universidad Catélica Madre y Maes- tra, Santiago, Dominican Republic. The University’s Research Center is most appreciated for its support in realizing field ex- cursions to plant population localities in the United States, Lesser and Greater Antilles, and South America. Particular thanks are extended to Armando Hunziker (Cérdoba), Carmen and Antonio Krapovickas (Corrientes), Nuncia Tur (La Plata), Baltazar Tru- jillo (Maracay), Santiago Diaz-Piedrahita (Bogota), Gustavo Mo- rales L. (Bogota), Walter Judd (Gainesville), Jim Watson (Miami) and Alicia Lourteig (Paris) for their assistance in the field and fulfillment of special herbarium requests. I acknowledge the Department of Photolithography of the Com- pania Anonima Tabacalera (Santiago) for reducing the figures. The following institutions are recognized for their herbarium ser- vices: A, COL, CORD, CTES, F, FLAS, FTG, GH, GOET, LP, MO, MVFA, MVJB, MY, NY, OS, P, SI, US, VEN. LITERATURE CITED ASCHERSON, P. AND M. Girke. 1889. Hydrocharitaceae, pp. 238-258. In: A. Engler and K. Prantl, Eds., Natiirlichen Pflanzenfamilien, Vol. 2(1). BERNARDELLO, L. M. AND E. A. Moscone. 1986. The karyotype of Limnobium spongia (Hydrocharitaceae). Pl. Syst. Evol. 153: ie Bopte, M. J. 1986. American frog’s bit. Aquatics 8: 4 Bosc, L. 1807. Description de la Moréne a Eponge. kik spongia.) Ann. Mus. Natl. Hist. Nat. 9: 396-398, pl. 30. 1992] Lowden—Limnobium 133 CaTLING, P. M. AND W.G. Dore. 1982. Status and identification of Hydrocharis morus-ranae and Limnobium spongia (Hydrocharitaceae) in northeastern North America. Rhodora 84: —545. Cook, C. D. K. 1982. Pollination mechanisms in the Hydrocharitaceae, pp. 1—- 15. In: J.-J. Symoens, S. S. Hooper and P. Compére, Eds., pris on Aquatic Vascular Plants. Royal Botanical Society of Belgium, Brusse . Arevision of the xe oe including Hydromystria (Hydvochacitaceae), Aquatic Bot. 17: —— anp ———.. 1984. A revision of the genus att (Hydrocharitaceae) 2. The species el Eurasia, Australasia and America. Aquatic Bot. 20: 177 . Gut, E. M. Rix, J. SCHNELLER AND M. SEITZ. 1974. Water Plants of the World. Ww. . Junk, The Hague. Danby, J.E. 1959. the subfamilies, tribes, and genera) pp. 540-541. In: I. Hutchinson, Ed., The Fuuilies of Flowering Plants, Vol. 2. Clarendon Press, Oxford. Deam, C.C. 1940. Flora of Indiana. Dept. Conservation, Div. Forestry, Indi- Diaz-MiRANDA, D., D. PHtLcox AND P. Denny. 1981. Taxonomic clarification of Limnobium Rich. and Hydromystria G. W. F. Meyer (Hydrocharitaceae). J. Linn. Soc., Bot. 83: 311-323. Diaz-PIEDRAHITA, S. 1985. Flora de la Real Expedicién Botanica del Nuevo Reino de Granada (1783-1816), promovida y dirigida por José Celestino Mutis. Tomo 3(1), S. Diaz-Piedrahita identificé las laminas y redacté los textos de la Hidrocaritaceas, p. 43 + 3 plates (Jar. Bot. Madrid, Vols. 25 9, 260 and 260A). Ediciones Cultura Hispanica, Madrid. Fassett, N.C. 1957. A Manual of Aquatic Plants, with — Appendix by E. C. Ogden. University of Wisconsin Press, Madiso Hauman, L. 1915. Note sur Hydromystria stolonifera Ni Anales Mus. Argent. Ci. Nat. “Bernardino Rivadavia” (Buenos Aires) 27: 325-331. Hoeune, F. C. 1948. Plantas Aquaticas. Instituto de Botanica, Secretaria da Agricultura, Sao Paulo. Howarp, R. A. 1979. Flora of the Lesser pn eae Vol. 3. Brooke Thompson-Mills, Jamaica Plain, Massachus UNZIKER, A. T. 1981. areata laevigata (Hydrocharitaceae) en el centro de Argentina. Lorentzia 4: 5-8. wen, 9087. hia biologicas y taxonomicas sobre Hydromystria lae- vigata (Hydrocharitaceae). Taxon 31: 472-4 INsTITUTO COLOMBIANO DE CULTURA HISPANICA (coleccion José Celestino Mutis, Vol. 2). 1958. Diario de observaciones de José Celestino Mutis (1760- 1790). Transcripcién, Prologo y _ - Guillermo Hernandez de Alba, Tomo 2. Editorial Minerva Ltda., Bog KauL, R. B. 1968. Floral morphology nie ee in the Hydrocharitaceae. Phytomorphology 18: 13-35, Figs. 1-77. Lancpon, K.R. 1959. Limnobium, ional g Proc. Oklahoma Acad. Sci. 40: 9. Louis-Marie, P. 1958. Quelques problemes de biologie végétale québécoise, bd i vctecntoanls morsus-ranae L., nouveau pour Am mérique. Rev. Oka Agron. Inst. Agric. 32(6): 149-150. tn the fara of Oklahoma 134 Rhodora [Vol. 94 Mackenzie, K. 1922. Th ds for Li bium spongia in the northern United States. Torreya 22: 102-104. Moscone, E. A. A AND te MM. BERNARDELLO. nes Chromosome studies on Hy- dromystria laevig ). Ann. Missouri Bot. Gard. 72: 480- RICHARD, L.C. publ. 1814 (dated 1812). Mémoire sur les Hydrocharidées. Mém. Classe Sci. Math. Phys. Inst. France (Paris), Année 1811(pt. 2): 1-81, 9 plates. Roberts, M. L., R. L. StucKEY AND R.S. MITCHELL. 1981. Hydrocharis morsus- ranae (Hydrocharitaceae): new to the United States. Rhodora 83: 147-148 RypserG, P. A. 1909. Hydrocharitaceae. North Amer. Fl. 17(1): Simmonps, N. W. 1967. Hydrocharitaceae. Flora of Trinidad and THEE 3(1): 4-6. EL HERBARIO “RAFAEL M. MOSCOSO” PONTIFICIA UNIVERSIDAD CATOLICA MADRE Y MAESTRA FACULTAD DE CIENCIAS Y HUMANIDADES SANTIAGO DE LOS CABALLEROS REPUBLICA DOMINICANA (ANTILLAS) | RHODORA, Vol. 94, No. 878, pp. 135-140, 1992 CHLORIS BARBATA SW. AND C. ELATA DESVAUX (POACEAE), THE EARLIER NAMES FOR C. INFLATA LINK AND C. DANDEYANA ADAMS JOHN T. KARTESZ AND KANCHEEPURAM N. GANDHI ABSTRACT Andropogon barbatus L. 1759 and 1760 refer to a Jamaican grass, whereas A, barbatus L. 1771, alater homonym, applies to an East Indian grass. A. polydactylon L. 1763, a renaming of A. barbatus L. 1759 and 1760, is a superfluous name. Chloris barbata Sw. 1797, which was based on A. barbatus L. 1771, must be considered to be a nom. nov. C. barbata Sw. and C. elata Desvaux are the earlier names for C. inflata Link and C. dandeyana Adams. Key Words: Andropogon barbatus, A. polydactylon, Chloris barbata, C. dandey- ana, C. elata, C. inflata, C. polydactyla INTRODUCTION In his monograph of the genus Chloris Sw., Anderson (1974) used C. dandeyana Adams instead of C. elata Desvaux [found in Florida, Mississippi, Texas? (see below), the Caribbean Islands, and South America] and C. inflata Link for C. barbata Sw. (widely distributed in warmer parts of the world; in North America, found in Texas and the Caribbean Islands). Anderson’s treatment of C. inflata and C. dandeyana has been followed by Gould (in Howard, 1979), Kartesz and Kartesz (1980), Soil Conservation Service (1982), and Hatch et al. (1990). Although Anderson (in Gould, 1975, p. 328, in the protologue of C. canterai Arech.) considered Gould and Box’s 1965 report of the occurrence of C. polydactyla (= C. elata) in Texas to be an error, Hatch et al. (1990) reinstated it for plants from Texas. Our analysis of the nomenclature follows. CHLORIS BARBATA SW., AN EARLIER NAME FOR C. INFLATA LINK Anderson (1974, p. 34-36) discussed the taxonomy and no- menclature of Andropogon barbatus L. (Linnaeus, 1759, p. 1305; 1760, p. 412; 1771, p. 302) and A. polydactylon L. (Linnaeus, 1763, p. 1483). Although the protologues of A. barbatus L. 175%, A. barbatus L. 1760, and A. polydactylon differed slightly from each other, Anderson demonstrated that the preceding three names 135 136 Rhodora [Vol. 94 were based on the same Jamaican specimen. However, A. bar- batus L. 1771 was based on an East Indian specimen, belonging to a different species of Chloris. Anderson (p. 36) considered A. barbatus L. 1771 as a later homonym. Since Swartz (1797, p. 200) based his C. barbata on A. barbatus L. 1771, Anderson rejected C. barbata Sw. and accepted C. inflata Link (published in 1821). Due to its wide distribution in the tropics, we decided to inves- tigate its post-1974 nomenclatural literature. Without discussion of the nomenclature and without referencing Linnaeus, Rama- moorthy (in Saldanha and Nicolson, 1976, p. 715) used the name C. barbata Sw. for this complex found in South India, whereas Cope (1982, p. 123) mentioned the illegitimacy of A. barbatus L. 1771 and used the name C. barbata Sw. for this species complex found in Pakistan. Our discussion follows. Some workers may not accept the illegitimate status assigned to Andropogon barbatus L. 1771. Since it was Linnaeus himself who used the name A. barbatus for both the Jamaican grass and the East Indian grass, such workers may believe that Linnaeus misidentified the East Indian grass as the Jamaican grass and that A. barbatus L. 1771 is a misapplied name. Consequently, based on ICBN Art. 33 Note 1 (Greuter, 1988), such workers may treat Chloris barbata Sw. as a new species (as suggested by one of the anonymous reviewers of this article). However, we refute this view, with the following discussion provided. First, in 1763, Linnaeus abandoned his binomial Andropogon barbatus L. 1759, and renamed it as A. polydactylon. We speculate that Linnaeus preferred the epithet po/ydactylon (= many fingers, referring to the 7-15, palmately arranged spikes) over the epithet barbatus (= bearded, perhaps referring to the pubescence of the lemma) for the Jamaican grass; nevertheless, the name A. poly- dactylon is superfluous. Second, in 1771, based on the belief that the epithet barbatus was still available for use in Andropogon L., Linnaeus proposed the name 4. barbatus for a new East Indian grass. The protologue of A. barbatus L. 1771 is wholly different from those of A. barbatus L. 1759, A. barbatus L. 1760, and A. polydactylon, and has no direct or indirect reference to any of these protologues. Linnaeus’ 1771 publication (titled Mantissa) mostly included additions to his previous publications; hence. Linnaeus did not list 4. polydactylon in his Mantissa. However: Willdenow (1806, p. 926-927), who revised Linnaeus’ Species Plantarum, recognized both Chloris barbata and C. polydactyla 1992] Kartesz and Gandhi— Chloris 137 as two distinct species. We assert that Linnaeus did not misiden- tify the East Indian grass, but correctly recognized it as a new species, distinct from A. polydactylon. Had Linnaeus, in 1763, not renamed A. barbatus L. 1759 as A. polydactylon, or had Lin- naeus, in the protologue of A. barbatus L. 1771, referenced any of the three earlier works, then A. barbatus L. 1771 could be classified as a misapplied name, but that is not the case here. Hence, A. barbatus L. 1771, which was based on a type different from that of A. barbatus L. 1759, must be classified as a later homonym, and thus illegitimate. Chloris barbata Sw. must be treated neither as a new species nor as a new combination, but rather as a nom. nov. (without parenthetical authorship), with its priority from 1795 (Art. 72 Note 1, Ex. 2). Regarding Anderson’s rejection of C. barbata Sw., we speculate that he was unaware of Art. 72 of the Code. As a legitimate name, C. barbata Sw. 1797 has priority over C. inflata (established in 1821). CHLORIS ELATA DESVAUX, AN EARLIER NAME FOR C. DANDEYANA ADAMS Adams (1971, p. 408) remarked that the type of Andropogon polydactylon is the type of A. barbatus L. 1759; hence, the former name is superfluous, and thus illegitimate (Art. 63.1). Instead of transferring A. barbatus L. 1759 to Chloris, Swartz (1788, p. 26) transferred A. polydactylon (= C. polydactyla). At that time, usage of the epithet barbata in Chloris was not pre-empted, since the name C. barbata Sw. 1797 (for the East Indian grass) appeared nine years later. Hence, in 1788, Swartz had the opportunity to use the name C. barbata (instead of C. polydactyla) for the Ja- maican grass. Moreover, according to the present Code (Arts. 55.1b, 63.2), Swartz ought to have adopted the epithet barbata, since he included the type of A. barbatus L. 1759 for his com- bination, but failed to do so. Had Swartz (in 1788) used the name C. barbata for the Jamaican grass, then (in 1797) he could have chosen a different name for the East Indian grass. We presume that Swartz decided to follow Linnaeus’ 1763 and 1771 treatments and ignored the 1759 and 1760 treatments. Nevertheless, the oo. C. polydactyla is also superfluous, and thus illegitimate (Art. 1) Nash (1898, p. 443-445) briefly discussed the nomenclature of this complex (including illegitimacy of Andropogon barbatus L. 138 Rhodora [Vol. 94 1771) and made the combination based on A. barbatus L. 1759: Chloris barbata (L.) Nash. Thus, Nash also failed to realize that his combination was illegitimate due to the existence of C. barbata Sw. Alternatively, Adams proposed the new name C. dandeyana for the Linnaean names of 1759, 1760, and 1763. Apart from his new name proposal, Adams provided no discussion on the tax- onomy of this complex. In addition to accepting Chloris dandeyana, Anderson cited three taxonomic synonyms [C. e/ata Desvaux, Opusc. Sci. Phys. Nat. 73. 1831; C. consanguinea Kunth (Rev. Gram. 1: 89. 1829, nom. invalid.) Enum. Pl. 1: 264. 1833; and C. arundinacea Nees ex Steudel, Syn. Pl. Glum. 1: 207. 1854], all of which are legitimate and have priority over C. dandeyana. Of these three synonyms, C. elata is accepted here to be the legitimate name in place of C. dandeyana. TAXONOMY Chloris barbata Sw., Fl. Ind. Occid. 1: 200. 1797. Andropogon barbatus L., Mant. Pl. 302. 177 1, non L. 1759. Type LocALity: India. No. 1211.21 (LINN; microfiche!). On the right side of the type specimen, the following are found: ““Konda-Pulli Rheed. Mal. XII. p. 95. t. 51.” (on the top of the sheet) and a sketch (on approximately middle right edge of the sheet) of unrecognizable meaning. In South Indian languages (Tamil and Malaya- lam), the word Kondai refers to a tuft of woman’s hair and the word pullu refers to grass. The remainder of the citation refers to Rheede, Hort. Malab. 12: 95, t. 51. 1693. However, none of these particulars are found in Linnaeus Mantissa. It is most likely that these data were entered by later workers. According to Nicolson et al. (1988, p. 307), Lamarck (1785, p. 376) was the first to associate Rheede’s element with A. barbatus L. 1771. C. inflata Link, Enum. Pl. 1: 105. 1821. “Type grown in the Berlin Botanic Garden from seed said to come from ‘California,’ probably from Mexico” (Hitchcock, 1936, p. 134). The type at B was presumably destroyed during World War II. Reputedly, a fragment of the type exists at us (fide Anderson, 1974, p. 53)- Chloris elata Desvaux, Opusc. Sci. Phys. Nat. 73. 1831. TYPE LOCALITY: Brazil. (P). Andropogon barbatus L., Syst. Nat. ed. 10: 2: 1305. 1759: Amoen. Acad. 5: 412. 1760, non L. 1771. TyPE Locauity: Jamaica. No. 1211.28 (LINN; microfiche!). A. polydactylon L., Sp. Pl. ed. 2. 2: 1483. 1763, nom. illegit. C. polydactyla Sw., Prodr. 26. 1788. 1992] Kartesz and Gandhi— Chloris 139 c eieoy (L.) Nash, Bull. Torrey Bot. Club 25: 443. 1898, non C. barbata Sw. cy a Adams, Phytologia 21: 408. 1971. ACKNOWLEDGMENTS We thank John McNeill (TRT), Dan H. Nicolson (US), Richard P. Wunderlin (USF), Bruce F. Hansen (USF), Rogers McVaugh (NCU), and Dennis E. Anderson (HSC) for a discussion on the nomenclature; we also thank Mary E. Barkworth (UTC), Paul A. Fryxell (TAES), Larry E. Brown (SBSC), Jimmy R. Massey (NCU), and two anonymous reviewers for helpful suggestions. LITERATURE CITED AMS, C. D. 1971. Miscellaneous additions i‘ revisions to the flowering plants of Jamaica III. Phytologia 21: 405-4 ERSON, D. E. 1974. Taxonomy of the genus Choris (Gramineae). Brigham Young Univ. Sci. Bull., Biol. Ser. 19(2): 1- Cope, T. A. 1982. Poaceae. In: E. Nasir and : “ Ali, Eds., Flora “4 Pakistan, No. 143. Department of Botany, University of Karachi, Pa kista Goutp, F.W. 1975. The Grasses of Texas. Texas Agric. Expt. Sta. Teas A&M University Press, College Station, TX. —— anpT. W. Box. 1965. Grasses of the Texas Coastal Bend. Texas A&M University Press, College Station, TX. Greuter, W. (Ch. Ed. Comm.). 1988. International Code of Botanical No- menclature. Adopted by the Fourteenth International Botanical Congress, Berlin, Jul-Aug. 1987. Regnum Veg. 118. Hatcu, S. L., K. N. GANDHI AND L. E. Brown. 1990. Checklist of the Vascular Plants of Texas. Texas Agric. Exp. Sta. MP-1655. The Texas A&M University System, College Station, TX. Hircucock, A. S. 1936. Manual of the Grasses of West Indies. U.S.D.A.., M.P. No. 243. Government Printing Press, Washington, DC Howarp, R. A. 1979. Flora of Lesser Antilles, Vol. 3: “Monocotyledoneae. Arnold Arboretum, Jamaica Plain, MA. TESz, J.T. AnD R. Kartesz. 1980. A Synonymized Checklist of the Vascular Flora of the United States, Canada, and Greenland. The University of North Carolina Press, Chapel Hill, NC. ck, J.B. A. P.M. D. 1785. Encyclopedie Methodique. Botanique, Vol. 1(2). Paris. Linnazus, C. 1759. Systema Naturae, 10th ed., Vol. 2. Stockholm. ———. 1760. Amoenitates Academicae, Vol. 5. Stockholm. ———. 1763. Species Plantarum, 2nd ed., Vol. 2. Stockholm. -———. 1771. Mantissa Plantarum Altera, Part 2. Stockholm Nasu, G. V, 1898. A revision of the genus Chloris and E ustachys. Bull. Torrey Bot. Club 25: 432-450. 140 Rhodora [Vol. 94 Nicotson, D. H., C. R. SURESH AND K. S. MANILAL. 1988. An Interpretation of Van Rheede’s Hortus Malabaricus. Regnum Veg. 119. SALDANHA, C. J. AND D. H. Nicotson. 1976. Flora of Hassan District, Kar- nataka, India. Amerind Publishers, New Delhi Sor, CONSERVATION SERVICE. 1982. National List of Scientific Plant Names, Vols. 1 and 2. USDA, SCS-TP 159. Government Printing Office, Washington, Swartz, O. 1788. Nova Genera & Species Plantarum Seu Prodromus. Stock- holm. ———. 1797. Flora Indiae Occidentalis, Vol. 1. WILLDENow, C. L. 1806. Caroli a Linne Species Plantarum, 4th ed., Vol. 4(2). rlin. THE NORTH CAROLINA BOTANICAL GARDEN DEPARTMENT OF BIOLOGY, COKER HALL UNIVERSITY OF NORTH CAROLINA CHAPEL HILL, NC 27599-3280, USA RHODORA, Vol. 94, No. 878, pp. 141-155, 1992 GEOGRAPHICAL DISTRIBUTION AND ECOLOGY OF LONG’S BULRUSH, SCIRPUS LONGII (CYPERACEAE) IN CANADA NICHOLAS M. HILL AND Marts E. JOHANSSON ABSTRACT Scirpus longii is a rare coastal plain plant traditionally regarded as a New Jersey Pine Barrens species. It had been known from only one site in Canada, in Nova reduced either by ice scour or flooding. Flooding, and consequent anaerobiosis, appears to be the main mechanism which reduces shrub growth in most S. longti habitats, but ice scour may be a more important factor on the shores of high watershed area lakes. Scirpus longii forms circular clones at all non-ice-scoured habitats. Above-ground production of shoots is restricted to the periphery of circular clones. Our data indicate that this circular phalanx may help the species to compete with shrubs. Populations in Nova Scotia appear to be secure; however, their persistence will be influenced by any factor which alters their =a dynamics with shrubs. Key Words: Scirpus longii, Long’s bulrush, rare plants, shrub competition, flood- ing, clonal plants, circular growth, coastal plain flora, Nova Scotia INTRODUCTION Scirpus longii Fern., or Long’s bulrush, was given specific status by Fernald (1911) after Bayard Long called his attention to an atypical Scirpus bling Scirpus atrocinctus Fern. At that time, S. longii was known from “marsh” in the Pinus rigida-dominated barrens of New Jersey, and from meadow bordering the Charles River in Massachusetts. The species is listed as Imperiled by the Nature Conservancy, a ranking which is supported in a recent report by Rawinski (1990). This report stated that the species 1s most abundant in the Pine Barrens of New Jersey and is poorly represented in New England (Massachusetts, Rhode Island and ew Hampshire), but that new findings from Maine show that wetlands bordering the Saco River support large S. longii pop- ations. The Canadian distribution rested upon Weatherby’s report of a single population of Scirpus longii in Nova Scotia (1942). This report was not further investigated until 1988, when Hill and Keddy (1992, in press) discovered two new sites during an ex- tensive survey of rare coastal plain plants. At several locations in 14] 142 Rhodora [Vol. 94 Nova Scotia, the distribution of S. /ongii and other globally-rare species (viz.: Lophiola aurea Ker-Gawl. and Lachnanthes caro- liniana (Lam.) Dandy) overlap; our findings for a portion of the species’ range of greatest value for conservation are included in Wisheu et al. (1992, in press). In this paper, we report the full distribution of S. /ongii in Nova Scotia, a species which was a distribution pattern typical of many Atlantic coastal plain species (Roland and Smith, 1969). We describe and attempt to classify the types of wetland habitat which support the species, and infer what environmental factors may be regulating the distribution. METHODS Identification Initially, Scirpus longii could not be found among 47 Nova Scotian lakes surveyed in 1988 for rare Atlantic coastal plain plant species (Hill and Keddy, 1992, in press), despite the fact that it was documented to occur at one of these lakes. When unidentified vegetative shoots of a sedge were compared with those belonging to S. longii, it was realized that S. longii occurred in several locations but did not flower. Schuyler (1963a) noted this phe- nomenon in S. /ongii in New Jersey and observed that flowering culm formation was rare except after fires. It is essential to be able to identify the plant in its vegetative state. Rhizomes of S. longii are longer lived and consequently stouter (usually 1.5-3 cm diam.) than those of S. cyperinus (L.) Kunth or S. atrocinctus (usually 1-1.5 cm diam.). In Nova Scotia, S. Jongii grows under conditions of low disturbance as large circular clones commonly -75 to 5 m in diameter, although we have observed clones up t0 50 m in diameter. In contrast, S. cyperinus rhizomes decay rapidly and small tussocks are formed (Schuyler, 1963b), usually less than 50 cm in diameter. At rich, mucky sites, the vegetative shoots of Carex rostrata Stokes ex With. may be confused with those of 5. longii; however, C. rostrata does not form superficial rhizomes or large circular clones. a Once it became possible to identify vegetative Scirpus long, the bogs and fens of the Medway and Tusket Rivers were inves- tigated in 1989 and 1990. Two other sites in southwestern Nova Scotia were discovered in 1990 while doing inventories of bogs- 1992] Hill and Johansson— Scirpus longii 143 Distribution of Scirpus longii in Bogs Scirpus longii was discovered in five types of wetlands which can be categorized as stillwater meadow, fen, “bay” and “barrier” bogs associated with lakes or rivers, and peaty lakeshores (de- scribed below). The within-bog distribution was studied at a lake “bay” bog at Shingle Lake. The relationship between cover of S. longii, water-logging and shrub growth was examined along five 50-m-long transects spaced at 10-m intervals. Transects ran from the lake edge of the bog inland to the rocky upper margin of the bog. Vegetation cover estimates were recorded according to the Braun-Blanquet scale and were transformed according to Maarel (1979). At each point along the transect, depth of standing water was recorded after sphagnum and litter were compressed. Measurements of depth to the watertable were made at three hollows in the fen at Eighteen Mile Brook on August 20, 1990. Twenty-cm-diameter holes were excavated to depths below the watertable and depth from the surface peat to the watertable was determined after five minutes. Vegetation Associated with Circular Clones Clones of Scirpus longii are usually circular, above-ground shoots are restricted to the perimeter of the clone which gives them a “doughnut’’-shaped appearance. Shrub growth was markedly low- er in the interior region of clones than it was in the zone imme- diately outside the perimeter of leafy shoots of S. /ongii. We documented this phenomenon by measuring shrub height and cover in the interior region of clones of S. /ongii (Figure 2a, zone A) and ina ring-shaped area outside the perimeter of the leafy S. longii phalanx (Figure 2a, zone B). Dimensions of the outer ring area were calculated in the field so as to approximate the circular area surveyed inside the phalanx by considering the outer ning area to be a rectangle whose width was calculated by dividing the area of the interior circle by the outside circumference of the clone. Average error between these two sample areas (1.€., the interior circle and outer ring) for 21 circular colonies was 4.8%. 144 Rhodora [Vol. 94 66° 64° 62° 60 4g 6 a 100 km, —L6 16 = al oy L5- 7e Scirpus longii Mey 4h ew ads in Nova Scotia bh — 66 4 «J 64° 62° 60° 0 l | ee See 5 “Ble 2 athe ae Ay: Stiat Lake stip ‘ (an S ha : mn (EA mn PN tt : OK: eth 7% \ © Moosehorn a Real ~) Eighteen Mile \. mle we Brook —* 3 Aa te ue Figure 1. a. Distribution map of Scirpus longii in Nova Scotia. b. Map of above inset area showing sites in the Medway River watershed where S. longii 1s a dominant species over 10-15 ha. (&), over 5 ha. (A), and in less than 1 ha. (a). Ly bi Vi wy bY 1 , AA (( i Ho WW s BK y, UCU Aib y 4 ‘( d 4 Fania et scheme for vegetation associated with circular clones hatched a . = interior region, B = outer ring-shaped area surrounding the used to obtain ™ ich represents the phalanx of S. longii shoots. Measurements from the insid equivalent areas for A and B are r’, clone radius; and r, distance captain e margin of the vegetated phalanx to center of clone. b. Simplified f halanx ofa class 1 clone (1 m diam.) of S. /ongit (positions of shoots traced from photograph of clone). 146 Rhodora [Vol. 94 RESULTS AND DISCUSSION Wetland Habitats of Scirpus longii At all of the wetland habitats supporting Scirpus longii in Nova Scotia, S. /ongii grows on peat, but the hydrology of the sites varies greatly. Stillwater meadows border slowly-moving, tea- colored rivers at the outflow of Shingle Lake; they become in- undated as water levels rise in winter and may remain flooded until late May. The sizes of large, discrete circular clones of S. longii in these stillwater meadows range from .75 m diam. (small- er, intact circular clones have not been found) to 20 m diam. The stillwater meadow is dominated by a matrix of graminoids (largely Carex stricta Lam., C. oligosperma Michx., C. bullata Schkuhr and Spartina pectinata Link), and contains large discrete clones of S. longii and the shrubs Myrica gale L. and Salix pedicellaris Pursh The “fens” bordering Eighteen Mile Brook have a more diverse shrub community scattered throughout hummock and hollow microtopography, with small trees (< 2.5 m) of Acer rubrum and Larix laricina and discrete circular clones of Scirpus longii which range from .75 m to 10 m in diameter. These fens are flooded in winter, but in summer the watertable drops 15-20 cm below the peat surface in hollows. Lakeshore and riverside “bogs” can be classed either as “bay bogs” or “barrier bogs.” Bay bogs form when peat accumulates in sheltered bays of lakes, eventually filling in the entire bay. Tall Myrica gale dominates the water’s edge of these bay bogs, but shrub growth is depressed in the central waterlogged region behind the tall shrub zone. Shrub height and cover recovers toward the terrestrial rocky margin of the bay bog, which is the original lake shoreline. Scirpus longii is most commonly found in the cent region of bay bogs, where the amount of standing water is greatest. “Barrier bogs,” on the other hand, are small wetland areas separated from lakes (or rivers) by a rocky barrier; they become flooded in winter when the water level of the neighboring watet- body rises. Because of the impermeable barrier of rock, these bogs remain flooded after the water level of the lake or river has fallen in spring. Muskrats have overwintered at both barrier bogs suP- porting Scirpus longii, and consume its culm base/rhizome region. Unlike the foregoing boggy habitats where Scirpus longii aP- 1992] Hill and Johansson— Scirpus longii 147 ars to be restricted to waterlogged depressions away from the actual lake shoreline, at high watershed area lakes, populations of S. longii do occur directly on the peaty shores of lakes near the water’s edge. Populations at the south of Ponhook Lake consist mainly of scattered clumps of less than 20 individual ramets which appear to be the remnants of larger clones fragmented by ice scouring. Such clumps occur most often on the downstream side of promontories at the outflow to Ponhook Lake. In contrast, at Lac D’Ecole and at Little Ponhook Lake (a small embayment of Ponhook Lake), discrete, intact circular clones of S. longii occur at the water’s edge in peaty muck. Interestingly, judging from the lake size and the amount of mucky organic matter on respective shorelines, it appears that these populations experience less ice scour than is normal on the shorelines of high watershed area lakes. We do not know yet how old the clones are at these various sites. Large clones of 5 m or more in diameter occur only in the extensive peatland sites, the stillwater meadows and the fen beside Eighteen Mile Brook. The smallest clones occur directly on the peaty shores of high watershed area lakes. This pattern suggests that extensive clone development may occur only where distur- bance is minimal. New Occurrences of Scirpus longii We confirmed that Scirpus longii still occurs at Ponhook and Moosehorn Lakes (Figure 1, sites 6 and 9-12). Weatherby (1942) had documented its occurrence at Ponhook Lake and had found “a battered individual, probably of this species” at Moosehorn Lake. We found only four small clones (each < 1 min diameter) of S. longii at the boggy outflow of Moosehorn Lake. At Ponhook Lake, several small populations (each with less than 1000 indi- vidual ramets) occur on the peaty shore of the lake. Large pop- ulations (> 5000 individual ramets) occur at site 9 in marsh beside the lake; the largest of these populations occurs over one hectare ofa bay bog where approximately 200 clones contain an estimated total of 40,000 individual ramets (systematic sampling over grid- ded area). Once we were able to identify non-flowering clones of Scirpus longii (primarily from their circular growth form and their stout 148 Rhodora [Vol. 94 superficial rhizomes), we discovered that the largest populations of S. /ongii in Nova Scotia had gone unnoticed and occur over about 25 ha of stillwater meadow near Shingle Lake (Figure 1, sites 1 and 2), and in 15 ha of fen beside Eighteen Mile Brook (Figure 1, sites 7 and 8), where they rarely produce flowering culms. Weatherby had not visited Shingle Lake in 1941, but did visit the fen at Eighteen Mile Brook where he recorded Salix pedicellaris var. hypoglauca. The fen must have had quite a dif- ferent aspect in 1941. Judging from our preliminary estimates of the annual radial expansion rate of S. /ongii (4 cm yr.~'), and assuming that established clones have a low mortality rate, only the few clones that are now two meters in diameter would have been present in 1941. All of the lakeshore bogs we found supporting Scirpus longii in Nova Scotia are quite small ranging from .1 to 1 ha. In addition to those reported for Ponhook Lake (Wisheu et al., 1992, in press), we have found S. /ongii in two lakeshore bay bogs and in one barrier bog at Shingle Lake (site 3), and in a barrier bog at Riv- ersdale on the Medway River (discovered by C. Keddy and I. Wisheu). The above sites for Scirpus longii belong to the Medway River watershed area. Lakes of this river system and of the Tusket River have the highest species richness of rare coastal plain plants in Nova Scotia. Many of the rare species are found in both river systems, but others are found in one river and not in the other (e.g., Sabatia kennedyana Fern. and Coreopsis rosea Nutt. in Tusket River, and Lophiola aurea and Lachnanthes tinctoria 10 Medway River). It appeared that S. longii might be restricted to the Medway watershed until we found two small populations on two lakes of the Tusket River, 80 km away from the closest site in Medway watershed. The population at Wilson’s Lake (43°56'N, 66°53’W) is in a bay bog behind the water’s edge line of shrub, while at Lac D’Ecole (44°56'N, 65°49'W) circular clones of S. longii grow at the water’s edge on organic muck. We have since discovered two isolated populations in fenland between the Medway and Tusket watersheds, at Dunraven Bog (44°05'N, 65°08’W) and at the southern end of Quinns Meadow (43°40’N, 65°29’W). The small population at Dunraven Bog 4P- pears stable, judging from the robust circular clones (ca. | ™ diam.) of Scirpus longii, but that at Quinns Meadow is composed of small groups of individual shoots (total 100) in a wet meadow 1992] Hill and Johansson— ‘Scirpus longti 149 of Carex bullata, C. stricta and C. oligosperma which has been disturbed by all-terrain vehicles. There are undoubtedly many other locations where Scirpus longii grows in Nova Scotia which have yet to be discovered. It is apparent that the province has not been thoroughly botanized; at two of the new sites for S. /ongii (Shingle Lake site 3 and Dunraven Bog), we found the endangered and showy Lophiola aurea, noticeable immediately to any field botanist, but hitherto unreported. Relationship with Shrub Growth At bay bogs on Shingle Lake and at Wilson’s Lake, shrub height was greatest at the water’s edge and at the terrestrial margin of the bog near the edge of woods (see Figure 3, Shingle Lake). Conversely, cover of Scirpus longii was greatest at the waterlogged central region of the bog, although this relationship was not sig- nificant due to the limited number of occurrences within the transect. Cover of S. Jongii in the Shingle Lake bog was signifi- cantly (t-test, P = .05) greater in areas without shrubs (.63 + 1.02, & + SD, n = 19) than in those with shrubs (.11 + .38, % + SD, n= 46). The inverse relationship between shrub height and cover and of Scirpus longii found at this bay bog may be due to an inability of Myrica gale, the dominant shrub, to endure prolonged water- logging. The same within-bog distribution of S. /ongii was ob- served at every bay bog whose lake margin was dominated by shrubs: standing water collected in their interior region where shrubs were sparse and S. /ongii most abundant. Similarly, at barrier bogs, S. Jongii was most abundant in the central, depressed region where the watertable is highest (standing water in center until mid-May) and shrub growth is lowest. Since shrubs grow vigorously at lake margins of bay bogs where there should be free exchange of oxygenated water between the lake and shrub roots, it may not be flooding per se which restricts shrub growth, but rather stresses associated with anaerobiosis (Jackson and Drew, 1984; Kozlowski, 1984; Dionigi et al., 1985). Keddy and Wisheu (1989) have suggested that as a group, rare coastal plain plants are poor competitors and hence need to oc- cupy habitats in which the importance of competition is reduced 150 Rhodora [Vol. 94 80 - Terrestrial margin Lake Gi edge 5 | A = 40- SD ) i S = 20- & ep) (fizlj~ 0 rs ‘ AA 0 —_— ae a ae Standing Water __ cm Figure 3. Relationship between standing water (upon compression of sphag- num) and shrub height for nd Positions along 5 transects of Shingle Lake bay bog. Note that the data point for a different symbol (%) and plotted off-scale on the X-2 -axis because lake depth at the edge of the bog is much greater than the maximum a water depth in the bog. Vertical bars represent standard errors of mean values by stress or disturbance (sensu Grime, 1979). Scirpus longii occurs in bogs and on peaty lakeshores where different mechanisms may permit it to avoid competition from shrubs; such shrubs appear to be eliminated from the shores of high watershed area lakes by 1992] Hill and Johansson— Scirpus longii 151 disturbance or stresses associated with flooding. Hill and Keddy (1992, in press) showed that richness of rare coastal plain plants was largely accounted for by watershed area, a measure which was strongly correlated with the over-winter flooding of lakes. Scirpus longii occurs in wetlands having a wide range of watershed areas, but it only occurs directly (i.e., not in a bog above the waterline) on the shores of high watershed area lakes (Ponhook Lake = 109,280 ha; Lac D’Ecole = 81,641 ha). The exact mech- anism of how shrub growth is checked by flooding on the shores of high watershed lakes is not clear. One hypothesis is that there is greater ice scouring during the winter in areas where waterlevel fluctuations and river currents are greatest. In this case, ice scour- ing would not only remove nutrients from the site as litter is removed, but ice movement would sever above-ground growth. Although severe ice scouring could eliminate both shrubs and S. longii clones from lakeshores, zones of moderate scouring (e.g., in coves, behind promontories) can be observed on lakeshores where the above-ground woody shoots of shrubs have been largely removed, yet stout superficial rhizomes of other species remain. At such sites (e.g., 10 and 12 at Ponhook Lake), thick superficial thizomes of Osmunda regalis are found with those of S. longii. In contrast to populations of Scirpus longii at the water’s edge of high watershed lakes, it appears that in bogs it is local rather than regional hydrology which determines distribution of the spe- cies. In bay and barrier bogs, water movement between the de- Pressed central region of the bog and the larger water body (lake Or river) is limited; here, shrubs are protected from ice scour but their growth appears to be depressed by anaerobiosis. Vegetation Association with Circular Clones _ In areas of low disturbance, Scirpus longil clones are circular In outline, and above-ground shoots are confined to the periphery of the clone (Figure 2b). This growth pattern has been observed in Spartina townsendii H. & J. Groves (Caldwell, 1957) and in Larrea tridentata (Sesse & Moc ex DC.) Coville (Vasek, 1980). Bell and Tomlinson (1980) suggested that in the absence of specific environmental constraints, clone shape is innate and is deter- mined by the precise rhizome branching pattern of the species in question. More specifically, Caldwell (1957) hypothesized that 152 Rhodora [Vol. 94 Table 1. Vegetation structure inside and outside Scirpus longii clones of two size Classes. Small clones range from 1-2.5 m; large clones from 5-6. Standard error appears in parenthesis below its mean value Small Clones (7 = 16) Large Clones (n = 6) Outer Outer Vegetation Interior Ring Interior Ring Shrub Height 12.20 17.40 20.00 29.20 (cm) (1.22) (0.86) (0.63) (1.42) Shrub Cover 1.48 3.40 3.66 7.83 Units* (0.36) (0.36) (0.50) (0.32) Species 6.31 8.38 16.7 212 Richness** (0.81) (0.87) (1.75) (2.70) S. longii 0.44 0.38 2.67 0.17 Cover Units (0.16) (0.15) (0.34) (0.17) * Braun-blanquet scale transformed according to Maarel (1979). ** All vascular plant species; inner and outer areas are equivalent for each class size. circular growth patterns occur only in areas where the importance of competition is low. The circular form of S. /ongii clones ob- served in areas that we believe are flood-stressed (e.g., stillwater meadow, barrier and bay bogs) is consistent with Caldwell’s hy- pothesis. Competitive interactions between S. /ongii and shrubs appear to be small at the most water-logged locations where shrub cover is low; however, we nae observed what appears to be a physical interference bet ircular clones of S. /ongii and shrubs in fens, the driest of the S. longii habitats in summer. Scirpus longii clones of ca. 1 m or more in diameter are circular and present an outer phalanx of attached litter and densely packed leafy shoots. Rhizome sections to the inside of the outer phalanx produce very few above-ground shoots (Figure 2b); however, this interior region of clones frequently has little shrub growth in comparison with that immediately to the outside of the leafy perimeters of S. /ongii clones. An indication of the possible in- terference of Scirpus longii of shrubs at Eighteen Mile fen was obtained by comparing the cover and height of shrubs in the interior region of circular clones (Figure 2a, zone A) with those measures in an equivalent ring-shaped area just outside the clone (Figure 2a, zone B). The height and cover of shrubs was markedly lower in the interior than in the outer ring (Table 1). Interestingly, the above-ground cover of S. Jongii in the interior region was less 1992] Hill and Johansson—Scirpus longii 153 than the cover of shrubs in this region (Table 1). Shrub cover and height were also lower in the interior region of large clones (5-6 m diam.); however, in this case, vegetative cover of S. longii was substantial in the interior region and may account for the reduc- tion in shrub growth. With both small and large clones, total species richness was higher in the interior than in the exterior region of the clone (Table 1). We do not know how small clones of Scirpus longii exclude shrubs from their barren interior regions. Preliminary investi- gations reveal that the rhizomes in the interior of the clones are mostly dead but that they persist and form a fretwork which is overlain by decaying litter and moss. Shrubs may be excluded in part by the outer leafy phalanx of S. /ongii shoots because shrub shoots invading the phalanx would have to establish their roots in layers of S. Jongii litter on top of the matrix of dead rhizomes. Shrub growth may also be reduced if soil nutrients in the clone interior have been depleted by S. /ongii. Most of the wetland habitats occupied by Scirpus longii in Nova Scotia are not immediately threatened, but it is important that the high population density sites in fen and river meadow receive protection. Distributional evidence suggests that this plant con- forms with the generalization that coastal plain plants require habitats where competition is reduced (Keddy and Wisheu, 1989). Alteration of bogland hydrology changes the vegetation (Larsen, 1982; Thibodeau and Nickerson, 1985), such changes may favor aggressive species to the detriment of coastal plain flora, as was Observed in the New Jersey Pine Barrens (Ehrenfeld, 1983). The Status of this plant in Nova Scotia could easily be made more secure by safeguarding the small watersheds of the stillwater ane and fen sites which support the largest S. longii popu- ations. ACKNOWLEDGMENTS We appreciate the taxonomic advice of Paul Catling and loan of Scirpus longii specimens from PH. Field work was greatly assisted by the hospitality of the Sheppards in Caledonia and of landowners Harry Freeman and Louis Waterman. We are grateful for funding from the Special Places Programme of the Nova Scotia Museum and from the World Wildlife Fund. 154 Rhodora [Vol. 94 LITERATURE CITED Bet, A. D. AND P. B. ToMuINSoN. 1980. Adaptive architecture in rhizomatous plants. J. Linn. Soc. Bot. 80: 125-160. ALDWELL, P. 1957. The spatial development of Spartina colonies growing without competition Ann. Bot. (London) 21: 203-216. Dionici, C. P., I. A. MENDELSSOHN AND V. I. SULLIVAN. 1985. Effects of soil waterlogging on the energy status and distribution of Salix nigra and Salix exigua (Salicaceae) in the Atchafalaya River basin of Louisiana. Amer. J. Bot. 72: 109-119. EHRENFELD, J. G. 1983. The effects of changes in land-use on swamps of the New Jersey Pine Barrens. Biol. Conserv. 25: 353-375 FERNALD, M. L. 1911. A new species of Scirpus from Massachusetts and New Jersey. Rhodora 13: Grime, J. P. 1979. Plant Strategies and Vegetation Processes. Wiley and Sons, Chichester Hii, N. M. AND - A. Keppy. 1992. Predicting rarities from habitat variables: the Atlantic Coastal Plain flora on lakeshores in Nova Scotia. Ecology (in ress). p Jackson, M. B. AND M. C. Drew. 1984. Effects of flooding on growth and metabolism of herbaceous plant. Jn: T. T. Kozlowski, Ed., Flooding and Plant Growth. Acad. Press, Harcourt Brace Jovanovich, New York. ar Fi A. AND I. C. WisHEu. 1989. Ecology, biogeography, and ¢ Lagasse a Scotian wetlands. ‘Rhodora 91: 72-94, Koztowski, T. T. 1984. Responses of woody plants to flooding. In: T. T. Koz- lowski, Ed., Flooding and Plant Growth. Acad. Press, Harcourt Brace Jovano- vich, New York. Larsen, J. A. 1982. Ecology of the Northern Lowland Bogs and Conifer Forests. Acad. Press, Harcourt Bra ce Jovanovich, New York. MAAREL, E. VAN DER. 1979. Transformation of cover-abundance values in phy- WINSKI, T. J. 1990. Final status survey report: the distribution and — of Long’s Bulrush (Scirpus longii). U.S. Fish and Wildlife Service (unpub report). ROLanpD, A. E. AND E. C. SmitH. 1969. The Flora of Nova Scotia. Nova Scotia Museum, Halifax. 4 ScHuyLer, A. E. 1963a. Sporadic culm formation in Scirpus longii. Bartonia 32: |- : 1963b. A biosystematic study of the Scirpus cyperinus complex. Proc. Acad. Nat. Sci. Philadelphia 115: 283-311 THIBODEAU, F. R. AND N. H. NICKERSON. 1985. Changes in a wetland plant association induced by impoundment and draining. Biol. Conserv. 33: 269- 279. VASEK, F. C. 1980. Creosote bush: long-lived clones in the Mohave Desert. Amer. J. Bot. 67: 246-255. WEATHERBY, C. A. 1942. Two weeks in Nova Scotia. Rhodora 44: 229-236. Wisueu, I. C., C. J. Keppy, P. A. Keppy anp N. M. Hut. 1992. Disjunct Atlantic coastal plain species in Nova Scotia: distribution, habitat and con- servation priorities. Biol. Conserv. (in press). 1992] Hill and Johansson— Scirpus longii DEPARTMENT OF BIOLOGY MOUNT ST. VINCENT UNIVERSITY, HALIFAX, NOVA SCOTIA CANADA, B3M 2J6 M.E. J. DEPARTMENT OF ECOLOGICAL BOTANY UMEA UNIVERSITY S-901 87 UMEA, SWEDEN 155 RHODORA, Vol. 94, No. 878, pp. 156-166, 1992 THE FLORA OF LIMESINK DEPRESSIONS IN CAROLINA BEACH STATE PARK (NORTH CAROLINA) DAvip J. SIEREN AND KAREN R. WARR! ABSTRACT Seven selected limesink depressions in Carolina Beach State Park (New Hanover County, North Carolina) were floristically surveyed in 1990. One state record, 11 county records, and 28 species of special interest or concern are among the 108 vascular plant species collected or observed. Key Words: Vascular plants, limesink depressions, distribution records, North Carolina INTRODUCTION As part of a general floristic survey of the vascular plants of Carolina Beach State Park (New Hanover County, North Caro- lina), seven selected limesink depressions were studied from Jan- uary through December 1990. Such depressions are reportedly formed by the slumping of surface soils following fracture and dissolution of underlying limestone (McDonald et al., 1981; Tag- gart and Dickerson, 1980); those chosen exhibited a variety of types, from very dry to very wet, and with and without woods. For purposes of identification and description, the depressions were assigned the following names: (1) Sawgrass Pond (a small, elliptical depression ca. 29 m x 53 m at its widest points with a population of sawgrass in its center), (2) Loblolly Pond (a shallow, dry depression ca. 40 m x 56 m with numerous small loblolly pines scattered throughout), (3) Lily Pond (a large pond ca. 66 m x 150 m, normally aquatic with standing water and water lilies), (4) Cypress Pond (a partially wooded depression with numerous pond cypresses. This depression consists of east and west sections which in most years are separated by a dry ridge and which often have standing water in their centers. In wet years such as 1991, they can be one continuous pond; together they are ca. 61 m wide x 218 m long), (5) Dry Pond (ca. 25 m x 84 m, drier and more shallow than the others, usually without standing water), (6) Gum Pond (a nearly round depression ca. 48 m in diameter, partially ‘ Present address: 5385-J New Centre Dr., Wilmington, NC 28403 156 1992] Sieren and Warr—Limesink Flora 157 wooded with numerous pond cypresses and black gums), and (7) Grass Pond (ca. 41 m x 109 m, dominated by grasses and other herbaceous vegetation). All of the limesinks surveyed are sur- rounded by longleaf pine-scrub oak communities, and they grade from their moist or aquatic centers or low areas upward to sur- rounding xeric sand ridges, often having distinct vegetation zones with increasing elevation. The limesinks were surveyed on 26 weekly trips, primarily during the active growing season, March through October; known species were recorded and unknowns were collected and identi- fied. The 249 voucher specimens have been deposited in wNc; nomenclature follows that of Kartesz and Kartesz (1980). In the species list, synonyms from Radford et al. (1968) are given in brackets following those taxa with name changes. SPECIES LIST Pteridophyta BLECHNACEAE Woodwardia virginica (L.) Smith LYCOPODIACEAE Lycopodium alopecuroides L. Lycopodium appressum (Chapman) Lloyd & Underwood Lycopodium carolinianum L. OSMUNDACEAE Osmunda regalis L. var. spectabilis (Willd.) Gray Gymnospermae PINACEAE Pinus taeda L. TAXODIACEAE Taxodium ascendens Brongn. 158 Rhodora [Vol. 94 Angiospermae ACERACEAE Acer rubrum L. ANACARDIACEAE Toxicodendron radicans (L.) Kuntze. [Rhus radicans L.] APIACEAE Centella asiatica (L.) Urban Oxypolis filiformis (Walt.) Britt. AQUIFOLIACEAE Tlex cassine L. Ilex glabra (L.) Gray CEAE Coreopsis falcata Boynt. Erechtites hieracifolia (L.) Raf. ex DC. Eupatorium capillifolium (Lam.) Small Eupatorium leptophyllum DC. [Eupatorium capillifolium vat. leptophyllum (DC.) Ahles] Eupatorium recurvans Small Euthamia tenuifolia (Pursh) Greene [Solidago microcephala (Greene) Bush] Pluchea rosea Godfrey BROMELIACEAE Tillandsia usneoides (L.) L. BURMANNIACEAE Burmannia biflora L. Burmannia capitata (Walt.) Mart. CAMPANULACEAE Lobelia nutallii Roemer & Schultes 1992] Sieren and Warr—Limesink Flora 159 CLUSIACEAE [HYPERICACEAE] Hypericum cistifolium Lam. Hypericum reductum P. Adams CYPERACEAE Cladium jamaicense Crantz Cyperus polystachyos Rottb. var texensis (Torrey) Fern. Eleocharis equisetoides (Ell.) Torr. Eleocharis melanocarpa Torr. Psilocarya scirpoides Torr. Rhynchospora chalarocephala Fern. & Gale Rhynchospora chapmanii M. A. Curtis Rhynchospora corniculata (Lam.) Gray Rhynchospora filifolia Gray Rhynchospora inundata (Oakes) Fern. Rhynchospora pleiantha (Kiikenthal) Gale Rhynchospora plumosa Ell. Rhynchospora tracyi Britt. Rhynchospora wrightiana Boeck. Scleria georgiana Core Scleria reticularis Michx. var. pubescens Britt. LLACEAE Cyrilla racemiflora L. DROSERACEAE Drosera capillaris Poiret Drosera intermedia Hayne EBENACEAE Diospyros virginiana L. ERICACEAE Lyonia lucida (Lam.) K. Koch Lyonia mariana (L.) D. Don Vaccinium corymbosum L. [Vaccinium atrococcum (Gray) Por- ter] 160 Rhodora [Vol. 94 ERIOCAULACEAE Eriocaulon compressum Lam. Lachnocaulon beyrichianum Sporleder. ex Koern. GENTIANACEAE Bartonia paniculata (Michx.) Muhl. Bartonia verna (Michx.) Muhl. Sabatia difformis (L.) Druce HALORAGIDACEAE Proserpinaca pectinata Lam. ELIDACEAE Liquidambar styraciflua L. JUNCACEAE Juncus abortvus Chapman Juncus scirpoides Lam. LAURA Persea borbonia (L.) Sprengel. LENTIB CEAE Pinguicula caerulea Walt. Utricularia juncea Vahl. Utricularia purpurea Walt. Utricularia subulata L. LINACEAE Linum floridanum (Planch.) Trel. var floridanum {Linum vir- ginianum L. var. floridanum Planch.] MAGNOLIACEAE Magnolia virginiana L. MELASTOMATACEAE Rhexia cubensis Griseb. Rhexia mariana L. var. mariana 1992] Sieren and Warr—Limesink Flora 161 MYRICACEAE Mpyrica cerifera L. NYMPHAEACEAE Nymphaea odorata Ait. NYSSACEAE Nyssa sylvatica Marsh var. biflora (Walter) Sargent ONAGRACEAE Ludwigia linifolia Poir. Ludwigia suffruticosa Walt. ORCHIDACEAE Pogonia ophioglossoides (L.) Juss. POACEAE Andropogon capillipes Nash [A. virginicus L., in part] Andropogon perangustatus Nash Andropogon virginicus L. var. virginicus Aristida affinis (Schultes) Kunth. Aristida purpurascens Poir. Aristida virgata Trin. Coelorachis rugosa (Nutt.) Nash [Manisuris rugosa (Nutt.) Kuntze 5 Dichanthelium acuminatum (Sw.) Gould & Clark var. longi- ligulatum (Nash) Gould & Clark [Panicum longiligulatum Nash] Dichanthelium dichotomum (L.) Gould var. ensifolium (Baldw. ex Ell.) Gould & Clark [Panicum ensifolium Baldwin ex EIl.] Eragrostis refracta (Muhl.) Scribn. Erianthus giganteus (Walt.) Muhl. Panicum rigidulum Bosc ex Nees [Panicum agrostoides Spreng. ] Panicum tenerum Bey. ex Trin. Panicum verrucosum Mubhl. Panicum virgatum L. Paspalum praecox Walt. Sacciolepis striata (L.) Nash b Setaria magna Griseb. 162 Rhodora [Vol. 94 | POLYGALACEAE | Polygala cruciata Nutt. Polygala cymosa Walt. Polygala lutea L. POLYGONACEAE Polygonum opelousanum Riddell ex Small [Polygonum hydro- piperoides Michx. var. opelousanum (Riddell ex Small) Stone] Polygonum persicaria L. RUBIACEAE Diodia virginiana L. SALICACEAE Salix nigra Marsh SAXIFRAGACEAE | Itea virginica L. SCROPHUL ARIACEAE Agalinis fasciculata (Ell.) Raf. Agalinis linifolia (Nutt.) Britt. Agalinis virgata Raf. SMILACACEAE Smilax laurifolia L. VIOLACEAE Viola lanceolata L. XYRIDACEAE Xyris ambigua Bey. ex Kunth. Xyris caroliniana Walt. Xyris jupicai L. Xyris smalliana Nash 1992] Sieren and Warr—Limesink Flora 163 DISCUSSION Following Schafale and Weakley (1990), limesinks in the Park can be classified as either Depression Pocosins, Small Depression Ponds, or Vernal Pools. Depression Pocosins are characterized by a dense to fairly dense shrub layer and a sparse to fairly dense tree layer. Of the limesinks surveyed, only Gum Pond with its canopy of black gum, pond cypress, and red maple fits that de- scription. Small Depression Ponds are permanently or nearly per- manently flooded in the center and may have scattered pond cypresses and black gums. Cypress Pond is an example of a small depression pond, although a portion of its east section has char- acteristics of a vernal pool. Vernal Pools are seasonally flooded depressions dominated by herbs, although there may be a few wetland trees or shrubs in the depression interior. In the park, vernal pools include Grass, Dry, Lily, Loblolly, and Sawgrass Ponds. In many cases, classification of limesinks is difficult when characteristics overlap two or more of these community types. According to Schafale and Weakley (1990), “Differences in hy- droperiod seem to be responsible for the different vegetation of the community types. The three types may represent different Stages in primary succession, which is proceeding at different rates in different depressions.” Seasonal fluctuation in water levels and its variation among years appears to be an important factor in determining presence of plant species. Because the ponds have no surface exits, they hold water fairly well in wet years. The deepest ponds may have several feet of standing water; in dry years they may be completely dry. During the growing season that these limesinks were surveyed, most of the depressions had some standing water from January to April, then gradually dried and were without standing water for the remaining months, al- though soils in the deepest parts remained moist. The water table was reported to be several inches lower than usual in southeastern North Carolina during that year (D. M. DuMond, pers. comm.); as a result, some species such as Eleocharis equisetoides which had been reported previously were not encountered. It is likely that in a wet year additional species would be apparent. For example, Utricularia purpurea Walt., not seen in 1990, was re- Ported to be in the Grass Pond in 1991 (J. Fowler, pers. comm.) and was observed in the Cypress Pond; the summer of 1991 was quite rainy. On the other hand, many rare species found in the 164 Rhodora [Vol. 94 ponds would not be apparent during a rainy season because they would be under water. The limesinks were observed again in September 1991 and most were still full of standing water. Dry Pond had none, but its soil was quite moist; Loblolly Pond had shallow water covering about one-half its total area. These limesinks are especially rich in species of special interest or concern because of rarity. The state record (with voucher ci- tation in parentheses) is Andropogon perangustatus (Sieren 4120), which was found growing abundantly in moist parts of Gum Pond and the east section of Cypress Pond. Hitchcock (1951) indicated its habitat and range as “Bogs and moist pine woods, Florida and Mississippi.”” New Hanover county records include Aristida affinis (Sieren 4095), A. purpurascens (Sieren 4127), A. virgata (Sieren 4136), Bartonia paniculata (Sieren 4094), Eragrostis refracta (Sieren 4105), Eupatorium leptophyllum (Sieren 4097), Linum floridanum var. floridanum (Sieren 4044), Lycopodium caroli- nianum (Sieren 4057), Paspalum praecox (Sieren 4049), Rhyn- chospora inundata (Sieren 4028), and Utricularia juncea (Sieren 4108). Fifteen species are on the Natural Heritage Program List of the Rare Plants of North Carolina (Weakley, 1990): Agalinis linifolia (occasional to locally abundant in all seven ponds), 4. virgata (scattered to locally abundant in Cypress, Dry, Grass, Gum, and Sawgrass Ponds), Aristida affinis (occasional to frequent in Cy- press, Grass, Gum, and Sawgrass Ponds), Eleocharis equisetoides (previously observed and photographed in Cypress Pond), EL. mel- anocarpa (frequent and abundant in Loblolly Pond), Eupatorium leptophyllum (scattered under pond cypresses in Cypress Pond), Lachnocaulon beyrichianum (frequent in the woods/pond eco- tones around Cypress, Dry, Grass, Gum and Loblolly Ponds), Ludwigia linifolia (scattered to abundant in Cypress and Grass Ponds), L. suffruticosa (frequent in Cypress and Sawgrass Ponds), Panicum tenerum (occasional in Cypress Pond), Psilocarya scir- poides (abundant in Lily Pond), Rhexia cubensis (scattered to numerous in Dry, Cypress, Grass, and Gum Ponds), Rhynchos- pora pleiantha (abundant in patches in Dry, Grass, and Cypress Ponds), R. tracyi (several plants in Cypress and Sawgrass Ponds), and Scleria georgiana (scattered in Dry and Sawgrass Ponds). An additional 13 species are on the North Carolina Watch List which includes plant species that appear to be rare or otherwise threat- ened with serious decline, but which have not yet been placed on 1992] Sieren and Warr—Limesink Flora 165 the Rare Plant List of North Carolina (Weakley, 1990). These species are Agalinis fasciculata (one plant in the Sawgrass Pond), Andropogon capillipes (scattered to abundant on the dry margins of Dry, Cypress, Loblolly, and Sawgrass Ponds and also observed in a pocosin/savannah ecotone and in other scattered depressions in the park), Bartonia paniculata (scattered in Cypress, Grass, and Sawgrass Ponds), B. verna (frequent to numerous in Grass, Gum, and Sawgrass Ponds), Burmannia biflora (several plants in Sawgrass Pond), Coelorachis rugosa (few to numerous in Gum and Sawgrass Ponds), [/ex cassine (occasional to numerous on the dry margins of all seven ponds), Juncus abortivus (scattered to locally numerous in Cypress, Dry, Grass, and Sawgrass Ponds and also observed in two pocosins in the park), Paspalum praecox (occasional in Sawgrass Pond), RAynchospora fiifolia (locally abundant in Cypress and Gum Ponds), R. inundata (abundant in Lily Pond, the dominant herb in mid-summer), R. wrightiana (locally abundant in Loblolly and Sawgrass Ponds), and Xyris smalliana (occasional in Grass Pond). Cypress Pond, which has the greatest variety of habitats, also has the largest number of rare or watch list species (18), followed by Sawgrass Pond with 15, Grass Pond with 12, Dry and Gum Ponds with 8 each, Loblolly Pond with 5, and Lily Pond with 3. Vegetation of the limesinks is clearly denser and more diverse than that of the surrounding longleaf pine-scrub oak communi- ties. Mesic depressions have the most diversity, and it is assumed that an appropriate amount of moisture is significant in the dis- tribution of limesink species. The presence of so many rare species in the depressions suggests that there may be a unique combi- nation of factors responsible. Beal (1977) provided habitat data for many marsh and aquatic vascular plants of North Carolina, including eight of the rare species found in the limesinks. An analysis of Beal’s data for the pH of the habitats for those eight species indicated that most were growing in sites that were neutral to alkaline. Three of the species were collected at least once in conditions which were acid (pH 44.9), but two of those were also collected in alkaline sites. Similarly, chloride content for most species was between 0-1 ppm, but in two cases those limits were exceeded. A study of limesinks for these and other environmental factors should provide a greater understanding of conditions con- trolling the distribution of limesink species. 166 Rhodora [Vol. 94 ACKNOWLEDGMENTS We thank the North Carolina Division of Parks and Recreation for permission to conduct this study, H. Leo Dillard and Phoebe Wahab of Carolina Beach State Park for their cooperation and assistance, Kelly W. Allred for several Aristida identifications, and Jimmy R. Massey and K. N. Ghandi for aid with the deter- minations of the Andropogon perangustatus specimens. LITERATURE CITED BEAL, E.O. 1977. A Manual of Marsh and Aquatic Vascular Plants of North Carolina with Habitat Data. Tech. Bul. No. 247. N.C. Agricultural Experi- ment Station, Ralei Hitcucock, A. S. 1951. oo of the Grasses of the United States, Rev. 2nd ed. by Agnes Chase. United States Department of Agriculture Miscellaneous Publication No. 200. U.S. Government Printing Office, Washington. .T. ANDR. Kartesz. 1980. A Synonymized Checklist of the Vascular Flora of the United States, Canada, and Greenland, Vol. 2: The Biota of North America. Univ. of North Carolina Press, Chapel Hill. McDonaLp, C., A. AsH AND J. Fussett. 1981. Natural Areas Inventory of Craven County, North Carolina. Report No. 15, North Carolina Coastal Energy Impact Program, Raleigh. FORD, A. E., H. E. AHLEs AND C. R. Bett. 1968. Manual of the Vascular Flora of the Carolinas. Univ. of North Carolina Press, Chapel Hill. ALE, M. P. AND A. S. WEAKLEY. 1990. Classification of the Natural Com- munities, Third Approximation. North Carolina Natural Heritage Program, Division of Parks and Recreation, N.C. Department of Environment, Health, and Natural Resources, Raleigh. TaGcarr, J. AND T. Dickerson. 1980. Carolina Beach State Park Natural ‘esta Report to the North Carolina Division of Parks and Recreation, Raleigh. Weak ey, A. S. 1990. Natural Heritage Program List of the Rare Plant Species of North Carolina, 2nd ed. North Carolina Natural Heritage Program. Di- vision of Parks and Recreation, N.C. Department of Environment, Health, and Natural Resources, Raleigh. DEPARTMENT OF BIOLOGICAL SCIENCES UNIVERSITY OF NORTH CAROLINA AT WILMINGTON WILMINGTON, NC 28403 —EE RHODORA, Vol. 94, No. 878, pp. 167-170, 1992 THE NORTHEASTWARD SPREAD OF MICROSTEGIUM VIMINEUM (POACEAE) INTO NEW YORK AND ADJACENT STATES Davip M. HUNT AND ROBERT E. ZAREMBA ABSTRACT Microstegium vimineum is a weedy species which has recently spread from New Jersey into New York and Connecticut. It has been collected at two sites along the Hudson River in New York and reported from three sites in Connecticut. Because of its ability to invade rapidly into disturbed areas, this grass may pose a threat to rare native associate species. Key Words: Microstegium vimineum, grass, weed, exotic species, geographical distribution, New Jersey, New York, Connecticut Microstegium vimineum (Trin.) A. Camus [= Eulalia viminea (Trin.) Kuntze] was introduced from Asia into Tennessee about 1919 (Fairbrothers and Gray, 1972). By 1960 this grass had nat- uralized and spread northward to Ohio and Pennsylvania, and eastward to all the Atlantic coastal states from Florida to New Jersey. Its geographical range has steadily expanded northeast- ward from Tennessee, reaching Pennsylvania in 1938, Delaware in 1942 and New Jersey in 1959 (Fairbrothers and Gray, 1972). Since the first report of this grass from New Jersey in 1959, it has increased in abundance and spread throughout the state. Ac- cording to L. Mehrhoff and D. Snyder (pers. comm.), Microstegi- um vimineum was uncommon in New Jersey ten years ago, but is now widespread. Collections from two sites in northern New Jersey during 1989 are deposited at CHRB (Hunt NJ42, Morris Co., Jefferson Township, along Russia Brook; Hunt NJ43, Som- erset Co., Bernards Township, along tributary of Peapack Brook). Additionally, this grass was observed between 1988 and 1990 by the senior author at seven other sites in northern and central New Jersey (Hunterdon Co.: Readington Township; Middlesex Co.: Cranbury Township; Somerset Co.: Bedminster, Warren and Ber- nards Townships). These observations were invariably along riv- €fs Or small tributaries, including Millstone River, Passaic River, Chambers Brook and Muddy Run. _ Fairbrothers and Gray (1972) noted that Microstegium vimi- neum occupies various habitats including creek banks, river bluffs, floodplains, damp fields, swamps, lawns, woodland thickets and 167 168 Rhodora [Vol. 94 roadside ditches. In New Jersey, this species was found in habitats ranging from forested, scrub-shrub and emergent wetlands to pas- tures, early successional fields, and forested and disturbed up- lands. It was commonly observed in transitional floodplain forests dominated by Acer rubrum L., Ulmus americana L., Impatiens capensis Meerb. and Onoclea sensibilis L., in shallow emergent wetlands along small streams with Cornus amomum Mill., Rosa multiflora Thunb. and Polygonum sagittatum L., and in wet meadows with Carex lurida Wahl., Juncus effusus L., Phalaris arundinacea L. and Euthamia graminifolia (L.) Nutt. Barden (1987) noted that this grass readily invades and occurs most abun- dantly in areas which have undergone natural (e.g., flood scouring) and artificial (e.g., mowing, tilling) disturbance. Microstegium vimineum has most recently spread northeast- ward from New Jersey into New York. Fertile collections made in September 1987 from two sites registered by The Nature Con- servancy (TNC) in the Hudson River Valley are deposited at NYS (Zaremba 4254, Dutchess Co., Town of Fishkill, red maple woods near a man-made berm on the S Side of Fishkill Creek near its mouth; Zaremba 4255, Rockland Co., Town of Stony Point, moist open forested wetland on Round Island). These specimens may be the only collections of this species from New York since this species is absent from the recent Checklist of New York State Plants (Mitchell, 1986). The spread of Microstegium vimineum has also continued northeastward into Connecticut. Populations were recently ob- served at a site owned by TNC in the Town of Haddam (Mid- dlesex Co.) along the Connecticut River (L. Mehrhoff, pers. comm.), and in New Haven (New Haven Co.) and New London (New London Co.) (G. Tucker, pers. comm.). These reports may be the first of the occurrence of M. vimineum in Connecticut and New England, as this grass is not listed in Seymour’s (1989) Flora of New England. Because Microstegium vimineum is apparently lacking in up- lands which intervene between the Hudson and Connecticut Riv- ers, it is hypothesized that this species may be nearing its northern limit due to a lack of cold hardiness. Barden (1987) noted the susceptibility of North Carolina populations to late frost. This species may continue to spread northeastward to a limited extent along estuaries of Connecticut, Rhode Island and southeastern 1992] Hunt and Zaremba— Microstegium 169 Massachusetts where the local climate is warmer than adjacent inland areas. It may also spread farther upstream to the tidal limits of the Hudson and Connecticut Rivers and their tributaries and increase in abundance in the lower Hudson and Connecticut River Valleys. A second possible explanation for the distribution of Microste- gium vimineum is that it occurs primarily in areas of red clay soils. New Jersey collections were from the Piedmont physiographic province, noted for such soils. This species was commonly found in alluvial soils composed of sandy clay loam associated with crimson-colored Brunswick shale. This grass is considered a char- acteristic Piedmont species (Godfrey, 1980) and its distribution in Georgia (Jones and Coile, 1988) and the Carolinas (Radford et al., 1968; Barden, 1987) corresponds closely with the limits of the Piedmont in these states. The New York and Connecticut populations occur in areas where red clay is locally abundant. If soil type is limiting the spread of M. vimineum, then this species may be approaching its maximal geographic extent in New York and New England, since large areas of red clay are uncommon outside the Hudson and Connecticut River Valleys. In this re- spect, M. vimineum may parallel other taxa whose range is closely linked to the distribution of red clay soils such as Quercus mari- landica Muenchh., Q. stellata Wang., Diospyros virginiana Le Pinus echinata Mill. and Asclepias viridiflora Raf. Microstegium vimineum is colonial in nature, rooting at the nodes and forming dense monotypic stands. A rapid rate of spread has been documented for this species, especially in disturbed areas (Barden, 1987). Although it may not be as aggressive and form colonies as dense and as tall as Phragmites australis (Cav.) Trin. ex Steud., Lythrum salicaria L. or other exotic species, it could threaten uncommon native wetland species along the Hudson and Connecticut Rivers. It may become necessary to establish an ac- live monitoring and eradication program to ensure survival of these rare species. The North Carolina office of TNC is preparing an “Element Stewardship Abstract” for M. vimineum which sum- marizes the ecology of the species, the threat it poses to native vegetation, and known means of control. It is the intent of this article to increase awareness of the potential danger of this weedy species in the northeast with regard to possible future extirpation of rare native associate species. 170 Rhodora [Vol. 94 LITERATURE CITED BARDEN, L. S. 1987. Invasion of Microstegium vimineum (Poaceae), an exotic, annual, shade-tolerant, C, grass, into a North Carolina floodplain. Amer. Midl. Naturalist 118: 40-45. FAIRBROTHERS, D. E. AND J. R. GRAY. 1972. Microstegium vimineum (Trin.) A. Camus (Gramineae) in the United States. Bull. Torrey Bot. Club 99: 97-100. Goprrey, M. A. 1980. The Piedmont, A Sierra Club Naturalist’s Guide. Sierra Club Books, San Francisco, CA MiTcHELL, R.S. 1986. A Checklist of New York State Plants. New York State Museum Bull pao Albany, NY. Jones, S. B., JR. N. C. Come. 1988. Distribution of the Vascular Flora of Georgia. Cuveeny of Georgia Botany Department, Athens ; Raprorp, A. E., H. E. AHLEs AND C. R. BELL. 1968. Manual of the Vascular Flora of the Carolinas. University of North Carolina Press, Chapel Hill, NC. Seymour, F.C. 1989. The Flora of New England, 2nd ed. Privately printed. NEW YORK FIELD OFFICE THE NATURE CONSERVANCY 1736 WESTERN AVE. ALBANY, NY 12203 | RHODORA, Vol. 94, No. 878, pp. 171-209, 1992 NATURAL PLANT COMMUNITIES OF BERKSHIRE COUNTY, MASSACHUSETTS! PAMELA B. WEATHERBEE AND GARRETT E. CRow ABSTRACT The natural plant communities and habitats of Berkshire County, Massachu- setts, are described. Communities are delineated by a description of the physical characteristics of the habitat and typical vegetation. An extensive list of species most closely associated with each community is provided. Distribution of the communities within the county is discussed, and examples are given. Other factors, both biotic and abiotic, that infl devel t of each ity are noted. Key Words: Natural plant communities, habitats, Berkshire County, Massachu- setts INTRODUCTION Berkshire County covers 260,860 ha and is located at the west- er end of Massachusetts. Varied topography and bedrock have resulted in diverse habitats that support a flora of 1777 taxa (Weatherbee, 1990), both native and introduced, including 117 State-listed rare taxa (Sorrie, 1989; they are indicated by an as- terisk in this discussion). The plant communities described below for Berkshire County are based in large part on Rawinski’s (1983) outline of a classi- fication of natural communities in New England. In his classifi- cation, the communities are described by a brief description of Physical characteristics of the habitat and physiognomy of the community, and are identified by a few of the most characteristic plant and animal species. Although less detailed, his descriptions of communities are similar in scope to that of Reschke (1980) for adjacent New York State, but we found Rawinski’s classification more applicable to New England. For this study, information on species composition was derived from extensive field notes of the senior author made during the Preparation of a Flora of Berkshire County (Weatherbee, 1990), from notes made by Bruce Sorrie, former Massachusetts Natural eritage and Endangered Species Program Botanist, and from theses and published studies conducted in Berkshire County. Us- ' Scientific contribution No. 1725 from the New Hampshire Agricultural Ex- Periment Station. 171 172 Rhodora [Vol. 94 ing this information, the authors were able to provide a much more detailed account of the vascular plant species that are char- acteristic of each community than that given by Rawinski (1983), and to describe the structure of those communities in Berkshire County. The most important factors determining habitat and plant com- munity relate to bedrock geology (Figure 1) and to the impact of glaciation, climate, and plant migration. Geologic events have influenced the development of soil, topographic relief, drainage patterns, and elevation (Figure 2). Bedrock described as calcar- eous includes the Stockbridge Formation, which consists of var- ious types of marble, and the Walloomsac Formation which con- sists of calcitic and schistose marble (Zen, 1983). Regional climate determines the overall type of vegetation, but there is some varl- ation in climate within the county due to elevational variation (176 m to 1064 m) and microclimatological differences produced by local topographical diversity. Post-glacial migration of species from various refugia south of the glacial margin or from the exposed coastal plain has played an important role in determining the pool of species which ultimately colonized available habitats and developed our plant communities. The following 35 natural plant communities are recognized by us for Berkshire County: Forests and Woodlands Mesic Northern Conifer Forest Mesic Northern Hardwood Forest Rich Mesic Forest Mesic Acidic Oak/Conifer Forest Dry Acidic Oak/Conifer Forest Dry Calcareous Oak/Conifer Forest Floodplain Forest Pitch Pine/Scrub Oak Barrens Rocky Summit and Cliff Communities Southern Acidic Rocky Summit Southern Calcareous Rocky Summit Southern Acidic Cliff Southern Calcareous Cliff Serpentine Outcrop VERMONT SCHIST GNEISS AND QUARTZITE CALCAREOUS ROCK SERPENTINE tS Ss a — CONNECTICUT Figure 1. Bedrock geologic map of Berkshire County (redrawn from Zen, 1983). 174 Rhodora [Vol. 94 LEGEND 305 - 457m Ee) 457-610m 610 - 762 m 762-914m WE 914-1067m f Berkshire County (adapted gur T r ao pl Ps 1 ¥ J i ao pl from Technical Planning Associates, 1959). 1992] Weatherbee and Crow— Plant Communities |W ge Lakes and Ponds Clear Softwater Lake/Pond Acidic Brownwater Lake/Pond Moderately Alkaline Lake/Pond Highly Alkaline Lake/Pond Rivers, Streams and Springs High Gradient Stream Medium Gradient Stream Low Gradient Stream Spring and Spring Run Wetlands Acidic Conifer Swamp Acidic Hardwood Swam Circumneutral Hardwood Swamp Acidic Shrub Swamp Circumneutral Shrub Swamp Robust Emergent Marsh Acidic Graminoid Marsh Circumneutral Graminoid Marsh Level Bog Forested Fen Shrub Fen Sloping Graminoid Fen Lake Basin Graminoid Fen Calcareous Seep FORESTS AND WOODLANDS Mesic Northern Conifer Forest Community The montane spruce-fir forest (Oosting and Billings, 1951, Cog- bill, 1987), the boreal forest (Siccama, 1974) or the Northeastern Spruce-Fir Forest of Kiichler (1964) are alternate terms for this community dominated by Picea rubens and Abies balsamea. It is best developed above 914 m on Mt. Greylock summit and along the adjacent ridge of Saddle Ball where these two species occur with Betula cordifolia and B. alleghaniensis. The summit, 176 Rhodora [Vol. 94 although drastically altered now, was described in 1799 as being densely covered with stunted Abies balsamea (Dwight, 1822). Other associated species are Sorbus americana, S. decora*, Luzula parviflora ssp. melanocarpa*, Amelanchier bartramiana*, Gaultheria hispidula, Solidago macrophylla*, Viburnum lantanoi- des, Ribes glandulosum, Oxalis acetosella, Cornus canadensis, Dryopteris intermedia, D. campyloptera and Lycopodium lucidu- lum Many needle-leaved evergreen trees are well adapted to high elevations where a shorter frost-free season, colder average tem- peratures, high winds, heavy snow, lower light level and formation of rime ice due to frequent cloud cover, and shallow soil create difficult growing conditions (Marchand, 1987). In the spruce-fir zone of the Green Mountains, Siccama (1974) noted that the number of frost-free days was 93. He suggested that the eleva- tional limit of low-hanging clouds corresponds to the elevation of the transition zone between the montane boreal forest and the northern hardwoods zone. Abies balsamea trees on the summit of Mt. Greylock exhibit a “table tree” form, as described by Marchand (1987). Ice particles, wind-driven over the surface of the snow, remove foliage above the snow while branches below the snow remain luxuriant. Mesic Northern Hardwood Forest Community This widespread community covers the slopes and summits of all but the highest mountains and also the low hills and valley sides of the northern half of the county. Within the broader bound- ary of this community, other communities occur locally on edaph- ically, topographically or climatically distinct areas. The upper elevational limit of this forest ranges from 800-850 m, which closely agrees with Siccama’s (1974) estimate for this community type in northern Vermont of 792 m. Towards the southern part of the county, particularly in the valley region, this forest type grades into the Mesic Acidic Oak/Conifer Forest. The co-dominant tree species Acer saccharum, Fagus grandi- folia and Betula alleghaniensis occur in varying proportions throughout the community, with B. alleghaniensis becoming dominant at higher altitudes. Other trees that commonly occur are Fraxinus americana, Betula lenta, B. papyrifera, Quercus ru- ee 1992] Weatherbee and Crow— Plant Communities 177 bra, Prunus serotina and Acer rubrum. Except for Fraxinus amer- icana, these species may indicate a past history of disturbance, both natural and human. Quercus rubra, although often an im- portant component of the mature forest, does not regenerate well in shade. Very large individuals of Q. rubra have been observed on some slopes of Mt. Greylock, but the regeneration around the trees is all Acer saccharum. Tsuga canadensis is seen in small patches in moist ravines with thin soils, steep rocky, north-facing slopes and along moist streambanks. The Mesic Northern Hard- wood Forest becomes more common on the Berkshire Plateau; there it becomes intermixed with Picea rubens and Abies balsa- mea. Acer pensylvanicum, A. spicatum, Ostrya virginiana, and Ham- amelis virginiana, which are adapted to growing under the closed canopy, form a characteristic understory layer. Typical shrubs include Taxus canadensis, Viburnum lantanoides, V. acerifolium, Sambucus pubens and Lonicera canadensis. The vernal flora is somewhat sparse, but Claytonia caroliniana, Erythronium amer- icanum, Viola rotundifolium and Trillium erectum are common in the community. The ground cover includes Lycopodium spp., Mitchella repens, Aster acuminatus, A. divaricatus and the ferns Dryopteris intermedia and Polystichum acrostichoides. Factors that influence the extent of this forest are climate and soil. Siccama (1974) found the frost-free season in the Green Mountain northern hardwoods forest to average about 144 days, 51 days longer than in the montane boreal forest. Generally the soil is well-drained, but moist throughout the year, and most often derived from a bedrock of schist (Figure 1), and less often from gneiss or quartzite. In studies of the forests of New Hampshire, Leak (1978) found typical northern hardwoods occurred on fine glacial till. Kudish (1979) indicated that Acer saccharum and other deciduous trees need at least a 30-90 cm layer of till to grow well, while conifers can survive on only 8-15 cm of till. His studies in the Catskills showed that till depth was more limiting than pH or elevation. This factor appears to be important in the distri- bution of this forest community in Berkshire County. High ele- vation (Siccama, 1974) with its concomitant cloud cover and shorter growing season seems to be the most important factor in limiting development of this community type on mountain slopes. Drier soils may account for the community’s decreasing abun- dance in the southern section of the county. 178 Rhodora [Vol. 94 Rich Mesic Forest Community Occurring in smaller, discontinuous areas throughout the coun- ty and typically surrounded by other forest community types, sites of the Rich Mesic Forest correlate with, and are influenced by, calcareous bedrock (Figure 1) and associated alkaline ground- water. The best examples of this community are situated on slopes below calcareous outcrops, on talus below these outcrops or on fairly level sites where bedrock is near the surface. These sites tend to be on the low hills edging valleys, but they also occur higher (822 m) on the slopes of Mt. Graylock. The community is distinguished by a dominance of Acer sac- charum in most examples, sometimes to the near exclusion of other tree species. Other trees, in order of decreasing closeness of association with the community, are Fraxinus americana, Tilia americana, Carya cordiformis, Betula lenta, B. alleghaniensis and Fagus grandifolia. Leak (1978) found that in the White Mountains of New Hampshire, Acer saccharum became dominant only on enriched sites, and that Fraxinus americana was abundant only on these sites. Individuals of Quercus rubra may attain large stat- ure and remain part of this community for a long time, but the species does not regenerate well. Dirca palustris, a shrub, is Tre- stricted to this habitat, and is thus an important indicator species. Additionally, Ostrya virginiana, an understory tree, iS abundant in these rich sites. In the Rich Mesic Forest the herbaceous species are abundant and diverse, and the community has an especially rich vernal flora. Table 1 lists these species in approximate order of decreasing degree of association with the community. The most important determining factors in this community are the presence of calcareous rock and adequate moisture. The soil is derived from the soft, easily weathered rock. The higher pH of the soil facilitates microbial activity and decomposition of litter. Leaves of Acer saccharum are rich in bases (Dickinson and Pugh, 1974), thus favoring rapid decomposition and recycling of nutri- ents. An adequate supply of moisture is also crucial to this process and to development of mull soils typical of this community (Bor- mann and Buell, 1964). The litter layer is thinner than that of Northern Hardwood Forest because of rapid turnover, thus en- abling the small vernal species to receive enough sunlight to emerge 1992] Weatherbee and Crow— Plant Communities 179 able 1. Herbs of the Rich Mesic Forest, listed in approximate order of de- creasing degree of association with the community. * = State-listed rare species. Caulophyllum thalictroides (L.) Michx. Adiantum pedatum L. var. pedatum Dicentra canadensis (Goldie) Walp. Dicentra cucullaria iF ) Bernh. Uvularia grandiflora S Dryopteris goldiana (acl ex Goldie) Gray Viola selkirkii Pursh ex Goldie Galearis spectabilis (L.) Raf. & ypripedium calceolus L. var. pubescens (Willd.) Correll a hirsuta (Pars Benth. var. hirsuta *Panax quinquefolius L Oryzopsis racemosa (Sm. ) Ricker Carex amphibola ie var. rigida (Bailey) Fern. Carex rosea Schkuhr e illd Sanicula a Bick, *Sanicula gregari *Waldsteinia meer (Michx.) Tratt. Thalictrum dioicum Thalictrum (Anemonella) thalictroides (L.) Eames & Boivin | § Li. Botrychium — (L.) Sw. Viola canade Solidago Fads L: Carex hirtifolia Mackenzie anguinaria canadensis L. Actaea pachypoda El. Actaea rubra (Ait.) Willd. Laportea canadensis L. Osmorhiza claytonii (Michx.) Clarke 180 Rhodora [Vol. 94 early enough in the spring to complete their life cycles before the canopy leafs out. Mesic Acidic Oak/Conifer Forest Community This community is found in the southern half of the county and the southern third of the Berkshire Plateau (Egler, 1940) usually in rocky, acidic soil. This community is more difficult to define. It grades into the Mesic Northern Hardwood and Dry Acidic Oak/Conifer Forests, and in some areas may represent a stage rebounding from earlier disturbance (McVaugh, 1958). Kichler’s (1964) designation of this area as transitional between Mesic Northern Hardwood and Appalachian Oak, indicates the variability of this community. Westveld et al. (1956) used the term Transition Hardwoods to indicate the overlap of northern hardwoods with oaks and hickories. Quercus rubra, Acer rubrum and Pinus strobus share domi- nance, but Tsuga canadensis and Fagus grandifolia are important, and may become more dominant as the forest matures, as Egler (1940) describes for the southern section of the Berkshire Plateau. Prunus serotina, Betula lenta and B. papyrifera are common. Understory trees and shrubs include Acer pensylvanicum, Kalmia latifolia, Amelanchier arborea, Hamamelis virginiana, and Vi- burnum acerifolium. The ground cover layer is often somewhat sparse, but typical species include Lycopodium obscurum, Polys- tichum acrostichoides, Medeola virginiana, Maianthemum can- adense, Aster divaricatus, Polygonum cilinode, Aralia nudicaulls, and Carex pensylvanica. Soil derived from acidic bedrock, such as quartzite or gneiss (Figure 1), tends to be rocky, shallow, well- drained and poor in nutrients. These soil characteristics seem to be important in the distribution of this community. Kudish (1979) States that the tree species characteristic of this community can subsist on minimum till depths of 20-45 cm. Leaves of Quercus rubra do not decompose quickly, resulting in the development of a thick litter layer; they also contribute to the development of acidic soil. The same is true of the needles of Pinus strobus and Tsuga canadensis (Dickinson and Pugh, 1974). Acidic, shallow soil is not favorable for growth of Acer saccharum, thus this species is often absent. Betula alleghaniensis is often scarce as well. Since the canopy of Quercus rubra is more open than a canopy of Acer saccharum, the light levels that reach the forest PS 1992] Weatherbee and Crow— Plant Communities 181 floor favor reproduction of Quercus rubra and Acer rubrum, spe- cies that require higher light levels for seedling growth. Dry Acidic Oak/Conifer Forest Community This community, dominated by species of Quercus, occurs on soils derived from acidic rock, typically from quartzite (Figure 1), on upper slopes where the soil is thin and excessively well- drained. Sites occur throughout the Berkshires, particularly on south-facing slopes, but the community is more common in the southern half of the county. Quercus prinus, along with Q. velutina, Q. alba and Pinus rigida, are the typical tree species, but Pinus strobus, Betula papyrifera, Acer rubrum and B. lenta are also very common. Castanea dentata is frequently present, and may attain small tree size before suc- cumbing to Chestnut blight. Carya glabra occurs most commonly in southernmost towns, and Sassafras albidum may be locally common. The canopy is quite open, and ericaceous shrubs, such as Vaccinium angustifolium, V. pallida, Gaylussacia baccata, d Kalmia angustifolia form a dense shrub layer. Gaultheria procures and Epigaea repens are com- mon evergreen ground cover species in more open areas. Her- baceous species are few and are represented by Pteridium aquili- num, Melampyrum lineare, Cypripedium acaule, Lysimachia quadrifolia and, uncommonly, Isotria verticillata. The thin, dry soil that develops over acidic bedrock with many fractures and joints providing drainage, is an important factor in the distribution of this community. Because decomposition of litter and development of soil is slow in acidic sites (Dickinson and Pugh, 1974), these soils tend to be thin and infertile (Mc- Intosh, 1959). The openness of the forest canopy, coupled with an often south-facing slope, allows high insolation of the soil, which results in increased dryness and higher soil temperatures (Cantlon, 1953). Succession to a more mesic and closed-canopy forest may occur (Little, 1973), but this process may easily be set back by the occurrence of an occasional fire, to which this com- munity is particularly susceptible. Fire, which reduces ground cover and stimulates the opening of serotinous cones (Elias, 1987), is important to the regeneration of Pinus rigida. This community shares many species with, and gradually grades into, the Southern 182 Rhodora [Vol. 94 Acidic Rocky Summit Community, a community that usually occupies the ridges above this community. Dry Calcareous Oak/Conifer Forest Community This community, occupying well-drained slopes or low ridges underlain with calcareous rock (Figure 1), occurs more commonly in the southern half of the county. In the north, it is found occa- sionally on south- or southwest-facing slopes, but lacks some of the characteristic species of the community type as it occurs in the southernmost towns. These more southerly species are des- ignated by a plus sign (+) in the following discussion. Quercus alba, Q. velutina and Q. muhlenbergii** are charac- teristic of this community, along with Carya ovata and Ostrya virginiana. Pinus strobus and Quercus rubra occur commonly, although scattered. The understory trees Cornus florida and Staphylea trifolia*, flourish under this more open canopy. Carya ovata and Ostrya virginiana often form a distinct association, with which two rare grasses occur: Poa languida* and Sphenopho- lis nitida*. The community is rich in herbaceous species, which are listed in Table 2 in approximate order of decreasing degree of association with the community. Bedrock, slope and aspect play an important part in the com- position of this community. The soil, a fine sandy or clayey loam, is strongly influenced by the calcareous bedrock, which may be near the soil surface, or may occur as an outcrop. In contrast to the Rich Mesic Forest, there is no seepage of groundwater. A topographical barrier, a ridge of 426 m elevation, rises north of Lanesborough, and may have prevented those species common in the southern part of the county from spreading to northern sites, in addition, there is little suitable habitat. The high number of herbaceous species may relate to the less dense Quercus SPP. canopy and high soil fertility derived from calcareous bedrock. Floodplain Forest Community This well-defined community occurs on deep, alluvial soil bor- dering the low-gradient sections of the major rivers: the Housa- tonic, Hoosic, Green (in Great Barrington), Williams, Konkapot and Deerfield Rivers. The community which experiences periodic flooding, most frequently in the spring, may occur as a narrow Seb ea ea tin aS 1992] Weatherbee and Crow— Plant Communities 183 Table 2. Herbs of the Dry Calcareous Oak/Conifer Forest, listed in approxi- mate order of decreasing degree of association with the community. * = Species restricted to southern-most towns. * = State-listed rare species. Asclepias quadrifolia Jacq. Arabis canadensis L. + Lespedeza violacea (L. , Lespedeza intermedia (S. Wats.) Britt. Rosa carolina L. Aureolaria flava (L.) Farw. *Aureolaria virginica (L.) Pennell Linum virginianum L. *Ranunculus hispidus Michx. var. hispidus **Chamaelirium luteum (L.) Gra *Muhlenbergia tenuiflora (Willd.) BSP. * Silene caroliniana Walt. var. pensylvanica (Michx.) Fern. *Paronychia canadensis (L.) Wood *Desmodium rotundifolium DC. Hieracium venosum var. nudicaule (Michx.) Farw. *Sphenopholis nitida (Biehl.) Scribn. *Poa languida Hitchc. * Helianthus divaricatus L. Carex eburnea Boott Desmodium paniculatum (L.) DC. var. paniculatum Uvularia perfoliata L. band along the river or as extensive flat areas, and include oxbow ponds and associated marshes. In some areas there is a step-wise series of terraces back from the river, the highest level being the driest. Silt bars and mud banks provide small pioneer habitats, colonized by a few annual species or by seedlings of the Floodplain Forest trees. Many of these rich alluvial habitats have been cleared for agriculture, leaving only a narrow strip, if any, of the original vegetation along the river. Salix nigra, Platanus occidentalis and Populus deltoides are early colonizers of banks and silt bars (Nichols, 1916) and are more often found nearest the river. Acer negundo is more common as a floodplain tree along the Hoosic River, while A. saccharinum is dominant on floodplains in the Housatonic River watershed. Acer nigrum* is locally common. Ulmus americana, formerly 184 Rhodora [Vol. 94 common along floodplains, is now more typically represented by bare, dead trunks than by living trees. Fraxinus pennsylvanica is usually found on riverbanks, while Jug/ans cinerea, Tilia amer- icana and Celtis occidentalis typically occupy the upper alluvial terraces of the floodplain. There are few shrubs characteristic of the floodplain; however, introduced Lonicera morrowii has become weedy. Several woody vines, Vitis riparia, Parthenocissus quinquefolia and Menisper- mum canadense, and the herbaceous vines, Echinocystis lobata and Sicyos angulatus, flourish on the floodplains. Many of the herbaceous species (Table 3) occur restricted to floodplain forests or are more commonly associated with this community. Inundation is the major factor influencing in development of this community. Plants growing in a typical floodplain, which 1s usually inundated for a significant length of time only during the spring (Nichols, 1916), must be adapted to withstand a reduced oxygen supply to their roots during that time. Likewise, these species must tolerate a high water table most of the year. The impact of inundation may account for the lack of a shrub layer. Ice floes scour banks and silt bars, resulting in many newly dis- turbed sites, open for colonization. The cutting of a new channel may shift the river, resulting in somewhat isolated floodplain remnants. The environment of a river is one of constant change. The soil, silt that has been deposited by the river, is typically deep, without distinct horizons and contains considerable organic matter (Rawinski, 1983). The woody species occupying these sites typically grow to a large size, and herbaceous growth can likewise be exceedingly vigorous and tall. Calcareous rock underlies por- tions of these river valleys, and probably influences the higher fertility of the soil. Pitch Pine/Scrub Oak Barren Community This community once occurred extensively on the sandy out- wash plains along the Housatonic and Konkapot Rivers in the southernmost towns of Sheffield and New Marlborough. Pinus rigida and Quercus ilicifolia dominated an open woodland with a dense shrub layer of Vaccinium angustifolium, V. pallidum and Gaylussacia baccata. Grassy clearings were dominated by Schi- zachyrium (Andropogon) scoparium, and Quercus prinoides OC 1992] Weatherbee and Crow— Plant Communities 185 Table 3. Herbs of the Floodplain Forest, listed in approximate order of de- creasing degree of association with the community. * = State-listed rare species. *Arisaema dracontium (L.) Schott *Carex davisii Schwein. & Torr. *Carex grayi Carey *Carex trichocarpa Schkuhr hx. Polygonatum biflorum (Walt.) Ell. var. commutatum (Schultes f.) Morong Allium canadense L. var. canadens Xanthium strumarium L. var. pnt oe (P. Mill.) Torr. & Gray Ambrosia trifida L. Helenium autumnale L. var. autumn Teucrium canadense L. var. ras “ae Eat. *Aster prenanthoides Muhl. Elymus wiegandii om Elymus riparius Wie *Elymus villosus Ne ex Willd. Bromus altissimus Pursh Matteuccia struthiopteris i ) Todaro Carex hirtifolia Mackenz eum laciniatum eage var. trichocarpum Fern. Scrophularia marilandica L. Rorippa palustris (L.) Bess. var. fernaldiana (Butters & Abbe) Stuckey Carex spregnelii Dewey ex Spreng curred as scattered a along with the herbaceous species char- acteristic of this habita At present, ele and residential housing fragment these natural habitats. While formerly dominant, Pinus rigida now 1s being displaced by Pinus strobus, Quercus alba, Q. rubra, Q. coc- cinea, Acer rubrum and Prunus serotina. Quercus ilicifolia now occurs primarily on roadsides and edges of this successional forest. Vaccinium angustifolia, V. pallida and Gaylussacia baccata per- sist as the shrub layer, along with species more commonly found in Mesic or Dry Acidic Oak/Conifer Forest. The following species typical of grassy clearings are now found along roadsides or in Clearcuts in the sandplain area: Quercus prinoides, Carex brevior, Dichanthelium oligosanthes, Schizachyrium (Andropogon) sco- 186 Rhodora [Vol. 94 parium, Cyperus filiculmis, Bulbostylis capillaris, Juncus secun- dus, Rhus copallina, Helianthemum canadense, Hypericum gen- tianoides, Linaria canadensis, Aster linariifolius and Krigia virginica. Species that were listed by Hoffmann (1922) as occa- sional or common on the sandplain, but have not been re-located since in Berkshire County, include Helianthemum bicknellii, Lu- pinus perennis, Crotalaria sagittalis, Asclepias amplexicaulis and A. tuberosa. These communities are strongly influenced by dry, sandy soil, and are fire dependent. Breakdown of litter is slow because of dryness and low pH of the soil, which results in considerable accumulation of litter, a condition conducive to fire. Fire removes litter, creating conditions that allow sunlight to reach the soil (Henderson, 1982). Bare mineral soil is often necessary for the regeneration of species adapted to fire, such as Pinus rigida. How- ever, due to development, incidents of fire are suppressed so that it is unlikely to be a factor in the future. The absence of burning allows conditions conducive to a vegetational change favoring establishment of Quercus spp., other hardwoods, and Pinus stro- bus, as noted by Seischab and Bernard (1991) in New York State. ROCKY SUMMIT AND CLIFF COMMUNITIES Localities occupied by rocky summit and cliff communities are limited in extent, as exposed rock is rare in this generally mesic region. Acidic bedrock, either gneiss, schist or quartzite, most often is found on summits and ridgetops. In calcareous regions the rock is easily weathered, resulting in few distinct outcrops or vertical cliffs, especially at higher elevations. Differences between the summit and cliff communities may be difficult to define as ridgetops quickly grade into more shaded lower slopes and ledges. Southern Acidic Rocky Summit Community Occupying mountain summits and ridges, this community is more Common on the southern Taconic summits of the elevation range from 600-792 m, notably Mt. Everett, Mt. Race and Alan- der Mountain in the southwest portion of the county, on East (533 m) and Monument Mountains (533 m) and on Pine Cobble Mountain (560 m) in the north. Vegetationally, this community is closely related to the Dry 1992] Weatherbee and Crow— Plant Communities 187 Acidic Oak/Conifer Forest, but is characterized by larger areas of open bedrock, scattered small trees and dense low shrub growth. Pinus rigida is always present, and on the southern Taconic ridges growth of Quercus ilicifolia may be very dense. Other trees, such as Betula papyrifera, Quercus rubra and Acer rubrum, may appear in a rather stunted form. Juniperus communis occurs occasionally. The shrub layer is a dense growth of ericaceous shrubs including Vaccinium angustifolium, Gaylussacia baccata and Rhododen- dron prinophyllum. On a few southern summits, Arctostaphylos uva-ursi occurs in Open sunny areas at the edges of outcrops. Other shrubs that grow mixed with these species are Aronia mela- nocarpa, Amelanchier stolonifera and, uncommonly, Prunus pumila var. susquehanae. Potentilla tridentata appears frequently in Openings, as it does on more northern, higher, open ridges. Diversity of herbaceous species is low: Woodsia ilvensis occurs occasionally in southern locations; Deschampsia flexuosa and Schizachyrium scoparium are common tuft-forming grass species; Carex pensylvanica forms extensive patches; Corydalis semper- virens is frequent in shallow soil pockets on open rock. Physiognomy of this community is usually one of extensive low shrubs with few trees. However, an unusual dwarf “forest” of Pinus rigida, with Quercus ilicifolia, is found on the flat summit of Mt. Everett and on a level ridge in Clarksburg on the southern extension of the Green Mountains; Quercus ilicifolia is not present in the latter site. The dwarf trees average about 1 m in height, and present a flat-topped, laterally-growing aspect. Keith (1912) quotes Professor Edward Hitchcock’s description of Mt. Everett's summit in 1839: “ta naked eminence with numerous yellow pines, two or three feet high.” McIntosh (1959) notes a similar dwarf Pinus rigida forest on a flat-topped summit in the Shawangunk Mountains of New York. These summits, unlike most others in the county, have thin soil overlying the schist or quartzite, and what little organic ma- terial is there accumulates slowly. There is slight inflow of nutri- ents, such as might occur on a slope. Conditions are quite xeric, and the lack of a tree canopy allows extremes of heat and dryness to occur. Wind is probably not an important factor in keeping the vegetation low, as there are many higher ridges that are oc- Cupied by taller forests; however, fire may be a factor. McVaugh (1958) commented on the possible succession of this community and concluded that it appeared stable. Hoffman (1922) observed 188 Rhodora [Vol. 94 the same species that are presently seen, particularly an Arcto- staphylos uva-ursi population which was known to Dewey (in Field, 1829). Areas dominated by Vaccinium angustifolium exist on some northern mountain ridges, but these appear to be aban- doned pastures that were colonized by V. angustifolium. At one time, these areas were burned to maintain the blueberry fields, but they are now being slowly invaded by forest. Southern Calcareous Rocky Summit Community Rare throughout the county, usually this community occurs on ridgetops of low hills edging the main valleys, but there are a few high ridges whose rocks, while not true limestone or marble, have a distinct calcareous influence. West Stockbridge and Tom Ball Mountains are good examples. These open, dry rock outcrops support a vegetation whose spe- cies are intolerant of much shade, and require calcareous soil. Many of the plants that are specific for these situations are adapted to growing in crevices with little soil, Carex eburnea will often be abundant in such sites. Clematis occidentalis* is frequent at these sites, while Trichostema (Isanthus) brachiatum*, Minuartia michauxii* and Lonicera hirsuta* are rare. More northerly ele- ments, rare Rosa acicularis* growing with Senecio pauperculus, occupy a high ridge (532 m) in the northern third of the county. Tom Ball and West Stockbridge Mountains, sharp north-south trending ridges, support a mixture of both lime- and acid-loving species. Hedyotis longifolia* and Arabis lyrata* grow with Wood- sia ilvensis, an acidophile, in crevices on open rock. Amelanchier sanguinea* forms small patches close to Gaylussacia baccata, Pinus rigida, and Potentilla tridentata. Pinus resinosa, a native species in Berkshire County, occurs only on these ridges. Vibur- num rafinesquianum* tends to grow in more shaded situations on summits and also on rocky south-facing slopes. The tendency of trees to uproot and pull away from the steep- sided outcrops maintains the open aspect. Rocks splitting off from sharp ridges also renew open areas, and dry, shallow soil dis- courages woody growth. Southern Acidic Cliff Community Suitable habitat for the establishment of this community is widespread. The vegetation is sparse, and not chacterized by 4 i > ee ee ae SE —— 1992] Weatherbee and Crow— Plant Communities 189 specific or distinctive group of plants, unlike the Southern Cal- careous Cliff Community. Polypodium virginianum is always present in crevices and forms large mats on boulders, along with Aralia hispida, Dryopteris marginalis, Polygonum cilinode, Dier- villa lonicera and Parthenocissus quinquefolia, all plants found commonly in acidic situations. Two of the few rare plants found in this habitat are Asplenium montanum*, which occurs in two localities on quartzite, and Ad/umia fungosa*, found on boulder talus below cliffs. This habitat is cooler, moister and more shaded than the Sum- mit Community. The acidic rock produces little soil and nutrients for vegetation, and when shaded becomes unfavorable for many species. Southern Calcareous Cliff Community This uncommon community is found on steep cliff faces or sloping rock outcrops. Frequently, the ridge is capped with resis- tant rock while the soft limestone beneath has eroded into ledges and outcrops. The habitat, while found throughout the western part of the county, is best represented in Sheffield, particularly on Bartholomew’s Cobble. The vegetation, which includes many of the county’s rarer species, is distinct and specific to the habitat. The small ferns, Pellaea atropurpurea, Asplenium ruta-muraria*, A. trichomanes, A. platyneuron and Woodsia obtusa colonize sun- ny or slightly shaded crevices, Asplenium rhizophyllum occurs on moss-covered rock faces, while Selaginella rupestris and Cam- panula rotundifolia prefer dry, shallow soil pockets on open rock. Parietaria pensylvanica colonizes dry to moist rock outcrops; Ar . abis lyrata*, A. laevigata*, A. hirsuta and Saxifraga virginiensis colonize shaded, moist, deeper pockets of soil, and Cystopteris bulbifera and C. tenuis are common on moist, shaded calcareous rock. Of the few northern elements, Cryptogramma Stelleri* is found on dripping cliffs and Woodsia glabella* inhabits shaded crevices in a high, northern cliff. Few trees or shrubs can grow on these steep rock faces, therefore there is sufficient light and sites for the small species that are adapted for survival on little soil. This diverse community of plants is in contrast to the species poor, sparse vegetation of the Acidic Cliff Community sites. 190 Rhodora [Vol. 94 Serpentine Outcrop Community There are five or six locations for ultramafic rock in the north- eastern part of the county (Figure 1), only two of which occur as exposed outcrops. This bedrock is part of a narrow belt of ser- pentine rock occurrences along the east flank of the Appalachian Mountains that extends from Alabama to Newfoundland (Zika and Dann, 1985). Serpentine is found as intrusive rock in quartz- ite, gniess and schist on the Berkshire Plateau. he sharp contrast between vegetation of serpentine and non- serpentine soils is well-documented (Walker, 1954). This com- munity of serpentine outcrops is characterized as open, with sparse and stunted vegetation, and by the presence of serpentine species, which typically have disjunct distributions, or are even endemic (Zika and Dann, 1985). The outcrops in the county are not large enough to support an extensive serpentine vegetation. M oehringia macrophylla* is the only taxon on these outcrops that is restricted to serpentine outcrops. Cerastium arvense is most abundant on serpentine outcrops, but is also found occasionally in open rocky woods. Other taxa frequent on these serpentine formations in- clude Asplenium trichomanes and Campanula rotundifolia. The most extensive outcrop, in the town of Florida, also supports other taxa that are considered somewhat calciphilic, such as Se- laginella rupestris and Aquilegia canadensis. In other areas where the slope is not steep, soil may overlie the ultramafic rock, and the vegetation is not notably distinct. Infertility of serpentine soils is the result of complex factors, and plants endemic to serpentine have developed a variety 0 adaptations to these conditions. Serpentine rock and its derived soils are high in magnesium, iron and the heavy metals chromium and nickel, but usually deficient in major nutrients such as cal- cium, nitrogen and phosphorus (Walker, 1954; Brooks, 1987). Lack of organic matter and droughty characteristics of the soil contribute to infertility. LAKES AND PONDS These aquatic habitats are classified according to the alkalinity of the water, the most important factor influencing the distribu- tion of species. Hellquist (1980), in his study of the distribution of Potamogeton in New England, classified them into six groups. 4 a. ss eaediiigeee aoe inci e+ BS a tccctabbeaiinecai te iasean etc —— SS OS ii 1992] Weatherbee and Crow— Plant Communities 191 and found that the occurrence of most Potamogeton species was strongly correlated with certain ranges of alkalinity. Some species are restricted to a narrow range, but others many tolerate a wide range (Hellquist, 1980). Other aquatics | that correlate with alkalinity of the waters, which - is affected by its Origin in either acidic or calcareous bedrock. Alkalinity ranges of many aquatics are given in a series of papers on aquatic plants of New England (Crow and Hellquist, 1981, 1982, 1983, 1985; Hellquist and Crow, 1980, 1981, 1982, 1984). Depth of water, bottom conditions, water clarity, velocity of flow and habitat elevation are also important factors that determine species com- position of an aquatic habitat. Clear Softwater Lake/Pond Community Water bodies with extremely acidic waters occur in regions of acidic bedrock, usually schist or gneiss, at elevations between 448 and 622 m. Most acidic waters occur on the Berkshire Plateau where the poorly-drained, rolling topography has resulted in cre- ation of many lakes and ponds; a few occur on the uplands of the Taconic Range. Vegetation in general is characterized by submersed species with a rosette growth form and Potamogeton spp. that are re- stricted to very acidic waters. Isoetes echinospora, Eriocaulon pellucidum (= E. septangulare), Sagittaria graminea, Elatine minima and Lobelia dortmanna all have a submersed rosette of leaves. Potamogeton confervoides, P. bicupulatus and P. oakesi- anus are found only in these extremely acidic waters (Hellquist, 1980). Potamogeton spirillus occurs, but is not restricted to the most acidic waters. Floating-leaved components of this com- munity are the common and widespread Nuphar variegata, the infrequent Sparganium fluctuans and the uncommon Nym- Dhoides cordata. Myriophyllum tenellum and M. humile may also be present in acidic waters, but their occurrence in the Berkshires is rare. Acidic bedrock and shallow, rocky basins are important factors in the development of this community. In these waters, alkalinity measurements of up to 18 mg/liter HCO, (Hellquist, 1980) in- dicate low alkalinity; the pH may be as low as 5.0. The waters are also low in nutrients. Lake and pond bottoms are typically rocky, gravelly or sandy and therefore not suitable for deep-rooted 192 Rhodora [Vol. 94 plants. Margins are usually rocky, but may have peat or small areas of sphagnum bog along the edges. Conditions of clear and generally shallow water are conducive to growth of rosette-leaved plants. South Pond in Savoy State Forest has an excellent ex- amples of this community. Acidic Brownwater Lake/Pond Community Like the preceding habitat, this community is found at the higher elevations in regions of acidic bedrock on the Berkshire Plateau. It is usually associated with bog habitats, or boggy areas that may have been altered or flooded. The vegetation includes more species of the floating-leaved form, such as Nuphar variegata, Nymphaea odorata, Brasenia shreberi, Sparganium fluctuans, S. angustifolium and the rare Nuphar pumila*. Utricularia radiata is quite common on the Plateau along with U. vulgaris, a widespread species present 1n a variety of habitats. Utricularia purpurea may occur in both the Clearwater and Brownwater habitats, while U. geminiscapa 1s more restricted to bog waters. Potamogeton epihydrus vat. ra- mosus is fairly common in this habitat. Eriocaulon pellucidum and Sagittaria graminea occur in shallow waters at the margins. Ecological factors here are similar to those of clearwater hab- itats, but depth of water, light penetration, and substrate differ Alkalinity levels may range up to 30 mg/liter HCO, (Hellquist, 1980). If the basin of the water body is deep, organic matter will accumulate and provide suitable substrate for deep-rooted aquatics. The water becomes discolored with accumulated dis- solved organic acids and fine particles of organic matter, which reduce the amount of sunlight available for growth of plants ofa submersed growth form (McVaugh, 1958). Lack of drainage also tributes to in d organi lation. Bog Pond in Savoy State Forest and the Spectacle Ponds in Sandisfield State Forest are good examples of this aquatic community. Moderately Alkaline Lake/Pond Community The aquatic community of this and the following type are found in regions with calcareous bedrock in the Central Valley Region and occur at elevational range of 221-391 m. Potamogeton species that more often occur within an alkalinity 1992} Weatherbee and Crow— Plant Communities 193 range of 18-73 mg/liter HCO, include Potamogeton spirillus, P. epithydrus var. ramosus and P. robbinsii, which also inhabit less alkaline waters (Hellquist, 1980). Potamogeton amplifolius, P. gramineus, P. natans, P. praelongus, P. obtusifolius, P. zosteri- formis and P. epihydrus var. epihydrus are found in progressively more alkaline waters. The pH of these waters ranges from 7.0 to 9.0. Other aquatics that occur in moderately alkaline waters are Cerotophyllum demersum, Najas flexilis, Elodea canadensis, and occasionally E. nuttallii, Vallisneria americana and Ranunculus longirostris. Lemna trisulca is found more often in alkaline sit- uations, while L. minor and Spirodela polyrhiza are common and widespread in most water. Nuphar variegatum and Nymphaea odorata are commonly found here also, Myriophyllum sibiricum (= M. exalbescens) is occasional, and the adventive M. spicatum can become a troublesome weed. The higher alkalinity of the water, which derives from streams and springs flowing through calcareous bedrock, is the most im- portant factor in this community; edaphic factors are important also. Situated mainly in the valleys, where till, outwash or alluvial soil is abundant and more or less alkaline, ponds and lakes have accumulated thick layers of muck on the bottom. Examples of this habitat are Lake Garfield, which is less alkaline. Lakes Pon- toosuc and Onota in Pittsfield, and Prospect Lake in Egremont have higher, yet moderate alkalinity levels. Highly Alkaline Lake/Pond Community This rich and diverse aquatic community is similar in most respects to the preceding one and shares most of the vegetational elements. The alkalinity levels range from 73 to more than 109 mg/liter HCO, (Hellquist, 1980). The pH may range up to 10.7. This community is distinguished by several Potamogeton species that are more typically found in these waters. Potamogeton pusil- lus var. pusillus, P. foliosus and P. illinoensis are common, while P. friesii* is very rare. The alien Potamogeton crispus inhabits nutrient-enriched, sometimes polluted waters. Potamogeton no- dosus may be present, but is more often found in quiet sections of rivers and streams. Potamogeton pectinatus 1s restricted to the most alkaline waters, those with levels above 109 mg/liter HCO; (Hellquist, 1980). The same is true of P. Aillii*, which has the greatest concentration of populations, world-wide, in Berkshire 194 Rhodora [Vol. 94 County (Hellquist, 1984), and is frequent locally in cool waters of beaver ponds and small streams. Heteranthera dubia is com- mon in many highly alkaline sites while Megalodonta beckii, which is fairly uncommon, also occurs in this habitat, at least in Berk- shire County. Often plants in these waters are encrusted with marly deposits. Good examples of highly alkaline water bodies are Shaker Mill and Cranberry Ponds in West Stockbridge, Lake Buel in Monterey and Mill Pond in Egremont. RIVERS, STREAMS AND SPRINGS Rivers and streams may be characterized as high-, medium- or low-gradient, depending on steepness of slope; most have stretch- es of both fast and slow-moving waters. Rapid sections are often at the headwaters, which originate on the slopes of mountains and on the Berkshire Plateau. In the valleys, streams typically meander, and are bordered by wide floodplains. Vegetation of both the waters and immediate banks is discussed here. High-gradient Stream Community There is no distinctive vascular aquatic vegetation in these streams. Usually the riverbank vegetation reflects the surrounding forest type, but may tend to have more mesic species. Carex torta is commonly found among rocks along the shore. A few elements ofa flood-scoured bedrock community, as described by Rawinski (1986), occur along the Deerfield River, such as Sanguisorba can- adensis, Aster johannensis, Andropogon gerardii, Rosa blanda and Trisetum triflorum ssp. molle. Medium-gradient Stream Community These rocky or gravelly streams, which are typically quite shal- low, also contain few — vascular aquatic plants. Callitri- aes ites in the shallow, quieter waters of these streams, and a few floodplain tree species, such as Plat- anus occidentalis and Salix nigra, may occur on streambanks. Lobelia cardinalis, Helenium autumnale and Rudbeckia laciniata seem to be generally restricted to streambanks, while numerous gravel bars support a few weedy species. 1992] Weatherbee and Crow— Plant Communities 195 Low-gradient Stream Community The aquatic flora of deep, slow-moving waters, with alluvial banks and silty bottoms, is similar to that of the Moderately and Highly Alkaline Lake/Pond Communities, but fewer species are present. The Housatonic watershed has the highest alkalinity of any in New England (Hellquist, 1980). The Hoosic River and its tributaries, which are part of the Hudson River watershed, has the third highest alkalinity in New England. Potamogeton epihy- drus var. ramosus and P. amplifolius inhabit less alkaline streams, while P. nodosus, P. hillii* and P. perfoliatus occur in highly alkaline streams. The introduced P. crispus is common in more polluted streams. Ceratophyllum demersum and Elodea cana- densis occur in shallow backwaters, while 4 ‘yriophyllum spicatum tends to be found in deeper water. Shallow, muddy waters may be colonized by Sparganium americanum and S. emersum, both of which often exhibit long, ribbon-like floating leaves in faster moving currents, while becoming emergent and reproductive nearby on the stream margin. The silt banks and bars associated with this stream type host a distinctive pioneer community limited to annuals, due to the instability of these bars. Commonly found species are Lindernia dubia, Bidens cernua, Xanthium strumarium, Eragrostis pectina- cea, E. hypnoides, Rorippa palustris and Cyperus aristatus; less common are Gratiola neglecta, Veronica anagallis-aquatica, and the rare grass Eragrostis frankii*. Spring and Spring Run Community This community is found throughout the county wherever wa- ter wells up as springs and forms small streams. Shaded springs have little vegetation. However, the adventive Nasturtium offic- inale, and native species such as Veronica americana, Chrysosple- nium americanum and Mentha arvensis, are typically abundant in springs of sunny locations and along spring-fed streams. WETLANDS There is a great variety of wetlands in Berkshire County, ranging from forested swamps to open graminoid wet meadows and level bogs. Many elements of these communities have a considerable 196 Rhodora [Vol. 94 latitude of occurrence. A community may contain elements of both acidic and calcareous iations; one community may grade into another as a series of vegetation zones. For example, a For- ested Fen may surround a Lake Basin Graminoid Fen, with Shrub Fen elements extending into both communities. Or, very com- monly, there may be a mosaic of communities occurring in small patches throughout a wetland area. Important in influencing the occurrence of these communities is the degree of acidity or alkalinity of the water, which is influ- enced by its source, the underlying bedrock, and the soil type. Soil depth and amount of accumulated organic matter also sig- nificantly influences the type of vegetation. Drainage through the wetland, whether very little, as in bogs, or considerable, as in swamps, marshes and fens, is an important factor affecting the amount of organic accumulation and nutrient supply (Dansereau and Segadas-Vianna, 1952). In the following discussion, swamps include both forested wet- lands and shrub swamps; marshes are open wetlands dominated by emergent or graminoid species; bogs and fens are treated sep- arately. SWAMPS Acidic Conifer Swamp Community These swamps occur on the Berkshire Plateau (420-600 m elevation) on flats adjacent to bogs or streams. The dominant species are Abies balsamea and Picea rubens, or Picea mariana; Tsuga canadensis is also common. The ground is typically cov- ered with a deep layer of Sphagnum spp. and other mosses, shrubs and herbaceous plants are sparse. Among the shrub species are Nemopanthus mucronata and Ilex laevigata. The small evergreen groundcover taxa include Gaultheria hispidula, Linnaea borealis, and Coptis trifolia. Cornus canadensis, Carex disperma, C. echi- nata, C. folliculata and C. trisperma are common on the mossy mat; Smilacina trifolia is occasional. Although the soil surface is peaty, often there is little accu- mulation, and the mineral soil layer is typically thin and rocky, unless the forest has developed over the margin of an old bog. Pools of standing water occur between roots and in low places. Poor drainage, coupled with acidic bedrock, are factors that con- 1992] Weatherbee and Crow— Plant Communities 197 tribute to the low nutrient condition of the soil. Wolf Swamp in New Marlboro is an example ofa closed bog with a well-developed Acidic Conifer Swamp along the edge. Acidic Hardwood Swamp Community This community, often referred to as the Red Maple Swamp Community, is frequent in the county at lower elevations in poor- ly-drained acidic sites. It is dominated by Acer rubrum, which although abundant, does not typically develop into large trees as in mesic terrestrial sites. Other tree species include Ulmus amer- icana, and usually Pinus strobus on the drier margins. Quercus bicolor and Nyssa sylvatica may occur in this community, but only in the extreme southern part of the county. The forest canopy is somewhat open, and a variety of shrubs and herbaceous species are associated with this community. Typ- ical shrubs include lex verticillata, Viburnum lentago, V. cassi- noides and Lyonia ligustrina, and occasionally Nemopanthus mucronata. Salix discolor and S. sericea may occur in openings. Common herbaceous species include Osmunda regalis, O. cin- namomea, Dryopteris cristata, Onoclea sensibilis and Carex stric- ta. Important factors influencing this community are acidic water, derived from acidic substrate, poor drainage, and nutrient-poor soils. Few tree species can tolerate both high water levels and low nutrient conditions. Tree species, such as Acer rubrum and Pinus strobus, which need sunny openings to regenerate, are well-adapt- ed to this habitat and find little competition here. Although quite acidic, there is no accumulation of peat in this habitat. Circumneutral Hardwood Swamp Community Given the extensive calcareous deposits in Berkshire County, this community and other more calcareous ones are common throughout the county, except on the Berkshire Plateau. These swamps may occur along rivers or streams or at headwater sources of a stream. The dominant trees are Fraxinus nigra and Acer rubrum, and occasionally Quercus bicolor in the extreme southern portion of the county. The understory tree Carpinus caroliniana is common while Lindera benzoin is usually the dominant shrub. Of the 198 Rhodora [Vol. 94 herbaceous plants, the most typical are Symplocarpus foetidus, Saxifraga pensylvanica, Veratrum viride, Solidago patula, Equi- setum Sylvaticum, Osmunda cinnamomea, Carex bromoides, Ru- bus pubescens and Platanthera psycodes. Important factors influencing the community are rate of flow and pH of the water. In this community, ground water approxi- mating a neutral pH seeps to the surface and flows slowly toward a stream, constantly supplying nutrients to the plants and creating a seepage swamp (McVaugh, 1958). Levels of alkalinity higher than preceding wetland communities facilitate nutrient uptake and decomposition of organic material. Acidic Shrub Swamp Community This community is widespread in regions where waters and soils are acidic, thus is more extensive on the Berkshire Plateau; it occurs on margins of streams and in poorly-drained basins. The dominant species is usually A/nus incana ssp. rugosa, but a num- ber of other shrubs occur, including //ex verticillata, Spiraea lati- folia, Vaccinium corymbosum, Cephalanthus occidentalis, Myrica gale, Lyonia ligustrina, Viburnum cassinoides and V. recognitum. A few tree species such as saplings of Acer rubrum, or, if on the Berkshire Plateau, Abies balsamea or Picea rubens, are usually present, as well as the small tree Amelanchier arborea and the vine Clematis virginiana. eg species include Onoclea sensibilis, Osmunda cinnamo , O. regalis, Glyceria striata, Carex gynandra and C. stricta. Additionally, Aster puniceus and Eupatorium maculatum occur in openings. This habitat is similar to that of the Acidic Hardwood Swamp, shrub swamps may be transitional to a forested type. Flooding or water level fluctuation may retard or reverse the tendency toward development of a swamp forest. Circumneutral Shrub Swamp Community This community typically occurs on the low margins of streams or ponds, where the alkalinity of the water is nearly neutral. It is widespread in the county, but is more common in the valley region. Dominant shrubs are Cornus amomum, C. stolonifera, Salix discolor, S. sericea and S. eriocephala. Lindera benzoin is important as a good indicator of less acid conditions, and tends 1992] Weatherbee and Crow— Plant Communities 199 to be more abundant when slightly shaded, unlike the previously- named shrubs. Rosa palustris grows thickly along marsh edges, Alnus incana ssp. rugosa is present, but plants tend to be scattered, and the vines Clematis virginiana and Solanum dulcamara can be abundant. Among the herbaceous species, Symplocarpus foe- tidus, Saxifraga pensylvanica and Solidago patula are most dis- tinctive of the community. Other common species are Osmunda cinnamomea, Dryopteris cristata, Viola obliqua, Caltha palustris, Ranunculus hispidus var. caricetorum, Carex lacustris and Gly- ceria grandis. Tree species, such as U/mus americana and Acer rubrum may also occur. As in other wetland communities, alkalinity of water and to- pography are the main factors affecting the composition of this community. The habitat is similar to that of Circumneutral Hard- wood Swamp. This community may be successional, as described by McVaugh (1958), but extremely wet or flooded conditions, such as those caused by beavers, may preclude extensive tree growth, or even kill existing trees. MARSHES The vegetation of marshes is characterized by tall, emergent graminoid species and some broad-leaved species growing either in water or in saturated soil. There is adequate movement and drainage of water to supply nutrients and remove organic acids. The soil is a mixture of organic and mineral sediments (Dansereau and Segadas-Vianna, 1952). Robust Emergent Marsh Community This community is found in shallow water along streams, rivers, ponds and lakes throughout the county. It is dominated by tall emergents that are rooted in a thick layer of organic muck or deeper mineral soil. Waters and soil may vary from acidic to fairly alkaline, with the most common condition being inter- mediate, making it difficult to delineate two separate communi- ties. Typha latifolia is most common and typically present in most marshes. Typha angustifolia occurs in more alkaline sites or along roadsides experiencing significant salt runoff from winter road Salting. Other common species more often seen in acidic marshes 200 Rhodora [Vol. 94 are Scirpus cyperinus, S. pungens, Sparganium americanum, Eleocharis smallii, Spiraea latifolia, S. tomentosa and Aster puni- ceus. With increasing alkalinity, Scirpus tabernaemontanii (= S. validus), S. acutus, Sagittaria latifolia, Pontederia cordata, Pel- tandra virginica, Alisma subcordatum, Equisetum fluviatile, Po- tentilla palustris, Sium suave, Cicuta bulbifera, Carex comosa, C. lacustris, Asclepias incarnata, Iris versicolor and Lysimachia ter- restris will be observed, along with the preceding species. Acorus calamus and Sparganium eurycarpum occur in the more alkaline habitats, as does Sagittaria cuneata*, which is found only in occasionally flooded oxbow ponds. Depth and flow of water, and to a lesser extent the acidity or alkalinity of the water and substrate, as well as depth and texture ofthe trate, are import tant thi community. Maximum depth of water for Scirpus cyperinus is approximately 50 cm (Kadlec, 1958), for Typha latifolia, approximately 60 cm (Grace and Wetzel, 1981). Higher alkalinity increases organic and nutrient ilability. A slow water-flow allows accumulation of mineral and organic soil favoring growth of rhi- zomes and roots of these and other species associated with this community. Acidic Graminoid Marsh Community This community, associated with rivers, streams and ponds, becomes established in very moist to saturated soil that is usually above water level in the latter part of the growing season. It occurs more often in the poorly-drained areas of the Berkshire Plateau, where broad, shallow margins of acidic ponds and streams pro- vide much of this habitat type. The vegetation is dominated by a variety of grasses and sedges, the most typical being Calamagrostis canadensis, Glyceria can- adensis, Carex stricta, C. folliculata, C. lurida, Dulichium arun- dinaceum, Scirpus cyperinus and S. atrocinctus. Other common species are Triadenum virginicum, Juncus canadensis, J. brevi- caudatus and Eupatorium perfoliatum. Smaller species that may exist in openings are Eleocharis obtusa, Hypericum boreale, Ly- copus uniflorus and Gratiola neglecta. Flat stream valleys with acidic bedrock, acidic waters and in- fertile soils contribute to the formation of this vegetation type. 1992] Weatherbee and Crow— Plant Communities 201 Many of these marshes develop from silted-in, abandoned beaver ponds. Circumneutral Graminoid Marsh Community This community, more often found along streams and rivers in the valley region where substrate and water tend to be more alkaline, is more diverse than that of the previous community, with a greater variety of sedges and forbs. The common grasses include Calamagrostis canadensis, Phal- aris arundinacea, Glyceria grandis, G. borealis and Leersia ory- zoides. Common sedges are Carex lacustris, C. stipata, C. vul- pinoidea and C. comosa, as well as the rarer C. trichodcarpa*, common only in Hoosic River swales. Eleocharis intermedia* may occur on open mud margins. /ris versicolor and Angelica atropurpurea are common in this, and many other, wetland types. Acorus calamus and Juncus nodosus occur in nutrient-rich sites. Many smaller forbs occur between the graminoids or at marsh edges, such as Thelypteris palustris, Boehmeria cylindrica, Scu- tellaria galericulata, Campanula aparinoides, Penthorum se- doides, Verbena hastata, Mimulus ringens, Lysimachia thyrsi- flora, Asclepias incarnata and Solidago uliginosa. Water and soil of circumneutral alkalinity, topography and de- velopment of thick swamp muck, are important factors influenc- ing the distribution of the community. BOGS Level Bog Community The acidic bog is a well-defined, often-studied community dis- tinguished by a floating Sphagnum mat and dominated by eri- caceous shrubs. This ombrotrophic peatland is most often found on the Berkshire Plateau in poorly-drained basins associated with acidic bedrock at elevations averaging 520 m. At least one bog occurs in a sandy outwash underlain by clay at an elevation of 213 m in Sheffield (Oltsch, 1974). The vegetation may occur in a regular series of zones from an open water pond in the center to a forested zone at the edge of the bog (Crow, 1969). When lacking an extensive Sphagnum mat, elements of these zones may be clustered in an irregular pattern 202 Rhodora [Vol. 94 around the outer edge. The species composition of these zones varies from one bog to another, but certain species are almost always present. Chamaedaphne calyculata and Decodon verticil- lata are common at the inner edge along the pond edge, their intertwined stems and roots forming a base for the beginning of a Sphagnum mat. If there is a broad Sphagnum mat, sedges such as Carex limosa, C. paupercula, C. canescens and C. lasiocarpa occur along with Eriophorum virginicum, E. tenellum, E. vagina- tum, Cladium mariscoides and Scheuchzeria palustris*. Other species common on the mat are Vaccinium macrocarpon, V. oxy- coccos, Sarracenia purpurea, Rhynchospora alba and Drosera ro- tundifolia. Pogonia ophioglossoides is fairly common, while Cal- opogon tuberosus and Platanthera blephariglottis are less so. In open, peaty, shallow pools, Xyris montana, Drosera intermedia, Utricularia cornuta, and occasionally U. gibba occur. Picea mar- iana and Larix laricina occur as scattered, stunted trees, and in small clumps on the mat, but become denser and taller at the outer bog edge. The bog heaths, Andromeda glaucophylla, Kalmia polifolia and Ledum groenlandicum may form a large shrub zone nearer to the outer edge along with Myrica gale. Calla palustris commonly grows in Sphagnum among the heaths, while Smila- cina trifolia occurs occasionally. There may be an abrupt change to upland woods, or bog shrubs may give way gradually to swampy coniferous woods; a moat may also occur at the outer edge. Several factors affect formation of the typical level bog, in- cluding acidity, nutrients, water flow, basin depth, and climate. Ombrotrophic bogs have little inflow or outflow, and are depen- dent on rainfall for the few nutrients that enter the system (Hein- selman, 1963). Berkshire bogs range from those with absolutely no inflow to those fed by small streams from a limited watershed. Lack of drainage leads to accumulated organic acids that retard decomposition of organic material (Dansereau and Segadas-Vian- na, 1952). Sphagnum moss actively produces acidic conditions (Andrus, 1980), and strongly influences the vegetation patterns. Characteristic of cold basins, temperature measurements taken in a Massachusetts bog (542 m elevation) by Moizuk and Liv- ingston (1966) showed that the bog mat experienced approxi- mately one-third the number of frost-free days as was recorded in the surrounding forest. They also determined that nutrient deficiencies hindered Acer rubrum from invading the mat. Those conditions, and the instability of the bog mat, may retard or 1992] Weatherbee and Crow— Plant Communities 203 prevent succession to a closed bog forest. Larsen (1982) indicates that succession may take an irregular course, which can be re- versed, or can diverge toward a variety of associations. FENS Fens differ from bogs in several ways. Cold water, seeping up through calcareous bedrock, becomes highly alkaline and moves Slowly through the community toward a small stream (Schwint- zer, 1978, 1981). High levels of minerals are constantly supplied. A floating mat may develop, but it will be dominated by sedges and grasses. Sphagnum is typically absent, but may appear as small patches and hummocks, accompanied by other bog vege- tation. There are a variety of fen community types, ranging from those that are peatland communities to those lacking peat, and they may exist as a mosaic or in definite vegetation zones. The species composition of fens is rich and diverse, containing many plants rare for the state. A preponderance of fen communities in Massachusetts is found in southern Berkshire County, because of its extensive calcareous bedrock and more level terrain. Forested Fen Community Influenced by highly calcareous water, this peatland community is found more frequently in the southern half of the county, and is usually associated with other fen communities. These fens tend to occupy edges of level basins or margins of extensive, flat areas along gently flowing streams in open valley bottoms. . Species composition of the tree layers shows considerable vari- ation from site to site, there being both hardwood and conifer elements that are important. Larix /aricina is the dominant co- nifer, and in the valley region it seems to be an indicator of fen conditions. Fraxinus nigra is always present, and Acer rubrum commonly occurs as well. Quercus macrocarpa* is common in the southern part of the county. In some fens Tsuga canadensis is one of the dominant trees. Thuja occidentalis*, while never common, is very rare at present, possibly because it was thor- oughly harvested in earlier times. In forested fens, trees are typically scattered, thus the canopy is usually open. Fen shrubs such as Salix candida and S. serissima, Rhamnus alnifolia and Potentilla fruticosa occur in open areas. 204 Rhodora [Vol. 94 These species are good indicators of highly calcareous situations. Ribes triste* occurs in shaded wet areas. Shrubs typical of acidic bogs such as Andromeda glaucophylla and Ledum groenlandicum occur only on Sphagnum hummocks raised above, and isolated from, the calcareous water. In muddy, bare openings, small her- baceous species such as Lobelia kalmii, and rarities such as Ma- laxis brachypoda* and Rhynchospora capillacea* occur. Other species associated with Forested Fens are Cypripedium calceolus var. parviflorum*, C. reginae*, Pyrola asarifolia*, Conioselinum chinense* and very rarely Petasites frigidus var. palmatus*. The presence of a number of northern species found in fens may be due to the cold groundwater, which has provided a stable, cool habitat. Fens along Schenob Brook in Sheffield are among the most extensive and significant. Shrub Fen Community This community is found in association with other peatland fen communities, either intermixed or as a zone at the edge. The dominant and characteristic shrubs are Salix candida, S. seris- sima* and Potentilla fruticosa, an important indicator plant of any calcareous situation, wet or dry. The small shrub Rhamnus alnifolia is also a reliable indicator of fen conditions; other com- mon shrubs are Toxicodendron vernix and Ribes hirtellum. Lo- nicera villosa occurs most often in association with fens. The small tree Betula pumila* occurs at the edge of shrub fens and in open graminoid fens, along with small specimens of Larix /aricina. The shrub layer has many openings where herbaceous species such as Geum rivale, Lysimachia thyrsiflora, Galium labradoricum*, Sol- idago patula, Cirsium muticum, Symplocarpus foetidus and Carex interior occur. Carex castanea*, a rare northern species, occurs at edges of shrub fens. Sloping Graminoid Fen Community This peatland community, often referred to as a calcareous wet meadow, occurs in calcareous areas throughout the central valley region of the county, but more often in the southern two-thirds, and is found on the sloping edge of other fen communities. Many of its species, restricted to this community, are rare in the state. —— SSS a = a SS a i 1992] Weatherbee and Crow— Plant Communities 205 Sedges dominate this community, with the common species being Carex flava, C. granularis, C. lanuginosa and C. aurea. Rare species include Carex alopecoidea*, C. chordorrhiza*, C. sterilis* and C. tetanica*. Scirpus pendulus* may be locally abun- dant in these rich meadows. The most common cotton-grass is Eriophorum viridicarinatum; less common is E. gracile*. Muh- lenbergia glomerata is typical of these fens. The rare Cardamine pratensis var. palustris* occurs in small pools between sedge tus- socks. Other indicative herbaceous species are Geum rivale, Gali- um boreale*, Lobelia kalmii, Parnassia glauca, Gentianopsis crin- ita and Solidago purshii. Both Liparis loeselii and Sisyrinchium mucronatum® are restricted to, yet uncommon in, this commu- nity. Usually, small individuals of the shrub fen are also present here, such as Rhamnus alnifolia, Salix candida and S. serissima and saplings of Larix laricina, indicating that without some form of disturbance, shrub fen elements would become more common. Many of these sloping fens are presently maintained as pasture- land or by mowing at various times. Lake Basin Graminoid Fen Community This rare community occurs on extensive, flat, former lake basins that have become filled with peat, and is influenced by highly alkaline water from both ground water and adjacent small streams. Elements of the sloping graminoid fen, the shrub fen and forested fen communities surround the open, flat sedge-domi- nated mat. Bog elements are also present. The open flats adjacent to pond and stream are dominated by Carex aquatilis and C. lasiocarpa, typical of minerotrophic fens (Schwintzer, 1978; Lar- sen, 1982). Scirpus acutus and Cladium mariscoides occur at the pond edge. Typha latifolia occurs scattered throughout the sedge flats. Small, open, muddy sites contain Eriophorum alpinum (= Scirpus hudsonianus) and a rare, northern species, E leocharis pau- ciflora*. Utricularia intermedia occurs in shallow channels. The shrub Myrica gale is also present. Shrubs typical of fens, such as Salix pedicellaris, S. candida and Potentilla fruticosa occur scat- tered, while Betula pumila* is part of the shrub fen that borders the sedge flat. Bog species, such as Calopogon tuberosus, Pogonia ophioglossoides, Sarracenia purpurea, Andromeda glaucophylla, Carex limosa and Vaccinium macrocarpon occur in scattered patches of Sphagnum. The one excellent example of this com- 206 Rhodora [Vol. 94 munity in the county is Kampoosa Fen in Stockbridge, from which this description is drawn. The broad basin topography allows the accumulation of a large expanse of peat, enabling the community to form an extensive and fairly uniform association. The resulting flat topography may not be as well drained as in other fen systems, and areas that are more remote from water flow may develop accumulations of Sphagnum and other acidophilic bog species (Larsen, 1982; Schwintzer, 1981). Calcareous Seep Community This is a small, unusual and distinctive community found throughout the county where cold, calcareous water seeps to the surface and forms small rivulets that, at times, flow just under the soil surface. It usually occurs on slopes in rocky or gravelly soil; no peat accumulates. The vegetation consists mainly of sedges and other low grami- noids and herbaceous species. Parnassia glauca is an indicator plant. Equisetum hyemale or E. variegatum often forms dense, pure stands; E. scirpoides* occurs on cool, moist, mossy surfaces. Various species characteristic of other fen communities are also present here, such as Carex aurea, C. granularis, C. leptalea, Muhlenbergia glomerata, Malaxis brachypoda*, Cypripedium re- ginae*, Spiranthes cernua, S. romanzoffiana*, Lobelia kalmii and Petasites frigidus var. palmatus*. The slope and flowing ground water create an unstable situation where trees seldom persist beyond the sapling stage, therefore perpetuating the sunny openings that can be colonized by small fen species requiring at least some sun. Cypripedium reginae*, for example, may be shaded out in forested fen habitats, while in this community, it may persist indefinitely. SUMMARY The floristic diversity of the county is illustrated by the rec- ognition of 35 major plant communities. Of primary importance influencing the development of these diverse communities are the underlying bedrock and the topographical heterogeneity of the county, climate, local microclimatic conditions, groundwater pH and alkalinity, and soil nutrient levels. Important also is the di- eee 1992] Weatherbee and Crow— Plant Communities 207 versity of the pool of species available to colonize these habitats, and the post-glacial history of plant migrations into New England. ACKNOWLEDGMENTS We thank A. Linn Bogle and T. D. Lee for helpful comments on the manuscript; B. Sorrie and P. Swain provided data and comments on the manuscript. Suggestions of three anonymous reviewers are gratefully acknowledged. LITERATURE CITED Anprus, R. E. 1980. Sphagnaceae (Peat Moss Family) of New York State. New York State Mus. Bull. No. 442. Albany, NY BoRMANN, F. H. AND M. F. BuELL. 1964. Old-age stand of hemlock-northern hardwood forest in central Vermont. Bull. Torrey Bot. Club 91: 451-465. Brooks, R.R. 1987. Serpentine and its Vegetation. Dioscorides Press, Portland, OR CANTLON, J. E. 1953. Vegetation and microclimates on north and south slopes of Cushetunk Mountain, New Jersey. Ecol. Monogr. 23: 241-270. CocsiLt, C. V. 1987. The boreal forests of New England. Wild Flower Notes (New 7 Wild cisaver sceEsy) ae lida Crow, G.E. 19 An uthern Michigan bog. Michigan Bot. 8: re Crow, G. E. AND C. B. HELLquist. 1981. Aquatic vascular plants of New En- gland: part 2. tition ina Sparganiaceae. New Hampshire Agric. Exp. Sta. Bull. 517 1982. Aquatic vascular plants of New England: part 4. Juncaginaceae, Scheuchzeriaceae, Butomaceae, Hydrocharitaceae. New api nag Agric. Exp. Sta. Bull. 520. 1983. Aquatic vascular plants of New England: part 6. Stee Haloragaceae, Hippuridaceae. New Hampshire Agric. Exp. Sta. Bull. 524. AND 985. Aquatic vascular plants of New England: part 8. Lentibulariaceae. New Hampshire Agric. Exp. Sta. Bull. 528. DANSEREAU, P. AND F. SEGADAS-VIANNA. 1952. Ecological study of the peat bogs of eastern North America. Canad. J. Bot. 30: 490-520. Dickinson, C. H. AND G. J. F. PuGH. 1974. Biology of Plant Litter Decom- position, Vol. 2. Academic Press. New York, NY. Dwicut, T. 1822. Travels; in New-England and New-York, III. Timothy Dwight, New face en. CF. EGier, F.E. 1940. Berkshire Plateau vegetation, Massachusetts. Ecol. Monogr. 10: 147-192. Euias, T. S. 1987. The Complete Trees of North America. Gramercy Publ. Co., New York, NY. Fietp, D. D. 1829. A History of the County of Berkshire by Gentlemen of the County. Samuel Bush, Pittsfield, MA. 208 Rhodora [Vol. 94 Grace, J. B. AND R. G. WetzeL. 1981. Habitat partitioning in cattails (Typha): experimental field studies. Amer. Naturalist 118: 463-474 HEINSELMAN, M. L. 1963. Forest sites, bog processes, and i Cig in the glacial Lake Agassiz region, Minnesota. Ecol. Monogr. 33: 327 HE.iquist, C. B. 1980. Correlation of alkalinity and the poche Pota- 4 : 84. Observations of Potamogeton hillii Morong in North America. Rhodora 86: 101-111. HELLQutst, C. B. AND G. E. Crow. 1980. Aquatic vascular plants of New En- gland: part 1. Zosteraceae, Potamogetonaceae, Zannichelliaceae, Najadaceae. New heel ag Agric. Exp. Sta. Bull. 515. 19 Aquatic vascular plants of New England: part 3. cmt New Hampshire Agric. Agr. Exp. Sta. Bull. 518. AND ———. 1982. Aquatic vascular plants of New England: part 5. Araceae, Lemnaceae, Xyridaceae, Eriocaulaceae, and Pontederiaceae. New sie ca Agric. Exp. Sta. Bull. 523. 1984. Aquatic vascular plants of New England: part ak rakes N and Ceratophyllaceae. New Hampshire Agric. Exp. Sta. Bull. 527. HENDERSON, R. 1982. Vegetation-fire ecology of Tallgrass prairie. Nat. Areas J. 2: 17-26. HoFFMANN, R. 1922. Flora of Berkshire County. Proc. Boston Soc. Nat. Hist. 36: 171-382. Kaptec, J. A. 1958. An analysis of a woolgrass (Scirpus cyperinus) community in Wisconsin. Ecology 39: 327-3 Keitu, H. F. 1912. History of Taconic and Mount Washington. Berkshire Couri- er Printers., Great Barrington, M Kicuier, A.W. 1964. Potential Natural Vegetation of the Conterminous United States. Amer. Geogr. Soc. Special Publ. 36. New York. KupisH, M. 1979. Catskill soils and forest history. The Catskill Center for Conservation and Development, Hobart : LarsEN, J. A. 1982. i of the Northern Lowland Bogs and Conifer Forests. Academic Press, New Leak, W. B. 1978. Relationship of species and site index to habitat in the White Mountains of New Hampshire. USDA For. Serv. Res. Pap. NE-397. Littte, S. 1973. Eighteen-year changes in the composition of a stand of Pinus echinata and P. rigida in southern New Jersey. Bull. Torrey Bot. Club 100: 4-102. MARCHAND, P. J. 1987. Life in the Cold. Univ. of New England Press, Hanover, NH McIntosn, R. P. 1959. Presence and cover in pitch pine-oak stands of the Shawangunk Mountains, New York. Ecology 40: 482-485. McVauau, R. 1958. Flora of the Columbia County Area, New York. New York State Mus. a No. 360. Moizuk, G. A. AND R. B. Lrvincston. 1966. Ecology of red maple (Acer rubrum L.) in a Massachusetts upland bog. Ecology 47: 942-950. NicuHots, G. E. 1916. The vegetation of Connecticut. V. ea societies along rivers and streams. Bull. Torrey Bot. Club 42: 235-26 1992] Weatherbee and Crow— Plant Communities 209 OxtscH, F. M. 1974. The bogs of Berkshire County. M.S. Thesis, Univ. of Massachusetts, Amherst, OostinG, H. J. AND W. D. Bittincs. 1951. A comparison of virgin spruce-fir forest in the northern and southern Appalachian system. Ecology 32: 84— 103. Rawinski, T. J. 1983. Classification of natural communities in New England. The Nature Conservancy. Photocopy. Boston, MA. 1986. Classification of major natural communities of New England (Draft). The Nature Conservancy, Photocopy, Boston, RESCHKE, C. 1980. Ecological Communities of New ae State. ‘Kew York State Department of Environmental Conservation, Lath Scuwintzer, C. R. 1978. Vegetation and nutrient aia “ee northern Michigan fens. Canad. J. Bot. 56: 3044-3051. 1981. Vegetation and nutrient status of northern Michigan bogs and conifer swamps with a comparison to fens. Canad. J. Bot. 59: 842-853. SEISCHAB, F. K. AND J. M. BERNARD. 1991. Pitch pine (Pinus rigida Mill.) communities in central and western New York. Bull. Torrey Bot. Club 118: 412-423. SiccaMA, T. G. 1974. Vegetation, soil and climate on the Green Mountains of Vermont. Ecol. Monogr. 44: 325-3 SorriE, B. A. 1989. Rare Native Plants ae Massachusetts. Massachusetts Di- vision of Fisheries and Wildlife, Boston, MA. TECHNICAL PLANNING ASSOCIATES. 1959. Berkshire —e Massachusetts, A Regional Planning Study. Vol. 2. New Haven WaLKer, R. B. 1954. Factors affecting plant secanvatl on serpentine soils. Ecology 35: 259-266. WEATHERBEE, P.B. 1990. Flora of Berkshire County, Massachusetts. M.S. Thesis, University of New Hampshire, Durham, WESTVELD, M., R. < AsHMAN, H. I. BALDwIN, R. P. Howpswortn, R. S. JOHNSON, . H. Lampert, H. J. Lutz, L. Swan AND M. STANDISH. 1956. Natural Pinwak oak zones of New England. J. Forest: (Washington) 56: 332- 338. Zen, E., Ep. 1983. Bedrock geologic map of Massachusetts. Prepared in coop- eration with the Commonwealth of ao Dept. of Public Works and Joseph A. Sinnott, State Geologist. Shee ZikA, P. F. AND K. T. DANN. 1985. Rare ace on as soils in Vermont. Rhodora 87: 293-304. DEPARTMENT OF PLANT BIOLOGY UNIVERSITY OF NEW HAMPSHIRE DURHAM, NH 03824 RHODORA, Vol. 94, No. 878, pp. 210-211, 1992 NEW ENGLAND NOTE PERSISTENCE OF CAREX CARYOPHYLLEA (CYPERACEAE) IN MASSACHUSETTS L. A. STANDLEY The European sedge, Carex caryophyllea Latourr. (section Mi- tratae Kukenth.), has been introduced into several areas of the northeast. Herbarium records indicate that this species became widely established in Massachusetts (Essex County, Norfolk County, Suffolk County, and Worcester County), with single rec- ords from Maine (Portland), Washington, DC, New York (Dutch- ess County), and New Brunswick (Albert County). The majority of herbarium records date from 1860 to 1905, but the earliest collections appear to be from the early 1800’s in Massachusetts. With one exception (Palmer, 1935) I have seen no records for collections made after 1905 from Massachusetts sites. In conjunction with treatment of the section for The Flora of North America, | attempted to determine whether this introduced species was persistent in Massachusetts by searching all localities for which there was sufficient label information. Ten sites were examined in Essex County, Suffolk County, and Norfolk County; of these, populations appear to have been extirpated in Salem, Westwood, Needham, and West Roxbury, but were re-located at Dedham, Norwood, and Boston localities. No new localities were found during 1991. Two small patches of Carex caryophyllea were located near the Downey School, Downey St., Dedham (Standley 1749), a site for which the previous record was in 1897 (S. K. Harris s.n.). A large population was located near the Xaverian High School on Clap- boardtree St. in Norwood (Standley 1750), for which the previous record was 1884 (T. O. Fuller s.n.). Asmall population was located in the Arnold Arboretum, Jamaica Plain, on Peters Hill in the Sorbus collection (Standley 1761), for which the previous record was 1932 (E. J. Palmer 40185). Vouchers are deposited at NEBC. Carex caryophyllea is widely distributed in Europe in dry, grassy or rocky soils (Chater, 1980), although it is generally restricted to chalk or limestone grassland in Britain (Jermy et al., 1982). Where it persists in Massachusetts, it occurs in habitats similar to i. pensylvanica Lam., including dry roadsides, margins of woods, 210 1992] New England Note 211 old dry pastures, and areas of poor soil in fields from which taller grasses (Dactylis glomerata L., Poa pratensis L.) appear to be excluded. Where it has been extirpated, development or other disturbance appears responsible; where not eliminated by distur- bance, C. caryophyllea has been able to persist successfully for over 100 years, but has not colonized new sites. This finding Suggests that reproduction is primarily vegetative, and that despite abundant flowering and seed set, this introduced species has not become weedy. LITERATURE CITED Cuater, A. O. 1980. Carex, pp. 290-323. In: T. G. Tutin et al., Eds., Flora Europaea, Vol. 5. Cambridge University Press, Cambridge Jermy, A. C., A. O. CHATER AND R. W. Davin. 1982. Sedges of the British Isles. Bot. Soc. Br. Isles, London. PALER, E. J. 1935. Supplement to the spontaneous flora of the Arnold Arbo- retum. J. Arnold Arbor. 16: 81-98. VHB, INC. P.O. BOX 9151 101 WALNUT ST. WATERTOWN, MA 02272 THE NEW ENGLAND BOTANICAL CLUB 22 Divinity Avenue Cambridge, MA 02138 The New England Botanical Club is a non-profit organization that promotes the study of plants of North America, especially the flora of New England and adjacent areas. The Club holds regular meetings, has a large herbarium of New England plants, anda library. It publishes a quarterly journal, RHODORA, which is now in its 94th year and contains about 400 pages a volume. Membership is open to all persons interested in systematics and field botany. Annual dues are $35.00, including a subscription to RHODORA. Members living within about 200 miles of Boston receive notices of the Club meetings. To join, please fill out this membership application and send with enclosed dues to the above address. Regular Member $35.00 Family Rate $45.00 For this calendar year For the next calendar year —— Name Address City & State Zip Special interests (optional): Vol. 94, No. 877, including pages 1-109, was issued March 9, 1992. aie EEA Pic tet eye Slt Pat Ree LI PAP ES oe ACRE 300 a ne oe SECS = a Ca ae ae ee ene ally ae eet gee a en : - = : nabs 3 Sap ee TN a a oe ORES CaREES WY eh ee ee ae INFORMATION FOR CONTRIBUTORS TO RHODORA Submission of a manuscript implies it is not being considered for publication simultaneously elsewhere, either in whole or in part Manuscripts should be submitted in triplicate (an original and two xerographic copies) and must be double-spaced (at least 4”) throughout including tables, figure legends, and literature cita- tions. The list of legends for figures and maps should be provided On a separate page. Footnotes should be used sparingly. Do not indicate the style of type through the use of capitals or under- scoring, particularly in the citation of specimens. Names of genera and species may be underlined to indicate italics in discussions. Specimen citations should be selected critically, especially for common species of broad distribution. Systematic revisions and similar papers should be prepared in the format of “A Monograph of the Genus Malvastrum,” S. R. Hill, Rhodora 84:1-83, 159- 264, 317-409, 1982, particularly with reference to indentation of keys and synonyms. Designation of a new taxon should carry a Latin diagnosis (rather than a full Latin description), which sets forth succinctly just how the new taxon is distinguished from its congeners. Papers of a floristic nature should follow, as far as Possible, the format of ““Annotated List of the Ferns and Fern Allies of Arkansas,” W. Carl Taylor and Delzie Demaree, Rho- dora 81: 503-548, 1979. For bibliographic citations, refer to the Botanico-Periodicum-Huntianum (B-P-H, 1968) which provides Standardized abbreviations for journals originating before 1966. All abbreviations in the text should be followed by a period, except those for standard units of d direction ( For standard abbreviations and for guidance in ‘other matters of biological writing style, consult the CBE Style Manual, 5th ed. (original title: Style Manual for Biological Journals). In preparing figures (maps, charts, drawings, photos, etc.) please remember that the printed plate will be 4 x 6 inches; be sure that illustrations are proportioned to reduce correctly, and indicate by blue pencil the intended limits of the figures. (Some “turn-page” figures with brief legends will be 32 x 6 in.) Magnification/reduction values given in text or figure legends should be calculated to reflect the actual printed size. An Abstract and a list of Key Words should be supplied at the beginning of each paper submitted, except for a very short article or note. All pages should be numbered in the upper right-hand corner. Brevity is urged for all submissions. RHODORA April 1992 Vol. 94, No. 878 CONTENTS Floral variation and taxonomy of Limnobium L. C. Richard (Hydrochari- taceae DescHatd TGWEN re Sc a en ee 111 Chloris barbata Sw. *e C, elata = Adem the earlier names for C. inflata Link and C. dandeyana A John T. Kartesz a ‘Kancheepuram N. RINE eae a fk ales 135 Geographical distribution and ecology of Long’s Bulrush, Scirpus longii (Cyperaceae) in a a Nicholas M. Hill and Mats E. Johansson ...........0.000000+ 0000! 141 The flora of limesink depressions in Carolina Beach State Park (North Carolina) Debra J Sion tds Haaren Warr oc no oc eee ee ees 156 The northeastward spread of Microstegium vimineum (Poaceae) into New York and adjacent states David M. Hunt and Robert E. Zaremba ..............--6 0600000 e> 167 Natural unities of Berkshire County, Massachusetts Pamela B. tile amet ee Carell ©. Crow oe eee ete 171 NEW ENGLAND NOTE Persistence . Carex caryophyllea (Cyperaceae) in Massachusetts EE a ga SOR oS NES Go ee nee Oe i 210 Pum I i ober creme pureeueeen fer Comtvetess. 8... inside back cover ee eee OS SOE eS Tee ee Ee a ee eee SROPR Te ee sha: | 4 | OK Hovdova Rug JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Che New England Botanical Club, Inc. 22 Divinity Avenue, Cambridge, Massachusetts 02138 RAODORA NORTON H. NICKERSON, Editor-in-Chief Associate Editors DAVID S. BARRINGTON RICHARD A. FRALICK A. LINN BOGLE GERALD J. GASTONY DAVID E. BOUFFORD C. BARRE HELLQUIST CHRISTOPHER S. CAMPBELL MICHAEL W. LEFOR WILLIAM D. COUNTRYMAN ROBERT T. WILCE GARRETT E. CROW RHODORA (ISSN 0035-4902). Published four times a year (January, April, July, and October) by The New England Botanical Club, 22 Divinity Ave., Cambridge, MA 02138 rae printed by Allen Press, Inc., 1041 New Hampshire St., Lawrence, KS 66044. hag class postage paid at Boston, MA and at additional mailing office RHODORA is a journal of botany devoted primarily to the rs of North America. Scientific papers and notes relating to this area and floristically related areas, and articles concerned with systematic bot- any and cytota my in their broader implications, will be consid- ered. Articles are subjected to peer review. RHODORA assesses page charges. SUBSCRIPTIONS: $45.00 per calendar year, net, postpaid, in funds payable at par in the United States currency at Boston. Remittances payable to RHODORA. Send to Treasurer, ” Divinity Ave., Cam- bridge, MA 02138. a Regular $35; Family $45. Application form printed tres BACK OL AND SINGLE COPIES: Some available; infor- ation and prices will be furnished upon request to the Treasurer. ADDRESS CHANGES: In order to receive the next number of RHO- DORA, changes must be received by the Treasurer prior to the first day of faniary April, July, or October. POSTMASTER: Send address changes to RHODORA, 22 Divinity Ave., Cambridge, MA 02138. pppoe: FOR CONTRIBUTORS: Inside back cover, January il MANUSCRIPTS: Send to: Joan Y. seco Mana: Edito: wet al Pidneen | aCraix oe bari Dept. of Biology, Tufts adverse Medford, MA 02155 Cover Illustration Rhexia virginica L., meadow beauty, is found from Nova Scotia to Geo rgia, but is rare a northern limits ¢ of its range. The only northern outlier of the . one ees 3-vei 1 lea’ ves and geniculate stamens. Original artwork by sith Ewing. Tbodora OURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 94 July 1992 No. 879 SYMPOSIUM PROCEEDINGS NEW ENGLAND PLANT CONSERVATION: THE SCIENTIFIC BASIS FOR EFFECTIVE ACTION March 21, 1992 Bentley College, Waltham, MA presented by THE NEW ENGLAND WILD FLOWER SOCIETY THE NEW ENGLAND BOTANICAL CLUB with additional support from THE SURDNA FOUNDATION Symposium Committee Lisa A. Standley, Chair William E. Brumback W. Donald Hudson Leslie J. Mehrhoff MISSOURI BOTANICAL DEC 9 1992 RHODORA, Vol. 94, No. 879, pp. 215-217, 1992 NEW ENGLAND PLANT CONSERVATION: THE SCIENTIFIC BASIS FOR EFFECTIVE ACTION FOREWORD Lestiz J. MEHRHOFF, President New England Botanical Club and WILLIAM E. BRUMBACK, Conservation Director New England Wild Flower Society On March 21, 1992, the New England Botanical Club and the the New England Wild Flower Society co-sponsored a symposium entitled New England Plant Conservation: The Scientific Basis for Effective Action. This symposium was held at Bentley College in Waltham, Massachusetts, and was attended by 175 botanists, conservationists, and others concerned with the welfare of our native flora from all over the northeast. The four papers from the symposium are presented in this issue of Rhodora. As with any successful symposium, much work went into its preparation. Almost two years ago, the New England Botanical Club asked Lisa Standley to chair a committee to plan this sym- posium, jointly sponsored by the Club and the New England Wild Flower Society. In addition, the committee consisted of William Brumback, W. Donald Hudson and Leslie Mehrhoff. The com- mittee was given a valuable boost when the New England Wild Flower Society hired Frances Clark as Conservation Program Officer. Frances kept us all on track, aided in the correspondence and served as tireless agent for the committee. The symposium initiative was given further impetus by the recently formed New England Plant Conservation Program (NEPCoP), a voluntary collaboration of representatives from over sixty organizations, agencies and universities already working to protect endangered flora throughout the region. NEPCoP ex- amines both in situ and ex situ conservation strategies and en- courages cooperative action to prevent regional extirpation and to promote recovery of endangered species in the wild. The sym- posium committee’s intention was to bring the most up-to-date information to four areas of the decision-making process where 215 216 Rhodora [Vol. 94 science is most needed. This focus is the rationale for the sequence of papers. We begin with those all-too-familiar taxonomic ques- tions that plague most conservationists, proceed to the difficulties of deciding which taxa to protect, address the processes involved in effective habitat management, and finally finish with the “sticky wicket” of ex situ conservation and reintroductions. Unlike many symposia with lots of speakers and little time for interaction between the speakers and those in attendance, from the onset this symposium revolved around the idea that much would be gained from offering ample time for audience partici- pation. Each of the speakers would be part of a panel, with a moderator and two or three panelists to respond; after the re- sponses, the moderator was to take questions for any panelist from the floor. This format facilitated discussion and set the groundwork for further discussions during the breaks which fol- lowed each presentation. While time and expediency prohibited us from including panel- ist’s response and comments from the audience with each pre- sentation, speakers had the option of incorporating these re- sponses or comments into his or her paper. We would like to thank the following formal participants for their input to the symposium: Taxonomic Issues and Rare Plant Protection: Gregory J. Anderson (moderator), David Barrington and Susan von Ottingen Setting Priorities for Regional Plant Protection Programs: Jeanne Anderson (moderator), Anne Hecht, Leslie J. Mehr- hoff and Robert E. Zaremba Habitat Management: W. Donald Hudson (moderator), Peter W. Dunwiddie, Susan C. Gawler and Harry R. Tyler Reintroduction Dilemma: Barbara St. John Vickery (moderator), William E. Brumback and Mary J. Parkin The New England Plant Conservation symposium was sup- ported by the New England Botanical Club, the New England Wild Flower Society, and a grant from the Surdna Foundation. Publication costs were supported by the New England Botanical Club with generous contributions from the speakers and panel participants. 1992] Taxonomic Issues CONNECTICUT GEOLOGICAL AND NATURAL HISTORY SURVEY BOX U-42 THE UNIVERSITY OF CONNECTICUT STORRS, CT 06268 WEE. GARDEN IN THE WOODS 180 HEMENWAY ROAD FRAMINGHAM, MA 01701 RHODORA, Vol. 94, No. 879, pp. 218-242, 1992 Symposium Paper No. 1 TAXONOMIC ISSUES IN RARE SPECIES PROTECTION LisA A. STANDLEY ABSTRACT eg EES ee 2 of PEE en, ES fe *baeab-t | BE REO B 1 Pp p y £ the t for listing and b i i inf ti itical to plant protection. As endangered species laws become stronger (or more restrictive, depending on the point of view), it will become increasingly important that lists “7 endangered species be based on a solid scientific foundation and incorporate net seceuppisoregisaaber Thee factors are also nnperen ona aes basis and efforts. However, 0661 PW td qd jenied T1O-WA durwing Pequos Aqtfiqtsuajap -o1d >? AY spoou osn pur] po -TIQRIA JUaT[I9 SUOIS aed -1091]01duN 1nq ([e419 -X9 ‘S[eNpIAIp -S900NS§ -o1ddeu] SUONUDIUT POOF =— = Pa) NNN = -UT QOO‘OI-I00I 6861 tW cd ad SO t1S-Wd po19a1 soi0e -o1dun 9wos +009¢ 19A0 ‘drysioumo s1ods [R90] Jo ONL 20s a1eAd # UI ‘syenprlAIp ‘Pasa1SIBa1 IUIOG ajdninw = -U QOO'OT-000$ 0661 CW Id Vv SOX cOS-Vd aoueu -O, UBL S]USUID9I3e ([B90]) oI] peo juaWwoseueUr -qnd pue uOoIS ‘$]B914) y9oS — pad} a1eaud se -S99NS dIBO[OIpAH] “SIBOI S19B.11 ISO adniny “NPIAIpUul +007 = 8861 CW Id d SoA 900-f[N snes diysi9umO snqe1s Poasos S]JUJUIWOD sJUusUIWOD yURY UdISag # uol199101g -dO /A9u931(-) /KJUIB1 90udd dus 90ual dus sey quoUI uONdD101g = -INIIO -INIIO -oseury « ponunuoy ‘¢ 31gey [Vol. 94 Rhodora 270 “WBISOIg IdEILIDH [PIMEN AAA PUR BBRILIOHY [RINIEN JO UOISIAIG: WA ‘AJOIUQAU] AlISIZAIG] [RIMIeN Vd ‘weidolg adeWoH [einen FN ‘AJOWUSAU] dBeILIOY [RIMIVN HN “Weis01g Wea jeIMIeN FW “wesso1g saiads posasuvpug pue aseqLiayY [eINIEN WW ‘eseqeieg Alsi9Aiq [RIMIeN LD ‘swesB0ig IWeILAH [eINIeN AVI WO LING , 1b} “Iqey peqinisip ONL a1eauid -un ul suIa}s Aq poumo Ayiied adninw 000°000°I < 1861 aa Cd Vi soyetuied TOO-AMA snieis diysioumMO, smeis POAIOS S]JUZUIWOD s]JUIWWOD yuRYyY UsISag a UuOT]N9101g -qQ /AauadiQ - /AQUa81Q, 20ual aus aoual aus ise] quUoUI uolda1OIg == -ANDIQ -INdIO -odeuryy x PanuNUoy “¢ I3QRL 1992] Buttrick— Habitat Management il cessful in a number of southern states (Pearsall, 1984). Informal, small protection planning meetings have been successful in our northeast region. Protection planning meetings should focus on reviewing and evaluating the status of all occurrences, determining the appro- priate needs and next actions for each occurrence (whether it be monitoring, active management or land protection), and assigning responsibilities for all actions. To be truly effective, these meetings should be held on a regularly scheduled basis so that progress toward accomplishing assigned actions (we should never under- estimate the importance of peer pressure in getting things done) and the effects of those actions can be evaluated. Next, actions should be recorded in a networked database. The Nature Con- servancy’s Biological and Conservation Data system, used by the Natural Heritage Program Network and a number of federal agen- cies, can track actions and responsibilities and link these actions to individual species, populations and tracts of land. PROTECTION PLANNING AT THE POPULATION LEVEL When a decision is made to focus conservation resources on a particular site more information needs to be collected and another decision making process needs to be developed. Figure | outlines the recommended process for protecting and managing rare plant habitat. This process emphasizes the need to understand the dy- namics of the species and its habitat and the processes that directly affect them, the need to carefully develop and articulate conser- vation goals, and the need to develop monitoring programs to track our success at meeting these goals in order to evaluate and modify our ecological and management assumptions. THE ECOLOGICAL MODEL The goal of in situ conservation is to protect the plant popu- lation in the context of the habitat in which it is found. We are not trying to create wildland gardens, where plants are carfully tended and seeds germinated and planted. Wildland gardening is often the result of a lack of information on how the biological System works. A case in point is Peter’s Mountain mallow, Iliamna corei 2i2 Rhodora [Vol. 94 Ecological Model (Data Collection) (Threats Assessment) (Info Gap Analysis) site = | pn el Design 4 v : Research Goal & + | <—¢——_-—-—__——-g P Setting Baseline Data Biological Monitoring \ Collection agement Planning and Implementation Figure 1. Habitat management planning process. (Sherff) Sherff, known from only one site in southwestern Virginia (Williams et al., 1992). When this herbaceous perennial was dis- covered in 1927, there were approximately 50 individuals. In the 1960’s and 1970’s the population precipitously declined, and to- day there are only four individuals left at the site. There has been no recruitment in the population and there are now high levels of flower abortion and low levels of seed production. Management has been essentially wildland gardening focused on maintaining these last four plants, including caging to prevent deer browse, and watering during periods of drought. Seeds collected from the plants were scarified, germinated and grown at Virginia Poly- technic Institute and State University. ; What was lacking in the above scenario was an understanding of how biotic and abiotic factors affected the growth and repro- BRAN Liman 1992] Buttrick— Habitat Management nig duction of the mallow. True in situ conservation (the preservation of viable populations in functioning habitats) requires an under- standing of the life cycle of the species targeted for protection, the relationship between the species and its habitat, and the abi- otic and biotic processes acting on both. It is the description of this dynamic relation that I am calling an ecological model. This model might be simple (and initially it will be) or, as our under- standing of the system develops, complex. It can be qualitative or quantitative, textual or diagrammatic. Whatever form this model initially takes, it requires land managers to put down on paper their knowledge and assumptions on how the system op- erates. Regardless of complexity or completeness, the model must have two traits: it must be descriptive, showing our understanding and assumptions of how the system works, and it must be pre- dictive. To develop the model the land manager needs to collect all available information on the species, the population of interest and the site itself. If appropriate, information on more common but closely related species found in similar habitats can be col- lected to supplement available data on the rare plant (Baskin and Baskin, 1986). Information on the site and the population should emphasize disturbance and management history. A careful as- sessment of the threats listed in Table 1 should be made. Because our knowledge of the ecological dynamics within a system is never complete, the ecological model will never be complete and will only reflect the current state of our knowledge. In the case of I/iamna corei, research initiated in 1986 by The Virginia Department of Agriculture and Consumer Services, the US Fish and Wildlife Service, the Nature Conservancy, Virginia Polytechnic Institute and State University, and recent work by Baskin and Baskin (1990) have found that the species is self- Incompatible and cross pollination among the four plants is prob- ably infrequent, that a large seedbank exists at the population, that seed germination is triggered by fire, and that plant growth is negatively affected by shading and neighboring vegetation (Wil- liams et al., 1992). With this information, a model of how the system works begins to form (Figure 2) and leads us to ask further questions: What was the frequency and timing of fire in this habitat? Are there other unexpressed seedbanks? Management Predictions can be made as to appropriately time prescribed burns 274 Rhodora [Vol. 94 to stimulate the seed bank, and overstory thinning to promote development of individuals. Research is currently underway to determine the fire history of the habitat and a research proposal, developed by Caren Caljouw of the Virginia Division of Natural Heritage, has been funded to determine, on site, the effect of prescribed burning on added seed and the existing seed bank. Development of the ecological model drives the rest of the protection planning process at a site (Figure 1). Specifically, the model should point out areas of weakness in the information base, and thus where research or baseline data are needed. Additionally, the more we know about the threats and needs of specific pop- ulations and their habitats as well as the biotic and abiotic pro- cesses affecting them, the better we are able to design preserves that will allow these processes to function and reduce identified threats. Similarly, this ecological model will guide management activities aimed at maintaining or reintroducing critical processes such as fire, flooding and herbivory, and at eliminating or reducing threats such as exotics, succession and negative land management practices. Storage and management of both the sources of information used to develop the model, and the model itself, need to be care- fully considered. The Biological and Conservation Data system’s source, stewardship and species and community characterization datafiles provide not only a place and format for assembling and documenting this critical information but also a system for dis- seminating much of this information so that everyone interested in the conservation of a particular species can benefit from the accumulated knowledge of others and not end up reinventing the wheel. SITE DESIGN The ecological model should provide much of the information needed to develop a site design. A well-thought-out site design is probably the most critical step in the habitat management plan- ning process; it minimizes the need for extensive interventionist management and maximizes the chances for conservation success. Without a well-designed site plan, fatal flaws in our ability to protect and manage. the critical natural resources occurring at a site might not be detected until after a major investment of con- servation time and money. A complete site design has three com- ponents. The first, which was mentioned at the beginning of this Buttrick— Habitat Management 275 1992] ‘ainsojo Adoueo pue asy 0} uoNeai ul sormeUAp uoNeindod Zulmoys [apow jemdaou0y °z oin3ly ADUBAJBSUOD BIMBN BY) '@91}}O Plaly B1U/BIIA, “SWIM 824849 4Qg Aq peredaig S3NI1930 NOILV1NdOd | | MOT Y3AO9 LS3Y¥04 | NV G33S 10S 39yVv1 SINI1030 LNdLNO NOILONGOYdSY | SIHSINI1d3Y ANVE G33S 10S | ANVE G33as 10s | HDIH Y3AO9 LS3YO4 S3SV3YONI Y3AOD LSIYOS | NOILONGOYd G3I3S/ONINSMO1d | WOYS LNSWLINYOSY | LNADSAN3S NOILVINdOd asiInd gurd asind duld 4AZIS NOILW1NdOd ISHOD VNWVITI HOIH 276 Rhodora [Vol. 94 paper, is the identification and delineation of the land area needing some level of protection or protection consideration to ensure the security and defensibility of the targeted population and its hab- itat, to moderate or eliminate threats, and to maintain or reinstate the natural processes needed to maintain or enhance the species and its habitat. While a single species often directs conservation attention to a particular site, we are usually protecting more than one endangered plant there. Most rare species do not occur in isolation and many occur within rare or exemplary communities. The design must take into account all resources to be protected at the site. Site design must consider the major processes acting on the habitat. It needs to look at the habitat in relation to the larger landscape in which it occurs and evaluate current and potential activities within that landscape that could impact the resources we are trying to protect. In the case of a rare plant population in a coastal plain pond in southeastern Massachusetts, obvious vis- ible threats to the system might include development along pond- shores and inappropriate use such as all-terrain vehicles (ATV’S). But these issues by themselves are not enough to consider in the site design. The maintenance of the coastal plain pond habitat is controlled by the hydrology of the region. A critical protection goal will have to be the maintenance of groundwater quality and quantity, a goal that needs to be addressed in the site design and cannot be resolved by land acquisition. Direction and speed of groundwater flow need to be determined to help evaluate potential threats from septic systems and to help calculate pondshore set- backs. Overall water quantity entering the site needs to be deter- mined to limit quantity of groundwater removal by pumping at nearby municipal wells. The needs of active management, such as prescribed burning, also must be considered in site design. The land area necessary to implement safe and effective prescribed burns and address smoke management issues must be taken into account. Plans for active management can be difficult or impossible to execute with- out a well-designed site. The second component of a complete site design is a protection plan or strategy for the land. All of the above design work should have been accomplished without regard to ownership patterns. Now ownership information must be collected and appropriate 1992] Buttrick— Habitat Management ait levels of protection, from fee ownership to simple landowner contact to zoning considerations, need to be identified. Finally, management issues need to be reviewed and an outline of man- agement needs prepared. This work should all be done before any land protection work begins. The site design, then, should outline the scope and cost of the conservation activity needed at a par- ticular site. SETTING CONSERVATION GOALS A critical step between development of the ecological model and develop t and implementation of a management program is establishment of conservation goals (Figure 1). The need for establishment of goals has been well articulated by Schroeder and Keller (1990). As managers of biological resources, we are con- stantly making decisions. Even the decision to do nothing, often called passive management, is a management plan with real con- sequences to ecological processes. This approach to habitat man- agement planning requires us to state the potential management results that would reinforce our commitment to management strategies, and the results that would lead us to alter our ideas about how stewardship should proceed. Before implementing a management program, land managers must ask themselves what they are trying to accomplish through this management. Setting of these conservation goals needs be done early in the planning Process to avoid misdirected management and wasted resources. Conservation goals also need to be clear, concise and measur- able. Generally, the goal for a rare plant population is to maintain or improve it. Goals stated this way are not useful because “main- tain” and “improve” are undefined and not measurable. Quan- tifiable goals are not easy to develop and often require a level of information which we do not have, or they force managers to state assumptions that are unproven or of which they are unsure. We seldom have complete information concerning species life histories, dynamics of populations and communities, or environ- mental relationships. Additionally, each population is unique, making it difficult to extrapolate from available species infor- mation. In this case, however, uncertainty should not be a bar to action. By articulating our goals we enable our colleagues to ex- amine our operating premises, and to critique them. We also help 278 Rhodora [Vol. 94 assure continuity despite personnel changes. And most impor- tantly, we can design and implement monitoring that will, over time, allow us to refine these goals and our site management plans. Goal setting should be viewed as a dynamic and flexible process and representative of our current understanding of the system. Setting conservation goals requires four steps. The first step is the identification and selection of features or parameters that best indicate the health and status of a population and/or habitat. This goal might reference size of the population or size of the area occupied within a range over time; or the size (age) class distri- bution of a population. Frequency is often used, especially in the West (Smith et al., 1986). For Carex polymorpha, a clonal species, individuals can be counted but density needs to be better defined. Are we talking about density of ramets or genets? Our ecological model shows that as the canopy closes, flowering and seed set decrease (Standley and Dudley, 1991). A conservation goal based on number of ramets would not really represent the health of the population. Possibly an estimate of the number of ramets and percentage of these setting seed would be better indicators of the status of Carex at a site. If canopy cover is a factor, then a goal needs to be set for this parameter also. The second step is to develop a baseline. To set a goal it is necessary to know the current status of the population and habitat in terms of the selected parameters. For parameters that are vari- able or fluctuate over time, such as density in an annual species, baseline data need to be collected over a long enough period to understand natural fluctuations and how they might relate to habitat factors such as drought. In this case, a baseline might reference the average density over a five-year period or the max- imum number of flowering stems within any four-year period. Once the parameters are identified and the baseline established, quantitative goals should be articulated as the third step. For an occurrence of Carex polymorpha, the goal might be a total pop- ulation size of more than 5000 ramets with at least 25% setting seed and a canopy cover of 10-30%. Zaremba (1992) established a conservation goal for a population of the annual Agalinis acuta at a site in New York, of maintaining an average of 1000 plants over a five-year period at 8 to 10 subpopulations. These subpopu- lations should occur in good quality natural communities that are maintained by management that simulates natural processes. 1992] Buttrick— Habitat Management 279 The last step in setting conservation goals is to establish a “‘red flag,” or lower threshold, for the parameters measured that will alert the manager to problems and the need to re-evaluate man- agement strategies or to review conservation goals. This re-eval- uation is an important step because the manager needs to know when to intercede in the management process in order to avoid extirpation of the population. SITE MANAGEMENT PLANS Site management plans drafted as part of the site design process should now be fleshed out. The purpose of the plan is to document our ecological and management assumptions and to spell out the management policies and actions that need to be implemented to accomplish the conservation goals. I am not proposing a rigid format for these plans. What is most important is that they be dynamic, capable of being reviewed and updated on a regular basis. A management plan should never be bound! It might be useful to think of a management plan as a file drawer where information can be added, replaced or updated easily. Informa- tion in this plan can be classified as archival and active. Archival information would include items such as species lists, resource maps (soils, vegetation, etc.), land use history, research results and legal documents, while the active portions of the plan would include the ecological model, goals, policies, management pre- scriptions, monitoring plans and results, lists of next actions or action plans, and budgets. Once again, the format of the plan 1s not critical; documenting our ecological and management as- sumptions and spelling out management policies and actions are. BIOLOGICAL MONITORING The last stage of the management planning process which I wish to address and emphasize because of its importance and role as a feedback mechanism is biological monitoring. Monitoring means many things to many people but virtually all monitoring activities share (or should share) three common traits. First, mon- itoring actions are repeated over time; second, monitoring results are interpreted by comparison to some standard or objective and finally, the results of monitoring are always subject to interpre- 280 Rhodora [Vol. 94 tation. Biological monitoring indicates what is happening but does not answer any hypotheses; it does not tell you why. Taking one’s temperature is a form of monitoring. The standard or goal a person is trying to achieve is 98.6° F; a high temperature indicates something is wrong, but not what. For the past six years The Nature Conservancy has been using and promoting a very narrow definition of monitoring in the context of the management process discussed in this paper. The Conservancy defines biological monitoring as tracking or auditing of species and communities relative to stated conservation goals. Biological monitoring, then, has the explicit objective of describ- ing a species’ status at a point in time, and its trend in status over time. A monitoring program allows land managers to evaluate whether their management decisions, including the decision to leave the system alone and do nothing, are correct (whether con- servation goals are being attained) and similarly whether the status and trend in status of the species indicate that it is time to review the ecological and management assumptions and either revise the management program or revise the conservation goals. This mon- itoring will not tell us why we did not reach our management goals. Similarly, this monitoring is not designed to set manage- ment prescriptions. The distinction is important. I consider the questions of causation and comparisons of management prescrip- tions to be research and to be addressed by a research program. A monitoring program is an ecological audit whose function is simply to identify where things are working and where they are not. All in situ management programs should have a monitoring program to evaluate their success. Palmer (1986) conducted a review of monitoring programs in a number of states. The Nature Conservancy followed up with a review of their own programs. Based on results of these two surveys a number of recommen- dations can be made which will help ensure a successful biological monitoring program and wise use of limited conservation re- sources. 1. Set conservation goals and red flags. Conservation goals need to be developed for each monitoring program. Without measurable goals, biological monitoring will simply detect change, and change in any biological system is a given. With- out a goal it is difficult to determine what changes are pos- 1992] Buttrick— Habitat Management 281 itive and what negative. Finally, without a goal it is difficult to decide what parameters need to be measured, resulting in monitoring programs that attempt to look at everything or look at the wrong parameters. “Red flags” warn the man- ager when it is time to reevaluate management assumptions and initiate new management actions, including research. . Keep methods simple and focused on goals. Methods need to be focused on selected parameters; if goals are clear and well formed, methods tend to fall into place. Ill-conceived goals result in weak methods which become subject to fre- quent change as goals are reinterpreted. Methods need to be as simple as possible to measure status of the key parameters. Methods capable of being carried out by volunteers are the best and increase our capacity. . Document methods. Each monitoring program should have a clear plan that describes the methods and goals. Long- term continuity of the program depends on development of a monitoring plan. . Plan for analysis of data. Too many monitoring programs do not address how the data will be analyzed. Methods of analysis need to be stated up front in the plan to ensure that data collection methods are adequate. . Get monitoring plans reviewed. It is always valuable to get outside review and opinions. Of course, without a mont- toring plan there is nothing to review. Provide for data storage. Storage of raw and processed data needs to be considered in a monitoring plan. Processed data should be stored and managed so that results can be easily accessed, communicated and summarized. NR Ww os wn * The Biological and Conservation Data system (BCD) used by the Natural Heritage Program Network in all New England states Provides for the tracking of the status and trends of species and community occurrences through monitoring datafiles that can be linked to all species and communities in the Heritage Data Net- work. BCD, then, can be used to record every monitoring event (Ecomonitoring Visit Record) and to summarize these events (Ecomonitoring Record) to provide a picture of the overall status and trend in status of the species or community monitored. For- Mats are shown in Figure 3. 282 Rhodora [Vol. 94 ECOMONITORING RECORD A Monitoring Subject: ion: 3. Goals bes Objectiv: anagement 3 Plans: Y/N Monitoring Plan: Y/N Monitoring Level: 1,2,3 Parameter Goal Red Flag 1 io a ze n. 4. Monitoring Procedures Methods: Sampling Frequency: Visit Dates: 1. = 2; | From Ecomonitoring Visits Record n. ees | Reco: nm Short Term Trend: Long Term Trend: Trend In Trend Update: Management Recommendations: Monitoring Recommendations: Location of Raw & Processed Data: ——— ECOMONITORING VISITS RECORD Visit a Visit a Observer Person Hours: Effort aeiccere Parameter Quantitative S 'y Qualitative Note is 2: n. Other Observations: “One record produced per monitoring event Figure 3. Formats of Ecomonitoring and ee Visits records used in the Biological and Conservation Data Syste 1992] Buttrick — Habitat Management 283 SUMMARY AND RECOMMENDATIONS A balanced conservation program with a goal of protecting rare plants needs to address in situ as well as ex situ protection strat- egies. To succeed, an in situ program requires more than setting aside land for conservation purposes; many of the threats faced by plant populations and their habitats require a carefully planned program of habitat management that addresses both the appro- priate uses of the site and the need for active biological manage- ment. Almost 50% of the most ecologically significant sites in New England occur, at least in part, on public land. Since use and management goals of many publicly owned sites do not reflect their ecological significance or specifically address conservation of rare species, my first recommendation is: 1. The New England conservation community should support and promote programs aimed at: a) alerting public land management agencies and land managers of the occurrence and significance of rare plant populations on their holdings, b) raising the level of protection on these lands, and c) pro- viding habitat management information and assistance. Protection and habitat management decisions for specific pop- ulations of rare species are best made after looking at the range and status of the entire species. Each occurrence needs to be evaluated on the basis of its contribution to protection of the species. This evaluation is critically important because of a need to focus limited conservation resources where they will have the greatest effect. My second recommendation is: 2. The New England conservation community should support and promote a range-wide approach to conservation plan- ning for individual plant species by utilizing data and data structures found in and produced by Natural Heritage pro- grams in each New England state. Planning for the conservation of a species requires a coordi- nated and non-redundant effort on the part of many conservation players—members of the conservation and academic communi- ties as well as private and public land managers. My third rec- ommendation is: 3. Following existing models, the New England conservation community should promote use of protection planning 284 Rhodora [Vol. 94 meetings to include, as needed, land managers, conserva- tionists and scientists capable of significantly contributing to protection of a species and its habitat. These meetings should focus on and result in conservation action by iden- tifying needs, assigning responsibility to specific individuals and agencies for next-step type actions and tracking progress toward accomplishing these actions. When a decision is made to focus conservation resources on a particular plant population, a habitat management planning pro- cess needs to be implemented. This process begins with collection of all available information/data on the species, the population of interest, its habitat and the site in general. This information should be summarized into an ecological model or description of the life cycle of the species, dynamics of the species and its habitat, and biotic and abiotic factors affecting them. This model is used to identify land needing protection attention, to develop specific measurable conservation goals, to identify information gaps and research needs, and to guide specific habitat management activ- ities. 4. The New England conservation and academic communities should focus resources on applied research that will con- tribute to our understanding of biotic and abiotic factors affecting endangered species and their habitats. This work should include studies to determine minimum viable pop- ulations and result in development or refinement of ecolog- ical models to advance in situ protection efforts. Development of the above ecological model improves our abil- ity to design sites which are defensible, require minimum man- agement intervention and provide the best chance for long-term protection and viability of the population in the context of its habitat. Site designs should precede any land protection efforts. 5. For all populations of rare plants targeted for in situ con- servation, whether on public or private land, the New En- gland conservation community should promote develop- ment of detailed site designs driven by our understanding of system function, and threats to and processes acting on the population and its habitat. These designs should include maps of lands needing protection attention, land protection 1992] Buttrick— Habitat Management 285 strategies, preliminary habitat management plans and cost estimates. Development of concrete measurable conservation goals that articulate what we are trying to accomplish at a site is necessary, for we can never determine whether our protection and manage- ment efforts are successful unless we can articulate what success is. Site management plans need to be written to document man- agement policies and actions and the assumptions behind them. The last step in the habitat management process is development of biological monitoring programs which will measure our success at meeting the stated conservation goals, and if necessary lead us to reevaluate our ecological and management assumptions as well as the conservation goals themselves. This process leads to my last four recommendations: 6. For each plant population targeted for in situ conservation (including reintroduction efforts), measurable conservation goals should be developed. These goals should be unam- biguously stated and revolve around one or more parameters of the population and habitat that best indicate health. Neg- ative, or “red flag” goals should also be developed which will warn the land manager that management is not working and ecological and management assumptions need to be reevaluated. . . Foreach site, management plans need to be developed which document ecological assumptions (ecological models) that the manager is working with, and spell out management policies and management actions. While no format is pro- posed, these plans should be dynamic, changing as infor- mation becomes available. . Biological monitoring programs need to be developed at each site targeted for in situ protection. These monitoring programs must be documented to ensure long-term conti- nuity and aimed at collecting the minimum amount of data needed to evaluate progress at meeting conservation goals. All of the data driving and derived from the planning process (including the programs, needs, procedures and results of biological monitoring, management and research, species and community dynamics, and results of site designs) should be stored and managed within the Network of Natural Her- itage Programs and their cooperators (e.g., USFWS, USFS, ~— oo © 286 Rhodora [Vol. 94 TNC). These programs have the data structures in place to manage this information and facilitate data exchange and communication. LITERATURE CITED Baskin, J. M. AND C. C. BASKIN. 1986. Some considerations in evaluating and — 2 aaa of rare plants in successional environments. Natural Areas J. 6: 26-30. AND . 1990. Breaking dormancy in seeds of //iamna corei. Unpubl. rept. submitted to Virginia Dept. of Agriculture and Conservation Services, Richmond, VA. Fak, D. A. 1990. The theory of int ies for biological diversity, pp. 5-10. Jn: B. S. Mitchell, C. J. Sheviack and D. J. Leopold, Eds., Ecosystem Management: Rare Species and Significant Habitats. New York State Mus. Bull. 471. Lapin, B. P. 1991. Natural area defensibility in populated areas. Natural Areas Master, L. L. 1991. Assessing threats and setting priorities for conservation. Cons. Biol. 5: 559-563. PALMER, M. E. 1986. A survey of rare plant monitoring: programs, regions and species priority. Natural Areas J. 6: 27-42. PEARSALL, S. 1984. Multi-agency planning for natural areas in Tennessee. Public Administration Review. January/Febru SCHROEDER, R. L. AND M. E. KELier. 1990. Setline objectives—a prerequisite of heey management, pp. 1-4. Jn: B. S. Mitchell, C. J. Sh heviack and D. J. Leopold, Eds., Ecosystem Management: Rare Species and Significant nde New York State Mus. Bull. 471. Smitu, S. D., S. C. BUNTING AND M. HironakA. 1986. Sensitivity of frequency plots for detecting vegetation change. Northw. Sci. 60: 279- 286 STANDLEY, L. A. AND J. L. DupLey. 1991. Vegetative and sexual reproduction in the rare sedge Carex polymorpha. Rhodora 93: 268-290. Wuirte, P. S. AND S. P. BRAaTTON. 1980. After — philosophical and practical problems of change. Biol. Cons. 18: 241 WiiiaMs, C. E., T. F. WieEBoLDT AND D. M. al 1992. Recovery of the cleneseal Peters Mountain mallow, J/iamna corei. Natural Areas J. (in press). wi R. 1992. Unpublished site design for a population of Agalinis acuta n New York. The Nature Conservancy, New York Regional Office, Alba- ny, NY. THE NATURE CONSERVANCY EASTERN REGIONAL OFFICE 201 DEVONSHIRE ST., 5TH FLOOR BOSTON, MA 02110-1402 RHODORA, Vol. 94, No. 879, pp. 287-315, 1992 Symposium Paper No. 4 SCIENTIFIC AND POLICY CONSIDERATIONS IN RESTORATION AND REINTRODUCTION OF ENDANGERED SPECIES DONALD A. FALK AND PEGGY OLWELL “Restoration, hang thy medicine on my lips.” — King Lear, Act IV, Scene 3 Like Cordelia’s impassioned prayer in the closing moments of Shakespeare’s greatest tragedy, the world of conservation is turn- ing increasingly to the healing powers of ecological restoration for salvation. At least three national organizations—Society for Ecological Restoration, Natural Areas Association, and Restoring the Earth—are devoted to promoting intensive habitat manage- ment and ecological restoration as strategies for conserving bio- logical diversity. Federal land-managing agencies as well as the largest private land conservation organizations, notably The Na- ture Conservancy, are engaged increasingly in various forms of reintroduction and restoration practice. Many members of the national Center for Plant Conservation network are playing an active role in replacing extremely rare and threatened species back into natural areas. Even the mainstream media have taken notice (Nash, 1991), as have many corporations whose activities have a major impact on the land and the life it supports. . But is reintroduction always a feasible or appropriate tool with which to perpetuate biological diversity, ecological processes and the grand dance of evolution? Experience reveals that reintro- duction efforts frequently encounter substantial technical obsta- cles and may entail such a high degree of uncertainty that their value as a tool for conservation can be questionable. The biolog- ical value of reintroduction remains largely a matter of informed Speculation, although the evidence supports cautious optimism. And the strategic and political implications of an improved ability to translocate plants are just beginning to be understood, including the chilling possibility that reintroduction will, through its ap- plication in mitigation programs, rationalize and even facilitate the destruction of existing natural areas. . In the face of such uncertainty and peril, why are conserva- onists thinking at all about moving or transplanting rare species? 287 288 Rhodora [Vol. 94 At least three environmental conditions compel consideration of this activity: 1. Continued Destruction of Habitat Despite decades of heroic effort, the continent’s remaining un- disturbed natural areas continue to be strip mined, logged, paved and fragmented. The Nature Conservancy—the most successful and most important private land conservation organization in the history of the country—has managed to protect on the order of 5 million acres, amounting to approximately .22% of the land area of the United States. Many states with the highest indices of plant diversity, including Hawaii, California, Texas, Florida and Puerto Rico, have major areas of habitat (encompassing whole endemic community types) on land that is unprotected and under active commercial development. In the state of Hawaii alone, as many as 95 new golf courses are proposed and awaiting construc- tion permits to consume some of the last remaining level lowland habitat in the islands (Hawaii Office of State Planning, 1991). Wallace (1990) describes an analogously critical context for the sand scrub communities of central peninsular Florida. Pressure is increasing in many states to open public lands for large-scale strip mining for minerals; substantial lands are owned privately by mining corporations with plans for operations covering hun- dreds of thousands of acres. Equally intense pressure exists in tropical regions (e.g., Holden, 1986). In the face of such destruction, rare plants are inevitably the first to suffer. Of the approximately 20,000 plant taxa native to the United States, 4412 (22%) are in the top endangerment cat- egories in the ranking systems of the U.S. Fish and Wildlife Ser- vice, The Nature Conservancy, or the Center for Plant Conser- vation (Center for Plant Conservation 1992). Of these plants, 2768 (about two thirds) are restricted to a single state or province. Hawaii has 547 endemic rare taxa. The California Floristic Prov- ince includes 4450 taxa of plants, of which 2140 (48%) are en- demic to the region (Myers, 1990). On a continental basis, 1183 (39% of the total flora) are known from five or fewer populations or from 1000 or fewer individuals. Not surprisingly, these taxa have the least margin of survival in the face of habitat loss and the greatest vulnerability to extirpation of populations and the loss of evolutionary viability. Even where entire species have not 1992] Falk and Olwell— Considerations 289 been driven to extinction, many species are becoming severely depleted due to loss of genetically distinct populations. As Ehrlich (1991) observes, “extirpation of populations is the dominant el- ement of the extinction crisis in temperate regions today and most severely threatens ecosystem services in these areas.” 2. Habitat Degradation Even where land is formally protected from outright destruc- tion, significant deterioration of habitat can occur and with it the loss of the ability of many plants to adapt and survive. For in- stance, well over two-thirds of the total land area of the United States west of the Rocky Mountains is owned and managed by various agencies of the Federal government—Bureau of Land Management, U.S. Forest Service, National Park Service, De- partment of Defense—as well as the significant acreage reserved for Native American Indians, primarily in the southwest. Many of these lands are severely overgrazed, eroded, logged and overrun with recreation vehicles. Biological invasion by exotic organisms can so thoroughly alter habitat characteristics that the composi- tion of entire communities is altered (Vitousek et al., 1987; Brock- ie et al., 1988: Vermeij, 1991). Millions of acres of Great Basin and northern Great Plains grasslands, for example, have been invaded and altered by exotic grasses such as red brome (Bromus tectorum) and other exotic species, reducing the native perennial bunchgrass communities. Many rare plants are endemic to specific community types, or restricted primarily to specific seral stages. Fiedler (1986; Fiedler and Ahouse, 1992) and others (Kruckeberg and Rabinowitz, 1985) have described a variety of patterns of rarity, including such eco- logical and evolutionary factors as the relationship of edaphic endemism and restricted ecological amplitude to geographic rar- ity. Habitat endemics, a category that includes a large proportion of rare American plants, are often sensitive to alteration of phys- ical or biotic characteristics of site and hence vulnerable to decline even if the site has not been completely or visibly “destroyed.” 3. Large-Scale Climate Change As Peters (1988) has described, current trends of climate change may render many protected areas unsuitable to maintain the spe- 290 Rhodora [Vol. 94 cies that presently occur within their boundaries. Correspond- ingly, as the appropriate climatic zones for species migrate toward the poles, the optimal range for many plants will shift into areas of land that are currently unprotected. Much of this change iS projected to occur at a rate tens or hundreds of times faster than plant populations can “‘migrate” generationally across the land- scape (Schwartz, 1992). These effects will likely be most severe for species that are restricted in their geographic distribution and ecological range. For example, species found in coastal areas, wetlands, montane- alpine and arctic biomes, and microclimatic refugia (such as the Appalachicola River Basin of Florida and southern Georgia) stand to experience drastic reduction or movement of their preferred habitat. The same would be true for the Hawaiian flora, which is more than 95% endemic to the Islands; many taxa are found on only a single island, at specific elevations and aspects to prevailing wind directions. Species occurring in few sites or in small pop- ulations will be the most vulnerable, a category that (as noted above) includes as many as two-thirds of all sensitive species in the United States. If these species are to survive at all, serious and active intervention into their management, including inten- tional alteration of their geographic distribution, will be essential. Combined with the effects of habitat destruction and degradation, many of these plants will not survive without such measures. These three factors, among others, suggest that reintroduction and restoration will play a vital role in conservation strategies in the coming decades (Falk, 1990b; Edwards, 1990; IUCN, 1992; McMahan, 1990). Limited experience suggests that these methods may provide an enormously positive supplement to land protec- tion, especially in helping to preserve individual populations. Such developments, however, do not take place in a political vacuum. Many of the societal forces contributing to the destruc- tion of natural areas are seizing on the ability to translocate OT- ganisms as a way to rationalize or facilitate further destruction of habitat, rather than as a tool to repair damage that has already been done. Continuing threats to the Endangered Species Act, and to the basic legislative provisions for the protection of land in its natural state, will make such misuses of reintroduction even more dangerous. Certain regions of the country, notably southern California, peninsular Florida, and coastal Hawaii, are already 1992] Falk and Olwell— Considerations 291 experiencing such overwhelming population growth and devel- opment pressure that the few regulatory restrictions protecting land seem like a fragile dam about to burst. When this happens, whether in a continuing trickle of incremental loss of habitat or a catastrophic deluge, the ability to reintroduce species will play a pivotal role in the survival of many species. The question will then become: when all of the suitable land has been destroyed, where can they be introduced? REINTRODUCTION AND THE U.S. ENDANGERED SPECIES ACT The Endangered Species Act of 1973, as amended, provides the legislative authority for the use of reintroduction as a con- servation tool in the recovery process for federally listed endan- gered and threatened species (U.S. Fish and Wildlife Service, 1988). However, reintroduction has been controversial from both the economic and scientific perspectives. It was the economic conflicts and concerns that prompted Congress in 1982 to amend the Act to include a special provision for experimental (reintro- duced) populations (Drabelle, 1985). Within Section 10(j) of the Act (U.S. Fish and Wildlife Service, 1988), experimental populations are allowed to further the con- servation of the species. An experimental population may be classified as either “essential to the continued existence of the species” or “nonessential.” If the reintroduced population is deemed “essential,” then the population is treated (for purposes of the Act) as though it is a threatened species. Thus, the popu- lation is awarded all the protection of a threatened species under the Act. However, if it is deemed “nonessential,” then it is treated as though it is a proposed species under the Act, thereby receiving less protection than the essential population. These designations allow the U.S. Fish and Wildlife Service more options in man- aging the reintroduced populations. Since 1982 when these amendments were developed, only sev- €n experimental populations have been listed under the Act; all have been for vertebrate animals (J. Sheppard, pers. comm.). However, plant reintroductions are occurring regularly without formal listing as experimental populations under the Act. Instead, they are being conducted without the legal designation of the Act through multi-agency and/or private cooperative efforts. 292 Rhodora [Vol. 94 Table 1. Recovery plan status of 239 federally listed plants, as of 1990.* Plan status No. (%) Approved 119 (50) Draft 71 (30) None 49 (20) Total 239 (100) * Source: U.S. Fish and Wildlife Service (1990a). In a review of plant recovery plans for their report to World Widlife Fund, Cook and Dixon (1989) found that the predomi- nant threat to rare plants, not surprisingly, is habitat disturbance or destruction. Sixty-six percent of the species they reviewed were threatened by loss of habitat. Thus, site preservation should be the first priority in a recovery plan. However, the authors also indicated that the second priority in developing a recovery pro- gram for rare plants should be “the identification of potential habitat and the development of techniques to establish new pop- ulations.” Obviously, Cook and Dixon view the reintroduction of endangered plants as a viable means of enhancing the conser- vation of these species. There are others who believe reintroduction can be beneficial to the recovery of endangered plant species (see Jones, 1990). For example, the U.S. Fish and Wildlife Service’s 1990 report to Congress revealed that for the 239 listed plants, only 50% (119) had approved recovery plans and 30% (71) had draft recovery plans (Table 1). However, reintroduction was either in progress or planned as part of the recovery efforts for 56 listed plants (Table 2). Additionally, of those 56 species, 42 had reintroduction as one recovery criterion for downlisting or delisting the species. Therefore, almost 25 percent of the federally listed plants, as of 1990, have reintroduction as a tool for the recovery of the species (U.S. Fish and Wildlife Service, 1990a). This number of species with reintroduction as a major focus or effort for recovery is surprisingly large, especially considering that the U.S. Fish and Wildlife Service does not have a specific policy on rare plant reintroductions or guidelines on how to conduct rare plant rein- troductions (U.S. Fish and Wildlife Service, 1990b). Essentially, the Service’s only policy guideline on reintroductions is that they should not occur outside the historic range of the species unless approved by the Director (U.S. Fish and Wildlife Service, 1990b). 1992] Falk and Olwell— Considerations 293 Table 2. Federally listed plant taxa with reintroduction projects, as of Septem- ber, 1990.* Down/ De- Cur- listing Taxon rent! Planned? Criteria’ Amorpha crenulata Amsinckia grandiflora Amsonia kearneyana Arctostaphylos ania ssp. ravenil Asimina tetram 4 mK mK mK Astragalus robbinsii var. jesupi Banara vanderbiltii Betula uber Boltonia decurrens Buxus vahlii x Calyptronoma rivalis x Crescentia portoricensis Cyathea dryopteroides Daphnopsis hellerana icerandra immaculata x Enceliopsis nudicaulis var. corrugata ¥ x eum radiatum Goetzea elegans Grindelia fraxinopratensis Harperocallis flava - Hibiscadelphus distans Hymenoxys acaulis var. glabra x texana x xK aes ata tues Ivesia eremica >> KK OK KKK KM Kokia cookei Lesquerella le latris hell Mahonia pene leucophylla Nitrophila mohavensis Oxypolis canbyi x nicum carteri Pedicularis sal x x ediocactus knowlto Pediocactus aie var. pebblesianus Peperomia wheeleri ~*~ xxx 294 Rhodora [Vol. 94 Table 2. Continued. Down/ ur- listing Taxon rent! Planned? Criteria’ Phacelia argillacea x X Potentilla robbinsiana x x 4 Ranunculus acriformis var. aestivalis > 4 Rhus michauxii x 4 Sarracenia oreophila ».4 xX Sidalcea pe x Solidago spithamaea X Pcahpemnileg malheurensis x Styrax tex xX aden ing oe Xx Torreya taxifo X Trichilia ot nae X Trifolium stoloniferum Xx Zanthoxylum thomasianum xX TOTALS 18 Le oe * Source: U.S. Fish and Wildlife Service (1990a). ' Any reintroduction project occurring in the past few years (1986-1990) or since any previous reports. 2 Any reintroduction project to be continued or initiated during 1991 and 1992. 3 Any reintroduction project that is included in the criteria of an approved or draft recovery plan. SCIENTIFIC POLICY CONSIDERATIONS IN SPECIES REINTRODUCTIONS AND ECOLOGICAL RESTORATION The term “policy” carries as many uses as there are users, but in one form or another generally involves asking, ‘“What should we do? How do we propose to act? What are our goals? (Faludi, 1973). In the present instance, identifying a clear long-term ob- jective for plant reintroductions is of central importance—an ex- ercise of considerable difficulty. For example, do we intend to maintain ail species in their current abundance and distribution, or to reduce some and augment others? Do we wish to return all native species to their pre-Columbian distribution and eliminate all exotic invaders? Is the distribution of species intended to re- main static or to change over time (and if so, in what way)? Is 1t our intention, as population biologist John Harper once asked a conservation meeting, “to make rare plants common”? Do we propose to manage habitat actively, or to allow ecological and 1992] Falk and Olwell— Considerations 295 Table 3. Nomenclature employed in published literature describing activities involving movement of living organisms to and from habitat, shown with the generally associated level of biological organization. Level of Biological Term Organization Relocation, translocation, transplantation Individuals, populations Augmentation, enhancement Populations Introduction, reintroduction Populations, species Revegetation unities Restoration, rehabilitation, reconstruction Communities, ecosystems evolutionary processes to operate without interference (other than the colossal disruption we are already causing by environmental degradation)? As these questions are meant to illustrate, a true societal con- sensus on any of these points is highly unlikely. In fact, even within (and between) the conservation and scientific communities there is considerable disagreement on almost every point. Some conservation biologists oppose any moving of organisms, as a disruption of the natural distribution of genetic (and hence evo- lutionary) variation. Others endorse replacing species in histori- cally documented localities, but only where populations have been recorded previously. Still others encourage placement of threat- ened species on any suitable protected land, as a pragmatic crisis i to prevent further anthropogenic extinction (see IUCN, 987) “Policy” is not made in the abstract; it represents the collective decisions and intentions of any organized group in society, wheth- er a botanical club, conservation group, mining corporation, trade lobbying organization, or the Federal government. Furthermore, Policy is meant to organize behavior in such a way as to promote fulfillment of that group’s objectives and mission. But since the objectives and interests of various groups affecting nature may e fundamentally different (and even mutually exclusive), the Priority accorded to different goals—such as protecting biological diversity as compared to returning a profit to stockholders—will vary dramatically and predictably. Moreover, one cannot pre- sume to make policy for another group; one can only articulate one’s own, and to describe the considerations one believes should be taken into account. Consideration of these questions for biodiversity management 296 Rhodora [Vol. 94 is further complicated by inconsistent and rapidly changing ter- minology. The literature includes references to introduction, re- introduction, relocation, translocation, transplantation, revege- tation, restoration, rehabilitation, augmentation, enhancement, and other terms. Many of these carry specific meanings, while others are used variously or interchangeably. Although the un- tangling of this nomenclature is beyond the scope of this paper, it is useful to note that each term generally refers to a particular level or range of biological organization (Table 3). For purposes of this paper we will employ the terms “reintroduction” and “restoration”’ to refer to the relevant actions for populations of rare plants and their associated communities and ecosystems. Good policy is ultimately dependent upon asking good ques- tions, which are the sine qua non of clarity. For that reason, we attempt here to outline some of the basic questions that should be asked, or considerations taken into account, in preparing a reintroduction program. The questions are categorical; the an- swers will be case-specific. That is the work of planning a good reintroduction effort. As a contribution to this process, the Center for Plant Conser- vation is presently undertaking a national study of reintroduction and mitigation policy with respect to rare plants. This will include a comparative review of the policies of major Federal and State land-managing agencies and private land-owning organizations, and the formulation of proposed guidelines. In assessing reintroduction as a strategy for conserving rare plants, a series of five areas of inquiry emerge. First, we should always be clear why we are undertaking a reintroduction program, and consider carefully what our long-term goal is with respect to the proposed project. Next, we need to consider the consequences of a reintroduction effort, including the possible implications for the protection status of existing populations and habitat. Third, we must assess where the reintroduction takes place and who will be involved in maintaining and managing the population. These considerations have a major bearing on the probable long-term viability and success of the reintroduction effort. Fourth, we need to understand how reintroductions should be conceived and car- ried out from a biological perspective, and the prospect that these actions will contribute to the achievement of a long-term con- servation goal. And finally, we must ask if reintroduction in a 1992] Falk and Olwell— Considerations 297 particular case is even technically possible; if not, the option is eliminated from consideration until further research opens the possibility. These five general areas of consideration may be de- scribed respectively as (i) objective, (ii) strategic, (iii) managerial, (iv) biological, and (v) technical. We propose these factors not only as a conceptual framework for thinking about reintroduction and restoration issues, but also as a decision-making sequence for well-conceived practical pro- grams. Although we discuss them as separate considerations, we acknowledge that they are highly interactive, and frequently dif- ficult to separate. For example, the absence of a pollinator or seed dispersal agent may raise questions both about the technical fea- sibility of a reintroduction project, and about its long-term bio- logical significance. Likewise, the strategic consequences of a re- introduction project for the protection of other natural populations may be deeply influenced by administrative factors such as land Ownership and the commitment of the land-managing agency to Proper ongoing management. Nonetheless, we maintain that rein- troductions should above all be well-conceived and well-planned in order to be successful, cost-effective, and contributory to the Overall objectives of conservation, and that approaching the plan- ning process in the order described will contribute to prospects for success. In our experience, many reintroduction efforts begin with care- ful study of technical feasibility and biological considerations, and only later (or incidentally, or not at all) address the strategic and political consequences of the action. We suggest that this is a Significant error, both tactically and procedurally, for several reasons. First, reintroductions should always be undertaken in relation to a clearly stated long-term objective; otherwise, they are simply stochastic events with little prospect for cumulative effect. Second, careful planning can ensure that the project will be carried out efficiently and cost-effectively. Managerial and stra- tegic considerations have a particularly large bearing on prospects for Success, both for the immediate reintroduction itself and for Protection of other similar populations or communities. And last- ly, technical and biological obstacles are the most amenable to resolution by research, experimentation, or simple trial and error; In fact, their resolution may be built into the implementation sequence for the project itself. 298 Rhodora [Vol. 94 1. Objectives Although it frequently runs counter to our established habitual ways of acting, we submit that thinking about reintroduction programs should start with clarity regarding objectives and stra- tegic implications. Once these overall contextual decisions are made—is reintroduction a good idea? does it meet our long-term objectives?—the more immediate technical and biological con- cerns may be addressed and incorporated into the project design. As Millar and Libby (1989) observe, reintroduction and res- toration projects can recreate viable, complex, sustainable com- munities—or “Disneyland” simulations. The determination of long-term objectives is a non-trivial undertaking for anyone work- ing to protect biodiversity in the face of continued development pressure and economic constraints. The days when conservation- ists could realistically envision large areas of the continent in a wild and natural state are gone, perhaps forever. Instead, we must reorient our efforts to a different set of goals involving the inte- gration of human activities into a matrix of natural biological diversity and functions. As Diamond (1987) notes, even such watchwords as “natural” and “self-sustaining” need to be ex- amined critically and realistically. For many areas of the continent the objective may become to re-establish diversity in a perpetually managed setting—a far cry from self-sustaining, undisturbed sys- tems (Diamond, 1985) Goal-related issues concerning a reintroduction program might include the following: — . Are there clearly articulated long-term goals for conserva- tion of the species or community type? What are these goals? . Is the goal to re-establish “pre-disturbance” distribution and abundance? If so, what point in the species’ or community's natural history do we propose to re-establish? Are distri- bution and abundance stable, trending, or undergoing long- term oscillation (Falk, 1990a)? . What forces influenced the distribution and abundance of the species or community prior to the “disturbance’’? . Is elimination or amelioration of the cause of decline or threat to species part of the project? . Once reintroduction or restoration has taken place, what will happen? Is there a commitment to provide intensive N es) BS A) 1992] Falk and Olwell—Considerations 299 management indefinitely, or eventually to allow natural suc- cessional and ecological forces to prevail? Do we propose to maintain every species and community in perpetuity, or to allow some to change or disappear over time? If the latter, what criteria will we use to make these decisions? - As is evident, some of these questions border on what many practitioners would consider philosophical issues beyond the realm of conservation, or at least beyond their mandate and responsi- bility. Nothing could be further from the truth; unless these basic issues are addressed and at least partly answered, the efforts of individual agencies with particular sites, species, or communities will have no organized relationship to a long-term goal. Answers may be elusive and difficult to pin down, and the criteria seem- ingly impossible to establish, but a reintroduction plan will be greatly strengthened by inclusion of these factors in its reasoning, to the extent that they can be addressed. 2. Strategic Considerations By their very nature, most reintroductions take place in the context of conflicts over land use, enhancing or re-establishing populations destroyed by incompatible uses. Usually at issue 1s the priority accorded protection of biological diversity versus resource extraction, recreation activities, or other extensive mod- ifications of the landscape. While some of these decisions are made internally by a single agency thinking about its own land, such as the multiple-use paradigm of the Bureau of Land Man- agement or the U.S. Forest Service, most involve conflicting man- dates and interests of more than one organization, often in an adversarial relationship. The most serious contextual consideration for the current prac- tice of reintroduction in the United States is the practice of mit- igation. Mitigation almost always involves a trade of some kind: land for land, land for money, land for long-term management support. Typically, a developer will propose to establish “new habitat equivalent to that being destroyed by the project. Vari- ables may include the amount of land involved, financial support, long-term ownership and management responsibility for the cre- ated habitat, quality and success assurances, and the relationship 300 Rhodora [Vol. 94 to large-scale bioregional strategies for maintaining ecological di- versity. Formally speaking, to ““mitigate’’ means to reduce or ameliorate the impact ofan action, to lessen its severity and alleviate or abate its consequences. Many “mitigation” projects are anything but; instead of reducing the impacts of development pressure, in reality they have become a means of justifying increased invasion of protected areas. In certain parts of the United States, regulatory agencies now receive hundreds of requests for mitigation projects involving the destruction or substantial modification of natural habitat, in exchange for the creation of “‘new”’ habitat elsewhere. On a regional basis, mitigation appears to be most heavily prac- ticed in southern California, peninsular Florida, and parts of the southwest and southeast. Specific habitats, however—notably, wetlands and riparian areas—are being heavily affected in all parts of the country. Wetlands habitats are naturally among the most controversial subjects for mitigation proposals. Given their com- plex energetic and nutrient webs, species composition, and hy- drology, wetlands represent extraordinary challenges for habitat creation or even individual reintroductions (Jarman et al., 1991; Munro, 1991). Yet Dobberteen (1989) found over 1000 wetlands creation projects in Massachusetts alone in just seven years be- ginning in 1983! A frequent problem with many mitigation projects is that they are undertaken in the absence of a consensus on long-term goals. For instance, there is considerable evidence that re-created hab- itats may take decades to become fully established; until this has occurred, there is little assurance that the reintroduced population or created habitat will serve the same ecological and evolutionary functions as those that were destroyed. Long-term issues of land protection may remain unresolved, leaving the reintroduced pop- ulation or created site vulnerable to future disruption or even destruction. This is especially true for lands under private or corporate ownership. Most critical, however, is the relationship of mitigation pro- grams to the conservation of existing natural areas. Again, this is fundamentally a matter of agreeing upon objectives. For a de- veloper, the objective may be to build a desired project, and to provide mitigation tradeoffs only to the extent required by law, regulation, political expediency, or a desire to maintain a good public image. Conservationists, on the other hand, may approach | 1992] Falk and Olwell—Considerations 301 mitigation primarily as a strategy to assist the protection of eco- logically valuable or unique communities, species, and popula- tions. The regulatory agency involved may operate strictly by the letter of the law, with no particular vision for the future. The overlap between these three different agendas can prove to be narrow ground indeed on which to build a sound future for hab- itats or threatened species. By and large, standards for the practice of mitigation projects involving endangered species do not presently exist at the national level. Individual agencies have been developing policies incre- mentally, but even the largest Federal agencies manifest consid- erable internal variability in how they approach mitigation pro- posals. In the meantime, individual states, counties and private corporations all develop, evaluate and implement mitigation pro- posals in increasing numbers, without the benefit of either a true consensus on goals, or clear national guidelines. A reasonable starting point for such standards would be that mitigation projects should never entail the destruction of existing significant and irreplaceable natural areas. Reintroduction and restoration should, in other words, be employed to heal damage that has already been caused, not to rationalize further destruction. This is analogous to the Hippocratic oath taken by physicians to “do no harm” in using their skills and knowledge (E. Guerrant, pers. comm.). In part, this position reflects a realistic assessment of the uncertainties of ecological restoration; the probability is extremely high that a created habitat or reintroduced population will be different in some essential aspects from the one it replaced, if it succeeds at all. More profoundly, it acknowledges that we still understand very little about the life history, interactions and evolution of most organisms in their natural state, so that a degree of humility is in order before we proclaim “‘success.” The strategic relationship between existing diversity and pro- Posed reintroductions is especially important for habitats or pop- ulations that may be difficult to replicate; in this respect there 1s a strong linkage between the degree of knowledge about the bi- ology ofa species or community proposed for reintroduction, and the strategic implications in undertaking such a project. For ex- ample, “no net loss” wetlands policies assume that the technology and scientific understanding reliably exist to create replacement wetlands, when in fact no such knowledge base exists. As Zedler and Langis (199 1) observe, decades may be required to ascertain 302 Rhodora [Vol. 94 the successful establishment of coastal wetland communities; even then, full functional parity with undisturbed reference wetlands may never occur, due to basic differences in soil or hydrology. For example, their study of natural and constructed wetlands in San Diego Bay found that after five years the constructed wetland areas provided only slightly more than half of the ecological func- tional equivalency of the natural site, in terms of soil character- istics, nutrient availability, biomass, and species diversity. Fur- thermore, they saw little evidence that this performance would improve over time. The same holds true for reintroduction projects with individual rare species. Reintroduction can play a useful role in helping to enhance damaged populations of rare species, to restore extirpated populations, or to establish new ones (Falk, 1990c; Falk and McMahan, 1988; Olwell et al., 1987, 1990). However, it can also be used intentionally by organizations with different objectives to justify the destruction of natural populations. In Western Aus- tralia, for instance, the successful results ofa state-funded research project into propagation techniques for one of Australia’s rarest terrestrial orchids at the Kings Park and Botanical Garden were used in part to rationalize the destruction of one of the few re- maining wild populations (Dixon, pers. comm.). In terms of establishing precedent for the future, mitigation represents a slippery slope. If carried to its logical extreme, any existing natural area could theoretically be subjected to a miti- gation tradeoff. Success in restoration and reintroduction work is thus a mixed blessing: while we may celebrate the successful es- tablishment of a new community or population, our very success may be used as a weapon against existing natural areas which we (but not others) are committed to protecting. Strategic issues that should be raised in evaluating a restoration or reintroduction project thus may include: 1. Is the project consistent with large-scale regional strategies for protecting land and biological diversity (Jenkins, 1989)? Will the project establish a precedent that might indirectly weaken protection for other natural areas? Will the project directly involve the destruction of any ¢x- isting natural area? If so, what is the comparative quality of the proposed sacrifice and replacement areas? What are the chances that unique biological or landscape values will be i 1992] Falk and Olwell— Considerations 303 lost if the project proceeds? Are there any alternatives to this course of action? What criteria were used in selecting the sacrifice area? Are they biologically sound and contribute to overall conser- vation objectives as identified in Step 1, Objectives, above? . Can impacts be directed to areas already disturbed and away from previously undisturbed sites? How much acreage is involved in destruction and creation? Have sufficient areas been included for buffer zones, large- scale ecological processes, and movement of communities over time? Has the project been costed-out over a realistic long-term financial schedule (King, 1991)? Is there a substantial net gain in funds available for conservation projects? How will unexpected future costs be borne, and by whom? Have per- formance bonds been posted that correspond to the time required to assess biological success? . Are there binding commitments by the land-managing agen- cies for permanent protection of the created habitat or re- introduced populations? What provisions have been made for monitoring and management (see Step 3, below). ww = wm al ~ 3. Land Management and Administration Introductions will most likely be of permanent conservation value if they are undertaken on land that is securely protected and managed for biological values. Long-term management In some form is essential for the maintenance of natural diversity On most protected areas in the continental United States, includ- ing even our largest natural areas and parks (e.g., Clark and Har- vey, 1988). The “managed natural area” is no longer considered an Oxymoron, but rather the dominant mode of preserving land and ecological values. _ Management considerations are especially important for re- introduced populations of rare plants. Many species inhabit suc- cessional communities, or habitat parcels so severely fragmented that only intensive management can maintain their integrity. Ke- Introduction experience with Pediocactus knowltonii, Penstemon barrettiae, Dicerandra immaculata, Stephanomeria malheurensis, Arctostaphylos uva-ursi var. leobreweri, Amsonia kearneyana, tyrax texana, Amsinckia grandiflora and other species confirms 304 Rhodora [Vol. 94 the importance of having a long-term management plan for a reintroduction site. The case of Stephanomeria malheurensis (malheur wire-lettuce) provides a good case in point. The plant, which is of considerable scientific interest because of evidence that it is recently speciated (Gottlieb, 1973, 1991), is known only from its type locality in south-central Oregon. The population was virtually destroyed in the 1980’s by grazing, fire and invasion by exotic species. Using seed collected at the time of the species’ original description, the Berry Botanic Garden collaborated with the U.S. Fish and Wild- life Service, Bureau of Land Management, Center for Plant Con- servation and others to re-establish the plant on the original site. Transplants were undertaken using four experimental ground cov- ers, with the area fenced by the BLM. For the foreseeable future, the species will have a chance only if management measures are maintained; otherwise, the same forces that brought about its demise once would almost certainly do so again (Parenti and Guerrant, 1990). Such cases are more likely to be the rule than the exception with rare species reintroductions. Several private companies have introduced or reintroduced rare species onto their lands following major disturbances from mining or construction operations. Such work is frequently mandated by state or Federal law and generally carried out as part of a master plan for post-mining land reclamation (Gillis, 1991). However, there is frequently a reluctance to establish, in the words of the environmental officer for one such firm, “species that carry a regulatory burden”’—in other words, restrict the company’s op- tions for future land use. Given the weakness of the Endangered Species Act in its present form to protect plant species on private land, reintroduced populations on non-conservation land should probably be regarded as ephemeral unless the company is willing to sign a long-term conservation contract with a public agency or conservation group. On the other hand, reintroduction can play a significant role in the recovery of endangered or threatened species, if integrated into long-term planning. An example is Trifolium stoloniferum (running buffalo-clover), for which a strategic recovery plan has been developed by the U.S. Fish and Wildlife Service, integrating legal status, site protection, establishment of reintroduced or en- hanced populations and research (U.S. Fish and Wildlife Service, 1989). 1992] Falk and Olwell— Considerations 305 Administrative and management considerations in planning reintroductions may include the following: —" . Doesa recovery plan exist? If so, did the agencies responsible for implementation play a part in its formulation, and have they made a commitment to carry it out? . Is the reintroduction site under secure long-term protection? Has the original threat(s) to the species or community been eliminated, or are there plans to manage the threat(s)? What is the commitment of the land-managing agency to conservation and biological management? Is such activity basic or incidental to its primary interest? . Will there be on-going monitoring and management of the site? Will such activities maintain habitat for the species in question? wr > nN 4. Biological Issues Despite elevated public interest and attention in rare species, the base of knowledge about rare plants remains small. The first comprehensive volume on the subject (Falk and Holsinger, 1991) Includes discussion of the genetics, population biology, demog- Taphy and ecology of rare plants to the extent they are known, but in general one is struck by how little published data exist for most rare species. Symposia organized by the California Native Plant Society (Elias, 1987), Natural Areas Association (Mitchell et al., 1990), and Society for Ecological Restoration (Bowles and Whelan, 1992) have contributed a great deal of new information, as have research reports in the conservation biology literature. Specific aspects of the biology of rare plants in some genera— Amsinckia, Astragalus, Calochortus, Cercocarpus, Clarkia, He- lianthus, Pediocactus, Pedicularis, Platanthera and Stephano- meria, for example—have been studied closely by one or more researchers, and there are now several hundred published citations of specific rare plant studies. But with over 4400 rare species 1n the United States alone, distributed in nearly 1000 genera, this 1S a thin knowledge base from which to successfully manage this diversity of species. uch information about rare species remains anecdotal and unrecorded except for the internal documents and contracted re- Ports of managing agencies. Recovery plans of the U.S. Fish and 306 Rhodora [Vol. 94 Wildlife Service help to draw together references for individual species, but here again fewer than 6% of species of national con- servation concern have been listed under the Endangered Species Act, and only half of those as noted above, have recovery plans. The detailed empirical observations of sanctuary and preserve managers, agency field staff, conservation horticulturists and am- ateur botanists are all of great value, but all too often these insights go unrecorded, and hence remain unavailable to all but a few people. Intensive research studies (such as Bowles, 1983) are ex- tremely useful for reintroduction work; however, few rare plant species have ecological studies in sufficient detail. Clearly, it will often be impractical to assemble a complete biological profile of every species or community before we begin engaging in conservation work. Consequently rare plant rein- troductions will involve a substantial element of trial and error experimentation (Harper, 1987; Griffith et al., 1989). Unlike well- studied crops or range grasses, field reintroductions of rare plants involve species whose developmental, reproductive and ecolog- ical characteristics are little known. Information critical to un- derstanding how to design an appropriate reintroduction, includ- ing the original biogeographic distribution, pollinators and seed dispersal agents, genetic architecture and long-term patterns of population growth, may be unknown. Present patterns of distri- bution may be so severely disrupted by anthropogenic causes that even the plants’ preferred soil type may not be reliably deter- mined. For example, the tallgrass prairie flora of the Great Plains now remains in such a small fraction of its original extent, and in such fragmented parcels of habitat, that it is difficult to know even which species were naturally rare and which were reduced to rarity by habitat modification. In Illinois, less than seven hun- dredths of one percent of the original prairie remains intact (White, 1978). Some entire communities, such as the tallgrass savannas, were virtually invisible because they had become so overgrown and invaded (Packard, 1991). A classic New England example of the need for deeper biological understanding is the attempted translocation of Isotria medeo- loides (small whorled pogonia) from a naturally occurring (but unprotected) site on the west side of Lake Winnipesaukee, New Hampshire, to nearby protected state land (Wilson, 1987). Orig- inally undertaken to salvage plants from a condominium devel- opment on private land beginning in 1985, the reintroduced plants 1992] Falk and Olwell— Considerations 307 have been monitored for recruitment, mortality and other pa- rameters. The species is known to undergo long-phase oscillations in reproductive population (Mehrhoff, 1989), further complicat- ing assessment of the size or status of the population at any given time; it is frequently difficult to distinguish such long-phase os- cillations from linear trends for decline or increase in population size (Brumback, pers. comm.). Thus, a proposal to confidently relocate this species as part of a mitigation project would have to be met with considerable skepticism, patience, or both. Most rare plant reintroductions involve outplanting of a subset of the genetic variation found within the species. As such, these efforts constitute empirical tests of several important hypotheses in population biology related to the genetic and evolutionary con- sequences of small population size: founder events and genetic bottlenecks, the effects of close inbreeding and the vulnerability of small populations to various types of stochasticity (Barrett and Kohn, 1991; Menges, 1991; Lewin, 1989). Although a few rare plant reintroductions (such as Amsinckia grandiflora by Pavlik, Stephanomeria malheurensis using material provided by Gott- lieb, and Hymenoxys acaulis by De Mauro) have been carried Out with baseline genetic data on the reintroduced individuals, until recently few have included such data or any provision to monitor changes in the genetic structure of the population over time. The result is two-fold: (i) collective knowledge of rare plants 1S not expanding as rapidly as it might; and (ii) many reintroduc- tions will fail biologically for reasons that will not be fully un- derstood. In general, the rare plant literature supports the view that ge- netic variation found among populations is significant reproduc- uvely, ecologically and evolutionarily, and should be maintained Conservation work (Holsinger and Gottlieb, 1991; Templeton, 1991; Hamrick et al., 199 1). Small differences among populations May represent incipient ecotypic differentiation or even the be- ginning of the speciation process. Moreover, a significant fraction of the total diversity in plants is found among populations (Ham- Mick et al., 1991). Approximately half of the allozyme loci in Plants, on the average, are polymorphic (although endemic species are somewhat lower, with approximately 40% of their loci poly- morphic). Twenty two percent of this diversity is distributed among Populations, influenced to some degree by ecological and life- history factors. Huenneke (1991) notes the ecological significance 308 Rhodora [Vol. 94 of genetic variation within and among populations for micro- habitat differentiation, resistance to pathogens and herbivores, and overall ecological amplitude. And Menges (1991) has de- scribed the correlation between low genetic variability in popu- lations and increased vulnerability to genetic, environmental, cat- astrophic and demographic stochastic events. Although direct experimental confirmation of these assertions remains scattered, genetic variation should be a major factor in designing reintroduction programs (Templeton, 1990). As Hol- singer (1991) observes, although the genetic architecture of most rare species has not been studied directly, the existing literature does permit reasonable generalizations. For example, outcrossing species on the whole exhibit a higher proportion of polymorphic loci (p) and more alleles per polymorphic locus (A) than do selfing species; conversely, selfing species distribute a far higher propor- tion of their variation among populations (G,,) than do outcross- ers. These and other general observations should permit the de- sign of reasonably effective genetic sampling programs, especially given the small number of natural populations for most rare plants. Propagules should be drawn from several populations, and from a sufficient number of individuals within each population, to ensure that the majority of genetic variation at polymorphic loci has been captured. Several sampling strategies have been pro- posed for rare plants (see: Brown and Briggs, 1991; Center for Plant Conservation, 1991; Guerrant, 1992). Reintroduction plans should reflect conscious decisions about the size of reintroduced populations; artificially small populations may be considerably more vulnerable to stochastic events, drift and inbreeding de- pression, particularly in species that are not adapted to small population size in the wild. Guerrant (1992) observes that many reintroduction programs should involve the establishment of sev- eral experimental populations, to guard against environmental and genetic stochasticity, and to exploit differen Among the major biological considerations in rare plant rein- troductions are: 1. What is known about the genetic structure of the species, within and among populations? Can assessment of this variation be built into the project design? 2. What is known about the population biology and demog- raphy of the species, including effective breeding popula- 1992] Falk and Olwell— Considerations 309 tion size? What short- and long-cycle fluctuations in these values are observed in natural populations? . How many experimental outplant sites should there be? . From which populations should propagules be drawn? Should genetic material from several populations be com- bined or kept isolated? . Is the species endemic to certain edaphic types? Can it survive elsewhere if competition is controlled? Do the edaphic tendencies appear to be obligatory? Are these ob- servations consistent for all populations? . What are the essential symbionts: pollinators, seed dis- persants, mycorrhizal fungi, others? . What is the species breeding system? Is propagation largely sexual, asexual, mixed, or variable? What is the likelihood that the population will be able to establish itself repro- ductively on the new site? . What are the viability characteristics of the seeds? What dormancy mechanisms exist, and how is dormancy bro- ken? What proportion of the genepool remains as unger- minated seed in the soil seed bank (Glass, 1989)? . What are the optimal environmental conditions for seed- ling growth? How much tolerance does the species have for conditions different from the optima? . Is the species characteristic of successional environments? Can it survive elsewhere if competition is controlled? . What is the species’ ‘‘natural” range? What factors, natural and anthropogenic, limit its distribution? wn & WwW ON oo ~— \O — i=) od — 5. Technical Feasibility Consideration of the finer points of context and strategy will be largely academic if a reintroduction project is not technically Possible. Feasibility certainly represents an absolute prerequisite for any actual on-the-ground activity and should be borne in mind throughout the planning process. Nonetheless, we have Te- Served this section for last because we believe that the feasibility of a project should not necessarily determine its priority or de- Sirability. In other words, simply because reintroduction can be done does not mean it necessarily should be done. Technical considerations grade strongly into the basic issues of biology discussed in the preceding step. For instance, lack of 310 Rhodora [Vol. 94 understanding about seed dormancy, pollinators, or edaphic re- quirements may raise questions about the biology of a project, but such unknowns also create uncertainty about its feasibility. Moreover, short-term success may not mean that long-term tech- nical problems have been solved. For instance, a recent study of California coastal sage scrub indicated that, although there was evidence of scrub seedling regeneration in the first few years of the project, the true success of the reintroduction could not be reliably assessed for 30-50 years. Similar caveats have been raised or community restorations and population reintroductions in prairies, woodlands and riparian areas (see essays in Jordan et al., 1987). Technical feasibility of the off-site cultivation phase may also play a part in the planning and reintroduction process for sub-tropical species, as Wallace (1990) has described for sand scrub endemics of peninsular Florida. A review of reintroduction and restoration literature reveals a wide assortment of technical considerations for successful rein- troduction practice. Most considerations are highly specific to a species or habitat type (e.g., Tipton and Taylor, 1984), but a few general considerations emerge as significant: — . Does a reliable source of reintroduction material exist, either in a natural population or off-site conservation collection? If the latter, are provenance, genetic composition and cu- ratorial history known? If collecting propagules from a wild population is required, do plants exist in sufficient numbers? Is the population re- producing, and if so are seeds or vegetative propagules pro- duced in sufficient quantity to permit removal without in- terfering with reproduction? Can material be reliably propagated to a stage that permits transplanting? Is there a dormancy requirement before seeds will germinate? Are propagules disease-free? Is there any possibility of im- porting diseases inadvertently into the new habitat, or if augmenting a population, into existing habitat? Are techniques for transplanting material known? Is there evidence that transplants will survive? What sort of interim management measures, such as watering, will be required in the early stages? . Is the recipient site free of threats or disturbances that might Ss e * ON 1992] Falk and Olwell— Considerations 311 compromise the prospects for success? Is the site stabilized physically from erosion? Is the recipient community invad- ed by exotic organisms that might reduce the likelihood of success? CONCLUSION Increasing destruction, disturbance and fragmentation of re- maining pristine habitat require that alt tive strategies be sought rapidly and incorporated into strategic planning for biodiversity. Reintroduction and restoration represent important tools for the preservation of biological diversity in its ecological and evolu- tionary context (Falk, 1992), as well as suggesting a more balanced and creative relationship of human society to nature than indus- trial societies have achieved to date (Jordan, 1986). It is imper- ative that practitioners of these important techniques remain clear about the strategic implications of their actions, to ensure that the work of restoration contributes positively to the continuation of the diversity of life on Earth. ACKNOWLEDGMENTS Preparation of this paper was supported in part by a grant from The Joyce Foundation. LITERATURE CITED Barrett, S.C.H. AND J.R.Konn, 1991. Genetic and evolutionary consequences of small population size in plants: = for conservation, pp. 3-30. In: D. Falk and K. Holsinger, Eds., op. cit. Bowies,M.L. 1983. The tallgrass ceaide bint Platanthera leucophaea (Nutt.) Lindl. and Cypripedium candidum Muhl. ex Willd. Some aspects of their Status, biology, and ‘ae and implications toward management of natural areas. Natural Areas J. 3: 14-38. ——— AND C. J. WHELAN. . Recovery and Restoration of Endangered Species. Cambridge University Press, Cambridge, England (in press). BROcKIE, R. E., L. L. Loops, M. B. USHER AND O. HAMANN. 1988. Biological invasions of island nature preserves. Biol. Cons. 44: 9-36. Brown, A. H. D. AND J. D. Briccs. 1991. Sampling strategies for genetic vari- ation in ex situ collections of aes plant species, pp. 99-119. In: D. K. E. Holsinger, Eds., JlzZ Rhodora [Vol. 94 . 1992. Database of Rare Plants of the United States. Missouri Botanical Garden, St. Louis, MO. CLARK, T. W. ANDA. H. Harvey. 1988. Management of the Greater Yellowstone males Northern Rockies Conservation Cooperative, Jackson, Cook, R. E. AND P. Dixon. 1989. A Review of Recovery Plans for Threatened and Endangered Plant Species, A Report for the World Wildlife Fund. World Wildlife Fund, Washington, DC. D1amonp, J. 1985. How and why eroded ecosystems should be restored. Nature 313: 629-630. . 1987. Reflection on goals and on the relationship between theory and practice, pp. 329-336. In: W. Jordan, III, N. Gilpin and J. Aber, Eds., op clt. DopperTEEN, R. A. 1989. Scientific analysis and policy evaluation of wetland replication in Massachusetts. Ph.D. dissertation, Tufts University, Med- ford, MA. DRABELLE, D. 1985. The Endangered Species Program, pp. 73-90. Jn: R. L. DiSilvestro, Ed. Audubon Wildlife Report 1985. Dartmouth ie Com- York Epwarps, S. W. 1990. The role of botanical gardens in a deteriorating global environment. Fremontia 18(1): 3-8. EHRLICH, P.R. 1991. Population diversity and the future of ecosystems. Science 254175. Euias, T. S., Ed. 1987. Conservation and Management of Rare and Endangered Plants. California Native Plant Society, Sacramento, CA. FALK, D. A. 1990a. Discovering the future, creating the past: some reflections of restoration. Restor. and Manag. Notes 8: 71-72. 1 Re hepa be Spray are species: a strategy for conservation, me 328-334. In: J. , Ed. Environmental aero: Science and Strategies for oe ie Earth. Island Press, Washington, DC. . 1990c. Ape strategies for conserving plant aie diversity. Ann. Missouri Bot. Gard. 77: 38-47. 92. Anevolutionary perspective on plant conservation. J. Linn. Soc., Bot. (in press). > K. E. Hotsincer. 1991. Genetics and Conservation of Rare Plants. Oxford University Press, New York. — anpD L.R. McMauan. 1988. Endangered species: managing for diversity. Natural Areas J. 8: 91-99. Fatupi, A. 1973. A Reader in Planning Theory. Pergamon Press, Oxford, U.K. FIEDLER, P. L. 1986. Concepts of rarity in vascular plant species, with special reference to the genus Calochortus Pursh (Liliaceae). Taxon 35: 502-518. AND J. J. AHOUsSE. 1992. Hierarchies of aes toward an understanding of rarity in vascular plant species, pp. 23-47. In: P. L. Fiedler and S. Jain, — Conservation Biology: The Theory a Practice of Nature Conserva- n, Preservation, and Management. Chapman and Hall, New York. thea - M. 1991. Bringing back the land: ecologists evaluate ei nalla success on western coal lands. BioScience 41(2): 68-71 Grass, S. 1989. The role of soil seed banks in restoration and management. Restor. and Manag. Notes 7: 24-29. Gottues, L. D. 1973. Genetic differentiation, sympatric speciation, and the origin of a diploid species of Stephanomeria. Amer. J. Bot. 60: 5 545-553. ea ee — - a ¢ 1992] Falk and Olwell—Considerations 313 ———. 1991. The malheur wire-lettuce: a rare, recently evolved Oregon species. Kalmiopsis 1: 9-13. GrirritH, B., J. M. Scott, J. W. CARPENTER AND C. REED. 1989. Translocation as a species conservation tool: status and strategy. Science 245: 477-480 GuerrRANT, E. O. 1992. Genetic and demographic considerations in sampling and reintroduction of rare plants, pp. 321-344. In: P. L. Fiedler and S. Jain, Eds., op. cit. Hamrick, J. L., M. J. W. Gopt, D. A. MuRAwSKI AND M. D. LoveELess. 1991. Correlations between species traits and allozyme diversity: implications for conservation biology, pp. 75-86. In: D. A. Falk and K. E. Holsinger, Eds., op. cit. Harper, J. L. 1987. The heuristic value of ecological restoration. Jn: W. R. Jordan, III, M. E. Gilpin and J. D. Aber, Eds., op. cit. Hawaii OFFICE OF STATE PLANNING. 1991. Golf course development in Hawaii: impact and policy recommendations. Honolulu, Hawaii (draft Houpen, C. 1986. Regrowing a dry tropical forest. Science 234: 809-8 10. Hosincer, K.E. 1991. Conservation of genetic diversity in rare and endangered plants, pp. 626-634. In: The Unity of Evolutionary Biology. Dioscorides Press, Portland, OR. AND L. D. Gottuies. 1991. Conservation of rare and endangered plants: principles and prospects, pp. 195-208. Jn: D. A. Falk and K. E. Holsinger, Eds., op. cit. HuENneKE, L. F. 1991. Ecological implications of genetic variation in plant populations, pp. 31-44. In: D. A. Falk and K. E. Holsinger, Eds., op. cit. INTERNATIONAL UNION FOR THE CONSERVATION OF NATURE AND NATURAL RE- SOURCES. 1987. The IUCN Position Statement on Translocation of Living Organisms. Species Survival Commission, IUCN, Gland, Switzerland. —. 1992. Guidelines for reintroductions. Species Survival Commission, Reintroduction Specialist Group. Royal Botanic Gardens, Kew, U.K. (in press) s). JARMAN, N. M., R. A. DoBBERTEEN, B. WINDMILLER AND P. R. Levito. 1991. Evaluation of created freshwater wetlands in Massachusetts. Restor. and Manag. Notes 9: 26-29. Jenkins, R. E. 1989. Long-term conservation and preservation complexes. The Nature Conservancy Mag. 39: 4-7. Jones, S. R., Ed. 1990, Endangered Species Update 8: 1-88. JorDAN, W. R., III. 1986. Restoration and the reentry of nature. Orion 5: 14-25. ——, M. E. Gipn anp J. D. Aer, Eds. 1987. Restoration Ecology: A Syn- thetic Approach to Ecological Research. Cambridge University Press, Cam- bridge, U.K. Kinc, D. M. 1991. Costing out restoration. Restor. and Manag. Notes 9: 15-21 KruckeperG, A. R. AND D. RABINOwITz. 1985. Biological aspects of endemism in higher plants. Ann. Rev. Ecol. Syst. 16: 447-479. _ . Lewm,R. 1989, Hidden complexities in the risks of extinction. Science 243(4896): 1 4. McManan, L. R. 1990. Propagation and reintroduction of imperiled plants, and the role of botanical gardens and arboreta. Endangered Species Update 8: 47. 314 Rhodora [Vol. 94 Menruorr, L. A. 1989. Thed f decli lati f dangered orchid, Isotria medeoloides. Ecology 70: 783-786. Mencess, E. S. 1991. The application of minimum viable pee theory to plants, pp. 45-61. Jn: D. A. Falk and K. E. Holsinger, Eds., cit. Mixiar, C. I. AND W. J. Lipsy. 1989. Disneyland or native os genetics and the an Restor. and Manag. Notes 7: 18-24. Reprinted in Deu walie 17(2): 3 Mircnerrs,. R. S., C. J. cas AND D. J. LeopoLp, Eds. 1990. Ecosystem Management: Rare Species os Significant Habitats. New York State Mu- seum Bulletin 471, Alban Munro, J. W. 1991. Wetland SS eee in the mitigation context. Restor. and Manag. Notes 9: 80-86. Myers, N. 1990. The biodiversity challenge: expanded hot-spots analysis. The Environmentalist 1(4): 243-256. Nasu, J. M. 1991. Learning how to revive the wilds of Eden. Time 14(October): 62-64. Owe Lt, P., A. C. Cutty, P. J. KNIGHT AND S. G. BRAcK. 1987. Pediocactus knowltonii recovery efforts, pp. 519-522. In: T. S. Elias, Ed., op cit. AND 1990. The establishment of a new population of Pediocactus knowltonii: third year assessment, pp. 189-193. In: R. S. Mitch- ell, R. C. Sheviak and D. Leopold, Eds., op. cit. PACKARD, S.G. 1991. Justa few oddball species: restoration and the rediscovery at ag Giles savanna. = R. Nilsen, Ed., Helping Nature Heal: A Rein- Restoration. Whole Earth Catalogue/Ten Speed Press, cee CA. PARENTI, R. L. AND E. O. GueRRANT. 1990. Down but not out: reintroduction of the fees a wirelettuce, Stephanomeria malheurensis. Endan- gered Species Update 8 Ss Peters, R. L. 1988. Effects i. global ——- on species and habitats: an over- view. Endangered Species Update 5: 1-8. SCHWARTZ, M. W. 1992. Modeling the effect of fragmentation on the ability of trees to respond to climatic warming. Biodiversity and Conserv. (in press). TEMPLETON, A. R. 1990. The rate of genetics in captive breeding ee reintro- duction for species conservation. Endangered Species Update 8: 14-17. Off-site breeding of animals and implications for aa conser- vation strategies, pp. 182-194. Jn: D. A. Falk and K. E. Holsinger, Eds., op. Tipton, J. L. AND R. M. Taytor. 1984. T sotebush (Larrea tridentata (D.C.) Cav.) from native stands. J. Faviron. Hort. 2: 83-85. U.S. Fish AND WILDLIFE SERVICE. 1988. Endangered Species Act of 1973 As Amended through the 100th eng U‘S. Department of the Interior, Fish and Wildlife Service, Washingt 1989. Trifolium eee et — Plan. Twin Cities, MN. 1990a. Report to Congress: Endangered and Threatened Species Re- covery Program. U.S. Department of the Interior, Fish and Wildlife Service, Washington, DC. ——. 1990b. Policy and Guidelines for Planning and Coordinating Recovery 1992] Falk and Olwell— Considerations ato of Endangered and Threatened Species. U.S. Department of the Interior, Fish and Wildlife Service, Washington, DC. VeRMEU, G. J. 1991. When biotas meee understanding biotic exchange. Science 253: 1099-1104. Virousek, P. M., L. R. WALKER, L. D. WHITEAKER, D. MUELLER-DOMBOIS AND P. A. Matson. 1987. Biological invasion ah Myrica faya alters ecosystem development in Hawaii. Science 238: 802-8 WALLACE, S. R. =e eas 1 Fl y a he pi Endangered Species Update 8 Waite, J. 1978. ome Natural Areas Inventory Technical Report, Vol. 1. Survey Methods and Results. Illinois Dept. of Cons., Springfield, IL. Witson, M. V. 1987. New England orchid endangered. Boston Globe. 12 Oc- tober. ZEDLER, J. AND R. Lanais. 1991. Comparisons of ecnonarap and natural salt marshes of San Diego Bay. Restor. and Manag. Notes 9: CENTER FOR PLANT CONSERVATION MISSOURI BOTANICAL GARDEN P.O. BOX 299 ST. LOUIS, MO 63166 RHODORA, Vol. 94, No. 879, pp. 316-318, 1992 HYMENOPHYLLUM WILSONIT HOOKER (HYMENOPHYLLACEAE) IN THE IBERIAN PENINSULA F. X. SoNorA, S. ORTIZ AND J. RODRIGUEZ-OUBINA We found Hymenophyllum wilsonii Hooker (SANT 19335) in the basin of the River Seixo de Landoi in the Serra da Capelada (Ortigueira, A Corufia, N.W. Spain; UTM 29TNJ8637; Figure 1). Hitherto, this small fern, which Lainz (1986) included among the species inquirenda in the Flora Iberica, in Europe has only been reported in northern and western Great Britain, Ireland, the Faroes, S.W. Norway, N.W. France, Madeira, the Azores and Canaries (Jalas and Suominen, 1972). The site at which Hymenophyllum wilsonii was found is located at an altitude of 350 m a.s.l. on a substrate of serpentines and granulites. According to the chorological typology of Rivas-Mar- tinez (1987), it is in the Eurosiberian region, Cantabro-Atlantic province, Galaico-Asturian sector. The climate, of Atlantic Eu- ropean [V(VI)] type according to Allte’s classification and of tem- perate maritime type according to Papadakis’ (Carballeira et al., 1983), is ideal for this relictual pteridophyte: the annual mean temperature is 12.5°C, the mean annual rainfall 1500 mm, and there are no frosts or summer drought (Giaccobbe dryness index 7). Bioclimatically, this site lies in the hyperhumid ombroclime hill belt of Rivas-Martinez (1987). Hymenophyllum wilsonii carpets the wet rocks of the water- course with H. tunbrigense (L.) Sm., in a shady, humid environ- ment engendered by a canopy of Quercus robur L. This environ- ment likewise harbors other biogeographically interesting ferns of Macaronesian type according to the biogeographical typology of Pichi-Sermolli et al. (1988), including Vandenboschia speciosa (Willd.) Kunkel and, in particular, Culcita macrocarpa K. Presl. The latter, with specimens exceeding 4 m in height, spreads almost uninterruptedly along the shaded bank of the river over an area of more than half a hectare, making this population certainly the largest known one in continental Europe. LITERATURE CITED CARBALLEIRA, A., A. DevesA, C. RETUERTO, R. SANTILLAN AND F. Uciepa. 1983. Teicictinaicivea de Galicia. Fundacién Barrié de la Maza, A Corufia, Spain. 316 ee 1992] Sonora et al. —Hymenophyllum wilsonii 317 “ 4 f La D JA g x@ 3 ‘ - +a ) 7 — a = " Figure 1, Location of Hymenophyllum wilsonit Hooker in c ontinental Europe (Jalas and Suominen, 1972). Native ocurrence: @; probably extinct: %; new locality: yr. 318 Rhodora [Vol. 94 JaLas, J. AND J. SUOMINEN. 1972. Atlas Florae Europaeae 1. Helsinki, Finland. Lainz, M. 1986. Hymenophyllum Sm., pp. 73-75. In: S. Castroviejo et al., Eds. Flora Iberica. C.S.I.C., Madrid, Spain. PicHI-SERMOLLI, R. E. G., a EspANA AND A. E. SALVo. 1988. El valor biogeo- grafico de la pteridoflora ibérica. Lazaroa 10: 187-205. Rivas-MArRTINEz, S. 1987. Memoria del mapa de las series de vegetacion de Espana. ICONA, Madrid, Spain. DEPARTAMENTO DE BIOLOXIA VEXETAL LABORATORIO DE BOTANICA FACULDADE DE FARMACIA UNIVERSIDADE DE SANTIAGO 15706 SANTIAGO DE COMPOSTELA GALICIA—ESPANA - —- ete | RHODORA, Vol. 94, No. 879, pp. 319-322, 1992 CHOOSING THE CORRECT NAME FOR ACONOGONON (POLYGONUM SECT. ACONOGONON) IN ALASKA KENTON L. CHAMBERS Polygonum alaskanum Wight ex Hultén (1944) represents the earliest validly published name for the species that was called Aconogonon hultenianum (Yurtz.) Tzvel. in a recent monograph (Hong, 1991). In many earlier works the authorities for P. alas- kanum were given as “(Small) Hultén.” However, Small should not be cited as the parenthetical author because the latter’s earlier- published varietal name Polygonum alpinum var. alaskanum Small (1895) is an avowed substitute for P. alpinum var. lapa- thifolium Cham. & Schlecht., and hence it is illegitimate. The publication by Chamisso and Schlechtendal, however, does pro- vide the description and type for the legitimate new name F, alaskanum published by Hultén. DISCUSSION Polygonum alaskanum Wight ex Hultén is widespread in Alas- ka and northwestern Canada. The purpose of this note is to call attention to the legitimacy of the above name (Hultén, 1944), and to correct the nomenclature for this species given in a recently published taxonomic treatment (Hong, 1991). The taxon in question was first named at the varietal level as Polygonum alpinum var. lapathifolium Cham. & Schlecht. (Lin- naea 3: 38. 1828). Small recognized this variety in his monograph of Polygonum in North America (1895), but gave it the new name P. alpinum var. alaskanum Small, while citing the name by Cha- misso and Schlechtendal as a synonym. Small pointed out the Prior existence of the species name P. lapathifolium L., based on a different type, and he may have been following the precedent of some other taxonomists of his time, who disallowed more than One use of a given epithet within a genus—even at different levels in the taxonomic hierarchy. In any case, the varietal name pub- lished by Small is an avowed substitute for the earlier var. /a- pathifolium of Chamisso and Schlechtendal; it is therefore no- menclaturally superfi and illegitimate under the International Code of Botanical Nomenclature (Greuter, 1988, Art. 63; Kartesz and Gandhi, 1990). ate 320 Rhodora [Vol. 94 Hultén (1944) was the first to propose a name at the species rank for this taxon. In his publication of Polygonum alaskanum, he cited ‘“(Small) Wight nov. comb.” as authorities for the name, and he gave “Polygonum alpinum Alaskanum Small, Monogr. N. Amer. Spec. Polyg. (1895) p. 33” as the basionym. Hong (1991) cited the authorities for P. alaskanum as “‘(Small) Wight ex Hul- tén,” and he claimed that this name was illegitimate. It is evident from Hultén’s publication, however, that his intent was to provide a name at the species level for the taxon that had earlier been named P. alpinum var. lapathifolium Cham. & Schlecht. He stated (p. 612): “P. alaskanum was transferred from variety to species by Wight in the manuscript flora of Alaska by Standley, and I therefore keep this name, although it was not published, especially as the first varietal name of the plant, /apathifolium, cannot be used, as it is already occupied” [note: cf. Polygonum lapathifolium L.]. Article 41.3 of ICBN (Greuter, 1988) specifies that “In order to be validly published, a name of a species must be accompanied . (b) by a reference to a previously and effectively published description or diagnosis of a species or infraspecific taxon. This requirement is met by Hultén’s citation both of‘ ‘Polygonum alpinum y lapathifolium Cham. & Schlecht. in Linnaea 3 (1828) p. 38” and “Polygonum alpinum Alaskanum Small, Monogr. N. Amer. Spec. Polyg. (1895) p. 33.” A regulation in ICBN which mentions the option of using previously illegitimate epithets in a new position or sense is Art. 72, Note 1: ‘When a new epithet is required, an author may adopt an epithet previously given to the taxon in an illegitimate name if there is no obstacle to its em- ployment in the new position or sense; the resultant combination is treated as the name of a new taxon or as a Nomen novum, as the case may be.” At the time of Hultén’s work, no other specific epithet had been published for this taxon; therefore, Art. 72, which deals with rejected names and nomina nova, does not strictly apply. I mention it here as an example of a rule that sanctions the use of illegitimate epithets when forming new names in par- ticular cases. In his revision of Aconogonon (Meisn.) Reichenb. (Polygonum sect. Aconogonon Meisn.), Hong (1991) rejected the name “Po- lygonum alaskanum (Small) Wight ex Hultén” as illegitimate. Instead, Hong took up Aconogonon hultenianum (Yurtz.) Tzvel. (Novit. Syst. Plant Vasc. 24: 77. 1987), which is based on Polyg- onum alaskanum ssp. hultenianum Yurtz. (Bot. Zurn. 59: 1452. 1992] Chambers—Aconogonon 321 1974; type from Fairbanks area, Alaska, collected in 1962). Hong also stated that the synonym P. alpinum ssp. alaskanum (Small) Welsh” (1968), with the basionym attributed to Small rather than to Hultén, was illegitimate. The original publication refers to “Polygonum alaskanum (Small) Wight” as a “nov. comb.” rather than a “sp. nov. ” Despite Hultén’s choice of phrases, we must retrospectively credit him with having named a new species by reference to a previously published description of the taxon at the varietal level. There appears to be no barrier to the acceptance of Polygonum alas- kanum Wight ex Hultén as the earliest validly published name for the species. Polygonum alaskanum is typified by plants of the pubescent phase of the species [to which Hong (1991) gave the name Acon- ogonon hultenianum var. lapathifolium (Cham. & Schlecht.) S.- P. Hong]. The usually glabrous, more eastern phase has as its earliest varietal name P. alaskanum var. glabrescens Hultén (1944). Because I favor use of the generic name Polygonum for these taxa, I leave to other workers the option of a new combination placing var. glabrescens in Aconogonon. The taxonomic assignment of Polygonum alaskanum varies in recent floristic works. It was merged with P. phytolaccifolium Meisn. ex Small by Hitchcock (1964) and with P. alpinum All. by Welsh (1968). The name “Aconogonon alaskanum (Small) Wight” mentioned by Hong (1991, p. 331) cannot be traced; however, there does exist the name A 2 lash (Wight ex Hulten) Sojak [as “(Small) Sojak”], Preslia 46: 150. 1974. The name Polygonum alaskanum (Small) Wright [sic!] ex Harshberger (1928) is a nomen nudum published without description or bibliographic reference. ACKNOWLEDGMENTS I thank Stanley L. Welsh and James L. Reveal for helpful com- ments on the manuscript. LITERATURE CITED Greuter, W. (Chairman, Editorial Committee). 1988. International Code of Botanical Nomenclature. Regnum Veg. 118. HARSHBERGER, J. W. 1928. Tundra vegetation of central Alaska directly under the Arctic Circle. Proc. Amer. Philos. Soc. 67: 715-23 322 Rhodora [Vol. 94 Hitcucock, C. L. 1964. ig tag pp. 139-168. In: poeaanay Plants of the Pacific Northwest, pt. 2. Univ. of Washington Press, Seat Hone, S.-P. 1991. A revision of ge ahi =Polygonum oe Aconogonon, Polygonaceae) in North America. Rhodora 93: 322-346. Huttén, E. 1944. Polygonum. In: Flora of Alaska and Yukon. Acta Univ. Lund. F. Avd. 2, Bd 40 (no. 1) 4: 609-624. Kartesz, J. T. AND GANDHI, K. N. 1990. Nomenclatural notes for the North American flora. No. 2. Phytologia 68: 421-427. SMALL, J. K. 1895. A monograph of the North American species of the genus Polygonum. Mem. Dept. Bot. Columbia Coll. 1: 1-183. We su, S. L. 1968. Nomenclature changes in the Alaskan flora. Great Basin Naturalist 28: 147-156. DEPARTMENT OF BOTANY AND PLANT PATHOLOGY OREGON STATE UNIVERSITY CORVALLIS, OR 97331 RHODORA, Vol. 94, No. 879, pp. 323-325, 1992 COMMENTS ON THE PHENOLOGY OF CARPHEPHORUS CORYMBOSUS (COMPOSITAE) Davip T. COREY The genus Carphephorus includes 4 species (Correa and Wilbur, 1969: King and Robinson, 1987). There is no published infor- mation on the phenology of any Carphephorus species. This report gives the phenology for Carphephorus corymbosus (Nutt.) Torr. & Gray. Carphephorus corymbosus is a perennial herbaceous composite. Plants of this species consist of an acaulescent rosette of spatulate to oblanceolate leaves from thickened, fibrous roots, and a long green terete stem which branches into an inflorescence with purple to lavender flowers (Correa and Wilbur, 1969). Stems have al- ternate leaves which become progressively smaller distally. Carphephorus corymbosus grows in open, sandy areas in sand pine scrub, sandhills, and open pinewood barrens from south- eastern Georgia throughout most of peninsular Florida (Correa and Wilbur, 1969; Wunderlin, 1982). The species is an indicator species of sandhills. I examined 3848 basal rosettes at 12 sandhill communities in north and central Florida from 24 August to 8 November 1988 (Figure 1). Carphephorus cor -ymbosus was absent from 2 of these study sites (Suwannee River State Park, Suwannee Co. and San Felasco Hammock Preserve, Alachua Co.). Its density at the 10 remaining sites ranged from .004 basal rosettes per m? at Wekiwa Springs State Park, Orange Co. to _470 at Spruce Creek Preserve, Volusia Co. (mean = .086 basal rosettes per m2? for the 10 study sites). Density was determined in late summer and early fall. Shoots emerge from vertical roots in early March. Each root may produce from 1 to 5 basal rosettes (< = 3), which arise a few mm below the surface and probably from the root/hypocotyl. These basal rosettes often arise from locations on the rootstock different from those of the previous year. Although 1-5 basal rosettes may be produced, usually only one produces a flowering stem (range: 42.8-136.2 cm, ¥ = 74.9 cm, ” = 197; stems were measured only after buds had opened). Rarely, 2 basal rosettes always much older rootstocks may produce more ) more stems. More research is needed in this area. 323 324 Rhodora [Vol. 94 Figure 1. Range of Carphephorus corymbosus in Florida (@) and 12 study site locations (@): Suwannee River State Park, Suwannee Co.; O’leno State Park, Columbia Co.; San Felasco Hammock and Morningside Nature Center, Alachua Co.; Interlachen, Putnam Co.; Spruce Creek Preserve and Orange City, Volusia Co.; Wekiwa Springs State Park, Orange Co.; Janet Butterfield Brooks Preserve and Sandhill Boy Scout Reservation, Hernando Co.; Starkey Well Field Area, Pasco Co.; Bok Tower Gardens, Polk Co. Of the 3848 basal rosettes examined, 9.25% produced inflo- rescences. Terete stems first appear in late May and reach full length by early June. Around mid-June buds develop on the ends of the corymbs; flowers appear in mid-August. Leaf fall occurs 1n December and the stems and corymbs dry out but remain standing well into the next year, and often may be seen alongside new basal rosettes with stems. On several occasions, a root mass was dug up (reburied after examination); it bore a new basal rosette with a stem and an old dried stem from the previous year. 1992] Corey — Carphephorus corymbosus 325 ACKNOWLEDGMENTS I am grateful to I. Jack Stout and Joseph A. Beatty for their assistance in this study. Robert Wilce, George Ellmore, and Loran Anderson improved an earlier draft of this manuscript. I thank Richard P. Wunderlin and Loran Anderson for the county records used in Figure 1. I thank Ellis Collins, Fred Hunt, The Nature Conservancy, Florida Department of Natural Resources, South- west Florida Water Management District, Morningside Nature Preserve, Bok Tower Gardens and Sandhill Boy Scout Preserve for allowing access to the study sites. LITERATURE CITED Correa, A. AND R. L. Witaur. 1969. A revision of the some Carphephorus (Compositae—Eupatorieae). J. Elisha Mitchell Sci. Soc. 85: 79-91. Kina, R. M. AND H. Rosinson. 1987. The genera of the Eupatorieae (Asteraceae). WUNDERLIN, R. P. 1982. Guide 6 the Vascular Plants of Central Florida. Uni- versity Press of Florida, Tampa DEPARTMENT OF ZOOLOGY SOUTHERN ILLINOIS UNIVERSITY CARBONDALE, IL 62901 THE NEW ENGLAND BOTANICAL CLUB 22 Divinity Avenue Cambridge, MA 02138 The New England Botanical Club is a non-profit organization that promotes the study of plants of North America, especially the flora of New England and adjacent areas. The Club holds regular meetings, has a large herbarium of New England plants, anda library. It publishes a quarterly journal, RHODORA, which is now in its 94th year and contains about 400 pages a volume. Membership is open to all persons interested in systematics and field botany. Annual dues are $35.00, including a subscription to RHODORA. Members living within about 200 miles of Boston receive notices of the Club meetings. To join, please fill out this membership application and send with enclosed dues to the above address. Regular Member $35.00 Family Rate $45.00 For this calendar year For the next calendar year scceceeoese Name Address City & State Zip Special interests (optional): Vol. 94, No. 878, including pages 111-212, was issued June 17, 1992. 326 THE NEW ENGLAND BOTANICAL CLUB Elected Officers and Council Members for 1992-1993 President: Leslie J. Mehrhoff, Connecticut Geology and Natural eg Survey, U-42, University of Connecticut, Storrs, CT 68 Vice-President (and Program Chair): C. Barre Hellquist, De- , partment of Biology, North Adams State College, North Adams, MA 02147 Corresponding Secretary: Harold G. Brotzman, Department of Biology, North Adams State College, North Adams, MA 02147 Treasurer: A. Linn Bogle, Department of Plant Biology, Nesmith Hall, University of New Hampshire, Durham, NH 03824 Recording Secretary: W. Donald Hudson, Jr. Curator of Vascular Plants: Raymond Angelo Assistant Curator of Vascular Plants: Cathy A. Paris Curator of Non-Vascular Plants: Anna M. Reid Librarian: Lisa A. Standley Council: Consisting of the Elected Officers, Associate Curator, Editor of Rhodora and — Councillors: David S. Barrington (Past President) Nancy M. Eyster-Smith °93 David S. Conant ’94 William E. Brumback °95 ] Marybeth Deller, (Graduate Student Member) °92 0: See RHODORA July 1992 Vol. 94, No. 879 CONTENTS SYMPOSIUM New England Plant Conservation: The Scientific Basis for Effective Action FOREWORD Leslie J. Mehrhoff, President NEBC, and William E. Brumback, Conservation Director NEWFS SYMPOSIUM PAPERS axonomic issues in rare species protection Lisa A, Standley 218 Habitat management: a decision-making process ETE Go TRMEN IC os a a ek ees Sees 258 Scientific and policy considerations in restoration and reintroduction of endangered species Donald A. Falk ant apt — ik oi chee a 287 OTHER PAPERS IN THIS — cia Hooker (Hymenophyllaceae) in the Iberian penins re, Sek S. Ortiz and J. Rodriguez-Oubifia .............++++: 316 Choosing the correct name for Aconogonon (Polygonum sect. Aconogo- non) in Alas Kenton L. Chambers 319 omments on the phenology of Carphephorus corymbosus (Compositae) DRE TC a es ees 323 NEBC Membership Application Form ....................-.--2++-+:: 3 BC Officers Council Members...........-........ inside back cover ‘Hovova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vv ol. 94 October 1992 No. 880 The New England Botanical Club, Inc. 22 Divinity Avenue, Cambridge, Massachusetts 02138 RAGBORA NORTON H. NICKERSON, Editor-in-Chief Associate Editors DAVID S. BARRINGTON RICHARD A. FRALICK A. LINN BOGLE GERALD J. GASTONY DAVID E. BOUFFORD C. BARRE HELLQUIST CHRISTOPHER S. CAMPBELL MICHAEL W. LEFOR WILLIAM D. COUNTRYMAN ROBERT T. WILCE GARRETT E. CROW RHODORA (ISSN 0035-4902). Published four times a year (January, April, July, and October) by The New England Botanical Club, 22 Divinity sho Cambridge, MA 02138 and printed by Allen Press, Inc., 1041 New Hampshire St., Lawrence, KS 66044. Second class postage ons I t Boston, MA and at agains sii offices. any and cytotaxonomy in their broader implications, will be consid- ered. Articles are subjected to peer review. RHODORA assesses page charges. po ciearab antes $40.00 per calendar year, net, postpaid, in funds yable at par in the United States currency at Boston. Remittances payable - ees Send to Treasurer, 0 Divinity Ave., Cam- bridge, MA 02138. MEMBERSHIPS Regular $35; Family $45. Application form printed BACK VOLUMES AND SINGLE COPIES: Some available; infor- mation and prices will be furnished upon request to the Treasurer. ADDRESS CHANGES: In order to receive the next number of RHO- DORA, changes must be received by the Treasurer prior to the first day of January, April, July, or October POSTMASTER: Send eorer changes to RHODORA, 22 Divinity Ave., Cambridge, MA 02138. INFORMATION FOR CONTRIBUTORS: Inside back cover, January and April. MANUSCRIPTS: Send to: — Y. Nickerso Managing Editor, "RHODORA Phippen-LaCroix Her Dept. of Biology, Tufts U Jaivessity Medford, MA 0215 Cover Illustration Rhexia virginica L., meadow beauty, is found from Nova Scotia to Georgia, but is rare at the northern limits of its range. The only northern o outlier of the istinctive 3-veined leaves and geniculate Tbodora JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 94 October 1992 No. 880 RHODORA, Vol. 94, No. 880, pp. 327-339, 1992 = CLIMATE AND THE DISJUNCT DISTRIBUTION OF POLYSTICHUM ALFARII (CHRIST) BARR.., COMB. NOV. IN MESOAMERICA DAvip S. BARRINGTON ABSTRACT Polystichum alfarii (Christ) Barr. is here recognized as a species with a wide disjunction from the montane regions of northern Mesoamerica to the mountains of Costa Rica and Panama. Morphological circumscription and a meiotic chro- mosome number of n = 41 bivalents are provided. Comparison of habitat and distribution with other Central American species of Polystichum suggests that the 620 km disjunction of P. alfarii, noteworthy for a lower-altitude Polystichum species in Mesoamerica, is related to i i fe fe lly dry forests at altitudes between 1300 and 1700 m. Key Words: Polystichum alfarii, plant nomenclature, disjunction, biogeography, climate and distribution, Mesoamerica INTRODUCTION Polystichum, though not the most diverse fern genus in Latin America, has presented unique problems in recognizing species boundaries. Many of these problems are biological: environmen- tally induced and developmental variation often obscures evo- lutionary relationships in the genus (Barrington, 1985). Species unrecognized for strictly historical reasons (taxonomic and no- menclatural) have also made it difficult to understand evolution- ary relationships in the genus. In 1980, Alan Smith described a distinctive new species of Polystichum from Oaxaca, Mexico, which he named after John Mickel, co-author of the recent fern flora for that region (Mickel and Beitel, 1988). Smith excluded a : 327 syxssouR! PO ANICAl 328 Rhodora [Vol. 94 small set of materials that differed from P. mickelii A.R. Smith in having a true indusium (A. R. Smith, 1981: 200). Recognition of these materials as a distinct species of the Mexican endemic center, also including a large series of plants from southern Me- soamerica, is the subject of this contribution. METHODS The work reported here is based on a morphological analysis of a set of about 60 collections (often represented by one or more duplicates) from the following herbaria: BM, CR, DUKE, F, GH, IJ, MO, NY, UC, US, VT, and YU. Thanks are extended to these herbaria for providing loans of materials for this project. This work was carried out in the course of preparing a treatment of Polystichum for the Flora Mesoamericana. In the course of on- going fieldwork in Costa Rica, I encountered the new species (P. alfarii) in two sites in the Valle Central; these sites have made additional work with living plants possible. I have not seen the plant in the field at any Mexican or Guatemalan sites. In January 1991, I collected a sample of 23 living sporophytes from the Cerro Tablazo population, which served as the material for a report of basic information on chromosome number and behavior in meiosis I. Cytological technique largely followed Bar- rington, 1990. RESULTS Systematic Section Polystichum alfarii (Christ) Barr. comb. nov. (Figures 1-10) Polystichum aculeatum (L.) Roth var. alfarii Christ, Bull. Herb. Boissier II. 4: 963 (1904). HoLotyre: Costa Rica, Alajuelito, 1300 m, March 1902, Alfaro 16471, P not seen; Isotype: us! Polystichum aculeatum (L.) Roth, var. flavidum Rosenstock, Repert. Spec. Nov. Regni. Veg. 22: 9 (1925). Syntypes: Costa Rica, Carthago, 1400 m, 16.1V.1910, Brade 555, uc!; Costa Rica, ge ae 10.1V.1908. Brade 26a, NY! Rhizome 2-6 cm diam., unbranched; proximal petiole scales lanceolate, center coriaceous, dark reddish brown, border char- taceous, orange-yellow, marginate throughout, the margin irreg- 1992] Barrington—Polystichum Figures 1-10. Polystichum alfarii (Christ) Barr. 1. Median pinna, Barrington 1242 (vt). 2. Basal pinna, Barrington 1242 (vt). 3. Basal petiole scale, Steyermark 35194 (F). 4, Detail of edge of basal petiole scale near the center as indicated in Figure 3. 5. Distal petiole scale, Steyermark 35194 (F). 6. Small scale from petiole, Alfaro 137 (us). 7 his, Steyermark 35194 (us). 8. Flange of true indusium from above, Alfaro 137 (us). 9. Larger, more dissected pinnule, Steyermark 35194 (r). 10. Smaller, less dissected pinnule, Jiménez 395 (CR). ular, with a few short rigid to flaccid cilia throughout; distal petiole scales similar, but broad-lanceolate, uniformly papyraccous, Or- ange-yellow or reddish orange with an orange-yellow border, con- form or marginate except at base; small scales of petiole mostly Ovate-caudate or long-lanceolate (less often deltate or broad-lan- ceolate), color various but generally dark, marginate at least at base; lamina 37-100+ cm, 2-pinnate-crenate to 2-pinnate-pin- natifid, acuminate: rachis without a proliferous bud, rachis scales light yellowish brown: pinnae 10-16 x 2-2.5 cm, attached at More or less right angles to rachis; pinna-rachis scales filiform or 330 Rhodora [Vol. 94 less often narrow-lanceolate, with a few basal cilia, rare elsewhere; pinnules mucronate, flat, the basal acroscopic pinnules longer than the next distal; veins free, spinules developed only at pinnule tip, weak; sorus round, terminal on vein; true indusium 0.5-0.7 mm diam., irregularly peltate with a margin of narrow, radiate cells and a medulla of regular, slightly elongate cells, most often brownish orange. The name Polystichum alfarii (Christ) Barr. honors Anastasio Alfaro, Costa Rican naturalist and first director of the Museo Nacional de Costa Rica, who made many important collections in the productive ‘““Golden Age” of work on the flora of Costa Rica (1885-10) (see Gomez P., 1977). Alfaro provided early in- sights into the character of the Costa Rican flora as well as a concise catalog of species (Alfaro, 1887). The type collection, made by Alfaro in March of 1902, is from Alajuelita—now a densely populated suburb just south of San José. Both Christ and Rosenstock recognized elements of Polysti- chum alfarii from Costa Rica as varieties of Polystichum aculea- tum (L.) Roth, in keeping with the then popular solution to the problem of recognizing clear species boundaries in the genus Po- lystichum. These varietal names were never consistently applied to Polystichum alfarii; the species has most often been left un- determined in herbaria. When collections of P. alfarii have been identified, they have been interpreted as any of several species with Mexican affinities, most often P. ordinatum (Kunze) Liebm. (a superficially similar species endemic to Mexico and Guate- mala), but also as P. muricatum (L.) Fée (a remotely allied species occurring sporadically throughout northern Latin American and the Antilles). A. R. Smith (1981: 200) first drew attention to materials ot fF. alfarii from Chiapas and Guatemala by setting them apart as unnamed collections, which he described as resembling P. mick- elii. Mickel and Beitel (1988: 315) suggested that this same set of plants from Chiapas and Guatemala might pertain to P. smithii Mickel & Beitel, since the two share characters of the indusium and petiole scales. However, P. smithii differs from P. alfarii in being a much smaller fern with conform (not strongly marginate) petiole scales and fertile pinnules attached at right angles to the pinna-rachis. Judging only from the history of collecting (see Ap- pendix), P. alfarii is rarer in the northern part of its range. At the SE 1992] Barrington—Polystichum 331 southern limit for the species, the Panama collections of P. alfarii have completely escaped recognition until now. The key features of Polystichum alfarii that allow identification are its small (0.5—0.7 mm diam.) but clearly marginate indusium, its strongly marginate but not ovate-caudate or basally thickened petiole scales with dark reddish brown centers and orange-yellow borders, and its lustrous, mucronate pinnules attached obliquely to the pinna-rachis. Most prominent among features varying with- in the species is lamina dissection. The leaves vary from 2-pinnate crenulate to 2-pinnate pinnatisect, in contrast to most Meso- american species, in which leaf dissection of reproductive adults is relatively constant. The complex variation of this species and the superficially similar P. mickelii is best represented in a di- chotomous key to the two taxa, especially because of the unusual petiole-scale variation in Polystichum mickelii. KEY TO POLYSTICHUM ALFARII AND P. MICKELII 1. Proximal petiole scales strongly marginate, edge irregular, without superficial cilia, marginal cilia rare, not darkened; distal petiole scales strongly marginate, edge irregular, with a very few short, flaccid cilia; indusium present. 0.50.7 mm P. alfarii eS Sle She See eee we ee eee & ed cer BI Be) a eee P. mickelii, form with superficial cilia Ecology Unlike almost all Polystichum species in Mexico and Central America, this species is most common in moist forests, not wet R32 Rhodora [Vol. 94 forests and rain forests (moisture adjectives are used in the quan- titative sense of Holdridge, 1967); it has been reported from moist and dry forests and it persists in remnant forests. Among the few Polystichum species with a relatively low altitude preference, P. alfarii has been encountered between 1140 and 2200 m: about three-quarters of these collections are from between 1300 and 1700 m, with an overall mean altitude of 1540 m. Hence, this species is almost always found below the altitudes where cloud cover itself strongly influences climate (Troll, 1968). Only three out of the 22 species of Polystichum known from Mesoamerica (Barrington, 1993) occur at these lower altitudes. A typical Po- lystichum, P. alfarii shows a marked preference for the slopes above streambanks in steep terrain, perhaps indicative of a ten- dency for gametophytes to pioneer on the moister, unstable slopes near streams. The two study sites where I have seen this species in Costa Rica provide further insight into its ecology and hence its distribution. t the Cerro Tablazo site above Tablon, in the central valley of Costa Rica (9°50’N, 84°01'W), a large population including plants in numerous size classes occupies an area of at least several hectares in the shade of an evergreen forest at 1600 m, near the median altitude for the species. There, plants are most frequent on very steep (40-60°) slopes in sight of and occasionally at the edge of the swift-moving Rio Purires. Younger plants are es- pecially frequent along vertical exposures of bare soil presumably persisting from past stream disturbances and slope wastage at the site. Older plants (often with leaves in excess of 1.5 m long) do not show the same preference for earthen exposures; presumably the disturbances they originally occupied have since been stabi- lized and obscured. Climate at the Tablon site is markedly different from that typ- ical of Polystichum sites in Costa Rica— Herrera (1985) maps the site as “‘clima subhumedo, caliente, con una estacién seca modera- da (35-70 dias con déficit de agua).”” The most frequent com- panion pteridophyte at the Cerro Tablon site is Anemia phyllitides (L.) Swartz, a species able to withstand considerable water stress. In contrast, typical Costa Rican sites for Polystichum species pres- ent much less demanding water-availability regimes. In addition, most sites provide less heat. Herbarium-sheet data suggest that P. alfarii is commonly found in regimes such as that encountered at Tablon. By contrast, I have also seen Polystichum alfarii at Porrosati 1992] Barrington — Polystichum 333 (10°06'N, 84°05’W), on the slopes of Volcan Barba, one of the series of volcanoes lying north of the central valley in Costa Rica. There, at the highest site at which the species has been encoun- tered, the species is extremely rare—only one immature individ- ual was seen in January, 1991. The Porrosati site is a 5 m-deep ravine cut in recent volcanic sediments by the Rio Ciruelas, which is surrounded by agricultural land cleared from cloud forest. The plants seen there all grew on the earthern walls of the ravine in deep shade. The poorer performance of this species at this higher, wetter site again suggests that, unlike other Costa Rican species, P. alfarii is a plant of drier, warmer forests. These forests, on the gentler lower slopes of the montane regions of the country and near the major population centers, are among those most dis- turbed by the human population. Distribution The most striking feature of Polystichum alfarii’s distribution is a range disjunction of 620 km, from the volcanic slopes above San Ramon in Costa Rica to Volcan de San Vicente in El Salvador (Figure 11). Similar disjunctions characterize the Central Amer- ican distribution of three other Polystichum species (Table 1). The distributional hiatus defines two centers of distribution, one northern (Chiapas-El Salvador) and the other southern (Costa Rica—Panama). Also evident is a distribution unique for Poly- stichum along the Pacific (lee) slopes of the two montane regions. The single exception to the Pacific-slope rule is the set of now threatened central-valley populations in Costa Rica, where rela- tively dry moist-forest conditions are enforced in the lee of a volcanic range. Cytology Two plants from the Tablon population yielded counts of n= 41 bivalents at Meiosis I prophase (Figures 12, 13). These counts are based on four sporocytes for Barrington 1976 (vt) and two sporocytes for Barrington 1991 (VT). Hybridization William R. Maxon has collected the hybrid between Polysti- chum alfarii and P. platyphyllum (Willd.) Presl, a widespread 334 Rhodora [Vol. 94 Table 1. Range disjunctions in the widespread Mesoamerican species of Polystichum (order is altitudinal range). aximum Location of Typical Disjunc- Disjunction (or Species Altitudes Habitat tion Notes on Range) P. platyphyllum 600-1500 m_ Tropical & — Missing from premontane Guatemala wet forests and El Salva- dor P. alfarii 1100-2200 m_ Premontane 620 km_ El Salvador - moist forest Costa Rica P. muricatum 1100-2700 m_ Lower mon- se Missing only ane wet for- from El Sal- vador P. hartwegii 1200-2700 m_ Lower mon- _ Present tane wet for- throughout est Mesoamerica P. fourniericom- 1900-3100 m Montane rain 550 km Honduras - plex fe Costa Rica P. speciosissi- 2600-3500 m High-montane 1010 km Guatemala - rain paramo Costa Rica P. orbiculatum 3200-3600 m High-montane 1070 km Mexico - Costa rain paramo Rica exindusiate species of lower-altitude wet and rain forests, from the east end of the Central Valley in Costa Rica. (Collection data: Costa Rica, Pr. Cartago, vicinity of Santiago, humid forest floor, 1050 m, April 20, 1906, W. R. Maxon, 100, us 575664!.) This hybrid combines the diagnostic subapical bud and attenuate apex of P. platyphyllum with the small indusium of P. alfarii. Sporangia of this hybrid rarely open and are largely collapsed. Altitudinally the hybrid is from near the median altitude for P. platyphyllum and just below the lowest altitude recorded for P. alfarii (Table L}. — Figure 11. Distribution of Polystichum alfarii (Christ) Barr. Contour intervals are 1000 m. Scale bar is 100 km. Only a few representative collections from the Valle Central of Costa Rica are plotted. Localities are at tip of line opposite the circle. Barrington— Polystichum 335 1992] } oe WOOT - 0 [| WQOOT - “OOOT WQOOE - U000Z (HH + WQ00E x 336 Rhodora [Vol. 94 13 Figures 12 and 13. Meiotic chromosome number for Polystichum alfarii (Christ) Barr. Two cells in prophase of meiosis I, n = 41 pairs, D. S. Barrington 1976 (v7). DISCUSSION Comparison of the disjunct distributions of Central American Polystichum species (Table 1) reveals a relationship between al- titude, rainfall regime, and the precise nature of the disjunction. The greatest disjunctions are characteristic of the species with the highest altitude preferences, e.g., P. orbiculatum (Remy) Fée and P. speciosissimum (Kunze) Tryon & Tryon; both species are con- fined to alpine or high-montane forest habitats that are limited in distribution in Mesoamerica. In contrast, two species with altitudinal ranges much like P. alfarii’s show no significant dis- junctions because of the relative continuity of lower montane wet and rain forests where these species are common. Polystichum alfarii shows a disjunct distribution like that of evolutionary lin- eages such as the P. fournieri Smith species group from higher (cooler), more island-like montane habitats of Central America, presumably because of its adaptation to drier forests than the other species at equivalent altitudes. Tryon (1972) portrayed fern species diversity in tropical Amer- ica in the context of endemic centers. These centers evidence high total numbers of species and high percentages of endemic species; they are separated from others by regions poor in total species and endemic species. Several of his inquiries revealed patterns consistent with an origin of diversity via an evolutionary migra- tion scheme, in which lineages reached different endemic centers 1992] Barrington— Polystichum 337 via long-distance dispersal and then diverged from source species in the new region (e.g., Sphaeropteris; Tryon, 1971). This scheme is the continental equivalent of inferred histories for insular en- demics such as the genus Diellia in Hawaii (Wagner, 1952). Understanding the process by which species diverge in tropical montane regions is dependent on the resolution of species and vicariant species pairs showing disjunct distributions between en- demic centers; Polystichum alfarii is such a species. More gen- erally, the genus Polystichum is of particular interest as a system for the study of the process of evolutionary migration because it is a montane group including a diversity of species with a variety of climatic preferences. The apparent relationship between this variety and geographic distribution suggests that the genus has the potential to provide insight into the determinants of successful diversification in endemic centers. ACKNOWLEDGMENTS The University of Vermont provided support for field work in Costa Rica through its Institutional Grant #BSCI90-8. Cathy Par- is contributed valuable insights and a critical review. An anon- ymous reviewer added substantially to the quality of this work. LITERATURE CITED ALFARO, A. 1887. Lista de las plantas encontrada hasta ahora en Costa Rica y en los territorios limitrofes, extractada de la “Biologia Centrali Americana” Anales Inst. Fis.-Geogr. Nac. Costa Rica 1: 1-101. BARRINGTON, D.S. 1985. The present evolutionary and taxonomic status of the fern genus Polystichum: the 1984 ae ioe of America Pteridophyte Section symposium. Amer. Fern. J. 75: ——.. 1990. Hybridization and moral in Cental cytological and isozyme documentation. Ann. Missouri 305. 1 American Polystichum: Bot. Gard. 77: 297- 1993. he arg Tre Re @ Moran and R. Riba N., Eds. Flora Me- Gomez P., L. D. 1977. Contribuciones a la Pieridologia costarricense XI. Her- 13; 25— HERRERA, =o 1985. Clima de Costa Rica. Editorial Universidad Estatal a Dis- tancia, San José, Costa Rica. (Vegetacion y Clima de Costa Rica, Ed. L. D. Gomez P., Vol. 2. Hovpripce, L. R. 1967. Life Zone Ecology. Tropic Costa Rica al Science Center, San José, 338 Rhodora [Vol. 94 Micke, J. T. AND J. M. BEITEL. 1988. Pteridophyte Flora of Oaxaca, Mexico. Mem. New York Bot. Gard. 46: 1-568. Smitn, A. R. 1980. New taxa and combinations of Pteridophytes from Chiapas, Mexico. Amer. Fern. J. 70: 15-27. . 1981. Part 2. Pteridophytes. Jn: D. E. Breedlove, Ed., Flora of Chiapas. California Acad. Sci., San Francisco. TROLL, C. 1968. Geo-ecology of i Regions of the Tropical Amer- icas. F. Diimmlers Verlag, Bonn. Tryon, R. M., Jr. 1971. The American tree fe llied to Sph is horrid Rhodora 73: 1-18. . 1972, Endemic areas and geographic speciation in Tropical American ferns. Biotropica 4: 121-131. Wacner, W. H., Jr. 1952. The fern genus Diellia. Univ. Calif. Publ. Bot. 26: 1-212. PRINGLE HERBARIUM DEPT. OF BOTANY AND BIOCHEMICAL SCIENCE UNIVERSITY OF VERMONT BURLINGTON, VT 05405-0086 APPENDIX Exsiccatae for Polystichum alfarii (Christ) Barr. »MEXICO. ‘Chiapas: Finca Irlanda. C. A. Purpus 7231, 7232 (us), 7243 (GH, MO, us), 7243A (BM). ~GUATEMALA.- Dept. Quetzaltenango: along Rio Samala, between Santa Maria de Jesus and Calahuaché, 1200-1300 m, J. A. Steyermark 33834 (F, us); slopes of Volcan Santa Maria, 1 mi. below Santa Maria de Jesus, 1520 m, J. A. Steyermark 34367 (F); Finca Pirineos, between Santa Maria de Jesus and Calahuaché, 1350 m, J. A. Stevermark 35194 (F). VEL SALVADOR’ Pr. Ahuachapan: Sierra de Apaneca, region of Finca Colima, P. C. Standley 20135 (Gu, us)! Pr. San Vicente: Volcan San Vicente, 1450 m, W. Létschert 84 (us); Volcan San Vicente, P. C. Standley 21489 (us). Y¥ COSTA RICA:'Pr. Alajuela: La Ventolera, S slope of Volcan Poas, 1700 m, P. C. Standley 34646 (us).’Pr. Cartago: Cartago, 1300 m, A. Alfaro s.n., dist. as Hi Donnell Smith 6961 (us); just above Tablén on road up E side of Cerro Tablazo, 1600 m. 9°50’N x 84°01’W, D. S. Barrington 1242 (cr, vt), 1976, 1991 (vT); Carpintera, 1700 m, A. Brade & A. C. Brade 26 (Gu); Malavassi coffee finca, S of Tres Rios, 1200 m, F. M. Givens 3195 (F); in Lankester Garden, 4750’ (native fern), E. Scamman 5925 (Gu); Pacayas, foot of Volcan Turrialba, 1400 m, E. Scamman 7080 (us); above Tres Rios, 5000’, E. Scamman & L. R. Holdridge 7945 (us); Cerro de la Carpintera, 1500-1850 m, P. C. Standley 34227 (us); Dulce Nombre, 1400 m, P. C. Standley 35956 (us); La Carpintera, 1800 m, R. Torres R. 170 (us); Tablazo, 1877 m, M. Valerio on 8-1-27 (cR 33362); Hacienda Linda 1992] Barrington—Polystichum 339 Vista, just S of Dulce Nombre de ee SE of Cartago, 1300 m, E. J. Judziewicz 4370 (cr, Mo). Pr. Herédia: 1.7 km E of Porrosati, on the Rio Ciruelas, 2200 m, 10°06'N x 84°05'W, D. S. oe 1213 (cr, vt). Pr. San José: Rio Herradu- ra, tributary of the Rio Chirrip6 del Pacifico, NW of Canaan, General Valley, 1600 m, 9°29'N x 83°37’W, W. C. Burger & R. L. Liesner 7084 (cr, F); Hacienda La Verbena, Rio Tiribi, Alajuelita, 1140 m, O. Jiménez L. 395 (cR, Us). Rio Torres: San Francisco de Guadalupe, O. Jiménez 360 (cr), San Francisco de Guadalupe, 1150 m, H. Pittier 7153 ice: us), 7155 (cr); Cerro de Piedra Blanca, above Escasa, P. C. Standley 32501, 32533 (us), San Francisco de Guadalupe, A. Tonduz 9837 dist. as J. Donnell Smith 7216 (Gu, us); El Copey, Rio Pedregoso, 1800 m, 4. Tonduz 11853 p.p. (cr; the us collections of this number are P. concinnum Lellinger ex Barr A. Pr. Chiriqui: vic. “New Switzerland,” ag valley of Rio Chiriqui Viejo, 1800-2000 m, P. H. Allen 1334 (us); Guadalupe, Cerro Punta, 2000 m, R. Caballero 149 (mo); between Holcomb’s trail and Meutice plantation, above El Boquete, 1500-1725 m, E. P. Killip 5086 (us); Piedra de Lino, El Boquete, 1400-1600 m, E. P. Killip 5410 (us); N side of Rio Caldera, Horqueta - Bajo Mano, NW of Boquete, J. L. Luteyn 4581 (DUKE); above Sabana de El Salto, trail to Camp Aguacatal, E slope of Volcan Chiriqui, 1500-1750 m, W. R. Maxon 5267 (us); km 10.7 on Volcan-Cerro Punta Rd., G. R. Proctor 32005 (u). RHODORA, Vol. 94, No. 880, pp. 340-347, 1992 A NATURALLY-OCCURRING POPULATION OF PUTATIVE ARISAEMA TRIPHYLLUM SUBSP. STEWARDSONII x A. DRACONTIUM HYBRIDS IN MASSACHUSETTS LAuRIE L. SANDERS AND C. JOHN BURK ABSTRACT A 1 : c oe: oo Oe - fame spay FF (T Cp} hk fb \ ‘oe stewardsonii (Britt. ) Huttleston and A. dracontit (A id in a disturbed floodplain forest in Hampshire ae Northampton, MA di in ile sit spring, 1988. The hybrids are intermediate Lekweees the parental oe in leaflet number, spadix length, spathe length They also ne apparent hybrid vigor and occur in a near-classic example of a hybrid habita Key Words: Natural hybrid, Arisaema I] b stewardsonii, A. dracon- tium, intermediacy, hybrid vigor, hybrid habitat, Hampshire County, Massachusetts INTRODUCTION A population of apparent hybrids between Arisaema triphyllum (L.) Schott subsp. stewardsonii (Britt.) Huttleston and Arisaema dracontium (L.) Schott (Araceae) was encountered in June, 1988, in the course of a distributional study of Arisaema dracontium (Sanders, 1989). The putative hybrids were identified by leaf types and spadix lengths that are intermediate between those of 4. triphyllum subsp. stewardsonii and A. dracontium. The hybrids tended to be taller than plants of either putative parent and were morphologically unique. Both A. triphyllum subsp. stewardsonii and A. dracontium were present and flowering in the immediate vicinity. Arisaema triphyllum subsp. stewardsonii is one of the three major subspecies of eastern North American jack-in-the- pulpits that comprise the Arisaema triphyllum complex (Huttleston, 1948, 1949, 1953: Kartesz and Kartesz, 1980; Treiber, 1980.) It has one or two trifoliolate leaves, prominent ridges or fluting on the outer spathe, and a cylindrical spadix that is enclosed within the spathe. The throat of the spathe is marked with green, whitish, or purple-maroon striping and the spadix is pale-green or pur- plish-maroon. 340 | 1992] Sanders and Burk— Arisaema 341 Figure 1. Representative specimens: Arisaema triphyllum subsp. stewardsonli (left), a putative hybrid (center), and A. dracontium (right) collected June, 1988. Arisaema dracontium, green dragon, has a single pedately 7- 17 foliolate leaf, a smooth spathe, and a relatively lengthy, ex- serted slender spadix (Huttleston, 1948, 1953). The spathe is uniformly lime-green in color and the spadix bright yellow. The apparent Arisaema hybrids are distinguished by their single pedately 5-7 foliolate leaves and their spadix lengths, which are intermediate between the shorter spadices of A. triphyllum subsp. stewardsonii and the longer ones of A. dracontium (Figure 1). Their Outer spathes are purplish green and the inner spathes patterned with green and maroon; their spadices are yellow-green with ma- Toon checkering. Although Arisaema triphyllum subsp. stewardsonii and A. dra- contium occur sympatrically throughout much of the northeast Portion of their ranges (Huttleston, 1948, 1953), reports of natural hybridization between any subspecies of 4. triphyllum and A. dracontiumare virtually nonexistent. Thompson (1911) described a colony of A. triphyllum growing in a central Connecticut swamp that “showed peculiarities that suggested its close relationship to A. dracontium.” The plants had a strict growth habitat, slender spathes that were atypically erect at their summits, and five-parted leaves. Huttleston (1948, 1953), working near Varna, New York, 342 Rhodora [Vol. 94 observed a number of plants of A. triphyllum which possessed slender spathes that were held more erect than usual and lateral leaflets that tended to be lobed, sometimes to the midrib. He concluded that this variation may possibly have resulted from past hybridization with A. dracontium, but that present evidence for this was unconvincing. Similarly, Pittillo et al. (1972) noted that a previous report of A. dracontium from Henderson Co., NC represented “‘a peculiar form (possibly a hybrid) of A. triphyllum.” Records of artificial crosses between Arisaema triphyllum and A. dracontium are limited to brief mentions of artificial hybrids by MacDougal (1901) and Rennert (1902), both of whom describe the germination and seedling characteristics of the parent species but do not describe either how the hybrid cross was made or the appearance of the mature hybrid. The purpose of the present study is to characterize the flood- plain site at which probable hybridization occurred, to describe the putative hybrids more fully, and to compare the hybrids with nearby plants of Arisaema triphyllum subsp. stewardsonii and A. dracontium. THE STUDY SITE The population of putative hybrids, which totals approximately 800 plants, occurs within a narrow, two hectare strip of floodplain forest along the Connecticut River in Northampton, Hampshire Co., MA. The hybrids occur together in clusters of 1 0-20 flowering culms, often accompanied by an equal number of depauperate vegetative stems, at elevations that are normally flooded several times per year. The underlying soils are Pootatuck fine sandy loam and Limerick silt loam (Swenson, 1981). An additional population containing approximately 1000 putative Arisaema hy- brids was discovered during summer, 1992 in a similar floodplain forest about 1 km distant from the original site. The study site has been severely affected by natural distur- bances, including the formation of an oxbow on the main stem of the Connecticut River in 1840, and by human activities, in- cluding the rerouting of an adjacent tributary of the Connecticut River as early as 1710 (Robinton and Burk, 1971; Holland and Burk, 1982, 1984). Railroad and highway constructions have also occurred in the vicinity, and agricultural activities continue on land adjacent to the study site. 1992] Sanders and Burk— Vaccinium is the most likely, although admittedly decisive textual proof is wanting. ACKNOWLEDGMENTS I thank B. Verstraete for his careful perusal of this manuscript; his expert comments were much appreciated, nevertheless any remaining errors and omissions are strictly mine. LITERATURE CITED BRowninG, R. 1983. Medieval and Modern Greek, 2nd ed. Cambridge Univ. Press, Cambridge, pp. 26-31. Drypen, J. 1880. Works of Virgil. American Book Exchange, New York, NY. FaircLoucu, H. R. 1960. Virgil, Vol. 1. Eclogues, Georgics, Aeneid I-VI. Har- vard University Press, Cambridge, MA. FERNALD, M. L. 1950. Gray’s Manual of Botany, 8th ed., American Book Com- pany, New York, NY. Gare, P. G. W. 1982. Oxford Latin Dictionary. Fascicle § Sopor-Zythum. Clarendon Press, Oxford Gray, A. 1848. Manual of Botany. James Munroe and Co., Boston and Cam- bridge, MA. Haapsma, R. A. AND J. NUCHELMANS. 1963. Précis de Latin Vulgare. Walter, Groningen, Nederland, pp. 26-31. JOHNsoNn, T. 1636. Gerards Herbal. Norton and Whitakers, London McGratu, J. W. 1977. Dyes from Lichens and Plants. Van Nostrand Reinhold Ltd., New York, NY. Pace, T. E. 1960. Virgil: Bucolics and Georgics. MacMillan and Co. Ltd., Lon- don REHDER, A. 1927. Manual of Cultivated Trees and Shrubs. MacMillan Co., New York, NY. Smitu, J. E. 1817. Vaccinium. In: Rees’s Cyclopedia, Vol. 36. Longman, Hurts, Rees, Orme and Brown, London. BIOLOGY DEPARTMENT ACADIA UNIVERSITY WOLFVILLE, NOVA SCOTIA CANADA BOP 1X0 RHODORA, Vol. 94, No. 880, pp. 374-380, 1992 ECOLOGICAL ASPECTS OF ARETHUSA BULBOSA, CALOPOGON TUBEROSUS AND POGONIA OPHIOGLOSSOIDES (ORCHIDACEAE) IN EASTERN NEWFOUNDLAND. II. PARTITIONING OF THE MICROHABITAT J. Topp BOLAND AND PETER J. SCOTT ABSTRACT Three sympatric species of peatland orchids were investigated for differences in their ability to partition microhabitats. Species distributions at the study sites reflected differences in the hydrology and microtopography, indicating microhab- itat partitioning. Pogonia was restricted to the wettest areas while Ca/opogon and Arethusa were found on drier sites. Key Words: Peatland orchids, Arethusa, Calopogon, Pogonia, microhabitat, New- foundland INTRODUCTION Peatlands are a significant vegetation type of Newfoundland, with 18 of the 32 species of Newfoundland orchids inhabiting these peatland environments. Various plants within a particular peatland may appear to be evenly distributed; however, upon closer inspection, the species is seen to have a specific distribution pattern. This pattern often reflects microhabitats within a peat- land. Differences in the growth rates of various peatland plants, es- pecially Sphagnum spp. and sedges, create a hummocky terrain. These hummocks often remain above the watertable year-round, while in hollows the watertable is usually at the surface. The species within a peatland community are often distributed relative to the watertable. Thus, while cranberry (Vaccinium macrocarpon Ait.) and sheep laurel (Ka/mia angustifolia L.) commonly co- occur on a particular peatland, the former species is distributed near the watertable while the latter is often confined to the drier hummocks (Wells, 1981). Arethusa bulbosa L., Calopogon tuberosus (L.) BSP and Pogonia ophioglossoides (L.) Ker are three sympatric Newfoundland or- chids which frequently grow on the peatlands of the Avalon Pen- insula (Luer, 1975). This study was undertaken to determine how these three species 374 1992] Boland and Scott— Orchid Ecology als are distributed when they co-occur. Do they partition the micro- habitats of the peatlands they inhabit? Microhabitat partitioning is defined in this study by the species’ distributions in relation to hydrology and microtopography of the peatland. METHODS AND MATERIALS Three study sites were located about 40 km west of St. John’s, Newfoundland in the vicinity of the Witless Bay Line. Site 1: elevation 216 m; 47°22’00’N; 53°02’51”W; site 2: elevation 231 m; 47°20'35”N; 52°59'14”W; site 3: elevation 197 m; 47°20'19’N; 52°55'52”W. Two study sites were slope bogs, the most common peatland type on the Avalon Peninsula (Wells, 1981). These tree- less bogs are rarely more than 2 m deep, with slopes from 5-15%. The watertable is usually at or near the surface. Underlying bed- rock is mostly acidic; seepage waters are generally nutrient-poor. The third study site was a ribbed fen, an uncommon peatland type on the Avalon Peninsula (Wells, 1981). These severely-wind- swept peatlands occur on exposed slopes, usually over 200 m in elevation. They are treeless and shallow, mostly less than | m in depth. Ribbed fens may contain exposed rock outcrops and glacial erratics. Flashets or small pools in ribbed fens are common and are oriented at right angles to the slope. These fens are generally more minerotrophic than slope bogs as the nutrients are released by high soil-frost activity and continual downslope water move- ment (Wells, 1976). Field work was conducted from June to September, 1988. A 10 m2 plot was delineated at each site and each plot was subdi- vided into 400 0.5 m? quadrats. Arethusa, Calopogon and Pogonia plants were counted and mapped, and elevation above the wa- tertable was determined in each quadrat. . Microtopography of each plot was determined. Mean height above the watertable for the orchids was estimated with the sur- face of the nearest flashet taken to be height zero. This estimation was based on the quadrat’s mean height for all quadrats containing at least one orchid species. Two peat samples were taken from each plot. Since Pogonia was noted to occur separately from Arethusa and Calopogon, one sample was taken near a group of Pogonia and the other near a group of Arethusa and Calopogon. Cores were analyzed for nu- trient content, pH and water content. 376 Rhodora [Vol. 94 To determine if two orchid species within a site were distributed separately from one another, 0.5 m?* quadrats that contained at least one individual were examined by a z-test (Wetherill, 1967). A z-score of 1.97 or higher was taken to indicate that the observed number of individuals of a species was significantly higher than expected for that quadrat. If over 5% of the total quadrats within a site showed significant z-scores, then the two species were taken to be significantly separate from each other. RESULTS Analysis of peat samples showed that the nutrient content was variable from site to site but pH and percent of available nitrogen were less variable. There was no significant relationship between the distribution of the three orchid species and availability of nutrients. Based on z-scores, in sites 1 and 3, Pogonia was significantly clumped apart from Arethusa and Calopogon; Calopogon and Arethusa overlapped in their distribution. In site 2, none of the orchids was significantly separate from the others. Mean water content of the substrate near Pogonia was 92.4% (range = 91.3-93.5%) at all sites. In sites 1 and 3, Pogonia was found to be restricted to the perimeters of flashets which had gently sloping sides (a slope of 1-2%). Site 2, however, had Po- gonia abundant throughout, distributed from 1 to 11 cm above the watertable (mean = 4.9 cm, SD = 1.96). The frequency of quadrats containing Pogonia reached a maximum for those quad- rats which were 5 cm above the watertable (Figure 1). Calopogon and Arethusa were found close to steep-sided flash- ets and on the sides and tops of peat hummocks, especially at sites 1 and 3. At site 2, Arethusa was restricted to the few hum- mocks present but Ca/opogon was found in many of the depres- sions as well. The mean water content near these species was 78.2% (range = 75.4-80.7%). Microtopography of site 1 had elevations ranging from 0 cm (surface of the flashets) to 36 cm above the watertable. Orchids of this site were restricted to 1 to 29 cm above the watertable. Sites 2 and 3 had elevations above the watertable ranging from 0 to 32 cm and 0 to 36 cm respectively. At site 2, orchids were restricted to 1 to 23 cm above the watertable while at site 3, they were distributed from 1 to 25 cm. ee el 1992] Boland and Scott— Orchid Ecology ATT MMB Arethusa Calopogon Pogonia n 100 - = u m b e 80 |-----—— a as ee — r fe) — f 60 aie’ _ a i —_ aaa sear — oe-prapcoa > oO T | | 9 1 9 21 23 25 27 29 height above watertable (cm) Figure 1. Frequency of quadrats containing Arethusa, Calopogon or Pogonia at various heights above the watertable. Arethusa was found from 5 to 25 cm above the watertable (mean = 12.1 cm, SD = 4.72). The frequency of quadrats containing Arethusa reached a maximum for those quadrats which were 9 cm above the watertable (Figure 1). Calopogon ranged from 5 to 29 cm above the watertable (mean = 12.9 cm, SD = 5.96). The two orchids had a bimodal distri- bution with the frequency of quadrats containing them peaking at 7 and 17 cm above the watertable (Figure 2). DISCUSSION Analysis of the nutrient content in areas of Pogonia, Calopogon and Arethusa within each site did not show any patterns. Distri- bution of orchids in these particular sites did not appear to be influenced by the availability of nutrients. The distribution of Pogonia individuals in relation to the mi- crotopography showed them to be mostly in the wettest areas of the three sites (mostly 3 to 9 cm above the watertable). Ca/opogon and Arethusa individuals were found close to steep-sided flashets 378 Rhodora [Vol. 94 - Arethusa Calopogon 30 +~oO -OOR3CS Qeopnanmcecno 1 ao -s ¢£ 5 11 8% 6 17 0 21 23 25 27 2 height above watertable (cm) Figure 2. Frequency of quadrats containing Arethusa or Calopogon at various heights above the watertable. but were mostly distributed in somewhat drier areas on the sides and tops of peat hummocks. Case (1964) found all three species co-occuring in the Great Lakes region. He noted Pogonia as being most abundant in the wetter bogs and swamps, Arethusa frequently growing on sedge tufts among flashets as well as on isolated sphagnum hummocks, and Calopogon being common on the edge of hummocks and around the base of stunted spruce. Petrie (1981) also noted that while these three orchids often co-occur, Pogonia was usually most common in the wettest part ofa bog, and its orchid companions, although nearby, were mostly distributed in somewhat drier situations. The distribution pattern of the orchids in this study can be partly explained by physical features of the orchid’s root-systems. Pogonia individuals have a fine, fibrous root-system which grows paige thus Pogonia is not tolerant of drought conditions (Luer, 1975). Gentle shorelines of flashets and low depressions are < areas of a peatland which have the least possibility of becoming droughty. Calopogon and Arethusa have root-systems which bear small, rounded bulbous corms (Luer, 1975) which can act as water-storage organs. Thus, these orchids apparently can 1992] Boland and Scott— Orchid Ecology 379 tolerate the drier conditions found on the sides and tops of peat hummocks. Statistics from this study show the distribution of Pogonia to be separate from Calopogon and Arethusa. The latter two orchids overlap in their distribution; mean height above the watertable for Calopogon and Arethusa was 12.9 cm and 12.1 cm, respec- tively. However, Calopogon and Arethusa were not evenly dis- tributed throughout their overlapping microhabitat. Both orchids appear at 5 cm above the watertable. Below 9 cm and above 15 cm, Calopogon is the most abundant, while between 9 and 15 cm above the watertable, Arethusa is more common. While the dif- ferences in the distribution of Ca/opogon and Arethusa are subtle, the pattern of distribution found in this study suggests that Cal- Opogon is more tolerant of wetter and drier conditions than is Arethusa. Factors other than hydrology of these peatlands may be affecting distributions of Calopogon and Arethusa in particular. Further investigation is necessary to ascertain what other factors might affect the distribution of these orchids. This study suggests that in eastern Newfoundland, hydrology is a major factor affecting partitioning of microhabitats for these three sympatric orchids. Arethusa, Calopogon and Pogonia have overlapping distributions throughout the northeastern United States and eastern Canada from the Great Lakes to Newfound- land. While hydrology may be a major factor affecting the par- titioning of microhabitats for the three orchids throughout their geographic range, other factors may also affect their distribution within a given peatland. These other factors may include the hydrology pattern of a Particular peatland type. Watertable patterns may vary consid- erably between a ribbed fen and a basin bog (Wells, 1981). Dif- ferent plant associations within a peatland site have different key species, which may prevent one or more of the orchids from becoming established. For example, Sheviak (1974) found Po- gonia to be absent from areas where leatherleaf, Chaemadaphne calyculata, was present. Throughout North America, much of the emphasis on orchid ecology is based on their pollination ecology (Thien and Marcks, 1972). Orchids which co-occur ata particular site are often studied to determine differences in their breeding systems. Few studies have been done on how the orchids co-exist, based on the ecology 380 Rhodora [Vol. 94 of their particular site preferences. Such factors affecting the par- titioning of microhabitats between sympatric orchids should be the subject of further investigations within the Orchidaceae. ACKNOWLEDGMENTS We thank Edward Woodrow of the Land Resource Centre for his help with identifying the peatlands and organic profiles of the sites, Don Trenholm and his staff at the Provincial Department of Forestry for chemical analysis of the peat samples, Terry Hed- derson for statistical advice and for identifying the mosses, and Roy Ficken for the photographic preparations. LITERATURE CITED Case, F. W. 1964. Orchids of the Western Great Lakes Region. Cranbrook Institute of Science., Bloomfield Hills, MI. Luger, C. A. 1975. The Native Orchids of the United States and Canada (ex- cluding Florida). New York Botanical Garden, Bron Petrie, W. 1981. Guide to Orchids of North America. aesit House Pub- lishers Ltd., Vancouver, B.C. SHEviAK, C. J. 1974. An introduction to the ecology of the Illinois Orchidaceae. Illinois State Museunt Scientific Pape me Urbana, I THIEN, L. B. AND B. G. Marcxks. 1972. The floral aes af Arethusa bulbosa, Calopogon tuberosus and Pogonia Pins (Orchidaceae). Canad. J. Bot. 50: 2319-2325. Weis, E.D. 1976. A classification of peatlands in Eastern Newfoundland. M.Sc. thesis, Memorial University, St. John’s, Newfoundland. 1981. Peatlands of Eastern Newfoundland: distribution, morphology, vegetation and nutrient status. Canad. J. Bot. 59: 1978-1997 WETHERILL, T. D. 1967. Elementary Statistical Methods. Methuen, London. DEPARTMENT OF BIOLOGY MEMORIAL UNIVERSITY OF NEWFOUNDLAND ST. JOHN’S, NEWFOUNDLAND CANADA AI1B 3X9 RHODORA, Vol. 94, No. 880, pp. 381-382, 1992 TWO NEW COMBINATIONS AND A NAME CHANGE FROM THE LONG EXPEDITION OF 1820 GEORGE J. GOODMAN AND CHERYL A. LAWSON ABSTRACT A new combination is provided for a Scutellaria and an Engelmannia. A name change in Mimulus is explained. Key Words: Scutellaria, Engelmannia, Mimulus, Long Expedition The Long Expedition to the Rocky Mountains in 1820 included on its scientific staff Edwin James, M.D., as botanist, geologist and surgeon. In our work on the botany of this expedition, we find it necessary to make some new combinations and a name change preparatory to our extensive publication on the expedi- tion. Scutellaria parvula Michx. var. missouriensis (Torr.) Goodm. & Laws., comb. nov. #338 S. ambigua Nutt. 8 missouriensis Torr., Ann. Lyc. Nat. Hist. N.Y. 2: 232.1827. “Council Bluff, on the Missouri.” The type was collected by James. Torrey’s description of var. missouriensis, an overlooked, ear- lier epithet, fits the current concept of Scutellaria parvula var. leonardi (Epling) Fernald, Rhodora 47: 172.1945. This latter name and its basionym S. /eonardi Epling, Amer. J. Bot. 26: 20.1939, now become synonyms of var. missouriensls. Engelmannia peristenia (Raf.) Goodm. & Laws., comb. nov. Silphium peristenium Raf., Atl. Jour. 1: 146.1832. “S[ilphium] anon T[orrey].239.” Based on the following: #239 Silphium n. sp. Nutt. mss. in Torr., Ann. Lyc. Nat. Hist. N.Y. 2: 215.1827. The type was collected by James but cannot be located at ny. Torrey de- scribed James’s plant as “A singular species, with pinnatifid leaves, and the scales of the involucrum very narrow. Mr. Nuttall found the same on the 1” Red River. Engelmannia pinnatifida Nutt., Trans. Am. Phil. Soc. ser. 2, 7: 343.1840. The plains of Red River.” This collection was made in Oklahoma by Nuttall in 1819. Engelmannia pinnatifida T. & G. in Nutt., Lc. fide Torrey & Gray, Fl. No. Am. 2: 283.1842. “Silphium, n.sp. (Nutt.) Torr. in ann. lyc. New York, 2. p.215. On the Canadian, Dr. James! Red River, Arkansas, Nuttall!.... 381 382 Rhodora [Vol. 94 Although the authorship of Enge/mannia pinnatifida has noth- ing to do with our new combination, it is interesting to note that in some twenty floras and manuals of this century that include this genus, about half use Torrey and Gray, several use Nuttall, some use Gray, and one uses Gray ex Nuttall. Following Article 57.3 of the 1988 International Code of Bo- tanical Nomenclature (often referred to as the Demoulin Rule), the plant Mimulus glabratus HBK. var. fremontii (Benth.) Grant should again be called Mimulus glabratus HBK. var. jamesii (T. & G.) Gray The autonym var. jamesii was established by Mimulus jamesii var. fremontii Benth. This autonym has priority over the epithet var. fremontii. The pertinent synonymy is as follows: Mimulus glabratus HBK. var. jamesii (T. & G.) Gray, Syn. FI. No. Am. ed. 2, 2(1): Suppl. 447.1886. M. jamesii T. & G. ex Benth. in DC., Prodr. 10: 371.1846. Type: James (NY). Along the Missouri River in Pottawatamie Co., Iowa, May 27, 1820. M. jamesii T. & G. ex Benth. var. fremontii Benth. in DC., per 10: 371.1846. ype: Fremont (Ny). Laramie Co., Wyoming, July 14, 42. M. glabratus HBK. var. fremontii (Benth.) Grant, Ann. Mo. Bot. ent 11: 190.1924. DEPARTMENT OF BOTANY AND MICROBIOLOGY UNIVERSITY OF OKLAHOMA NORMAN, OK 73019 ' RHODORA, Vol. 94, No. 880, pp. 383-386, 1992 ADDITIONS TO THE FLORA OF NEWFOUNDLAND. II STUART G. HAy, ANDRE BOUCHARD, LuC BROUILLET AND MARTIN JEAN ABSTRACT Two significant additions have been made to the native vascular flora of New- foundland as a result of explorations on the Great Northern Peninsula: Cardamine bellidifolia L. and Salix cordata Michx. Key Words: Vascular plant flora, additions, Newfoundland Our understanding of the vascular flora of Newfoundland rests largely on the work of M. L. Fernald (1911, 1926, 1933), and of E. Rouleau (1978, unpublished distribution maps and compiled herbarium records). As a result of their considerable contribu- tions, and a long history of botanical exploration by other bota- nists that began with the voyage of Sir Joseph Banks in 1766, today, the flora of the island is relatively well known. Even so, parts of the island remain poorly explored, particularly the more inaccessible interior plateau regions, and exciting new discoveries will continue to be made as botanists further explore this vast territory. Our research on the rare vascular plants of the island (Bouchard et al., 1991), and our earlier work on the flora of the west coast and Gros Morne National Park, have resulted in the addition of many new records to the native flora of the province (Hay et al., 1990). In 1991, we pursued our investigations on the rare plants and general flora in Port au Choix National Historic Park and the region of Canada Bay on the Great Northern Peninsula. In- cluded in this report are notes on the habitat and distribution of two remarkable species that were collected for the first time on the island during these field explorations. Both species must be added to the list of rare plants of Newfoundland (Bouchard et al., 1}. Cardamine bellidifolia L. SPECIMENS. White Bay North Distr.: Chimney Bay, Cloud Hill, 1991/08/08, Bouchard, Hay, Brouillet & Jean 91417 (CAN, MT). 383 384 Rhodora [Vol. 94 Cardamine bellidifolia is a circumpolar, arctic-alpine plant ranging as far south in the Cordillera of western North America as northern California, whereas in the east, it lies mostly to the north of 55 latitude N (northern Québec/Labrador Peninsula, Ellesmere Is., Greenland). Rare disjunct occurrences are found southward in Maine (Mt. Katahdin, West Baldwin), New Hamp- shire (Mt. Washington), and in the Shickshock Mountains of Qué- bec’s Gaspé Peninsula (Crow, 1982; Scoggan, 1978-79). At Chimney Bay, Cloud Hill is a high, barren, wind-swept mountain (alt. 1147 ft.) lying along the northern reaches of the Long Range Mountains. At several sites near the summit, Car- damine bellidifolia was discovered growing in moist cracks of the barren siliceous bedrock. The tiny plants were most inconspicuous and fewer than fifty individuals were observed. The discovery of this species on the Great Northern Peninsula of Newfoundland adds yet another species to an already impres- sive list of wide-ranging, arctic-alpine plants that are restricted to the Long Range Mountains or to the northern tip of the Pen- insula; some of these include Lycopodium alpinum L., Oxyria digyna (L.) Hill, Ranunculus pedatifidus J.E. Sm. ex Rees, Salix herbacea L., Saxifraga rivularis L., and Sibbaldia procumbens L. These arctic-alpine species have ranges extending southward in the eastern part of the continent to isolated or disjunct alpine habitats such as occur in the mountains of the Long Range of western Newfoundland, the Shickshocks of Québec, and outposts in northern New England. Salix cordata Michx. [syn.: S. adenophylla Hook., S. syrticola Fern.] SPECIMENS. St. Barbe South Distr.: Pointe Riche Peninsula, pond south of Port au Choix Cove, 1991/08/01, Bouchard, Hay, Brouillet & Jean 91106 (CAN, GH, MT). The range of Salix cordata extends from southern Hudson Bay and James Bay (Ontario, Québec), southward to the lower Great Lakes (Ontario, Illinois, Michigan), with isolated occurrences eastwards in Québec, southern Labrador and northern Maine (Soper and Heimburger, 1982). Although S. cordata has been attributed to the flora of Newfoundland, its presence there has never been substantiated (G. Argus, pers. comm.). All previous reports, including those of Fernald, are based on misidentifica- i ee ee a ———— : ee 2 eck 1 eee eee Pe 1992] Hay et al.—Newfoundland Flora 385 tions of specimens presently referred to S. eriocephala Michx. (S. rigida Muhl.; S. cordata Muhl., not Michx.) or S. glaucophylloides Fern. (S. myricoides Muhl.), and on misapplication of the name S. cordata Michx. to specimens identified as S. cordata Muhl., a synonym of S. eriocephala. The discovery of Heart-leaved willow in Port au Choix National Historic Park is the first authentic report for the island of New- foundland. Despite the confusion which appears to reign among herbarium specimens, Salix cordata is an unmistakably distinc- tive willow, particularly in the field, where its form and foliage are quite striking. The population, which appears to consist of a single clone of about fifty individuals, was found in an open, minerotrophic larch fen, bordering a large pond. e Port au Choix Peninsula harbors several other rare or noteworthy species of willows that occur in different habitats on the limestone barrens, including Salix arctophila Cockerell, S. ballii Dorn, S. lanata L. ssp. calcicola (Fern. & Wieg.) Hult., S. myrtillifolia Anderss., S. pedunculata Fern., S. reticulata L., and S. wiegandii Fern. ACKNOWLEDGMENTS Funding for the study of the rare plants of Port au Choix Na- tional Historic Park came from Parks Canada, Ottawa. G. Argus (Canadian Museum of Nature, Ottawa) kindly confirmed our identification of Salix cordata. LITERATURE CITED Boucuarp, A., S. Hay, L. BRoumiet, M. JEAN AND I. SAUCIER. 1991. The rare vascular plants of the island of Newfoundland/Les plantes vasculaires rares de l’ile de Terre-Neuve. Syllogeus 65. Crow, G. E. 1982. New England’s Rare, Threatened, and Endangered Plants. United States Dept. of the Interior, Fish and Wildlife Service, Northeast Region, [Newton Corner, MA]. FERNALD, M. L. 1911. A botanical expedition to Newfoundland and southern Labrador. Rhodora 13: 109-162. . 1926. Two summers of botanizing in Newfoundland. Rhodora 28: 49- 63, 74-87, 89-111, 115-129, 145-155, 161-178, 181-204, 210-225, 234- 241 ae 1933. Recent discoveries in the Newfoundland flora. Rhodora 35: |- 16, 47-63, 80-107, 120-140, 161-185, 203-223, 230-247, 265-283, 298- 315, 327-346, 364-386, 395-403. 386 Rhodora [Vol. 94 Hay, S. G., A. BOUCHARD AND L. BRouILLET. 1990. Additions to the flora of the island of Newfoundland. Rhodora 92: 277-293. RouLeAu, E. 1978. List of the vascular plants of the province of Newfoundland (Canada). Oxen Pond Botanic Park, St. John’s, Newfoundlan OGGAN, H. J. 1978-79. The Flora of Canada. 4 parts. Natl. Mus, Canada, Nat. Sci. Publ. Bot. No. 7. Soper, J. H. AND M. L. HEIMBURGER. 1982. Shrubs of Ontario. Royal Ontario Museum, Toronto. INSTITUT DE RECHERCHE EN BIOLOGIE VEGETALE JARDIN BOTANIQUE DE MONTREAL 4101 EST, RUE SHERBROOKE MONTREAL, QUEBEC, CANADA HIX 2B2 RHODORA, Vol. 94, No. 880, pp. 387-390, 1992 ARMORACIA LACUSTRIS (BRASSICACEAE) REDISCOVERED IN OHIO James S. McCoRMAC Key Words: Armoracia lacustris, rediscovery, ecology, Ohio Armoracia lacustris (A. Gray) Al-Shehbaz & V. Bates (Lake Cress) is a heterophyllous mustard of aquatic habitats which oc- curs over an extensive area in eastern North America, ranging from Pennsylvania and Virginia west to the 95th meridian, south to Florida and Texas, and north to southern Canada. Within this broad distribution, 4. /acustris is irregularly dispersed, and has been reported from widely scattered localities in nineteen states and two provinces. It has been proposed as a candidate for listing as federally endangered or threatened (Federal Register, 1990) due to its apparent rarity. This taxon has a confusing nomencla- tural history, having been placed in six different genera. In recent years the name Armoracia aquatica (Eaton) Wieg. had been widely used for Lake Cress, but Al-Shehbaz and Bates (1987) demon- strated that this name is invalid, and proposed the new combi- nation Armoracia lacustris (A. Gray) Al-Shehbaz & V. Bates. Lake Cress was last collected in Ohio on 10 June 1936 by Floyd Bartley near Circleville, Pickaway County (Cooperrider, 1982). There have been no reports of this plant in Ohio since that time; as of 1990 it was listed as presumed extirpated by the Ohio De- partment of Natural Resources (Div. of Natural Areas and Pre- serves, 1990). Historical locations, almost all of which date from the late nineteenth century, are known from eight Ohio counties (Coshocton, Erie, Licking, Lorain, Lucas, Madison, Perry, Pick- away) based on surveys of CLM, GH, KE, MICH, MU, NY, OS, and PH. Published county dot maps of Armoracia lacustris in Ohio by Al-Shehbaz and Bates (1987) and Easterly (1964) included Clark County and omitted Perry County. I could not locate a specimen to substantiate the Clark County report, and os contains an 1895 collection from Perry County. On 19 June 1991, M. A. Moser, Jr., S. J. Stine and I were ducting field investigations of riparian habitats along St. Marys River in west-central Ohio. In an Ash-Maple floodplain woods in Mercer County, we located a population of Armoracia lacustris which contained ca. 180 fertile plants and numerous sterile ro- settes. The plants were growing in wet soil of a seasonally inun- 387 388 Rhodora [Vol. 94 dated, diassociated channel of the river, and were flowering and fruiting vigorously. St. Marys River is a moderately large stream of low gradient (2.9’ per mile) (Div. of Water, 1960) which originates in western Auglaize County, Ohio and flows west and north to its confluence with the Maumee River, a major drainage of the Lake Erie basin. In Ohio, the St. Marys is characterized by a broad, well-forested floodplain with numerous oxbows and buttonbush swamps; it is one of the few remaining unchannelized rivers in western Ohio. Near Fort Wayne, Indiana, a low-lying divide known as the Mau- mee Terrace separates the St. Marys and Wabash drainages (Thornbury, 1958). During periods of extreme flooding, waters from the St. Marys drainage unite with the Mississippian drainage of the Wabash River through this divide (Greene, 1935), thus representing a potential migratory corridor for plants between the Mississippi River and Lake Erie drainages. Armoracia lacustris has been collected at several locations in the Wabash system in central and eastern Indiana (Deam, 1940). Additional field surveys of the St. Marys River area during June and July by Moser and myself located a total of six separate populations of Armoracia lacustris (Figure 1), distributed in all three Ohio counties through which the St. Marys flows (Auglaize, Mercer, Van Wert). These populations range in size from ca. two dozen plants to 1000+ plants scattered over two to three acres at a site in Mercer County. All populations occur in old river channels now isolated from the St. Marys except during periods of flooding, when they are temporarily inundated. The oxbows in which A. lacustris grows are in semi-open riparian woodlands dominated by Acer saccharinum L. and Fraxinus pennsylvanica Marsh.; Cephalanthus occidentalis L. is always present in the wettest sections of these oxbows. Typical herbaceous associates include Carex crus-corvi Shuttlew. , C. lupulina Muhl., C. muskin- gumensis Schwein., C. tribuloides Wahlenb., Leersia lenticularis Michx., L. oryzoides (L.) Swartz, Ludwigia palustris (L.) Ell., Poly- gonum hydropiperoides Michx., Proserpinaca palustris L., Ro- rippa sessiliflora (Nutt.) Hitche., Samolus floribundus HBK., and Saururus cernuus By mid-June, surface water was no longer present in the oxbow habitat, and Armoracia lacustris was growing in a soft, thick, muddy substrate in a loosely defined zone along the periphery of 1992] McCormac—Armoracia 389 a 7 | Figure 1. Distribution of Armoracia lacustris in Ohio showing pre-1937 rec- ords (@), and 1991 collection sites (@). each oxbow. Sterile rosettes greatly outnumbered flowering and fruiting plants in all populations. Few mature seeds could be found by random examination of apparently well-developed fruits; re- production in these populations was mostly, if not entirely, by vegetative means. Cauline leaves easily detach from the plants; soon after falling onto the soft muck, they send out roots and eventually form rosettes. LaRue (1942) described this method of regeneration in detail; leaves I collected and placed on moist potting soil also rooted and formed rosettes. The percentage of these rosettes which survive and develop into mature plants in the wild is not known. Voucher specimens of Armoracia lacustris collected in this study are deposited in CLM, GH, KE, MICH, MU, NY, OS and PH. 390 Rhodora [Vol. 94 ACKNOWLEDGMENTS I thank A. W. Cusick and W. P. Stoutamire for their helpful suggestions, and gratefully acknowledge the comments of two anonymous reviewers. LITERATURE CITED AL-SHEHBAZ, I. A. AND V. BATEs. 1987. Armoracia lacustris (Brassicaceae), the correct name for the North American Lake Cress. J. Arnold Arbor. 68: 357- 359. COOPERRIDER, T. A., ED. 1982. ye aoe and threatened plants of Ohio. Ohio Biol Surv. Bull. Notes No. 16, Columbus, OH. Deam, C. C. 1940. Flora of Indiana. Dept. a Conserv., Div. of Forestry, Indi- anapolis, DIVISION OF Wacuss AREAS AND PRESERVES. 1990. Rare native Ohio plants. 1990-91 status list. Ohio Dept. of Nat. Res., Columbus Division OF WaTER. 1960. Gazeteer of Ohio streams. Ohio ei, of Nat. Res., Columbus, OH. EasTERLy, N. W. 1964. Distribution patterns of Ohio Cruciferae. Castanea 29: 164-173. FEDERAL REGISTER. 1990. Dept. of the Interior, U.S. Fish and Wildlife Serv. Vol. 50, No. 35, Washington, DC. GREENE, C. W. 1935. The distethaition of Wisconsin fishes. Wisconsin Conser- vation Commission, Madison, WI. LaRue, C. D. 1942. Gea ks in Radicula aquatica. Pap. Michigan Acad. Sci. 28: 51-60. THORNBURY, W. D. 1958. The geomorphic history of the Upper Wabash valley. Amer. J. Sci. 256: 449-469. OHIO DEPARTMENT OF NATURAL RESOURCES DIVISION OF NATURAL AREAS AND PRESERVES 1889 FOUNTAIN SQUARE, BLDG. F COLUMBUS, OH 43224 RHODORA, Vol. 94, No. 880, pp. 391-392, 1992 NEW ENGLAND NOTE UTRICULARIA INFLATA WALTER (LENTIBULARIACEAE) IN MASSACHUSETTS Bruce A. SORRIE On 11 October 1990, I traveled to Federal Pond, on the border of Carver and Plymouth in Plymouth County, Massachusetts to verify a report of black spruce, Picea mariana (Miller) BSP. Ac- cording to Austin Mason, Regional Forester with the state Di- vision of Forests and Parks, a small stand of this spruce occurred on the east shore of the pond. If verified, it would represent only the second extant population of P. mariana in southeastern Mas- sachusetts. After a short search, four small trees were found in deep peat at the edge of a cove just within the boundary of Plym- outh, but none elsewhere, including several peat islands offshore. Of even more interest, however, were a few flowering bladder- worts in another cove that obviously were related to Utricularia radiata Small, a common species in acid lakes and ponds in Massachusetts, but which were much larger. The scape of one measured 23 cm and the floats spanned 18 cm! Particularly dis- tinct, in comparison to specimens of U. radiata, were the very large flowers, long floats that tapered to the axis, thick peduncle and stem, and massive system of bladder-bearing branches. These plants keyed easily in Godfrey and Wooten (1981) and Schnell (1976) to U. inflata Walter, a species heretofore unknown north of southern New Jersey. A return visit on 20 October found U. inflata abundant around much of Federal Pond, especially in the lee of peat islands. Associates included Utricularia purpurea Wal- ter, Nymphaea odorata Aiton, Myriophyllum heterophyllum Mi- chaux, M. humile (Raf.) Morong, Lobelia dortmanna L., Juncus militaris Bigelow, and Scirpus subterminalis Torrey. Specimens (BAS 5317 and 5332) have been deposited at GH, MASS and in my personal herbarium. This discovery marks the 13th Utricularia for Massachusetts and New England. As with other disjunctions, one must consider the possibility of human-aided introduction, particularly in this case since Federal Pond is used as a reservoir for a cranberry growing firm. Mr. David Parks of the Federal Furnace Cranberry Company assured me that although New Jersey and Massachu- 391 392 Rhodora [Vol. 94 setts have extensive cranberry cultivation, there is no movement of plant material or soil between the states; rarely a piece of machinery may be transported. Griffith (1913) stated that Federal Pond was created in 1793 by damming Crane Brook, in order to mine bog iron. Certainly there has been ample time since then for propagules to be transported, most likely via waterfowl or herons. One might argue that since there are hundreds of ponds in southeastern Massachusetts, U. inflata should be widespread. However, there are dozens of rare species that defy such logic, occurring in but a tiny fraction of the apparently suitable habitat in the Plymouth/Cape Cod area. Such is true of some of the other inhabitants of Federal Pond: Carex lanuginosa Michaux (here at its sole southeastern Massachusetts site), Carex striata Michaux (formerly C. walteriana Bailey), red-bellied turtle (Chrysemys rub- riventris), and the black spruce. ACKNOWLEDGMENTS Access to Federal Pond, and a pair of oars, was generously provided by the Federal Furnace Cranberry Company of Carver, MA. A boat was provided by the MA Division of Fisheries and Wildlife. Field assistance was ably provided by Ann Buckley, Ted Hendrickson, and Chuck Katuska. The staff of Gu is thanked for access to their holdings and library resources. Inventory was per- formed when I was Botanist for the Massachusetts Natural Her- itage and Endangered Species Program. LITERATURE CITED Goprrey, R. K. anp J. W. WooTen. 1981. Aquatic and Wetland Plants of Southeastern United States. Dicotyledons. University of Georgia Press, Ath- ens, GA. GriFFiTH, H.S. 1913. History of the Town of Carver, Massachusetts. Historical Review 1637-1910. E. Anthony & Sons, New Bedford, MA. SCHNELL, D.E. 1976. Carnivorous Plants of the United States and Canada. John F. Blair, Winston-Salem, NC 202 HOLLY PINES DRIVE PINEHURST, NC 28374 RHODORA, Vol. 94, No. 880, pp. 393-394, 1992 BOOK REVIEW MacKinnon, A., J. Pojar and R. Coupe, Eds. 1992. Plants of Northern British Columbia. 344 pp. British Columbia Min- istry of Forests and Lone Pine Publishing, 206, 10426-81 Ave., Edmonton, Alberta, Canada T6E 1X5. ISBN 1-55105- 015-3. (Price: $19.95 Canadian, paper). This outstanding book is the only field guide to the plants of northern interior British Columbia and, to my mind, the best available field guide for the interior of Alaska. As a field guide, itis limited to the more common plants, but includes trees, shrubs, graminoids, ferns and bryophytes in addition to the showy her- baceous angiosperms generally covered in such guides. Approx- imately 550 species are illustrated, but the book enables the user to identify at least fifty additional species. Forty species and sub- species of willow are described! The book is organized by family, and emphasizes the Ericaceae, Salicaceae, Rosaceae, Liliaceae, Asteraceae, Scrophulariaceae, Fa- baceae, Ranunculaceae, Orchidaceae and Saxifragaceae, although common members of other families are included. A table of con- tents directs the user to the family, to keys, and to conspectuses. For each species, the authors provide clear descriptions of habit, leaves, flowers, fruits and ecology. Notes discuss characters that distinguish similar species, and provide ethnobotanical infor- mation. Each species is illustrated with color photographs of flow- ers and/or fruits, and with line drawings showing habit, leaves or fruits. The line drawings of trees and shrubs are charmingly il- lustrated with an appropriate mammal for scale. There are some errors, notably the transposition of photos for several species of graminoids and incorrect illustrations for three lichens, for which the publisher has provided a page of errata. Several unique features make this an exceptionally user-friendly field guide. Difficult groups of plants are provided with easy-to- use, clear dichotomous keys (trees, Ribes, genera of the Aster- aceae, Orchidaceae, Arnica, Platanthera, Pedicularis, Salix). An outstanding picture-key to the grass genera makes identification relatively simple. In addition, species in large and difficult groups (Ribes, Vaccinium, Potentilla, Viola, Pedicularis, Liliaceae) are distinguished in a “conspectus” —a two-page table allowing the user to directly compare the stems, leaves, flowers, or habitat of 393 394 Rhodora [Vol. 94 up to fifteen species. This table is especially useful in large groups where space has not permitted the authors to provide illustrations or complete descriptions for all species. The authors frequently use leaf silhouettes in keys and comparative tables where these aid identification (Fabaceae, Petasites, Potentilla, Ranuncula- ceae), and have augmented the key to Salix with leaf photocopies showing venation as well as outline. This little book is, without a doubt, the most usable and most complete field guide I have encountered. It enables the amateur or professional to easily identify almost any vascular or non- vascular plant encountered—even sedges. The excellent illustra- tions, use of leaf silhouettes, and the innovative “‘conspectus”’ for large genera simplifies identification dramatically. I highly rec- ommend this book for anyone planning travel to British Colum- bia, the Yukon, Mackenzie District, or to Alaska, as well as to anyone planning to write a field guide. LISA A. STANDLEY VANASSE HANGEN BRUSTIN, INC. 101 WALNUT STREET, P.O. BOX 9151 WATERTOWN, MA 02272-9151 RHODORA, Vol. 94, No. 880, p. 395, 1992 IN MEMORIAM WILLIAM HOLLAND DRURY 1921-1992 William Holland Drury, Professor at The College of the Atlantic in Bar Harbor, Maine, died on March 26, 1992. Born in Newport, RI, he was a 1942 magna cum laude graduate of Harvard College. He served with the United States Navy from 1942 until 1945, serving tours of duty both in the North Atlantic and the South Pacific. He returned to graduate study at Harvard after the war and was awarded his doctorate in biology and geology in 1952. Dr. Drury was a member of the New England Botanical Club for forty-five years, first joining in 1947. He was elected to serve as Vice-president from 1962-65, and as President from 1965-68. One of the country’s best ecologists, Dr. Drury, among other accomplishments, showed, along with his co-worker Dr. Ian Nis- bet, that the ideas of ecological succession found in most ecology textbooks are based upon wishful thinking, rather than on ob- served facts. He taught ecology and evolutionary biology at Har- vard for many years. He was an active member of the Massa- chusetts Audubon Society, and served as its Director of Education and Research from 1972 to 1976. The esteem in which Dr. Drury was held by some individuals who knew him is largely the reason that Massachusetts Audubon today occupies Drumlin Farm, which was donated to the Society to help further his research on birds and ecology. He traveled widely, and was learned about many subjects, one of which was the flora of Alaska. An active conservationist, Dr. Drury helped sound the early alarms about pesticides and other environmental insults. He served as a member of the President’s Science Advisory Council during both the Kennedy and Nixon administrations. He was instru- mental in reintroducing the peregrine falcon to the Northeast, and had recently spent much time on studying and enhancing diversity among shorebird populations around the Gulf of Maine. In 1976, Dr. Drury joined The College of the Atlantic, where he was a teacher, researcher, and director of advanced studies until shortly before his death. He was proud of his role there, which he summarized to me as, “Whatever else we do for stu- dents, we certainly teach them how to think.” —[Norton H. Nick- erson] Be RHODORA, Vol. 94, No. 880, p. 396, 1992 IN MEMORIAM HERMAN ROYDEN SWEET 1911-1991 Dr. Herman R. Sweet died on October 2, 1991, at his home in Florida, shortly after completing a journey in which he visited the Boston area. A memorial service in his honor was held at Goddard Chapel, Tufts University, Medford in early November. Interment was in Mount Auburn Cemetery, Cambridge. Volume 92 of RHODORA was dedicated to Dr. Sweet on the occasion of his retirement from service to the New England Bo- tanical Club; he served as its Treasurer for twenty-three years, and had been a member since 1959. An account of his many activities and duties for the Club, for Tufts University where he was a long-time faculty member, and for the American Orchid Society of which he was an active member, may be found in the frontispiece of RHODORA, Vol. 92, (No. 869, January, 1990). —[Norton H. Nickerson] 396 a NEBC AWARD FOR THE SUPPORT OF BOTANICAL RESEARCH The New England Botanical Club will offer an award of $1000 in support of botanical research to be conducted in relation to the New England flora during 1993. This award is made to stim- ulate and encourage botanical research on the New England flora, and to make possible visits to the New England region by those who would not otherwise be able to do so. The award will be given to the graduate student submitting the best research pro- posal dealing with field studies in systematic botany, biosyste- matics, plant ecology, or plant conservation biology, although proposals for research in other areas of botany will also be con- sidered. This award is not limited to graduate students at New England institutions, nor to members of the New England Bo- tanical Club. Papers based on this research must acknowledge the NEBC’s support, and it is encouraged that they be submitted to RHODORA for possible publication subject to standard review processes. The 1992 New England Botanical Club Award for the Support of Botanical Research was awarded to Mr. Francois Lutzoni, a graduate student at Duke University, in support of his Ph.D. research, entitled “Phylogenetics of Omphalina (Basidiomyco- tina, Agaricales) and the Evolution of Lichenization.” Applicants should submit a proposal of no more than three double-spaced pages, a budget (the budget will not affect the amount of the award), and a curriculum vitae. Three copies of the application must be submitted. Two letters in support of the proposed research, one from the student’s thesis advisor, are also required. Proposals and supporting letters must be received no later than March 1, 1993. The recipient of the award will be notified by April 30, 1993. Proposals should be sent to: Awards Committee The New England Botanical Club 22 Divinity Ave. Cambridge, MA 02138 397 U.S. NATIONAL ARBORETUM HERBARIUM The Herbarium of the U.S. National Arboretum (NA), Wash- ington, DC announces the installation of a new Spacesaver com- pactor filing system. The project, to be installed in three modules, was Started in January, 1992 and should be finished in July, 1992. The transfer of the herbarium specimens will take place at the completion of each of the three modules, with the transfer of specimens into the final module to be completed in late 1992. When the shift of the specimens is completed, scientists and stu- dents will be welcome to visit and use the new system. The National Arboretum herbarium collection of over 600,000 specimens concentrates on extant and potential economic plants from indigenous and cultivated sources on a world-wide basis. The collection is especially rich in woody landscape plants—spe- cies, cultivars and clones—cultivated in the United States. Major genera include: Amaranthus, Salix, Ilex, Viburnum, Pyr- acantha, Coffea, Trifolium, Medicago, Magnolia, among many other agricultural and nursery crop plants. Major geographical areas represented by collections of wild-occurring and cultivated plants include: North America, Japan, the Peruvian Andes, Tierra del Fuego (Argentina), Juan Fernandez Islands (Chile), The Peo- ple’s Republic of China, and The Republic of Korea. Large por- tions of the Isaac C. Martindale Herbarium (puiL) a 19th century world-wide collection, were acquired in 1964. Sizable sections of the Catholic University Herbarium (Lcu) and the Morris Arbo- retum Herbarium (Moar) were incorporated into NA in 1984. The National Arboretum Herbarium welcomes exchange of herbarium specimens of economic plants and their wild-occurring relatives—landscape, food, forage, commercial, industrial, weed, etc.—both indigenous and cultivated. The herbarium is also pleased to send material world-wide on loan for study by spe- cialists in various groups. Dr. T. R. Dudley became Lead Scientist in charge of the Her- barium and the taxonomic and nomenclatural research unit at the National Arboretum, effective 1 October 1991. Mr. P. M. Mazzeo, Botanist, is responsible for herbarium curatorial activ- ities. Dr. Frederick G. Meyer, formerly Botanist in charge of the Herbarium, retired effective 30 September 1991. 398 1992] 399 Contact either Dr. T. R. Dudley or Mr. P. M. Mazzeo for exchanges and loans: U.S. National Arboretum Herbarium 3501 New York Avenue, NE Washington, DC 20002-1958, U.S.A. Dr. Dudley’s phone: 202-475-4842 Mr. Mazzeo’s phone: 202-475-4841 FAX: 202-475-5694 THE 1992 JESSE M. GREENMAN AWARD The 1992 Jesse M. Greenman Award has been won by Sharon Elaine Bartholomew-Began for her publication “A morphogenetic re-evaluation of Haplomitrium Nees (Hepatophyta),” published as Volume 41 of Bryophytarum Bibliotheca. This study is based on a Ph.D. dissertation from Southern Illinois University at Car- bondale, under the direction of Dr. Barbara Crandall-Stotler. The Greenman Award, a certificate and a cash prize of $500, is presented each year by the Missouri Botanical Garden. It rec- ognizes the paper judged best in vascular plant or bryophyte sys- tematics based on a doctoral dissertation published during the previous year. Papers published during 1992 are now being ac- cepted for the 25th annual award, which will be presented in the summer of 1993. Reprints of such papers should be sent to: Dr. P. Mick Richardson Greenman Award Committee Missouri Botanical Garden P.O. Box 299 St. Louis, MO 63166-0299, U.S.A. In order to be considered for the 1993 award, reprints must be received by 1 June 1993. RHODORA, Vol. 94, No. 880, p. 400, 1992 REVIEWERS OF MANUSCRIPTS JULY 31, 1991-JULY 31, 1992 The editors of RHODORA are grateful to each of the following specialists for their participation in the review process. Kelly W. Allred Loran C. Anderson Stephen C. Buttrick David S. Conant Michael O. Dillon Peter W. Dunwiddie George S. Ellmore Ronald L. Hartman Walter M. Hewitson Richard A. Howard David B. Lellinger J. K. Massey 400 Richard S. Mitchell Guy L. Nesom Wayne B. Powell Anton A. Reznicek Sally C. Rooney Charles J. Sheviak Scott W. Shumway Tod F. Stuessy Francis R. Trainor S. P. Vander Kloet Peter F. Zika ‘Hodova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB NORTON H. NICKERSON, Editor-in-Chief JOAN Y. NICKERSON, Managing Editor Associate Editors DAVID S. BARRINGTON RICHARD A. FRALICK A. LINN BOGLE GERALD J. GASTONY DAVID E. BOUFFORD C. BARRE HELLQUIST CHRISTOPHER S. CAMPBELL MICHAEL W. LEFOR WILLIAM D. COUNTRYMAN ROBERT T. WILCE GARRETT E. CROW VOLUME 94 1992 The New England Botanical Club, Inc. Harvard University Herbaria, 22 Divinity Ave., Cambridge, Mass. 02138 , ” 7 , =. = |; i 7 . Sk Aiea - SS ha RHODORA, Vol. 94, No. 880, pp. 403-408, 1992 INDEX TO VOLUME 94 New scientific names are in bold face A(risaema) dracontium, Arisaem . phyllum subsp. stewardsonii x, brids in Massachusetts. A See occurring population of putative, 340-347 Aconogonon (Polygonum sect. Acono- gonon) in Alaska. Choosing the cor- rect name for, 319-322 Additions to the flora of Newfound- land. II. 383-386 Alabama. A new species of etre (Lamiaceae) from northern, I- Alaska. Choosing the correct name is Aconogonon (Polygonum sect. Aco- nogonon) in, 319-322 Alpine flora of the northeastern United States. Contributions to the, 15-37 ANNOUNCEMENTS Jesse M. Greenman Award 108, 399 NEBC Award for Support of Botan- ical Research 106, 397 NEBC and New England Wildflower Society Symposium 99 New England Botany Graduate Stu- dents Meeting 107 Ronald L. Stuckey Herbarium En- dowment Fund 107 U.S. National Arboretum Herbari- um Arethusa bulbosa, _Calopogon tubero- titioning the microhabitat. 374-380 Arisaema ag oe subsp. — sonii x A. dracontium hybrids poet A naturally-occur- ring population of putative, 340-347 Armoracia lacustris (Brassicaceae) re- 0 ditional reports and comments on the cytogeography and status of some species of, 48-62 Asteraceae of the Guianas, III: Verno- nieae and the restoration of the genus Xiphochaeta. The, 348-361 Barrington, David S. Climate and the disjunct distribution of Polystichum alfarii (Christ) Barr., comb. nov. in Mesoamerica. 327-339 Berkshire County, Massachusetts. Nat- ural plant communities of, 171-209 Blephilia (Lamiaceae) from northern Alabama. A new species of, 1-14 ape ig subnuda Simmers & Kral, sp. 2 bietendl, J. Todd and Peter J. Scott. Eco- ing the microhabitat. 374-380 K REVIEWS Phytogeography and Vegetation Ecology of Cuba. 103-105 Plants of Northern British Columbia. —394 Rare Plants of Prince Edward Island. e, 102 ere ne ae pony Soca Surface, rsity Based on . Electron ta Studies. Bouchard, Andre see Hay, Stuart G: Brouillet, Luc see Hay, Stuart G. Burk, C. John see Sanders, Laurie L. Buttrick, Steven C. Habitat manage- ment: a decision making process. 258- 2s C3 te dandeyana Adams. C. inflata and, Chloris barbata Sw. a ndC. ee eee (Poaceae), the earlier names for, 135-14 C.(hloris) elata Desvaux, Chloris bar- 403 404 bata Sw. and, (Poaceae), the earlier names for C. inflata Link and C. dan- deyana Adams. 135-140 C.(hloris) inflata Link and C. dandey- ana Adams. Chloris barbata Sw. and C. elata Desvaux (Poaceae), the ear- lier names for, 135-140 Caldwell, Fredricka A. and Garrett E. Crow. A floristic and vegetation anal- ysis of a freshwater tidal marsh on the Merrimack River, West New- ry, Massachusetts. 63-97 Calopogon tuberosus and Pogonia ophioglossoides (Orchidaceae) in eastern Newfoundland. Ecological aspects of Arethusa bulbosa, II. Par- titioning the microhabitat. 374-380 Canada. Geographical distribution and ecology of Long’s Bulrush, longil i. in, Cape Cod. An enous populatio of Clintonia ae (Liliaceae) on, Carex caryophyllea (Cyperaceae) in Massachusetts. Persistence of, 210- l Carolina Beach State Park (North Car- olina). The flora of limesink depres- sions in, 156-166 Carpheporus corymbosus (Composi- tae). Comments on the phenology of, 323-325 Chambers, Kenton L. Choosing the correct name for Aconogonon (Po- lygonum sect. Aconogonon) in Alas- ka. 319-322 Chloris barbata Sw. and C. elata Des- vaux (Poaceae), the earlier names for C. inflata Link and C. dandeyana Ad- ams. 140 Chmielewski, Jerry G. see Semple, John C. Choosing the correct name for Acono- gonon (Polygonum sect. Aconogo- non) in Alaska. 319-322 Chromosome number determinations e ments on the cytogeography and sta- Rhodora [Vol. 94 tus of some species of Aster and Sol- idago. 48-62 Chromosomes of Mexican Sedum VI. Section Sedastrum. 362-370 Climate and the disjunct distribution of som ipo alfarii (Christ) Barr., n Mesoamerica. 327-339 Clintonia baat (Liliaceae) on Cape Cod. An indigenous population of, 98-99 Comments on the phenology of Car- pheporus corymbosus (Compositae). 323-325 Contributions to the alpine flora of the northeastern United States. 15-37 Corey, David T. Comments on the phe- nology of Carpheporus corymbosus (Compositae). 323-325 Cross, Shirley G. An indigenous po ulation of Clintonia borealis (Lili- aceae) on Cape Cod. 98-99 Crow, Garrett E. see Caldwell, Fred- ricka A. ric Crow, Garrett E. see Weatherbee, Pam- la B ela B. Cytogeography and status of some spe- cies of Aster and Solidago. Addition- al reports and comments on, Chro- mosome number determinations in Fam. Compositae Tribe Astereae. IV. 48-62 Distribution, disjunct, of Polystichum alfarii (Christ) Barr., comb. nov. 1n Mesoamerica. Climate and the, 327- 339 Ecological aspects of Arethusa bulbosa, Calopogon tuberosus and Pogonia ophioglossoides (Orchidaceae) in eastern Newfoundland. II. Partition- ing the microhabitat. 374-380 Endangered species. Scientific and pol- icy considerations in restoration and reintroduction of, 287-31 Engelmannia peristenia (Raf.) Goodm. & Laws., comb. nov. 381 Etymology of Vaccinium L. On the, 371-373 1992] Index to Volume 94 405 Falk, Donald A. and Peggy Olwell. Sci- entific and policy considerations in restoration and reintroduction of en- dangered species. 287-315 Flora of limesink depressions in Car- olina Beach State Park (North Car- olina). The, 156-166 Floral variation and taxonomy of Lim- nobium L. C. Richard (Hydrochan- taceae). 111-134 Floristic and vegetation ee of a ca are tidal marsh on the Mer- ack River, West earners Mas- ian setts. A, Freshwater tidal marsh on the Merri- mack River, West Newbury, Mas- sachusetts. A pe and vegetation analysis of a, 63- Gandhi, Kancheepuram N. see Kar- tesz, John T. Geographical egg and ecology of Long’s Bulrush, Scirpus longii Pe aerigecl in Canada. 141-155 odman, George J. and Cheryl A. sn Two new combinations and Guianas, Asteraceae of the, III: Ver- nonieae and restoration of the genus Xiphochaeta. The, 348-361 Habitat management: a decision-mak- ing process. 258-286 Hay, Stuart G., Andre Bouchard, Luc Brouillet and Martin Jean. Additions to the flora of Newfoundland. II. 383-386 Herndon, Alan. Nomenclatural notes on North ag oa an Hypoxis (Hy- Poxidaceae). 7 | Hill, Nicholas and Mat s E. Johans- ecology of Long longii ( )in Canada. 141- 55 Holsinger, Kent E. Setting priorities for regional plant conservation pro- grams. 243-257 Hunt, David M. and Robert E. Zarem- crostegium vimineum New York and adjacent states. 167- 170 Hymenophyllum wilsonii Hooker (Hy- in ing in the Iberian pen- insula. 316-318 Hypoxis (Hypoxidaceae). Nomencla- tural notes on North American, 43447 Iberian peninsula. Hymenophyllum wilsonii Hooker oe Pcs: in the, 316- William H. Drury 395 Herman R. Sweet 396 Indigenous population of Clintonia bo- realis (Liliaceae) on Cape Cod. An, 8-9 Jean, Martin see Hay, Stuart G. Johansson, Mats E. see Hill, Nicho- las M. Kartesz, John T. and Kancheepuram N. Gandhi. Chloris barbata Sw. and Keenan, Philip E. A new form of Triphora trianthophora (Swartz) Rydberg, and part 3 of observations of the ecology of Triphora trian- thophora (Orchidaceae) in New Hampshire. 38-42 Kral, Robert see Simmers, Richard W. Lawson, Cheryl A. see Goodman, George J. Limesink depressions in Carolina Beach State Park (North Carolina). The flo- ra of, 156-1 Limnobium L. C. Richard (Hydro- charitaceae). Floral variation and taxonomy of, 111-134 Limnobium spongia (Bosc.) Steudel, subsp. laevigatum (Humboldt & 406 Bonplant ex Willdenow) Lowden, comb. nov. 129 Long Expedition of 1820. Two new combinations and a name change from the, 381-382 Long’s Bulrush, Scirpus longii (Cyper- of Limnobium L. C. Richard (Hydrocharitaceae). 111- eo freshwater tidal, on the Merri- k River, West Newbury, Mas- ah A floristic — vegetation analysis of a, 97 Massachusetts. A floristic and vege tion analysis of a freshwater tidal marsh on the Merrimack River, West Newbury, 63-97 Massachusetts. A naturally- “occurring Massachusetts. Utricularia inflata Wal- ter (Lentibulariaceae) in, 391-392 M eg oe ames S. Armoracia lacus- ris ee rediscovered in Ohio 387-390 bieeitenenit River, West Newbury, Massachusetts. A floristic and vege- tation analysis of a freshwater tidal marsh on the, 9 Mesoamerica. Climate and the disjunct distribution of Polystichum alfarii (Christ) Barr., comb. noy. in, 327- 339 Mexican Sedum VI, Chromosomes of, Section Sedastrum. 362-370 Microhabitat. Ecological aspects of Ar- Partitioning the, 374-380 Rhodora [Vol. 94 Microstegium vimineum (Poaceae) into New York and adjacent states. The northeastward spread of, 167-170 Natural plant communities of Berk- shire County, Massachusetts. 171- 209 Naturally-occurring population of pu- tative Arisaema triphyllum subsp. nil i brids in Massachusetts. A, 340-347 OTES An indigenous population of Clin- tonia borealis (Liliaceae) on Cape Cod. 98-99 Persistence of Carex caryophyllea (Cyperaceae) in Massachusetts. 10-211 Utricularia inflata Walter (Lentibu- lariaceae) in Massachusetts. 391- 392 New Hampshire. A new form of to trianthophora (Swartz) Newfoundland. Additions to the flora f, II. 383-386 Newfoundland. Ecological aspects of Arethusa bu oe tural notes o can Hypoxis caution 43447 North American Hypoxis (Hypoxida- ceae). Nomenclatural notes on, —47 North Carolina. The flora of limesink depressions in Carolina Beach State Park, 6-166 Northeastward spread of Microstegium 1992] Index to Volume 94 407 vimineum (Poaceae) into New York and adjacent states. The, 167-170 Ohio. Armoracia lacustris (Brassica- ceae) rediscovered in, 387 Olwell, Peggy see Falk, Datla A. Ortiz, S. see Sofiora, F. X. Persistence of Carex caryophyllea (Cyperaceae) in Massachusetts. 210- 211 Phenology of Carpheporus corymbosus (Compositae). Comments on the, 323-325 Plant conservation programs. Setting priorities for regional, 243-257 Pogonia ophioglossoides (Orchidaceae) in eastern Newfoundland. Ecological aspects of Arethusa bulbosa, Calo- pogon tuberosus and, II. Partitioning the microhabitat. 000-000 eae alfarii (Christ) Barr., comb. Poostichu ci (Christ) Barr., comb. M rica. Climate and the ae 5 aia of, 327-339 Rare species protection. Taxonomic is- sues in, 218-242 Restoration and reintroduction of en- Reviewers, List of, jee 1991-1992. Robinson, Harold. Asteraceae of the Guianas, III: Vernonieae and resto- ration of the genus Xiphochaeta. 348-361 Rodriguez-Oubiiia, J. see Sofiora, F. X. Sanders, Laurie L. and C. John Burk. A naturally-occurring population of putative Arisaema triphyllum subsp. x rids in Massachusetts. 340-3 Scientific and policy considerations in restoration and reintroduction of en- dangered species. 287-315 Scirpus longii (Cyperaceae) in Canada. Geographical distribution and ecol- ogy of Long’s Bulrush, 141-155 Scott, Peter J. see Boland, J. Todd Scutellaria parvula Michx. var. mis- souriensis (Torr.) Goodm. & Laws., comb. nov. 381 Sedum, Chromosomes of Mexican, VI. 0 aceae) from northern A Siren, David J. and Karen R. Warr. The flora of limesink depressions in Car- olina Beach State Park (North Car- olina). 156-166 Solidago, Aster and u , Chromosome Fam. cytogeography and status of some species of, 48-62 Sofiora, F. X., S. Ortiz and J. Rodri- guez-Oubifia. Hymenophyllum wil- sonii Hooker (Hymenophyllaceae) in the Iberian peninsula. 316-319 Sorrie, Bruce A. Utricularia inflata Walter (Lentibulariaceae) in Massa- chusetts. 391-392 Standley, Lisa A. Persistence of Carex caryophyllea (Cyperaceae) in Mas- sachusetts. 210-211 Standley, Lisa A. Taxonomic issues in rare species protection. 218-242 SYMPOSIUM— March, 1992. New England Plant Conservation: The Scientific Basis for Effective Action. 213-315 Foreword. Leslie J. Mehrhoff, Pres- ident NEBC and William E. Brum- 408 back, Conservation Director NEWFS. 215-217 Taxonomic issues in rare species pro- tection. 218-242 ‘ciple: trianthophora (Swartz) Ryd- i forma albidoflava Keenan forma no 39 Cries trianthophora (Swartz) Ryd- berg, A new form of, and part 3 of observations on ay ie of me change from the Long Expedition of 1820. 381-382 Uhl, Charles H. Chromosomes of Mex- ican Sedum VI. Section Sedastrum. 362-370 United States, northeastern. Contri- butions to the flora of the, 15-37 Utricularia inflata Walter (Lentibular- iaceae) in Massachusetts. 391-392 Rhodora [Vol. 94 Vaccinium L. On the etymology of, 371-373 Vander Kloet, S. P. On the etymology of Vaccinium L. 371-373 Vernonieae and restoration of the genus Xiphochaeta. The Asteraceae of the Guianas, III:, 348-361 Warr, Karen R. see Siren, David J. Weatherbee, Pamela B. and Garrett E. Crow. Natural plant communities of Berkshire County, Massachusetts. 171-209 Xiang, Chunsheng see Semple, John C. Xiphochaeta. The Asteraceae of the Guianas, III: Vernonieae and resto- ration of the genus, 348-361 Zaremba, Robert E. see Hunt, Da- vid M. Zika, Peter F. Contributions to the flora of the northeastern United States. 15-37 THE NEW ENGLAND BOTANICAL CLUB 22 Divinity Avenue Cambridge, MA 02138 The New England Botanical Club is a non-profit organization that promotes the study of plants of North America, especially the flora of New England and adjacent areas. The Club holds regular meetings, has a large herbarium of New England plants, anda library. It publishes a quarterly journal, RHODORA, which is now in its 94th year and contains about 400 pages a volume. Membership is open to all persons interested in systematics and field botany. Annual dues are $35.00, including a subscription to RHODORA. Members living within about 200 miles of Boston receive notices of the Club meetings. To join, please fill out this membership application and send with enclosed dues to the above address. Regular Member $35.00 Family Rate $45.00 For this calendar year For the next calendar year eee Name Address City & State Special interests (optional): Vol. 94, No. 879, including pages 213-326, was issued September 9, 1992. 409 INFORMATION FOR CONTRIBUTORS TO RHODORA Submission of a manuscript implies it is not being considered for publication simultaneously elsewhere, either in whole or in part Manuscripts should be submitted in triplicate (an original and two xerographic copies) and must be double-spaced (at least ¥”) throughout including tables, figure legends, and literature cita- tions. The list of legends for figures and maps should be provided on a separate page. Footnotes should be used sparingly. Do not indicate the style of type through the use of capitals or under- scoring, particularly in the citation of specimens. Names of genera and species may be underlined to indicate italics in discussions. Specimen citations should be selected critically, especially for common species of broad distribution. Systematic revisions and similar papers should be prepared in the format of ““A Monograph of the Genus Malvastrum,” S. R. Hill, Rhodora 84:1-83, 159- 264, 317-409, 1982, particularly with reference to indentation of keys and synonyms. Designation of a new taxon should carry a Latin diagnosis (rather than a full Latin description), which sets forth succinctly just how the new taxon is distinguished from its congeners. Papers of a floristic nature should follow, as far as possible, the format of “Annotated List of the Ferns and Fern Allies of Arkansas,’’ W. Carl Taylor and Delzie Demaree, Rho- dora 81: 503-548, 1979. For bibliographic citations, refer to the Botanico-Periodicum-Huntianum (B-P-H, 1968) which provides standardized abbreviations for journals originating before 1966. All abbreviations in the text should be followed by a period, except those for standard units of measure and direction (compass points). For standard abbreviations and for guidance in other matters of biological writing style, consult the CBE Style Manual, Sth ed. (original title: Style Manual for Biological Journals). In preparing figures (maps, charts, drawings, photos, etc.) please remember that the printed plate will be 4 x 6 inches; be sure that illustrations are proportioned to reduce correctly, and indicate by blue pencil the intended limits of the figures. (Some “‘turn-page”’ figures with brief legends will be 312 x 6 in.) Magnification/reduction values given in text or figure legends should be calculated to reflect the actual printed size. An Abstract and a list of Key Words should be supplied at the beginning of each paper submitted, except for a very short article or note. All pages should be numbered in the upper right-hand corner. Brevity is urged for all submissions. a ne Statement =f Ownership, - Management and tion =a 8 6Circula SEPSES cubed by 39 U.S.C, 9685) 2. Date of Filing 1A. Title of Publication setts 9/27/92 3. Frequency of Issue E % A m4 Annually Quarterly Four $40.00 New England Botanical Club, Inc., 22 Divinity Ave., Cambridge, MA. 02138 New England Botanical Club, Inc., 22 Divinity Ave., Cambridge, MA. 02138 Sane Publisher Tea oant Conia Main New England Setanicad Ta, Inc., 22 Divinity Ave., Cambridge, MA. 02138 Prof. Norton H. Nickerson, Tufts University-Biology, Medford, MA. 02155 jor (Name and Complete Marling Joan ¥. Nickerson, Tufts gS. Medford, MA. 02155 7. Owner mned by a corporation, ‘and address must be stated and also immediately thereunder the names and addresses of pice meng. dar pron eng Lfoalhipeigha har onic oir ero or mae aon ed see or other unincorporated firm, odin Sip as well as that of each individual mast be given. If the publication is is published hy a nonprofit organization. and sain ig ae pep ch be completed.) The purpose ay (2) hanged During Changed During changed, publisher must submit explanation of LibPreceding 12 Months (Uiteceaing 12 Months with this statement.) xtent and Nature of Circulation ‘Average No. Copies Esch tssue During| Actual No. Copies of Single Issue ag aig yal escape Preceding 12 Months Published Nearest to Filing Date A. Total No. Copies (Wer Press Run} 813 800 8. Paid and/or Requested Circulation 1. Sales through dealers and carriers, street vendors and counter saies | oO ° Pe ect eet Db le ind eT 2 Mos Subscription Jal 754 ee I i ee eae png nerod 1082) — 741 734 REPRE ies dearer Sacanidancaene doe SUD om es TT eee RCI ee Ser D. Free Distribution by Mail, Carrier or Other Means 22 22 Samples, Complimentary, and Other Free Copies E. Totat Distrit and D) 763 776 F. a 50 24 . Office use, left over, unaccounted, spoiled after printing 2. Return from News Agents 0 0 G. TOTAL (Sum of E. Fl and 2—should equal net press run shown in A) 800 oS ‘ piece igs Paden Greauste Burne Morpe sea sa PS Form 3526, January 1991 RHODORA October 1992 Vol. 94, No. 880 CONTENTS Climate and the disjunct distribution of Polystichum alfarii (Christ) Barr., comb. nov. in Mesoamerica Oe A sis ies ek ORT AS eh 327 A naturally-occurring population of — Arisaema triphyllum subsp. stewardsonii x A. dracontium hybrids in Massachusetts La The Asteraceae of the Guianas, III: Vernonieae and restoration of the genus Xiphochaeta iarle BORGO os ee ee OE ene ee 348 Chromosomes of Mexican Sedum VI. Section Sedastrum CT as a oe es i pe ee 362 On the etymology of Vaccinium L es et cee ths 371 Ecological aspects of Arethusa bulbosa, Calopogon tuberosus and Pogonia ophioglossoides — in eastern Newfoundland. II. Partitioning e micro! J. Todd Roland a PE MeO is ns 374 Two new combinations and a name change from the Long Expedition of 1820 George J. Goodman and Cheryl A. Lawson .............0++0++00°° 381 Additions to the flora of Newfoundland. II Stuart G. Hay, Andre Bouchard, Luc ee _ Martin Jean .... 383 Armoracia I. ) n Ohi ; James S. MoCormac a a ey oro ee tee 387 NEW ENG Utricularia acs Walter (Lentibulariaceae) in Massachusetts MORE FE ON i we eh ee ke 391 BOOK REVIEW Plants of Northern British Columbia TM a a ee ee es 393 IN MEMORIAM WE ee I cae as bos wet ee 395 WGI eg eer eet 396 ANNOUN' NEBC Award for the Support of Botanical Research .........----- +: 397 U.S. National Arboretum Herbarium ................-------:+-09°* 398 1992 Jesse M. Greenman Award 22. eee eee 399 List of Mesvaets a ee 400 Index to Vole ea 401 NEDC Memhershig Form. ok i 409 Information for Comtribaters = oo oe eee 0 f&Ovnnle |. 8 a inside back cover