Dodova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 85 January 1983 No. 841 The New England Botanical Club, Inc. Botanical Museum, Oxford Street, Cambridge, Massachusetts 02138 Conducted and published for the Club, by NORTON H. NICKERSON, Editor-in-Chief Associate Editors A. LINN BOGLE GARRETT E. CROW WILLIAM D. COUNTRYMAN RICHARD A. FRALICK GERALD J. GASTONY NORTON G. MILLER ROBERT T. WILCE RHODORA.—Published four times a year, in January, April, July, and October. A quarterly journal of botany, devoted primarily to the flora of North America. Price $20.00 per year, net, postpaid, in funds payable at par in the United States currency at Boston. Some back volumes, and single copies are available. For information and prices write RHODORA at address given below. Subscriptions and orders for back issues (making all remittances payable to RHODORA) should be sent to RHODORA, Botanical Museum, Oxford Street, Cambridge, Mass. 02138. In order to receive the next number of RHODORA, changes of address must be received prior to the first day of January, April, July or October. Scientific papers and notes, relating to the plants of North America, and floristically related areas, will be considered by the editorial com- mittee for publication. Articles concerned with systematic botany and cytotaxonomy in their broader implications are equally acceptable. All manuscripts should be submitted in duplicate, and must be double- (at least % of an inch) or triple-spaced throughout. Please conform to the style of recent issues of the journal. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to: Russell R. Walton Managing Editor, RHODORA Harvard University Herbaria Building 22 Divinity Avenue Cambridge, Mass. 02138 Second Class Postage Paid at Boston, Mass. PRINTED BY THE LEXINGTON PRESS, INC. LEXINGTON, MASSACHUSETTS Cover illustration Trollius laxus Salisb. is very rare as it nears its northeastern limit in Connecticut. Some old records indicate that it has grown in parts of New Hampshire and Maine, so it may yet be found in appropriate habitat. Original artwork by Tess Feltes, Illustrator. Thodora (ISSN 0035-4902) JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 85 January 1983 No. 841 THE TAXONOMY OF VACCINIUM § OX YCOCCUS S. P. VANDER KLOET In Eastern North America, both Fernald (1950) and Gleason (1952) recognized but two specific entities in Vaccinium § Oxycoccus (Hill) Koch, namely, V. oxycoccus L., the small cranberry and V. macrocarpon Aiton, the large cranberry. However, taxonomists who have dealt with the group in the boggy places of the circum- boreal taiga have, as a rule, split V. oxycoccus sensu lato into two or more segregate species, for example, V. oxycoccus sensu Stricto and V. microcarpum (Turcz ex Rupr.) Schmal. (Tutin et al., 1972); Oxy- coccus! quadripetalus Gilib and O. microcarpus Turcz ex Rupr. (Bobrov & Bush, 1967); O. microcarpus and O. palustris Pers. (Polunin, 1959); O. microcarpus, O. quadripetalus, and O. ovalifo- lius (Michx.) Porsild (Scoggan, 1979; Porsild, 1938). The basis for this splitting of Vaccinium oxycoccus s.1. into segre- gates stemmed from the discovery that this species is comprised of populations which have different chromosome numbers: some are diploid, 2n= 24 (Hagerup, 1940; Darrow, et al., 1944; Love & Love 1956; and Jorgensen, et al., 1958); others tetraploid, 2n = 48 (Hagerup, 1940; Darrow, et al., 1944; Hara, 1956; Jorgensen, et al., 1958; Love & Love 1966; and Pojar, 1974); and still others are hexaploid, 2n = 72 (Hagerup, 1928; Tischler, 1934; Rohweder, 1937: Newcomer, 1941; and Jorgensen, et al., 1958). However, gross morphological characters which run parallel to these ploidy levels have been difficult to find. Pedicel indumentum in conjunction with leaf dimension have been used most frequently to separate the 'I will not address the question of generic rank for Oxycoccus at this time but will follow the classification of Sleumer (1941) and Stevens (1969). l Table |. Comparison of morphological characters used by various authors to differentiate among the small cranberries. taxa character V. OXYCOCCUS S.S. V. microcarpum Reference O. quadripetalus O. microcarpus O. ovalifolius O. palustris pedicels pubescent + glabrous + glabrous Porsild (1938) puberulent glabrous Porsild & Cody (1979) puberulent glabrous Tutin et al. (1972) pubescent + glabrous Polunin (1959) puberulent usually glabrous Ohwi (1965) + pubescent glabrous Bobrov & Bush (1967) locations of pedicel bracts below, at, or above the middle near the middle above the middle below the middle usually below the middle below the middle below the middle Porsild (1938) Polunin (1959) Ohwi (1965) Bobrov & Bush (1967) filaments densely hairy puberulent Bobrov & Bush (1967) Tutin et al. (1972) vlopoyy $8 IOA) Table | (cont'd) leaf length 6-10 mm 2-6 mm 6-8 mm Porsild (1938) 4-9 mm 2-4 mm Porsild & Cody (1979) 6—10(—15) mm 3-8 mm Tutin et al. (1972) 2-10 mm 2—6(—8) mm Polunin (1959) 5-12 mm 3-6 mm Ohwi (1965) 8-16 mm 3-7.5 mm Bobrov & Bush (1967) leaf width 2-5 mm 1.5—2 mm 2-3 mm Porsild (1938) 3-6 mm 1-2.5 mm Tutin et al. (1972) 1.5-5 mm 1—2(—3) mm Polunin (1959) 2-2.5 mm Ohwi (1965) 3-6 mm 1—2.5 mm Bobrov & Bush (1967) berry diameter 8-12 mm 5-7 mm 10-12 mm Porsild (1938) 8-14 mm 5-10 mm Porsild & Cody (1979) (6)8—10(—15) mm 5-8 mm Tutin et al. (1972) 7-13 mm 5—7(-—8) mm Polunin (1959) 10 mm 6-7 mm Ohwi (1965) 10-18 mm 5-10 mm Bobrov & Bush (1967) chromosome 2n = 48 2n= 24 Tutin et al. (1972) number 2n = 48 2n = 24 2n = 48 Darrow et al. (1944) [E86 SNIDODAXO § WINIUIDDVA — JOO; y JopueA 4 Rhodora [Vol. 85 diploid V. microcarpum from the tetraploid V. oxycoccus 8.8. (Table 1). The hexaploids (2n = 72) occur infrequently and are thought to be sterile hybrids between V. oxycoccus and V. micro- carpum (Hagerup, 1940; Camp, 1944). But as Table | shows, these characters (as interpreted by the various authors) do not unerringly separate the diploids from the tetraploids nor are pedicel indumen- tum and leaf dimensions invariably correlated: there are plants with glabrous pedicels which have the leaf size of V. oxycoccus s.s. and these have been referred by Porsild (1938) to O. ovalifolius; and then there are those plants which have pubescent pedicels and very small inrolled leaves and which have been referred to O. palustris ssp. microphyllus (Lge.) Love & Love, or by Camp (1944) to the “microcarpoid” phase of O. quadripetalus. In her study on the variability of several vegetative and inflores- cence characters in 50 Vaccinium oxycoccus specimens gathered from six bogs near Durham, New Hampshire, Rodrigues (1963) found that stem color varied from light brown to black; that leaf width varied from | to 4mm and length varied from 3 to 10 mm and had mean dimensions of 2+ 0.5 mm X 6+ | mm which is interesting insofar as these values fall just at the boundary between Porsild’s (1938) concept of V. microcarpum and V. oxycoccus (see table 1); that the leaf blades of several specimens were flat, others slightly revolute, and yet others revolute (conditions she attributed to rela- tive exposure of the plants); and that the leaf apices were mostly pointed and that only a few had slightly rounded apices. She also showed that a significant source of the leaf variation described above arose from between plants rather than within plants. Furthermore, Rodrigues (1963) described all the pedicels as pubescent, bearing a pair of green or red bracts (a few pedicels had none at all) whose shape varied from long and narrow to leaf-like to very small scales. Pedicel number varied from one to four and in 21 specimens arose from the axils of the uppermost reduced leaves, thereby forming an apparent terminal cluster (figure 1); in the remainder, however, the pedicels were in the axils of the lowermost reduced leaves of a normal leafy branch: a condition I have also observed in Cape Breton, Nova Scotia (figure 8) and which results in an inflorescence type usually associated with Vaccinium macro- carpon. Indeed, as Boivin (1967) has noted, “the diagnostic characters [for the small cranberry taxa] are not quite constant and various recom- 1983] Vander Kloet — Vaccinium § Oxycoccus 5 binations of characters occur here and there. He who would here accept two species will eventually be led to accept four, then perhaps eventually eight...!” Another difficulty arises whenever herbarium specimens are examined. A sheet of several fragments (see figures 1, 2, 3, especially their annotations) each bearing one or two pedicels, several of which may be quite glabrous, others quite puberulent, may be the result of a mixed gathering; a likely event since, according to Bobrov & Bush (1967) Oxycoccus microcarpus often occurs with O. quadripetalus in boreal peat bogs. Nonetheless Camp (1944) argued that diploid Vaccinium micro- carpum, which is found in Iceland, Europe, Asia, and northwestern North America but is unknown from Labrador and Greenland, is quite homogeneous and is distributionally disjunct from its sister group V. macrocarpon. Furthermore, according to Camp (1944) Oxycoccus quadripe- talus is a mixture of tetraploid hybrids and their segregates which resulted from the contact O. microcarpus made with V. macro- carpon when it migrated southward in response to the advancing Pleistocene ice sheets. The purpose of this paper is to test Camp’s hypothesis, that in Vaccinium § Oxycoccus there are in fact two rather homogeneous diploid taxa, V. macrocarpum and V. microcarpum respectively, and a tetraploid-hexaploid population comprising a heterogeneous assemblage of hybrids and their segregates, using numerical and experimental taxonomic techniques. MATERIALS AND METHODS To assess morphological discontinuity 26 likely cranberry habi- tats, mostly in eastern North America, such as bogs, raised bogs, boggy barrens, high moors, wet meadows, headlands, and low tun- dras were surveyed between 1969 and 1980 (for geographical details see Appendix). If, after a careful search, a site had more than ten cranberry colonies present, three 200 m grid lines were laid out ina random fashion. At every 10m, the nearest cranberry plant was flagged and identified by number. If there was no plant within | m of the 10-m mark of the tape, the point was declared empty. For every 10 plants flagged, three were drawn at random and perma- nently tagged so that they could be revisited during the flowering Rhodora [Vol. 85 e a E P's ~ eo w 4 rs > 7 & “. e | ia é | wy | / PLANTS OF ALASKA - Figure |. Herbarium specimens showing continuous variation in pedicel pubes- cence and leaf size in V. oxycoccus. 1983] Vander Kloet — Vaccinium § Oxycoccus 7 " ~ Zz ¥ O ws 4 w x a “ F ' PLANTS OF A Figure 2. Herbarium specimens showing continuous variation in pedicel pubes- cence and leaf size in V. oxycoccus. 8 Rhodora [Vol. 85 ya ra ~ pe s f ‘ , a > r af eye »* — * - : een ny & zs “ ~ : y « ay a <7 « FY FISHER SCIENTIFIC Figure 3. Herbarium specimens showing continuous variation in pedicel pubes- cence and leaf size in V. oxycoccus. 1983] Vander Kloet — Vaccinium § Oxycoccus 9 j XZ Qo Fe 2 w 3 a FISHER oe ome acslehacloredien aes Wee Peticede e Me om ® 4 . - Tray ad PLANTS OF ALASKA V dete nags tril ifaw ware ) Monk in t n of 4 junctior r 9 ' VIRION os om ners ao rome sn an Figure 4. Herbarium specimens showing continuous variation in pedicel pubes- cence and leaf size in V. oxycoccus. 10 Rhodora [Vol. 85 io * = 4 s o w 4 w = e & aw Figure 5. Herbarium specimens showing continuous variation in pedicel pubes: cence and leaf size in V. oxycoccus. 1983] Vander Kloet — Vaccinium § Oxycoccus 1] 4 Pi & Zz . Q ¥# x i 4 e _* fay Figure 6. Herbarium specimens showing continuous variation in pedicel pubes- cence and leaf size in V. oxycoccus. 12 Rhodora [Vol. 85 - z = ; FISHER HERBARIUM OF THE NOVA SCOTIA MUSEUM OF SCIENCE Figure 7. Herbarium specimens showing continuous variation in pedicel pubes- cence and leaf size in V. oxycoccus. 1983] Vander Kloet — Vaccinium § Oxycoccus 13 and fruiting season for at least two reproductive bouts. Flowers, fruits and a few mature vines from each number were collected and dried. Voucher specimens are at ACAD?. Soil samples were taken and ecological habitat notes were made for each collection. In addition, a few, usually five or more, berries were harvested from each plant so that germination trials could be carried out and the growth mor- phology of the seedlings compared with that of the female parent. Table 2. Morphological comparison between selected attributes of V. macrocarpon and V. oxycoccus character Ve macrocarpon — V. oxycoccus (n = 136) (n = 246) length of flowering shoot 5.5+3.4 2.3 £0.8 (cm) width of pedicel bracts 1.30.4 0.4+0.2 (mm) leaf width (mm) 22035 2207 leaf length (mm) 8+1.3 ee ie Be ratio w:1 1:2.66 1:2.50 leaf shape narrowly elliptic ovate berry diameter (mm) 12+2 9+2 large seeds / berry 17 +8 8 +5 seed weight (mg/ 100) 91 +28 68 + 17 Next, 114 dried specimens were scored for the following 12 fea- tures: (1) length of fertile shoot; (2) width of bracts on pedicel at anthesis; (3) pedicel indumentum (pubescent, intermediate, or gla- brous); (4) calyx lobes (pubescent at the tip or glabrous); (5) stamen filaments (pubescent, pubescent along the margins, or glabrous); (6) leaf blade (revolute, intermediate, or flat); (7) mean leaf width; (8) mean leaf length; (9) leaf length to width ratio; (10) leaf shape (ovate, narrowly ovate, or narrowly elliptic); (11) indumentum in ?Acronyms follow /ndex Herbariorum (Holmgren & Keuken, 1974), 14 Rhodora [Vol. 85 twigs of the current season (puberulent or glabrous); (12) berry diameter. However, since my sampling space was restricted to northeastern North America, where according to Camp (1944) and Porsild (1938) V. microcarpum is absent, I decided to augment the basic data matrix described above with selected herbarium specimens (cited under specimens examined) from the Northwest Territories where according to Porsild & Cody (1979) V. microcarpum is frequent and V. oxycoccus is rare, known to occur only from a few muskegs adjacent to the Liard and upper Mackenzie Rivers. Furthermore, because my sampling procedure was unbiased towards the inclusion of rare “morphs”, I also added several her- barium specimens which had been identified as Vaccinium oxycoc- cus var. microphyllum or V. oxycoccus var. intermedium or V. oxycoccus Var. ovalifolium or looked somewhat dubious. Whenever a herbarium specimen had a missing feature (e.g. no ripe berries) | would dub in the appropriate value using the Flora of the region where the plant had been collected: thus if fruit was absent from a Keewatin specimen which had been referred to V. microcarpum, | would check Porsild & Cody (1979) for berry size and use their median value for the data matrix. After the discretely scored characters had been converted to multi- State characters, a non-metric coefficient Soy= a was calcu- culated for each pair of specimens; then the similarity coefficients were sorted by way of an unweighted pair-group clustering method based on arithmetic averages (UPGMA). Sneath & Sokal (1973, p. 132 & p. 230) furnish details of these procedures. Since dendrograms are only useful at depicting nearest neighbour relationship and are less reliable at the higher internodes, i.e. at the group level, I also subjected the data matrix to Principal Compo- nent Analysis. In this procedure individuals are ordered along uncor- related axes, while variation in all characters among all the indi- viduals is considered simultaneously (Sneath & Sokal, 1973, pp 245-247). Any major groupings in the sample data are elucidated through this sorting technique. In summary then, UPGMA is very useful in describing the relationship between pairs of individuals, while PCA detects major patterns in the data. For the actual processing, I used the numerical taxonomy system (NT-SYS) written by F. J. Rohlf, J. Kishpaugh and D. Kirk (1974, Figure 8. 419673). 1983] Vander Kloet — Vaccinium § Oxycoccus ee are ~< Oo. Lea ] {] 1 SX ; A a y iH { a 4s ue Add oy MA AT A Yd fd A 5 d sa \af/ tf Be =< ‘Z AN IN Nf } a Aiea a = — V. oxycoccus bearing a “normal” flowering shoot (Vander Kloet 15 16 Rhodora [Vol. 85 State University of New York, Stony Brook). In the PCA routine, however, the character states were not converted to multistate char- acters but were read as continuous, which is justifiable since all the characters scored were ordered. Once the germinated seedlings had had their chilling requirements fulfilled in a cold frame, and had begun to produce new shoots and/or flower buds, several actively growing stem tips or developing floral primordia were fixed, stained, and squashed according to a procedure described by Hall & Galletta (1971) so that the chromo- some number for a number of plants might be ascertained. Finally, to substantiate distribution data, herbarium specimens from North America were examined at A, ACAD, ALTA, BM, CAN, DAO, GH, K, LINN, NEBC, NHA, NY, NYS, PH, QK, TRT, UAC, UBC, and V. RESULTS In the first dendrogram (figure 9), which is based on field data only, initial separation occurred at a similarity coefficient of 0.38 and resulted in the formation of two groups: one whose members fell within the circumscription of Vaccinium macrocarpon and the other V. oxycoccus sensu Fernald (1902 & 1950) and Rodrigues (1963). Inspection of some of the smaller clusters within these spe- cific groups shows that the V. macrocarpon cluster contains more variability than the V. oxycoccus. Thus OTU’s 68, 57, 75, 69, 63 and 66 form a subgroup of + glabrous members in V. macrocarpon, whilst OTU’s 79, 54, 82, 71, 81, 78, 55, 62, 58, and 65 form a subgroup of rather robust and pubescent plants. When the herbarium specimens were added to the matrix, initial separation occurred at a similarity coefficient of 0.35 (figure 10) and again resulted in the formation of two groups, namely Vaccinium macrocarpon and V. oxycoccus s.1. OTU 10 was the only specimen added to the V. macrocarpon cluster; this specimen had been mis- identified as V. oxycoccus. Within the V. oxycoccus cluster several new groups appeared: the most noteworthy of these being OTU 6, a specimen which can be referred to Oxycoccus ovalifolius since it has glabrous pedicels and large flat ovate leaves, and OTUs 2, 3, 44, 9, and 43, a mixture of specimens which had been identified as V. microcarpum and V. oxycoccus var. microphyllum. The remaining V. oxycoccus var. microphyllum specimens (OTU’s 45, 42, 46, 41, 47, and 7) were linked more closely to V. oxycoccus per se. Table 3. Difference in germination of autumnal and vernal collections of Vaccinium § Oxycoccus seeds. Radicle Collection number of % emerges Dicotyledons True leaves Time taxa samples seeds germination — days days days Vo oxvcoccus 19 1067 21 19+ 4 30 + 3 38 +7 Autumnal harvest V. macrocarpon 17 2434 5 26+ 10 38 + 12 47+ 14 V. oxycoccus 18 1057 92 iO=Z 19+ ] 24+ 3 Vernal harvest V. macrocarpon 19 2082 ae 9+ | e22 Px Be a Note: + = one standard deviation. [E861 Snd909AXO § WNIUIDDB A — JOO; Yy JopueA LI 0.2 0.3,- SIMILARITY (2) oO T 0.6 07h i 0.8 0.9}- 7 75696366 28 22 3) 16 30 20 48 23 4039 13 2) 32 I7 18 15 38 29 39/9 14 33 8! 827954 7) 78 55 62 58 65 505273 51 536474 496! 70 76 59 BO G7 72 56 776083 685 Figure 9. Dendrogram depicting relationshi ps among 57 OTU’s of Vaccinium § Oxycoccus specimens from 25 collecting sites in North America. (Details in text.) 81 eIopoyuy $8 1OA] 1983] Vander Kloet — Vaccinium § Oxycoccus as | l Sam-NSonwmnsm TMA-NM———M—-—M- Dendrogram depicting relationships among 83 OTU’s of Vaccinium § Oxycoccus based on both field and Ot—--Mn non -OR@ORH OW WO Figure 10. herbaria collections. (Details in text.) 19 20 Rhodora [Vol. 85 N iva re) | a oO cuit Bownt. CA Stecouns gu Aetna yet ain é A Gate MG. Dumais, Colt fi pw T7é — (i a Figure 12. Herbarium collection containing more than one vine of V. oxycoccus. 1983] Vander Kloet — Vaccinium § Oxycoccus 23 transplanted vines with revolute leaves from exposed hummocks to the shady side of black-spruce clumps and to terraria in the green- house and in both places the next set of innovations had flat leaves. Moreover, seedlings, greenhouse grown from open pollinated ber- ries gathered from V. oxycoccus plants with strongly revolute leaves, produced quite flat leaves during the winter which became progressively more inrolled as the seasons progressed. Chromosome counts revealed no surprises: collections of Vaccin- jum macrocarpon pollen mother cells from Nova Scotia and Onta- rio were consistently diploid (n= 12) and all pollen mother cells of V. oxycoccus from Nova Scotia, Quebec and Ontario were consistently tetraploid (n = 24). Provenance of plants from which counts were made are marked with an asterisk in the appendix of collecting sites. Vouchers are at ACAD. In fine, these analyses suggest that Vaccinium § Oxycoccus con- sists of two rather heterogeneous groups, discriminated primarily on the basis of size (figure 11 and table 2), one of which fits into V. macrocarpon Aiton and the other into V. oxycoccus L. That these clusters ought to be recognized at the species level is further enhanced by the lack of crossing success between the groups. In the areas of sympatry, the groups are temporally separated by two to three weeks at anthesis as well; for example, in Nova Scotia V. oxycoccus blooms on the 27th of June + 13 days and V. macro- carpon on the 19th of July + 9 days. Similar differences in the time of anthesis were observed by Bell & Burchill (1955), who also found that in the resting stage, V. oxycoccus bract, sepal, and stamen primordia are differentiated only, there is no sign of stamen and carpel primordia; whilst in V. macrocarpon the stamen and carpel primordia show some signs of differentiation. TAXONOMY Vaccinium § Oxycoccus (Hill) Koch, Fl. Germ. and Helv. 474. 1837. Oxycoccus [Tourn] Hill, Brit. Herb. 324. 1756. Oxycoccus [Tourn] Adams. Fam. 11. 164. 1763. Oxycoccus [Tourn] Persoon, Syn. Plant. 1.419. 1805. Vaccinium § Oxycoccus [Tourn] A. Gray, Manual Ist ed. p. 260. 1848. Trailing vines; branches woody, slender, flexible, terete, glabrous or pubescent; leaves alternate, persistent, almost sessile, margin 24 Rhodora [Vol. 85 Figures 13,17. Glaucescence of the abaxial leaf surface. 13. Wax rodlets, Vaccin- tum macrocarpon. 17. Wax rodlets, V. oxycoccus. 1983] Vander Kloet — Vaccinium § Oxycoccus 25 entire; flowers 4-merous, solitary or in small clusters, axillary or apparently terminal, nodding on long slender pedicels; corolla white to dark pink, lobes nearly cleft to the base, strongly recurved at anthesis; anthers 8, awnless but with long slender tubules; ovary 4-celled, fruit a red, several seeded, berry. KEY TO THE SPECIES A. Leaves narrowly elliptic, largest usually > 1 cm long; pedicels with leaf-like bracts > 1mm wide ........ 1. V. macrocarpon AA. Leaves ovate, largest usually < Icm long, margin often inrolled; pedicels with red scale-like bracts, < | mm wide, SOIMELMICS AUSONE 5 402.446 bok’ vw wk ole ake 2. V. oxycoccus 1. Vaccinium macrocarpon Aiton, Hort. Kew. ed. i.ii.13. 1789. V. oxyvcoccus var. oblongifolium Michx. Fl. Bor. Am. 1: 228. 1803. V. oblongifolium (Michx) Hort. ex Dun in D.C. Prod. vii. 576. 1824. V. propinquum Salisb. Prod. 291. 1796. V. macrocarpon f. eburna MacKeever, Rhodora 64: 351. 1962. Oxycoccus macrocarpus (Aiton) Pers. Syn. Plant. 1. 419. 1805. Trailing woody vine; innovations from axillary buds, frequently erect or ascending, 4-15 cm high. Twigs of the current season terete, golden brown, glabrous, rarely pubescent. Leaves narrowly elliptic, elliptic, rarely oblong; (2)3—4(5) mm wide, (5)7—10(18) mm long; green above, glaucous below with the wax so thick that it frequently obscures the stomata (figures 13, 14, 15); margin entire, scarcely revolute. Flowers borne singly in the axils of reduced leaves at the base of current shoots. Pedicels slender, 2—3 cm long, glabrous or pubescent, bearing a pair of green bracts |—2 mm wide. Calyx lobes 4, very small. Corolla lobes 4, white to pink, strongly reflexed at anthesis. Stamens 8, filaments usually stiffly pubescent along the margins, rarely entirely pubescent or glabrous; anther sacs awnless, tubules long slender 1-2 mm long; pollen tetrads 32—37 wu in di- ameter. Style 5-7 mm long, glabrous. Berry red, 9-14 mm in diame- ter, locules 4, each of which has 2 to 7 large brown seeds (figure 16). Chromosome number 2n = 24. RANGE: Newfoundland west to central Minnesota south to northern [linois, northern Ohio and central Indiana and in the Appalachian 26 Rhodora [Vol. 85 1983] Vander Kloet — Vaccinium § Oxycoccus 27 Mountains to Tennessee and North Carolina. Its mass centre lies between 40° and 50°N and 70° and 80°W (figure 17). The only report of V. macrocarpon occurring north of 50° latitude comes from Sir Joseph Banks (see Lysaght 1971: 69 and 341 where it is argued that the Banksian specimen came from the Bay of Isles (near Cornerbrook) via Wilkinson from James Cook in 1767 rather than from “The Illettes” near Hare Bay in the vicinity of St. Anthony). Aside from the dubious Banksian record, there are in the various herbaria consulted no other specimens extant to support this claim, nor after a diligent 2-day search of the headlands near St. Anthony did I find the plant. Currently the species is quite rare if not extirpated in Illinois, Ohio, Maryland, North Carolina and Tennessee. Introduced and adventive along the eastern shore of Maryland (Brown & Brown, 1972) as well as on Lulu Island near Vancouver, British Columbia and several localities in Washington and Oregon. In Europe also the species has been introduced and has become feral in Britain, Germany, Switzerland, and the Netherlands (Popava in Tutin et al. 1972). TYPE LOCALITY: “Native of North America”. Introduced at Kew in 1760 by Mr. James Gordon. Type at BM, seen in 1975. According to Lysaght (1971:341) Aiton’s description was based on material from James Gordon as well as on Ehret’s plate and not on Banksian material. HABITAT: Open bogs, swamps, mires, wet shores, headlands, and occasionally on poorly drained upland meadows. Occurrence is restricted to acid soils and peat. The pH for 27 soils and peaty substrates tested ranged from a low of 4.8 in bogs to a high of 6.1 in old fields. In open bogs, V. macrocarpon is frequently associated with the Carex-Chamaedaphne calyculata community at the edge of the floating mat. The species of Carex associated with V. macro- carpon varies, however: for example, Carex canescens (Pers.) Poir in southern New Hampshire (Barrett 1964); Carex rostrata Stokes in the Cranberry Glades of West Virginia (Darlington, 1943); Carex lasiocarpa Ehrh. in northern Michigan (Gates, 1942): Carex nigra (L.) Reichard in Nova Scotia (Hicks, et al., 1968). Figures 14, 15,18. Glaucescence of abaxial leaf surfaces. 14, 15. Vaccinium macro- carpon. 14, Glaucescence thick enough to obscure stomata; 15, Removal of glau- cescence reveals stomata. 18. V. oxycoccus: glaucescence thick enough to obscure several stomata. 28 Rhodora [Vol. 85 1983] Vander Kloet — Vaccinium § Oxycoccus 29 PHENOLOGY: south of 45°N flowering occurs between mid June and mid July and north of 45°N between mid July and mid August. Reproduction is amphimictic, the flowers perfect, protandrous, and entomophilous. Roberts (1978) has found that solitary bees are most efficient at removing pollen from Vaccinium macrocarpon flowers, followed by Bombus spp. with Apis mellifera as the least efficient pollinator; the latter is also inefficient at removing nectar. According to Moeller (1978) in unfertilized flowers the petals turn rosy and persist for several weeks but prompt fertilization causes the petals to drop in a few days. After pollination, 80-85 days are required for seed set; seed pro- duction is a function of pollinator activity. Moeller (1978) found that the plant from which bees were excluded had 6 seeds/ berry compared to 12-14 seeds/ berry where bees had been permitted to forage. Hall & Aalders (1965) found that berry weight is a function of seed number with each additional seed contributing some 36 mg to total berry weight. Although the berries turn a deep red in the autumn, and the large seeds are plump and brown, germination studies by Devlin, et al. (1974, 1976) and Devlin and Karczmarczyk (1975, 1977) among others have shown indubitably that seeds harvested in the autumn require heat, high light intensities, ultraviolet, gibberellic acid, or scarification to induce germination. Under natural conditions, the berries hold through winter, and seeds taken from these berries germinate promptly and en masse without pretreatment unless they have been exposed to frequent inundations of salt water during the winter (tables 3 and 4). 2. Vaccinium oxycoccus Linnaeus, Sp. Pl. 351. 1753. V. palustre Salisb. Prod. 291. 1796. V. microcarpum (Turcz.) Hook, f. Trans. Linn. Soc. xxiii:334. 1861. V. microcarpum (Turcz. ex Rupr.) Schmalt. Trudy-Imp.S. Peterb. Obs¢. Estestv. 2: 149. 1871. V. hagerupii (L. & L.) Ahokas, Ann. Bot. Fenn. 8: 255. 1971. © oxyvcoccus Var. ovalifolium Michx. Fl. Bor. Am. 1: 228. 1803. V. oxycoccus var. intermedium Gray. Syn. Fl. N. Am. ed 2.2 pt. 1: 396. 1886. V. oxyvcoccus var. microcarpum (Turcz.) “~ oxycoccus var. microphylla (Lange) Rousseau & Raymond, Nat. Can. 79: 82. 1952. -_ —_ Figures 16, 19. Seeds and seed coats. 16. Vaccinium macrocarpon. 19. V. OXVCOCCUS 30 Rhodora [Vol. 85 V. oxycoccus f. parvifolia Kurtz. Engl. Bot. Jahrb. 19: 393. 1894. V. oxycoccus f. obovatum Lepage, Nat. Can. 81: 259. 1954, V. macrocarpon f. dahlei Finlay & Core, Castanea 38: 408. 1973. Oxycoccus quadripetala Gilib. Fl. Lituan i:5. 1781. O. europaeus Pers. Syn. Plant. 1: 419. 1805. . palustris Pers. Syn. Plant 1:.419. 1805. . vulgaris Pursh, Fl. Am. Sept. 1814. O. microcarpus Turez. ex Rupr. in Beitr. Pfl. Russ. Reich 4: 56. 1845. O. oxyvcoccus (L.) MacM. Bull. Torr. Bot. Club 19: 15. 1892. O. intermedium (A. Gray) Rydb. Fl. Rocky Mts. 646. 1065. 1917. O O O ee) . ovalifolius (Michx.) Porsild, Can. Field-Nat. 54: 116. 1938. . hagerupii Love & Love, Bot. Not. 114: 40. 1960. . oxycoccus intermedius (A. Gray) Piper, Contrib. U.S. Nat. Herb. 11: 444. 1906. O. palustris ssp. microphylla (Lange) Love & Léve, Univ. Colo. Stud. Biol. Ser. no. 17: 28. 1965. O. palustris var. intermedius (A. Gray) Howell, Fl. N.W. Am. 1:413. 1901. O. oxycoccus var. intermedius (A. Gray) Farwell, Papers Mich. Acad. Sci. 2: 35. 1923. O. quadripetala var. microphylla (Lange) M. P. Porsild, Medd. Gronl. 77: 42. 1930. O. palustris var. ovalifolius (Michx.) Seymour, Am. Mid. Nat. 40: 935. 1953. O. palustris f. microphylla Lange, Consp. Fl. Gronl 2: 267. 1887. Trailing wiry woody vine; innovations from axillary buds, often ascending 1—3cm high. Twigs of the current season terete, very slender, dark brown to red, glabrous to pubescent. Leaves ovate, occasionally elliptic, (1)2—3(5) mm wide, (3)5—6(10) mm long; green above, glaucous below with wax so thick that it frequently obscures the stomata (figures 18, 19); margin entire; frequently revolute and often strongly so. Flowers borne singly in the axils of reduced leaves at the base of current shoots (figure 8), however on most vines the leafy portion of the fertile shoot especially north of 50° latitude does not develop giving the illusion that V. oxycoccus has an inflor- escence consisting of a short rachis bearing 1—4 flowers on long slender pedicels (figures 1, 2). Pedicels slender, 2-3 cm long, glabrous to pubescent, bearing (0)2(5) reddish scaly bracts << 1 mm wide. Calyx lobes 4, very small. Corolla lobes 4, white to deep pink, strongly reflexed at anthesis. Stamens 8, filaments usually stiffly pubescent along the margins, occasionally pubescent, rarely glab- rous; anther sacs awnless, tubules slender | mm long; pollen tetrads 34—46 w in diameter. Style 3-4 mm long, glabrous. Berry, at first punctate, later turning deep red, 6-12 mm in diameter, locules 4, 31 Vander Kloet — Vaccinium § Oxycoccus 1983] No. 102 NORTH AMERICA 2 180 os ©" 200 | 400) 400” 400” 1000 1200 MOO RILOMETERS. S WEST Loner vor Figure 20. Distribution of Vaccinium macrocarpon. Table 4. Germination characteristics of Vaccinium macrocarpon Aiton. Coll. date date #seeds germination first radicle emerges first dicotyledon first true leaf # collected sown sown (%) (days) (days) (days) NS 1 . 27979* 9.XI1.80 25 4 30 53 55 NS 4 . 27979 ‘ 40 0 - =~ -— NS 5 . 27980 . 48 0 ~- =~ NS 1. 71079 i 107 0 =~ -- -- NS 2 . 71079 " 87 0 -- -- ~~ NS 3 . 71079 : 90 0 - - - NS 1 . 81079 7 14] 0 - -- -- NS 1 . 111080 " 142 2 31 4] 54 NS 1 . 131080 e 131 0 ~ -- -- NS 2 . 131080 id 124 3 36 44 60 NS 3 . 131080 7 131 0 -- -- -- Ont. 2 . 281080 . 79 82 14 25 31 Ont. | . 311080 i 49 100 17 27 33 NS1 . 1481 1.1V.81 ae 96 8 17 px NS 1. 1481 . 34 76 9 20 23 NS 1. 1481 . 23 100 9 19 25 NS2 . 1481 60 100 8 17 23 ct vlopoyy $8 10A] NS 1. 17481 21.1V.81 263 99 9 20 25 NS2 . 17481 . 230 93 7 20 a NS 4 . 18481 . 74 94 8 17 24 NS5 . 18481** ‘ 414 22 9 17 24 NS6 . 18481** . 203 9 7 20 24 NS7 . 18481** a 225 I 7 20 24 NS 8 . 18481** i 207 6 9 21 30 NS2 . 27481 28.1V.81 6 100 10 2 26 NS 3. . 27481 . 39 100 10 20 25 NS 1 . 201080*** . 167 44 18 30 36 Ont. | . 311080*** id 19 78 15 24 30 Ont. 3 . 6581 15.V.81 59 40 1] 19 24 Ont. 4 . 6581 . 24 95 1] 21 23 NOTE — _ *seeds stored in a sealed mason jar at 1°C until sown. **berries submerged in brackish pools along headlands. *** ceeds kept in berries and frozen at —20°C until sown. [E861 SNIDOIAXO § WINIUIDDvVA — 190], y JopuRA 33 34 Rhodora NORTH AMERICA [Vol. 85 | | | wm | | Z i | SS | —— if | ia — — i | ad | } | | — ? Le 1 \ SCALE ee Cam ote tae eee wee oe abe eNouerens LAMBERT S AZIMUTHAL EQUAL-AREA PROJECTION SS —___ —1— Figure 21. Distribution of Vaccinium oxycoccus. 1983] Vander Kloet — Vaccinium § Oxycoccus 35 each of which has |—4 large brown seeds (figure 20). Chromosome number: 2n = 24, 48, 72. RANGE: Circumboreal, in North America the species is absent from the Arctic Archipelago including Baffin Island, and extends south- wards to Central Oregon in the Cascades and to Virginia in the Appalachians (figure 21). TYPE LOCALITY: Sweden, Type at LINN, #497./8, seen in 1975, HABITAT: half buried in Sphagnum hummocks of bogs, the muskegs of the taiga and the low arctic tundra. This habitat is best described as oligotrophic with low pH (2.9—3.8), few exchangeable cations (Grandtner, 1960) and little available nitrogen and phosphorus (Small, 1972). PHENOLOGY: Anthesis occurs from mid-May to mid-June in New Jersey, eastern New York, Connecticut, Massachusetts, south- eastern Ontario, Washington and southern British Columbia; from mid-June to mid-July in the Appalachians, Pennsylvania, New Hampshire, Maine, Vermont, northern New York, central Ontario, adjacent Quebec and the Maritimes as well as the West in that area west of 100° longitude and south of 65°N. In the remainder of the continent the species flowers from mid-July to mid-August. Repro- duction is amphimictic, the flowers perfect, protandrous and ento- mophilous. Bombus spp. and solitary bees are the most frequent pollinators. After pollination, 85—90 days are required for seed set in New England, the Maritimes and southern Ontario and Quebec, but only 35-45 days in the low arctic regions. Seed set and seed weight are invariably less than in V. macrocarpon even where the two species share the same bog (table 2). As in Vaccinium macrocarpon, V. oxycoccus berries hold over winter and are dispersed by water, birds, and mammals in early spring. As in V. macrocarpon, seeds collected in the spring germi- nate most readily (table 5), except that the rate of germination is significantly higher in V. oxycoccus in both the vernal and autumnal trials. Moreover berries collected in late October and early Novem- ber contain seeds ready to germinate so whatever the after ripening requirement may be, it has been met before the actual onset of winter in V. oxycoccus but not in V. macrocarpon, which is to be expected since it flowers 3 weeks later therefore also sets seeds 3 weeks later than V. oxycoccus. Table 5. Germination characteristics of Vaccinium oxycoccus L. Coll. date date #seeds germination first radicle emerges first dicotyledon first true leaf # collected sown sown (%) (days) (days) (days) NS! . 16979* 9.X11.80 36 0 -- -- -- NS2_. 16979 . 65 0 —- -- ~~ NS3.. 16979 i 53 3 19 33 50 NS2_. 27980 . 59 5 17 33 59 NS3_ . 27980 . 50 0 -- -- -- NS 2. 81079 " 33 0 -- ae -- NS 3. 81079 " 57 0 -- -- a NS2_. 91079 " 63 3 19 29 36 NS6_ . 131080 ” 75 2 21 32 39 NS7_ . 131080 if 68 4 27 33 44 NS 1. 141080 id 19 5 21 33 39 P.Q.1 . 261080 id 39 79 14 25 29 Ont. 1 . 281080 . 26 26 17 30 36 P.Q.1 . 21180 . 174 \o nN .S i) N N ‘oO 9 elopoyy $8 190A] NS I. 18481 21.1V.81 62 96 9 20 25 NS2.. 18481 . ab) 100 9 20 29 NS3_. 18481 267 89 9 20 29 NS 1. 27481 7 24 100 12 21 30 P.Q.1 . 261080** . ey 98 16 22 re P.Q.1 . 4581 15.V.81 63 99 11 19 23 P.Q.2 . 4581 ‘J 185 96 10 19 23 Ont.! . 6581 ” 214 83 1] 19 23 Ont. 1 . 6581 7 65 84 1] 19 24 PO.1 .. 123681 . 16 87 1] 19 24 PO: 12581 7 106 100 7 16 22 NOTE — *seeds stored in a sealed mason jar at 1°C until sown. **seeds kept in berries and frozen at —20°C until sown. [E861 SNIDOIAXO § WNIUIDDR A — JI0,y JOpurA Lt 38 Rhodora [Vol. 85 ACKNOWLEDGMENTS I wish to thank Dr. I. V. Hall for his critical reading of the manuscript and his valuable suggestions. Secondly, my thanks go to all those curators of herbaria cited above who gave me free access to their Vaccinium collections. This study was supported by NRC grant No. A9559. REFERENCES BARRETT, P. E. 1966. The relationship of some physical and edaphic factors to plant succession in a southern New Hampshire peat bog. M.Sc. Thesis, Univer- sity of New Hampshire. Bett, H. P., & J. BurcHILL. 1955. Flower development in the lowbush blueberry. Can. J. Bot. 33: 251-258. Botvin, B. 1967. Flora of the Prairie Provinces. Phytologia 15: 121-159, 329-446; 16: |—47. Brown, R. G., & M. L. BRown. 1972. Woody plants of Maryland. Distributed by Student Supply Store, University of Maryland, College Park. Camp, W. H. 1944. A preliminary consideration of the biosystematy of Oxycoc- cus. Bull. Torrey Bot. Club 71: 426—437. DarRLINGTON, H. C. 1943. Vegetation and substrate of Cranberry Glades, West Virginia. Bot. Gazette 104: 371-393. Darrow, G. M., W. H. CAmp, H. E. FiscHer, & H. DERMEN. 1944. Chromosome numbers in Vaccinium and related groups. Bull. Torrey Bot. Club 71: 498-506. Devin, R. M., & S.J. KARCZMARCZYK. 1974. The effect of light on cranberry seed germination. Cranberries 38: 3 & 16. 1975. Effect of light and gibberellic acid on the germination of ‘Early Black’ cranberry seeds. Hort. Res. 15: 19-22. 1977. Influence of light and growth regulators on cranberry seed dor- mancy. J. Hort. Sci. §2: 283-288. Devin, R. M., S.J. KARCZMARCZYK, & K. H. Deupert. 1976. The influence of abscisic acid in cranberry seed dormancy. Hort. Science 11: 412-413. FERNALD, M. L. 1902. The variations and distribution of American cranberries. Rhodora 4: 231-237. 1950. Gray’s manual of botany. 8th ed. American Book Co., New York. Gates, F. C. 1942. The bogs of northern lower Michigan. Ecol. Monogr. 12: 215-254. Gieason, H. A. 1952. The new Britton and Brown illustrated flora of the north- eastern United States and adjacent Canada. New York Botanical Garden, New York. GRANDTNER, M.M. 1960. La forét de Beauséjour, comté de Lévis, Québec. Fonds rech. forest, Univ. Laval, Contr. No. 7, 62 p. HaGerup, O. 1928. Morphological and cytological studies of Bicornes. Dansk. Bot. Arkiv 6: 1-26. 1940. Studies on the significance of polyploidy. IV. Oxycoccus. Heredi- tas 26: 399-410. 1983] Vander Kloet — Vaccinium § Oxycoccus 39 HALi, 1. V., & L. E. AALDeERS. 1965. The relation between seed number and berry weight in the cranberry. Can. J. Plant Sci. 45: 292. Hatt, S. H., & G. J. GALLETTA. 1971. Comparative chromosome morphology of diploid Vaccinium species. J. Am. Soc. Hort. Sci. 96: 289-292. Hara, H. 1956. Contributions to the study of variations in the Japanese plants closely related to those of Europe or North America. Pt. 2. Jour. Fac. Sci. Tokyo Imp. Univ., Sect. 3, Bot. 6: 343-391. Hicks, J. L., 1. V. HALL, & F. R. ForsytuH. 1968. Growth of cranberry plants in pure stands and in weedy areas under Nova Scotian conditions. Hort. Res. 8: 104-112. JORGENSEN, C. A., T. H. SORENSEN, & M. WESTERGARD. 1958. The flowering plants of Greenland: a taxonomical and cytological survey. K. Danske Videns- kab. Selskab. Biol. Skrift. 9: 1-172. Love, A. & D. Love. 1956. Cytotaxonomical conspectus of the Icelandic Flora. Acta Horti Gothoburgensis 20: 65-291. LysaGuT, A. M. 1971. Joseph Banks in Newfoundland and Labrador, 1766. Faber & Faber, London. MoeELLER, F. E. 1979. How long must honey bees be present to effectively set a crop of cranberries? Proc. IVth Int. Symp. on Pollination, Oct. 11-13, 1978. Md. Agric. Exp. Sta. Spec. Misc. Publ. 1: 171-173. Newcomer, E. H. 1941. Chromosome numbers of some species and varieties of Vaccinium and related genera. Proc. Am. Soc. Hort. Sci. 38: 468-470. Posar, J. 1974. The relation of the reproductive biology of plants to the structure and function of four plant communities. Ph.D. Thesis, University of British Columbia, Vancouver. PoLuNIN, N. 1959. Circumpolar Arctic flora. Clarendon Pr., Oxford, Eng. 514 p. PorsiILp, A. E. 1938. The cranberry in Canada. Can. Field-Nat. 52: 116-117. ,& W. J. Copy. 1980. Vascular plants of continental Northwest Territories, Canada. National Museum of Natural Sciences, Ottawa. 667 p. Roperts, R. B. 1979. Energetics of cranberry pollination. Proc. [Vth Int. Symp. on Pollination Oct. 11-13, 1978. Md. Agric. Exp. Sta. Spec. Misc. Publ. 1: 431-440. Ropricues, L. A. 1963. A systematic study of variation in the small-fruited cran- berry. M.Sc. Thesis, University of New Hampshire. ROHWEDER, H. 1937. Versuch zur Erfassung der mengemassigen Bedeckung des Darss und Zingst mit polyploiden Pflanzen: ein Beitrag zur Bedeutung der Polyploidie bei der Eroberung neuer Lebensraume. Planta 27: 501-549. ScoGGANn, H. J. 1979. The flora of Canada. pt. 4. Dicotyledoneae (Loasaceae to Compositae). National Museum of Natural Sciences. Publ. in Bot. 7(4). SHISHKIN, B. K. & E.G. Boprov. 1967. Flora of the U.S.S.R. XVIII. Metachla- mydeae. In: Academy of Science of the U.S.S.R. Botanical Institute im V. L. Komarov. Flora of the U.S.S.R. Transl. from Russian. Israel Program for Scientific Translations, Jerusalem. SLEUMER, H. 1941. Vacciniodeen Studien. Bot. Jahrb. Sonder.-Abdr. 71: 375-510. SMALL, E. 1972. Ecological significance of four critical elements in plants of raised bogs. Ecology 53: 498-503. SNEATH, P. H. A., & R. R. SOKAL. 1973. Numerical taxonomy. W. H. Freeman, San Francisco, CA. 40 Rhodora [Vol. 85 STEVENS, P. F. 1969. Taxonomic studies in the Fricaceae. Ph.D. Thesis, Univer- sity of Edinburgh. Tiscuter, G. 1934. Die Bedeutungen der Polyploidie fur die Verbreitung der Angiospermen, erlautert an den Arten Schleswig Holsteins, mit Ausblicken auf andere Florengebiete. Bot. Jahrb. 67: 1—36. Turin, T. G., V. H. Heywoop, eT AL. 1972. Flora Europaea. V. 3. Diapensiaceae to Mvoporaceae. Cambridge University Press, London. DEPARTMENT OF BIOLOGY ACADIA UNIVERSITY WOLFVILLE, NOVA SCOTIA. APPENDIX I. LOCATION OF SAMPLING SITES *Chromosome number determined for one or more plants. 1. Jet. of Webbs Mill Brook and SR539, Ocean County, New Jersey; mire; V. macrocarpon. 2* Mast Way, Lee, Strafford County, New Hampshire; bog: V OXVCOCCUS (N = 24) 2 plants. 3. Quoddy Head, Washington County, Maine: hummocks on high moor; V. OXVCOCCUS,. 4. Drakes Island, Rachel Carson Wildlife Refuge, York County, Maine; red maple swamp: lV. macrocarpon. S* Mud Lake, | km west of Black River Lake, Kings County, Nova Scotia; bog; V. oxveoccus and VF. macrocarpon (n= 12) 6 plants. 6* 4km E of Kingston along Hwy. I, Kings County, Nova Scotia: boggy depres- sions; V. macrocarpon (n = 12) | plant. 7. Little Lorraine, Wild Cove, Cape Breton County, Nova Scotia: rocky head- lands: V. macrocarpon. 8. Leap Frog Lake, Port Maitland, Yarmouth County, Nova Scotia; boggy mar- gins V. oxyvcoccus. 9. &km west of Birchtown along Hwy. 3, Shelburne County, Nova Scotia: raised bog and boggy barrens: V. oxycoccus. 10. Skm N of exit 46 along Hwy. 104, Richmond County, Cape Breton Island, Nova Scotia; raised bog; V. oxycoccus. 11. 1.5km N of West Dover along Hwy. 333, Halifax County, Nova Scotia; boggy barrens: VW. oxveoccus. 12. French Mountain, Cape Breton Highlands, Inverness County, Nova Scotia; extensive high moor; V. oxycoccus. 13. Red Head, Port Maitland, Yarmouth County, Nova Scotia: mire: V OXVCOCCUS; headlands, . macrocarpon. 14* 4km west of West Branch, Pictou County, Nova Scotia: abandoned, poorly drained Festuca ovina meadow; V. macrocarpon (n = 12) | plant. I5* Portuguese Cove Headlands, Halifax County, Nova Scotia; headlands: V macrocarpon (n= 12) 2 plants; boggy depressions, V. oxycoccus (n=24) | plant. 1983] Vander Kloet — Vaccinium § Oxycoccus 4] 16* Kennington Cove, Cape Breton County, Nova Scotia; raised bog, V. oxvcoccus (n = 24) 2 plants; headlands, V. macrocarpon (n = 12) 5 plants. 2km SW of Bonavista, Massive headlands, Newfoundland: boggy depression between headlands; V. oxycoccus. RT185 at P.Q. - N.B. border, Temiscouata County, Quebec: spruce woods and extensive bogs; V. oxvcoccus (n = 24), 19* Villeroy, Route 20, Sortie 158, Lotbiniere County, Quebec; raised bog: V. oxy- coccus (n = 24) 3 plants, V. macrocarpon. Along SR8& at Graphite, Warren County, New York; lake margin: V. macro- carpon. Westport Bog, 4km NE of Westport, Leeds County, Ontario: bog: V. oxveoc- cus, WV. macrocarpon. 22* Byron Bog, suburban London, Middlesex County, Ontario: bog; V. oxvcoccus (n = 24) 3 plants; V. macrocarpon (n = 12) | plant. 1km W of Heart Lake, near Snelgrove, Peel County, Ontario: bog; V. macrocarpon. Mer Bleue, beyond Borthwick Road, Carleton County, Ontario: raised bog: V. oxveoccus, Vo macrocarpon. 25* Hebert Bog, Upper Rock Lake, Frontenac County, Ontario: bog: V. oxvcoccus (n = 24) 6 plants; V. macrocarpon (n= 12) 4 plants. Fraser River Delta at Richmond Rd #5 and the CNR right-of-way, Lulu Island, British Columbia; disturbed raised bog; V. macrocarpon. APPENDIX II Citation of Specimens GREENLAND. Ameralik Fj. Ameragdla Arm. Equaluit, 64°09’N, 50° 22’W, A. E. Porsild 8406 in 1941 (CAN); oTU 2. Amitsuarssuk, 60° 08’N, 44°45’W, Carlo Hansen, Lars Kliim-Nielson, and Ben- Jamin Mlgaard 67—408 in 1967 (CAN); OTU 3. Frederikshaab, 62°, J. Eugenius in 1928 (CAN); oTU 47. Godthaab Fj. S. point of Sadlen Island, A. FE. Porsild 12000 in 1942 (CAN); oTU 42. Godthaab Fj., Qornog Island, 64° 30’N, 51°05’W, A. E. Porsild 8698 in 1942 (CAN): OTU 41. CANADA. Alberta. Anzac, S.E. tip of Gregoire Lake, S.E. of McMurray, M. G. Dumais and K. Anderson 2692 in 1968 (ALTA); OTU 9. 1/2 mi. S. of Ma-Me-O Beach, G. H. Turner MD. 8332 in 1953 (ALTA); OTU 8. Primrose Lake, 6 mi S of tip of lake, M@M. G. Dumais and C. G. Rankin 1287 in 1967 (ALTA); OTU 7. British Columbia. Cape Beale, Lake Kihha, S. Hartwell 62503 (UAC). Georgie Lake, Port Hardy, J. Hett and W. Armstrong 303 in 42 Rhodora [Vol. 85 1964 (uvic). Lulu Island, Vancouver, J. W. Eastham 11,434 in 1944 (ALTA). Wolverine Range, near Manson Creek village, N. of Ft. St. James, 55° 40’N, 124°22’W, J. A. Calder, D.B.O. Saville and J. M. Fergusson 13690 in 1954 (DAO). Nova Scotia. Colchester County, Earltown, A. R. Prince 206 in 1927 (ACAD); oTU S$. St. Paul Island, J. S. Erskine 53,693 in 1953 (DAO). N.W.T. Dumpy Lake, near Eldorado Mine, Port Radium, E end of McTavish Arm, Great Bear Lake, 66°05’N, 118°02’W, Hansford T. Shacklette 2766 in 1948 (CAN); oTU 43. Grassy Island, Thelon River, John S. Tener 259 in 1952 (CAN); oTU 44. Hornby’s Bend. Thelon River, John S. Tener 258 in 1952 (CAN); oTU 45. Keewatin, mouth of Windy River, F. Harper 2347 in 1947 (CAN); oTU 46. MacKenzie Lowlands, Liard R. Valley, 10 mi. N of B.C. boundary, W. W. Jeffrey 84 in 1959 (CAN). Ontario. De Grassi Pt., Lake Simcoe, E. M. Walker in 1894 (TR1). Durham County, Newtonville Bog, J. C. Krug and J. E. Purchase 484 in 1963 (TRT). Grey County, Stewart Lakes, 44°23’N, 80°55’W, G. R. Thaler 182 in 1965 (tRT). Holland Swamp, near Newmarket, W. C. McCalla in 1896 (ALTA); oTU 10. London and Parry Sound, T- J. W. Burgess and D. Burgess in 1880 and 1881 (TRT). Mador, Snake Lake, Hastings in 1907 (TRT). Middlesex, Byron Bog, W. W. Judd in 1956 (TRT). N.W. of Nuelton Lake, mouth of Windy Lake, S.W. Keewatin, Francis Harper 2347 in 1947 (CAN). Wellington County, Brisbane, 3/4 mi. S. just off Hwy. 24, 43°44’N, 80°05’W, G. R. Thaler 119 in 1965 (TRT). Wellington County, Puslinch, J. J. Stroud in 1939 (TRT). Wentworth County, | mi. S.E of Copetown, J. H. Soper and J. K. Shields 4735 in 1950 (tRT). York County, Holland R. Marsh, Roy F. Cain in 1930 (tr). York County, Pottageville, W. R. Watson 2668 in 1926 (TRT). Yukon. Halfway Lakes, 15 mi. N of Mayo, 63°48’N, 135°48’W, W. A. Calder, J. M. Gillett and D. A. Mitchell 4156 in 1949 (Dao). Jensen Flats, 6 mi. on rd. to Paris fr. Granville, 63° 42’N, 138° 33’W, J. A. Calder and L. G. Billard 3163 in 1949 (pao). Peavine Camp, tributary of Porcupine River, 65°48’N, 138°35’W, J. A. Calder and J. M. Gillett 26243 in 1960 (DAO). UNITED STATES Alaska. Eagle River, near Juneau, J. P.- Anderson 635] in 1940 and 1941 (pao). Lake Spenard, Anchorage, Dutilly, Le Page and 1983] Vander Kloet — Vaccinium § Oxycoccus 43 O'Neill 20,274 in 1947 (DAO). 2 mi. N of Seward-Kenai-Anchorage rd. junction, Kenai Peninsula, 60°34’N, 149°35’W, J. A. Calder 5757 in 1951 (pAo). Naknek, W. B. Schofield 195] in 1952 (pao). Sterling Hwy. between Kasilof and Homer, Kenai Peninsula, 60° 20’N, 151° 16’W, J. A. Calder 5417 in 1951 (Dao). Maine. Aroostook Co., Crystal, A. R. Hodgdon and William Countryman 19285 in 1971 (NHA). Cumberland Co., Perley Pond, Sebago, M. L. Fernald, Bayard Long and A. H. Norton 11854 in 1916 (NHA). Washington Co., on hwy. 191 near S.W. corner of Cathance Lake, Edw. & Susan Johnson 279 in 1970 (NHA). Wash- ington Co., Jonesport, C. A. Cheever in 1905 (NHA). New Hampshire. Carroll Co. Mystery Pond, Ossipee, 4. R. Hodgdon and T. Marsdon Jones 3376 in 1939 (NHA). Carroll Co. Ossipee, Heath Pond Bog, /rene M. Storks 328 in 1978 (NHA). Car- roll Co. Tamworth, A. R. Hodgdon, Anthony J. Hodgdon, and Alec Lincoln 7819 in 1954 (NHA). Coos Co. Pittsburg, A. R. Hodg- don and P. Allen, et al. 1194] in 1960 (NHA). Grafton Co. Heath Pond Bog, Paul Barrett and A. R. Hodgdon 15534 in 1965 (NHA). Hillsborough Co., Greenfield, Hogback Pond, Greenfield State Park, A. R. Hodgdon, Peter Allen, Henry Baldwin 12724 in 1963 (NHA). Rockingham Co. Kingston Cedar Bog Pond, 4. R. Hodgdon and F. L. Steele 20002 in 1973 (NHA). Rockingham Co. Kingston Jct Rt. 125 and Rt. 111, Garrett E. Crow 2284 in 1976 (NHA). Strafford Co. Madbury, Barbadack Pond, Lee Rodrigues and A. R. Hodgdon 12415 in 1962 (NHA). Strafford Co. Barrington, near Winkley Pond, A. R. Hodgdon and H. A. Giddings 4612 in 1943 (NHA). Strafford Co. Lee, A. R. Hodgdon and H. Clapp 3383 in 1940 (Nua). Vermont. Windham Co. Whitingham, “floating Is.” anchored eri- caceous bog in Lake Sedoga S. of Wilmington, A. R. Hodgdon 19954 in 1973 (NHA). FINLAND South Savo, Sulkava, Partalansaari, Auvila, S. W. shore of Lake Kikolampi, Marjatta Isoviita in 1976 (ALTA), KALMIA ERICOIDES REVISITED. WALTER S. JUDD ABSTRACT Specific limits within the Cuban taxa of Kalmia (Ericaceae) are reinvestigated. A single species, K. ericoides, with two geographically isolated varieties is recognized. Kalmia ericoides var. ericoides is limited to Pinar del Rio while var. aggregata occurs only on the Isle of Pines. The taxa may be readily separated by the indumentum of their young twigs. Ka/mia simulata is placed in synonymy under var. aggregata. The delimitation of species within the Ka/mia populations of western Cuba (prov. Pinar del Rio) and the Isle of Pines has varied widely (Table 1). Some botanists, e.g., Wood (1961) and Ebinger (1974) have considered all the Cuban Kalmias as members of a single species, Kalmia ericoides (with or without recognized varie- ties), while others, e.g., Roig & Acufia (1957), Alain (1946 & speci- men identifications), and Southall and Hardin (1974) have divided the group into three species, 1.e., K. ericoides, K. aggregata, and K. simulata, based upon variation in pubescence (see Figure 1), com- pactness of the inflorescence, and length of the leaves and/or calyx lobes. Small (1914) recognized only two species, however: at that time K. simulata had not been described (see Britton, 1920). Dis- agreements have also existed as to the geographic distribution of the various recognized taxa (Table |) and the value of the various char- acters used in their identification. Therefore the taxonomic relation- ships between the diverse Cuban populations of Ka/mia are reassessed here through a detailed study of herbarium materials. Kalmia ericoides sensu stricto has usually been characterized by the pilose (i.e., with long-celled hairs) and stipitate-glandular pu- bescence of its leaves and stems. However, the leaves are frequently described as being smooth above (Small, 1914; Roig & Acuna, 1957). Other distinguishing characters are the lax inflorescence (Small, 1914; Roig & Acuna, 1957; Southall & Hardin, 1974; Ebin- ger, 1974), ciliate calyx lobes (Roig & Acufia, 1957), and pubescent filaments (Small, 1914). Kalmia aggregata, described by Small in 1914, is usually distinguished by the densely puberulent and stipitate- glandular pubescence of its leaves and stems, “crowded” leaves (Ebinger, 1974), + compact inflorescence (Small, 1914; Jennings, 1917; Roig & Acuna, 1957; Southall & Hardin, 1974; Ebinger, 45 46 Rhodora [Vol. 85 Table |. Comparative treatment of Cuban taxa of Kalmia by various botanists. (PR = Pinar del Rio; IP = Isle of Pines.) “ericoides” “simulata” “aggregata” botanist variant variant variant Small (1914) K. ericoides — K. aggregata PR IP Jennings (1917) _ K. ericoides — K. aggregata PR IP Roig & Acuna K. ericoides K. simulata K. aggregata (1957) PR, IP IP IP Wood (1961) K. ericoides K. ericoides K. ericoides (geographical distribution of variants not discussed) Southall & K. ericoides K. simulata K. aggregata Hardin (1974) oe bg IP PR, IP Ebinger (1974) K. ericoides K. ericoides K. ericoides var. ericoides var. ericoides var. aggregata PR IP IP this study K. ericoides K. ericoides K. ericoides var. ericoides var. aggregata__—- var. aggregata PR IP IP 1974), and elongate calyx lobes (Southall & Hardin, 1974). Small (1914) considered the filaments to be glabrous, but Jennings (1917) correctly pointed out that the filaments of this taxon are slightly pubescent near the base, as are those of K. ericoides. This species is often stated to have larger leaves than either K. ericoides or K. simulata (see Small, 1914; Ebinger, 1974). Finally, K. simulata, a species described by Britton and Wilson (Britton, 1920), is typically characterized as having nearly glabrous to sparsely stipitate- glandular leaves but puberulent stems. Additional characters in- clude the lax inflorescence (Britton, 1920; Roig & Acufia, 1957: Southall & Hardin, 1974; Ebinger, 1974) and short calyx lobes (Southall & Hardin, 1974; Ebinger, 1974). Although the interpreta- tion of the geographical distribution of these three taxa has varied (see Table 1), a careful study of available herbarium specimens revealed that the plants fitting the general pattern of K. ericoides are 1983] Judd — Kalmia ericoides 47 limited to Pinar del Rio, while individuals agreeing with the descrip- tions of K. aggregata or K. simulata occur only on the Isle of Pines. The range of variation and geographical correlation of the above mentioned characters (along with several additional characters) were surveyed through a study of herbarium specimens from throughout the range of this group. The results of this investigation are summarized in Table 2 and in the taxa descriptions. The overall diversity in the Cuban taxa of Kal/mia is comparable to that exhibited by many North American species of the genus (see Ebinger, 1974). The three described Cuban species differ only in a few indumentum characters, slight (and not too consistent!) differ- ences in leaf and calyx lobe length, and inflorescence structure. Thus Table 2. Variation in selected morphological characters within the Cuban taxa of Kalmia. (LC = long-celled hairs; GH = glandular-headed hairs) character “ericoides” variant “simulata” variant ‘“aggregata” variant unicellular hairs on stem multicellular hairs on stem number of lvs. per cm on stem unicellular hairs on adaxial leaf surface multicellular hairs on adaxial leaf surface leaf length inflorescence unicellular hairs on pedicels unicellular hairs on abaxial sur- face of calyx lobes calyx length filaments lacking or very sparse scattered GH; moderate to dense LC ca. 4-25 lacking to sparse along midvein scattered GH and LC (3-) 4-9 mm lax lacking lacking or witha few near apex or margin 3-5 mm pubescent + dense scattered GH; sometimes a few LC ca. 10-25 lacking to sparse along midvein scattered GH, sometimes also LC (3-) 3.5-8.5 mm lax sparse to dense moderate to dense throughout 3—4.5 mm pubescent + dense scattered to dense GH; often scat- tered LC ca. 9-25 moderate to dense (occasion- ally only sparse) scattered to dense GH, sometimes also LC 5-14 mm lax to compact dense dense throughout (3-) 4-6 mm pubescent 48 Rhodora [Vol. 85 all the Cuban populations of Ka/mia are here considered as a single species: Kalmia ericoides. Within this assemblage the plants of the Isle of Pines were found to differ consistently from those of western Cuba in that their young stems are + densely covered with unicellu- lar hairs (vs. lacking to very sparsely covered with such hairs). Slight, but fairly consistent, differences in indumentum were also found on the pedicels, calyx lobes, and adaxial leaf surfaces (see key). Kalmia ericoides sensu lato thus is considered here to be com- posed of two geographical varieties—the first, var. ericoides, en- demic to Pinar del Rio, and the second, var. aggregata, endemic to the Isle of Pines. Ka/mia simulata is included within K. ericoides var. aggregata because no combination of characters will unambig- uously separate these two supposed taxa—all characters show extensive overlap (Table 2). The leaves of these Isle of Pines plants may lack unicellular hairs on the adaxial surface, be slightly pubes- cent along the midvein, slightly or moderately pubescent through- out, or densely pubescent; the multicellular pubescence varies from a sparse to dense covering of long to short-stalked glandular-headed hairs, sometimes with a few long-celled hairs intermixed. No mor- phological gap exists between the often nearly glabrous “simulata” variant and the densely unicellular-pubescent “aggregata” variant (see especially Killip 45385 and Ekman 12492). In fact, both forms have been collected at the same locality (see Killip 42882, Leon & Marie-Victorin 17852, 17853, 18857, Marie-Victorin & Alain 77, 77a, all from Los Indios), and sometimes even mixed together on the same herbarium sheet. Clearly the Ka/mia populations of this island (or at least the Los Indios area) are variable in leaf pubes- cence. Continuous variation in leaf length, internode length, com- pactness of inflorescence, and calyx length is also present. Ebinger (1974) considered Kalmia simulata to be a synonym of Kalmia ericoides var. ericoides based upon characters of indu- mentum, leaf and calyx lobe length, and compactness of the inflo- rescence. However, as indicated in Table 2 (and in discussion above) these characters do not support Ebinger’s placement of these plants. Thus the basic dichotomy within the Cuban Kalmias is between those populations of Pinar del Rio which have young stems essen- tially lacking unicellular hairs but with a moderate to dense covering of long-celled hairs and scattered glandular-headed hairs, and those of the Isle of Pines which have stems with a + dense covering of unicellular hairs along with a sparse to dense covering of short to 1983] Judd — Kalmia ericoides 49 long-stalked glandular-headed hairs and often a few long-celled hairs. A single plant (Marie-Victorin & Alain 77a) from the Isle of Pines completely lacks unicellular hairs on the young twigs. How- ever, this plant grew together with densely pubescent plants (see Marie-Victorin & Alain 77) and can be identified by its leaf indumentum. Keys and descriptions of the Cuban taxa of Kalmia follow. Kalmia ericoides Wright ex Griseb. Cat. Pl. Cubensium 51. 1866. An evergreen erect to decumbent-spreading, sparsely branched shrub to | (—1.4) m tall with a thickened or burl-like stem just below soil surface; frequently sprouting from base after fire or disturbance. Bark of larger branches dark brown and longitudinally furrowed. Twigs terete, light gray to reddish, slender to stout, densely covered with unicellular hairs to lacking such hairs, sparsely to densely covered with multicellular, multiseriate, short- to long-stalked glandular-headed hairs, and lacking to densely covered with multi- cellular, multiseriate, long-celled hairs. Buds minute, to 0.3 mm long, densely covered with glandular-headed and long-celled hairs. Leaves alternate and sparsely to densely distributed along stem (ca. 4-25 (—30) leaves per cm); blade linear to ovate, (3—) 3.5-l14 mm long, 0.5-3 (4) mm wide, coriaceous; apex acute; base cuneate to rounded; margin strongly revolute; adaxial surface densely covered with unicellular hairs to lacking such hairs, very sparsely to densely covered with short- to long-stalked glandular-headed hairs, and lacking to sparsely covered with long-celled hairs; abaxial surface sparsely to densely covered with unicellular hairs, occasionally such hairs completely lacking, with scattered short- to long-stalked glandular-headed hairs and sometimes long-celled hairs; petiole essentially lacking to 1.5 mm long with a unifacial vascular bundle. Flowers 5-merous, solitary (or occasionally in fascicles or short racemes), in the axils of leaves or bract-like leaves at ends of branches, thus forming a pseudoterminal cluster. Pedicels 4-14 mm long, densely covered with unicellular hairs to lacking such hairs, with scattered glandular-headed hairs and sometimes also long- celled hairs; bracteoles two, opposite and basal, ovate-triangular to narrowly triangular, |—-3 mm long, with unicellular hairs and glandular-headed hairs. Calyx lobes narrowly triangular to ovate- triangular, 3-5 mm long, 0.8-2.3 mm wide, green, with slightly 50 Rhodora [Vol. 85 acuminate to acute apices, tardily deciduous in fruit; adaxial surface densely covered with unicellular hairs to lacking such hairs, with or without scattered glandular-headed and/or long-celled hairs; abax- ial surface with a few unicellular hairs near apex to densely covered with such hairs, with scattered glandular-headed hairs, sometimes also long-celled hairs. Corolla with a short cylindrical tube extend- ing into a shallowly lobed and rotate limb, with saccate depressions in which the anthers are held under tension, 6-12 mm long, 8-17 mm wide, white to pink; adaxial surface sparsely covered with uni- cellular hairs toward base (tubular portion); abaxial surface with glandular-headed and sometimes also long-celled hairs, occasionally Figure |. Pubescence of Kalmia ericoides. a—c, Cross-sections of leaves, X 25: a, K. ericoides var. ericoides. b, K. ericoides var. aggregata, “glabrous” extreme. c, K. ericoides var. aggregata, pubescent extreme. d-i, Hairs (adaxial leaf surface), X 75: d, unicellular hairs. e-h, glandular-headed hairs. i, long-celled hair. 1983] Judd — Kalmia ericoides 51 also with a few unicellular hairs. Filaments sparsely to moderately covered with unicellular hairs near base and glabrous above, 3—5 mm long; anthers ovoid, 0.7—1.2 mm long, opening by large, + terminal, slit-like pores; pollen released in tetrads with viscid strands. Ovary with axile placentation and centrally located, bi- lobed placentae; style 3.5-7 mm long. Fruit a septicidal capsule, subglobose to ovoid, |.7-3 mm long, 24 mm wide, sparsely to densely covered with glandular-headed hairs, sometimes also with a few unicellular hairs. Seeds brown, ovoid, 0.4—-0.7 mm long, the testa of slightly elongated and minutely pitted cells, not extending past the ends of the seed; embryo minute. KEY TO VARIETIES 1. Stems lacking unicellular hairs to very sparsely unicellular- pubescent; pedicels lacking unicellular hairs; abaxial surface of calyx lobes lacking unicellular hairs or with only a few near apex or along margin; adaxial surface of leaves lacking unicellu- lar hairs to sparsely covered with such hairs along mid-vein; lee te WON peg vedeaeew eu yed ih en Bas var. ericoides I. Stems + densely covered with unicellular hairs; pedicels very sparsely to densely covered with unicellular hairs; abaxial surface of calyx lobes moderately to densely covered with uni- cellular hairs throughout; adaxial surface of leaves lacking uni- cellular hairs to densely covered with such hairs; [Isle of Pines] were eee ee rer eee er ee var. aggregata Kalmia ericoides Wright ex Griseb. var. ericoides Chamaedaphne ericoides (Wright ex Griseb.) Kuntze, Rev. Gen. Pl. 2: 388. 1891. Kalmiella ericoides (Wright ex Griseb.) Small, North Amer. FI. 29: 54.1914. Type: Cuba, Pinar del Rio: Guane, near La Grifa, C. Wright 2199 (HOLOTYPE: GOET, not seen; ISOTYPES: GH, MO!, NY (3 sheets)!, Us). Stems lacking unicellular hairs or very sparsely covered with such hairs, with a moderate to dense covering of long-celled hairs and glandular-headed hairs. Leaves (3—) 4-9 mm long, 0.5—1.7 (—3) mm wide; adaxial surface lacking unicellular hairs or only sparsely covered with such hairs on mid-vein (especially near base), with scattered glandular-headed hairs and conspicuous, long-celled hairs (especially along margin near base). Pedicels lacking unicellular hairs, usually longer than subtending leaves, thus giving inflores- cence an open appearance. Calyx lobes 3—S mm long; abaxial sur- 52 Rhodora [Vol. 85 face with a few unicellular hairs near apex and along margin, with scattered glandular-headed and long-celled hairs. Capsules lacking unicellular hairs. (Figure 1; see also Southall & Hardin, 1974). inflorescence a compact to open appearance. Calyx lobes 3-6 mm long; abaxial surface moderately to densely (or occasionally only sparsely) covered with unicellular hairs throughout, with a sparse to dense covering of glandular-headed hairs, often also with long- celled hairs. Capsule with or without unicellular hairs. (Figure 1; see also Jennings, 1917; Marie-Victorin & Léon, 1944; Roig & Acufia, 1957; Southall & Hardin, 1974.) DISTRIBUTION AND ECOLOGY. Cuba, Isle of Pines (Figure 2), in whitesand savannas and pinelands (dominants: Pinus tropicalis, P. caribaea, Colpothrinax wrightii, and Acoelorraphe wrightii); for detailed discussion of vegetation along with lists of characteristic species see Jennings (1917), Marie-Victorin & Leon (1944), Alain (1946). Flowering from November through May. DISTRIBUTION AND ECOLOGY. Cuba, prov. Pinar del Rio (Figure 2), in white-sand savannas; associated species briefly discussed by Marie-Victorin & Leon (1944). Flowering from November through May (June). REPRESENTATIVE SPECIMENS: Cuba, prov. Pinar del Rio. Sabanalamar, El Sabalo, Bro. Alain 1326 (GH, Us); Arroyo Mantua, Damuji, near Rincon del Prado, Ekman 11024 (Ny); La Grifa, Laguna Larga, Ekman 18/65 (Ny, us); Laguna de Alcatraz Grande, Remates de Guane, Bros. Leon & Marie-Victorin 18706 (GH, US). Kalmia ericoides Wright ex Griseb. var. aggregata (Small) Ebinger, Rhodora 76: 389. 1974. Kalmiella aggregata Small, North Amer. Fl. 29: 54. 1914. Kalmia aggregata (Small) Copeland, Amer. Midl. Nat. 30: 571. 1943. Type: Cuba, Isle of Pines: Los Indios, 17 May 1910, Jennings 324 (HOLOTYPE: NY!; ISOTYPE: MO!). Kalmiella simulata Britton & Wilson, Mem. Torrey Bot. Club 16: 93. 1920. Kalmia simulata (Britton & Wilson) Southall, Jour. Elisha Mitchell Sci. Soc. 90: 22. 1974. Type: Cuba, Isle of Pines, vicinity of Los Indios, 13 Feb. 1916, Britton, Britton & Wilson 14205 (HOLOTYPE: NY!). Stem + densely covered with unicellular hairs, with sparse to dense covering of glandular-headed hairs and often with long-celled hairs. Leaves (3—) 3.5-14 mm long, 0.5—3 (+4) mm wide; adaxial 64 Figure 2. Distribution of Kalmia ericoides var. ericoides (circles) and K. ericoides var. aggregata (dots). [€861 Saploole eluyey — ppne es 54 Rhodora [Vol. 85 surface lacking to densely covered with unicellular hairs through- out, with sparse to dense covering of glandular-headed hairs and sometimes a few long-celled hairs. Pedicels very sparsely to densely covered with unicellular hairs, variable in length, thus giving the REPRESENTATIVE SPECIMENS: Cuba, Isle of Pines: San Pedro, Britton, Britton & Wilson 14146 (F, GH, MO, NY, US); Santa Barbara, Westport, Ekman 12096 (Ny); Loma Dagnillo, Ekman 12492 (Ny); Los Indios, Killip 42882 (F, GH, NY, US); Playa Roja, Killip 43001 (Gu, NY, US); between Mina de Oro and Playa del Soldado, Killip 45385 (GH); Los Indios, Bros. Leon & Marie-Victorin 17852 (GH, US), 17853 (GH), 18857 (GH, US); Los Indios, Bros. Marie-Victorin & Alain 77 (GH, NY, US), 77a (US). ACKNOWLEDGMENTS I thank the curators at F, GH, MO, NY, and us for the loan of specimens for this study. LITERATURE CITED ALAIN, HNo. (=A. H. LioGierR) 1946. Notas taxonomicas y ecologicas sobre la flora de Isla de Pinos. Contr. Ocas. Museo Hist. Nat. Colegio “De La Salle” 7: 13-106. Britton, N. L. 1920. Description of Cuban plants new to science. Mem. Torrey Bot. Club 26: 57-118. EBINGER, J. E. 1974. A systematic study of the genus Ka/mia (Ericaceae). Rho- dora 76: 315-398. JENNINGS, O. E. 1917. A contribution to the botany of the Isle of Pines, Cuba. Ann. Carnegie Museum 15: 19-290. MARIE-VICTORIN, FRERE & FRERELEON. 1944. Itineraires botaniques dans I'Ile de Cuba. Contr. Inst. Bot. Univ. Montreal 50: 11-410. RoiG, J. T. & J. ACUNA. Ericaceae in: Hno. Léon & Hno. Alain. 1957. Flora de Cuba, vol. 4. Contr. Ocas. Museo Hist. Nat. Colegio “De La Salle” 10: 91-106. SMALL, J. K. 1914. Ericaceae. North American Flora 29: 33-102. SOUTHALL, R. M. & J. W. HARDIN. 1974. A taxonomic revision of Kalmia (Eri- caceae). Jour. Elisha Mitchell Sci. Soc. 90: 1-23. Woop, C. E., Jr. 1961. The genera of Ericaceae in the Southeastern United States. Jour. Arnold Arbor. 42: 10-80. DEPARTMENT OF BOTANY UNIVERSITY OF FLORIDA GAINESVILLE, FLORIDA 32611 THE NATIONAL HISTORICAL DISTRIBUTION OF PLATANTHERA PERAMOENA (A. GRAY) A. GRAY (ORCHIDACEAE) AND ITS STATUS IN OHIO Davip M. SPOONER AND JOHN STEPHEN SHELLY ABSTRACT An updated county distribution map of Platanthera peramoena was made utilizing data from state natural heritage programs and similar organizations. Herbarium and field surveys were conducted by Ohio Division of Natural Areas and Preserves staff for Ohio localities of the species. These studies identified 41 sites of P. peramoena in Ohio, and habitats were identified where new populations are likely to be found. In Ohio, P. peramoena is a plant of wet acidic soils on the Illinois till plain and in the unglaciated Allegheny plateau. While not common in the state it is not in immi- nent danger of extirpation. Platanthera (Habenaria) peramoena (A. Gray) A. Gray (Purple Fringeless Orchid) was placed under review as a threatened species by the U.S. Fish and Wildlife Service (the Service), as a result of the 1974 Smithsonian Institution’s Report on endangered and threa- tened species of the United States. This report was requested by Congress in the Endangered Species Act of 1972 (Public Law 93-205, Approved Dec. 29, 1973). Subsequent to this proposal, the Service received additional information on this species from various states, and on December 15, 1980, P. peramoena was formally with- drawn from consideration because it was found to be more wides- pread than was previously believed (U.S. Fish and Wildlife Service, 1980). The formal review of this and other species by the Service focused attention on their possible rarity. Partly as a result of this listing, Platanthera peramoena was included on many newly developing state and regional rare plant lists. Attention was further focused when the Service contracted various state natural heritage programs and other similar agencies to study the status of plants listed within their respective states. The Ohio Department of Natural Resources, Division of Natural Areas and Preserves, manager of the Ohio Nat- ural Heritage Program, entered into a formal two-year study of this species. a ) 56 Rhodora [Vol. 85 METHODS Ohio records of P. peramoena were obtained from the Ohio Nat- ural Heritage Program data base. These records were obtained from a survey of 26 Ohio herbaria and the U.S. National Herbarium. Field work was planned utilizing these records as guides to previous occurrences of the species. When the locational data was specific enough, the old records were field checked. The following organizations were consulted to obtain updated distributional information outside of Ohio: The Arkansas Natural Heritage Inventory The Indiana Natural Heritage Program The Illinois Department of Conservation The Kentucky Nature Preserves Commission The Maryland Natural Heritage Program The Mississippi Natural Heritage Program The Missouri Department of Conservation The Conservation and Environmental Studies Center, Inc. of New Jersey The North Carolina Natural Heritage Program The South Carolina Heritage Trust Program The Tennessee Natural Heritage Program The Tennessee Valley Authority Regional Natural Heritage Project The Virginia Natural Diversity Information Program The Western Pennsylvania Conservancy The West Virginia Natural Heritage Program In addition, recent state and regional floras (references 3, 9, 10, 11, 16, 19, 21, 25, 27, 30, 33), state endangered species publications (4, 12, 18, 20, 26, 31, 34), and regional treatments of orchids (7, 8, 13, 14, 23, 24), were consulted for relevant information. RESULTS Total Distribution The only published county dot map of Platanthera peramoena for its entire range is that of Ayensu (1974). Literature references and heritage information added new records. New maps were made for the distribution of this species in Ohio (Map 1), and for its national historical range (Map 2). 1983] Spooner & Shelly — Platanthera a7 om pick ® a ie an @) a ® [UL ® Map |. Ohio distribution of Platanthera peramvoena. The lines to the north and west are the limits of Wisconsin glaciation. The numbers in the counties represent the number of stations reported in the counties. Uncircled numbers represent popula- tions located since 1978: circled numbers represent populations not relocated since this date. The date is the most recent record for the county. B refers to a reference from Braun (1967). Platanthera peramoena is often listed or diagrammed as occurr- ing in New York. This is apparently based on Hotise (1924). Zenkert (1934), however, pointed out that this report was due to an error in transcription, P. peramoena being substituted for P. fimbriata. She- viak (personal communication, 1981) checked the New York State Museum (NYS), and found no New York specimens of P. peramo- ena. Another possible source of error may have been misidentifica- 58 Rhodora [Vol. 85 Map 2. Total historical distribution of Platanthera peramoena. 1983] Spooner & Shelly — Platanthera a tions with P. psycodes or P. grandiflora. Both of these errors were found in old New York specimens deposited at The Ohio State University (0s). Platanthera peramoena is also often listed or diagrammed as occurring in Georgia. Correll (1950) lists it for Georgia, citing S. B. Buckley as the collector. Duncan and Kartesz (1981) also list it for Georgia. Many of Buckley’s labels read “in montibus Carolinae et Georgiae.” We have records from 26 Ohio herbaria, and sent in- quiries to the New York State Museum (Nys), the University of Georgia (GA), the U.S. National Herbarium (us), the Missouri Botanical Gardens (Mo), the Philadelphia Academy of Natural Sciences (PH), and Yale University (vy). Correll suggested the last four herbaria, as he obtained many of his records there. We have found no Georgia record for this species. If it exists in another herbarium, it can only doubtfully be ascribed to Georgia if the locational data is like the one above. Status in Ohio Over the past three years, the Division of Natural Areas and Preserves field staff has found many new sites for Platanthera peramoena. In many cases, old records have been relocated. Post- 1978 populations are now known from 41 locations in nine counties. Herbarium specimens exist from ten other counties. In addition, Braun (1967) maps an Adams County record. We were not able to locate any specimen from this county, or pre-1967 specimens from Brown and Scioto counties. In Ohio, Platanthera peramoena grows in areas that are wet in the spring, but which may become droughty later in the year. It grows most often in full sun or partial shade, and is occasionally found in deep shade. The most common habitat is an open swampy area or low wet depression, and in such situations populations can include scores of individuals. It is infrequent on exposed moist slopes, and in such situations both individuals and populations are generally smaller. One individual was found in an area that was extensively trampled and grazed by cattle. The most vigorous populations are known from moderately disturbed sites, such as periodically mowed or lightly grazed wet fields, or recently logged swampy areas. All presently known populations occur on seasonally wet acidic soils of the unglaciated Allegheny Plateau, or on the highly leached acidic soils of the Illinois Till Plain, south of the Wisconsin Till 60 Rhodora [Vol. 85 Plain border. Swampy habitats on the unglaciated Allegheny Pla- teau occur on lacustrine silts and clays deposited in many of the deeper pre-glacial valleys of the Teays River system and its tribu- taries. These lacustrine deposits were laid down during a damming of the Teays River by the Kansan or pre-Kansan glacier, forming a vast lake, which innundated the lower elevations of much of south- ern Ohio and adjacent West Virginia and Kentucky (Stout et al., 1943). DISCUSSION Platanthera peramoena is not presently an exceptionally rare plant in Ohio. There are a number of reasons, however, which may explain why it is not commonly seen and collected. First, it has a rather brief blooming period, and when not in bloom it is very inconspicuous among the surrounding tall vegetation. Most Ohio records are from mid-July to mid-August, but the blooming period in any one year lasts only about two weeks, and it quickly goes out of bloom. Second, even at a short distance, it is very similar in appear- ance to Phlox paniculata L. and Phlox maculata L. The size, gen- eral form of the inflorescence, flower color, and habitat of these plants are similar, and Platanthera peramoena is often found grow- ing with them. These phloxes are very common and widespread in Ohio, and P. peramoena is easily overlooked. Third, many popula- tions of P. peramoena occur in swampy, weedy habitats that are not particularly pleasant to explore. Fourth, and perhaps of greatest importance, populations do not bloom every year. This phenome- non is known for many orchid species and dramatic changes in number of blooming plants per population have been noted for P. flava (Buker, 1980) and P. peramoena (Henry & Buker, 1955; Henry, Buker, & Pearth, 1975). The reason for these yearly changes is not understood, but there are various possible explanations. Case (1964), Sheviak (1974), and Stoutamire (1974) point out that disturb- ance can cause rapid changes in wild orchid populations. This rapid colonization could come about by the reseeding of an area from an outside source, well documented for other orchid species (Stouta- mire, 1974). Another possible reason for shifts in the population size may be the delayed germination of orchid seeds, but this is not well documented. Another reason is the possible periodic shift of the plant from an autotrophic to a temporary heterotrophic state. This 1983] Spooner & Shelly — Platanthera 61 phenomenon is known for Triphora trianthophora (Sw.) Rydb., and Sheviak (1974) has observed it for other orchid species under horticultural conditions. Case (1964), however, doubts that this is a cause for population shifts in nature. This autotrophic to heterotro- phic shift might very possibly be triggered by environmental condi- tions that are as yet very poorly understood. Sheviak (1964) points out that these conditions may operate on the mycorrhizae of the orchid. Clearly, this is a fertile area for research. The general distribution of Platanthera peramoena on the Illinois Till Plain of Ohio parallels the situation in Illinois, where a similar Wisconsinan-Illinoian edaphic transition exists. Similar wet acidic soils developed on silts of lacustrine origin in southeastern Ohio. The occurrences of P. peramoena in Franklin County (one record, 1896-“Clintonville”), and in Wayne County (one record, 1906- county only mentioned) are apparently disjunct. Soil surveys of Franklin County (McLoda & Parkinson, 1980) and of Wayne County (Bureau & Scherzinger, in press) were checked to see if there were any areas that might be favorable for the growth of this spe- cies. There were indeed many sites in these counties that seemed to have similar conditions of moisture and acidity. While the majority of Wisconsinan-age soils in Ohio are generally calcareous, there are also many locally wet acidic soils. It is not understood why P. peramoena apparently has not colonized them. Platanthera peramoena has varying degrees of rarity in different parts of its range. It is believed extirpated from Delaware (Tucker et al., 1979), South Carolina (pers. comm., SC Heritage staff), and possibly New Jersey (Snyder & Vivian, 1981). Reveal and Broome (1981) believe it to be in danger of extirpation from Maryland due to a combination of habitat modification and overcollecting. It is listed as a “special concern” species (due to vulnerability of its habi- tat) in Alabama (Freeman et al., 1979), Tennessee (pers. comm., TN Heritage staff), and Virginia (pers. comm., VA Heritage staff). Penn- sylvania lists it as threatened, but it is widespread throughout the southern part of the state (Wiegman, 1979). It recently has been found to be much more widespread than was originally believed in Kentucky (pers. comm., KY Heritage staff), and West Virginia, where it has been taken off the West Virginia state list (pers. comm., WV Heritage staff). Sheviak (1974) mentions it as widespread in southern Illinois, and believes it to be the most common member of its genus in southern Illinois. 62 Rhodora [Vol. 85 In Ohio, Platanthera peramoena is not in imminent danger of extirpation. Populations are believed to have disappeared after cul- tivation of an area, and drainage of the habitat would very likely eliminate populations, but habitats exist in southern Ohio that sup- port apparently healthy populations. We believe that moderate dis- turbance has increased its numbers in the state and that there are many as yet undiscovered stations for it in southern Ohio. The increasing awareness of endangered habitats and species is stimulating the development of rare species lists and their refine- ment through selective and planned field surveys. Many lists are initially based on randomly collected herbarium data. Data bases of the occurrences of these species are under various stages of devel- opment in different states, and new records are being obtained by these selective field surveys. Studies of species being considered for federal listing, such as Platanthera peramoena, could especially benefit from heritage program data. They are studied throughout their range, and not just in peripheral areas, where they are consid- ered rare. The information gained through natural heritage program surveys will prove to be rich sources of data in future years. ACKNOWLEDGMENTS We thank Kenneth Powell and Alexander Ritchie, Ohio Depart- ment of Natural Resources, Division of Lands and Soils, for help with Ohio soils information; Andrew Robinson, U.S. Fish and Wildlife Service, and Donovan Correll, Fairchild Botanical Garden, for information about the Georgia record; Charles Sheviak, New York State Museum, for pointing out Zenkert’s 1934 correction and searching for possible New York specimens; Warren Stoutamire, University of Akron, and Barbara Andreas, Cuyahoga Community College, for useful suggestions and critique of this paper; and the various regional and state heritage programs for the new data on county occurrences. LITERATURE CITED 1. Ayensu, E.S. 1975. Endangered and threatened orchids of the United States. Am. Orchid Soc. Bull. 44: 384-394. .& R.A. DeFivipps. 1978. Endangered and threatened plants of the United States. Smithsonian Institution World Wildlife Fund, Inc. Washington, D.C. 403 p. 3. Braun, E. L. 1967. The Monocotyledonae [of Ohio]: Cat-tails to orchids. With Gramineae by C. G. Weishaupt. Ohio State University Press, Columbus, Ohio. 464 p. 2. 1983] Spooner & Shelly — Platanthera 63 4. 22. Broome, C. R., J. L. REveAL, A. O. TucKer, & N. H. Ditt. 1979. Rare and endangered vascular plants of Maryland. U.S. Fish and Wildlife Service, New- ton Corner, Massachusetts. 64 p. . Buker, W. 1980. Population explosions among the orchids. Castanea 45: 144— 145. . Bureau, M. F., & R. J. SCHERZINGER. (to be published in 1983). Soil survey of Wayne County, Ohio. USDA Soil Conservation Service, U.S. Government Printing Office, Washington, D.C. . Case, F. W., Jr. 1964. Orchids of the western Great Lakes. Cranbook Insti- tute of Science, Bloomfield Hills, Michigan. 147 p. . Correct, D. S. 1950. Native orchids of North America north of Mexico. Chronica Botanica Company, Waltham, Massachusetts. 399 p. . Cusick, A. W., & G. M. SILBERHORN. 1977. The Vascular Plants of Ungla- ciated Ohio. Ohio Biol. Surv. Bull. (New Series) Vol. 5, No. 4. 157 p. . DeamM, C. C. 1940. Flora of Indiana. Burford Printing Co., Indianapolis. 1236 p. . Duncan, W. H., & J. T. KArtesz. 1981. Vascular Flora of Georgia. The University of Georgia Press, Atlanta. 143 p. . FREEMAN, J. D., A. S. Causey, J. W. SHort, & R. R. Haynes. 1979. Endan- gered, threatened, and special concern plants of Alabama. Department of Botany and Microbiology, Agricultural Experiment Station, Departmental Ser- ies No. 3, Auburn University, Auburn, Alabama. 25 p. . Henry, L. K., & W.E. Buker. 1955. Orchids of Western Pennsylvania. Ann. Carnegie Mus. 33: 299-346. ,&D.L. PEARTH. 1975. Western Pennsylvania Orchids. Cas- tanea 40: 93-168. . Houst, H.D. 1924. Annotated list of the ferns and flowering plants of New York State. New York State Museum Bulletin, The University of the State of New York. 759 p. . Lowe, E.N. 1921. Plants of Mississippi. A list of flowering plants and ferns. Mississippi State Geological Survey, Bulletin 17. 292 p. . McLopa, N. A., & R. J. PARKINSON. 1980. Soil Survey of Franklin County, Ohio. USDA Soil Conservation Service, U.S. Government Printing Office, Washington, D.C. 188 p. + 69 pl. . Massey, J.R.,& R.D. WHETSTONE. 1978. Threatened and endangered vascu- lar plants of western North Carolina. Southeastern Forest Experiment Station, Highlands Biological Station, Cooperator. . MOHLENBROCK, R.H.,& D.M. Lapp. 1978. Distribution of Illinois vascular plants. Southern Hlinois University Press, Carbondale and Edwardsville. 282 p. . Porter, D.M. 1979. Rare and endangered vascular plant species in Virginia. U.S. Fish and Wildlife Service. . RApForD, A. E., H. E. AHLES, & C. R. Bett. 1964. Manual of the vascular flora of the Carolinas. The University of North Carolina Press, Chapel Hill, North Carolina. 1183 p. ReveaL, J. L. & C. R. Broome. 1981. Minor nomenclatural and distribu- tional notes on Maryland vascular plants with comments on the state’s proposed endangered and threatened species. Castanea 46: 50-82. 64 Rhodora [Vol. 85 23. REED, C. F. 1964. Orchidaceae of Maryland, Delaware, and the District of Columbia. Castanea 29: 77-109. 24. SHEVIAK, C.J. 1974. An Introduction to the ecology of the Illinois Orchida- ceae. Illinois State Museum, Scientific Papers XIV. Springfield, Illinois. 89 p. 25. SMITH, Epwin B. 1978. An atlas and annotated list of the vascular plants of Arkansas. University of Arkansas Bookstore, Fayetteville, Arkansas. 592 p. 26. SNYDER, D. B., & V. E. Vivian. 1981. Rare and endangered vascular plant species in New Jersey. U.S. Fish and Wildlife Service, Newton Corner, Massa- chusetts. 98 p. 27. STEYERMARK, J. A. 1977. Flora of Missouri. The lowa State University Press, Ames, Iowa. 1728 p. 28. Stout, W., K. VER STEEK, & G. F. Lams. 1943. Geology of water in Ohio. Bull. 44, Ohio Division of Geological Survey. 694 p. 29. STOUTAMIRE, W. P. 1974. Terrestrial orchid seedlings. /n: Withner, The Orchids, Scientific Studies. John Wiley and Sons, New York. p. 101-128. 30. STRAUSBAUGH, P.D. & E.L. Core. 1970. Flora of West Virginia. Part |, Ed. 2. West Virginia University Bulletin, Series 70, No. 7-2. 273 p. 31. Tucker, A. O., H. H. Ditt, C. R. Broome, C. E. PHILLIpPs, & M. J. MACIA- RELLO, 1979. Rare and endangered vascular plant species in Delaware. U.S. Fish and Wildlife Service, Newton Corner, Massachusetts. 89 p. 32. U.S. FisH & WILDLIFE SERVICE. 1980. Endangered and threatened wildlife and plants: Review of plant taxa for listing as endangered or threatened species. Federal Register 45: No. 242. 33. WuerryY, E. T., J. M. FoGG, Jr. & H. A. WAHL. 1979. Atlas of the flora of Pennsylvania. Morris Arboretum of the University of Pennsylvania. 390 p. 34. WIEGMAN, P.G. 1979. Rare and endangered vascular plant species in Pennsyl- vania. U.S. Fish and Wildlife service. 94 p. 35. ZENKERT, C. A. 1934. The flora of the Niagara frontier region. Ferns and flowering plants of Buffalo, New York, and vicinity. Bull. Buffalo Soc. Nat. Sci. 16. 313 p. D.M.S. DEPARTMENT OF BOTANY THE OHIO STATE UNIVERSITY 1735 NEIL AVENUE COLUMBUS, OHIO 43210 J.S.S. DEPARTMENT OF BOTANY AND PLANT PATHOLOGY OREGON STATE UNIVERSITY CORVALLIS, OREGON 97331 WIND DISPERSAL OF SOME NORTH AMERICAN SPECIES OF ANDROPOGON (GRAMINEAE) CHRISTOPHER S. CAMPBELL! Abstract: The plumed diaspores of the genus Andropogon are wind-dispersed. The dispersibility of Andropogons helps account for their success as colonizers both in their infestation of appropriate sites and their spread to newly available sites. The wind may also be responsible for the long-distance dispersal of A. bicornis L. from the Caribbean to southern Florida. This is the first record of this widespread New World taxon in the United States. Many species of the cosmopolitan genus Andropogon are remar- kably successful colonizers. Wherever there is full sun and in all but the poorest and driest of soils, they rapidly form large, dense popu- lations. Some species usually grow away from frequent disturbance of the habitat by man and rely upon fire and other natural pheno- mena for an open canopy. Others are of limited range in man-made habitats like roadsides and old fields. The commonest species, such as A. virginicus L., tend to dominate all other vegetation in early stages of secondary succession (Keever, 1950; Golley, 1965; Bazzaz, 1975). The colonizing success of Andropogons comes, at least in part, from their dispersibility. Andropogons flower at the end of the first or second year of growth. The number of flowers per stem varies from as few as 100 in Andropogon gyrans Ashe (long incorrectly known as A. elliottii Chapman) to as many as 10,000 in A. glomeratus (Walter) B.S.P. The number of stems per individual may be as high as 60. Fruit set depends upon the favorability of the habitat for growth, the extent of insect predation and other natural adversities (e.g. infestation by the smut Sorosporium), which vary considerably and can reduce fruit set greatly as in other grasses (Roos & Quinn, 1977). The caryopsis is small (I—4 mm long and less than 0.5 mm in diameter) and light (0.2—0.9 mg). It remains within the spikelet after matura- tion. Dispersal of the fruit begins with the disarticulation of the racemes of spikelets into units (Figure |) consisting of the internode of the rachis of the raceme, one spikelet (bearing one fruit), and a pedicel bearing a second spikelet, which is either vestigial or stami- \Present address: Department of Botany and Plant Pathology, University of Maine, Orono, Maine 04469 65 66 Rhodora [Vol. 85 Figure |. Diaspore of Andropogon glomeratus var. pumilus, many hairs of cal- lus, pedicel and rachis internode omitted. a — awn (only base shown), c — callus hairs, p — pedicel (pedicelled spikelet completely suppressed), r — rachis internode, s — spikelet (showing lower glume with marginal prickle hairs). nate and caducous. The spikelet itself is glabrous except for short (less than 5 mm) hairs on the callus, but the rachis internode and pedicel of many species bear long hairs (1-13 mm, especially toward the apex). These hairs, as well as the rachis internode and pedicel spread horizontally at maturity and, when the atmosphere is dry, forma plume. This unit of dispersal, or diaspore, is then ready to be carried away by the wind. The dispersal distance of plumed dia- spores increases as height of release of the diaspores, wind velocity, and convection (vertical air movement) increase. Dispersal decreases as terminal or fall velocity of the diaspore, density and height of surrounding vegetation, and atmospheric humidity and precipita- tion increase. Diaspores released from a higher point will take longer to fall to the ground and will be subjected to the atmospheric forces which may carry them about for a longer time. Doubling the height of release will double the dispersal distance in a constant wind and convection velocity (Sheldon & Burrows, 1973). The arrangement of the diaspores and hence their height of release is determined by the shape of the inflorescence. In Andropogon virgini- cus the spikelets are more or less evenly spaced over the upper 1983] Campbell — Andropogon dispersal 67 two-thirds of the plant. The spikelets of A. bicornis L. and A. glomeratus in contrast, are usually densely clustered toward the top of the stems. The average height of release of diaspores from the latter species is greater than from the former. The terminal velocity of the diaspores of four species of Andropo- gon and two other species have been computed (Table 1). Taraxa- cum officinale Weber is included as a commonly known reference point. The value obtained here is close to the value (35.7 cm/sec) determined by Sheldon and Burrows (1973). Confidence in the bio- logical importance of these mean values lies in their reproducibility (compare the values for two different plants of Andropogon gyrans obtained on different days) and apparent genetic base (compare the two populations of A. /ongiberbis Hackel). The terminal velocity of an object reflects its coefficient of aerodynamic drag. The lower the terminal velocity the greater is the force opposing the acceleration due to gravity. That the main source of the aerodynamic drag is the plume of Andropogon diaspores is shown by their performance after the hairy rachis internode and pedicel have been removed. The mean times required to fall through two m by the intact diaspores of A. virginicus and A. longiberbis (Gilchrist County population) are 6.1 and 5.4 seconds respectively. After the rachis internode and pedicel have been removed, the fall times are 1.0 and 1.2 seconds respectively. In these species there does not appear to be any strong relationship between terminal velocity and weight of the diaspores. The most important factors are the length and density of hairs on the rachis internode and pedicel. The diaspore of Schizachyrium scoparium (Michx.) Nash (Andro- pogon scoparius Michx.) resembles that of Andropogons in struc- ture. The shortness of the hairs on the rachis internode and pedicel (2-4 mm long) accounts for its relatively high terminal velocity (Table 1). If the only movement of the air is constant and parallel to the ground at, for example, 150 cm/sec (5.4 km/hr), a diaspore of Andropogon glomeratus released from 100 cm above the ground will travel 750 cm away from the plant before reaching the ground (assuming the diaspore moves at the same velocity as the wind). If there is any convection, the rate of fall of the diaspore will be reduced and its potential dispersal distance increased. If the ratio of convection velocity to terminal velocity is greater than or equal to one, there will be no downward movement of the diaspore, and it 68 Rhodora [Vol. 85 Table 1. Terminal velocity (x + S.D.) and weight of plumed diaspores of several species. (N = 30) Taxon Weight Collector, number Terminal velocity (mg) State, County (cm/sec) mean Andropogon virginicus 32.3 + 4.6 0.35 Campbell, 3782 Florida, Jackson A. glomeratus var. pumilus 20.1 + 1.8 0.32 Vasey Campbell, 3739 Florida, Lee A. gyrans plant | 26.9 + 2.7 ae Campbell, 3872 plant II 29.8 + 3.3 0.75 Florida, Jackson A. longiberbis 37.8 + 4.5 0.65 Campbell, 3910 Florida, Gilchrist A. longiberbis 41326 Campbell, 3734 Florida, Dade Schizachyrium scoparium 101¢ + 12 0.750.984 V. L. Cory, 11,348 Texas, Galveston Taraxacum officinale 31 + 6.2 Campbell Massachusetts, Middlesex 4In all cases except Schizachyrium scoparium (see note c) the measurements of terminal velocity where made by dropping the diaspores in still air from about 2 dm above a line two m above the floor. A stop watch was started as the diaspore passed through the two meter height and stopped when it reached the floor. It appeared that all these diaspores reached terminal velocity rapidly, before reaching the two meter line. Sheldon and Burrows (1973) showed that plumed diaspores of the Compositae with terminal velocities of the same range as these reached termi- nal velocity quickly. bMean weight was determined by weighing 20 diaspores together and dividing the total by 20. €Terminal velocity determined by dropping diaspores from two and one m separ- ately and constructing a graph of height of release against time of fall. The slope of the line of the two data points is the terminal velocity (Sheldon & Burrows, 1973). The value given here is near that recorded for the same species (91 cm/sec) by Rabinowitz & Rapp (1981). dFrom Roos and Quinn, 1977. 1983] Campbell — Andropogon dispersal 69 could be dispersed an infinite distance on the slightest breeze (Bur- rows, 1973). In the Compositae, the pappus, which forms the plume, is sensi- tive to humidity (Sheldon & Burrows, 1973). At high humidities the pappus closes and is thereby protected from damage by raindrops and less likely to be released from the plant when raindrops could increase the downward movement of the diaspores. Andropogon diaspores are also hygroscopic. Diaspores of Andropogon gyrans suspended over water in a closed container radically change their configuration. The rachis internode, pedicel, and their many hairs quickly become tightly appressed to the spikelet. Although their terminal velocity was not determined in this condition, it is assumed that it would be about the same as a spikelet without rachis inter- node and pedicel. When these diaspores were placed under a lamp, the rachis internode, pedicel, and hairs spread to their horizontal position in a few seconds. There are two aspects to the dispersal of a colonizing group. First, rapid spread in a local area which is edaphically and otherwise appropriate is the sine qua non of colonization. Observations of Andropogon dispersal in the eastern United States clearly indicate that the density of diaspores falls rapidly away from the plant. Dispersal appears to follow the pattern of Senecio jacobaea (Com- positae) in which 60% of the plumed diaspores fell at the base of the plant, only 0.39% travelled beyond 4.6 m from the edge of the plants, 0.08% beyond 9 m, 0.02% beyond 18 m, and 0.005% greater than 36 m from the plot containing the plants (Harper, 1977). Some habitats, like roadsides, may be more or less continuous, but most sites colonized by Andropogons, like fields, are discontinu- ous. Hence the second aspect of dispersal in connection with colonization is the transport of diaspores away from one locality to another suitable one some distance away. As Harper (1977) points out, “successional species are doomed in their present habitats, and their continued survival depends on escape and establishment else- where.” The general process of colonization, then, is like that of Composites with specialized mechanisms for wind dispersal in which there are “isolated individuals over a great distance which may subsequently act as foci for new infections”. Under normal atmospheric conditions dispersal of diaspores away from the population to a distant and suitable new site may be a rather unlikely event. Long-distance dispersal becomes a greater 70 Rhodora [Vol. 85 likelihood in extreme weather conditions such as hurricanes, tropi- cal storms, and thunderstorms. From 1885 to 1971 inclusive, 160 tropical storms entered or significantly affected Florida (Anon, 1974). A little over half of these were of hurricane intensity (winds greater than 74 mph). The frequency of occurrence of hurricane force winds in any given year varies considerably: one in six around Miami, one in twenty-five in Tampa, one in eight in Tallahassee, and one in one hundred in Jacksonville, the least affected region in the state (Anon, 1974). Florida is the most strongly hit state in the southern United States. The usual hurricane tracks after passing through Florida — either to the northeast along the Atlantic Coas- tal Plain, or to the west into the Gulf of Mexico (Gentry, 1974) pass over the center of greatest diversity of species of Andropo- gon in the United States and the area of their greatest prominence in the overall flora. The tropical storm season of eastern North America overlaps with the period of diaspore dissemination in Andropogon. The average frequency of tropical cyclones (winds 39-73 mph) and hur- ricane peaks from August to October (Dunn & Miller, 1960). There is an average of 1.5, 2.6, and 1.9 tropical storms in each of these months respectively and 0.9, 1.6, and 0.8 hurricanes per month in the North Atlantic region in these three months respectively. Andro- pogon diaspores mature in late August, September, and especially October and November. As pointed out earlier, the dispersal capac- ity of the diaspores falls markedly when humidity is high. So the diaspores must be picked up by winds which are relatively dry (or very powerful). Once the diaspore is inside a storm itself, it is likely to be carried for a great distance. Thunderstorms, though less powerful and more local, occur fre- quently enough to be potentially important modes of dispersal for Andropogon diaspores. There is no direct evidence of the extent to which these diaspores are transported by storms. However, there is one possible example of long-distance dispersal. Andropogon bicornis is a common spe- cies of South and Central America and the West Indies. It has never been reported for the United States, and I have never seen it in my field work in South Florida. There is one specimen of this plant collected by F. C. Craighead in northern Monroe County in October of 1962 (deposited in the herbarium of the Fairchild Tropi- cal Garden in Miami, Florida). It was identified as A. floridanus 1983] Campbell — Andropogon dispersal 71 Scribner, but the lack of awns and the conspicuous, staminate pedi- celled spikelets leave no question of its being A. bicornis. It is possi- ble that it was transported from the West Indies during Hurricane Donna in 1960. This was the most powerful hurricane in Florida’s history, and its track passed within 30 miles of the site where Craighead found the plant. The nearest known localities of A. bicornis are in Cuba, over 300 km from Craighead’s collecting site. Long-distance wind dispersal is also suspected in another genus of the Andropogoneae, Bothriochioa Kuntze, in which 15 of the 18 New World Species show major disjunctions between North and South America (Allred, 1981). After the diaspore becomes estab- lished following long-distance dispersal it must reproduce itself if the new site is to be colonized. The capacity for apomictic seed production has not been reported for Andropogon sensu stricto (Connor, 1979; Campbell, 1983). Most North American Andropo- gons are cespitose, so colonization depends upon sexual seed pro- duction. If these plants were self-incompatible, a second genotype would have to be present within the range of pollen dispersal for fertilization and seed production to be effected. The probability of two plants of differing genotypes arriving at the same place within one life span is much less than that of the establishment of a single individual. The fact that most North American Andropogons (and certainly the most vigorous colonizing species) are self-compatible and frequently cleistogamous (Campbell, 1982) increases the likeli- hood of the effectiveness of long-distance dispersal (Baker, 1953). ACKNOWLEDGMENTS R. C. Rollins, R. M. Tryon, and C. E. Wood, Jr. read an earlier draft. This work was supported by grants from the Anderson and Fernald Funds of Harvard University, the National Science Founda- tion (DEB 77—17317), and the Rutgers University Research Council (07-2200). LITERATURE CITED ALLRED, K. 1981. Cousins to the South: Amphitropical disjunctions in Bothri- ochloa (Graminae). Abstract from Bot. Soc. Amer. meetings, Bloomington, Indiana. ANON. 1979. Climates of the States. 2 vols. Officials of the National Oceanic and Atmospheric Administration, U.S. Dept. of Commerce. Water Information Center, Inc. New York. 72 Rhodora [Vol. 85 Baker, H.G. 1953. Race formation and reproductive effort in flowering plants. Symp. Soc. Exptl. Biol. 7: 114-143, Bazzaz, F. A. 1975. Plant species diversity in old-field successional ecosystems in Southern Illinois. Ecology 56: 485-488. Burrows, F.M. 1973. Calculation of the primary trajectories of plumed seeds in steady winds with variable convection. New Phytol. 72: 647-664. CAMPBELL, C.S. 1982. Cleistogamy in Andropogon (Gramineae). Amer. J. Bot. 69:1649-1658. 1983. Systematics of the Andropogon virginicus complex (Gramineae), J. Arn. Arb. (In press). Connor, H. E. 1979. Breeding systems in grasses: a survey. New Zealand J. Bot. 17: 547-574. Dunn, G.E.,& B. I. Miter. 1960. Atlantic hurricanes. Louisiana State Univer- sity Press. 326 pp. Gentry, R. C. 1974. Hurricanes in South Florida. p. 73-81 /n: Gleason, P. J. (ed.). Environments of South Florida: past and present. Memoir 2, Miami Geological Society. 452 pp. Goutey, F. B. 1965. Structure and function of an old-field broomsedge commu- nity. Ecol. Monogr. 35: 113-137. Harper, J. L. 1977. Population biology of plants. Academic Press. London, 892 pp. Keever, C. 1950. Causes of succession on old fields of the Piedmont, North Carolina. Ecol. Monogr. 20: 229-250. Rapinowitz, D., & J. K. Rapp. 1981. Dispersal abilities of seven sparse and common grasses from a Missouri prairie. Amer. J. Bot. 68: 616-624. Roos, F. J., & J. A. QuINN. 1977. Phenology and reproductive allocation in Andropogon scoparius (Gramineae) populations in communities of different successional stages. Amer. J. Bot. 64: 535-540. SHELDON, J. C., & F. M. Burrows. 1973. The dispersal effectiveness of the achene-pappus units of selected Compositae in steady winds with convection. New Phytol. 72: 665-675. BOTANY DEPARTMENT RUTGERS UNIVERSITY NEWARK, N.J. 07102 ON THE TAXONOMIC STATUS OF LOPHIOLA AUREA KER-GAWLER MICHAEL ZAVADA,! XUE-LIN XU, AND J. M. EDWARDS Familial and tribal treatments of the Haemodoraceae have been inconsistent, and there has been disagreement among various authorities over the inclusion of Lophiola aurea in the family (Geer- inck, 1969). Lophiola aurea = (L. americana (Pursh) Wood; see Robertson, 1976), which is the only species of Lophiola, grows in acid, pine- barren bogs from New Jersey to Florida, with a disjunct Nova Scotian population. The species has been variously treated, and placed in the tribe Conostyleae of the Amaryllidaceae (Pax, 1930), in the tribe Haemodoreae (Geerinck, 1969), and the Conostyleae of the Haemodoraceae (Hutchinson, 1973). Ornduff (1979), basing his conclusion on gross morphology and chromosome numbers, sug- gested that 1. aurea is more closely allied with the tribe Haemodo- reae and possibly with the genus Lachnanthes, the only other North American member of the Haemodoreae. However, Robertson (1976) investigated vegetative morphology, Simpson and Dickison (1981) investigated anatomy, and Simpson (1981) embryology, of Lach- nanthes and Lophiola, and they found few similarities between these taxa. Our phytochemical and palynological studies of various taxa of the Haemodoraceae further suggest that Lophiola aurea is not closely related to other genera of Hutchinson’s Haemodoreaceae. MATERIALS AND METHODS Pollen from living and dried herbarium material was used in this study. Live pollen was prepared for transmission electron micro- scopy (TEM) by fixation in cacodylate-buffered gluteraldehyde- formaldehyde followed by fixation with osmium tetroxide, dehydra- tion in an ethanol series, and embedding in Dow Epoxy Resin-334 (DER-334). Pollen removed from herbarium specimens was aceto- lyzed and prepared for TEM by dehydration in an ethanol series, and embedded in DER-334. Sectioning was done on an LKB-! ultramicro- tome; the sections were post-stained in uranyl acetate-lead citrate, ‘Present address: Department of Botany, Indiana University, Bloomington, Indiana 47401 73 74 Rhodora [Vol. 85 and viewed on a Philips EM-300. Pollen was prepared for scanning electron microscopy (SEM) by mounting the pollen on stubs with the high vacuum wax Apiezon W-100, coated with gold-palladium, and viewed on a Coates and Welter Field Emission Electron Microscope. Dried, defatted, above ground parts of Lophiola aurea (Ed- wards, s.n., New Jersey, CONN) were extracted with 95% ethanol. Thin-layer chromatography of the extract (Silica gel, ethyl acetate- benzene, 1:2) indicated the presence of two phenolic compounds (positive color reaction with diazotized p-nitroaniline) having Rg s 0.3 and 0.1. Column chromatography of the mixture over silica gel with ethyl acetate and increasing amounts of ethanol (95%) resulted in the isolation of the less polar compound as a pale yellow solid of mp 219—223° (ethyl acetate-petrol ether). RESULTS Palynology Hutchinson (1973) includes ten genera in the tribe Haemodoreae. Three genera and three species were investigated palynologically; Lachnanthes caroliana (Lam.) Dandy, (Weatherley s.n., Connecti- cut, CONN); Wachendorfia paniculata Burm., (Zavada 501, Cult., CONN); and Xiphidium caeruleum Aubl., (Wolfe 234, Surinam, CONN). Pollen of this tribe is monosulcate (Erdtman, 1952; Radu- lescu, 1973; present study). Pollen of Lachnanthes caroliana aver- ages 36-44 um along its long axis and sculpturing is scabrate to verrucate (fig. 1). Wall structure is atectate and infrequently tra- versed by minute channels (fig. 2). No endexine is evident. Pollen of Wachendorfia paniculata averages 55—58 um along its long axis and sculpturing on the proximal and distal faces of the pollen grain is scabrate to pustulate (fig. 5). Separating the proximal and distal faces of the pollen grain is a psilate ridge (figs. 4, 5). Wall structure on the proximal and distal faces is granular with large and small spherical to irregularly shaped granules closely appressed to one another (fig. 3; cf. “structure grenue” of Van Campo and Lugardon, 1973). The wall structure of the psilate ridge differs in having a thick tectum with a granular layer beneath (fig. 4). No foot layer or endex- ine is evident. Pollen of Xiphidium caeruleum averages 40-55 um along its long axis and pollen grains are psilate (figs. 6, 7). Wall structure is tectate with the tectum fused to irregularly shaped rod- like structures which are often interspersed with granules (fig. 7). 1983] Zavada, et al. — Lophiola aurea ID The tectum is occasionally traversed by minute channels (fig. 7). No foot layer or endexine Is evident. Hutchinson (1973) includes six genera in the tribe Conostyleae. Three genera and three species were investigated palynologically; Anigozanthos flavus D.C. ex Red., (Edwards s.n., “1977”, CONN); Conostylis setosa Lindl., (Wolfe s.n., Australia, CONN); and Lophi- ola aurea Ker-Gawler, (Edwards s.n., New Jersey, CONN). Pollen of this tribe is 2—8-porate, except Lophiola aurea which is monosulcate (Erdtman, 1952; Radulescu, 1973; present study). Pollen of Anigo- zanthos flavus averages 36—44 um along its long axis and is diporate with the pores located opposite one another along its long axis (fig. 8). Sculpturing is roughly scabrate and wall structure is atectate (figs. 8, 9). No endexine is present. Pollen of Conostylis setosa is triporate averaging 22—29 um along its long axis and pollen wall sculpturing 1s roughly scabrate (fig. 10). Wall structure is atectate and no endexine is present (fig. 11). Pollen of Lophiola aurea is monosulcate averaging 29 um along its long axis. Pollen grains are finely reticulate (fig. 13). Wall structure is tectate-columellate with a thin foot layer and no endexine (fig. 12). Phytochemistry The yellow-colored compound isolated from silica gel chromatog- raphy had Amax (EtOH) 227 (Loge 4.49), 293 (4.51), and 330 sh (3.90) nm. ymax 3270, 1650, and 1520 cm.' 6 (persilyl ether in CDC1; 60MHz) 7.22 (2H, d, 9Hz), 7.12 (2H, d, 9Hz), 6.76 (2H, d, 9Hz), 6.61 (2H, d, 9Hz), 5.94 (2H, dd, 2.5Hz), 5.76 (1H, s), 5.66 (1H, ad, 262),:3.22 (10,0, 12 and SHz)..4.36 (1H; d)I2hiz), 3.00-2.78 (2H, m). M* (%) 542 (.02); 416.885, Cy4H1607 (65); 296.0317, Ci6HgO¢ (52); 270.0520, CysH Os (3); 126 (40) and 94 (100). The MS of the persilyl ether had M* 974 indicating the presence of six hydroxyl groups in the molecule. Comparison of these spectral data with those published for the biflavanone GB-—la (Jackson et al., 1967 and Bandaranayake et al. 1975) indicated that the compound was 5, 7, 4’, 5’, 4’”’ -hexahydroxy [3,8’’] biflavanone. DISCUSSION The pollen data underscore the unique status of Lophiola aurea in the Haemodoraceae. Lophiola aurea is the only species investigated exhibiting the tectate-columellate wall structure and reticulate exine 76 Rhodora [Vol. 85 1983] Zavada, et al. — Lophiola aurea 77 sculpturing. The five other genera investigated have atectate or tectate-granular wall structure. The monotypic species Lanaria plumosa Ait., placed in the Haemodoraceae by Hutchinson (1973), is the only other genus reported to have reticulate exine sculpturing (Erdtman, 1952). The reticulate exine pattern is usually accom- panied by the tectate-columellate wall structure. Thus, Lophiola and Lanaria are the only genera of the Haemodoraceae, as assigned by Hutchinson, exhibiting palynological features not found among the other genera of the family (Erdtman, 1952; present study). Their palynological features are more similar to those of some members of the closely related family Tecophilaeaceae (sensu Hutchinson). However, data on vegetative morphology (Robertson, 1976), anat- omy (Simpson and Dickison, 1981), embryology (Simpson, 1981), and palynology and chemistry (present study) more firmly establish the differences between 1. aurea and other members of the Haemo- doraceae than the taxonomic affinities of L. aurea. This is due to the paucity of detailed morphological, anatomical, palynological and chemical data on the Tecophilaeaceae and other closely related fam- ilies to the Haemodoraceae. Placement of L. aurea in any of these families would make that designation as tenuous as its present status. Excluding Lophiola aurea (and Lanaria plumosa) from the Hae- modoraceae highlights the stenopalynous nature of the remaining genera in each tribe. This lends support to a treatment of the Haemo- doraceae similar to Hutchinson’s, save the inclusion of Lophiola and Lanaria. The tribe Haemodoreae then only includes taxa with monosulcate pollen, a more primitive situation than the 2—8-porate Figures 1-7, Pollen of the Haemodoreae. Figure 1. Lachnanthes caroliana, aceto- lyzed, SEM X 1,530. Figure 2. L. caroliana, unacetolyzed, TEM showing atectate wall structure, minute channels (arrows), and thin intine, ¥ 17,000. Figure 3. Wachen- dorfia paniculata, unacetolyzed, TEM showing spherical to irregularly shaped granules comprising the wall of the distal face of the pollen grain, note thin intine, 14,200. Figure 4. W. paniculata, unacetolyzed, TEM showing wall structure in the region of the psilate ridge, note thick tectum and granular layer beneath resting on a thin intine, *14,200. Figure 5. W. paniculata, acetolyzed, SEM showing scabrate and pustulate exine sculpturing on the proximal (P) and distal (D) faces of the pollen grain, note psilate ridge separating these regions, X 2,850. Figure 6. Yiphidium caeruleum, acetolyzed, SEM showing sulcus, * 2,850. Figure 7. ¥. caeruleum, aceto- lyzed, TEM showing granular infrastructure, also with irregularly shaped rods, note the minute channels (arrows) traversing the relatively thick tectum, X25,000. 78 Rhodora [Vol. 85 1983] Zavada, et al. — Lophiola aurea ie pollen characteristic of the tribe Conostyleae. The evolutionarily advanced status of the entire family is supported by the occurrence of the atectate and granular wall structure: wall structural types occurring in the more advanced taxa of other monocot orders (Walker & Doyle, 1975; Zavada, manuscript in preparation). Investigations of the yellow, orange, brown, and purple pigments present in the colorful root systems, flowers, and seed capsules of eight genera of the Haemodoraceae (Anigozanthos, Conostylis, Haemodorum, Lachnanthes, Macropidia, Phlebocarya, Wachen- dorfia, and Xiphidium) have shown them to be derivatives of either 9-phenalenone or naphthoxanthenone. Furthermore, secondary metabo- lites containing the phenalenone nucleus, or having structures which can reasonably be presumed to have been derived from an intact phenalenone, are rare in nature; their occurrence seems to be res- tricted to four genera of hyphomycetes (Fungi Imperfecti), one genus within the class Discomycetes, and one family of higher plants. These chemical compounds seem to be chemotaxonomic markers for the Haemodoraceae (Cooke & Edwards, 1981). Lophiola aurea lacks obvious pigmentation except for its pale yellow flowers, and previous investigation of the plant (Edwards, et al., 1970) has shown it to be devoid of phenalenone and related pigments. The present study, while confirming the absence of phe- nalenones, has identified the biflavanoid 5, 7, 4’, 5’, 7’, 4’’’-hexa- hydroxy [3,8’’] biflavanone (GB la) in extracts of L. aurea, a compound which has been isolated previously from species of Gar- cinia (Guttiferae) (Jackson, et al., 1967; Bandaranayake, et al., 1975). Biflavonoid compounds have been isolated from some 20 plant families, none of which is monocotyledonous (Geiger & Quinn, 1975); furthermore, no flavonoids have been isolated from the Haemodoraceae. There is an account of the chromatographic Figures 8-13, Pollen of the Conostyleae. Figure 8. Anigozanthos flavus, aceto- lyzed, SEM showing the two opposing pores, X1,760. Figure 9. A. flavus, unaceto- lyzed, TEM showing atectate wall structure and thick intine, X 17,100. Figure 10. Conostylis setosa, acetolyzed, SEM showing triporate condition, X 2,370. Figure 11. C. setosa, unacetolyzed, TEM showing atectate wall structure and thin intine, X 20,100. Figure 12. Lophiola aurea, unacetolyzed, TEM showing tectate-columellate wall structure, X 20,100. Figure 13. L. aurea, acetolyzed, SEM showing reticulate exine sculpturing on the distal face of the pollen grain, a sculpturing type found in the monotypic genus Lanaria plumosa of the Haemodoraceae, X 4,220. 80 Rhodora [Vol. 85 identification of flavonoids in a Haemodorum sp. (Bate-Smith, 1968; Gornall, et al., 1979), but no flavonoids have been isolated; the phenalenone pigments are so abundant in Haemodorum that the identification of flavonoids by chromatographic methods alone must be regarded as tentative at best. We feel that the chemical findings reported here: absence of phenalenone and naphthoxanthenone pigments and the presence of a biflavonoid, in conjunction with the palynological data, provide good chemotaxonomic and palynological evidence against the inclusion of Lophiola aurea in the Haemodoraceae. LITERATURE CITED BANDARANAYAKE, W. M., S. S. SertrAn, M. U. S. SULTANBAWA, & W. D. OLLIS. 1975. Biflavonoids and xanthones of Garcinia terpnophylla & G. echino- carpa. Phytochemistry 14: 1878-1880. Bate-SMITH, E.C. 1968. The phenolic constituents of plants and their taxonomic significance. J. Linn. Soc. (Bot.) 60: 325-356. Cooke, R. G., & J. M. Epwarps. 1981. Naturally occurring phenalenones and related compounds. Fortschr. Chem. Org. Naturstoffe 40: 158-190. Epwarps, J. M., J. A. CHURCHILL, & U. Weiss. 1970. A chemical contribution to the taxonomic status of Lophiola americana. Phytochemistry 9: 1563-1564. ERDTMAN, G. 1952. Pollen morphology and plant taxonomy. Angiosperms. Chronica Botanica Co., Waltham, Mass. GEERINCK, D. 1968. Considerations taxonomique au sujet des Haemodoraceae et des Hypoxidaceae. (Monocotyledones). Bull. Soc. Roy. Bot. Belgique 101: 265—278. GeiGerR, H., & C. QUINN. 1975. Biflavonoids. Jn: “The Flavonoids,” J. B. Har- borne, T. J. Mabry, and H. Mabry, Eds., Academic Press, New York. pp. 692-742. GornaLl, R. J., B. A. Boum, & R. DAHLGREN. 1979. The distribution of flavo- noids in the angiosperms. Bot. Not. 132: 1-30. HuTCHINSON, J. 1973. The families of flowering plants. 2nd Edition, Vol. II, Monocotyledons. Oxford Univer. Press, Oxford. Jackson, B., H. D. LocksiLey, F. SCHEINMANN, & W. A. WOLSTENHOLME. 1967. The isolation of a new series of biflavonones from the heartwood of Garcinia buchananii Baker. Tetrahedron Lett. 787-792. OrnpburF, R. 1979. Chromosome numbers and relationships of certain African and American genera of Haemodoraceae. Ann. Missouri Bot. Garden, 66: 577-580. Pax, F. 1930. Haemodoraceae. Nat. Pflanzenfam. ed. B. 15a: 386-391. Rapucescu, D. 1973. La morphologie du pollen chez quelques Haemodoraceae. Lucrarile Gradinii Botanice Din Bucuresti, 1972-73, pp. 123-132. ROBERTSON, K. R. 1976. The genera of Haemodoraceae in the southeastern United States. J. Arnold Arbor. 57: 205-216. 1983] Zavada, et al. — Lophiola aurea 81 Simpson, M. G. 1981. Embryological development of Lachnanthes caroliniana (Lam.) Dandy and Lophiola aurea Ker-Gawler (Haemodoraceae) and its taxo- nomic significance. Bot. Soc. America, Misc. Series, Pub. 160, p. 78 (Abstract). Simpson, M.G., & W.C. Dickison. 1981. Comparative anatomy of Lachnanthes and Lophiola (Haemodoraceae). Flora 171: 95-113. VAN Campo, M., & B. LUGARDON. 1973. Structure grenue infratectale de l’ec- texine des pollens de quelques gymnospermes et angiospermes. Pollen et Spores 15: 171-197. WALKER, J. W., & J. A. Doyte. 1975. The basis of angiosperm phylogeny: Paly- nology. Ann. Missouri Bot. Garden 62: 664-723. DEPARTMENTS OF BIOLOGICAL SCIENCES AND PHARMACY UNIVERSITY OF CONNECTICUT STORRS, CONNECTICUT 06268 THE FERN GENERA VITTARIA AND TRICHOMANES IN THE NORTHEASTERN UNITED STATES! DONALD R. FARRAR, JAMES C. PARKS, AND BRUCE W. MCALPIN Abstract. Vitraria is reported from 5 counties in Pennsylvania; Trichomanes from 24 counties in Connecticut, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, and Vermont. All are new state records except Tricho- manes in Massachusetts. These ferns occur as gametophytes only, but in large con- spicuous populations in crevices and grottos of moist, non-calcareous rock outcrops in canyon topography. The northernmost distribution of Trichomanes has not been established, but may parallel that of the oak-hickory forest type. Vittaria was not found north of the southern terminus of Pleistocene glaciation; its northern distribu- tion may have been truncated by that event. Vitraria sites in northwestern Pennsyl- vania are well north of apparently suitable glaciated habitats in eastern Pennsylvania wnich house Trichomanes but not Vittaria. Field botanists of the southern Appalachian Mountain and Pla- teau regions of the Eastern U.S. are familiar with the occurrence of conspicuous populations of fern gametophytes totally devoid of the sporophyte phase of the fern life cycle. Best known of these is the Appalachian Vittaria gametophyte which produces dense, vivid green colonies often dominating the flora of heavily shaded, moist outcroppings of non-calcareous rocks (Farrar, 1978). Less familiar but equally common are the filamentous gametophytes of the genus Trichomanes which form velvety cushions and tufts in the same habitats as Vittaria (Farrar, 1967; Wagner & Evers, 1963). Gameto- phytes of both genera have the ability to reproduce themselves vegetatively by means of gemmae and thus exist independently of and beyond the range of their sporophyte counterparts. Sporo- phytes of Vitraria are totally absent from Eastern U.S. uplands, and gametophytes of Trichomanes are known far north of the range of sporophytes of 7Trichomanes petersii and Trichomanes boschianum (with northernmost occurrences in Sevier Co., Tennessee and Hock- ing Co., Ohio respectively). It was, in fact, the recent discovery of Trichomanes gametophytes in Massachusetts (McAIpin & Farrar, 1978) that prompted this study. 'This study was supported by a grant from the National Geographic Society and made possible through a Faculty Improvement Leave from Iowa State University to the senior author. 83 84 Rhodora [Vol. 85 In considering the origin of the Appalachian Vitraria gameto- phytes, Farrar (1978) proposed that they were possibly relicts of a Tertiary subtropical flora. In support of this hypothesis he cited the basic tropical affinity of the genus, the morphological and physio- logical differentiation of the Appalachian plants from all other known species of Vittaria, and the restriction of their occurrence to rock exposures predating and lying south of the borders of Pleis- tocene glaciation. It was also suggested that Trichomanes shared many of these characteristics and may have experienced a similar history. The distribution of Virtaria and Trichomanes in the Northeastern U.S. has not been carefully studied previously. Since appropriate habitats here are abundant and continuous through the southern terminus of Pleistocene glaciation, this is an important area in which to investigate the factors limiting the distribution of these unusual plants and providing clues to their origin. They also con- stitute significant new additions to the floras of the northeastern states in which they occur. METHODS Field search for gametophytes of Vittaria and Trichomanes was conducted in portions of all states north of the Virginias, except Delaware and Rhode Island. Sites for investigation were identified through consultation with local field botanists and naturalists, anal- ysis of physiographic maps, and checking of herbarium records of plants commonly associated with the gametophytes in more south- ern locations. Available time permitted investigation of only a frac- tion of the sites so identified. Herbarium vouchers were collected and prepared for deposit in the herbaria of lowa State University, The New York Botanical Garden, and Millersville State College of Pennsylvania. RESULTS AND DISCUSSION Vittaria gametophytes were collected from 6 sites in 5 counties of Pennsylvania, and Trichomanes gametophytes were collected from 32 sites in 24 counties of Pennsylvania, Maryland, New Jersey, New York, Connecticut, Massachusetts, Vermont, and New Hampshire. These constitute new state records for all states listed except Massa- chusetts. A listing of collection sites and habitat substrates is given in Appendix 1. 1983] Farrar, et al. — Fern gametophytes 85 Habitats of Vitraria and Trichomanes gametophytes in the North- east are very similar to those in the Southeast: heavily shaded, moist outcroppings of non-calcareous rock. A significant difference is that occurrence of the gametophytes in the north seems to require devel- opment of these habitats on a larger scale. Gorges must be deeper and narrower, cliffs and waterfalls must be higher, and sites in general must offer greater buffering from climatic extremes than is required to support these plants in the south. This extreme of habitat and microclimate specificity is well illus- trated by the habitat of Vittaria at its northernmost station at Jake’s Rocks in Warren County, Pennsylvania (Figure 1). This cave-like habitat was created by the cleavage and slumping of a huge sand- stone block from an undercut cliff face. The massive, porous sand- stone provides continuous moisture while moderating temperature through radiation and blockage of wind currents. Sufficient light (about 100 ft. c.) penetrates the “cave” opening to allow growth of Vittaria. Trichomanes may also require wetter habitats in its northernmost localities. In the Catskill Mountains of New York and Green Moun- tains of Vermont it was collected only from crevices of rocks which were saturated with moisture, even during the relatively dry period Pottsville Sandstone Figure |. Diagrammatic cross section of the habitat of Appalachian | itraria gametophytes at Jake’s Rocks on the Allegheny Reservoir in Warren Co., Pennsyl- vania. Arrow indicates the microhabitat of Virraria. 86 Rhodora [Vol. 85 when collections were made. By contrast, in Pennsylvania and more southern areas, Trichomanes gametophytes often occur in habitats that are periodically quite dry. The currently known distribution of Vittaria and Trichomanes in the Northeastern U.S. is presented in Figure 2. The distribution of Trichomanes clearly is not governed by the limits of glaciation. The northern limit of Trichomanes was not determined in this study as no sites north of the collections in New York and Vermont were investigated. A search for Trichomanes in habitats with suitable physical characteristics in the White Mountains of New Hampshire (Franconia Notch area) and Maine (Grafton Notch area) was unproductive. However, forest vegetation of these sites is decidedly more northern than the Central Hardwoods—Oak-Hickory forest (Westveld, 1956; Little, 1971) typical of Trichomanes sites. On the Figure 2, The known distribution of Vitraria (stars) and Trichomanes (dots) in the northeastern United States. Each symbol represents a county in which the plants were collected. Hatched lines represent the approximate southern limits of Wisconsin (upper line) and Illinoian (lower line) glaciation. 1983] Farrar, et al. — Fern gametophytes 87 basis of its association with this forest type, a northern limit for Trichomanes may be predicted to occur in the mountains of south- ernmost Quebec. Vittaria gametophytes were not found in any habitats covered by Wisconsin glaciation. They were found in habitats just beyond the southern terminus of Wisconsin glaciation, and in one site, approx- imately one kilometer within the boundary of Illinoian glaciation (Gray, 1960). Some Vittaria sites are as far north as glaciated sites harboring Trichomanes but not Vittaria. Since the two genera have nearly identical habitat requirements, frequently sharing the same habitat, it is reasonable to conclude that Pleistocene glaciation has influenced the current distribution of Vittaria. In considering the origin of gametophytes of Vittaria and Trich- omanes, it is important to analyze the nature of their habitats and probable means of dispersal. Their habitats are invariably situated within canyon-like topography and surrounded by a dense mature forest canopy. The microhabitats in which they grow are crevices or grottos deeply recessed under cavernous rock outcrops. Wind cur- rents in these microhabitats are minimal so that wind dispersal of propagules into or out of these habitats seems unlikely. More plau- sible dispersal agents are the variety of insects and other inverte- brates which actively seek out these same habitats. Another aspect of their distribution suggests that they are not presently being dispersed beyond their immediate habitats. They are not found in man-made habitats such as roadcuts, tunnels, bridges or quarries, although these often present all conceivable combina- tions of light and moisture with appropriate substrate, and are in proximity to existing populations. Many such habitats have been in existence 100 years or more. If the gametophytes are not currently dispersing and thus not invading adjacent available habitats, when did they reach their cur- rent northeastern distribution? In the case of Trichomanes, this was certainly post-Pleistocene, perhaps during a hypsithermal period (Cushing, 1965; Davis, 1965) when a warmer climate may have facilitated northward migration. At this time, sporophytes of Trich- omanes may also have extended their distribution northward into this area and contributed to the spread of the gametophytes. A pre-Pleistocene distribution of Vitraria appears to have been truncated by Pleistocene glaciation. This conclusion is evidenced by the coincident northern limits of the plants and southern limit of 88 Rhodora [Vol. 85 glaciation from southern Indiana! to eastern Pennsylvania. This coincidence holds despite the fact that habitats near the more north- erly areas of the glacial terminus appear much less favorable to warm temperate plants than do some more southerly habitats lying north of glaciation. For example, the high, exposed rim rocks of the Allegheny reservoir in unglaciated northwestern Pennsylvania sup- port populations of Vittaria whereas the more protected narrow canyons and waterfalls of Rickets Glen and the Delaware Water Gap of glaciated cental and eastern Pennsylvania contain abundant Trichomanes but not Vittaria. Could the absence of Vittaria within the glaciated area be attrib- uted to glacial alteration of habitats rendering them permanently unsuitable for these plants? In the Appalachian Mountains there is little probability of this surmise as the mountain topography was not greatly altered by glaciation, and glacial drift has been largely removed from upland areas through subsequent erosion (Muller, 1965; Schafer and Hartshorn, 1965; Thornbury, 1965). Common acidiphilic fern associates of Vittaria, e.g. Asplenium montanum, and Trichomanes gametophytes, and mosses such as Hookeria acu- tifolia also show no limitations to unglaciated areas (Crum & And- erson, 1981; Wherry, 1942). In glaciated areas west of the Appalachian Mountains, calcareous tills generally overlie sandstone outcrops of the type suitable for Vittaria in unglaciated areas (Goldthwait, et al., 1965; Wayne & Zumberge, 1965; Thornbury, 1965) and such outcrops are less fre- quent and less continuous than they are beyond the glacial terminus. These factors are significant, but do not alone explain the absence of Vittaria in this area. Trichomanes occurs in Montgomery County, Indiana, suggesting suitability of habitat there, approximately 25 km within Wisconsin glaciation. Vittaria occurs in Martin Co., approximately 100 km from the Montgomery Co. site and just beyond the Illinoian glacial boundary. In Lawrence Co., Pennsylvania, Vittaria gametophytes occur in great abundance on Pottsville sandstone approximately | km within I]linoian glaciation and about | km beyond the Wisconsin glacial limit. Some post-Illinoian migration of the gametophytes has 'This study has yielded new county records for Vittaria gametophytes in Martin and Perry Counties, Indiana and Fairfield County, Ohio. It was previously known from Crawford County, Indiana and Hocking and Jackson Counties, Ohio. 1983] Farrar, et al. — Fern gametophytes 89 obviously occurred, perhaps made possible by the continuity of suitable habitats along Slippery Rock Creek from this area to well south of glacial limits. Similarly, in Hocking and Fairfield counties of Ohio, Vittaria is abundant on the common and continuous outcrops of Pottsville sandstone to within | km of the glacial terminus. However, it is absent from the relatively isolated outcrops of the same substrate in Licking Co., Ohio, 60 km to the northeast, although habitats here support populations of Trichomanes gametophytes, Asplenium pinnatifidum, and the acidophilous mosses Hookeria acutifolia and Bryoxiphium norvegicum, all frequent associates of Vitraria. These observations suggest that a large geographical extent and linear continuity of suitable habitats are important determinants of the occurrence of Vittaria, perhaps operating through probability of habitat continuity over time as well as provision of migration routes requiring minimal transport distances. It is difficult to explain this distribution of the Appalachian Vitta- ria gametophytes except by assuming a very limited dispersal capac- ity and consequently a Pleistocene occurrence very near the glacial terminus. The latter is not unlikely considering the bryophyte-like growth form of Vittaria, its ability to tolerate freezing (Farrar, 1978), and the strong climatic moderation afforded by the cave-like habitats of these plants (Fig. 1). Similar rock shelters in Alabama, Illinois, and Missouri contain remains of human habitation which have been dated at more than 9000 years b.p. (Jennings, 1968). Clearly these habitats are stable and very likely persisted with little modification throughout the Pleistocene. In such habitats a fre- quent associate of Vitraria in the northern, unglaciated portions of the Appalachian Plateau is the sword moss, Bryoxiphium norvegi- cum. Steere (1937) described this moss as exhibiting a relict distribu- tion pattern of a species which thrived near the periphery of pleistocene glaciers. In summary, it seems likely that both Trichomanes and Vittaria reached pre-Pleistocene and possibly interglacial northern limits more or less coincident with the distribution of the Central Hard- wood forests of eastern North America. Trichomanes has regained this distribution during post-Pleistocene time, possibly during an interval warmer than the present. Vittaria, by contrast, appears to have migrated very little beyond the areas to which it was restricted by Pleistocene glaciation. The greater post-Pleistocene migratory 90 Rhodora (Vol. 85 success of Trichomanes may be related to the more northerly occur- rence of sporophytes of this genus. Many prospective northeastern sites for these ferns remain un- investigated. Trichomanes may be expected wherever sizeable out- crops of non-calcareous rocks occur throughout the range indicated in Figure 2, and further significant range extensions are likely to be documented in the northeastern U.S. and possibly in southeastern Canada. Vittaria likely occurs throughout unglaciated Pennsylvania and possibly in unglaciated portions of New York and New Jersey. Our conclusions regarding the history and mobility of Vittaria do not exclude the possibility of its chance occurrence within the limits of Pleistocene glaciation.2 However, present information strongly suggests that if found in such areas, Vitraria there will not approach the abundance or frequency of its occurrence in unglaciated areas. LITERATURE CITED Crum, H.A.& L.E. ANDERSON. 1981. Mosses of Eastern North America. Vol. 2. Columbia Univ. Press, New York. CusHING, E. J. 1965. Problems in the Quaternary phytogeography of the Great Lakes region. Jn: Wright, H. E., Jr. and D. G. Frey, Editors. 1965. The Quater- nary of the United States. Princeton University Press, Princeton, N.J. Davis, M. B. 1965. Phytogeography and palynology of Northeastern United States. /n: Wright, H. E., Jr. and D. G. Frey, Editors. 1965. The Quaternary of the United States. Princeton University Press, Princeton, N.J. Farrar, D. R. 1967. Gametophytes of four tropical fern genera reproducing independently of their sporophytes in the southern Appalachians. Science 155: 1266-1267. 1978. Problems in the identity and origin of the Appalachian Vittaria gametophyte, a sporophyteless fern of the Eastern United States. Amer. J. Bot. 65: 1-12. GotpTHwaiT, R. P., A. DRreIMANis, J. L. Forsytu, P. F. KArrow, & G. W. Wuite. 1965. Pleistocene deposits of the Erie Lobe. Jn: Wright, H. E., Jr. and D. G. Frey, Editors, 1965. The Quaternary of the United States. Princeton University Press, Princeton, N.J. Gray,C. 1960. Geologic Map of Pennsylvania. Pennsylvania Geological Survey, fourth series. Commonwealth of Pennsylvania, Harrisburg. JENNINGS, J.D. 1968. Prehistory of North America. McGraw-Hill, New York. *Note added in proof: Collections of vittaria gametophytes have recently been made in Lake and Geauga Counties in northeastern Ohio by Allison Cusick of the Ohio Department of Natural Resources. The plants were collected from Sharon Conglo- merate exposed along the Lake Erie watershed. This site is approximately 80 km north of the Wisconsin glacial terminus and constitutes the first documented occur- rence of Vittaria within the boundaries of Wisconsin glaciation. 1983] Farrar, et al. — Fern gametophytes 91 Littte, E.L.,Jr. 1971. Atlas of the United States Trees. Volume |. Conifers and Important Hardwoods. U.S. Dept. Agric. Misc. Publ. 1146. Washington, D.C. MCALPIN, B., & D. R. FARRAR. 1978. Trichomanes gametophytes in Massachu- setts. Amer. Fern J. 68: 97-98. Mutter, E. H. 1965. Quaternary Geology of New York. /n: Wright, H. E., Jr. and D. G. Frey, Editors. 1965. The Quaternary of the United States. Princeton University Press, Princeton, N.J. ScCHAFER, J. P. & J. H. HARTSHORN. 1965. The Quaternary of New England. /n: Wright, H. E., Jr. and D. G. Frey, Editors. 1965. The Quaternary of the United States. Princeton University Press, Princeton, N.J. STEERE, W.C. 1967. Bryoxiphium norvegicum, the sword moss, as a preglacial and interglacial relic. Ecology 18: 346-358. THORNBURY, D. 1965. Regional geomorphology of the United States. Wiley, New York. Waaner, W. H., Jr., & R. A. Evers. 1963. Sterile prothallial clones (Tricho- manes?) locally abundant on Illinois sandstones. Amer. J. Bot. 50: 623. Wayne, W. J., & J. H. ZUMBERGE. 1965. Pleistocene geology of Indiana and Michigan. /n: Wright, H. E., Jr. and D. G. Frey, Editors. 1965. The Quarter- nary of the United States. Princeton University Press, Princeton, N.J. WESTVELD, M. & COMMITTEE ON SILVICULTURE, NEW ENGLAND SECTION, SOCIETY OF AMERICAN Foresters. 1956. Natural forest vegetation zones of New Eng- land. J. Forestry 54: 332-338. Wuerry, E. T. 1942. Guide to Eastern Ferns. ed. 2 . Science Press, Lancaster, Pennsylvania. DRF DEPARTMENT OF BOTANY IOWA STATE UNIVERSITY AMES, IOWA 50011 JCP DEPARTMENT OF BIOLOGY MILLERSVILLE STATE COLLEGE MILLERSVILLE, PA 17551 BWMCA THE NEW YORK BOTANICAL GARDEN BRONX, NY 10458 92 Rhodora [Vol. 85 APPENDIX | LIST OF COLLECTIONS! Vittaria2 PENNSYLVANIA: Huntingdon Co.: Orbisonia Gap, Juniata and Tuscarora sand- stone, 81-9-20-6, 81-9-20-7. Lancaster Co.: Kelley’s Run, Wissahickon schist, 81-9- 22-6; Tucquan Glen, Wissahickon schist, 81-10-2-5. Lawrence Co.: McConnell’s Mill—Gorge of Slippery Rock Creek, Pottsville sandstone, 81-10-21-4. Warren Co.: Jake’s Rocks on Allegheny Reservoir, Pottsville sandstone, 81-10-20-1. York Co.: Otter Creek, Wissahickon schist, 81-9-22-12. Trichomanes2 CONNECTICUT: Litchfield Co.: Campbell Falls—gorge south of falls, schist, 81-10- 17-2; Kent Falls, granite-gneiss, 81-10-17-3. MARYLAND: Cecil Co.: Susquehanna River—east side, | km south of Pa. state line, Peters Creek Schist, 81-9-23-1. MASSA- CHUSETTS: Franklin Co.: West of Mt. Toby along Hwy 47, 5 km north of Sunderland, Triassic conglomerate and shale, 81-10-16-1. NEW HAMPSHIRE; Cheshire Co.: Chester- field Gorge, quartzite, 81-10-15-1. NEW JERSEY: Warren Co.: Delaware Water Gap— Van Campen’s Brook, Bloomsburg shale and siltstone, 81-9-27-2. NEW YoRK: Green Co.: | km south of Westkill on Hwy 42, Catskill sandstone, 81-10-10-1. Schoharie Co.: Minekill Falls, sandy shale, 81-10-10-3. Sullivan Co.: Frost Valley, Catskill sandstone, 81-10-9-5. Ulster Co.: Frost Valley, Catskill sandstone, 81-10-9-8. West- chester Co.: Mianus River Gorge, schist, 81-10-9-2. PENNSYLVANIA: Blair Co.: Bell's Gap, Pocono sandstone, 81-9-19-1. Butler Co.: Jennings Environmental Education Center, Pottsville sandstone, 81-10-21-1. Cambria Co.: Bell's Gap, Pocono sand- stone, 81-9-19-3. Carbon Co.: Hickory Run, Pocono sandstone, 81-9-25-1. Hunting- don Co.: Juniata River—junction of US 22 and Pa 829, Catskill sandstone and shale, 81-9-20-1; Orbisonia Gap, Juniata sandstone, 81-9-20-5. Lancaster Co.: Kelly’s Run, Wissahickon schist 81-9-22-8; Pequea River, Wissahickon schist, 81-10-2-8: Furnace Hills—Seglock Run, Triassic conglomerate and shale, 81-9-30-1. Lawrence Co.: McConnel’s Mill—Gorge of Slippery Rock Creek, Pottsville sandstone, 81-10-21-7. Luzerne Co.: Rickett’s Glen—Kitchen Creek Falls, Pocono sandstone, 81-10-18-1. Monroe Co.: Delaware Water Gap—Slateford Creek, Martinsburg shale, 81-9-27-1. Pike Co.: Bushkill Falls, Marine Bed shale and sandstone, 81-9-26-1; Dingman’s Falls, Marine Bed shale and sandstone, 81-9-26-8. Tioga Co.: Colton Point—Pine Creek Gorge, Catskill sandstone, 81-10-19-1. York Co.: Susquehanna River at Accomac, Chickies quartzite. 81-9-30-1; Otter Creek, Wissahickon schist, 81-10-12-4. VERMONT: Addison Co.: 2 km west of Ripton, quartzite, 81-10-12-1; Moss Glen, schist, 81-10-12-2. Washington Co.: Mad River Glen, schist, 81-10-12-4. ‘Collection numbers are those of Farrar. ?All specimens are gametophytes only. ADIANTUM PEDATUM SSP. CALDERI, ANEW SUBSPECIES IN NORTHEASTERN NORTH AMERICA WILLIAM J. Copy The presence, in the Shickshock Mountains of the Gaspé Penin- sula, Quebec, of an Adiantum which differed substantially from Adiantum pedatum L. of rich deciduous woods in northeastern North America was pointed out by Fernald (1905). This plant he concluded was var. aleuticum Rupr., which is known from Califor- nia to Alaska and eastern Asia. Rugg (1922) later reported it from Belvidere Mountain in northwestern Vermont, and Mousley (1923) reported it from Stanstead County, Quebec, and western New- foundland. An examination of some 300 sheets of Adiantum pedatum ssp. pedatum from eastern North America and over 100 sheets of A. pedatum ssp. aleuticum (Rupr.) Calder & Taylor from western North America has shown that the plants of the Shickshock Moun- tains and parts of southeastern Quebec and adjacent Vermont and western Newfoundland are even more distinct than ssp. aleuticum is from ssp. pedatum. They may be distinguished from ssp. aleuticum and ssp. pedatum by their generally shorter stature, stiffly crowded stipes, glaucous fronds, smaller pinnules and conspicuous indusia. The plants are found on serpentine and dolomite talus slopes and tablelands. Of particular phytogeographic significance is the presence of similar plants on serpentine rocks in Washington and northern Cali- fornia. This distribution pattern is shared with another fern, Poly- stichum scopulinum (D.C. Eaton) Maxon. The only specimen of Adiantum pedatum in the Linnaean Her- barium at London (/252.2) was collected by Kalm in Virginia. It is the typical plant of our northeastern North American woodlands, in which the fronds have arched and strongly recurved branches. The type material of ssp. aleuticum came from Unalaschka and Kodjak. A fragment of the Chamisso collection from Unalaschka has been seen (NY). It is the characteristic plant found from Alaska to California. The following name is proposed for the plant of serpentine and dolmitic rocks: Adiantum pedatum L. ssp. calderi ssp. nov. 93 94 Rhodora [Vol. 85 A ssp. pedato et ssp. aleutico statura plerumque minore, stipitibus rigide aggregatis, frondibus glaucis, pinnulis semper parvioribus et indusiis conspicuis distinguendum. Ho.Lotype: cold sheltered ra- vines, Mt. Albert, Gaspe County, Quebec, J. F. Collins & M. L. Fernald, s.n. 14 August 1905 (DAO); IsoTYPE (GH) (Figure 1). A chromosome number of n = 29 (det. G. A. Mulligan) has been obtained from a collection from Mount Albert (C. Rousseau s.n., § Sept. 1965 (DAoO)). This is the same as for both ssp. pedatum and ssp. aleuticum. Many collections of this subspecies have been made by later col- lectors from the type locality, e.g. Fernald & Collins 270 (Gu, DAO), Marie-Victorin 17000 (DAO), Cody et al. 718 (DAO). The following are additional localities. CANADA. Quebec: Megantic Co., Black Lake, en grosses touffes le long des ruisselets sur les terrains magnésiens decouverts, Marie-Victorin et al. 46, 653 (DAO) plus several other collections; Caribou Lake east of Black Lake, shaded serpentine, Fernald & Jackson 11959 (CAN, GH); Caribou Hill, Black Lake, dry serpentine slopes and crests, Fernald & Jackson 11957 (CAN, GH); Coleraine, sur les collines de serpentine depuis longtemps denudeées, Marie-Victorin et al. 45, 434 (CAN); Thedford Mines, sur les serpentines pres des mines, Marie- Victorin et al. 4125 (CAN); Sherbrooke Co., Sawdust Lake, Chain Ponds, on rock ledges firmly rooted in narrow crevices, rocky bluff overlooking lake beneath young red pines (partly shaded), Terrill 7970 (pao); Wolfe Co., Lac Sunday, talus de’eboulis de serpentine, Blais et al. 11,428 (DAO). Newfoundland: Humber District: Goose Arm, Parson’s Cover, on dry limestone talus, Rou/eau 1751] (DAO); Serpentine (Coal) River, Red Barren Brook, dry serpentine shores, Rouleau 1893 (DAO, CAN); North Arm, woods on southerly slopes of dry serpentine ridge, Rouleau 872 (DAO); North Arm, Flagstaff Point, southerly slopes of dry serpentine ridge, Rouleau 842 (DAO); Cloudy Pond Mountain, on the dry limestone talus, Rouleau 200 (DAO, CAN); Bear Lake, limestone talus-slopes, Rouleau 4584 (DAO, CAN); Rileys Brook, barren serpen- tine top of hill, Rouleau 4423; Blomidon Mountains, serpentine and magnesian limestone barrens, Fernald & Wiegand 2308 (CAN); Benoit’s River, Humber Arm, 48°58’N, 58° 16’W, Tuomikoski 426 (CAN); St. Barbe District: Bonne Bay, Stanley- ville, on dry limestone talus, Rouleau 443 (DAO, CAN); Trout River Big Pond, Rocky Brook, dry serpentine slopes, Rouleau 3278 (DAO); Bonne Bay, the Tableland, dry serpentine barren slopes and tableland, Rouleau 3773 (DAO, CAN); Gros Morne National Park, Trout River Pond, 49°17’N, 58°03’W, Larix laricina scrub with Betula papyrifera on serpentine talus slope, Bouchard & Hay 73-5 (CAN); Gros Morne National Park, Winter House Brook, 49°28’N, 57°57’W, serpentine table- land, Bouchard & Hay 73-4 (DAO), Bonne Bay, serpentine tablelands, Fernald & Wiegand 2309 (CAN); St. Georges District: Bond Asbestos Mine, serpentine outcrop, Rouleau 4436 (DAO, CAN); White Bay District: St. Anthony, east slope of Eastern White Hills, moist hillside, Savile & Vaillancourt 2776 (pao); Flatwater Pond, 49° 48’N, 56° 19’W, on rocky lake shore, Shchepanek & White 2812 (CAN). u.s.A. Vermont: Belvidere Mt., Eden, Mrs. Jolley, s.n. July 1922 (Gu); Mont- gomery, Mrs. Jolley, s.n. in 1924 (Gu). California: Gasquet, Del Norte Co., moist 1983] Cody — Adiantum pedatum ssp. calderi 95 Figure 1. Photograph of type specimen of Adiantum pedatum ssp. calderi (Col- lins & Fernald, Mount Albert, Quebec, 14 Aug. 1905 (DAO)). 96 Rhodora [Vol. 85 spots in serpentine formation, Tracy 9520 (DAO). Washington: Wenatchee National Forest, Mount Stewart, in crevice in moist draw on steep slope at foot of Esmerelda Peak, greenstone, Cody 18454 (DAO). The subspecies is named for James A. Calder, a former colleague and co-author of Flora of the Queen Charlotte Islands (Calder & Taylor, 1968), who pointed out the distinctiveness of this fern. The following key may be used to separate the three subspecies of Adiantum pedatum: A. Branches of fronds widely divergent, with the lowermost arched and strongly recurved (rich deciduous woods, north- eastern North America) .............0008; ssp. pedatum A. Branches usually strongly ascending B. Fronds tightly clumped; pinnules glaucous, small, the mid- dle 7-12 (17) mm long, but little incised; indusia conspicu- ous (serpentines and dolomites, western Newfoundland, Gaspe, Megantic and Sherbrooke counties Quebec, north- ern Vermont, California and Washington) ............. ‘ves eee ne iene eeaeyadu Ra bbe e ea eae ssp. calderi Cody B. Fronds not tightly clumped; pinnules green, the middle larger (10) 12-20 (23) mm long, frequently deeply incised; indusia not conspicuous (rocky woods, Alaska to Cali- POTWIA Tic ds 243 3 3 ssp. aleuticum (Rupr.) Calder & Taylor The loan of specimens from National Museum of Natural Sci- ences, Ottawa (CAN), Gray Herbarium of Harvard University (Gu), and New York Botanical Garden (Ny), is much appreciated. LITERATURE CITED CaLper, J.A.& R.L. TAyior. 1968. Flora of the Queen Charlotte Islands, Part I. Systematics of the Vascular Plants. Canada Dept. Agriculture, Research Br. Monograph 4 Part |. 659 pp. FERNALD, M.L. 1905. An alpine Adiantum. Rhodora 7: 190-192. Moustety, H. 1923. The alpine maidenhair fern (Adiantum pedatum L. var. aleu- ticum Rupr.) at Hatley, Stanstead County, Quebec. Can. Field-Nat. 27: 84-85. RuGG, H.G. 1922. Adiantum pedatum var. aleuticum in New England. Amer. Fern Jour. 12: 128-129. BIOSYSTEMATICS RESEARCH INSTITUTE AGRICULTURE CANADA OTTAWA, ONT., CANADA KIA 0C6 CONTRIBUTIONS TO THE REPRODUCTIVE BIOLOGY OF PANAX TRIFOLIUM L. (ARALIACEAE)! C. THOMAS PHILBRICK Abstract: Flowering phenology, pollination, and morphological aspects of pollen and seeds of Panax trifolium L. are investigated. Plants are apparently androdioe- cious, but with very minor differences in pollen between floral forms (staminate and hemaphrodite). Light microscope and SEM studies were made of pollen and seeds. Panax trifolium L. (Dwarf ginseng) is a springtime herb of east- ern North America which perennates as a bulbous underground rhizome in moist forested areas. The genus is also represented in the region by the more economically important Panax quinquefolium. However, the latter is much less common due to over-collection for its reputed medicinal purposes. There is relatively little information available concerning the biology of either species of Panax. What is available deals largely with P. quinquefolium? and is derived mostly from cultivated mate- rial, rarely from plants in their natural habitat (Hu, et al., 1980). Hu, et al. discussed the general ecology and habitat of both species but pointed out the lack of information concerning their growth habit, population structure and longevity of individuals. There are also few published reports for either species concerning field observa- tions on flowering phenologies, breeding systems, and pollination. Morphologically Panax trifolium is polygamous, with staminate and hermaphroditic flowers occurring on separate plants. The func- tion of the gynoecium in the staminate flower is seldom questioned due to its very reduced nature; it completely lacks an ovary. How- ever, the functionality of the pollen in the seed-producing hermaph- roditic flowers is not as apparent and has not been investigated. A result of this lack of understanding is illustrated by the varied des- criptions of sexual condition in the literature; “mostly unisexual” (Hu, et al., 1980), “dioeciously polygamous” (Fernald, 1950), and “often unisexual” (Gleason & Cronquist, 1963). 'This study was supported by the Research Initiation Fund of the University of New Hampshire. Scientific Contribution No. 1162 from the New Hampshire Agricul- tural Experiment Station. 2While this paper was in press another article dealing with Panax quinquefolium was published: Lewis, W. H., & V. E. Zenger. 1982. Population dynamics of the American Ginseng Panax quinquefolium (Araliaceae). Amer. J. Bot. 69: 1483-1490. 97 98 Rhodora [Vol. 85 The purpose of this paper is to report on investigations of several aspects of the reproductive biology of Panax trifolium. These include flowering phenologies, pollination, and morphological as- pects of pollen and seeds. The results of these investigations suggest that Panax trifolium is androdioecious, a very uncommon and hence very interesting form of polygamy in which staminate and hermaphroditic flowers occur on separate plants (Lloyd, 1975; Charlesworth & Charlesworth, 1978). MATERIALS AND METHODS This study is based primarily on observations of two populations of Panax trifolium located in Durham, New Hampshire. For flower- ing phenologies 15 staminate and 15 hermaphroditic plants were selected, labelled with colored threads, and numbered. Each of these plants was then visited daily from 27 April to 4 June, 1981. Fruit development was subsequently followed in the same manner. Observations of pollinators were made during the above visits to the populations. Floral insect visitors were collected by applying fine strands of Tanglefoot (The Tanglefoot Co., Grand Rapids, Michigan) to inflorescences of 10 plants of each type. The trapped insects were then collected the next day and preserved in 95% ethyl alcohol for later identification. Pollinator exclusion experiments were conducted by covering 15 hermaphroditic plants with nylon screening before the flowers matured. The screens were removed after the flowering period and possible fruit development observed. For pollen morphology studies, material was collected from both populations, acetolyzed (after Faegri & Iversen, 1975), rinsed twice with distilled water, and transferred to glycerine jelly. Microscope slides were then prepared from this material for light microscope (LM) observations. For scanning electron microscope (SEM) observa- tions acetolyzed pollen was washed in two rinses of 100% acetone and transferred to aluminum stubs. The mounted material was then coated with approximately 200 angstroms of paladium gold. Seeds were collected from each population for morphological studies. Observations were made of air dried seeds with a binocular dissecting microscope. For sEM observations air dried seeds were mounted on aluminum stubs using double-stick tape and coated 1983] Philbrick — Reproduction of Panax trifolium 99 with paladium gold. All seM observations were made on an AMR 1000 Scanning Electron microscope at the University of New Hampshire. Descriptions of pollen and seeds are based on LM observations with supplemental information from sEM micrographs. All data in this study were evaluated statistically using an analysis of variance, or in the case of P/E ratios (pollen morphology), a Mann-Whitney U-test. The amount of pollen produced per anther was estimated by counting all the grains in one sac and multiplying by two, pollen: ovule ratios were then calculated. Pollen was stained on glass micro- scope slides by dissecting open anthers in a drop of the respective stain (Analine blue, Aceto-carmine, and Malachite green-acid fuch- sin). Cover glasses were then applied and observations made by examining most grains on the slide. Investigations of herbarium material were made in three herbaria; Hodgdon Herbarium, University of New Hampshire (NHA), The New England Botanical Club Herbarium (NEBC) and the Gray Her- barium (GH) at Harvard University. Observations were made by examining each reproductive specimen and noting the sexual makeup of the inflorescence. POLLEN MORPHOLOGY Pollen grains of Panax trifolium are radially symmetrical, isopo- lar and tricoloporoidate, with three long, narrow colpi reaching nearly to the poles. The poroid areas are formed by a distinct zone of nexine thickening which incompletely surrounds a thinner area in the equatorial region of the colpus (Fig. 2, D,E). The regions of thickening form two separate semi-circular areas on opposite sides of the thin area and aligned parallel to the equatorial axis (Fig. 3, E). The poroid areas are approximately 4 to 5 wm in width, mea- sured along the polar axis, and 5 to 6 um in length, measured along the equatorial axis. The nexine is also thickened along the margins of the colpi (Fig. 2, G; Fig. 3, A), thus forming a margo that is apparent when the grain is expanded but is less distinguishable when the colpi are invaginated. In cross section the tectate exine is composed of two distinct zones; the inner nexine and outer sexine (Fig. 2, C; Fig. 3, A, B). 100 Rhodora [Vol. 85 Fig. |: Two floral types of Panax trifolium: A, staminate, B, hermaphrodite. The nexine is approximately | um in thickness throughout, exclud- ing the previously described poroid regions. The sexine is thicker than the nexine (Fig. 2, C), ranging from about 2 um in the center of the equatorial zone to approximately 3 um in the central polar area. In both of these locations the sexine becomes thinner toward the colpi. The sculpturing of the exine is difficult to distinguish through LM analysis alone. Different characters of the exine become apparent when LM observations are compared with those from the sem. Under LM the grain surface appears reticulate (Fig. 2, A, B, F). SEM micro- graphs, however, illustrate a distinctly different striate tectum (Fig. 2, H,I; Fig. 3, B). This difference presumably results from a lack of Fig. 2: A -— E, series of LM micrographs focusing through a single grain in equatorial view. A, high focus showing columellae, appearing white (arrow) (1587X). B, slightly lower focus showing dark columellae (arrow) forming a reticu- late appearance (1587X). C, optical cross-section of grain showing thin area of exine in the equatorial region (1587X). D, optical cross-section of poroid areas (1587X). E, surface view showing one colpus and poroid area (1587X). F—G, LM micrographs of grain in polar view. F, high focus showing white columellae (1587X). G, optical cross-section (1587X). H—I, SEM micrographs of separate grains in equatorial view. H, showing two invaginated colpi and the striate nature of the sexine (1235X). 1, showing two colpi with slightly protruding colpus membranes in the poroid area (1100X). 1983] Philbrick — Reproduction of Panax trifolium 101 102 Rhodora [Vol. 85 ee a? et < ee *. oe 1983] Philbrick — Reproduction of Panax trifolium 103 relief on the tectum, therefore causing it to appear more or less transparent under LM analysis. This phenomenon is illustrated in Kapp’s (1969) LM description for grains of the species in which he reports the sexine to be reticulate. Thus the reticulate appearance Is a result of the orientation of the subtending columellae (baculae), not the sculpturing of the tectum itself (Fig. 3, A, B). Very careful LM analysis shows faint striate sculpturing in the peripheral regions of some grains. No distinct differences in exine sculpturing were found in pollen from the two floral types. However, upon comparison of size mea- surements some differences were revealed (Table 1). The mean equa- torial axis lengths of grains from the two floral types were shown to be similar, but mean polar axis lengths were significantly different. Grains from hermaphroditic flowers exhibited slightly longer polar axis lengths on the average. Also, upon comparison of the ratios of polar to equatorial axis lengths (P/E ratio of Erdtman, 1952) signif- icant differences were again shown (Table |). Grains from the stam- inate flowers possessed P/E ratios ranging from 0.97-1.63 (sub- oblate-prolate) with a mean of 1.30 (subprolate). In comparison those from the hermaphroditic floral form ranged from 1.00-1.70 (prolate spheroidal-prolate) with a mean of 1.45 (prolate). SEED MORPHOLOGY Seeds of Panax trifolium are white and somewhat reniform in shape. The length ranges from 2.5 to 3.5 mm and the width from 1.5 to 2.5 mm (Table 1). The seed coat is characterized by a reticulate network of ridges, from which arise abundant short stiff unicellular hairs (Fig. 3, C, D). Kapil, et al. (1980) state that seed hairs are usually adaptations for wind dissemination (anemochory), but note that they also function Fig. 3: A — B, SEM micrographs of sliced open grains. A, grain sectioned approximately in the equatorial plane showing the thickening of the nexine along the colpi (arrow), also showing the tectum (t) and columellae (c) (1655). B, grain sectioned obliquely, showing the two separate nexine thickenings (arrow) parallel to each other in the equatorial zone (2330X). © — E, SEM micrographs of seeds. C, whole seed showing unicellular hairs high-lighting the reticulate pattern (10%). D, close-up of the seed coat showing the unicellular hairs (75x). E, whole seed showing split seed coat (15x). F, LM micrograph showing an andromonoecious inflorescence; staminate flower (m), hermaphrodite flower (h) (1.5%). Table 1: Data for comparison of characters of staminate and hermaphrodite individuals. Staminate Plant Hermaphrodite Plant F-ratio Character Range Mean S.D. N= Range Mean S.D. N= (ANOVA) eA arbi sa 941 19.3 7.53 33 3-12 7.1 2.23 33 79.87 inflorescence Total duration of inflorescence 12-19 14.5 2.46 15 5-8 6.0 115 15 97.78 flowering (days) Duration from anthesis to fruit maturation -- --- --- --- 20-24 22.5 1.32 15 --- (days) Pollen measurements; Polar axis (um) 38-49 43.7 2.61 34 34-49 41.6 3.29 34 7.73 Equatorial axis (um) 25-40 31.9 3.83 34 27-38 30.5 2.12 34 3.53 P/E ratio 0.97-1.63 1.30 --- 34 1.00—1.70 1.45 -—- 34 --- No. of grains per anther sac (X2 for 180-253 210 22.40 22 40-243 157 47.00 22 22.57 total no. in anther) Percent stainability Analine Blue --- 99.3% --- 3 --- 93.9% --- 3 --- Aceto-carmine --- 91.3% --- 3 --- 99.7% --- 3 --- Mal Gr./ Ac. Fucs. --- 99.7% --- 3 --- 97.3% oo 3 --- Flower size (um) 1.0-1.5 1.2 1.75 30 1.5-2.5 1.30 0.25 30 149.76 Seed length (um) -_ -—- -—- -— 2.5-3.5 3.05 1.75 30 vOl elopoyy $8 1OA] 1983] Philbrick — Reproduction of Panax trifolium 105 in water (hydrochory) and animal (epizoochory) dispersal. These mechanisms of dispersal do not seem to apply to the seeds of Panax trifolium, which fall to the ground and filter into the leaf litter. Although the seeds of Panax trifolium germinate the next spring after they are shed, I have observed that the seed coat actually splits in the late fall (Fig, 3, E) (Table 2), not just before germination in early spring. FLORAL MORPHOLOGY The staminate flower of Panax trifolium (Fig. 1, A) is composed of a subconical floral tube which surrounds a persistent, reduced, solitary style. Five sepals and five petals alternate on the lip of the floral tube. Five two-celled anthers also arise from the margin of the floral tube opposite the sepals. The hermaphroditic flower (Fig. 1, B) is composed of a cylindrical floral tube fused below to a three-carpellate inferior ovary which is noticeably swollen at its base. Three styles arise from the top of the ovary. Five sepals and five petals alternate on the lip of the floral tube, along with five two-celled anthers located opposite the sepals. In the majority of cases the inflorescences of Panax trifolium are composed of only one or the other floral form: staminate or her- maphrodite. However, a survey of specimens located in several her- baria revealed that occasional andromonoecy exists; both staminate and hermaphroditic flowers on the same plant (Fig. 3, F). In a sample of 753 plant specimens, on approximately 260 sheets, 516 (69%) possessed only staminate flowers, 229 (30%) only hermaph- roditic flowers and 8 (1%) had both flower types on the same inflo- rescence. I have not observed andromonoeceous individuals in the field. FLOWERING PATTERNS AND SECONDARY SEX CHARACTERS Two distinct plant types occur in Panax trifolium. One produces morphologically and functionally staminate flowers. These possess a full complement of five anthers as well as a single short style, but lack a noticeable ovary. The other plant type produces morphologi- cally hermaphroditic flowers possessing one three-carpellate ovary, each carpel capable of producing a single seed. These flowers also contain the full five anther complement. 106 Rhodora [Vol. 85 Table 2: Pollen:ovule ratios; percent seeds with split seed coats; percent seed set; results of bagging. Percent seed set: Total number of fruits collected: 154 from 22 plants. Number with full (3/3) seed set 93 (60.4%) Number with less than full seed set 61 (39.6%) Percent seeds with split seed coat: Date coll. N= Number with split Number with non- seed coat split seed coat 4 Nov. 1981 53 34 (64.2%) 19 (35.8%) 30 Nov. 1981 33 31 (93.9%) 2 (6.1%) 23 Jan. 1981 16 13 (81.3%) 3 (18.7%) Pollen:Ovule ratio: total no. grains per staminate flower total no. ovules per hermaphrodite = 696:1 flower total no. grains per staminate and hermaphrodite flowers combined 1220:1 total no. ovules per hermaphrodite flower mean no. grains per staminate inflor. 986:1 mean no. ovules per hermaphrodite inflor. mean no. grains per staminate and hermaphrodite inflor. 2378: 1 mean no. ovules per hermaphrodite inflor. Seed set. No. seeds set No. plants observed Plants left uncovered (control) 25 10 Plants covered 0 15 1983] Philbrick — Reproduction of Panax trifolium 107 Both plant types of Panax trifolium bloom concurrently within the population, with floral development and maturation taking place while the inflorescence is in a drooping position. Lloyd and Webb (1977) reported that staminate individuals of a dimorphic spe- cies often begin flowering before the seed-producing plant. I have not found this to be the case with Panax trifolium. I have found several distinct differences between the two plant types (Table 3) that have been termed secondary sex characters (Lloyd & Webb, 1977). Distinct size differences of flowers, mea- sured in bud, between the two plant types is obvious (Table 1; Fig. 3, F). This distinction is directly related to the presence of a fully developed ovary in one floral form and not in the other. Differences in numbers of flowers per inflorescence and their pattern of flower- ing were apparent. Significantly more flowers were observed on staminate inflorescences than on hermaphroditic, averaging 19 and 7 respectively (Table 1). The sequence of flowering of the staminate inflorescence was from the outermost flowers inward (centripetal), in contrast with the hermaphroditic inflorescences in which the flowers all opened more or less simultaneously. As a result of the differences in flowering patterns, the duration of flowering of the respective inflorescences differed markedly. Individual flowers, whether staminate or hermaphrodite, remained open for about four to five days. However, the entire staminate inflorescence blooms about twice as long as the hermaphrodite, averaging about 15 days for the staminate versus 6 for the hermaphrodite (Table 1). On the Table 3: Secondary sex characters used in comparison of the two plant types. 1) Number of flowers per inflorescence. 2) Total duration of flowering per plant. 3) Sequence of flowering of respective inflorescence. 4) Amount of pollen produced per anther. 5) Percent stainability of pollen. 6) Flower size. 108 Rhodora [Vol. 85 population scale this prolongs the total time that pollen is available and no doubt plays an important role in the pollination biology of the species. During anthesis the staminate flower sheds its pollen and the anthers then fall. However, the perianth persists on these spent flowers and the flowers remain attached to the inflorescence until all the flowers have passed. After this time the inflorescence wilts and falls, or the entire stem dies. The persistent perianths may contribute to the attractiveness of the staminate inflorescence, drawing insects to the inner flowers that are still shedding pollen. Anthesis proceeds very differently in the hermaphroditic flower. As the flowers open the anthers dehisce while the styles remain pressed together. After anther dehiscence, the styles begin to separ- ate, becoming fully divergent after the anthers have fallen (pro- tandry). Occasionally some overlap exists between the time of anther dehiscence and style divergence. The styles remain divergent for three to four days while pollination takes place. After this time they discolor and wilt slightly, after which the perianth is shed. The significance of simultaneous blooming of the hermaphroditic inflorescence may be related to its protandrous nature. If non- simultaneous flowering occurred, selfing between flowers within an inflorescence (geitonogamy) would be more likely to take place. The simultaneous nature of flowering may increase the probability of outcrossing. At the time of stigma divergence there would be a better chance that all the pollen had been shed and the probability of outcrossing would be increased. Cruden (1977b) reported this same character in protandrous umbelliferous taxa. The staminate plant often lives for only a few days after anthesis. Due to the fruit development period, the hermaphroditic stem per- sists well past anthesis. The initiation of fruit development, indi- cated by ovary expansion, begins very soon after the styles have wilted, often before the entire perianth has fallen. The time from style wilting to fruit maturation and drop averaged 28 days (Table 1). In many cases fruits which lacked the full three seed complement were observed (Table 2). It is not known whether this lack of seed set is related to an insufficient pollen or nutrient supply or to some other factor. Upon maturation the fruit becomes slightly yellow and falls from the plant, a result of an apparent abscission zone at the base of the ovary. On the ground the fruit wall quickly breaks down, 1983] Philbrick — Reproduction of Panax trifolium 109 often along the septations between carpels. The seeds then fall from the fruit, sometimes forming small piles where several fruits have fallen together. My observations differed from those of Hu, et al. (1980) where they observed seeds being expelled from the fruits while the latter were still attached to the plant. After the fruits have fallen the stem lives for several more days, turns yellow and dies. Once shed, the seeds filter down into the litter. Germination takes place early the following spring. Differences also exist between the two floral types regarding the amount of pollen produced per anther; the staminate producing approximately 25% more pollen per anther than the hermaphrodite (Table 1). Although less pollen is produced in the hermaphroditic flowers, that which is produced seems significant in amount and therefore should be considered when assessing pollen flow within the population. The staining qualities of the pollen from the two types also differed somewhat. When stained, pollen from staminate flowers consistently showed a higher staining percentage (less grain abortion), about 100% for staminate and 95% for hermaphroditic (Table 1). I prefer to use the phrase “staining percentage” rather then “viability percentage” because the staining characteristics of pollen are only an estimate of viability, not a definitive test. Although the observations were not quantified and were some- what inconsistent, there seemed to be differences in the degree to which the different grains took up the stain: the grains from stami- nate flowers often stained slightly darker than those from the other floral type. These slight differences could signify fertility differences and are therefore noteworthy. If the breeding system of Panax trifolium is one of functional dioecy, distinct differences between functional versus non-function- al pollen produced by the two floral types would be expected. Bar- rett and Helenurm (1981) reported that the anthers of the function- ally female flowers of Aralia nudicaulis did not produce pollen, and Bawa (1977) found that anthers in similar functionally female flow- ers of Cupania guatemalensis produced pollen, but that anther deh- iscence failed to occur. Anderson and Gensel (1976) and Anderson (1979) reported the lack of aperture development in pollen from what were interpreted as functionally female flowers of a species of Solanum. | have not observed any of these phenomena in P. trifo- lium other than the size differences noted earlier in the discussion of pollen morphology. 110 Rhodora [Vol. 85 POLLINATION AND SEX RATIOS Although not referring to dimorphic taxa, Cruden (1977a) illus- trated a relationship between the ratio of pollen grains to number of ovules in a flower and its pollination mechanism. The more hapha- zard the mechanism the higher the pollen/ovule (P/O) ratio. This ratio can be estimated variously for Panax trifolium (Table 2). However, even in the estimate resulting in the largest ratio (2378.1), this ratio is still relatively small and suggests a rather predictable pollination mechanism. This agrees with the observations of polli- nators made throughout this study. Pollination in the species appears to be primarily entomophilous. Members of the insect families Empididae, Scatopsidae, and Syr- phidae (Diptera) were observed visiting, and were collected upon, flowers of P. trifolium. Members of the Syrphidae were observed visiting the staminate flowers, where they initially collected pollen from several anthers, cleaned themselves of excess pollen and then reached to the base of the floral tube for an apparent “drink of nectar.” This series of events was seen on several occasions and was also observed, but in less detail, on hermaphrodite flowers. During early morning visits to the populations mosquitos were often seen sitting on flowers of Panax trifolium as well as those of Coptis groenlandica and Anemone quinquifolia. Although mosqui- tos are known to take part in pollination of some orchids (Stoutam- ire, 1968, 1971) these insects were never observed visiting the reproductive organs of P. trifolium flowers and were presumably only “resting.” A random walk through one population revealed a sex ratio highly skewed to staminate individuals; 110 staminate versus 16 hermaphrodite (approximately 7:1). This, combined with the higher numbers of flowers in staminate versus hermaphrodite inflorescen- ces, indicates a relatively high pollen to ovule ratio at the population level. More detailed investigations into this phenomenon are re- quired before the true significance of this ratio can be appreciated. As a result of caging experiments in which 15 hermaphroditic plants were covered with nylon screening it was shown that signifi- cant anemophily probably does not occur (Table 2). In all cases no fruit development occurred on the covered plants whereas all unco- vered plants in the same area exhibited significant seed set. This also 1983] Philbrick — Reproduction of Panax trifolium 111 rules out mechanical self pollination but not necessarily self incompatibility. CONCLUSION The results of these investigations suggest that the pollen pro- duced in the hermaphroditic flower appears functional, although possibly to a lesser extent than that from the strictly staminate flower. This, combined with the protandrous nature of the her- maphrodite flower, suggests that androdioecy may better character- ize the breeding systems of Panax trifolium than previous designa- tions. However, actual confirmation of androdioecy will not be accomplished until detailed crossing experiments within and be- tween the two plant types are completed. Panax trifolium is unique in that it exhibits ideal characters needed to conduct a survey regarding age structure of staminate and hermaphroditic plants in a population. During the growing season rhizomes of staminate and hermaphrodite individuals can be unco- vered and their age determined by counting the persistent leaf scars. This allows a unique opportunity to estimate the longevity of the sexes, usually a very difficult undertaking. During the spring of 1981 a rhizome of P. trifolium was excavated which possessed approxi- mately 30 leaf scars on a portion about 4 cm in length. Because one leaf is produced per year the plant was therefore about 30 years old. The results of this study make a contribution to the current knowledge of several aspects of the reproductive biology of Panax trifolium. However, there remain many additional areas of investiga- tion which have yet to be addressed, regarding among others, sex ratios, confirmation of pollen fertilities, self-compatibilities, and population age structure. ACKNOWLEDGMENTS I would like to thank Dr. A. Linn Bogle, Dr. Mark A. Schless- man, Dr. Garrett E. Crow, and Dr. Thomas D. Lee for their critical review of the manuscripts. I also express my sincere thanks to Dr. Thomas D. Lee for his continued assistance in designing and carry- ing out this study. Ms. Marilyn Ecker of the University of New Hampshire Electron Microscopy Facility is thanked for her techni- 112 Rhodora [Vol. 85 cal expertise in producing the SEM micrographs. Dr. Siegfried E. Thewke is thanked for his identifications of the insects collected. I also thank the curators of the Gray (GH), New England Botanical Club (NEBC), and Hodgdon Herbaria (NHA) for allowing me to examine the specimens of Panax trifolium in their collections. LITERATURE CITED ANDERSON, G. J. 1979. Dioecious Solanum species of hermaphroditic origin is an example of broad convergence. Nature 282: 836-838. . & P.G. Genser. 1976. Pollen morphology and the systematics of Sola- num section basarthrum. Pollen et Spores 18: 533-552. Barrett, S.C. H. & K. HELENURM. 1981. Floral sex ratios and life history in Aralia nudicaulis (Araliaceae). Evolution 35: 752-762. Bawa, K. S. 1977. The reproductive biology of Cupania guatemalensis Rudbk. (Sapindaceae). Evolution 31: 52-63. CHARLESWORTH, B. & D. CHARLESWORTH. 1978. A model for the evolution of dioecy and gynodioecy. The Amer. Natur. 112: 975-997. CRuUDEN, R. W. 1977a. Pollen-Ovule Ratios: A conservative indicator of breeding systems in flowering plants. Evolution 31: 32-46. 1977b. Temporal dioecism: An alternative to dioecism? Evolution 31: 863-866. ERDTMAN, G. 1952. Pollen Morphology and Plant Taxonomy. Angiosperms. Almavist and Wiksell, Stockholm. FaeGri, K. & J. Iversen. 1975. Textbook of Pollen Analysis. Hafner Press, NY. FERNALD, M. L. 1950. Gray's Manual of Botany. eighth edition. D. Van Nos- trand, Co., NY. GLEASON, H. A. & A. Cronguist. 1963. Manual of Vascular Plants of Northeast United States and Adjacent Canada. D. Van Nostrand Co., NY. Hu, S. Y., L. RUDENBERG, & P. DEL Trepici. 1980. Studies of American Gin- sengs. Rhodora 82: 627-636. Kapit, R. N., J. Bor & F. BouMAN. 1980. Seed appendages in Angiosperms. I. Introducton. Bot. Jahrb. Syst. 101: 555-573. Kapp, R. D. 1969. How to know Pollen and Spores. Wm. C. Brown Co., Dubuque, Iowa. Lioyp, D.G. 1975. The maintenance of gynodioecy and androdioecy in angio- sperms. Genetica 45: 325-339. .. & C. J. Wess. 1977. Secondary sex characters in plants. Bot. Rev. 43: 177-216. STOUTAMIRE, W. P. 1968. Mosquito pollination of Habenaria obtusata (Orchida- ceae). The Mich. Bot. 7: 203-212. 1971. Pollination in Temperate American Orchids. /n: Proceedings of the sixth World Orchid Conference, Sydney, Australia, 1969. ed. M. J. G. Corrigan, Halstead Press. 1983] Philbrick — Reproduction of Panax trifolium 113 DEPARTMENT OF BOTANY AND PLANT PATHOLOGY NESMITH HALL UNIVERSITY OF NEW HAMPSHIRE DURHAM, NH 03824 MOUNTAIN LAUREL (KALMIA LATIFOLIA L.) DISTRIBUTION IN MASSACHUSETTS BRAYTON F. WILSON AND JOHN F. O’ KEEFE ABSTRACT Mountain laurel (Kalmia latifolia L.) presence along roads was mapped exten- sively for the whole state and intensively for selected quadrangles. Presence of moun- tain laurel in selected forests was determined from plots along transects. Frequency of mountain laurel was calculated as the per cent of road distance mapped, or of plots taken, that had laurel present. Mountain laurel occurred throughout the state on a range of sites, except for Cape Cod and eastern Plymouth County, in the high Berkshires and in the far west and north-west. Mountain laurel was uncommon in eastern Massachusetts and in a north-south strip through central Massachusetts, but was otherwise abundant. Absence of mountain laurel was probably due to either fire history, maritime influence, cold winters, or basic soils. Low frequencies were prob- ably due to a previous history of extensive land clearing for agriculture. Mountain laurel (Ka/mia /atifolia L.) is an evergreen shrub widely distributed in Massachusetts (Ebinger, 1974) even though it is near its northern limit (Kurmes, 1967). We have studied the details of local distribution related to site factors in central Massachusetts (O’Keefe, 1981). Our objective in the present study was to survey the whole state of Massachusetts to see where mountain laurel grows naturally. Our primary method was to map mountain laurel visible from automobiles while driving when deciduous species were leaf- less and the evergreen mountain laurel was conspicuous. We mapped the whole state on small scale maps and portions of the state on larger scale maps. We also incorporated data from site studies that used plots along transects in various forests. The results of the mapping and plot studies were then combined to interpret state-wide variations in the occurrence of mountain laurel. METHODS We mapped mountain laurel from November to April in 1979- 1980 and 1980-1981. One person drove at 40-60 km/hr and observed mountain laurel on both sides of the road, the other helped observe and marked mountain laurel presence on maps. The exten- sive survey of state-wide distribution used USGS 1:250,000 maps; the intensive survey of portions of the state used USGS 1:24,000 115 116 Rhodora [Vol. 85 maps (quadrangles). The resolution of the surveys, the minimum distance between mountain laurel plants that permits individual marks on the map rather than marking as continuous cover, was approximately 500 m for the extensive survey and 50 m for the intensive survey. For extensive mapping we tried to select represen- tative roads that sampled the non-urbanized areas of the state. For the intensive survey we selected quadrangles within easy distance of Amherst, MA and drove on all of the passable roads in each quadrangle. The per cent of mountain laurel on maps was determined by measuring on the maps the total distance marked as having moun- tain laurel and dividing by the total distance mapped. The extensive survey mapped only roads outside urbanized areas (yellow on the 1:250,000 maps). The intensive survey mapped only roads through wooded areas (green on the 1:24,000 maps). These percentages are only for comparative purposes on maps of the same scale. Plot data from forests were taken at regular intervals along equally spaced transects that sampled the entire forest. Data from Monroe and Hawley-Savoy State Forests are froma study by Hibbs (1978) using 0.01 ha plots. Data from the other 6 forests came from a study by O’Keefe (1981) using 0.02 ha plots. RESULTS In the extensive survey we mapped about 1,800 km of road and 21% had mountain laurel. The state was composed of regions of differing mountain laurel distribution (Fig. 1, Table 1). As generally noted by others (Ebinger, 1974), we found no mountain laurel on Cape Cod or eastern Plymouth County. In eastern Massachusetts, roughly east of Worcester and an urbanized area, mountain laurel was scarce, although it was observed on a few rocky hills to the north and in some swampy areas to the south. Mountain laurel was common (about 40%) west of Worcester to the Connecticut River (region III) and in the Berkshires (region V), with the marked excep- tion of a north-south strip in western Worcester county and south of the Quabbin reservoir west to the Connecticut river. There, moun- tain laurel was scarce generally, although it was locally abundant in some swampy areas and pond margins. Mountain laurel was less common in the northern Berkshires than in the southern Berkshires. In far western Massachusetts, the northern Housatonic valley, Hoo- ars Pees, mas Pe a SS sae ~ ~ - IITA Ba: ots / . sy, 7 / vhs ~~ \ Pr ~ ~ ~ . Fs 4 ‘ Ssy “Tit s. ~ ip ee a / tol QUABBIN Ss “Ss “Ss. | a RESERVOIRS, “s. ~ 1d ee i ae » de /s “~. ae se. s, / oe “ey se “. 5 s/ | le ~ ~ ~ ~ ( r - *s ie Ss 7 ~ pa EN, ee tet he ey 30 KM eS es Se | Figure 1. Distribution of mountain laurel in Massachusetts. The center of each quadrangle intensively mapped is marked by q a + and identified with a letter (G = Greenfield, T = Mt. Toby, S = Shutesbury, B = Belchertown, A = Athol). Forests sampled with plots are marked by numbers (1 = Monroe, 2 = Hawley-Savoy, 3= D.A.R., 4= Mt. Toby, 5 = Cadwell, 6 = Harvard, 7 = Erving, 8 = Leominster). Blank areas (I and VI) had no mountain laurel, stippled areas (II and IV) had mountain laurel, but it was uncommon, in hatched areas (III and V) mountain laurel was common at least along roads. [€86l [Jainey] UIeJUNO|P, — IJ9d¥,O W UOS|TIM ra | 118 Rhodora [Vol. 85 Table |. Distribution of mountain laurel in different regions of Massachusetts. See Figure | for boundaries of the regions. % mountain Total km Region laurel mapped I Cape Cod and E. Plymouth County 0 157 Il Eastern Mass. 3 506 III E. Worcester County and N. of Quabbin Res. 38 303 IV W. Worcester County + S. of Quabbin Res. 2 161 V_ Berkshires 45 367 VI Far west 8 ra sic valley, and Mount Greylock to the north, and the Taconic range along the New York border had no mountain laurel. Mountain laurel does occur on parts of a rocky ridge that extends north into the Housatonic valley south of Pittsfield (Fig. 1). Greenfield, Mt. Toby, and Shutesbury quadrangles were all in region III where mountain laurel was abundant. They had about 40% mountain laurel along the roads (Table 2). The Mt. Toby quad- rangle has about 50% agricultural land in the Connecticut river valley, therefore the total distance mapped is lower than in the other two quadrangles, but the percentage of mountain laurel in the wooded portions of the three quadrangles was about the same (Table 2). Belchertown and Athol quadrangles were primarily in region IV where mountain laurel was scarce. The intensive survey found only 11-14% mountain laurel in these quadrangles. The Athol quadran- gle showed a sharp discontinuity between the southwest corner, where mountain laurel was very abundant, and the rest of the quad- rangle where it was scarce (Table 2). This discontinuity was along the boundary between regions III and IV (Fig. 1) where mountain laurel was respectively abundant and scarce. In the areas where mountain laurel was generally scarce in the Athol quadrangle it was locally abundant in, or near, a few swampy areas. Three of the quadrangles could be subdivided into different phy- siographic areas (Table 2). The Greenfield quadrangle contains the Montague sand plain, an area of deep sands, frequently burned, with pitch pine (Pinus rigida) and scrub oak (Quercus ilicifolia) 1983] Wilson & O’Keefe — Mountain Laurel 119 stands similar to Cape Cod. Although mountain laurel was scarce on the sand plain, it did grow there. The Mt. Toby quadrangle contains extensive, flat, wooded swamps on the west side of the Connecticut river. Mountain laurel was just as abundant in these swamps as in the rest of the wooded portion of the quadrangle, predominantly rocky hills. In the Belchertown quadrangle moun- tain laurel was equally common in the rocky Pelham hills, where it is very dense in some areas, and in the valley floors of Granby and Belchertown. The Shutesbury quadrangle is almost entirely rocky hills and mountain laurel was abundant on the whole quadrangle. The data for mountain laurel occurrence in State Forests (Table 3) are not biased towards roadside observations as are the maps. Generally the data from plots were consistent with the location of the forest in the regions of the state delineated by the extensive mapping. Both Monroe and Hawley-Savoy State Forests are at high altitudes (700-800 m) in the northern Berkshires and neither had any mountain laurel. The D.A.R. state forest is in the central Berk- shires, at lower elevation (500m) and 13% of the plots had mountain laurel. The Harvard Forest is in region IV, where mountain laurel is Table 2. Distribution of mountain laurel in selected quadran- gles. See Figure | for location of quadrangles. % mountain Total km Quadrangle laurel mapped Greenfield 36 72 Sand plain 3 5 Remainder 38 67 Mt. Toby 37 as Swamp 38 11 Remainder 34 24 Shutesbury 42 58 Belchertown 14 63 Hills 14 50 Valleys 14 13 Athol 11 88 SW corner 64 11 Remainder 3 a7 120 Rhodora [Vol. 85 Table 3. Distribution of mountain laurel in selected forests. Numbers of the forests refer to their numbers on Figure 1. % mountain Total number Forest laurel of plots 1 Monroe 0 19] 2 Hawley-Savoy 0 213 3 D.A.LR. 13 270 4 Mt. Toby 34 395 5 Cadwell 32 620 6 Harvard 10 615 7 Erving 4] 470 8 Leominster 79 370 scarce, and it had only 10% mountain laurel. The other forests are in region III where mountain laurel was abundant and all had more than 30% of the plots with mountain laurel. In many parts of Leo- minster state forest, where mountain laurel was the most abundant we encountered, the mountain laurel formed almost impenetrable thickets covering many hectares. DISCUSSION Mountain laurel grows naturally in most of Massachusetts, with three exceptions. We did not observe it on Cape Cod or eastern Plymouth County, in the high portions of the northern Berkshires, or in northwestern Berkshire county and the Taconic range. It is generally assumed that mountain laurel is absent from Cape Cod due to the past history of severe fires. In Rhode Island, fire elimi- nated large mountain laurel and greatly reduced the relative density of small ones (Brown, 1960). We did observe mountain laurel grow- ing on a site similar to Cape Cod, the Montague sand plain, where fires are common, so additional maritime factors may be important on Cape Cod. Mountain laurel presumably cannot grow in the high Berkshires because of the cold winters. It does occur throughout most of the Berkshires at lower elevations. Limitation by the cold winters is also consistent with the fact that the northern limit for the species is in southern Vermont and New Hampshire. The lack of mountain laurel in northwestern Berkshire county is probably related to the general decreased acidity of the soil (USDA, 1973). Mountain laurel generally prefers acid soils (Braun, 1961). 1983] Wilson & O’Keefe — Mountain Laurel 121 Within those areas of Massachusetts where mountain laurel grows it can be found on almost any site, from deep sands, to swamps, to rocky hillsides. Yet, despite its adaptability, it is uncommon both east of Worcester and in a strip through western Worcester county and south of the Quabbin reservoir. In both areas where it is generally uncommon it can still be locally abundant. The major factor that probably makes mountain laurel generally uncommon in some areas is land use history. If land has been cleared of forest, mountain laurel is extremely slow to re-invade when the land is abandoned (O’Keefe, 1981). Mountain laurel has stringent seed bed requirements, usually moss, and the seeds do not spread far from the parent plant (Kurmes, 1961). As a result, moun- tain laurel is usually found either in or near areas that were never cleared of forest during the height of agriculture in the 1800’s (O’Keefe, 1981). Consequently, mountain laurel can provide a clue to past land use, but it is not, by itself, an adequate indicator of non-cleared land. For the same reason, mountain laurel may be abundant locally in swamps or on rocky slopes that, because they were never cleared, served as refugia for the mountain laurel. The eastern part of the state is, and has been, highly populated and relatively flat so most of the land has probably been cleared and mountain laurel distribution is severely restricted. The strip in cen- tral Massachusetts where mountain laurel is scarce is not highly populated, but it was extensively cleared in the 1800’s (Raup & Carlson, 1943). In the area of the Brookfields, south and east of the Quabbin rerservoir, there are still a large number of farms on top of the rolling hills. The present distribution of mountain laurel at the Harvard Forest coincided almost exactly with land that was never cleared in the past (O’Keefe, 1981). There may be other factors limiting mountain laurel distribution in this central portion of the state, but land use history is probably the major factor. It is difficult to determine the overall bias introduced by the map- ping techniques used in this study. A survey from automobiles gives a sample that is potentially biased in several ways: small plants cannot be seen from moving automobiles, the edges of roads have relatively high light intensities that favor the growth of mountain laurel, road banks may create favorable seeds beds for mountain laurel establishment, roads tend to pass through populated areas with more intensive land use history, roads tend to be in valleys rather than on hill tops. Interstate highways avoid some of these 122 Rhodora [Vol. 85 biases, but there are not enough of them and it is unsafe to drive slowly. Mountain laurel undoubtedly grows in some areas where we saw none from the roads. The results of mapping in this study agree quite well with the results of studies using plots, so we assume that the bias introduced by the mapping technique is relatively un- important. ACKNOWLEDGMENTS We thank Mark Dale, Anne Hine, Gary Kronrad, Karen Saun- ders, and Mary Alice Wilson for assistance in mapping. Supported by Massachusetts Agricultural Experiment Station Grant Mclntire- Stennis 32. LITERATURE CITED Braun, E. L. 1961. The Woody Plants of Ohio. Ohio State Univ. Press, Columbus. Brown, J. H. 1960. The role of fire in altering the species composition of forests in Rhode Island. Ecology 41: 310-316. Hipps,D.E. 1978. The life history and strategy of striped maple (Acer pensy/vani- cum). Ph.D. Thesis, University of Massachusetts, Amherst. EpinGer, J. E. 1974. A systematic study of the genus Kalmia. Rhodora 76: 315-398. Kurmes, E. A. 1961. The ecology of mountain laurel in southern New England. Ph.D. Thesis, Yale University, New Haven, Connecticut. 1967. The distribution of Kalmia latifolia L. Amer. Midl. Nat. 77: 525-527. O’KeeFe, J. F. 1981. Factors affecting the distribution and growth of mountain laurel (Kalmia latifolia L.) in central Massachusetts. MS Thesis, University of Massachusetts, Amherst. Raup, H. M. & R. E. Carson. 1941. The history of land use in the Harvard Forest. Harvard Forest Bull. 20. USDA. 1973. General soils report, Berkshire County. USDA Soil Cons. Ser. with Berkshire-Franklin Resource Dev. Project. DEPT. OF FORESTRY AND WILDLIFE MANAGEMENT UNIVERSITY OF MASSACHUSETTS AMHERST, MA 01003 NEW ENGLAND NOTE TRIPHORA TRIANTHOPHORA IN MASSACHUSETTS AND VERMONT PETER F. ZIKA Triphora trianthophora (Swartz) Rydberg is a small and incon- spicuous orchid that flowers in mid to late August and fruits in September in southern Vermont. Triphora reaches the northern limit of its range in New England. Crow et al. (1981) classify Tri- phora as rare in New England but not threatened or endangered. Coddington & Field (1978) included the species on the Massachu- setts list of rare and endangered plants. There is one historical Massachusetts station, from Conway, Franklin County, where the species was last collected in 1928 (Allis s.n., AMES). Triphora is also on the Vermont rare and endangered list (Countryman, 1978). All Vermont records are from Windham Co., the southeast corner of the state, where Triphora was last collected in 1903, in Dum- merston, (Howe s.n., VT). Vermont also has documented historical stations from Brattleboro and Newfane. Rugg (1950) reported the Brattleboro station was last seen circa 1925 and described a station he discovered in 1943 in Brookline, Vermont. Photographs of plants from each of the four known extant Tri- Phora populations in Massachusetts and Vermont were made in 1981 and are on deposit in the Pringle Herbarium (VT). MASSACHUSETTS: FRANKLIN COUNTY Greenfield. Station discovered by the author in 1981. Two plants, one in flower on 19 August. VERMONT: WINDHAM COUNTY Brattleboro. Roberta Poland and Herman Willey provided directions to find this colony. Seventeen plants total, four- teen flowering on 18 August, three fruiting on 20 Sep- tember. Brookline. This is Rugg’s 1943 station. Roberta Poland guided the author to the site. Thirteen plants, seven flower- ing on 18 August. Dummerston. Station discovered by Roberta Poland in 1981. Eleven plants, ten flowering on 18 August, six fruiting on 19 September. 123 124 Rhodora [Vol. 85 At each Triphora station the habitat was strikingly similar: an old-age or maturing forest, dominated by 7Tsuga canadensis (L.) Carriere and Fagus grandifolia Ehrhart. At the Massachusetts sta- tion the beeches were the largest trees, averaging 9-inch DBH. In Brookline the largest trees were hemlocks, averaging more than 14-inch DBH. Other prominent woody species in the woodlots were Acer saccharum Marshall, Acer rubrum L., and Ostrya virginiana (Miller) K. Koch. Under the dense shade of the late summer canopy few competing herbs were present. I observed widely scattered plants of Corallorhiza maculata Rafinesque, Maianthemum cana- dense Desfontaines, Monotropa uniflora L., and Epifagus virgini- ana (L.) Barton. Most of the Triphora were restricted to leaf-lined depressions on gentle slopes. Triphora’s hemlock and beech habitat is regenerating in western Massachusetts and southern Vermont, and this tiny orchid is likely to be found at several new stations if a search is made for it. The four sites I visited are on private land, and are threatened by road widening, logging, or development. My sincere thanks to Roberta Poland and Herman Willey for generously sharing their knowledge of Vermont orchids with me, and to Bruce Sorrie for providing the Massachusetts herbarium records. Funding for field work was supplied by the Pringle Her- barium, the Vermont chapter of The Nature Conservancy, and a grant from the Andrew Mellon Foundation. LITERATURE CITED CoDDINGTON, J. & K. G. Fietp. 1978. Rare and Endangered Vascular Plant Species in Massachusetts. The New England Botanical Club in cooperation with the U.S. Fish & Wildlife Service, Newton Corner, MA. 52 pp. COUNTRYMAN, W. D. 1978. Rare and Endangered Vascular Plant Species in Vermont. The New England Botanical Club in cooperation with the U.S. Fish and Wildlife Service, Newton Corner, MA. 68 pp. Crow, G. E., W. D. CountrYMAN, G. L. Cuurcu, L. M. EASTMAN, C. B. HELL- Quist, L. L. MEHRHOFF, & I. M. StorKs. 1981. Rare and endangered vascu- lar plant species in New England. Rhodora 83: 259-299. RuGG, H. G. 1950. Rare orchids in Vermont. Joint Bulletin of the Vermont Botanical and Bird Clubs 18: 51-55. PRINGLE HERBARIUM UNIVERSITY OF VERMONT BURLINGTON, VT 05405 LITERATURE FOR NEW ENGLAND BOTANISTS HELLQUIST, C. B. & G. E. Crow. 1980. Aquatic Vascular Plants of New England: Part 1. Zosteraceae, Potamogetonaceae, Zanni- chelliaceae, Najadaceae. New Hampshire Agricultural Experi- ment Station Bulletin 515. Durham, N.H. 03824. Crow, G. E. & C. B. HELLQuistT. 1981. Aquatic Vascular Plants of New England: Part 2. Typhaceae and Sparganiaceae. New Hampshire Agricultural Experiment Station Bulletin 517. Dur- ham, N.H. 03824. HELLQulisT, C. B. & G. E. Crow. 1981. Aquatic Vascular Plants of New England: Part 3. Alismataceae. New Hampshire Agricul- tural Experiment Station Bulletin 518. Durham, N.H. 03824. — Crow, G. E. & C. B. HELLQuist. 1982. Aquatic Vascular Plants of New England: Part 4. Juncaginaceae, Scheuchzeriaceae, Buto- maceae, Hydrocharitaceae. New Hampshire Agricultural Exper- iment Station Bulletin 520. Durham, N.H. 03824. HeELLQuist, C. B. & G. E. Crow. 1982. Aquatic Vascular Plants of New England: Part 5. Araceae, Lemnaceae, Xyricaceae, Eri- ocaulaceae, and Pontederiaceae. New Hampshire Agricultural Experiment Station Bulletin 523. Durham, N.H. 03824. This is a continuing series of reports on the aquatic and wetland plants of New England. There are technical keys to the species; habitat information, alkalinity and pH values and ranges. Detailed line drawings, and dot distribution maps for each species are valua- ble features of the series. MARY M. WALKER LIBRARIAN NEW ENGLAND WILD FLOWER SOCIETY FRAMINGHAM, MA 01701 125 NOTICES OF PUBLICATION BATSON, WADE T. 1982. Genera of the Western Plants. 207 pp. Wade T. Batson, 1120 Blake Drive, Cayce, S.C. 29033. (Price $8.50) The author provides dichotomous keys to the 1669 genera of western plants in this compact guide. The unique feature of the keys is that they are illustrated with line drawings showing important diagnostic features of each genus. Only the small size of drawings detracts from the usefulness of this feature. Of more interest to Rhodora readers may be the author’s earlier book: BATSON, WADE T. 1977. Genera of the Eastern Plants. John Wiley, New York, N.Y. (Price $6.95) SEYMOUR, FRANK CONKLING. 1982. The Flora of New England. 2nd. ed. xvii + 611 pp. Phytologia Memoirs V, H. N. Mol- denke, 303 Parkside Rd., Plainfield, NJ, 07060. (Available from the author, Frank C. Seymour, 264 Hixville Rd., N. Dart- mouth, MA, 02747 for $21.32, includes postage. New England botanists especially will welcome the fact that The Flora of New England, for sometime out of print, is again available. The second edition contains line drawings by Andrea Robbins, instead of photographs, and there are seven pages of additions and corrections. Vol. 84, No. 840, including pages 453-559, was issued November 26, 1982. 126 Joint Meeting: June 20, 21, & 22, 1983 The annual joint Field Meeting of The Northeastern Section of the Botanical Society of America, The Torrey Botanical Club, and The Philadelphia Botanical Club will be held at Aurora, New York, with Cornell University and Wells College as host organizations. There will be accommodations at Wells College; primary field leadership will be supplied by Cor- nell University. Tentative plans include: half-day trips to Great Gully, Zurich Mud Pond, Clark’s Reservation, Oneida Lake, Cornell Plantations, Taughannock Gorge, etc. Some possible optional trips for the 23rd include Geneva Experiment Station, an illustration of old-field suc- cession, Sapsucker Woods Bird Sanctuary, and a local winery. Full details will be available after 1 February 1983 by writing the chairman: Dr. Charles Burch Wells College Aurora, NY 13026 (Telephone: 315 364-7726) RHODORA January 1983 Vol. 85, No. 841 CONTENTS The taxonomy of Vaccinium § Oxycoccus S. P. Vander Kloet Kalmia ericoides revisited Walter S. Judd The national historical distribution of Platanthera peramoena (A. Gray) A. Gray (Orchidaceae) and its status in Ohio David M. Spooner and John Steven Shelly Wind dispersal of some North American species of Andropogon (Gramineae) Christopher S. Campbell On the taxonomic status of Lophiola aurea Ker-Gawler Michael Zavada, Xue-Lin Xu, and J. M. Edwards The fern genera Vittaria and Trichomanes in the northeastern United States Donald R. Farrar, James C. Parks, and Bruce W. McAlpin Adiantum pedatum ssp. calderi, a new subspecies in northeastern North America William J. Cody : : F ; ; : . Contributions to the reproductive biology of Panax trifolium L. (Araliaceae) C. Thomas Philbrick Mountain Laurel (Kalmia latifolia L.) distribution in Massachusetts Brayton F. Wilson and John F. O'Keefe NEW ENGLAND NOTES Triphora trianthophora in Massachusetts and Vermont Peter F. Zika ; Literature for New England Botanists ; Notices of Publication: Genera of the Western Plants; The Flora of New England, 2nd. edition 45 55 65 73 83 93 97 115 123 125 126 Announcement of Joint Meeting . . . . . . . , cover III JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB No. 842 April 1983 Vol. 85 Che New England Ratanical Club, Ine. Botanical Museum, Oxford Street, Cambridge, Massachusetts 02138 Conducted and published for the Club, by NORTON H. NICKERSON, Editor-in-Chief Associate Editors A. LINN BOGLE GARRETT E. CROW WILLIAM D. COUNTRYMAN RICHARD A. FRALICK GERALD J. GASTONY NORTON G. MILLER ROBERT T. WILCE RHODORA.—Published four times a year, in January, April, July, and October. A quarterly journal of botany, devoted primarily to the flora of North America. Price $20.00 per year, net, postpaid, in funds payable at par in the United States currency at Boston. Some back volumes, and single copies are available. For information and prices write RHODORA at address given below. Subscriptions and orders for back issues (making all remittances payable to RHODORA) should be sent to RHODORA, Botanical Museum, Oxford Street, Cambridge, Mass. 02138. In order to receive the next number of RHODORA, changes of address must be received prior to the first day of January, April, July or October. Scientific papers and notes relating to the plants of North America and floristically related areas will be considered by the editorial committee for publication. Articles concerned with systematic botany and cytotaxon- omy in their broader implications are equally acceptable. Brevity is urged whenever possible in all papers. Short items will be published on otherwise blank end pages as soon as possible, even if they appear ahead of longer articles already accepted. All manuscripts should be submitted in TRIPLICATE AND MUST BE DOUBLE (AT LEAST 3/8 OF AN INCH) OR TRIPLE- SPACED THROUGHOUT. Please conform to the style of recent issues of the journal. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to: Russell R. Walton Managing Editor, RHODORA Harvard University Herbaria Building 22 Divinity Avenue Cambridge, Mass. 02138 Second Class Postage Paid at Boston, Mass. PRINTED BY THE LEXINGTON PRESS, INC. LEXINGTON, MASSACHUSETTS Cover illustration Trollius laxus Salisb. is very rare as it nears its northeastern limit in Connecticut. Some old records indicate that it has grown in parts of New Hampshire and Maine, so it may yet be found in appropriate habitat. Original artwork by Tess Feltes, Illustrator. Tbhodora (ISSN 0035-4902) JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 85 April 1983 No. 842 THE GENUS ENCYCLIA HOOK. (ORCHIDACEAE) IN THE BAHAMA ARCHIPELAGO RUBEN P. SAULEDA AND RALPH M. ADAMS The neotropical genus Encyclia is represented by 13 species, 2 varieties, and 2 natural hybrids in the Bahama Islands, including the Turks and Caicos Islands. As such, it is the largest orchid genus, comprising more than 27 percent of all the orchid species distributed in the Bahama archipelago. For a complete inventory of the 52 orchid taxa that occur in the Bahama Archipelago, see Correll and Correll (1982). Of the 17 taxa of Encyclia in the Bahama archi- pelago, 7 are endemic, 3 share distributions only with Florida, 3 share distributions only with Cuba, and | shares its distribution only with Hispaniola. The remaining 3 taxa are more widely distributed (Table I). Adams, et al., (1982) described 9 distinct and recognizable habitats that occur on the islands of the Bahama Archipelago. Species of the genus Encyclia occur in all of these habitats except pine forests and saltwater marsh and tidal flats. While the majority of the taxa in the genus occupy several habitats, a definite habitat preference exists, with all 17 taxa occurring either in high coppices, low coastal coppices, or both (Table II). This paper represents the first modern and comprehensive taxo- nomic and ecological treatment of the genus in the archipelago. For the purposes of this floristic study, we have chosen to use the broader generic concept as proposed by Dressler (1961). This concept differs from the original concept of Encyclia as proposed by Hooker (1828) and later expanded by Schlechter (1915). The 127 128 Rhodora [Vol. 85 Dressler (1961) concept of Encyclia includes two species, Encyclia boothiana (Lind1.) Dressler and Encyclia cochleata (L.) Lemee, that do not correspond to the Hooker (1828) and Schlechter (1915) generic concept. TAXONOMIC TREATMENT Encyclia Hook., Bot. Mag. 55: t2831. 1828. Epiphytic or epilithic, rhizomatous plants. Stem pseudobulbous having one to many leaves at apex. Leaves entire, thinly coriaceous to rigid or semi-terete, slender and elongated. Inflorescence ter- minal, a simple raceme or panicle, often with an elongated peduncle, occasionally originating within a slender, conduplicate, foliaceous sheath. Flowers showy or inconspicuous, variously arranged, often non-resupinate. Sepals and petals free, spreading, or reflexed. Labellum free or variously adnate to column, entire or 3-lobed, disc variously callose. Column fleshy, without basal foot, cylindric, occasionally apically 3-to S5-dentate, commonly with auricles. An- ther terminal, operculate, incumbent, 2-celled. Pollinia 4 (8 in triandrous varieties), equal, compressed, waxy, connected by a viscidium, without a stipe or gland. Stigmata 2, confluent, trans- verse. Capsule ovoid to ellipsoid, with prominent ribs or broadly 3-winged. Type: Encyclia viridiflora Hook. KEY TO SPECIES IN THE BAHAMA ARCHIPELAGO 1. Inflorescence basally enclosed by scarious slender conduplicate POUMCOOUS ARGON: csc. vncakne views bas 5 eene recess ees 2 2. FLOWOIS PORUDINNIG 6000s 656 4G 35 285 oe een ee 88s 3 3. Column with one anther...1. E. boothiana var. boothiana 38, AON With THTEC QNINENS 6655656 5544au sins so ween dens 2, FIGWErs HON-TORU DINAN y o5ch sc kes eereyeenenen denn ens 4 4. Column with one anther....4. E. cochleata var. cochleata 4a. Column with three anthers ........... 00. e cece cece eee Pe nr rere ere eee een 5. E. cochleata var. triandra la. Inflorescence with basal foliaceous sheath absent........... 5 Table 1. The distribution of the species of Encyclia within and between islands of the Bahama Archipelago. The islands are listed, from left to right, by decreasing area. The abbreviations are: C, Cuba; CA, Central America and/or Mexico; F, Florida: H, Hispaniola; J, Jamaica; SA, South America 8 =) a zs > < os -E a: S eo ey ee £35 S2_8,28he8s3 Bese geds a ee 0 ws MP GYUNOSZZFOLS HBPA NRT HRCOVOTCE one voce unwox~w -a ene. eS ee Sc eS one SSSSPELAESABIT SR cSSEBSSERSES Additional SS ees SESS SESSESZZSESSSESSES Distributions Encyclia boothiana var. boothiana : L | C,H,CA Encyclia boothiana var. erythronioides | * | * F Encyclia caicensis _ . ° | d : | || Endemic Encyclia cochleata var. cochleata Ue sib Oe . | C,H,J,CA,SA Encyclia cochleata var. triandra : | | _| tid F Encyclia fehlingii ele] je *| | a Naa | Endemic Encyclia fucata ele) is | °| C Encyclia gracilis eleis ele ele] [ef [ele] | . . Endemic Encyclia hodgeana eleleie ° elelejlelele|l/elelele eel] re} | tt Re BS H Encyclia inaguensis “| | [+ fefef | [tT Endemic Encyclia plicata poled. ° C Encyclia rufa eleite i e el/el;e a) EI eal i ec de Hd elje elele Cc Encyclia selligera . | | Ltt || | | | | CA Encyclia tampensis refel | | LSS RENA:: Encyclia withneri . | | a _| Endemic Encyclia x bajamarensis — fel | [ Endemic Frncyclia x lucayana efe| | Endemic osvjadiyjy eweyeg Ul oyodouq — swepy x epeines = [e861 671 130 Rhodora [Vol. 85 5. Leaves non-deciduous, abscission layer at leaf base absent... CnTAREEAS EE ESO LANE EREL UE TORE OECTA eee aa eS 6 6. Labellum yellow with radiating purple lines; flowers May to IO dae eavaa sci eeenreerens 16. E. Xbajamarensis 6a. Labellum white with radiating purple lines; flowers July THIOURN SODICMDE! * ? a “., - ‘ gt, F a o* ~ ’ — ~*., ges % |§ / . =f * 34 . ~~ . 4s 4 * we, “es * 7 * —- Ni # ® # 4 : P _ on teen, Pl ; e ~ ~ 4 : Figures 12 and 13. Rhizoclonium tortuosum. Fig. 12, habit of filaments (Scale = 150 um). Fig. 13, reproductive cell with anisogametes (Scale = 10 ym). 1983] Blair — Chaetomorpha & Rhizoclonium 201 Taxonomic History The species was originally described by Dillwyn (1806) as Con- ferva tortuosa. No dimensional characters are given in the descrip- tion, other than the statement that “the filaments are fine as hair of the human head”, and that the cells are nearly twice as long as broad. A type specimen was not designated and the published draw- ing shows no scale. Chapman (1939) was unable to find a type specimen but succeeded in finding a specimen identified by Dillwyn as Conferva tortuosa. The filaments were 34 to 48 um diam. with a mean of 40 um, and were | to 2 times as long as broad. Kitzing (1845) transferred Conferva tortuosa Dillwyn to Rhizoclonium tor- tuosum, stating that its diameter was 1/70’ (’’’ = a ligne or 2116 pm) or 31 wm wide and | to 1.5 times as long as broad. Confusion arose between Rhizoclonium tortuosum and Kiitzing’s (1849) Chaetomorpha tortuosa. The confusion was examined by Chapman (1939) and will only be reviewed here. To examine this confusion we must look back to the origins of C. tortuosa and Rhizoclonium capillare (Kiitzing, 1847). The species R. capillare was characterized by cells measuring |/45’’’ (47 um) and | to 2 times longer than broad. Later Kiitzing (1849) described Chaetomorpha fortuosa stating it was characterized by cells 1/45’ to 1/40’”’ (47 to 57 um) in diameter, rigid, curled, tortuous, and | to 2 times as long as broad. Rhizoclonium capillare Kiitzing was placed as a synonym of Chaetomorpha tortuosa. The basionym cited by Kiitzing for both Chaetomorpha tortuosa and Rhizoclonium capillare is Conferva tortuosa J. Agardh, which is based on Conferva tortuosa C. Agardh (J. Agardh, 1842). Conferva tortuosa C. Agardh is in turn based on Conferva tortuosa Dillwyn (C. Agardh, 1824). However, Conferva tortuosa Dillwyn is the basionym for Rhizoclonium tortuosum (Dill- wyn) Kiitzing (1845). Thus, there had been an inadvertent establish- ment of two species on a single basionym. Therefore, either the two species are united under one name (brought into synonymy), or if two species are separate entities, the invalid name is discarded and the second species is renamed.: An examination of the original de- scription for R. capillare and R. tortuosum indicates that Kitzing’s original intent was to describe two separate taxa. The cells of Rhizo- clonium tortuosum were listed as 1/70’’’ (31 um) diameter and 1.5 times longer than broad (Kiitzing, 1845), while R. capillare cells were described as being 1/45’ (47 um) diameter and | to 2 times 202 Rhodora [Vol. 85 longer than broad (Kiitzing, 1847). Borgesen (1925) noted Kiitzing’s intent to describe two separate species, and the incorrect basionym declaration. Accordingly, he made the combination Chaetomorpha capillare (Kitzing) Bgrgesen citing R. capillare as the basionym. The continued confusion as to the proper naming and dimensional characteristics of C. capillare is evident in a variety of recent studies (Kornmann & Sahling, 1977; Christensen, 1975; Price, 1967) which cite the taxon as either C. tortuosa or C. capillare. However, recent work (Blair, 1978) has shown Chaetomorpha capillare (Kiitzing) Bérgesen and Rhizoclonium tortuosum to be conspecific on the basis of morphological continuity and indistinguishable habit. Thus, C. capillare (Kitz.) Borg. and R. capillare Kiitz. are syn- onyms of R. tortuosum (Dillw.) Kitz. Foslie (in Wittrock & Norstedt, 1877-1887) described Rhizo- clonium riparium (Roth) Harvey f. validum Foslie stating that the form was wider and had a greater length than R. riparium, being 26 to 36 wm wide, 0.25 to 2.33 times longer than broad, and without rhizoids (Koster, 1955). Rosenvinge (1893) elevated the form to varietal status, increased the width range to 30-50 um, and indi- cated that rhizoids could be numerous. Some authors (Stockmayer, 1898; Chapman, 1939) have reduced R. tortuosum to synonymy under R. riparium f. validum, while Koster (1955) preferred to syn- onymize R. riparium f. validum under R. tortuosum, which would be the correct synonymy on the basis of priority of publication. Therefore, R. riparium f. validum was also placed as a synonym of R. tortuosum (Koster, 1955). Hamel (1929) showed that Rhizoclonium lubricum Setchell et Gardner exhibited anisogamous reproduction. The character 1s unique within the Cladophoraceae and led to the establishment of Lola Hamel for unbranched cladophoralean algae with anisog- amous reproduction. Perrot (1965) later showed that Rhizoclonium implexum Harvey (= R. tortuosum (Dillwyn) Kiitzing; see Chap- man, 1939) has anisogamous reproduction. Accordingly he trans- ferred R. implexum to Lola. Presently the species is referred to as either Rhizoclonium tortuosum (Dillwyn) Kiitzing or Lola tortuosa (Dillwyn) Perrot. It should be noted that the establishment of a genus on a single characteristic (i.e., anisogamous reproduction) despite other similarities with the Rhizoclonium sp. is tenuous. The use of anisogamous reproduction as a generic character is further 1983] Blair — Chaetomorpha & Rhizoclonium 203 questionable in light of the fact that other green algal genera, (i.e., Sphaeroplea, Chlamydomonas) contain species that range from isog- amous to anisogamous. The present author, therefore, favors the retention of Rhizoclonium tortuosum in the genus Rhizoclonium. Rhizoclonium riparium (Roth) Harvey. 1846. Phyco. Brit. IV p. 1238 Basionym: Conferva riparia Roth. 1797. Catal. Bot. Fasc. 1. Type: Conferva riparia Roth leg. Mertens. 1803. Norder- ney, Germany ex Herv. Hooker (K). Type Locality: Morderney, Germany. Rhizoclonium jurgensii (Mertens) Kiitzing. 1843. Phyco. Gen. p. 261. R. lacustre Kiitzing 1847, Diag. Verner. odea Dritis. Alg. Bot Ziet. R. implexum (Dillwyn) Kitzing. 1845. Phyco. Germ. p. 206. Conferva implexa Dillwyn. 1806. Brit. Conf. p. 46 t. b. Rhizoclonium interruptum Kitzing. 1849. Sps. Alg. p. 384; 1853. Tab. Phyc. II] a Ae 2 R. kerneri Stockmayer. 1898. Uber die Algen. p. 583. R. kochianum Kitzing. 1845. Phyco. Germ. p. 206. The species is found in high tide pools and marshes on the open coast as well as within high marsh communities in estuaries. The filaments are lax and entangled, with a cell diameter of 8 to 45 um diameter. The LWRs vary between | and 4(< 6). Rhizoidal prolifer- ations are sparse to abundant in New England material, and consist of single or multiple cells. The plant has never been found attached (basally) in New England although it shows an attached stage in culture (Neinhuis, 1975). Rhizoclonium riparium exhibits an iso- morphic alternation of generations with isogamous gametes and bi- and quadriflagellate zoospores (Neinhuis, 1975). Distribution and Ecology Neinhuis (1975) and Koster (1955) have given extensive summar- ies of the ecology and distribution of the species. Accordingly only an overview Is given and reference 1s made to the above articles for a variety of details on the plant’s distribution and ecology. Rhizo- clonium riparium has a world-wide distribution, being recorded from the Netherlands, Europe and Japan. It is extremely euryhaline, being found in a broad range of estuarine and coastal habitats. Within the Great Bay estuary system R. riparium is usually found entangled among the bases of Spartina alterniflora on muddy sub- 204 Rhodora [Vol. 85 strata within the littoral zone, also reaching the headwaters of many riverine tributaries. Some samples of R. riparium may reach the dimensions of R. tortuosum, from which it may be separated by its broader estuarine distribution and differential reproductive mor- phology, i.e., isogamous for R. riparium and anisogamous for R. tortuosum. Representative specimens: Canada. NEW BRUNSWICK: Charlottes Co., Campobello Island, Hehre & Conway CI911 (NHA); South Wolf, Conway et al. W-77 (NHA). NEWFOUNDLAND: Bonne Bay, Hooper & Williams 7977 (NFLD), Bonavista Bay, Mathieson NF202 (NHA); Concep- tion Bay, South et al. 4509 (NHA). Tahiti. Society Islands, Setchell & Parks s.n., July 1922 (uc), Hoenneck & Meerlagen 478 (UC). United States. DELAWARE: Lewes, Mathieson et al. s.n., 14 August 1966 (NHA). MAINE: Aroostook Co., Eagle Island, Collins 2116 (NHA); Hancock Co., Camden, Hooper 95] (NHA); Washington Co., Lubec, Hehre 3/82 (NHA); York Co., Ogunquit, Searles s.n., 7 April 1966 (NHA); Eliot, Searles s.n., 30 April 1966 (NHA). NEW HAMP- SHIRE: Rockingham Co., Greenland, Mathieson & Hehre s.n., 14 September 1966 (NHA), Hehre et al. s.n., 8 July 1966 (NHA); Newington, Mathieson s.n., 9 September 1977 (NHA); Hampton, Mathieson s.n., 11 September 1969 (NHA); Portsmouth, Croasdale s.n., 28 July 1938 (NHA), Blair s.n., 17 August 1977 (NHA); Newcastle, Mathieson s.n., 20 April 1967 (NHA); Strafford Co., Durham, Mathieson s.n., 20 June 1966 (NHA); Dover, Hehre & Shipman s.n., 24 June 1966 (NHA); Durham, Mathieson s.n., 27 July 1977 (NHA); Newington, Reynolds NB38/ (NHA). Taxonomic History This species was first described by Roth (1793) as Conferva ripa- ria based upon its thin, twisted, bifurcating habit (“apice tantum divisa et pierumque bifia”’). No dimensional characters were given, other than that the width was one half the length, (“diametro sesqui- longioribus”). In addition, no type specimen nor drawing was indi- cated. Koster (1955) found a specimen of Rhizoclonium riparium collected by Mertens from the type locality of Norderney, Germany. She assumed that it was part of the type collection. The specimen contained a mixture of Rhizoclonium riparium (Roth) Harvey and Caldophora fracta (Dillwyn) Kiitzing f. haukii (Bérgesen) Slootweg. Koster felt the Cladophora fraction of this specimen pertained to Roth’s description of branching. Harvey (1845-51), when trans- ferring the species from Conferva to Rhizoclonium, selected only the R. riparium element from the mixture (Koster, 1955). Accord- ingly, Koster proposed that the specimen from Norderney, Ger- many, in the Kew Herbarium should be designated as a lectotype. 1983] Blair — Chaetomorpha & Rhizoclonium 205 The filaments are 18 to 30 wm wide and cells are 1.5 to 2.5 times longer than broad (Koster, 1955). Specimens that have been identified as Rhizoclonium riparium are commonly found, but the degree of rhizoidal proliferation is variable. As a result a large number of forms and varieties have been described depending upon the presence, absence and morphology of rhizoids. For example, R. riparium (Roth) Harvey var. polyrhizum (Lyngbye) Rosenvinge is described as having many rhizoids with one to few cells, while R. riparium (Roth) Harvey var. implexum (Dillwyn) Rosenvinge is recorded to have few or no rhizoids. Koster (1955) proposed the distinction of “status radicans” and “status arrhizum” for those specimens with a variable occurrence of rhi- zoids with the belief that rhizoidal proliferations were influenced by environmental factors. Nienhuis (1975) recorded extreme variability of filament and rhi- zoidal morphology for Rhizoclonium riparium populations in the Netherlands. After extensive field and culture studies, he found that filament diameter and the degree and size of rhizoidal proliferations were influenced by tidal elevations, firmness of substrates, and salin- ity. The rhizoidal variability he observed encompassed the mor- phologies recorded for R. implexum (Dillwyn) Kiitzing (Kiitzing, 1849), R. hieroglyphicum (C. Agardh) Kitzing (Kiitzing, 1849) and R. riparium (Roth) Harvey. As a result he synonymized R. im- plexum and R. hieroglyphicum with R. riparium (Roth) Harvey. DISCUSSION As reflected in the Taxonomic History sections of this paper, a great deal of confusion concerning specific delineation has existed within these genera. While the present report addresses some of the questions that have arisen, many more remain unanswered. For example, difficulties may arise in discerning non-reproductive Rhizo- clonium tortuosum and R. riparium within the region of dimen- sional overlap (24 to 45 um). Habitat information may aid in their determination as R. tortuosum is found in the low littoral to upper sublittoral while R. riparium is found in areas of extreme environ- mental fluxes, such as marshes and mudflats. The differential reproductive phenology, however, confirms the need for specific separation of the two species (Rhizoclonium tortuosum showing 206 Rhodora [Vol. 85 anisogamous reproduction while R. riparium reproduces isog- amously). Within the genus Chaetomorpha, questions arose as to the taxo- nomic validity of Chaetomorpha cannabina (Aresch.) Kjell. as no specimens examined (field or herbarium) fit the dimensions cited for the species. In general, the species is considered to vary in diameter from 75 to 100 zm with cell lengths being 3 to 8 diameters (500 to 600 xm) long (Taylor, 1957). The dimensions credited to this species have varied considerably from author to author (Table 1). Are- schoug (1843) described the species Conferva cannabina, and dis- tributed specimens of the species in the Algae Scandinavicea Exisccatae (1840). Areschoug (1843) referenced a previously pub- lished species, Conferva auricoma Suhr (Suhr, 1840) stating that through communication with Suhr, it was believed that the two plants were the same. If in fact the two species are the same then the correct designation of the taxon should be Chaetomorpha auri- coma. However, questions still exist as to the conspecificity of Suhr’s Conferva auricoma and Areschoug’s Conferva cannabina. At pres- ent, specimens of C. auricoma Suhr have not been examined. Spec- imens of Conferva cannabina Areschoug, distributed in Alg. Scand. Exiscc. (1840; #14) were examined through the courtesy of the Riks- museet (Ss) and Rijksherbarium (L) and represent a remotely branched member of the Cladophoraceae. However, until type material of Conferva auricoma Suhr can be examined, the proper generic placement and specific classification of the two species must be reserved. It should be noted that the herbarium specimens exam- ined from various institutions identified as Chaetomorpha can- nabina were referable to either Chaetomorpha brachygona or Rhizoclonium tortuosum, and none fit the description attributed to C. cannabina. Concern about possible conspecific relationships existing between some of the more narrow Chaetomorpha and Rhizoclonium species have previously been noted (p. 197). Finally, with regard to the separation of the genera Chaeto- morpha and Rhizoclonium, a review of the characters used to separate the two genera shows none of the characters to be valid criteria. The names Rhizoclonium and Chaetomorpha bring forth specific thoughts of shape and form for many taxonomists. How- ever, the characters that have been used to separate the genera have been shown to be erroneous criteria or based on environmentally 1983] Blair — Chaetomorpha & Rhizoclonium 207 variable characteristics. The genus Rhizoclonium was originally described with a single sentence “T7richomata parenchymatica, coelogonimica, ramus verticales, radicoutes emittentis” (Kitzing, 1843). The description of Chaetomorpha (Kitzing, 1845) was not much more elaborate but did point out the unbranched nature of the plant, and that it was a uniseriate filament with laminated cells that were as long as broad or longer, but less than 4 times the width. Previous authors have employed various characters to differenti- ate the genera: cell shape (Collins, 1909, Setchell & Gardner, 1920), mode of attachment (Bold & Wynne, 1978; Abbott & Hollenberg, 1976; Taylor, 1957; Setchell & Gardner, 1920; Collins, 1909), and number of nuclei per cell (Setchell & Gardner, 1920). The most common criterion for separation of the genera has been mode of attachment, with basal holdfasts in Chaetomorpha and rhizoids in Rhizoclonium. However, past studies have shown Rhizoclonium species to have basal holdfast stages at some point in their life history (R. tortuosum: Perrot, 1965; R. riparium (= R. implexum) (Dillw.) Kit. ; Nienhuis, 1975). The presence or absence of rhizoids has also been used as a diagnostic character for delineation of Rhizo- clonium. Nienhuis (1975) and Patel (1971b) have shown rhizoidal proliferation and complexity to be affected by light intensity, sub- stratum firmness, light quality, and temperature. The demonstrated variability of this character, and the fact that the New England populations of R. rortuosum often lack rhizoids, makes this charac- ter of little use in generic separation. Chapman (1939) cited the width of the filament as the distinguish- ing character between the species. The original description of the genera, however, gave no dimensional characteristics. Further, a continuum of cell widths is present between Chaetomorpha and Rhizoclonium (Table 2). Other criteria, such as pyrenoid number and number of nuclei per cell, have been shown to vary with the cell volume and are not specific or generic indicators in the unbranched Cladophoraceae (Prasad, et al., 1973; Price, 1967; Carter, 1919). It appears then, that the question must be raised as to the con- tinued separation of the two genera, at least on the basis of tradi- tional characteristics. Further information reflecting the close relationship between these two groups can be seen in the chromo- some numbers of the species within the genera (Table 3). As stated by Sinha (1953), there appears to be a polyploid series with a base 208 Rhodora [Vol. 85 Table 2. Cellular dimensions of representative species of Chaetomorpha and Rhizoclonium SPECIES CELL WIDTH CELL LENGTH Rhizoclonium riparium 8to 48 um_ 1|to4 times the width (Nienhuis, 1975) Chaetomorpha minima l0to 40 um_ | to 2 times the width (Taylor, 1960 & present study) R. tortuosum 27to 75 um_ 2 to 3 (1-4) times the (= Lola tortuosa) width (present study) Chaetomorpha brachygona 75 to 150 um_ | to 2 times the width (present study) C. linum 150 to 450 um _ 0.75 to 1.5 times the (present study) width Table 3. Chromosome numbers of various Chaetomorpha and Rhizoclonium species SPECIES 2n In AUTHOR Rhizoclonium riparium 36 18 Sinha (1958) R. tortuosum 24 — Sinha (1958) piel — Patel (1971b) 20 10 Perrot (1965) Chaetomorpha linum 36 18 Patel (197 1a) 36 18 Sinha (1958) C. aerea 24 12 Patel (1971a) 20 10 Hartman (1929) C. melagonium 24 12 Patel (1972) number of 6. Although enough information is available at present to question the validity of generic separation, additional information concerning the remaining species of Chaetomorpha and Rhizo- clonium must be collected before the final disposition of the two genera is made. 1983] Blair — Chaetomorpha & Rhizoclonium 209 ACKNOWLEDGMENTS I wish to thank Drs. A. C. Mathieson and D. P. Cheney for their assistance, discussions and criticisms during the various portions of this project, and to Dr. P. C. Silva for his helpful comments on the nomenclature of Chaetomorpha picquotiana. | would like to express my appreciation to Dr. W. R. Taylor and to the curators of the following herbaria for the loan of specimens: Atlantic Regional Laboratory, Halifax (NRcc); University of California, Berkeley (uc); Farlow Herbarium, Harvard University, Cambridge (FH); Memorial University of Newfoundland, St. John’s (NFLD); Univer- sity of Michigan, Ann Arbor (MICH); and the National Museum of Natural History, Paris (pc). Financial assistance provided by the Jackson Estuarine Laboratory, University of New Hampshire, is gratefully acknowledged. LITERATURE CITED AsgoTT, I. A., & G. J. HOLLENBERG. 1976. Marine algae of California. Stanford University Press. Stanford. AGARDH,C. A. 1824. System algarum. Lund. AGARDH, J.G. 1842. Algae maris mediterranei et Adriatici. Parisiis. ARESCHOUG, J. E. 1840. Algae scandinavicae exisccatae. 1843. Algarum (Phycearum) minus rite congnitarum pugillus secundus. Linnaea. p 268. 1850. Phycearum quae in maribus Scandinaviae Cresent, enumerato, sec- tic posterior ulvacaes contrien. Nov. Acta Reg. Soc. Sci. Uppsala. 14: 432-433. BLAIR, S. M. 1978. Biosystematic and taxonomic investigations of selected spe- cies of Rhizoclonium Kitzing and Chaeromorpha Kitzing in New England. M.S. Thesis, University of New Hampshire. Durham. .. A. C. MATHIESON, & D, P. CHENEY. 1982. Morphological and electro- phoretic investigations of selected species of Chaetomorpha (Chlorophyta; Cla- dophorales). Phycologia 21: 164-172. Botp, H. C., & M. J. WyNNE. 1978. Introduction to the algae: structure and reproduction. Prentice-Hall Inc. Englewood Cliffs. BORGESEN, F. 1925. The marine algae from the Canary Islands I. Chlorophyceae. K. Danske Vidensk. Selsk. Biol. Medd. 5S: 1-123. CarTER, N. 1919. The cytology of the Cladophoraceae. Ann. Bot. 33: 476-478. CHAPMAN, V. J. 1939. Some algal complexities. Rhodora. 41: 19-28. CHRISTENSEN, T. 1957. Chaetomorpha linum in the attached state. Bot. Tidsskr. §3: 311-316. 1975. Annotations to a distribution survey of Danish marine algae. Bot. Tidsskr. 69: 253-256. Cotiins, F.S. 1909. The green algae of North America. Tufts Coll. Stud. Vol. 2 no. 3. 210 Rhodora [Vol. 85 1918. The green algae of North America. Second supplement. Pap. Tufts Coll. Stud. — Sci. Ser. no. 37. .& A. B. Hervey. 1917. The algae of Bermuda. Pro. Am. Acad. Arts Sci. 53: 3-195. Dittwyn, L. W. 1809. Synopsis of British Confervae. London. EDELSTEIN, T., & J. MCLACHLAN. 1967. Investigations of marine algae of Nova Scotia IV. Species of Chlorophyceae new or rare to Nova Scotia. Can. J. Bot. 45: 211-214. Farrow, W. G. 1881. The marine algae of New England and adjacent coast. Report U.S. Fish. Comm. for 1879. HameL,G. 1929. Heterogamy of Cladophoraceae Lola (gen. nov.) /ubrica (Setch. & Gard.) C.r. hebd. Seanc. Acad. Sci., Paris 189: 1904—1906. HARTMAN, M. 1929. Untersuch urger uber die Sexualitat und den Generation- swechsel von Chaetomorpha und Entermorpha. Ber. It. bot. Ges. 47: 485-494. Harvey, W.H. 1846-51. Phycologia Brittanica. Vol. IV. 1857. Neris Boreali-Americana: Part III. HOLMGREN, P. K. & K. KEUKEN. (Ed.) 1974. Index Herbariorum I: The herbaria of the world. Regnum veg. Vol. 92. KJELLMAN, F.R. 1883. The algae of the Arctic Sea. Kongl. Sv. Vet-Akad. Hand. 20: 312. 1889. Om Beringhafuets Algflora. Kong. Sv. Vet.-Akad. Handl. 23(8): 58 pp, 7 pls. KORNMANN, P, 1972. Ein Beitrag zur Taxonomie der Gattung Chaetomorpha (Cladophorales, Chlorophyta). Helgolander wiss. Meeresunters 23: 1-31. .& P. H. SAHLING. 1977. Meeresalgen von Helgoland. Helgolander wis. Meeresunter 29: 1~289. Koster, J.T. 1955. The genus Rhizoclonium in the Netherlands. Publ. Sta. Zool. Naples, Italy. 27: 335-357. KOTZING, F. T. 1843. Phycologia generalis. Leipzig. 1845. Phycologia germanica. Nordhausen. 1847. Diagnosen und Vernerdugensuneyen oder dritischen algen. Bot. Ziet. Vol. 5. ———. 1849. Species algarum. Liepzig. 1853. Tabulae phycologicae. Nordhausen. Vol. 3. LINNEAUS, C. 1753. Species Plantarum. Vol. 2. Holmiae. Mazk, H., & A. SCHRAMM. Essai de classification des algue de la Guadeloupe. Second edition. Basse Terre. MonTAGne, J. E.C. 1849. Sixieme centurie de plantes cellulaires nouvelles, tant indigenes qu’exotiques. Decades II] 4 IV. Ann. Sci. Nat. 3 ser. Bot. 11: 33-66. MUCLLER, O. F. 1778. Icones Plantarum sponte nascentium in regris Davie et Norvegie et in Ducatibus slesvici. Holsatiae et Olbeburg, ad illustrandum opus de iisdem plantis, regio jussu esarandum. Florae Danicae 5: 721-780. Nienuuis, P. H. 1975. Biosystematics and ecology of Rhizoclonium riparium (Roth) Harvey (Chlorophyceae: Cladophorales) in the estuarine area of the rivers Rhine, Meuse, and Scheldt. Rotterdam. Pare, R.J. 197la. Cytotaxonomical studies on Chaetomorpha |. Chaetomorpha linum (Mill) Kiit. and C. aerea (Dillw). Kit. Phykos 11: 17-22. 1983] Blair — Chaetomorpha & Rhizoclonium 211 1971b. Cytotaxonomical studies on Rhizoclonium spp. |. Rhizoclonium implexum (Dillw.) Kiit. Seaweed Res. Util. 1: 8-18. 1972. Cytotaxonomical studies of British marine species of Chaetomor- Pha |. Chaetomorpha melagonium (Web. et Mohr) Kit. Phykos I; 17-22. Perrot, Y. 1965. Sur le cycle de reproduction d‘une Cladophoracee marine de la région de Roscoff: Lola implexa (Harvey). C.r. hebd. Seanc. Acad. Sci., Paris. 261: 503-506. PrasaD, B. N., S. Dutta, & V. JAIN. 1973. Cytological studies in the genus Rhizoclonium hieroglyphicum. Phykos 12: 106-113. Prick, W.M. 1967. Some aspects of the biology and taxonomy of the unbranched Cladophorales in the British Isles. Thesis, University of Liverpool. Rotu, A.G. 1797. Catalecta botanica plantae novae it minus cognitae discribun- ter atque illustrantur. Fasc. 1. SCAGEL,R.F. 1966. Marine algae of British Columbia and northern Washington. Part I: Chlorophyceae. National Museum of Canada. Bulletin 207. Biol. Ser. 74. SETCHELL, W. H., & N. L. GARDNER. 1920. The marine algae of the Pacific coast of North America. Part 2, Chlorophyceae. Univ. Calif., Berkeley. Publ. Bot. 8: 139-374, pls. 9-33. Sitva, P.C. 1950. Generic names of algae proposed for conservation. Hydrobio- logia 2: 252-280. SINHA, J. P. 1958. Chromosome numbers and life cycles in members of Clado- phorales. Br. Phyco. J. 1: 24-27. STAFLEU, F. A. (Ed.) 1978. International Code of Botanical Nomenclature. Reg- num veg. Volume 97. STOCKMAYER, S. 1898. Uber die Algengattung Rhizoclonium. Verh. Zool.-bot. Ges. Wien. 40: 571. Suur, J.N. 1840. Bietrage zur Algen Kunde. Flora 40: 289-320. Taytor, W.R. 1937a. Notes on North Atlantic Marine algae. Pap. Mich. Acad. Sci. 22: 225-233. ——_—. 1957. The marine algae of the northeast coast of North America. Univ. of Mich. Press. Ann Arbor. 1960. Marine algae of the eastern tropical and subtropical coasts of the Americas. Univ. Mich. Press. Ann Arbor, VicKEeRS, A. 1905. List des alg. mar. de la Barbade. Ann. Sci. Nat. Bot. 9: 45-66. Weser, F., & D. M. H. Monr. 1804. Naturhistorische Reise durch einen Teil Schwedens. Géttingeal. DEPARTMENT OF BOTANY AND PLANT PATHOLOGY UNIVERSITY OF NEW HAMPSHIRE DURHAM, NEW HAMPSHIRE 03824 212 Rhodora [Vol. 85 ERRATUM, JAN. 1983 ISSUE (VOL. 85, NO 841) “Kalmia ericoides revisited.” A block of text, comprising lines 4-16 on page 52 as published, was transposed from page 54. The mislocated text should start as line 5 of page 54, continuing the description and Distribution & Ecology sections of the discussion of Kalmia ericoides Wright ex Griseb. var. aggregata (Small) Ebinger. ANATOMY OF THE PERICARP OF CLIBADIUM, DESMANTHODIUM AND ICHTHYOTHERE (COMPOSITAE, HELIANTHEAE) AND SYSTEMATIC IMPLICATIONS Top F. STugessy AND Ho-Y1H Liu ABSTRACT The closely related genera, Clibadium, Desmanthodium, and Ichthyothere (Com- positae, Heliantheae) are distributed in Mexico, Central America, and Northern and Central South America. Some of the species of Clibadium (e.g., C. laxunr) have drupaceous achenes, which is unusual in the family. Comparative anatomical studies of fruits of the three genera show a strong similarity in the arrangement of the epidermis, hypodermis, and phytomelan and fiber layers. In most regards, Des- manthodium and Ichthyothere are more similar to each other than either is to Cliba- dium. The method of deposition of the phytomelan in all three genera begins with a series of tubes and cones from the fiber layer followed by patchy deposition of dark resistant material that eventually solidifies into a continuous layer. The phytomelan layer probably functions to protect the developing embryo and may also serve in regulating timing of germination. The genus Clibadium L. of the Compositae (tribe Heliantheae), with 30-40 species ranging from Mexico to Peru, is unusual among members of the family in that some species have drupaceous achenes. Only two other genera of the family are known to have similar fleshy fruits (Huber, 1898; Norlindh, 1977): Chrysanthe- moides Tourn. ex. Fabr. (tribe Calenduleae) and Wulffia Neck. ex Cass. (Heliantheae). Because Clibadium has already been studied for flavonoids (Bohm & Stuessy, 1981la), for sesquiterpene lactones and polyacetylenes (Czerson, et al., 1979), and for chromosome numbers and morphology for an eventual systematic synthesis by the senior author, an investigation of the anatomy of the unusual fruits of Clibadium would seem helpful. Especially useful would be indications of possible subgeneric groupings to test the sectional classification of Schulz (1912) and to compare with the broad flavo- noid differences documented recently (Bohm & Stuessy, 1981a). Clibadium is closely related to Desmanthodium Benth., which has eight species from Mexico and Central America, and to Ich- thyothere Mart., with 15 species in South America, especially Brazil (Stuessy, 1977). Although the achenes of these other two genera are not known to be fleshy, a comparative examination of fruit anat- ai 214 Rhodora [Vol. 85 omy would be useful to provide clues to evolutionary relationships both within and among all three taxa. This is especially important because the senior author (1977) has judged the three genera to be closely related (all placed in subtribe Milleriinae) whereas Robinson (1978, 1981) has recommended that each be placed in a separate subtribe (Clibadiinae, Desmanthodiinae, and Melampodiinae) of the Heliantheae. In addition, our preliminary studies on the peri- carp of Clibadium provided some insights on the method of forma- tion of the stony phytomelan layer. This dark brown or black layer is known to occur in the Compositae, primarily in the tribes Eupa- torieae and Heliantheae (Hanausek, 1912; Vaughan, 1970; Misra, 1972; Wagenitz, 1976; Hegnauer, 1977). It was hoped that under- standing the initiation and development of the phytomelan layer in the three genera, therefore, might provide even further insights to their evolutionary affinities as well as help reveal the mode of deposition of the phytomelan layer itself. A number of studies have been done on the origin of this layer in the Compositae (Hanausek, 1902, 1907, 1912; Vries, 1948; Politis, 1957; Misra, 1964, 1972; Pul- laiah, 1979, 1981), but they have been inconclusive as to events that result in its formation. The purposes of this paper, therefore, are to: (1) examine the achenial anatomy of different species within Clibadium, Desman- thodium, and Ichthyothere to learn if subgeneric groupings might be suggested within each; (2) postulate evolutionary relationships among the three genera based on pericarp anatomy; and (3) gain some insights on the development of the phytomelan layer in all three genera. MATERIALS AND METHODS Nineteen species of Clibadium (38 populations), three of Des- manthodium (five populations) and eight of /chthyothere (11 popu- lations) were examined (Table 1). Preparation of the achenes for studying the outer surface of the phytomelan layer was accom- plished by soaking in 10% NaOH for 4-14 hours followed by mechanical removal of the outer pericarp and finally by washing in distilled water. A few preparations were soaked in dilute HCI for one hour and then washed. This surface was then shadow-coated with gold and viewed in the SEM. 1983] Stuessy & Liu — Anatomy of pericarps 215 Table 1. Voucher specimens of taxa of Clibadium, Desmanthodium, and Ichthyo- there examined by SEM for external surface of the phytomelan layer and by light microscopy for pericarp anatomy. All vouchers cited here and elsewhere in this paper are at Os unless indicated otherwise (herbarium acronyms in parentheses after Holmgren, Keuken, and Schofield, 1981). SF = Stuessy & Funk, SG = Stuessy & Gardner, SJ = Stuessy & Jansen, SN = Stuessy & Nesom. F = freehand sections; p = paraffin-embedded sections; SEM =external SEM observa- tions; Ss = serial sections. Taxa and Vouchers Clibadium anceps Greenm., SG 45/8 [pss], Wilbur & Stone 10678 [SEM]; C. arbo- reum J.D. Smith, Dwyer //4/5 [sem], SG 4574 [F, SEM]; C. asperum (Aubl.) DC., Castaneda 6144 (Ny) [SEM], Killip & Smith 26834 (Gu) [SEM], McDaniel 2366 (us) [sem], SN 5899 [sem]; C. glomeratum Greenm., SG 4517 [SEM, ss], C. grandifolium S.F. Blake, Gentry 3034 [sem], SG 4533 [p,ss]; C. laxum S.F. Blake, SJ 4942 [p, SEM, ss]; C. leiocarpum Steetz in Seem., Almeda & Nakai 10678 [sem], SG 4456 [F]; C. micranthum O.E.Schulz, Killip & Smith 24822 (NY) [SEM]; C. parviceps S.F. Blake, Williams 10478 (Us) [SEM]; C. pentaneuron S.F. Blake, Pelaez 510 (us) [SEM] SF 5709 [ss]; C. peruvianum Poepp. ex DC., Killip & Smith 27263 (Us) [SEM], Mexia 6506a (Ny) [SEM]; C. pilonicum Stuessy, Hartman 3963 [SEM], C. pittieri Greenm., Cuatreca- sas 13709 (us) [SEM], Forero & Gentry 725 (COL) [SEM], Standley 45815 (us) [SEM], SG 4465 [ss]; C. psilogvnum S.F. Blake, Weberbauer 7864 (GH) [SEM]; C. sessile S.F. Blake, Hartman 3916 [SEM]; C. sprucei S.F. Blake, SN 481/ [ss]; C. surinamense L., Bristor 690 (us) [SEM], Cuatrecasas 14004 (us) [SEM], SG 445/ [sem], SN 5856 [ss], SPJ 4935 [sem]; C. terebinthinaceum DC., Cuatrecasas 23932 (F) [SEM], Fuchs et al. 21728 (COL) [SEM], SF 5737 [F]; C. trianae S.F. Blake, Cuatrecasas 6475 (F) [SEM], SN 4680 [F]. Desmanthodium fruticosum Greenm., King & Soderstrom 505 (uc) [SEM], SG 4115 [p, SEM, ss]; D. hondurense A. Molina, Hazlett 849 (Mo) [SEM], SG 4390 [ss]; D. perfoliatum Benth., SG 4306 [F, SEM]. Ichthyothere agrestis Baker in Mart., Hatschbach 19911 (F) [F]; /. cordata Malme, Maguire & Maguire 44510 (Ny) [F]; /. cunabi Mart. in Buchn., Dusen s.n. (MO) [F]; /. hirsuta, Irwin et al. 25876; (mo) [ss]; / latifolia Gardn., Irwin et al. 34716 (Mo) [F]; 7. rufa Gardn., Argent 6711] (Ny) [SEM]; / scandens S.F. Blake, Cuatrecasas 13313. (COL) [F]; 1. terminalis (Spreng.) Malme, Assis 157 (uC) [ss], Forero, Bastidas & Ramirez 882 (NY) [P, SS], Pereira 7535 (NY) [P, SEM, SS], Prance 8603 (NY) [SEM]. 216 Rhodora [Vol. 85 Sections of pericarp were made free-hand and by traditional paraffin embedding and sectioning (Sass, 1958). Stains used for the paraffin mounts were safranin and fast green. The serial sections were done in all paraffin material plus 12 other collections (Table 1) cut free-hand to determine the degree of constancy of pericarp anat- omy throughout the achene. RESULTS Structure of Pericarp The anatomy of the pericarp in Clibadium, Desmanthodium, and Ichthyothere is basically similar with five distinct zones observable (Fig. 1-6): epidermis, hypodermis, phytomelan layer, fiber zone, and internal parenchyma. Within each genus a constancy of ana- tomical features exists among the species examined. In addition, the organization of the pericarp is constant throughout the entire length of the achene. Slight variations occur in numbers of hypodermal cell layers and thickness of the phytomelan layer, but they are minor and do not obscure nor contradict patterns observed and docu- mented (Figs. 1-6; Table 2). The pericarp is the thickest in Clibadium, with Ichthyothere next, followed by Desmanthodium. The species of Clibadium with dru- paceous fruits (such as C. /axum) are such due to an inflated hypo- dermis with more and larger cells and cell layers. In all other respects they have the same structure as the other species of the genus with dry achenes. The epidermis and internal parenchyma provide no significant taxonomic characters for differentiating the three genera. The hypodermis, however, does differ, with paren- chyma being 2-5 cells thick and radially arranged in Clibadium and 3-6 cells thick and irregularly arranged in /chthyothere. Desman- thodium has only |-2 parenchyma cell layers. The phytomeian layer is different in the three genera (Figs. 1-3; Table 2), and this is Figures 1-6. Transections of mature (Fig. |-3) and immature (4-6) pericarps of Clibadium, Desmanthodium, and Ichthyothere. 1. C. laxum, SJ 4942. 2. D. frutic- osum, SG 4115. 3. 1. terminalis, Pereira 7535 (NY). E, epidermis; H, hypodermis; P, phytomelan layer; F, fiber zone; 1P, internal parenchyma. 1983] Stuessy & Liu — Anatomy of pericarps e sou pa Yi Table 2. Comparison of pericarp anatomy in Clibadium, Desmanthodium, and Ichthyothere. Region of pericarp Clibadium Desmanthodium Ichthyothere Hypodermis 2-5 cell layers of 1-2 cell layers of 3-6 cell layers of parenchyma in radial parenchyma parenchyma irregularly files arranged Phytomelan continuous and regular discontinuous, discontinuous, associated layer associated with with fiber bundles fiber bundles Fiber zone continuous, 4-10 cell discontinuous, discontinuous, with large layers thick bundles of 15-55 fibers (70-140) and small (20-40) bundles of fibers 8I1C vIOpOUy $8 1OA] 1983] Stuessy & Liu — Anatomy of pericarps 219 especially evident in surface configuration (Figs. 7-12). Clibadium has a thick continuous phytomelan layer, /chthyothere has the layer interrupted by eight longitudinal grooves, and Desmanthodium has numerous longitudinal grooves over the entire surface of the achene. In all taxa the phytomeian is intimately associated with the fiber zone immediately below. The fibers appear similar in the three gen- era, but in Clibadium they form a continuous layer (as with the phytomelan), in /chthyothere the fibers are interrupted eight times, and in Desmanthodium they are interrupted many times resulting in numerous small clusters. Variations in Upper Surface of Phytomelan Layer The upper surface of the phytomelan in Clibadium, Desmantho- dium, and Ichthyothere differs at the generic level with Desmantho- dium being nearly smooth and the former and the latter having more regular cellular sculpturing (Figs. 7-12). These cellular pat- terns result from the formation of phytomelan against the cells of the hypodermis; the phytomelan itself is acellular. Within each genus, minor variations of sculpturing exist, but they are not signifi- cant for the recognition of most species and they also show some developmental variation. Development of Phytomelan Layer Examination of cross-sections of young achenes of species of Clibadium, Desmanthodium, and Ichthyothere (Figs. 4-6) reveals general aspects of the system of development of the phytomelan layer. In all three genera the fiber zone below and hypodermis above serve as templates between which the phytomelan begins to be de- posited. A developmental sequence in Clibadium surinamense serves as an example for these details (Figs. 13-18). The upper cell layer of the fiber zone has tubular or cone-shaped extensions (Figs. 4 & 13), which touch the hypodermis. It is not clear whether the phytomelan derives from the fiber zone, the hypodermis, or both. The phytome- lan thickens and fuses laterally (Figs. 14-16), and eventually forms a hardened solid layer (Figs. 17 & 18). Ichthyothere and Desman- thodium develop in a similar fashion (Figs. 5, 6, 19, & 20 for similar early stages). 220 Rhodora [Vol. 85 Figures 7-12. External surface of phytomelan layer of mature achenes of Cliba- dium, Desmanthodium, and Ichthyothere. 7 &8. C. pittieri, Standley 45815 (us). 9 & 10. D. fruticosum, King & Soderstrom 5058 (uc). 11 & 12. 1. rufa, Argent 6711 (ny). All scale as in Figs. 7 & 8. DISCUSSION Systematic and Evolutionary Significance of Variations in Pericarp Structure Because of the initial interest in the unusual drupaceous achenes of some species of Clibadium, it was hoped that many differences in pericarp structure would be found among species of the genus. Although some minor variations do occur (e.g., a more elevated series of tubercles on the upper surface of the phytomelan layer in C. sychnocephalum), they are not generally useful at either the specific or subgeneric levels within any of the genera. The differences in pericarp structure among the three genera, however, are significant and help distinguish them more clearly (cf. differences listed in Table 2). A more important question is how 1983] Stuessy & Liu — Anatomy of pericarps 221 Figures 13-18. Development of phytomelan layer in Clibadium surinamense. 13. Surface of very immature achene with hypodermis partially removed, showing tu- bular extensions of fibers (FAA preservation). 14. Surface of young achene with hypodermis completely removed showing initiation of phytomelan around tips of extensions (FAA preservation. NaOH cleaning). 15. Inner view of hypodermis showing attachment of upper part of extensions with beginnings of envelopment by phytomelan and extensions fractured below (dry). 16. Surface of nearly complete and phytomelan layer (dry). 17. Surface of mature phytomelan layer (dry, with NaOH cleaning). 18. Transection of mature achene wall (dry) with solid phytomelan layer (Pp) between hypodermis (H) and fibers (F) below; compare with Fig. |. Figs. 13, 15, 16, 18, Stuessy & Gardner 4451; 14, Stuessy et al, 4935; 17, Bristor 690 (US). Figs. 13, 14, & 17, and 15 & 16 same scales, respectively. closely related the three genera are to each other. The structure and development of the pericarp is essentially the same and suggests a strong evolutionary tie. Particularly noteworthy are the extensions of the fibers around which the phytomelan accumulates. Although a number of Compositae have been analyzed for pericarp features (Hanausek, 1902, 1912; Briquet, 1916; Giroux, 1930, 1933; Vries, 1948; Misra, 1964, 1972; Carlquist, 1958; Dittrich, 1968a, b, 1969), it remains to be seen how relatively close Clibadium, Desmantho- No No No Rhodora [Vol. 85 dium, and Ichthyothere will be after many other genera of the Hell- antheae are investigated. Nonetheless, the similarity of pericarp features does not appear to support Robinson’s (1978, 1981) place- ment of Clibadium, Desmanthodium, and Ichthyothere in separate subtribes. Ihe strong similarities of pericarp structure in C/ibadium, Des- manthodium, and Ichthyothere make it difficult to determine the relationships of the genera one to another. The most significant point seems to be that Desmanthodium and Ichthyothere are similar in having discontinuous phytomelan as opposed to it being continu- ous in Clibadium. This suggests that the two are more closely related than either is to Clibadium. On the other hand, the similar cellular patterns of the external phytomelan surface in Clibadium and /chthyothere argue for their close relationship, in contrast to the smooth surface in Desmanthodium. The ontogeny of the phy- tomelan layer in all three genera is essentially the same, except that Desmanthodium and Ichthyothere have more conical and smaller fiber extensions whereas those in C/ibadium are larger and more tubular. Figures 19 & 20. Surface of young achene showing patchy development of phy- tomelan layer in /chthyothere terminalis (19. dry, NaOH cleaning; Prance et al. 8603, Ny) and Desmanthodium perfoliatum (20. FAA, HCI cleaning, Stuessy & Gardner 4306). Same scale. 1983] Stuessy & Liu — Anatomy of pericarps 224 Studies on the structure and development of the pericarp do not indicate which type is the most evolutionary primitive and which is more advanced. Because of the absence of comprehensive studies on genera of the Heliantheae, the basic pattern in the tribe is not yet known. This makes evolutionary directionality difficult to deter- mine. One might speculate that the structure of Clibadium may be more primitive simply because it is a more simple pattern. Chromo- some numbers give limited support to this idea because Clibadium is known as n = 16 and 24 (x = probably 8) (Turner & King, 1964; Coleman, 1968; Powell & King, 1969a, b; Powell & Cuatrecasas, 1970; Grashoff, Bierner, & Northington, 1972; Solbrig, et al., 1972), Desmanthodium as n = 18 (Fay, 1974; Keil & Stuessy, 1977), and Ichthyothere as n = 32 (Coleman, 1970). An ancestral base of x = 8 for the entire complex is the simplest explanation for these data which would make Clibadium and Desmanthodium, as the lowest level polyploids, more primitive than the others (Desmanthodium may have arisen by polyploidy from an x = 9 ascending aneuploid line). The flavonoid data give some suggestion that Clibadium has a more primitive biochemical profile than Desmanthodium, in that the latter accumulates more mono-, di-, and tri-methyl derivatives (Bohm & Stuessy, 1981b). Alternatively, however, the pattern of phytomelan interrupted by parenchymal rays as in Desmanthodium and Ichthyothere might be more primitive because this aspect also occurs (Hanausek, 1912; Vries, 1948; Vaughan, 1970) in Guizotia abyssinica, Helianthus annuus, Madia sativa, Rudbeckia fulgida, and Sclerocarpus uniserialis, taxa of the same tribe (Heliantheae). A continuous layer, however, is also known (Hanausek, 1912) in Eclipta alba, Engelmannia pinnatifida, Silphium trifoliatum, and Verbesina encelioides of the Heliantheae. More data are obviously needed from many other genera. Development of Phytomelan Layer A number of studies on the development of phytomelan in the Compositae exist (e.g., Hanausek, 1902, 1912; Vries, 1948; Politis, 1947; Misra, 1964, 1972; Pullaiah, 1979, 1981). These and our own studies have reported the following observations: (1) the phytome- lan forms between the hypodermis and the fiber zone; and (2) the phytomelan is non-cellular, very resistant, and whatever markings or sculpturing it has derives from the hypodermal cell layer in 224 Rhodora [Vol. 85 contact with it (e.g., the tubercles on the phytomelan surface of Clibadium pittieri reflect depressions at cell wall junctures of the hypodermis, Figs. 7 & 8). Previous workers (Vries, 1948; Misra, 1964, 1972; Pullaiah, 1979, 1981) have believed the phytomelan to be derived from exudates from the hypodermis. This may be so, but it could not be determined to our satisfaction. Alternatively, mate- rial may be exuded from the fiber zone or from both sides of the location of deposition. Detailed cytological and histochemical stu- dies will be needed to resolve this question. The tubes and cones in Clibadium, Desmanthodium, and Ichthyothere may function sim- ply as a framework around and within which the phytomelan solidi- fies, but they also may transport materials to the site of deposition (this seems especially possible in Clibadium with the numerous tubes, more than one would think necessary for just structural pur- poses; see Fig. 13). The chemical nature of phytomelan is still unknown although it has been proposed by Dafert and Miklauz (1912) to have a molecu- lar formula x(C,H,9O5)-yH,O and suggested by Hegnauer (1964, 1977) to be a polyacetylene. Our studies show that it is not modified by dilute NaOH or HCI treatment, whereas the cells of the epider- mis and hypodermis are easily softened and destroyed. Likewise, the adaptive value of the phytomelan layer is uncertain, but it probably functions as a protective device for the mature fruit and may also regulate the timing of seed germination. The fact that this layer is found principally within the tribes Eupatorieae and Heliantheae (Hanausek, 1912; Misra, 1972) also gives it special tax- onomic potential at the higher levels of the hierarchy in the family. The homologies of structure and developmental patterns will have to be determined before their full efficacy can be appreciated. ACKNOWLEDGMENTS Thanks go to: Mike Cichan and Vicki Funk for technical assist- ance with free-hand thin sections and sEM preparations, respectively, plus helpful discussions on the development of the early phases of the study; and the National Science Foundation for support under Grant No. DEB75-20819. 1983] Stuessy & Liu — Anatomy of pericarps 225 LITERATURE CITED Boum, B. A., & T. F. Sturssy. 198la. Flavonol derivatives of the genus Cliba- dium (Compositae). Phytochemistry 20: 1053-1055. 4 . 1981b. Flavonol derivatives of Desmanthodium (Composi- tae). Phytochemistry 20: 1573-1575. Briguet, J. 1916. Etudes carpologiques sur les generes de Composeés Anthems, Ormenis, et Santolina suivies de quelques conclusiones anatomiques et physio- logiques d’interét general. Ann. Conserv. et Jard. Bot. Geneve 19: 257-313. CarRLQuistT, S. 1958. Anatomy and systematic position of Centaurodendron and Yunquea (Compositae). Brittonia 10: 78-93. CoLeMAN, J. R. 1968. Chromosome numbers in some Brazilian Compositae. Rhodora 70: 228-240. 1970. Additional chromosome numbers in Brazilian Compositae. Rho- dora. 72: 94-99. CZERSON, H., F. BOHLMANN, T. F. Stuessy, & N. H. Fiscuer. 1979. Sesquiterpe- noid and acetylenic constituents of seven Clibadium species. Phytochemistry 18: 257-260. DarertT, F. W., & R. MIKLAUZ. 1912. Untersuchungen iiber die kohleahnliche Masse der Kompositen. Denkschr. kais. Akad. Wiss. Wien, Math.-Nat. KI. 87: 143-152. Dittrich, M. 1968a. Fruchtanatomische und cytologische Untersuchungen an einigen Arten der Gattung Rhaponticum Adans. und Leuzea DC. (Composi- tae). Osterr. Bot. Z. 115: 379-390. 1968b. Morphologische Untersuchungen an den Friichten der Subtribus Cardueae-Centaureinae (Compositae). Willdenowia 5: 67-107. 1969. Anatomische Untersuchungen an den Friichten von Carthamus L. und Carduncellus Adans. (Compositae). Candollea 24: 263-277. Fay, J.J. 1974. In: 1OPB chromosome number reports XLV. Taxon 23: 619-624. Grroux, M. 1930. Sur la carpologie de quelques Composées Nord-Africaines. Bull. Soc. Hist. Nat. Afr. Nord. 31: 161-189, 1933. Note sur la position systematique des Chrysanthemum cinerariifo- lium (Revy.) Vis. suivi de quelques remarques sur la caractéres carpologiques des Tanacetum. Bull. Soc. Hist. Nat. Afr. Nord. 24: 54 62. Grasnorr, J. L., M. W. Brenner, & D. K. NorTHINGTON. 1972. Chromosome numbers in North and Central American Compositae. Brittonia 24: 379-394. HANAUSEK,T.F. 1902. Zur Entwickelungsgeschicte des Perikarps von Helianthus annuus. Ber. Deutsch. Bot. Ges. 20: 449-454. 1907. Die “Kohleschicht” im Perikarp der Kompositen. Sitzungsber. kais. Akad. Wiss. Math.-Nat. K1., Abt. 1, 116: 3-32. 1912. Untersuchungen iiber die kohle’hnliche Masse der Kompositen. Denkschr. kais. Akad. Wiss. Wien, Math.-Nat. KI. 87: 93-142. HEGNAUER, R. 1964. Chemotaxonomie der Pflanzen, Band 3. Birkaiiser Verlag, Basel. 226 Rhodora [Vol. 85 1977. The chemistry of the Compositae, pp. 283-335. /n: V.H. Heywood, J.B. Harborne, & B.L. Turner (eds.), The Biology and Chemistry of the Com- positae. Academic Press, London. (cf. p. 289). Ho~MGreN, P. K.. W. Keuken, & E. K. Scuorietp. 1981. Index Herbariorum, ed. 7. Regnum Veg. 106: 1-452. Huper, J. 1898. Observacoes histologicas e biologicas sobre o fruto da Wulffia stenoglossa DC. (Jambu). Bol. Mus. Paraense Hist. Nat. 2: 96-101. Keir, D. J., & T. F. Sturssy. 1977. Chromosome counts of Compositae from Mexico and the United States. Amer. J. Bot. 64: 791-798. Misra, S. 1964. Floral morphology of the family Compositae II]. Development of the seed and fruit in Flaveria repanda. Bot. Mag. Tokyo 77: 290-296. 1972. Floral morphology of the family Compositae V. The seed coat and pericarp in Verbesina encelivides (Cavy.) Benth. and Hk. f. ex A. Gray. J. Indian Bot. Soc. 51: 332-341. NoRLINDH, T. 1977. Calenduleae—systematic review, pp. 961-987. /n: V. H. Heywood, J. B. Harborne, & B. L. Turner (eds.), The Biology and Chemistry of the Compositae. Academic Press, N.Y. Pouitis, J. 1957. Untersuchungen iiber die cytologische Bildung der Phytomelane bei einigen Zinnia-Arten. Protoplasma 48: 269-275. PoweL, A. M, & J. CUATRECASAS. 1970. Chromosome numbers in Compositae: Colombian and Venezuelan species. Ann. Mo. Bot. Gard, 57; 374-379, .R.M.KING. 1969a. Chromosome numbers in the Compositae: Colom- bian species. Amer. J. Bot. 56: 116-121. & . 1969b. Chromosome numbers in the Compositae: West Indian species. Sida 3: 319-320. PuLLAIAH, T. 1979. Embryology of Adenostemma, Elephantopus and Vernonia (Compositae). Bot. Notiser 132: 51-56. 1981. Studies in the embryology of Heliantheae (Compositae). Pl. Syst. Evol. 137: 203-214. Ropinson, H. 1978. Studies in the Heliantheae (Asteraceae). XIV. Validation of subtribes. Phytologia 41: 39-44. 1981. A revision of the tribal and subtribal limits of the Heliantheae (Asteraceae). Smithson. Contrib. Bot. 51: 1-102. Sass, J. E. 1958. Botanical Microtechnique, ed. 3. lowa State College Press, Ames, lowa. Scuutz, O. E. 1912. Beitrage zur Kenntnis der Gattung Clibadium. Engl. Bot. Jahrb. 46: 613-628. Sounric, O. T., D. W. Kynos, M. Power, & P. A. RAveN, 1972. Chromosome numbers in Compositae VIII: Heliantheae. Amer. J, Bot. 59: 869-878. Stussy, T. F. 1977. Heliantheae—systematic review, pp. 621-671. In: V. H. Heywood, J. B. Harborne, and B. L. Turner (eds.), The Biology and Chemistry of the Compositae. Academic Press, N.Y. TURNER, B. L., & R. M. KING. 1964, Chromosome numbers in the Compositae. VIII. Mexican and Central American species. Southwest. Nat. 9: 27-39. VAUGHAN, J. G. 1970. The structure and utilization of oil seeds. Chapman and Hall Ltd., London. 1983] Stuessy & Liu — Anatomy of pericarps 227 Vries, M.A. DE. 1948. Over de vorming van Phytomelaan bij Tagetes patula L. en enige andere Composieten. H. Buurman, Leiden. WAGENITZ,G. 1976. Systematics and phylogeny of the Compositae (Asteraceae). Plant Syst. Evol. 125: 29-46. DEPARTMENT OF BOTANY THE OHIO STATE UNIVERSITY 1735 NEIL AVENUE COLUMBUS, OHIO 43210 228 Rhodora [Vol. 85 CORRECT AUTHOR CITATION FOR SENECIO WERNERIAEFOLIUS LEONARD J. UTTAL The name Senecio werneriaefolius is generally cited as S. wer- neriaefolius A. Gray. It is so published by Barkley, No. Am. Flora ser II; 10: 88. (1978) and by Greenman, Ann. Mo. Bot. Gard. 5: 61. (1918). This is incorrect. The proper citation of this name is Senecio werneriaefolius (A. Gray). A. Gray. Gray first described this taxon as Senecio aureus var. werneriaefolius A. Gray, Proc. Acad. Phila. 1863: 68. (1864). He later proposed the new combination Senecio werneriaefolius, A. Gray, Proc. Am. Acad. 19: 54. (1883), based on the original varietal description with the same type: Colorado. Hall and Harbour 331. (GH). It is required, therefore, to cite the author as proposed above. DEPARTMENT OF BIOLOGY VIRGINIA POLYTECHNIC INSTITUTE AND STATE UNIVERSITY BLACKSBURG, VIRGINIA 24061 A CLARIFICATION OF THE STATUS OF CAREX CRINITA AND C. GYNANDRA (CYPERACEAE) LisA A. STANDLEY Abstract. Carex crinita and C. gynandra are considered by some authors to be conspecific on the basis of morphological similarities. Variation in morphology, foliar anatomy, chromosomal numbers, and habitat preference among 12 popula- tions of these taxa was studied to clarify their status and relationship. Two groups of populations are distinguished by non-overlapping qualitative differences in the mor- phology of perigynia, scales and sheaths. These groups appear to be reproductively isolated by differences in flowering time where they are sympatric. These taxa must be considered as distinct species on the basis of the morphological differences and reproductive isolation. Shared features of morphology, anatomy, and chromosomal number suggest that they are very closely related. Carex is the largest and most complex genus of the Cyperaceae in temperate regions. Estimates of numbers of species range from 793 (Kiikenthal, 1909) to 1500 (Airy-Shaw, 1973). Existing classifica- tions of this taxonomically difficult group are based on the mor- phology of features which appear to be highly reduced and conserva- tive. The plants are monoecious and anemophilous. A pistillate flower consists of a uniovulate gynoecium and a single style with 2 to 3 stigmas, and is included in a modified bract called a perigynium which is subtended by a single scale. The linear leaves are ligulate and have a closed tubular sheath. Variation in relatively few charac- ters, such as the shape, size, and color of the perigynia and subtend- ing scales, the dimensions of leaves and of involucral bracts, and differences in habit are used to distinguish species, although the extent and pattern of variation of these characters is often not understood. Various authors have used species concepts which are narrow (MacKenzie, 1935) or broad (Fernald, 1950) with regard to morphological variation. The construction of a classification of Carex which accurately represents the evolutionary relationships of taxa requires a more complete understanding of the patterns of variation within the genus. Morphological variation must be better understood and documented. Variations in anatomy, cytology, and reproductive biology which have been shown to be valuable in assessing relationships of taxa in the Poaceae (Stebbins & Cromp- ton, 1961) may also be important in the Cyperaceae. 229 230 Rhodora [Vol. 85 Sao ae : . LV EPA RZ Figure |. Representative perigynia, achenes, and scales. A. Pistillate scale of C. crinita. B. Perigynium of C. crinita. C. Pistillate scale of C. gynandra. D. Perigynium of C. gynandra. E-G. Achenes. (A, B, G; C. crinita, Standley 77-95; C, D; C. gynandra, Standley 76-104: E, F; C. gynandra Standley 77-76). One specific problem which may be resolved by the application of a biosystematic analysis is the relationship of two taxa of eastern North America which are assigned to section Cryptocarpae (Fer- nald, 1950). Carex crinita Lam. and C. gynandra Schwein. are tall, caespitose perennials of wet sites, with pendant unisexual spikes, biconvex perigynia, awned scales, and lenticular achenes which are often invaginated on one margin. They were originally distinguished by differences in the shape of the perigynia (Schweinitz, 1824) and later by additional differences in the vestiture of the sheaths of the proximal leaves and the shape of the scales (Boott, 1858). These two taxa have been recognized as distinct species (Weatherby, 1923, 1983] Standley — Carex crinita & C. gynandra 231 MacKenzie, 1935; Hermann, 1941) or as subspecies or varieties of a single, polymorphic species (Bailey, 1886; Fernald, 1950; Gleason & Cronquist, 1963). Complete descriptions and synonymy are pro- vided by MacKenzie (1935). Carex mitchelliana M.A. Curtis of the southeastern coastal plain is morphologically similar to C. crinita and C. gynandra, but is distinguished by differences in the shape of the achenes. Although this taxon is undoubtedly part of the Carex crinita complex, its status will not be discussed at this time as popu- lations have not been investigated in the field. The present study is an attempt to resolve the problems of classi- fication of these taxa by examining variation in morphology, foliar anatomy, chromosomal numbers, and habitats within populations in the vicinity of Ithaca, N.Y. This information, although from a limited geographic area, is useful in determining whether these two taxa should be recognized as separate species, and in clarifying their relationship. METHODS Eleven populations of Carex within a 20 km radius of Ithaca, N.Y. were sampled in this study (Table 1). Plants in each population were enumerated and a random sample of 10% of the plants in each population was chosen for study of morphological and anatomical features. For each character studied, average values are based ona single measurement per culm, with the following exceptions: the average of five measurements per culm was used for estimates of the dimensions of perigynia, scales, and achenes of each plant. To standardize measurements, only the proximal spike on a culm was chosen for study. Perigynia and their subtending scales were chosen arbitrarily from among those on the basal third of that spike. Herbarium specimens at CU, BH, GH, NY, and TRT were examined for the construction of distribution maps and to test hypotheses developed in the study of populations. In addition, photographs of the holotype of Carex crinita (Pp, photo GH!) and the holytype of C. gynandra (PH!) were examined to verify the use of these names. For anatomical study, a section 5 cm in length from the middle of the longest leaf of a randomly selected vegetative culm from each sampled plant was preserved in FAA. Transverse sections of leaves were cut with a razor blade, cleared in commercial bleach for 5 minutes, and mounted in Hoyer’s solution for study. Table 1. Descriptions of localities of populations of Carex crinita and C. gynandra studied in the vicinity of Ithaca, N.Y. Population Location Elevation (m) Habitat Taxon Connecticut Hill Tompkins Co., 7.5 km W of $27 marsh *C. gynandra Newfield Henderson Gulf Tompkins Co., 5km SW of 375 marsh *C. gynandra Newfield Carter Creek Tompkins Co., 0.5 km NE of 549 marsh *C. crinita Alpine Michigan Hollow Tompkins Co., 5 km S of 395 marsh, C. crinita Danby swamp Spencer Tioga Co., 2.5 km N of 330 marsh C. crinita Spencer Slaterville Tompkins Co., 5 km N of 375 marsh *C. crinita, Slaterville Springs C. gynandra Six-Mile Creek Tompkins Co., 7 km SE of 215 marsh C. crinita Ithaca Freeville Tompkins Co., 0.5 km N of 318 marsh, *C. crinita, Freeville swamp *C. gynandra Sapsucker Woods 5 km N of Ithaca 325 marsh *C. crinita Lansing Tompkins Co., 5 km W of 335 marsh C. crinita South Lansing *= chromosomal number determined for this population. CET elopouYy $8 190A] 1983] Standley — Carex crinita & C. gynandra 233 For study of chromosomes, collections of staminate spikes were made during the first half of May 1976. Spikes were fixed in Car- noy’s solution (3:1 absolute ethanol: acetic acid,v/v) for at least 24 hours at 10°C, and stored in 70% ethanol under refrigeration. An- thers were squashed onto a slide and stained with acetocarmine according to the methods outlined by Faulkner (1972). Counts of chromosomes were obtained from metaphase plates of the first mei- otic division of the microsporocytes. The majority of counts re- ported are based on at least five counts in each anther studied. RESULTS The majority of morphological characters studied exhibited little variation among populations or individuals. Dimensions of culms, leaves, bracts, spikes, scales, and achenes exhibit the same range of variation in both taxa (Table 2). These data were obtained from Table 2. Morphology of C. crinita and C. gynandra C. crinita C. gynandra Character Mean Range Mean Range Length of vegetative leaf 117 cm 82-165 107 cm 65-148 Width of vegetative leaf 10 mm 7- 13 10 mm 7- 13 Height of flowering culm 119 cm 73-155 116 cm 67-162 Length of involucral bract 34 cm 18- 60 26 cm 12- 46 Width of involucral bract 9mm 6- 11 7mm 5- 13 Length of staminate spike 6.3 cm 3.8- 9.0 45 cm 1.5- 63 Length of pistillate spike 7.2 cm 4,3-11.4 7.4 cm 5,5-10.4 Length of proximal peduncle 45 cm 2.0- 7.2 2.9 cm 1.5- 5.6 Length of perigynium 25mm _ 2.0- 3.0 3.1mm 2.4- 4.2 Width of perigynium 1.8mm _ — 1.5- 2.0 1.7mm _— 1.5- 2.1 Thickness of perigynium 1.0 mm 8- 1.2 1.5 mm 1.3- 1.7 Length of pistillate scale 6.0mm _ — 3.2-10.0 5.2mm _ — 3.5- 8.0 Length of awn 3.9 mm 1.0- 7.5 2.8 mm 1.5- 5.8 of pistillate scale Length of achene 1.6 mm 1.3- 2.0 1.8 mm 1.6- 2.1 Width of achene 3mm 1.1- 1.4 1.3mm 1.1- 1.4 Apex of scale retuse acute Vestiture of sheaths of glabrous scabrous proximal leaves Outline of perigynium obovate elliptical 234 Rhodora [Vol. 85 study of local populations, but are consistent with dimensions of herbarium specimens from the entire range of distribution of both taxa. Achenes are lenticular in outline but margins may be entire or invaginated along one or both margins (Fig. 1). The shape of achenes has been used to distinguish taxa within this group. I have found that the shape of the fruit commonly varies among perigynia of a single individual and is therefore not a reliable taxonomic character. Three morphological features, vestiture of the sheaths, shape of the scales, and shape of the perigynia exhibited variation between populations (Table 2). The sheaths of leaves, particularly the sheaths of the proximal leaves, are either glabrous or scabrous. Scabrous sheaths have an average of 20 silica prickle-hairs (0.15 mm in length) per mm2. Plants with glabrous sheaths lack prickles entirely. The scales which subtend the perigynia have a broad, 3- nerved midrib that is excurrent as an awn. The length of this awn varies within a spike, decreasing markedly toward the apex of the spike. The scale, exclusive of the awn, is obovate in outline and may be acute or retuse at the apex (Figure |). Perigynia may be obovate or elliptical in outline (Figure |), Obovate perigynia are rounded in their cross-sectional outline, with a characteristic width of 1.8 mm and thickness of 1.5 mm. The average length of an obovate perigy- nium is 2.4 mm. Elliptical perigynia are flattened in cross-sectional outline, with an average width of 1.7 mm and thickness of 1.0 mm. These have an average length of 3.0 mm. Length of the perigynia is correlated with differences in the characters of outline and cross- sectional shape, but the range of perigynium length does overlap, as shown in fig. 2, and is not a consistently reliable taxonomic character. Patterns of variation in the character-states of sheaths, scales and perigynia are correlated. Plants with scabrous sheaths have dorsi- ventrally flattened, elliptical perigynia and scales which are acute at the apex. Plants with glabrous sheaths have inflated, obovate peri- gynia and scales which are retuse at the apex. The scatter diagram (Fig. 2) illustrates these correlations. Two groups of populations, which correspond to the two taxa, are clearly distinguishable on the basis of these three morphological features. Leaves are shallowly plicate (Fig. 3) with an average of 40 (range, 28-46) vascular bundles separated by transversely oblong air- 1983] Standley — Carex crinita & C. gynandra 235 oO 4 L 0 C6 OO oO Oo ks 08 2 Oo E 3 ae * E 048 6 su « Oo ® s Et ax ee te ®@ = os ee c ee 3 e are al t.3° ae ! \ i 0.5 10 LS Thickness (mm] Figure 2. Scatter diagram of dimensions of perigynia. Open circles represent plants with scabrous sheaths (Carex gynandra). Closed circles represent plants with glabrous sheaths (Carex crinita). cavities. The epidermal surfaces are differentiated. The adaxial epi- dermal surface consists of cells which lack papillae. Cells of the abaxial surface are smaller, and have a single median papilla. Sto- mates occur only on the abaxial surface. Bulliform cells occur ina single group adaxial to the midvein, and consist of a layer of large, thin-walled cells subtended by a layer of smaller clear cells. The adaxial projection of the midvein, or keel, may be rounded or acute 236 Rhodora [Vol. 85 Figure 3. Features of transverse sections of leaves of Carex gynandra (Standley 77-76). A. Midrib, showing median vascular bundle. B. Secondary vascular bundle. C. One-half transverse section of leaf. in transverse section. Variation in the shape of the keel was not found to be correlated with variation in other features of anatomy or of morphology. No other variations in foliar anatomy were observed. Chromosomal numbers were obtained from five populations of plants with glabrous sheaths (Table 1). All individuals studied had 33 bivalents. No univalents or multivalents were seen, although Wahl (1940) observed both n = 33 and n = 32+3 in plants of Carex crinita. | examined voucher specimens for his counts at BH and GH. No morphological variation was correlated with these differences in pairing. Populations studied in upstate New York occur in roadside ditches and in two major types of undisturbed communities with waterlogged soils. The majority of populations occurred in wet open meadows (Table |) dominated by monocotyledons. Up to 20 spe- cies, including Typha latifolia, Acorus calamus, Glyceria grandis, Scirpus cyperinus, Eupatorium maculatum, Impatiens capensis, and Onoclea sensibilis frequently occur with Carex in these sites. Soils 1983] Standley — Carex crinita & C. gynandra 237 are glacial silt-loams or gravels. A few populations were located in wooded swamps dominated by Acer rubrum, Alnus rugosa, Vaccinium corymbosum, and Vibur- num dentatum. Fewer than ten herbaceous species typically occur with Carex on hummocks in these swamps. These species include Dryopteris cristata, Rubus pubescens, Symplocarpos foetidus, and Osmunda cinnamomea. Soils tend to be organic mucky silts. Although these two community types may be assumed to differ in environmental factors such as temperature, humidity, and the avail- ability of light, differences in habitat were not correlated with mor- phological or anatomical differences in the populations of Carex which were studied. Distribution maps of the two taxa (Figure 4) were constructed. Regional distribution maps are also provided by Braun (1967), Hermann (1941), Radford et al. (1968), Voss (1972), and Wheeler (1981). The two morphological forms have different ranges, al- though they are sympatric over a large part of the northeastern United States. Carex crinita occurs from Nova Scotia west to Wis- consin, and south to Georgia and Texas. It is uncommon in the mountains of the Appalachians and Adirondacks. Love and Léve (1981) report a new record of this species from Winnepeg, Mani- toba. Carex gynandra has a more limited geographic distribution. It occurs from Newfoundland west to the northern Great Lakes region, and south in the Appalachians to northern Georgia. This taxon generally occurs at higher elevations than Carex crinita, and is less common on the coastal plain. The taxa are geographically sympatric over most of their range of distribution. Populations in central New York, however, generally consist of only one morphological form. Sympatric populations of both taxa occurred only in two locations within this region, the Slaterville and Freeville populations (Table 1). As the two taxa have the same chromosomal number, they would be expected to be inter- fertile (Faulkner, 1973; Standley, 1981) and to hybridize in the absence of other barriers to interbreeding. An examination of popu- lations in these two localities yielded one individual which appeared to be intermediate in features of the scabrosity of sheaths, shape of the perigynia, and apical shape of the pistillate scales. No achenes developed on this plant, leading to the conclusion that it was a putative hybrid of reduced fertility. 238 Rhodora [Vol. 85 Figure 4. Distribution maps for Carex gynandra and C. crinita. 1983] Standley — Carex crinita & C. gynandra 239 Differences in the phenology of flowering may act as isolating mechanisms in sympatric populations. Observations at the Slater- ville site in May of 1977 indicated that plants with scabrous sheaths (Carex gynandra) initiated flowering at least one week later than plants with glabrous sheaths (C. crinita). This observation requires additional quantification, but suggests that there is little overlap of flowering time. DISCUSSION Carex crinita and C. gynandra do not differ with regard to most features of the morphology of leaves and of inflorescences, the anatomy of leaves, chromosomal numbers, and habitat preferences. Correlated variation in the morphology of perigynia, of pistillate scales, and of the sheaths of leaves is the basis for clustering popula- tions in this study into two distinct taxa. These sets of characters differ in their geographical distribution, although taxa are sympat- ric over part of their range. These patterns of variation are consis- tent with those observed in systematic studies of related species of Carex (Standley, 1981) which are distinguished by morphological differences of the perigynia, scales, and inflorescences. Chromo- somal numbers and foliar anatomy have been found to be most useful in clustering related species. Subspecies are defined as groups of populations which differ in their geographic distribution and which differ, although not consistently, in a few morphological characters. The two taxa distinguished in this study, Carex crinita and C. gynandra, are considered to be distinct yet closely related species in accordance with this model, as they differ consistently in morphology, overlap geographically, and appear to be reproduc- tively isolated through differences in the phenology of flowering. Speculation on the process of divergence of these taxa requires additional evidence on the evolutionary polarity of the character- states which distinguish them. Current studies of the systematics of the Cryptocarpae and Acutae groups will attempt to resolve this question. CONCLUSIONS The study of populations of Carex crinita and C. gynandra in the vicinity of Ithaca, N.Y. has provided information on variation in 240 Rhodora [Vol. 85 morphology, foliar anatomy, habitat, and chromosomal numbers. Details of foliar anatomy of both taxa and the chromosomal number of n = 33 for Carex gynandra are documented for the first time. Species are distinguished by the consistent correlated differ- ences in the morphology of the sheaths of leaves, the perigynia, and the pistillate scales, and appear to be reproductively isolated by differences in the phenology of flowering. This study has demon- strated that problems of species classification in Carex can be resolved through a better understanding of the patterns of variation and reproductive interactions within the genus. ACKNOWLEDGMENTS This research represents part of a master’s thesis submitted to the graduate school of Cornell University. I thank the curators at GH, NY, and TRT for use of collections, and R. T. Clausen, W. J. Dress, H. P. Banks, and C. H. Uhl for help and encouragement at various stages of this study. LITERATURE CITED AIRY-SHAW, H. K. 1973. A dictionary of the flowering plants and ferns. 8th edition. Cambridge University Press. BaILey, L. H. 1886. A preliminary synopsis of the genus Carex. Proc. Amer. Acad Sci. 22. Boott, F. 1858-1867. Illustrations of the genus Carex. London. Wm. Pamplin. Braun, E.L. 1967. The monocotyledonae; Cattails to Orchids. Columbia, Ohio State University Press. FAULKNER, J.S. 1973. Experimental hybridizations of northwest European spe- cies of Carex section Acutae (Cyperaceae). J. Linn. Soc., Bot. 67: 233-253. FERNALD, M.L. 1950. Gray’s Manual of Botany. 8th edition. New York, D. Van Nostrand. GLEASON, H. A., & A. CRONQUIST. 1963. Manual of vascular plants of the North- eastern United States and adjacent Canada. Princeton, Van Nostrand. HERMANN, F. J. - 1941. The genus Carex in Michigan. Amer. Midl. Nat. 25: 1-72. KOUKENTHAL, G. 1909. Cyperaceae: Caricoideae. Das Pflanzenreich (38 Heft) 1V:20. Love, A., & D. Love. 1981. Jn: A. Love, ed. Chromosome number reports LXXIII. Taxon 30: 845-849. MacKeEnzikg, K. K. 1931-1935. Cyperaceae: Cariceae. North American Flora 18: 1-478. RADFORD, A. E., H. E. AHLEs, & C. R. BELL. 1968. Manual of the vascular flora of the Carolinas. Chapel Hill, University of North Carolina Press. 1983] Standley — Carex crinita & C. gynandra 241 SCHWIENITZ, L. D. von. 1824. An analytical table to facilitate the determination of the hitherto observed North American species of the genus Carex. Ann, Lyceum N.Y. 1: 62-71. STANDLEY, L. A. 1981. The systematics of Carex section Carex in the Pacific Northwest. Ph.D. Thesis. University of Washington. STEBBINS, G. L., & B. CROMPTON. 1961. A suggested revision of the grass genera of temperate North America. Recent Advances in Botany 1: 133-145. Voss, E.G. 1972. Michigan Flora. Part |. Gymnosperms and Monocots. Bull. Cranbr. Inst. Sci. 55. WaHL, H. 1940. Chromosome numbers and meiosis in Carex. Amer. J. Bot. 27: 458-470. WEATHERBY, C. A. 1923. The identity of Carex gvnandra Schwein. Rhodora 25: 115-116. WHEELER,G. A. 1981. New records of Carex in Minnesota. Rhodora 83: 119-124. DEPARTMENT OF BIOLOGICAL SCIENCES WELLESLEY COLLEGE WELLESLEY, MA 02181 242 Rhodora [Vol. 85 NOTICE OF PUBLICATION Crow, GARRETT E. 1982. New England’s Rare, Threatened, and Endangered Plants. x+ 129 pp. U.S. Fish & Wildlife Service, Newton Corner, MA This is a comprehensive assessment of 101 of the rarest plant species in New England. Included for plants already listed or under review for Federal listing are: common & scientific names, “listing” status, a description of distinctive features, distribution, habitat, flowering period, endangerment, recommendations, taxonomic com- ments, and selected references. For other plants of national signifi- gance the habitat, distribution, and overall range are given. There are 12 full-color plates, as well as excellent pen and ink drawings of 75 species. This beautifully illustrated, hardbound book is available for $11, including postage and handling, from: Superintendent of Documents U.S. Government Printing Office Washington D.C., 20402 CHROMOSOMES OF MEXICAN SEDUM IV. HETEROPLOIDY IN SEDUM MORANENSE CHARLES H. UHL ABSTRACT At least 18 different chromosome numbers, from n = 21 ton = 153 are reported for 51 collections of Sedum moranense from 45 localities, most of them in east central Mexico. At least ten levels of ploidy, up to 16-ploid, are represented, mostly based on x = 18-21, including probable 5-, 7-, and 9-ploids. Dysploid variants occur at several levels of ploidy, and some other plants have extra chromosomes of standard size; at least six have B-chromosomes. In most collections the bivalents do not differ greatly among themselves in size, but four collections from the Sierra de Guanajuato have strongly asymmetrical karyotypes and also anomalous chromosome numbers (n = 52, 54). In spite of the great variation in their chromosomes, most collections are very similar in their morphology, and, except for the larger flowers of the 14-ploid subspe- cies grandiflorum, no correlation is apparent between morphology and chromosome number. This paper is the fourth in a series reporting the chromosomes of 100 or so species of Mexican Sedum. A general introduction, mate- rials and methods, and acknowledgments are given in the first paper (Uhl, 1976). Sedum moranense is singled out for special atten- tion here because it has the greatest cytological diversity of all the Mexican Crassulaceae, which as a group are distinguished for the great variety in their chromosomes. The Mexican Crassulaceae have an exceptionally broad array of chromosome numbers (heteroploidy), from n = 7 to n = ca. 320 (Uhl, 1972-1980). Polyploidy is common, ranging to at least 20-ploid, and very extensive dysploidy also occurs in some genera and in some groups of related species. In some cases a very broad range of chromosome numbers is found in plants that are very similar morphologically and that seem best regarded as all the same species. Thus construed, some species have four or more levels of polyploidy in different populations and many, perhaps most, wide- spread species have dysploid chromosome races (Uhl, 1972, 1978, 1980, and unpub.). For example, different populations of Sedum wrightii have five different levels of ploidy, n = 12, 24, 36, 48, and 60 (Uhl, 1972, and unpub.). On the other hand, the wide-ranging S. jaliscanum has at least eight different dysploid chromosome num- bers, ranging from n= 11 ton = 34(Uhl, 1976), and Clausen (1981) 243 244 Rhodora [Vol. 85 would also include plants having n = 9 and n = 10 as subspecies angustifolium of this species. The size of the chromosomes in S. jaliscanum varies inversely with their number, and it seems likely that the diversity of its chromosome numbers has resulted from rearrangements of chromosome parts on a massive scale, and that polyploidy is not present. Sedum moranense HBK is a relatively small “typical” repre- sentative of the genus (chamaephyte), with white flowers and small, ovate to lanceolate leaves that are 2-4 mm long and almost as thick as wide, closely set in five or six spiral ranks (Clausen, 1959). Clausen also reported that plants from six populations were “remarkably similar morphologically”, although environmental con- ditions, both in nature and cultivation, can significantly affect such characters as leaf size. One other population, subspecies grandiflo- rum Clausen, had larger flowers. Sedum moranense is one of the commonest and widest ranging species of the Crassulaceae in Mexico. It occurs as disjunct popula- tions, most often on volcanic rock above 2,000 meters, from the southern Sierra Madre Oriental in the states of Hidalgo, Puebla, and Veracruz west to the vicinity of Mexico City and Toluca and northwest to the Sierra de Guanajuato and to the interior moun- tains of San Luis Potosi; it is especially common in the state of Hidalgo (Fig. 1). The species also is common in cultivation in Mex- ico, and the collection reported here from Jalisco may have been introduced there. This paper presents the chromosome numbers of 51 collections of Sedum moranense from 45 populations in all parts of its range (Fig. 1). Collections are listed in Table | in order of increasing chromo- some number and then approximately from west to east and from north to south. At least 18 different chromosome numbers, from n = 21 (Fig. 2) to nm = 153 (Fig. 17) have been found in the species, including apparent 5-, 7-, and 9-ploids but not counting variants with one or a few extra univalents or B-chromosomes. These repres- ent apparent diploids, polyploids of probably nine different levels, and dysploids at several levels of ploidy. Many collections have meiotic irregularities in the form of one or more unpaired chromosomes of standard size at metaphase I and laggards at anaphase I. Most univalents are easily identified by their conspicuously shallower depth in focus than bivalents and multi- JAL Figure |. Central Mexico, showing locations of plants with various gametic chromosome numbers. [€861 AI lWnpas uroIxa JO samosowolyS — 14 SS 6 246 Rhodora [Vol. 85 Fig. 2-17. Chromosomes of Sedum moranense at metaphase I, all X2000. Thin lines in some figures indicate univalents or B-chromosomes. 2, U/557,n= 21; 3, UI871,n=21+ 1; 4, U2353, n= 24, 5, M/13396, n= 38; 6, M7654,n=40; 7, C44-109, n = 50+ 2B; 8, M/4736, n= 52; 9, MI0197,n=54, 10, U2069, n= 57, 11, U830, n= 60+ 3, 12, U2/28,n=62+2; 13,8d,n=63+6;, 14, M64/4, n=77+3; 15, U2487,n= 100; 16, M/4734,n = 140; 17, U1536,n = 153. 1983] Uhl — Chromosomes of Mexican Sedum IV 247 valents and by their strong tendency to lie at the margins or some- times off the plate at metaphase I. In one diploid (U/87/, n=21 +1, Fig. 3) the unpaired chromosome is the largest present; at pachytene a heterochromatic body of about the same size is attached to the nucleolus. Very small extra, unpaired chromosomes, apparently B-chromosomes, are present in at least six collections. One diploid from Hidalgo (U/870, n = 21 + 1), two probable heptaploids from Tlaxcala (7d and 8d), and one presumed octoploid from Hidalgo (U2707, n= 72 + 1) showed one or more chromosome bridges (up to 4) in most cells at anaphase I, presumably indicating heterozygosity for paracentric inversions. The basic or ancestral chromosome number in Sedum moranense appears to be in the range of 18 to 21. Most polyploids have multi- ples of one or another of these numbers, up to 14—ploids with n = 140 in subspecies grandiflorum Clausen (Fig. 16) and a presumed 16—ploid (U/536) with n = 153 or very close to that (Fig. 17). Some multivalent associations are present in most cells of polyploids, and these make exact counts difficult (e.g., Fig. 10-17). Counts of plants having multivalents are reported here in terms of their equivalents as bivalents and are believed to be accurate within 2% at all levels of ploidy. In each polyploid most cells at metaphase I generally produce a fairly constant number of bivalents, or their equivalents when multi- valents are allowed for. However, especially in many of the higher polyploids, additional univalents of standard size are common at metaphase I. The number of these univalents is not constant in the cells of the same anther, which probably means that the extra chromosomes sometimes participate in forming multivalents and therefore that they are extra representatives of standard chromo- somes of the genome. The species seems to be remarkably tolerant of chromosomal variation and unbalanced genetic dosages, espe- cially at the polyploid level. Doubtless this tolerance has made pos- sible the great variation that is found in the chromosomes of different populations. One collection from Hidalgo (U2343) regularly had 40 paired elements (bivalents and multivalents) but also had 5-11 univalents of standard size and may represent a pentaploid based on x = 20. Two plants from the same population in Tlaxcala (7d and 8d) had 60-62 paired elements plus 5-9 univalents (Fig. 13), and a plant from just 248 n=2I1 n= 24 n= 38 n= 40 Probable pentaploid n= 50 n= 52 n= 54 n= 57 Rhodora [Vol. 85 Table 1. Collections Studied. U2129 Qro. | km. W. of San Joaquin, 2360 m. U1870 Hgo. E. of Cuesta Colorada, 13 km. N. of Jacala, 1700 m. (1 = 21 + 1) U1871 Hgo. | km. N. of Minas Viejas, 2000 m. (n = 21 + 1) U1950 Hgo. | km. S. of Minas Viejas, 2040 m. U1557 Hgo. 11 km. E. of Tulancingo U2353 Pue. 10 km. SW of Aquixtla, 2750 m. U2708 Hgo. Ca. | km. N. of El Salto, km. 253 (R. T. Clausen CM P- ES) M13396 Hgo. Near summit of mountain SE of Santuario, 2560 m. (R. Moran and C.H. Uhl) C47-43 Hgo. Ca. 20 km. E. of Pachuca, 2320 m. (E.J. Alexander, via R.T. Clausen) C47-28 Tlax. Barranca near San Bernabe Amaxac de Guererro (Ibid.) M7654 SLP. Zaragoza (R. Moran) U2705 SLP. 5 km. NW of La Salitrera (R. T. Clausen 77-40) M7809 Hgo. San Vicente, ca. 3km. S. of Durango (R. Moran) (n = 40 + 3B) M7781 Hgo. Pachuca (R. Moran) (7 = 40 + | + 5B) U618, UC44-110 Cultivated U2343 Hgo. 4.3 km. S. of Durango, 2130 m. (n = 40 + 5-11) FL1419 Cultivated (n = 50 + 1B) C44-109 Cultivated (n = 50 + 2B) M14736 Gto. Santa Rosa, 2590 m. (R. Moran) U2288 Gto. 3 km. SW of Mesa San Jose and 22 km. NE of Guanajuato, 2360 m. U2267 Gto. 10 km. NE of Guanajuato, 2560 m. M10197 Gto. Near Santa Rosa (R. Moran) (n = 54+ IB) U2069 Jal. SW edge of Tapalpa, 2130 m. (F. C. Boutin & M. Kimnach) U1952 Qro. 5km. S. of Pinal de Amoles, 2560 m. U1854 Qro. 8.5 km. S. of Pinal de Amoles, 2600 m. U1476 Hgo. Presa Madero, 12 km. W. of Huichapan U1484, { Edo. Mex. 10.5 km. WNW of Soyaniquilpan (N.W. Uhl & W. U1485 Handlos) U1481 Hgo. 7.5 km. SE. of Tepeji del Rio (N.W. Uhl & W. Handlos) U1877 Hgo. 2 km. S. of Tepenene U1466 Hgo. Near Pueblo Nuevo C47-44 Pue. Near Honey. (H. Xolocotzi, via E. J. Alexander & R. T. Clausen) U1554 Pue. 22 km. NW of Zacatlan U1418 Edo. Mex. Hill N. of Toluca, 2740 m. 6d Ver. Near Las Vigas (R. T. Clausen) (n = 57 + 3) U1884 Ver. 27 km. NW of Jalapa, 2290 m. 1983] Uhl — Chromosomes of Mexican Sedum 1V 249 n= 60 U830 Hgo. El Chico, 3000 m. (H. E. Moore) (n = 60 + 2) n=63 2128 Qro. 13 km. W. of San Joaquin, 2330 m. (n = 63 + 1) U/I556 Pue. Puente Totolapa, 11 km. W. of Huauchinango U2706 Edo. Mex. El Zarco, Sierra de las Cruces (R. T. Clausen 55-169) Probable 7d, 8d Tlax. Atlihuitzia (R. T. Clausen) (7 = 60-62 + 5-9) heptaploids. n=72 U2707 Hgo. Cerro Jardin, W. slope, 2480 m. (R. T. Clausen 76-47) n=77 M6414 Hgo. Canon Venados (R. Moran) (n = 77 + 3) Probable U/548 Hgo. 2 km. S. of Epazoyucan (n = 80 + 10) 9-ploid n= 100 U2487 Pue. | km. W. of Nicolas Bravo, 2600 m. (one clump only) n= 140 10d SLP. Mountains (Sierra de Alvarez) E. of San Luis Potosi (R. T. Clausen) Clonotype of S. moranense ssp. grandiflorum) U1641 SLP. 2 km. W. of highway summit, Sierra de Alvarez, 2260 m. (W. Handlos) M14734 Gto. Picachos de la Bufa, cliffs NE of Guanajuato, 2290 m. (R. Moran) n= 153 U1536 SLP. 16 km. N. of Ahualulco west of Mexico City (U2706) formed about 57 paired elements and 5-8 univalents; these three are believed to be heptaploids based on x = 19-21. Another plant from Hidalgo (U/548) had probably 80 paired elements plus about 10 univalents and is thought to be 9- ploid. Presumably the number of chromosomes of the last (odd) set that participate in the formation of multivalents is variable, which leaves a variable number of them as univalents in different cells. In most plants the bivalent chromosomes at metaphase I do not differ greatly among themselves in size (Fig. 2, 5, 6, etc.), and any extra univalents that may be present also appear to be of the same size. However, four collections from the Sierra de Guanajuato dif- fered strikingly from nearly all other collections in having a wide range of sizes among their 52-54 bivalents at metaphase I (Fig. 8, 9). This condition (heterogeneous or asymmetrical karyotype) is gener- ally believed to have been derived by unequal translocations from ancestors having chromosomes more nearly equal in size (homo- geneous or symmetrical karyotype) (Stebbins, 1971). The numeri- cally anomalous diploid from Puebla (U2353, n = 24) also has bivalents that differ more than usual in size (Fig. 4), although not as much as they do in the plants from Guanajuato. Two collections with n = 50, both of unknown provenance from the wild, have 250 Rhodora [Vol. 85 bivalents that are more or less equal in size, although their | or 2 extra chromosomes are tiny and interpreted as B-chromosomes (Fig. 7). With so many chromosome races and so many levels of ploidy, some of them occurring close to each other, it is probable that considerable intercrossing has occurred between plants having dif- ferent chromosome numbers and that such intercrossing has been a source of still more cytological variation. The evident viability and tolerance of unbalanced chromosome dosages would make survival possible until new, more or less stable chromosome complements evolved. However, the phenotypic effects of this sort of variation, including some apparent cases of unbalanced dosages of chromo- somes, appear to be negligible. The effects, if any, on fertility or adaptation to the particular habitat are not clear. It is possible that some apomixis occurs here, but as yet there is no direct evidence for it. Qualitative and quantitative chemotaxonomic studies on these plants might prove interesting. According to Clausen (1959) Sedum liebmannianum of the Sierra Madre del Sur is the species most closely related to S. moranense, but it differs conspicuously in its persistent, withered leaf bases. Two collections of S. /iebmannianum had n = 34, but seven other collections were very irregular at meiosis and probably odd-ploid. A similar plant from Puebla with smaller leaves (U2358) had n = 30. Clausen (1959) also cited two other species as related to S. mora- nense: S. cupressoides (n = 29) of the Sierra Madre del Sur and S. parvum (n = 32, 34, 64) of northeastern Mexico, to which he later (1978) also reduced S. nanifolium (n = 26, 52). Chromosome numbers of these species (all from Uhl, unpublished) do not support any conclusion that they are closely related to S. moranense. Clausen (1959) studied a plant said to have come from high on the Nevado de Toluca that he thought might be a hybrid between Sedum moranense and Villadia batesii. The plant is intermediate in morphology and produces abortive pollen and ovules. Eight cells at metaphase I had 24 to 27 bivalents and multivalents plus 3 to 14 univalents, and the microspores after meiosis differ among them- selves in size and stainability. Nineteen collections of V. batesii from Puebla and Hidalgo to Michoacan, including one from the Nevado de Toluca, all had n = 25, which is considered to represent a diploid here (Uhl, unpublished). However, no collection of S. moranense 1983] Uhl — Chromosomes of Mexican Sedum 1V 251 from the Nevado de Toluca has been studied, and Clausen (1959) does not cite this as a locality for the species. The nearest S. mora- nense that has been studied came from just north of the city of Toluca, about 24km. north of the volcano, and had n = 57, consi- dered hexaploid. Clausen’s plant appears clearly to be a hybrid, but its chromosomes and its location make it very doubtful that S. moranense could have been one of its parents. Clausen (1981) suggested that Sedum furfuraceum (n = 34, Uhl, 1980), which is known from only one population in San Luis Potosi, might have originated at that site as a hybrid between Pachyphytum hookeri and S. moranense, both of which occur with it there. How- ever, S. moranense there (M7654) is tetraploid (n = 40), and P. hookeri there is decaploid (n = ca. 160, Uhl and Moran, 1973), and this suggested relationship seems most unlikely. Many hundreds of hybrids have been produced in cultivation within and between Sedum and various other Mexican genera. However, only one cross has been attempted with S. moranense, U1642 (S. retusum, n = 27, Uhl, 1980) X U/557 (S. moranense, n = 21), and this was unsuccessful. Analysis of chromosome pairing in hybrids of S. moranense having different chromosome numbers could allow their levels of ploidy to be identified, as has been done with Echeveria (Uhl, 1982ab). LITERATURE CITED CLAUSEN, R. T. 1959. Sedum of the Trans-Mexican Volcanic Belt. Cornell Univ. Press, Ithaca, NY. 1978. Sedum—seven Mexican perennial species. Bull. Torrey Bot. Club 105: 214-223. 1981. Variation of Species of Sedum of the Mexican Cordilleran Plateau. Ithaca, NY. STEBBINS, G.L. 1971. Chromosomal Evolution in Higher Plants. Edward Arnold, London. Unt, C. H. 1972. Intraspecific variation in chromosomes of Sedum in the south- western United States. Rhodora 74: 301-320. 1976. Chromosomes of Mexican Sedum I. Annual and biennial species. Rhodora 78: 629-640. 1978. Chromosomes of Mexican Sedum II. Section Pachysedum. Rho- dora 80: 491-512. 1980. Chromosomes of Mexican Sedum III. Sections Centripetalia, Fruticisedum and other woody species. Rhodora 82: 377-402. 1982a. The problem of ploidy in Echeveria (Crassulaceae) I. Diploidy in E. ciliata. Amer. J. Bot. 69: 843-854. 252 Rhodora [Vol. 85 1982b. The problem of ploidy in Echeveria (Crassulaceae) II. Tetraploidy in E. secunda. Amer. J. Bot. 69: 1497-1511. & Reip Moran. 1973. Chromosomes of Pachyphytum (Crassulaceae). Amer. J. Bot. 60: 648-656. SECTION OF PLANT BIOLOGY CORNELL UNIVERSITY ITHACA, NY 14853 PATCHES, CLONES AND SELFP-FERTILITY OF MAYAPPLES (PODOPHYLLUM PELTATUM L.) DaviD POLICANSKY'! ABSTRACT Mayapples (Podophy/lum peltatum L.) grow in patches on the forest floor. Con- trolled pollinations within and between patches, as well as self-pollinations, indicate that the mayapple is self-sterile in only part of its range, and that in this part of the range, the patches are not single clones, i.e. they consist of more than one genotype. The mayapple, Podophyllum peltatum L., is acommon rhizoma- tous, perennial herb in deciduous forests of eastern North America. It grows in well-defined patches over much of its range, and the form of these patches suggesis a clonal development. The patches have been called clones by Rust and Roth (1981), Taylor (1974), and Krochmal, et al. (1974), but those authors did not attempt to inves- tigate the population structure of the patches. Since there is consid- erable theoretical interest in the structure of populations of asexually reproducing herbs (e.g. Williams, 1975), | wanted to know whether mayapple patches are in general composed of only one genotype, or whether there is more than one genotype per patch. Mayapples proliferate shoots asexually (Sohn & Policansky, 1977; Holm, 1899), and thus there are fewer genotypes than shoots in a mayapple patch. Given seii-sterility, one would expect fewer fertile crosses within patches than between them, because some crosses would be between members of the same genotype. If the patches were single clones then all intra-patch crosses would be sterile. In this paper I report the results of such experimental crosses. METHODS I studied mayapples in Bowers Woods, north of Valparaiso, Indi- ana, in 1976; in the Institute for Advanced Studies, in Princeton, New Jersey, in 1976; in Oak Ridge, Tennessee, in 1975 and 1976; and at the Case Estates of the Arnold Arboretum, Weston, Massa- chusetts, in 1975 and 1981. At the Case Estates there is one very 'Current address: National Research Council, Commission on Life Sciences, 2101 Constitution Avenue, N.W., Washington, D.C. 20418. 253 254 Rhodora [Vol. 85 large, poorly-defined patch of probably introduced mayapples. This patch is bisected by a footpath. At the other sites there are numer- ous well-defined patches of various sizes. Flower buds were covered with numbered, brown paper bags. At anthesis, pollinations were made between shoots within and between patches, as well as self pollinations. Following the pollinations the bags were replaced. Some flowers were left bagged with no treat- ment throughout the experiments; others were destaminated, and left bagged. Later in the summer the fruits were examined. At the Case Estates in Massachusetts in 1981 1 counted the seeds in each fruit; in the other sites only presence or absence of seeds and fruit size were recorded. There is a good correlation between seed number and fruit size (Sohn & Policansky, 1977). RESULTS Some of the experiments in Oak Ridge, and those in Bowers Woods did not yield clear results. Some of the selfed plants pro- duced a few seeds, but most of them and many of the outcrossed plants did not. It appears that self-sterility is not complete at those sites. Clear results were obtained in one Oak Ridge site, in Weston, and Princeton; they are presented for comparison in Table 1. There was some degree of self-fertility in the populations in Wes- ton and Oak Ridge (Table 1). Additionally, in Weston in 1981, of 12 selfed flowers, 2 produced | seed each, and | produced 18 seeds. Four outcrossed flowers produced 0, 2, 46, and 53 seeds. Of 99 unbagged flowers, 63 failed to set seed, and of the 36 that did, the number of seeds ranged from | to 70, with a mean of 26.3. Table |. Proportion of experimental pollinations resulting in production of at least one seed. CROSSED LOCALITY CROSSED BETWEEN AND DATE SELFED WITHIN PATCH PATCHES Oak Ridge, 1975 4/6 7/10 — Weston, 1975 3/23 8/10 — Weston, 1981 3/12 3/4 _ Princeton, 1975 0/46 8/27 12/12 1983] Policansky — Mayapples Pay) In Princeton three clearly defined patches showed complete self- sterility. Of 13, 15, and 18 flowers selfed in the three patches none set any seeds, while 2 out of 6, | out of 11 and 5 out of 10 flowers outcrossed within the patches set seeds. The difference in proportion of flowers setting seed between selfed and outcrossed flowers is significant by chi-square test at the .01 level. Four flowers from each patch were cross-pollinated with 4 flowers from adjacent paches; all of these set seed. None of the destaminated flowers or those that were bagged with- out pollination set any seeds at any of the sites. The proportion of flowers setting seed in untreated populations ranged from zero ina central Tennessee population in 1975 to around 80% in Princeton in 1975 and 1976. DISCUSSION The analysis of patches in Princeton demonstrates that they are not single clones. Although self-pollinations never resulted in seeds, pol- linations within patches sometimes did, and pollinations between patches always did (Table 1). It is impossible to say how many genotypes there were in each of the three patches. There may have been only two, but there were definitely more than one. There appears to be considerable variability in the degree of self- fertility over the range of mayapples. The results of this study indicate that the mayapple is not completely self-sterile over all of its range, but that it is in Princeton. Recently Swanson and Sohmer (1976) made some inter- and intrapopulation crosses of mayapples in Wis- consin, to test the hypothesis that interpopulation crosses would be more fertile than intrapopulation crosses. This hypothesis was con- firmed. They did not report the results of any self-pollinations, but the low fertility of their intrapopulation crosses in Wisconsin suggests that the plants are probably self-sterile there also. Williams (1975) presented a model for the evolution of sex in asexually reproducing plants, such as the strawberry. He assumed that single clones formed patches of shoots that were better adapted to their own local environment, due to natural selection, than other clones in the general area. Mayapples grow very much like strawber- ries, and thus Williams’ model applies to them also. The demonstra- tion that at least some of the patches have more than one genotype means that one of the assumptions of the model may not be generally applicable. 256 Rhodora [Vol. 85 ACKNOWLEDGMENTS I thank Harvard University and the Institute for Advanced Stud- ies; Princeton, for access to study sites, and B. Dempsey, J. Endler, J. Sohn, and F. Taylor for discussions and field assistance. LITERATURE CITED Hoim,T. 1899. Podophyllum peliatum, a morphological study. Bot. Gaz. Chicago 27: 419-433. KROCHMAL, A., L. WILKINS, D. VAN LEAR, & M. CuieN. 1974. Mayapple. U.S. Forest Serv. Res. Pap. NE=296. 9 pp. Rust, R.W.,& R.R.RotH. 1981. Seed production and seedling establishment in the mayapple, Podophylium peltatum L. Amer. Mid]. Nat. 105: 51-60. Soun, J. J., & D. PoLicansky. 1977. The costs of reproduction in the mayapple Podophyllum peltatum (Berberidaceae). Ecology 58: 1366-1374. Swanson, S., & S. H. SOHMER. 1976. Reproductive biology of Podophyllum peltatum (Berberidaceae): the comparative fertility of inter- and intrapopula- tional crosses. Wisc. Acad. Sci., Arts & Lett. 64: 109-114. Taytor,F.G. 1974. Phenodynamics of production ina mesic deciduous forest, p. 237-254 In: H. Lieth (ed.) Phenology and seasonality modeling. Springer, NY. WituiaMs,G.C. 1975. Sex and evolution. Princeton University Press, Princeton, NJ. GRAY HERBARIUM OF HARVARD UNIVERSITY 22 DIVINITY AVENUE CAMBRIDGE, MASSACHUSETTS 02138 CHROMOSOMAL TYPIFICATION OF SISYRINCHIUM BERMUDIANA L. (IRIDACEAE)! L. MICHAEL HILL Literature reports for the chromosome number of Sisyrinchium bermudiana L. are 2n = 96 (Oliver & Lewis, 1962), 2” = 82, 84, 88 and 90 (Ingram, 1967), 27 = 80 and 96 (Ingram, 1968) and 2n = 32, 64 and 96 (Mosquin, 1970). Geographical data from voucher speci- mens which produced these numbers are from North America and Europe. While variation in chromosome number of a plant species is not uncommon, these data are perplexing in the light of the statements by Hemsley (1884), Bicknell (1896), and Britton (1918) that the name S. bermudiana L. should be applied only to the Bermudas. The latter author believed the species to be endemic to those islands. Ward’s (1968) thorough nomenclatural study revealed that S. bermudiana L. is applied only to the Bermuda population and then listed five names of northeastern North American species (S. angustifolium Mill, S. mucronatum Michx., S. arenicola Bickn., S. montanum Greene, and S. at/lanticum Bickn.) which were mor- phologically differentiated from each other, as well as S. bermudi- ana L. It therefore becomes necessary to determine the chromosome number of Sisyrinchium bermudiana L. in its probable singular location: Bermuda. Flower buds from several populations were sampled and cytologically examined using an aceto-orcein tech- nique described elsewhere (Hill & Rogers, 1970). The preparations were studied under oil at 1000 magnification using a Zeiss phase contrast microscope. Camera lucida drawings have been attached to the herbarium sheets of voucher specimens deposited in the herba- rium of Bridgewater College (BDWR). The chromosome number of all populations was n = 32. Sisyrinchium bermudiana in Bermuda easily shows morphologi- cal differentiation from the eastern North American species dis- cussed by Ward (1968). The taxon has these distinguishing charac- teristics: a large flower (17-20 mm in diameter), a stout stem with wings 2-3mm wide, leaves 5-7 mm wide and as stout as the stem, 'This paper represents contribution number 894 from the Bermuda Biological Sta- tion for Research, Ferry Reach I-15 West, Bermuda. 257 258 Rhodora [Vol. 85 and a prominent node which is at the base of both foliaceous and spathodal bracts. A re-examination of the vouchers cited by Oliver and Lewis (1962) and Mosquin (1970) is now in progress as part ofa continuing study of chromosome numbers of eastern North Ameri- can taxa of Sisvrinchium. This study was supported by NSF grant DEB-8008808 through the Bermuda Biological Station for Research and NSF grant TFI-8016238. LITERATURE CITED BICKNELL, E. P. 1896. The Blue-eyed grasses of the eastern United States (genus Sisyrinchium). Bull. Torrey Bot. Club 23: 130-137. Britton, N.L. 1918. Flora of Bermuda. Hafner Press, New York. HEMSLEY, W. B. 1884. Sisyrinchium bermudiana. J. Bot. 22: 108-110. Hint, L. MicHagL, & O. M. RoGeRS. 1970. Chromosome numbers of Aster bla- kei and A. nemoralis. Rhodora 72: 437-438. INGRAM, R. 1967. On the identity of Irish populations of Sisyrinchium. Watsonia 6: 283-289. 1968. Breeding barriers in some species of Sisyrinchium. New Phytol. 67: 197-204. Mosquin, T. 1970. Chromosome numbers and a proposal for classification in Sisyrinchium (Iridaceae). Madrono 20: 269-275. Oviver, Royce L., & WALTER H. Lewis. 1962. Chromosome numbers of Sisyrin- chium (Iridaceae) in Eastern North America. Sida 1(1); 42-48, Warp, D. B. 1968. The nomenclature of Sisyrinchium bermudiana and related North American Species. Taxon 17: 270-276. BIOLOGY DEPARTMENT BRIDGEWATER COLLEGE BRIDGEWATER, VIRGINIA 22812 THE TYPE LOCALITY OF SEDUM PUSILLUM MICHX. AND SENECIO TOMENTOSUS MICHX. LEONARD J. UTTAL The type locality of Sedum pusillum Michx. is cited in Michaux (1803) as “HAB. in Carolina septentrionali, loco dicto Flat-rock.”, that of Senecio tomentosus Michx. as “HAB. in Carolinae loco dicto Flat.-Roc.”. In the gazetteer of the introduction to the facsim- ile edition of the Michaux flora, Ewan (1974) identifies “Flat-rock” as near Hendersonville, Henderson County, North Carolina. This Flat Rock is a well-known resort village of the North Carolina mountains but it is really not the type locality of these two species. Neither occurs in the Appalachian highlands. Sedum pusillum is an endemic of piedmont granite outcrops of South Carolina with one station in adjacent North Carolina. Senecio tomentosus, while mainly a species of the southeast United States coastal plain, also occurs on piedmont rocks in South Carolina and Georgia. The true type locality of these species is another Flat Rock, in northern Ker- shaw County, South Carolina, not a community but a huge granite outcrop, now a quarry producing a beautiful monument stone highly esteemed in the industry. It is crossed by State Road 58, “Flat Rock Road”, 8.3 miles south of Heath Springs, Lancaster County, South Carolina (one-half mile south of the north Kershaw County; boundary). It is marked on the Kershaw County road map and on the Camden quadrangle topographic map. Highway 58 is the origi- nal main road between Camden and Lancaster, frequently travelled in colonial and early republican days, known then as “The Catawba Path”. Flat Rock and its neighboring crag to the north, Hanging Rock, were landmarks on this road. Michaux trvelled this road often en route between his headquarters in Charleston and regions in the mountains and further. Early geography of this region of South Carolina was gleaned from Hooker (1953), while Michaux’s routes were extracted from his journals as published by Sargent (1889). There is no evidence in his journals that Michaux ever passed through Flat Rock, North Carolina, but there are two entries which indicate he passed the Flat Rock of South Carolina. On 23 April 1795, he passed “par Flat rock, par Hanging rock Creek et couché a 259 260 Rhodora [Vol. 85 Cane Creek, Lancaster County. ..”. He passed early in the day, and there are no botanical notes for this date. On 6 April 1796, en route from Illinois to Charleston, Michaux passed the night near Hanging Rock and on 7 April he proceeded to Camden, collecting a little along the way. On this route he would have to pass Flat Rock. | believe he obtained the specimens of Sedum pusillum and Senecio tomentosus on this leg of his journey. The label of the specimen of Sedum pusillum flatly states “Flat rock” in Michaux’s hand. The plants are in prime anthesis, as they would be on 7 April. The label of Senecio tomentosus says “Sur un Rocher, Flat roc or hanging roc”. The languages are mixed. The specimen is mostly in bud; only a few of the lower heads have expanded rays. It is in a state of pre-anthesis as would be expected on that date. Thus it is believed that 7 April 1796 is the date of collection of both type specimens. Both are in p and were seen as photographs in the IDC Microfiche set 6211. It is generally accepted that Michaux, while given full credit for the Flora boreali-americana, did not write it. L. C. M. Richard is generally believed to be the anonymous author (Stafleu and Cowan, 1981). The variation in stated habitat between written label and printed citation in this case is not the only one in the Michaux flora. Kershaw, the type locality county of Senecio tomentosus, is not dotted on the map for this species in Radford, Ahles, and Bell (1968). A dot should also be placed on the map for Pickens County as the species was collected from Table Rock by Buckley in 1843-44 (GH). These are two resumed counties of record for South Carolina for this species, recently overlooked. On 19 and 20 April 1982 I visited the Flat Rock of South Carolina and two other granite bosses of that state long popular with field botanists. I did not find either type species discussed in this paper at Flat Rock. The extensive quarrying seems to have eradicated them as well as most of the other vegetation. | doubt if topotypes will ever be found. Outside the quarried area I did collect a specimen of Senecio anonymus Wood heavily suffused with the lanate tomen- tum usually associated with S. tomentosus. It might be a persistant introgressant, suggesting S. tomentosus had been here. Hybridiza- tion between these species occasionally occurs on the granites of South Carolina and Georgia. 1983] Uttal — Type Locality 261 The other bosses visited were 40-Acre Rock, in Lancaster County, and the huge one at the base of Caesar’s Head, in Greenville County. The endemic flora and other plants of the vernal pools and mossy pockets of these bosses are seriously endangered by heavy human pressure. The habitat is very fragile. | suppose this is true of other large outcrops I was not able to visit. It is hoped efforts will be made to preserve some of these unique areas and their threatened floral elements. ACKNOWLEDGMENTS The author wishes to thank the foreman of the Flat Rock Quarry for permission to explore the property, for useful local geographical information and for his personal botanical observations. The cura- tors of GH and Mo are thanked for the loan of specimens used in this study. Financial assistance was received gratefully from the Virginia Flora Committee. LITERATURE CITED Ewan, J. 1974. Introduction to facsimile edition of Flora boreali-americana by A. Michaux. Hooker, R. J.,ed. 1953. The Carolina backcountry on the eve of the revolution. Chapel Hill. MicHaux, A. 1803. Flora boreali-americana. Paris and Strasbourg. Raprorp, A. E., H. E. AHLES, & C. R. BELL. 1968. Manual of the vasular flora of the Carolinas. Chapel Hill. SARGENT, C.S. 1889. Portions of the journal of André Michaux. Proc. Am. Phil. Soc. 26: 1-145. STAFLEU, F. A. $ R. S. Cowan. 1981. Taxonomic Literature. II: Lh-O. The Hague. DEPARTMENT OF BIOLOGY VIRGINIA POLYTECHNIC INSTITUTE AND STATE UNIVERSITY BLACKSBURG, VA 24061 PHYSOSTEGIA VIRGINIANA VAR. ARENARIA SHIMEK, AN OVERLOOKED NAME PHILIP D. CANTINO AND THOMAS G, LAMMERS In a recent monograph of the genus Physostegia (Cantino, 1982), an obscure but validly published name was overlooked. Physostegia virginiana (L.) Benth. var. arenaria Shimek, Bull. Lab. Nat. Hist. lowa State Univ. 6 (2): 223 (1911), was described as a new variety in a footnote to an ecological treatise on Iowa prairies. There appears to have been only one mention of the name in print since its initial publication (Guldner, 1960). Even Shimek never again used it, either in print or on herbarium labels. From the description, Shimek’s variety is clearly referable to Phy- sostegia virginiana (L.) Benth. ssp. praemorsa (Shinners) Cantino and should be placed as a synonym under that taxon. However, if this subspecies were to be recognized at the varietal level, Shimek’s name would have priority. Because Shimek did not designate a type, a lectotype must be selected. The type locality is reported in the protologue as “the sand mound in Muscatine and Louisa counties [lowa] south of Musca- tine.” This locality, known today as Big Sand Mound, has been discussed by Brown and Brown (1939) and Guldner (1960). Only four Shimek collections of Physostegia virginiana ssp. praemorsa from this locality are known to us. None of these are deposited at IA, where Shimek’s personal collections are housed. Two of the speci- mens were collected after the publication of var. arenaria. Of the remaining two, the one that corresponds best to the original descrip- tion is here designated as the lectotype of Physostegia virginiana var. arenaria Shimek: lowa, Louisa Co., near Fruitland, sand mound, 25 September 1909, B. Shimek s.n., DAO! LITERATURE CITED Brown, M.E., & Brown, R.G. 1939 A preliminary list of the plants of the sand mound of Muscatine and Louisa Counties, lowa. Proc. lowa Acad. Sci. 46: 167-178. 263 264 Rhodora [Vol. 85 CANTINO, P.D. 1982 A mongraph of the genus PAysostegia (Labiatae). Contr. Gray Herb. 211: 1-105. GuLpner, L. F. 1960. The vascular plants of Scott and Muscatine Counties. Davenport Public Museum Bot. Publ. 1. P.D.C. DEPARTMENT OF BOTANY OHIO UNIVERSITY ATHENS, OHIO 45701 T.G.L. DEPARTMENT OF BOTANY THE OHIO STATE UNIVESITY COLUMBUS, OHIO 43210 NEW ENGLAND NOTES CAREX KOBOMUGI OWHI, AN ADVENTIVE SEDGE NEW TO NEW ENGLAND LisA A. STANDLEY Carex kobomugi Owhi is a dioecious sand-dune species native to Japan which was first collected in North America by C. W. Town- send in 1929 (Fernald, 1930) and was subsequently discovered on Cape Henry, Virginia, by Frank Egler in 1941. Fernald (1950) reported that it occurred from Ocean County, N.J., to eastern Vir- ginia, although the only collections in herbaria (GH or Ny) are from these two localities. Five additional populations from Virginia, and three at Sandy Hook, N.J. are currently known (Stalter, 1980). Naturalized populations of Carex kobomugi also occur in New Eng- land, at a distance of more than 200 miles north of the New Jersey population. Svenson (1979) reported finding this species in Fal- mouth, Massachusetts, ca 2 miles inland, in a disturbed sand pit (pers. comm.). Carex kobomugi was discovered in Rhode Island in early January, 1982, by Mr. Richard Champlin, and was checked by me on January 25th. This population is located on East Beach, Charleston, Rhode Island, in the Ninigret Conservation Area (Quo- nochontaug Quadrangle) | /2 mile east of the parking lot. The main population occupies an area of about 170 m? in a blow-out in the primary dune. Both staminate and pistillate plants are present. Scat- tered plants were found growing in loose sand inshore from the blow-out, and also at about 1/4 mile east of the parking lot. A few individuals of Ammophila breviligulata and Cakile edentula oc- curred near the margins of the population. The plants are quite conspicuous, with coriaceous yellow-green leaves that are 4-6 mm wide, up to 20 cm long, and somewhat contorted. The inflorescence of tightly clustered unisexual spikes forms an oblong head 4-6 cm long which is terminal on a trique- trous culm up to 20 cm tall. The erect, plano-convex perigynia with winged margins and dorsalsuture place Carex kobomugi in subge- nus Vignea, although its three stigmas are anomalous in this group. Carex kobomugi and the closely related species C. macrocephala Willd., which occurs along the Pacific Coast from northern Japan 265 266 Rhodora [Vol. 85 to Oregon, comprise section Macrocephalae Kiikenthal. These were not recognized as distinct species in the monographic treatments of Kiikenthal (1909) or MacKenzie (1935), despite the numerous, con- sistent, morphological differences between the western North Amer- ican/northern Japanese and the southern Japanese populations (Table 1). Stacey (1937) clarified the taxonomic history of these two taxa, and confirmed that Carex kobomugi, rather than C. macroce- phala, occurred in eastern North America. Akiyama (1939) described the differences in foliar anatomy which distinguish these taxa. Leaves of Carex kobomugi have a conspicu- ous multilayered adaxial hypodermis which is not found in C. macrocephala (or, to my knowledge, any other species of Carex). Leaves of plants from Seaside Park, N.J. and Charleston, Rhode Island have the multiple hypodermis characteristic of Carex kobo- mugi in Japan. Observations of the East Beach population suggest that Carex kobomugi is able successfully to invade disturbed areas of the pri- Table 1. Morphological differences separating Carex kobomugi and C. macrocephala. CHARACTER C. kobomugi C. macrocephala Perigynium orientation erect reflexed shape elliptical ovate shape of base actute cordate margins entire lacerate number of dorsal nerves ca 20 10-15 number of ventral nerves ca 20 6-9 Shape of apex of pistillate scales serrate entire Achenes length 4-7 mm 3-4.5 mm shape oblong, not elliptical, contorted contorted 1983] New England Notes 267 mary dunes which result, in many cases, from the destruction of Ammophila by trampling. Small (1954) suggested that this species, once established, is able to out-compete Ammophila. It would be interesting to evaluate the potential of Carex kobomugi to stabilize disturbed dunes, as it appears to survive in sites where Ammophila does not, and may provide more resistance to erosion by wind than does Ammophila. The range extension of this species may be coin- cident with increased disturbance of the coastal dunes. LITERATURE CITED AKIYAMA, S. 1939. On the systematic anatomy of the leaves of some Japanese Carices. XIX. Bot. Mag. (Tokyo) 53: 114-118. FERNALD, M. L. 1930. Carex macrocephala and C. anthericoides. Rhodora 32: 9-11. 1950. Gray’s Manual of Botany. 8th edition. D. Van Nostrand. New York. SMALL, J. A. 1954. Carex kobomugi at Island Beach, New Jersey. Ecology 35: 289-291. Stacey, J. W. 1937. Notes on Carex IX. Leafl. Western Botany 2: 30-31. STALTER, R. 1980. Carex kobomugi at Sandy Hook, New Jersey. Bull. Torrey Bot. Club 107: 431-432. Svenson, H. K. 1979. The Flora of Cape Cod. Cape Cod Museum of Natural History. DEPARTMENT OF BIOLOGICAL SCIENCES WELLESLEY COLLEGE WELLESLEY, MA. 02181 NOTE ON THE STATUS OF AGALINIS MARITIMA (RAF.) RAF. IN MAINE BARBARA ST. JOHN VICKERY AND PETER D. VICKERY Agalinis maritima (Raf.) Raf. (=Gerardia maritima) occurs on salt marshes along the eastern seaboard from Nova Scotia to Flor- ida, Mexico, and the West Indies. It was included in Rare and Endangered Vascular Plant Species in Maine (Eastman, 1978) because there were only three historical Maine records (Wells 1880, Wells 1916, Alna 1966) and there was little contemporary knowl- edge of the status of the species in the state. In 1981 the Alna site 268 Rhodora [Vol. 85 was relocated and evaluated by the Maine Critical Areas Program staff. The area was registered as a critical area because it was then the only known station extant in Maine. However, the Checklist of Vascular Plants of Maine (Bean, et al., 1966) lists Gerardia mari- tima from Washington and Cumberland Counties as well as York County. In addition, there were unvouchered reports of A. maritima present at a number of other localities along the southern Maine coast: Kennebunkport; Winnegance Creek, Phippsburg; Morse River marsh, Phippsburg; Bald Head, Georgetown; Reid State Park, Georgetown; Ocean Park, Old Orchard Beach. In the summer of 1982 an effort was made to reassess the abun- dance and distribution of Agalinis maritima in southern Maine. During August and September A. maritima was found regularly in salt marshes in Kittery Point, York, Wells, Kennebunk, Kenne- bunkport, Biddeford Pool, Scarborough, Phippsburg and George- town. In preliminary searches on more easterly marshes in Warren, South Thomaston, and Winterport, we did not locate A. maritima despite the presence of apparently suitable habitat. Agalinis maritima was most often encountered in the upper salt marsh in slight depressions where the vegetative cover, especially Spartina patens, was sparse. The most frequent associates of A. maritima were Glaux maritima, Salicornia europaea, Plantago jun- coides, Triglochin maritima, Limonium Nashii, and occasional Poten- tilla anserina and Solidago sempervirens. Seaside Gerardia was often patchily abundant with hundreds of plants in a few square meters. Although Maine individuals are typically shorter (10-15 cm) than plants in southern states, their numbers and seed production seem to indicate that the Maine populations are thriving and likely to persist. Due to its small size Agalinis maritima is not conspicuous on the marsh until it blooms. Its blossoms are short-lived, often falling off within a day. These factors, coupled with its annual habit, locally transient and patchy distribution, and the fact that it grows on salt marshes, a habitat infrequently visited by some botanists, make it likely that Seaside Gerardia is a species more overlooked and under- reported than genuinely rare in Maine. At least, it is a frequent member of the flora of most salt marshes along the southern Maine coast. The northeastern limit of the Maine distribution of A. mar- tima remains to be discovered. 1983] New England Notes 269 LITERATURE CITED BEAN, RALPH C., CHARLES D. RICHARDS, FAY HYLAND. 1966. Revised Checklist of the Vascular Plants of Maine. Bulletin of Josselyn Botanical Society of Maine. No.8, Orono, Maine. EASTMAN, L. M. 1978. Rare and Endangered Vascular Plant Species in Maine. Prepared by the New England Botanical Club in cooperation with the U.S. Fish and Wildlife Service (Newton Corners, MA). B.ST.J.V. MAINE CRITICAL AREAS PROGRAM MAINE STATE PLANNING OFFICE 189 STATE ST AUGUSTA, ME 04333 P.D.V. MASSACHUSETTS AUDUBON SOCIETY GREAT SOUTH RD. LINCOLN, MA 01773 270 Rhodora [Vol. 85 LITERATURE FOR NEW ENGLAND BOTANISTS CAMPBELL, CHRISTOPHER S. and LesLig M. EASTMAN, 1980. Flora of Oxford County, Maine. 244 pp., biblog., index. Orono, Maine, Life Sciences and Agricultural Experiment Station, University of Maine, Technical Bulletin 99. If each county in New England had a flora like this outstanding one, we would approach the English in their study of local botany— alas we are far from this ideal. Campbell and Eastman have used Arthur Stanley Pease’s excellent Flora of Northern New Hampshire (1964) as a model for their flora, and they have produced a solid work. A 29-page introduction discusses pertinent geographical fea- tures of the area and details the various plant associations. The occurrence of the taxa is documented with thoroughness of citation of herbarium specimens on a town by town basis. Frequency, and brief habitat notes are included for each species. McDOoNNELL, MARK J. 1979. The Flora of Plum Island, Essex County, Massachusetts. 110 pp., bibliogr., index. Durham NH, New Hampshire Agricultural Experiment Station, Uni- versity of New Hampshire, Station Bulletin 513. Maps, diagrams, and a description of the geographical features (climate, geology, land-use history), and plant associations preface this flora. Plum Island is an area long of interest to New England naturalists—especially birders and tide-marsh students. McDonnell’s flora summarizes and adds to the pre- vious botanical collections and studies of the island. It pro- vides for the first time a proper base for further additions and for documenting future coastal vegetation changes. MARY M. WALKER LIBRARIAN NEW ENGLAND WILD FLOWER SOCIETY FRAMINGHAM, MA 01701 BOOK REVIEW KulT, Jos. 1982. A Flora of Waterton Lakes National Park.xxiv + 684 pp. University of Alberta Press, Edmonton, Alberta. (Price $25; $15, waterproof paperback) This handsome flora should be of great use to botanists and botanically oriented visitors to Waterton Lakes Park and beyond to Glacier Park, east of the Continental Divide, and the Alberta Rock- ies south of latitude 50°. Further, 55% of the total species occuring in Alberta Province occur in the Park Flora, so the book is useful over a wide area. The introduction contains a brief description of the Park geology and the plant communities. There are 18 pages of master keys designed to lead one to the proper plant family. There are further keys to the genera within the family section and species keys under the genus description. Families, genera, and species are all arranged alphabetically, a feature that has both merits and drawbacks. The text is clearly written, and full; measurements are included, though technical terms are kept to a minimum to facilitate use by the inter- ested public. Each species is illustrated by a line drawing showing habit, and occasionally, details. The book concludes with a glossary, bibliography, and index. MARY M. WALKER LIBRARIAN NEW ENGLAND WILD FLOWER SOCIETY FRAMINGHAM, MA 01701 BOOK REVIEW BENSON, LYMAN. 1982. The Cacti of the United States and Canada. ix + 1044 pp. Stanford, CA., Stanford University Press. (Price $85.) The Cacti of the United States and Canada is a monument to a life-time of research on the subject, and what a monument! One zit 272 Rhodora [Vol. 85 opens the book expecting a short introduction to a descriptive monograph. This introduction (i.e. Part 1) turns out to be a book in its own right, a 247 page, most readable treatise on all aspects of the biology, taxonomy, and ecology of cacti. It includes a long section on the floras and floristic associations of the United States and Canada, listing the marker species (usually woody) plus the more important cactus species in each. A New England botanist browsing through Part | will quickly learn more about cacti than he ever knew before. Part 2 begins with interpretive notes to understanding the text and a key to the genera of Cactaceae. The text gives a comprehen- sive description of the 18 genera, 151 species, and 133 varieties which Benson recognizes as growing in this region. The descriptions include full technical details, large-scale range maps, numerous black and white photographs and 194 in color; and marvellous detailed line-drawings for many species by Lucretia B. Hamilton. There are keys to the species, and tables comparing the varietal characters. Next is a 60 page section of documentation for the scien- tific names. (Benson has reduced many species to varieties.) The book concludes with a glossary, references, illustration credits, ref- erence maps, and an index. In the Introduction (p. 8) under a photograph of George Engel- mann is a statement that he was an “astute botanist and the most outstanding student of cacti of all time”. Readers of this book will surely add “. . until the advent of Lyman Benson.” MARY M. WALKER LIBRARIAN NEW ENGLAND WILD FLOWER SOCIETY FRAMINGHAM, MA 01701 1983 SINO-AMERICAN BOTANICAL EXPEDITION TO WESTERN SICHUAN PROVINCE Three U.S. botanists have been invited to participate in a joint Chinese-American botanical expedition to western Sichuan Prov- ince, People’s Republic of China, in late summer and early fall of this year. The expedition will be conducted under the auspices of the Botanical Society of America’s Committee for Scientific Liaison with the P.R.C. and both the Institute of Botany, Beijing, and the Chengdu Institute of Biology, with the sponsorship of Academia Sinica. The expedition will visit Mt. Emei (Mt. Omei) and the Wolong Nature Reserve. The U.S. participants will try to fill spe- cific requests for research material (living plants, seeds, herbarium specimens, pickled material, wood samples, etc.) from this part of China. Requests for material can be sent to any of the three Ameri- can participants: Bruce Bartholomew, Department of Botany, Cali- fornia Academy of Sciences, Golden Gate Park, San Francisco, CA 94118; David E. Boufford, Harvard University Herbaria, 22 Divinity Avenue, Cambridge, MA 02138; and Paul L. Redfearn, Jr., Depart- ment of Life Sciences, Southwest Missouri State University, Spring- field, MO 65802. Vol. 85, No. 841, including pages 1-126, was issued January 31, 1983. 273 INSTRUCTIONS TO CONTRIBUTORS TO RHODORA Manuscripts should be submitted in triplicate (an original and two xerox copies) and must be double-spaced (at least 3/8 of an inch) throughout including footnotes, figure legends, and refer- ences. Please do not use corrasable bond. The list of legends for figures and maps should be provided ona separate page. Footnotes should be used sparingly. Do not indicate the style of type through the use of capitals or underscoring, particularly in the citation of specimens. Names of genera and species may be underlined to indi- cate italics in discussions. Specimen citations should be selected critically, especially for common species of broad distributions. Sys- tematic revisions and similar papers should be prepared in the for- mat of “A Monograph of the Genus Malvastrum”, S.R. Hill, Rhodora 84: 1-83, 159-264, 317-409, 1982, particularly with refer- ence to indentation of keys and synonyms. 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, Rhodora 81: 503-548, 1979. For bibliographic cit- ations, 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, 3rd ed. (original title: Style Manual for Biological Jour- nals). In preparing figures (maps, charts, drawings, photos, etc.) please remember that the printed plate will be 4 x 6 inches, be sure that your 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 3 1/2 x 6 in.) Magnifi- cation/reduction values given in text or figure legends should be calculated to reflect the actual printed size. Contrary to prior policy, Rhodora now requests that an abstract and a list of key words be supplied with all papers submitted except for very short articles and notes. RHODORA April 1983 Vol. 85, No. 842 CONTENTS The genus Encyclia Hook. (Orchidaceae) in the Bahama Archipelago Ruben P. Sauleda and Ralph M. Adams Taxonomic treatment of the Chaetomorpha and Rhizoclonium species (Cladophorales: Chlorophyta) in New England Stephen M. Blair ; : Erratum, Jan 1983 issue (Vol. 85, No. $41) . Anatomy of the pericarp of Clibadium, Desmanthodium, a Ichthyothere (Compositae, Heliantheae) and systematic implications Tod F. Stuessy and Ho- Yih Liu Correct author citation for Senecio werneriaefolius Leonard J. Uttal A clarification of the status of maak crinita an C. cuania (siabina Lisa A. Standley Notice of Publication: New England’s Rare, Threatened, and Endangered Plants Ce trie Le he te ; . Chromosomes of Mexican Sedum IV. — in Sedum moranense Charles H. Uh! Patches, clones, and self-fertility of sicouielas (Podophyllum peltatum L.) David Policansky Chromosomal typification of ae ee bermisiene L. tiriaacean) L. Michael Hill 2 : ; ; . ‘ ; ‘ The type locality of Sedum pusillum and Senecio tomentosus Michx. Leonard J. Uttal : ; F P ; ; : Physostegia virginiana var. arenaria Shimek, an overlooked name Philip D. Cantino and Thomas G. Lammers NEW ENGLAND NOTES Carex kobomugi Owhi, an advetive sedge new to New England Lisa A. Standley : : ‘ ; ‘ ‘ : Note on the status of Agalinis maritima (Raf.) Raf. in Maine Barbara St. John Vickery and Peter D. Vickery Literature for New England Botanists , Book Review: A Flora of Waterton Lakes National Park Book Review: The Cacti of the United States and Canada 1983 Sino-American Botalical Expedition 127 175 212 213 228 229 242 243 253 257 259 263 265 267 270 271 271 273 Instructions to contributors to Rhodora ; , ; ; . inside back cover JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB No. 843 July 1983 Vol. 85 Che Nef England Botanical Club, Inc. Botanical Museum, Oxford Street, Cambridge, Massachusetts 02138 Conducted and published for the Club, by NORTON H. NICKERSON, Editor-in-Chief Associate Editors A. LINN BOGLE GARRETT E. CROW WILLIAM D. COUNTRYMAN RICHARD A. FRALICK GERALD J. GASTONY NORTON G. MILLER ROBERT T. WILCE RHODORA.—Published four times a year, in January, April, July, and October. A quarterly journal of botany, devoted primarily to the flora of North America. Price $20.00 per year, net, postpaid, in funds payable at par in the United States currency at Boston. Some back volumes, and single copies are available. For information and prices write RHODORA at address given below. Subscriptions and orders for back issues (making all remittances payable to RHODORA) should be sent to RHODORA, Botanical Museum, Oxford Street, Cambridge, Mass. 02138. In order to receive the next number of RHODORA, changes of address must be received prior to the first day of January, April, July or October. Scientific papers and notes relating to the plants of North America and floristically related areas will be considered by the editorial committee for publication. Articles concerned with systematic botany and cytotaxon- omy in their broader implications are equally acceptable. Brevity is urged whenever possible in all papers. Short items will be published on otherwise blank end pages as soon as possible, even if they appear ahead of longer articles already accepted. All manuscripts should be submitted in TRIPLICATE AND MUST BE DOUBLE (AT LEAST 3/8 OF AN INCH) OR TRIPLE- SPACED THROUGHOUT. Please conform to the style of recent issues of the journal. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to: Russell R. Walton Managing Editor, RHODORA Harvard University Herbaria Building 22 Divinity Avenue Cambridge, Mass. 02138 Second Class Postage Paid at Boston, Mass. PRINTED BY THE LEXINGTON PRESS, INC LEXINGTON, MASSACHUSETTS Cover illustration Trollius laxus Salisb. is very rare as it nears its northeastern limit in Connecticut. Some old records indicate that it has grown in parts of New Hampshire and Maine, so it may yet be found in appropriate habitat. Original artwork by Tess Feltes, Illustrator. TRbodora (ISSN 0035-4902) JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 85, July 1983 No. 843 THE COMPOSITION AND SEASONAL PERIODICITY OF THE MARINE-ESTUARINE CHLOROPHYCEAE IN NEW HAMPSHIRE! ARTHUR C. MATHIESON AND EDWARD J. HEHRE In two previous accounts we have summarized a variety of floris- tic and phenological data on the Rhodophyceae (Hehre & Mathie- son, 1970) and Phaeophyceae (Mathieson & Hehre, 1982) of New Hampshire. A similar compilation for the marine Chlorophyceae is given here, based upon collections and observations made from 1965 to 1980. Monthly or seasonal collections of green algae were made at numerous estuarine and coastal sites throughout New Hampshire (Figs. | and 2), with herbarium voucher specimens of each taxon being prepared and deposited in the Albion R. Hodgdon Herbarium of the University of New Hampshire (NHA). The recent nomenclature of South (1976) is followed in most cases. In the present account, a synopsis of 58 chlorophycean taxa is given, with a statement of relative abundance, local distribution (vertical and horizontal), and seasonal occurrence of each taxon. PHENOLOGY AND DISTRIBUTION The monthly occurrence of chlorophycean taxa within estuarine- coastal waters in New Hampshire is summarized in Table I. The lowest numbers of taxa were recorded in December and January (i.e., 27 and 24 taxa respectively), while the highest numbers were \Scientific Contribution No, 1148 from the New Hampshire Agricultural Experi- ment Station; also issued as Contribution Number 120 from the Jackson Estuarine Laboratory. Zio 276 Rhodora [Vol. 85 | Maine > Piscataqua Little Squa re Rock prcey ee ne 42 58 _] fo’ White 70 38 o 70 36 l 4 | Figure 2. The Isles of Shoals, New Hampshire — Maine 278 Rhodora [Vol. 85 found in July and August (43 and 40 taxa, respectively). A few fresh-water green algae (Microspora, Oedogonium, Spirogyra and Stigeoclonium spp.) were collected intermittently throughout the year and primarily in riverine or innermost estuarine habitats. Forty-eight (83%) of the green algae recorded were considered to be annuals, while only 7 taxa (12%) were interpreted as perennials (Fig. 3). The longevity designations for Cladophora sericea, Derbe- sia marina, and Ulva lactuca require further study, as they may be either aseasonal annuals or pseudoperennials (sensu Knight & Parke, 1931). Specific details on the longevity and seasonal occur- ence of each taxa are summarized in Table I, as well as in the following annotated checklist. A summary of the local distribution of chlorophycean taxa in the four major coastal-estuarine areas in New Hampshire (Figs. | and 2) is also shown in Figure 3. The highest number of species was recorded from the near-shore open coast between Portsmouth and Seabrook (i.e. 47 taxa). The species diversity within the Great Bay System and at the Isles of Shoals were very similar (i.e. 46 and 42 taxa, respectively). In contrast to the similarity of species richness at the three areas, their species compositions were quite different (Table 1). For example, the fresh-water green algae mentioned ear- lier were only collected within riverine (i.e. innermost estuarine) habitats of the Great Bay Estuary System, and they were absent at both open coastal sites. The low species diversity (i.e. 18 taxa) within the Hampton-Seabrook Estuary System contrasts strongly with that of the Great Bay Estuary System. ANNOTATED CHECKLIST Fifty-eight taxa of chlorophycean algae are recorded in the following checklist from coastal and estuarine environments within the state. Twelve of these plants are newly recorded from New Hampshire: Acrochaete repens, Bolbocoleon piliferum, Chlorochy- trium moorei, Cladophora albida, C. pygmaea, C. rupestris, Enteromorpha flexuosa ssp. flexuosa, Gomontia polyrhiza, Mono- stroma leptodermum, Prasinocladus marinus, Pringsheimiella scu- tata and Stichococcus marinus. None of the above taxa represent range extensions on the northeast coast of North America. The new records are indicated by asterisks in the list. 1983] Mathieson & Hehre — Marine Chlorophytes 279 ] NH Open Coast Annual Great Bay Estuary System a0 : Hampton Seabrook Estuary System NO. TAXA a N Oo oO \ it it 1 1 | | Perennial cr Pseudoperennial = Aseasonal Annual 1 n 1 1 1 L ne tes oF Shools Longevity Local Distribution Figure 3. The longevity and local distribution of chlorophycean taxa in New Hampshire TAXA Acrochaete repens Acrochaete viridis Blidingia minima Bolbocodeon piliferum Brvopsis plumosa Capsosiphon fulvescens Chaetomorpha aerea Chaetomorpha brachygona Chaetomorpha linum Chaetomorpha melagonium Chaetomorpha minima Chaetomorpha picquotiana Chlorochytrium moorei Cladophora albida Cladophora pygmaea Cladophora refracta Cladophora rupestris Cladophora sericea Codiolum gregarium Codiolum petrocelidis Codiolum pusillum Seasonal Occurrence, Longevity, and Local Distribution Table 1: MONTHS J F xX xX xX xX xX** xX xX X X X X X X X xX xX X X X X X of Chlorophyceae M J J A x* xX xX XX X xX xX X X xX* xX xX X xX xX X X xX xX X X xX xX XK X X X xX X xX xX xX xX xX xX xX X X X X xX xX XX xX xX xX xX X xX xX X X X* X X X XxX X x KK KK x x Pad x x xm KKK KK KK x KK mK mx x x KK KK KK OK ~~ Ann. AAnn. AAnn. Ann. Ann. Ann. Per. Ann. (?) Per. Per. Ann. (?) Per. Ann. AAnn. Ann. Per. AAnn. or PPer. Ann. Ann. Ann. O87 e1Opoy y $8 10A] Derbesia marina Enteromorpha clathrata Enteromorpha compressa Enteromorpha flexuosa ssp. flexuosa Enteromorpha flexuosa ssp. paradoxa Enteromorpha intestinalis Enteromorpha linza Enteromorpha prolifera Enteromorpha torta Entocladia flustrae Gomontia polyrhiza Halicystis ovalis Microspora pachyderma Monostroma fuscum Monostroma grevillei Monostroma leptodermum Monostroma oxyspermum Monostroma pulchrum Mougeotia sp. Oedogonium sp. Percursaria percursa Prasinocladus marinus x KK * x KK KK mK x KK x KK xm mK KK mK mK x KK mx KK x x x x KK KK xxx ~*~ * x KKK KK x x ~ x KKK x x x KK KK ~ x x KKK KK ~ x KK x mK KK x KK x KKK xX* x KK KK AAnn. or PPer. Ann. AAnn. [€861 sadydolojyD suey — YydH W uosaiyi ep 187 TAXA Prasiola stipitata Pringsheimiella scutata Pseudendoclonium submarinum Rhizoclonium riparium Rhizoclonium tortuosum Spirogyra sp. Spongomorpha arcta Spongomorpha spinescens Stichococcus marinus Stigeoclonium sp. Ulothrix flacca Ulva lactuca Urospora_ penicilliformis Urospora speciosa Urospora_ wormskioldii MONTH x KK Dm me mK mK RK KK KK eum Table | (cont.) eli dP PP Pad xm mK mK 36 xm KK KKK mx 43 x KK X* 40 36 x KK x eK mK 34 27 AAnn Ann. (? AAnn AAnn AAnn. (? Ann Ann Ann Ann. Ann. Ann AAnn. or PPer Ann Ann Ann . 1,2,4 yl . 1,4 . 1-4 ) 1,2,4 .4 . 1,2,4 . 1,2,4 C87 viopoyy $8 109A] KEY TO TABLE | X = presence * = obtained in culture ** = residual basal material Longevity: Ann. = annual AAnn. = aseasonal annual Per. = perennial PPer. = pseudoperennial sensu Knight & Parke (1931) Local Distribution: 1 = Isles of Shoals 2 = Near-shore open coast 3 = Hampton-Seabrook Estuary System 4 = Great Bay Estuary System [€861 sayAydosojyD suey — aY4dH 2 uosaiyIep] €8¢ 284 Rhodora [Vol. 85 *Acrochaete repens N. Pringsheim Obtained in cultures from open coastal and estuarine water sam- ples (March and July), which were concentrated on glass fiber filters and grown in enriched seawater media (12-32 0/00) at 6-20C (R. Zechman, personal comm.). No in situ populations seen. Annual. Acrochaete viridis (Reinke) R. Nielsen Common on the open coast, less abundant and scattered within the Great Bay Estuary System. Endophytic within several fleshy red algae (0 to -8m), often turning them green. Nielsen (1979) trans- ferred the type species of Entocladia Reinke to Acrochaete based upon culture studies of materials from the type locality. Aseasonal annual. Blidingia minima (Nageli ex Kiitzing) Kylin A common “opportunitistic” species (see Prange, 1978) at diverse open coastal and estuarine sites. Growing on a variety of substrata (rock, wood, etc.). More abundant and exhibiting a broader zona- tion in estuarine (+1.0 to +2.3m) than open coastal areas (+2.0 to +2.5m). Plants found in riverine (i.e. innermost estuarine) sites are morphologically variable and should be designated as variety sub- salsa (Kjellman) Scagel, while the unbranched coastal plants are referrable to variety minima (see Abbott & Hollenberg, 1976; Scagel, 1966). Norris (1971) suggests that Blidingia marginata may be a juvenile stage of B. minima: we agree and have not differen- tiated it here. Aseasonal annual. *Bolbocoleon piliferum N. Pringsheim Obtained in cultures from open coastal water samples (March and July), which were concentrated on glass fiber filters and grown in enriched sea water (12-32 0/00) at 6~20C (R. Zechman, personal comm.). None of the cultured plants occurred as endophytes or epiphytes, although a variety of “host” plants were present in the same cultures. No in situ populations seen. Annual. Bryopsis plumosa (Hudson) C. Agardh Common in shallow estuarine areas, rare on the open coast. Found on rocks, shells and coarse algae (0.0 to —10m). Abundant 1983] Mathieson & Hehre — Marine Chlorophytes 285 during the summer and with some residual plants occasionally found during the fall-winter. Annual. Capsosiphon fulvescens (C. Agardh) Setchell et Gardner Occasional, found on a variety of substrata (+2.2 to +2.7m) in riverine or innermost estuarine habitats — often at the head-waters of tidal streams. Rare on the open coast, found growing in marshy ponds or pools, mixed with Enteromorpha spp., Cladophora seri- cea, and Rhizoclonium riparium. Collected from May-November. Annual. Chaetomorpha aerea (Dillwyn) Kitzing Common on the open coast and collected at one outer estuarine site on the Piscataqua River. Found on rocks and ledges (+1.0 to +1.8 m), usually in tide pools. Often epiphytized by Monostroma spp., Protectocarpus speciosus, diatoms and other algae. Blair, et al. (1982) give a detailed account of the taxon as well as its differentia- tion from C. /inum. Perennial. Chaetomorpha brachygona Harvey Occasional at several open coastal sites and scattered within the Great Bay Estuary System. Often mixed with other unbranched cladophoralean algae (+0.6 to —18.0m). Blair (1983) gives a detailed account of the taxon, as well as its differentiation from Chaetomor- pha cannabina and C. capillaris. Annual (?). Chaetomorpha linum (O. F. Miiller) Kiitzing Common in open coastal and estuarine environments; found de- tached and entangled amongst coarse seaweeds, including Chaeto- morpha picquotiana (+0.6 to —20.0m). Occasionally found attached (i.e. initially) by a single basal cell (Blair, 1983). More common than C. picquotiana in estuarine areas, where it often forms extensive festoons 3-5 m long in tidal rapids (Reynolds, 1971). Perennial. Chaetomorpha melagonium (Weber ef Mohr) Kiitzing Common on the open coast; rare within the Great Bay Estuary System where it often occurs as single filaments versus clumps of filaments on the open coast (Reynolds & Mathieson, 1975). Found on rocky substrata (+0.6 to —26.0m), often in tide pools. Rare in estuarine sites. Perennial. 286 Rhodora [Vol. 85 Chaetomorpha minima Collins es Hervey Found a few times at the Isles of Shoals during October and November; growing attached by a basal cell to Chaetomorpha aerea and C. melagonium (0.0 to +1.0m). Blair (1983) characterizes the taxon and discusses its possible interrelationships with Rhizoclo- nium riparium. Annual (?). Chaetomorpha picquotiana (Montagne) Kiitzing Common on the open coast, occasional and scattered in estuarine sites (Blair, 1983). Forming entangled masses amongst coarse sea- weeds (+0.6 to —20.0m), including Chaetomorpha linum. Blair (1983) gives a recent interpretation of this taxon which has pre- viously been referred to as C. atrovirens (Taylor, 1962). *Chlorochytrium moorei Gardner Found once (October) growing within the mucilaginous sheath of the colonial diatom Berkeleya rutilans. Present in a high (+2.0m) marshy tide pool on the open coast. Annual. *Cladophora albida (Hudson) Kiitzing Occasional at scattered open coastal and estuarine sites. Epilithic or epiphytic on coarse algae (0.0 to —3.0m). Present year-round. Aseasonal annual. *Cladophora pygmaea Reinke Occasional at scattered sites within the Great Bay Estuary Sys- tem. Found on rocks in association with a variety of encrusting taxa (e.g. Pseudolithoderma extensum, Hildenbrandia rubra, Rhodophy- sema elegans, Ralfsia spp.) within the sublittoral zone (0 to —10m). The plant may have been missed on the open coast because of its small size and dark green color (see South & Hooper, 1980; Wilce, 1970). Perennial. Cladophora refracta (Roth) Kiitzing Occasional on rocks at the Isles of Shoals (+2.0 to +2.5m), sometimes in tide pools. Found once within the mid-eulittoral zone within the Great Bay Estuary System. Annual. *Cladophora rupestris (Linnaeus) Kiitzing Occasional on rocks at the Isles of Shoals (0.0 to —1.0 m) — often in 1983] Mathieson & Hehre — Marine Chlorophytes 287 tide pools. Reported by Collins (Hoek, 1981) from Hampton, near the southern border of the near-shore open coast of New Hamp- shire. To date we have only collected the plant from the near-shore open coast in Massachusetts and Maine. Perennial (?). Cladophora sericea (Hudson) Kiitzing Common at diverse open coastal and estuarine sites. Growing on a wide variety of substrata (+2.6 to —6.0m). An aseasonal annual or a pseudoperennial (Knight & Parke, 1931) which is capable of regenerating upright filaments from basal, residual filaments. Codiolum gregarium A. Braun Occasional on the open coast and at the mouth of the Great Bay Estuary System. Also obtained in culture from estuarine water sam- ples (July), which were concentrated on glass fiber filters and grown in enriched seawater media (12-32 0/00) at 6-20C (R. Zechman, personal comm.). The in situ collections were found on rocks within the splash zone (+2.0 to +3.4m) of the open coast, often mixed with a variety of blue-green algae (Calothrix scopulorum, Lyngbya and Oscillatoria spp.). The morphology of Codiolum gregarium is only slightly different from C. pusillum (Hanic, 1965), and both of these plants represent stages in the life histories of Urospora spp. and other green algae (Kornmann, 1959; Kornmann & Sahling, 1977; Scagel, 1966). Annual. Codiolum petrocelidis Kuckuck Occasional on the open coast, endophytic within Petrocelis mid- dendorfii (0.0 to —8.0m). The plant is described as the “sporophyte” generation of Spongomorpha spinescens (Jonsson, 1958; Scagel, 1966). Annual. Codiolum pusillum (Lyngbye) Kjellman Common on the open coast. Also obtained in culture from estua- rine water samples (July), which were concentrated on glass fiber filters and grown in enriched seawater media (12-32 0/00) at 6-20C (R. Zechman, personal comm.). The in situ collections were found on high rocks (+2.0 to +3.4m) during the summer and fall; often mixed with Calothrix scopulorum, Urospora spp., Ulothrix flacca, and Bangia atropurpurea. As noted previously, the plant is very similar morphologically to C. gregarium, both plants represent 288 Rhodora [Vol. 85 stages in the life history of Urospora spp. (Hanic, 1965; Kornmann & Sahling, 1977; Scagel, 1966) and other green algae. Annual. Derbesia marina (Lyngbye) Solier Occasional on the open coast. Found on rocks (often sponge covered) within the sublittoral zone (0 to —15m). The plant repre- sents the sporophytic stage in a pleomorphic (specialized heteromor- phic) life history involving gametophytic Halicystis ovalis (Sears & Wilce, 1970). The vesicular stage of New England material has only been obtained in culture from Massachusetts plants (Sears, 1971) and from natural populations in New Hampshire (Mathieson & Burns, 1970). An aseasonal annual or a pseudoperennial, which is capable of regenerating from residual materials in sponge tissue (Sears, 1971). Enteromorpha clathrata (Roth) Greville Abundant in estuarine areas, entangled amongst coarse plants or attached to a variety of substrata (+1.0 to +2.0m). Less abundant on the open coast, occurring in high tide pools (+2.0 to +2.7m) mixed with Enteromorpha spp., Cladophora sericea, and Rhizoclonium riparium. Collected from April-December. Aseasonal annual. Enteromorpha compressa (Linnaeus) Greville Abundant at several widely distributed estuarine sites; less com- mon on the open coast. Growing on rocks, occasionally epiphytic (+1.5 to +2.5m). As noted by DeSilva and Burrows (1973) the plant is extremely variable morphologically. Present year-round. Asea- sonal annual. *Enteromorpha flexuosa (Wulfen ex Roth) J. Agardh subsp. flexuosa Bliding Uncommon at a few sites within the Great Bay Estuary System. Exposed to considerable fresh water. Found on rocks and entangled amongst coarse algae (+2.5m). Annual. Enteromorpha flexuosa (Wulfen ex Roth) J. Agardh subsp. paradoxa (Dillwyn) Bliding Occasional, found attached and entangled in high marshy tide pools (+2.2 to +2.7m) on the open coast, often mixed with Entero- 1983] Mathieson & Hehre — Marine Chlorophytes 289 morpha clathrata, Cladophora sericea and Rhizoclonium riparium. More abundant in estuarine than coastal areas, particularly in high marshy pannes in the former areas. Primarily collected during the summer and fall. Annual. Enteromorpha intestinalis (Linnaeus) Link Abundant, found ona variety of substrata (0.0 to +2.5m). Present in tide pools and in disturbed surfaces (see Daly & Mathieson, 1977) on the open coast, as well as. widely distributed in estuarine and riverine locales. Present year-round. Aseaonsal annual. Enteromorpha linza (Linnaeus) J. Agardh Common on the open coast, occasional and widely distributed in estuarine sites. Found on a variety of solid substrata, as well as epiphytic on coarse algae (+0.5 to —12.0m). Present year-round. Aseasonal annual. Enteromorpha prolifera (O. F. Miiller) J. Agardh Occasional on the open coast; found on a variety of substrata, as well as entangled in high marshy areas (+1.0 to +2.7m). Ubiquitous in estuarine and riverine areas, again on a variety of substrata (+1.0 to 2.0m). Extremely variable morphologically, and often mixed with a variety of other Enteromorpha species. Present year-round. Aseasonal annual. Enteromorpha torta (Mertens in Jurgens) Reinbold Occasional in outer estuarine sites within the Great Bay Estuary System and on the adjacent open coast. Found on rocks and entangled amongst coarse algae (+1.5 to +3.0m). Mixed with other Enteromorpha spp., Cladophora sericea, and Rhizoclonium ripa- rium. According to Nienhuis (1969) the morphological plasticity of E. prolifera may overlap with E. torta and the taxonomic difference between the two may not always be clear. Annual. Entocladia flustrae (Reinke) Batters Common on the open coast. Growing on the chitinous skeletons (i.e. hydrotheca) of Sertularia sp. (—0.3 to +2.2m) — often on vertical rock faces. Although only collected occasionally, it can probably be collected whenever it is specifically looked for (cf South & Hooper, 1980). Annual. 290 Rhodora [Vol. 85 *Gomontia polyrhiza (Lagerhiem) Bornet ef Flahault Uncommon, found twice (March and August) within mollusk shells on the open coast (0.0 to +2.0m). Probably more common than our collections indicate. The plant is described as the sporo- phytic generation of Monostroma grevillei (Kornmann, 1959). Annual (?). Halicystis ovalis (Lyngbye) Areschoug Uncommon, found twice during July at a single open coastal site; this is the only in situ collection of the plant on the northeast coast of North America (Mathieson & Burns, 1970). The vesicular thallus, which is the gametophytic stage in the life history of Derbesia ma- rina (Sears & Wilce, 1970), is found attached to a variety of crustose coralline algae (e.g. Clathromorphum circumscriptum, Lithotham- nium glaciale, and Phymatolithon lenormandii), on large rocks and boulders that are relatively free of sand or silt (-12 to —24m). According to Scagel (1966) new vesicles can be regenerated on suc- cessive years from the residual rhizomes. Perennial. Microspora pachyderma (Wille) Lagerheim Common on rocks at the head-waters of tidal tributaries (i.e. riverine sites) and occasionally found in more estuarine areas during spring thaw. The plant has also been extensively obtained in estua- rine and open coastal water samples (June, July and November) which were concentrated on glass fiber filters and grown in enriched sea water (12-32 0/00) at 6-20C (R. Zechman, personal comm.). A fresh-water green alga that has tentatively been identified as M. pachyderma, although it has not been identified previously from estuarine habitats (Collins, 1912; Prescott, 1962). Found mixed with Cladophora sericea and Blidingia minima var. subsalsa (+1.5 to +2.0m) during January to May, and with a single in situ collection obtained in October. Annual. Monostroma fuscum (Postels ef Ruprecht) Wittrock Common ata variety of open coastal and estuarine sites. Growing on rocks and epiphytic on coarse algae (+0.45 to —24.0m), often in tide pools. Present year-round and probably represented by several isomorphic sporophytic and gametophytic generations (Dube, 1967). Annual. 1983] Mathieson & Hehre — Marine Chlorophytes 291 Monostroma grevillei (Thuret) Wittrock Common at a variety of open coastal and estuarine sites; less abundant in the latter than the former areas. Found on rocks, mus- sels and as an epiphyte on coarse algae (+1.7 to —6.0m). First occurs as a saccate stage during the fall. Abundantly reproductive in May and June and absent during July-September. The interrelationship between Gomontia polyrhiza and Monostroma grevillei has been previously noted (see Kornmann, 1959). Annual. *Monostroma leptodermum Kjellman Abundant at outer estuarine sites — rare on the open coast. Us- ually epiphytic on Zostera marina (0.0 to —1.0m). Found from May to September. Annual. Monostroma oxyspermum (Kiitzing) Doty Common in estuarine areas and tolerant to pronounced salinity variations. Rare and only collected a few times on the open coast —particularly after a very warm summer. Found on wood as well as as epilithic and epiphytic (+1.0 to +2.5m). Most abundant during the summer but some residual populations survive periods of winter ice coverage. Annual. Monostroma pulchrum Farlow Common in open coastal and outer estuarine areas within the Great Bay Estuary System. Epiphytic on coarse algae (+1.0 to —|.3m). Collected during March-June, with reproductive plants abundant during May and June. Annual. Mougeotia sp. Collected once (July) on rocks at the head-waters of the tidal limits of the Oyster River (+1.5 to +2.0m). A fresh-water green alga, found mixed with Spirogyra sp., Cladophora sericea, Enteromorpha prolifera and Blidingia minima vat. subsalsa. Annual. Oedogonium sp. Collected twice (August and September) on rocks beneath a waterfall at the head-waters of the tidal limits of the Oyster River (+2.0m). A fresh-water green alga, found mixed with Cladophora sericea, Enteromorpha prolifera and Blidingia minima vat. sub- salsa. Annual. 292 Rhodora [Vol. 85 Percursaria percursa (C. Agardh) Rosenvinge Common in estuarine areas, occasional on the open coast (+2.0 to +2.7m). Growing amongst emergent halophytic plants in estuarine areas, mixed with Enteromorpha spp., Cladophora sericea, and Rhizoclonium riparium. Found in high marshy tide pools on the open coast, mixed with similar green algae as in estuarine areas. Most abundant during the summer, but present year-round. Annual. *Prasinocladus marinus (Cienkowski) Waern Obtained in cultures from open coastal water samples (July), which were concentrated on glass fiber filters and grown in enriched seawater media (32 0/00) at 6C (R. Zechman, personal comm.). The dichotomously branched colonial form was observed growing de- tached on the surface of the culture media. No in situ populations seen. Annual. Prasiola stipitata Suhr in Jessen Common on the open coast and at the mouth of the Great Bay Estuary System. Often forming a conspicuous green Coating in crev- ices and pockets on high rock faces (+2.7 to +3.5m), often asso- ciated with bird droppings. Present year-round. Aseasonal annual. *Pringsheimiella scutata (Reinke) Marchewianka Rare, found twice on the open coast (+0:5 to +1.5m), once epi- phytic on Chondrus crispus another time on Choreocolax polysi- Phoniae. Annual (?). Pseudendoclonium submarinum Wille Occasional in open coastal and estuarine areas. Found on rocks and shells (+2.0 to —10.0m). Probably more abundant than our collections would indicate. Aseasonal annual (Sears, 1971). Rhizoclonium riparium (Roth) Harvey Common in estuarine locations (+1.5 to +2.7m) and high marshy tide-pools (+2.2 to +2.7m) on the open coast, mixed with Clado- Phora sericea and Enteromorpha spp. Most common during the summer, but present year-round. Nienhuis (1975) placed R. im- plexum (Dill.) Kutz. (=R. kerneri) in synonomy with R. riparium based on extensive field and culture studies. Aseasonal annual. 1983] Mathieson & Hehre — Marine Chlorophytes 293 Rhizoclonium tortuosum Kiitzing Common on the open coast and with scattered estuarine popula- tions within the Great Bay Estuary System. Found entangled amongst coarse algae (+0.6 to —12.0m). Most abundant during the summer, but present year-round. Aseasonal.annual (°). See Blair (1983) for a detailed account of the taxon. Spirogyra sp. Collected once (July) on rocks at the head-water of the tidal limits of the Oyster River (+1.5 to +2.0m). A fresh-water green alga, found mixed with Mougeotia sp., Cladophora sericea, Enteromorpha pro- lifera and Blidingia minima var. subsalsa. Annual. Spongomorpha arcta (Dillwyn) Kiitzing Common on the open coast and outer sites within the Great Bay Estuary System. Usually epilithic, occasionally epiphytic on coarse seaweeds or entangled amongst dense populations of Myrilus edulis on the open coast (+0.9 to —11.0m). Most abundant during the winter and spring, but present year-round. Lamb and Zimmerman (1964) suggest that the plant is a perennial or pseudo-perennial being represented by reduced or inconspicuous basal filaments from late summer to the end of winter. We have observed the prolonged occurrence of basal remnants but no regeneration of upright fila- ments. Thus, we interpret it as an annual. As noted by South (1976), the size of the plant is extremely variable and it may encompass Spongomorpha sonderi (=S. lanosa) in Taylor (1962). Spongomorpha spinescens Kiitzing Common on the open coast and with scattered outer estuarine populations within the Great Bay Estuary System. Epilithic and occasionally epiphytic on coarse algae (+0.9 to —11.0m), often in tide pools. First appears in spring (April) and persists through the fall. Annual. *Stichococcus marinus (Wille) Hazen Collected at two sites on the open coast (Smuttynose and White Islands) during a single date in June. Epilithic in high tide pools (+3.5m), mixed with a variety of green algae. Annual. 294 Rhodora [Vol. 85 Stigeoclonium sp. Found once (March) on rocks at the head-waters of the tidal limits of the Squamscott (Exeter) River (+2.0 to +2.3m). A fresh- water green alga, found mixed with Blidingia minima var. subsalsa. Annual. Ulothrix flacca (Dillwyn) Thuret in Le Jolis Common at a variety of open coastal and estuarine sites, Found on rocks, wood, metal, fucoid algae, other coarse seaweeds and halophytic flowering plants (+0.8 to +2.8m). Primarily found during the winter and spring. Annual. Ulva lactuca Linnaeus Abundant at a variety of open coastal and estuarine sites. Epilithic and epiphytic on coarse algae (+0.9 to —24.0m). Present throughout the year. According to Scagel (1966) and Lamb & Zimmerman (1964) the holdfast of U/va may be perennial; we have not seen regeneration of the blades from the holdfast. We interpret the plant as an aseasonal annual or a pseudo-perennial. Urospora penicilliformis (Roth) Areschoug Common on the open coast, occasional in outer estuarine sites within the Great Bay Estuary System. Found on a wide variety of substrata (+1.2 to +2.4 m). Although it is most abundant in the winter and spring, it has been collected each month, except Sep- tember. Annual. Urospora speciosa (Carmichael ex Harvey in Hooker) Le Blond ex Hamel Uncommon on the open coast and within the Great Bay Estuary System. Epilithic and mixed with other Urospora spp., Ulothrix flacca, and Bangia atropurpurea (+2.5 to +2.8m). Collected during late winter and spring. Annual. Urospora wormskioldii (Mertens in Hornemann) Rosenvinge Common on the open coast and in outer estuarine sites within the Great Bay Estuary System. Found on a variety of rocky substrata (+1.2 to +2.4m), mixed with other Urospora spp., Ulothrix flacca, Codiolum pusillum, Blidingia minima var. minima, Bangia atro- 1983] Mathieson & Hehre — Marine Chlorophytes 295 purpurea, and various blue green algae. /n situ populations are abundant in the winter and spring (December-May), with a few residual populations found in the summer (July). The plant was also obtained in culture from open coastal water samples (August), which were concentrated on glass fiber filters and grown in enriched seawater media (12-32 0/00) at 6-20C (R. Zechman, personal comm.). Annual. DISCUSSION The chlorophycean flora of New Hampshire is characterized bya large number of annuals (82%) and a seasonally variable cycle of species numbers with a winter minimum and a summer maximum. Similar phenological patterns have been noted in other North Atlantic areas (MacFarlane & Bell, 1933; Lamb & Zimmerman, 1964: Coleman & Mathieson, 1975; Reynolds & Mathieson, 1975; Sears & Wilce, 1975), particularly sites with pronounced tempera- ture fluctuations (cf Chapman, 1964; Williams, 1948, 1949). The phaeophycean and rhodophycean taxa within the state (Hehre & Mathieson, 1970; Mathieson & Hehre, 1982) have more pronounced seasonal cycles and greater percentages of perennial taxa than the green algae. An evaluation of Table I suggests that there are two primary groups of chlorophycean annuals: aseasonal and seasonal. The former plants reproduce throughout the year and are represented by successive populations (see Mathieson et al., 1981), while the latter plants have a more restricted reproductive phenology. Some chlo- rophycean annuals (e.g. Bryopsis plumosa) showed a pronounced seasonal cycle of occurrence and abundance; others, such as Mono- stroma oxyspermum, were most abundant during one season (summer) and represented by residual populations at other times. Previous studies on the red and brown algae within the state (Hehre & Mathieson, 1970; Mathieson & Hehre, 1982) demonstrate a sim- ilar characterization of annuals. Detailed tagging studies of in situ populations would help to clarify the longevity and demography of many green algae, particularly those designated as aseasonal an- nuals and pseudoperennials (Knight & Parke, 1931). As noted earlier (Fig. 3), the highest number of species were recorded on the near-shore open coast (i.e. 47 taxa). The species diversity within the Great Bay Estuary System and at the Isles of 296 Rhodora [Vol. 85 Shoals were very similar (i.e. 46 and 42 taxa, respectively), while the lowest numbers of taxa were evident within the Hampton-Seabrook Estuary System (18 taxa). The high species diversity at the near- shore sites and the Great Bay Estuary System is probably associated with a greater variety of habitats and area than at the Isles of Shoals. Even so the Isles of Shoals represent a relatively “pristine” set of small islands with a very diverse chlorophycean flora. Overall, the near-shore open coast has a greater number of “estuarine” species than the Isles of Shoals. The species composition within the Great Bay Estuary System is very different than that at the Isles of Shoals, because of the presence of several fresh-water green algae, the enhanced number of estuarine taxa, and the reduced number of “open coastal” species (Table I). Even so, the presence of strong tidal rapids allows some “open coastal” forms to colonize estuarine areas like Dover Point (Reynolds, 1971: Reynolds & Mathieson, 1975). The differences in species diversity between the Great Bay and Hampton-Seabrook Estuary Systems may also be explained by a greater diversity of habitats, area, and substrata within the former estuary system. In addition to the taxa listed earlier, three other green algae are recorded from the Isles of Shoals in an unpublished checklist (Anon., 1975): Bryopsis hypnoides?, Rhizoclonium erectum, and Ulothrix laetevirens. We have neither collected these plants nor seen voucher materials of these taxa; hence they are not included in the present synopsis. ACKNOWLEDGMENTS We would like to acknowledge our sincere gratitude to the follow- ing people: “Ned” McIntosh, former captain of the R/V Jere A. Chase, who assisted with the diving operations; Dr. D. P. Cheney, formerly of the Jackson Estuarine Laboratory, for his constructive comments on the manuscript, as well as for his help with several collections; to a variety of marine phycology students (past and present) at the University of New Hampshire who have helped with many of the collections, including Stephen Blair, Richie Burns, Jan Chock, Marty Costa, Maureen Daly, William Flahive, Richard Fralick, Steven Fuller, Phelps, Fullerton, Eleanor Tveter-Gallagher, Barry Hutchinson, John Kilar, Mike Josselyn, Fred Murphy, Joan Conway-Lockhart, Cindy Mathieson, Chris Neefus, Richard Nie- 1983] Mathieson & Hehre — Marine Chlorophytes 297 meck, Timothy Norall, Chris Emerich Penniman, Clayton Penni- man, Norman Reynolds, John Shipman, Eric Sideman, and Tim Voorheis. Rick Zechman is also thanked for allowing us to cite several of his unpublished culture findings. REFERENCES Apsott, I. A. & G.J. HOLLENBERG. 1976. Marine Algae of California. Stanford Univ. Press, Stanford, California. xii + 827 p. AnonyMous. 1975. Checklist of the marine flora and fauna of the Isles of Shoals. Shoals Marine Laboratory. 13 p. (mimeograph). BLairR,S.M. 1983. Taxonomic treatment of the Chaetomorpha and Rhizoclonium species (Cladophorales: Chlorophyta) in New England. Rhodora 85: 175-211. ., A. C. MATHIESON, & D. P. CHENEY. 1982. Biosystematic investigations of selected species of Chaetomorpha (Chlorophyta: Cladophorales). Phycologia 21: 164-172. CHAPMAN, V. J. 1964. The Algae. Macmillan and Company Ltd., London. vill + 427p. CoLeMaAN, D. C. & A. C. MATHIESON. 1975. Investigations of New England marine algae VII: seasonal occurrence and reproduction of marine algae near Cape Cod, Massachusetts. Rhodora 77: 76-104. Coitins, F.S. 1912. The Green Algae of North America. Tufts College Studies, Sci. Ser. 2: 79-480. Daty, M. A., & A. C. MATHIESON. 1977. The effects of sand movement on intertidal seaweeds and selected invertebrates at Bound Rock, New Hampshire, U.S.A. Mar. Biol. 43: 45-55. De Sitva, M. W.R.N., & E. M. Burrows. 1973. Anexperimental assessment of the status of the species Enteromorpha intestinalis (L.) Link and Enteromorpha compressa (L.) Grev. J. mar. biol. Ass. U.K. 53: 895-904. Dupe, M. A. 1967. On the life history of Monostroma fuscum (Postels and Ruprecht) Wittrock. J. Phycol. 3: 64-73. Hanic, L. A. 1965. Life history studies on Urospora and Codiolum from southern British Columbia. Ph. D. Thesis. Univ. of British Columbia, Vancouver, B.C. 152 p. Henre, E. J., & A.C. MATHIESON. 1970. Investigations of New England marine algae III. Composition, seasonal occurrence and reproductive periodicity of the marine Rhodophyceae in New Hampshire. Rhodora 72: 194-239. HoEK, C. VAN DEN. 1981. A Taxonomic Revision of the American Species of Cladophora (Chlorophyceae) in the North Atlantic Ocean and Their Geo- graphic Distribution. Verh. K. Akad. Wetensch., Amsterdam, 236 p. Jonsson, S. 1958. Sur la structure cellulaire et la reproduction de Codiolum petrocelidis Kuck., Algue verte unicellulaire endophyte. C. R. Acad. Sci. (Paris) 247: 325-328. KNIGHT, M., & M. W. Parke. 1931. Manx Algae. Mem. Liverpool Marine Biol. Comm. 30: 1-147. 298 Rhodora [Vol. 85 KORNMANN, P. 1959. Die heterogene Gattung Gomontia |. Der sporangiale Anteil, Codiolum polyrhizum. Helgolander wiss. Meeresunters. 6: 229-238. .,& P.-H. SAHLING. 1977. Meeresalgen von Helgoland. Benthische Grun-, Braun- und Rotalgen. Helgolander wiss. Meeresunters. 29: 1-289. Lams, I. M., & M. H. ZIMMERMAN. 1964. Marine vegetation of Cape Ann, Massachusetts. Rhodora 66: 217-254. MacFar ane, C., & H. P. Bett. 1933. Observations of the seasonal changes in the marine algae in the vicinity of Halifax, with particular reference to winter conditions. Proc. Nova Scotian Inst. Sci. 18: 134-176. MATHIESON, A. C., & R. L. Burns. 1970. The discovery of Halicystis ovalis (Lyngbye) Areschoug in New England. J. Phycol. 6: 404-405. .. & E. J. Henre. 1982. The composition, seasonal occurrence and reproductive periodicity of the marine Phaeophyceae in New Hampshire. Rhodora 84: 411-437. .. & N. B. ReyNotps. 1981. Investigations of New England marine algae I. A floristic and descriptive ecological study of the marine algae of Jaffrey Point, New Hampshire. Bot. Mar. 24: 533-545. NIELSEN, R. 1979. Culture studies on the type species of Acrochaete, Bolbocoleon and Entocladia (Chaetophoraceae, Chlorophyceae). Bot. Notiser 132: 441-449, Niennuls, P.H. 1969. Enkele opmerkingen over het geslacht Enteromorpha Link op de schorren en slikken van Z. W. -Nederland. Overdruk uit Gorteria 4, no. 10, pp. 178-183. 1975. Biosystematics and Ecology of Rhizoclonium riparium (Roth) Harv. (Chlorophyceae: Cladophorales) in the Estuarine Area of Rivers Rhine, Meuse and Scheldt. Ryksuniversiteit te Groningen (Thesis). Rotterdam: Bronder Offset B. V. 240 pp. Norris, J. N. 1971. Observations on the genus Blidingia (Chlorophyta) in California. J. Phycol. 7: 145-149, PRANGE, R. K. 1978. An autecological study of Blidingia minima var. subsalsa (Chlorophyceae) in the Squamish Estuary (British Columbia). Can. J. Bot. 56: 170-179. Prescott, G. W. 1962. Algae of the Western Great Lakes Area. Revised edition. Wm. C. Brown Co. Publishers, Dubuque, Iowa, xiii + 977 p. ReyNotps, N. B. 1971. The ecology of a New Hampshire estuarine tidal rapid. Ph.D. Thesis. Univ. of New Hampshire, Durham, N.H., ix + 101 p. ..& A.C. MATHIESON, 1975. Seasonal occurrence and ecology of marine algae in a New Hampshire tidal rapid. Rhodora 77: 512-533. ScaGeL, R. F. 1966. Marine algae of British Columbia and northern Washington. I. Chlorophyceae (green algae). Bull. Nat. Mus. Can. 207: 1-257. Sears, J. R. 1971. Morphology, systematics and descriptive ecology of the sublittoral benthic marine algae of southern Cape Cod and adjacent islands. Ph.D. Thesis. Univ. of Massachusetts, Amherst, Mass. xiii + 295 p. -.&R.T. Wilce. 1970. Reproduction and systematics of the marine algae Derbesia (Chlorophyceae) in New England. J. Phycol. 6: 381-392. & . 1975. Sublittoral benthic marine algae of southern Cape Cod and adjacent islands: seasonal periodicity, associations, diversity, and floristic composition. Ecol. Monogr. 45: 337-365. 1983] Mathieson & Hehre — Marine Chlorophytes 299 Soutu, G. R. 1976. A check-list of marine algae of eastern Canada — First revision. J. mar. biol. Ass. U.K. 56: 817-843. .. & R. G. Hooper. 1980. A catalogue and atlas of the benthic marine algae of the island of Newfoundland. Memorial University of Newfoundland Occasional papers, 137 p. Taytor, W.R. 1962. Marine algae of the Northeastern coast of North America. 2nd edition. Univ. of Michigan Press, Ann Arbor. ix + 509 p. Witce,R.T. 1970. Cladophora pygmaea Reinke in North America. J. Phycol. 6: 260-263. WILLIAMS, L. G. 1948. Seasonal alternation of marine floras at Cape Lookout, North Carolina. Am. J. Bot. 35: 682-695. 1949. Marine algal ecology at Cape Lookout, North Carolina. Bull. Furman Univ. 31: 1-21. DEPARTMENT OF BOTANY AND PLANT PATHOLOGY and JACKSON ESTUARINE LABORATORY UNIVERSITY OF NEW HAMPSHIRE DURHAM, NH 03824 300 Rhodora [Vol. 85 The New England Botanical Club Council will be pleased to receive invitations to hold Club meetings at institutions in the New England area. It is expected that most of the meetings will be at Harvard, as has been the tradition, but at least two meetings will be scheduled for other Colleges or Universities each year. TAXONOMIC IMPLICATIONS OF ANEUPLOIDY AND POLYPLOIDY IN POTAMOGETON (POTAMOGETONACEAE) DONALD H. LEs ABSTRACT A synoptic study of chromosome number reports for 73 species of Potamogeton presents evidence supporting the conclusion that the genus is characterized by two different polyploid lineages (one based upon x =7 and the other upon x = 13). Analyzed comparatively, chromosome numbers also indicate that the diploid level of the genus is 27 = 14, and that the x = 13 lineage arose by aneuploidy from a 2n = 14 progenitor. Comparisons developed by juxtaposition of chromosome numbers with morphological characters and classification ranks indicate that subsections Pusilli, Oxyphylli, and Amplifolii may not represent entirely natural assemblages. The past use of chromosome numbers to demonstrate the primitiveness of submersed leaves in the genus is. challenged due to lack of correlation of these features when a larger number of species is considered. Chromosome numbers were found to correlate geo- graphically with the hypothetically proposed birthplace of the genus. Potamogeton is a large (ca. 100 species) and diverse genus of strictly aquatic monocotyledons (Cook, et al., 1974). Morphological diver- sity in Potamogeton is extensive, with many of the species exhibit- ing heterophylly (producing both floating and submersed foliage), whereas other species possess only leaves which are entirely sub- mersed. Researchers have often recognized two morphological groups within subgenus Potamogeton (Eupotamogeton Raunk.): the “linear-leaved” species (e.g., Fernald, 1932) and the “broad- leaved” species (e.g., Ogden, 1943). Both linear and broad-leaved species occur within several different subsections, thus neither group has been taxonomically well-defined. Considerably diversity within the genus is also apparent by the variety of fruit and stipule charac- ters found among the species. Potamogeton is also characterized by a complex array of polyploidy which has not yet been given a tho- rough systematic evaluation. While no generally acceptable classification has been developed for Potamogeton, taxonomists have relied for nearly 70 years upon the one developed by Hagstrém (1916). It has undergone only slight modifications (by St. John, 1916; Fernald, 1932; Ogden, 1943; Haynes, 1974; and others) in only portions of the genus. Hagstré6m’s classification has been criticized on several occasions for its un- wieldy nature. It included 2 subgenera, 5 sections, 26 subsections, 301 302 Rhodora [Vol. 85 138 species and multitudes of series, subseries, varieties, forms, and hybrids, yet neither a key to nor a synopsis of any of the ranks were given. St. John (1925, p. 462) emphasized this and stated, “There is no key of any kind. In order to identify a Potamogeton, it is neces- sary to read every page of this work, and practically memorize them.” There was no critique that the ranks in the classification were not unified by a phylogenetic framework, therefore, although similar taxa were grouped together, there was little indication of overall relationships among the groups. Hagstr6m approached clas- sification of Potamogeton anatomically, largely following Raunkia- er’s (1895-99) precedent. This approach was quite different from the morphologically oriented one used by Ascherson and Graebner (1907) to construct an earlier classification. Fernald (1932) later de- emphasized the anatomically oriented approach, which was only to be re-emphasized again by Ogden (1943), thus resulting in ambi- guity concerning which approach yielded the better classification. The classifications mentioned all predated the availability of reliable chromosomal data, hence the systematic value of chromosome number in Potamogeton has not been evaluated. The first known publication of a chromosome number for any Potamogeton was by Wiegand (1899) for P. foliosus, although Palm- gren (1939) erroneously cited an earlier paper by Wiegand (1898) as containing the first counts for the genus. Other early chromosome number reports for Potamogeton were made by Wisniewska (1931) and Kuleszanka (1934) but the first such study of any magnitude was by Palmgren (1939) and included counts for 18 species. Pub- lished counts remained relatively scarce throughout the 1940’s as evidenced by Ogden’s (1943) remarks. According to Goldblatt (1979), 66 species of Potamogetonaceae had published counts avail- able, which accounted for approximately 74% of the species. Because a substantial percentage of counts for species is now avail- able in the literature, the use of chromosome numbers to assist for- mulation of systematic relationships within Potamogeton is feasible. Additionally, replicate counts published for the same taxa assist in evaluating the credibility of a given chromosome number report. Incorporation of chromosome counts has facilitated progress in the phylogenetic reconstruction of other genera, but this technique has not yet so been used for Potamogeton. Stern (1961, p. 414) stated, “...reported chromosome numbers indicate that a thorough 1983] Les — Potamogeton 303 cytotaxonomic investigation of [ Potamogeton]...should yield inter- esting results, and would be a valuable adjunct to the previous extensive morphological and anatomical studies...” The purpose of this paper is to bring together chromosome number reports from the literature and to discuss them in the context of taxonomic and phylogenetic relationships within Potamogeton. METHODS A compilation of chromosome number reports published for Potamogeton was made from data available in the following sources: Bolkhovskikh, et al, (1969), Cave (1958-1965), Darlington & Wylie (1956), Goldblatt (1981), Love & Love ( 1974; 1975), Moore (1970-74; 1977), and Ornduff (1967-69). Morphological descrip- tions were located for each species for which counts were available. The descriptions were compiled from recent floristic treatments. Synonomy is an important issue regarding how counts are best apportioned to taxa. It is recognized that the author’s allocation of chromosome counts to a taxon identified as a particular species may not be in accord with some opinions. Most of the discrepancy appears to involve nomenclatural issues, however, and as such does not create a problem in terms of to which taxon a particular count should be assigned. Because of the scope of this study, however, no attempts were made to obtain verified voucher specimens which correspond with the chromosome counts reported in the literature. For the most part, synonomy presented in major floristic works was followed assuming that such works reflect the input of taxonomic expertise. The author is also aware of controversy regarding constitution of adequate methodology for conclusively recognizing hybrids in this genus and makes no claim of possessing the ability to ascertain correctly which taxa represent hybrids and which represent “pure” species. Therefore, the decision was made to eliminate “questiona- ble” taxa (i.e., those taxa widely considered to represent sterile hybrids) from further consideration in this study. Species names which the author accepts to represent the correct nomenclature and which are not thought to be sterile hybrids are listed in Table | along with the corresponding 2” chromosome numbers and numbers of studies corroborating each count. Names not appearing in Table | (and for which counts are reported) are 304 Rhodora [Vol. 85 listed in Table 2 along with reasons for their exclusion. The morpho- logical descriptions of the species given in Table | were used to construct Table 3. Morphological groups in Table 3 were delimited on the basis of the author’s perspective of general morphological groups occurring in Potamogeton. The groups divide linear-leaved species from broad-leaved species and further separate taxa on the basis of other leaf and stipule characters. In itself, Table 3 is not meant to infer taxonomic affinities between or among taxa. The classification followed in Table 4 is that of Hagstrém (1916), modi- fied only to include species described after. completion of that work. The 2n chromosome numbers are those from Table |. The morpho- logical groups (A-F) are from Table 3. Fruit types, either smooth (S) or keeled (K) were designated for each species. The smooth type refers to fruits in which the dorsal keel is absent or greatly reduced, whereas the keeled fruits possess prominent dorsal keels. The assignments of fruit types were made based upon descriptions given in the following references: Aalto (1970; 1974) Komarov (1937), Ogden (1953), and Ohwi (1965). Species for which the fruit type is unknown are designated by a question mark (?). RESULTS AND DISCUSSION Upon examination of Table 1, six discrete ploidy levels can be detected. It is further evident that the genus is characterized by two different euploid series, one based upon x = 7 and the other upon x = 13 as previously noted by Harada (1956). In Table 3 a correlation can be noted between chromosome numbers and the six morphological groups selected to characterize the genus. Most of the groups are dominated by one ploidy level, or at least by one of the two euploid series, which suggests a relation- ship between ploidy levels and general morphology of the species. Groups A and D are conspicuous exceptions because mixtures of both ploidy levels and euploid series occur. It is noteworthy that floating-leaved species occur in both of the euploid series, including the lowest levels for each (i.e., 2n = 26; 28). Furthermore, floating leaves are also produced by P. illinoensis which has the highest chromosome number known for any species of Potamogeton. In Table 4, a comparison of species from an anatomical stand- point is given, i.e., essentially the Hagstrém classification. Super- imposed on this classification are chromosome number, morpho- 1983] Les — Potamogeton 305 Table |. 21 Chromosome numbers of Potamogeton species incor- porated into this study (number of concurring references for count is given in parentheses) Species 2n Species 2n P. acutifolius 26 (2) P. nipponicus a (1) P. alpinus 52 (6) P. nodosus 52 (3) P. amplifolius 52 (1) P. oblongus 26 (2) P. asiaticus 28 (1) P. obtusifolius 26 (1) P. coloratus 26 (1) P. octandrus 28 (4) P. crispus 52 (9) P. orientalis 28 (1) P. cristatus 26 (2) P. oxyphyllus 26 (1) P. dentatus Sz (2) P. pectinatus 78 (9) P. distinctus §2 (2) P. perfoliatus Ba; 412) P. epihydrus 26 (2) P. polygonifolius 52 (2) P. filiformis 78 (6) P. praelongus a2 (6) P. foliosus 28 (1) P. pusillus 26 (6) P. franchetti 52 (1) P. richardsonil bs (3) P. friesii 26 (2) P. robinsii 52 (1) P. fryeri 42 (4) P. rutilus 26 (1) P. gramineus 2 (5) P. sibiricus 28 (1) P. groenlandicus 26 (1) P. strictifolius 52 (1) P. illinoensis 104 (2) P. trichoides 26 (3) P. lucens 52 (2) P. vaginatus 78 (4) P. maackianus 52 S62] P. vaseyi 28 (1) P. malaianus 52 (2) P. zosterifolius 26 (3) P. natans 52 (8) P. zosteriformis a2 (1) TOTAL: 44 species logical group and fruit type for each listed taxon. High concordance of chromosome numbers, morphological groups and fruit types is evident in the subgenus Coleogeton, the section Adnati, and the subsections Compressi, Lucentes, Perfoliati, Javanici, and Colorati. Considerable heterogeneity of these features, however, is notable in subsections Pusil/li, Oxyphylli, and Amplifolii. Species of the x = 13 chromosomal series possess both keeled and smooth fruits, whereas species of the x = 7 series have only keeled fruits (the latter type being more prevalent in the genus). More x = 13 species in group A 306 Species VUVUVUVUVVUVUUVUUU UD UV UU UU UU UU . alatus . anguillanus . apertus . berchtholdii biwaensis . compressus densus faurieri . flabellatus fluitans indicus kamogawaensis . leptocephalus . limosellifolius . longipetiolatus . malainoides . miduhikimo miyakezimaensis monoginous morongil numsakianus panormitanus . porsiaticus . subfluitans . subsessilifolius teganumensis . tenuifolius . tepperi . torquatus TOTAL: 29 species Rhodora [Vol. 85 Table 2. Disposition of problematic species having published counts Disposition = P. distinctus* excluded (lack of adequate description) excluded (alleged hybrid) = P. pusillus* excluded (alleged hybrid) = P. zosterifolius* excluded (= Groenlandia densa) excluded (alleged hybrid) = P. pectinatus* exluded (alleged hybrid) = P. nodosus* excluded (alleged hybrid) excluded (alleged hybrid) = P. octandrus* = P. distinctus* excluded (alleged hybrid) = P. octandrus* = P. malaianus* = P. trichoides* = P. natans* = P. octandrus* = P. pusillus* excluded (lack of adequate description) excluded (lack of adequate description) = P. fryeri* = P. dentatus* = P. maackianus* = P. franchetii* = P. fryeri* *Counts published under the respective problematic names were applied to these synonyms 1983] Les — Potamogeton 307 (linear-leaved species) have smooth fruits than keeled fruits; this relationship being particularly discernible in subsection Pusilli. These chromosome number discrepancies indicate the need for a re-evaluation of portions of the Hagstrém classification. The suggestion that morphological groups within the genus may each correlate with a different ploidy level has been made by Haynes (1974). Attempts to arrive at correlations between ploidy and morphology have led to conflicting hypotheses. An example is the question of whether species of Potamogeton possessing floating leaves are more primitive than their strictly submersed-leaved counterparts. Chrysler (1907, p. 183) considered the floating-leaved species to be more primitive than the submersed-leaved species, which he believed to have been derived from the former as “...a stage in the assumption of the aquatic life by the genus.” He further believed that Potamogeton evolved from a terrestrial ancestor, the classical interpretation applied to most aquatic plants by Arber (1920), Sculthorpe (1967), and others. Cronquist (1968; 1981), how- ever, considers Potamogeton to have evolved from an aquatic ancestor likely related to the dicotyledon order Nymphaeales. Haynes (1974) agreed with Cronquist and formulated a hypothesis of evolutionary progression in Potamogeton contradictory to the classical viewpoint. He (p. 582) invoked the “phyllode theory” (Candolle, 1827; Arber, 1920) for explaining the evolution of leaf morphology in the genus: “...one could infer that the floating leaves might have been derived from the submersed leaves by an increase in the amount of tissue between the veins near the tip of the petiole. This would imply that possibly the floating-leaved species were derived from some ancestral stock of submersed-leaved plants.” Haynes supported his hypothesis with cytological data which indi- cated (p. 582) “The submersed-leaved species are, for the most part, diploid, whereas the floating-leaved species are, for the most part, tetraploid.” His conclusion was that cytological data indicated (p. 583) “...the primitive condition probably was that of total vegeta- tive submergence.” Haynes also used chromosome numbers to sup- port his contention that anemophily is primitive in Potamogeton while hydrophily is derived due to the occurrence of the latter mech- anism only in species characterized by high chromosome numbers while anemophilous species were characterized by lower chromo- some numbers. Although the above arguments are reasonable, there Table 3. Basic morphological types in Potamogeton I. Floating leaves absent; submersed leaves various A. Submersed leaves linear, stipules free (Group A) 1. P.acutifolius (26) 2. P. foliosus (28) 3. P. friesii (26) 4. P. groenlandicus (26) 5. P. obtusifolius (26) 6. P. orientalis (28) 7. P. oxyphyllus (26) 8. P. pusillus (26) 9. P. rutilus (26) 10. P. sibiricus (28) ll. P. strictifolius (52) 12. P. trichoides (26) 13. P. zosterifolius (26) 14. P. zosteriformis (52) B. Submersed leaves broad, stipules free (Group B) 1. P. crispus (52) 2. P. dentatus (52) 3. P. lucens (52) 4. P. malaianus (52) 5. P. perfoliatus (52) 6. P. praelongus (52) 7. P. richardsonii (52) C. Submersed leaves linear, stipules fused (Group C) 1. P. filiformis (78) 2. P. pectinatus (78) 3. P. robbinsii (52) 4. P. vaginatus (78) Il. Floating leaves present; submersed leaves various D. Submersed leaves linear, stipules free (Group D) E. Submersed leaves linear, stipules fused (Group E) I. F. Submersed leaves broad, stipules free (Group F) N= RMWPwWn- SeCmIDRWAEWN— ( P. asiaticus P. cristatus P. epihydrus P. natans P. octandrus P. vaseyi P. maackianus P. alpinus . amplifolius . coloratus . distinctus . franchetii fryeri gramineus illinoensis . Nipponicus nodosus . oblongus . polygonifolius vv VI—IVVVVVVU number for Table |. ) denotes 2n chromosome (28) (28) (26) (52) (28) (28) (52) (52) (52) (26) (52) (52) (42) (52) (104) (52) (52) (26) (52) 80£ elopoy y $8 10A] 1983] Les — Potamogeton 309 is still a discrepancy from an adaptive standpoint. If wind pollina- tion is primitive, then why should the primitive state be character- ized by total vegetative submergence, and if water-pollinated species are advanced, then why do they all lack floating leaves? Ths Significance of Aneuploidy Curiously, the presence of two chromosomal series within Potamogeton has not lead to a re-appraisal of systematic relation- ships among the species despite the fact it has been realized since the 1950’s (Harada, 1956) and is still currently emphasized (e.g., Grant, 1982). From a systematic standpoint, it would be highly desirable to know whether each line represents a monophyletic assemblage. If this indeed apears to be the case, then the classification of the genus should reflect consistent segregation of species from the separate lineages, i.e., no one group should include species from both chromosomal series if a natural arrangement is to be depicted. Investigators usually regard the euploid series based upon x = 13 as typical for the genus, and treat species possessing chromosome numbers of 28 or 42 as “anomalous.” An example of this approach is illustrated by Haynes’ (1974) treatment of subsection Pusilli in which he hypothesized that P. foliosus (2n = 28) was derived froma 2n = 26 progenitor by aneuploidy. Since the other species within the subsection possess chromosome numbers based upon x = 13, his interpretation would appear reasonable. However, the more wide- spread occurrence of the 2n = 28 species must be considered. The presence of 2n = 28 species within the subsections Oxyphylli and Javanici (the former also including 2n = 26 species) would imply that aneuploidy had originated independently at least two separate times in the genus, or that the subsections are not natural. All counted species within subsection Javanici are 2n = 28 which implies that a single “aneuploid” event may have originated that group, but the progenitor of Javanici would have had to arise from a 2n = 28 species in another subsection, or from a 2n = 26 species by yet another aneuploid event. Furthermore, there is the question- able presence of P. fryeri (2n = 42) in a subsection characterized by 2n = 52 species. The “anomaly” would have to be explained by still another aneuploid event in the 2n = 26 progenitor of the group. It is unlikely, therefore, that recurring aneuploidy reasonably explains Table 4. Relationships of chromosome numbers and morphology to sub-generic ranks (see text for explanation of symbols) I. Subgenus: Coleogeton 1. Section: Connati a. Subsection: Filiformes |. P. filiformis 2. Section: Convoluti a. Subsection: Vaginati 1. P. vaginatus b. Subsection: Pectinati |. P. pectinatus Il. Subgenus: Potamogeton 1. Section: Adnati a. Subsection: Serrulati 1. P. robbinsii 2. P. maackianus 2. Section: A xi/lares a. Subsection: Pusil//i 1. P. foliosus 2. P. orientalis 3. P. obtusifolius 4. P. pusillus 5. P. friesii 6. P. rutilus 7. P. groenlandicus 8. P. strictifolius 2n 78 78 78 52 52 28 28 26 26 26 26 26 52 morphological group ™ >>>rr>>rr> fruit type a A VNNNN- WK Subsection: Lucentes . P. lucens . P. dentatus . P. malaianus . P. distinctus . P. gramineus . P. nipponicus 7. P. illinoensis RuOnk Wh — . Subsection: Praelongi 1. P. praelongus . Subsection: Perfoliati |. P. perfoliatus 2. P. richardsonii Subsection: Javanici 1. P. asiaticus 2. P. cristatus 3. P. octandrus 4. P. vaseyi Subsection: Amplifolii 1. P. amplifolius 2. P. franchetii 3. P. fryeri . Subsection: Natantes 1. P. natans Subsection: Hybridi 2n 52 52 52 52 52 52 104 52 52 52 28 28 28 28 52 52 42 52 morphological group wow ie*) TT NMNwOwOw UOD0D fruit type A KR KRKRRKR Ole RIOPpOUY $8 10A] b. Subsection: Monogyni 1. P. trichoides c. Subsection: Compress 1. P. acutifolius 2. P. zosterifolius 3. P. zosteriformis d. Subsection: Oxyphylli 1. P. sibiricus 2. P. oxyphyllus e. Subsection: Crispi 1. P. crispus 26 26 26 52 28 26 52 >> > RRA 1. P. epiphydrus . Subsection: Co/orati 1. P. coloratus 2. P. oblongus 3. P. polygonifolius . Subsection: A/pini 1. P. alpinus . Subsection: Nodosi 1. P. nodosus 26 26 26 52 52 52 se 9 Reap veo 3) (€861 uojasoulp10g — Sa] Ile 312 Rhodora [Vol. 85 the origin of 2m = 28 species. It is more likely that the 2n = 28 species are monophyletic, have resulted from a single aneuploid event and have been simply misplaced throughout portions of the classification. The question of the ancestral diploid number of the genus must be considered. Previous investigators (Stern, 1961; Haynes, 1974) have interpreted the chromosome numbers of 2n = 26 or 28 to represent the diploid level in Potamogeton. Goldblatt (1979), however, sug- gested that in Potamogetonaceae, the base number is likely to be x = 7 and, therefore, all counted species of Potamogeton having 2n chromosome numbers exceeding 14 would indicate polyploidy. This interpretation is in accord with Grant (1963) and Stebbins (1971), who believe that haploid numbers in excess of n = 10-13 are indica- tive of polyploidy. The suggestion of Ehrendorfer et al. (1968) that the progenitors of the angiosperms were characterized by a base number of x = 7 may also be pertinent. Since Potamogeton is usu- ally placed within the relatively primitive subclass Alismatidae (Cronquist, 1981; Takhtajan, 1969), it is conceivable that the base number for the genus would be x = 7. Evidence exists to support this interpretation. Four concurring studies have reported the chromosome number of P. fryeri to be 2n = 42 (see Table 1). If 2n = 28 is accepted as the diploid level, then P. Sryeri would repres- ent a triploid organism. If, however, 2n = 14 is accepted as the diploid level, then P. fryeri would represent a hexaploid (while 2n = 28 species would represent tetraploids). The latter interpreta- tion is here accepted to be more reasonable. Initially it was presumed that diploid (2n = 14) ancestors of the genus were extinct because counts of 2n = 14 for any species of Potamogeton were absent from the literature. There is some indica- tion to the contrary. A study by Wiegand (1899) reported the chromosome number of P. foliosus as 2n = 14. The credibility of that report may well be questioned on the basis of whether an . appropriate methodology was employed in that study. Early studies of chromosome numbers were often hampered by the use of section- ing methods (which Wiegand also used) that frequently resulted in overlooked chromosomes, but indications are that Wiegand’s counts were probably accurate. It is unlikely, even by sectioning, that he would have missed half (13-14) of the chromosomes, especially since he reported the same number in his replicate counts. His illus- 1983] Les — Potamogeton 313 trations clearly depict a chromosome number of 2n = 14. Most convincing is the observation that Wiegand’s count of 2n = 14 is in accord with the chromosomal series to which P. foliosus belongs (x = 7). Since only two North American species are known to belong to that series, it is difficult to believe that inaccuracy on Wiegand’s part would result in a count which was “coincidentally” within the proper numerical series. If the count of 2n = 14 for P. folio- sus is accepted as accurate, then it is evident that some species of Potamogeton may be characterized by different ploidy levels. As Tomlinson (1982) indicated, no cytological survey of large popula- tions of a species has been conducted tor Potamogeton which could clarify this point. The diploid level of Potamogeton is therefore recognized to be 2n = 14. From this assumption follows that x = 13 species were derived from a 2n = 14 diploid by descending aneuploidy. Aneuploid reduction is favored over the hypothesis of aneuploid addition to explain the patterns of chromosome number variation within Potamogeton. Stebbins (1966) stated that both polyploidy and aneuploid reduction are characteristic of many outcrossing species which occupy pioneer habitats. Aneuploidy is often favored in pio- neer habitats since the causative unequal reciprocal translocations tie linked adaptive gene complexes (Stebbins, 1974). The genus Potamogeton includes outcrossing species which frequently occupy pioneer habitats. Correlates to the Precedence of the x = 7 Series The hypothesis that the x = 7 lineage gave rise to the x = 13 species of Potamogeton can be evaluated by how well other factors correlate with it. One factor involves the place of origin of the species. On the basis of fossil evidence, Miki (1937) suggested that Potamogeton originated in oriental Asia. When the geographical distributions of extant x = 7 species are studied, an eastern Asian affinity becomes apparent. Six of the eight known x = 7 species occur in eastern Asia whereas the two remaining species are found in North America (see Table 5). The x = 7 species are absent from other regions (with exception of the wide-ranging P. octandrus), and overall they represent a fairly small portion of the total number of species of Potamogeton. On the basis of these observations, a reas- onable interpretation of evolutionary progression can be formu- 314 Rhodora [Vol. 85 Table 5, Geographical distributions of x = 7 species of Potamogeton Species 2n___— Distribution Reference P. asiaticus 28 Manchuria, USSR Komarov (1937) P. cristatus 28 China, USSR, Ohwi (1965) Korea, Formosa P. octandrus 28 China, USSR, Korea, Ohwi (1965) Formosa, India, Malaysia P. orientalis 28 = Japan, eastern Asia Hagstrém (1916) P. sibiricus 28 E. Siberia (endemic) Komarov (1937) P. fryeri 42 Japan, Korea Ohwi (1965) P. vaseyi 28 United States & Hellquist & Crow Canada (1980) P. foliosus 28 North America Haynes (1974) lated. The concentration of x = 7 species in eastern Asia is com- patible with Miki’s hypothesis that Potamogeton originated there if this lineage is considered to be primitive. It is probable that diploid (2n = 14) species occurred throughout eastern Asia and that some of them dispersed into North America, which would account for the few x = 7 species on the latter continent. The disjunction of x = 7 species in eastern Asia and temperate North America coincides with the same distributional pattern of other plant species known to occur in both regions (e.g., Hara, 1972). According to Raven and Axlerod (1974), the migration of plants between North America and Eurasia has been relatively unimpeded throughout much of the evo- lutionary history of the flowering plants, allowing migration of spe- cies across the North Atlantic region until about 49 millions years BP. When or where the x = 13 lineage arose is undetermined, but these species were apparently more successful than their x =7 counterparts as evidenced by the larger number and wider distribu- tions of extant x = 13 species. The successful radiation of x = 13 species throughout most continents of the world makes tracing their precise origin difficult. The relative scarcity of x = 7 species which survive today may be due to unsuccessful competition with their x = 13 descendants. Indeed, the persistence of P. foliosus (2n = 28) 1983] Les — Potamogeton Be in North America may be a result of the species high tolerances for various ecological extremes. In comparing the sensitivity of com- mon aquatic plants, Stuckey (1975, p. 29) listed P. foliosus as an “insensitive species with wide ecological tolerances” which has “remained the same or increased in abundance through time.” The greater success of x = 13 species is also evident from observations that this lineage includes higher level polyploids and is the only group possessing species exhibiting derived hydrophilous pollina- tion (subgenus Coleogeton). Fruit characters also reflect the ancestral nature of x = 7 species. Fruits of x = 7 species are prominently keeled (Table 4) whereas both keeled and keeless fruits occur among x = 13 species. The loss of keels appears restricted to the x = 13 lineage. Since keeled fruits occur within both chromosomal series, it is assumed that the keel represents the primitive condition. This interpretation is further supported by the occurrence of keeless fruits within subgenus Coleogeton. Subgenus Coleogeton is regarded as the most highly advanced group of the genus and the keeless fruits of the subgenus are interpreted to represent the derived state. Several species within subgenus Potamogeton possess fruits having small or highly reduced keels indicating a series of stages in the reduction of the structures. Chromosome Numbers and Classification The “working rank” in Potamogeton is the subsection and princi- pal segregation of species occurs at this level. To be taxonomically meaningful, the subsections should represent monophyletic assem- blages of species. An evaluation of constituent chromosome numbers is one method by which the naturalness of subsections of Potamo- geton can be tested; however, a rationale for the analysis must be presented. In this genus, variation in chromosome number results from two processes — aneuploidy and polyploidy. As discussed above, it is probable that x = 7 represents the ancestral lineage. From this starting point are two possible explanations for the origin of aneuploid (x = 13) species. Figure | presents the two hypotheses. One possibility (single origin hypothesis) is that the x = 13 series originated by a single aneuploid event in which a 2n = 14 diploid gave rise to a 2n = 13 diploid which persisted by chromosomal doubling (2n = 26). The higher polyploids in the x = 13 series would have been derived from the 2m = 26 tetraploids. In this 2n=26, 52, 78, 104 TO 2n= MS single origin hypothesis Sp. A sp.B sp.c sp.D sp. E sp. F X=7—»x-13 X=7—»x=13 x=7—»x=13 X=7 multiple origin hypothesis Figure 1. Hypothetical Origins of x = 13 Species SIE elopoyy $8 190A] 1983] Les — Potamogeton 317 instance, the affinities between species would be greatest within the same numerical series and least between the two numerical series. In the second hypothesis, however, affinities could be much stronger between species in different series (e.g., sp. A and sp. B) than between those in the same series (e.g., sp. B and sp. F). If the aneuploid species arose from multiple origins, then one would expect the grouping of closely related species (such as represented by a subsection) to include members from both numerical series as well as from different ploidy levels. If aneuploid species arose from a single event, then the uniformity of subsections regarding both numerical series and ploidy levels would be expected. Referring back to Table 4, it is evident that the subsections reflect homoge- neity in terms of both numerical series and ploidy levels of the constituents. This observation supports the single-origin hypothesis of aneuploid species. One difficulty in accepting this hypothesis is that subsections should be entirely characterized by species belong- ing to only one numerical series (x = 7 or x = 13). Although this is generally true for the representatives surveyed in this study, there are exceptions. In subsections Pusilli, Oxyphylli and Amplifolii occur species from both x = 7 and x = 13 series. This discrepancy suggests a misalignment of species. On the basis of anatomy and inflorescence characters, both subsection Pusi/li and subsection Oxyphylli were noted by Hagstrém (1916) to be closely related to subsection Javanici, the only subsection comprised entirely of x = 7 species. It is probable, therefore, that the x = 7 species in subsection Pusilli (P. foliosus and P. orientalis) along with P. sibiricus from subsection Oxyphylli should rightfully be placed within or close to subsection Javanici instead of in their present situation. Such an adjustment would also result in a greater uniformity of fruit types within the subsections. Potamogeton foliosus, P. orientalis and P. sibiricus are also geographically related (Table 5). Likewise, the removal of P. fryeri from subsection Amplifolii may also be war- ranted. This species may represent a highly advanced member of subsection Javanici or could also be placed close to that subsection. Naturally, more detailed comparative work is needed before such taxonomic decisions can be fully justified. With regard to polyploidy, most subsections appear to be charac- terized by one ploidy level, although there are a number of instances where a subsection will also include a species which is at a higher 318 Rhodora [Vol. 85 ploidy level (e.g., subsections Pusilli, Compressi, Lucentes, Colo- rati). In such situations, chromosome number alone cannot provide sufficient evidence for determining the evolutionary relationships among species. Perhaps a comparative analysis of somatic karyo- types, banding, etc., would be more revealing than simple compari- sons of numbers. Because higher polyploids occur infrequently in a subsection and yet are found in various subsections, it is likely that polyploidy has occurred independently in divergent lines of 2n = 26 species thus producing hexaploids and higher levels. In such in- stances, a higher level polyploid may represent an advanced member of a particular group. It is also unlikely that certain species represent transitional ploidy levels in a continuum from 2n = 26 to 2n = 52, 78, or 104 species, and thus could tie together phylogenetically cer- tain subsections which are characterized by these ploidy levels. There is the question of whether a particular ploidy level (e.g., 2n = 52) has resulted from one polyploid event or from parallel polyploidy arising in divergent groups. A satisfactory elucidation of the problem may be better achieved as a result of biochemical, cytogenetic, and hybridization studies. St. John’s (1916) revision of section Coleophyili appeared concur- rently with Hagstrém’s (1916) monograph of Potamogeton. Inter- estingly, St. John considered P. filiformis, P. pectinatus, and P. vaginatus (as P. moniliformis) to belong to a single section (Cole- ophylli), whereas Hagstr6m viewed each species as belonging to different subsections, and further divided those into two sections (Connati and Convoluti). St. John also differed from Hagstrém by including P. robbinsii in section Coleophy/li while Hagstrém con- sidered the species to belong in a different subgenus ( Potamogeton) in section Adnati. Potamogeton filiformis, P. pectinatus, and P. vaginatus have identical chromosome numbers (27 = 78) which differ from that of P. robbinsii (2n = 52). P. robbinsii also differs from the three aforementioned species by its keeled fruit, lack of hypodermis, and presence of mechanical tissue in the peduncle (Hagstrém, 1916). In this respect, Hagstrém’s inclusion of P. rob- binsii in subgenus Potamogeton may be justified. On the other hand, St. John’s combination of P. filiformis, P. pectinatus and P. vaginatus into one section is supported by the agreement of chromosome number, morphology and fruit type of these species. 1983] Les — Potamogeton 319 Floating-leaved Species — Primitive or Advanced? As previously mentioned, correlations have been attempted be- tween chromosome numbers and the presence of floating leaves. A perspective of the problem developed by examining a larger group of species fails to substantiate this correlation. The lowest repre- sentative ploidy level at which to make such a comparison Is tetra- ploid (2n = 26 or 28). At the tetraploid level, some species in each chromosomal series have floating leaves whereas others have only submersed leaves, thus presenting no clue to which morphological type arose first. Even the apparent loss of floating leaves in the advanced, high polyploid subgenus Coleogeton cannot be regarded as characteristic of high ploidy levels in general since floating leaves are still found at even higher ploidy levels (e.g., P. illinoensis). The primitiveness of floating leaves can be neither proven nor disproven solely on the basis of chromosome numbers, although other evi- dence may be brought to bear on the question. From an adaptive viewpoint, it is difficult to rationalize total vegetative submergence for an ancestor characterized by wind pollination. The arguments favoring an aquatic origin for Potamogeton have not been ade- quately substantiated although evidence for a terrestrial ancestry of the group is provided by Miki (1937) and den Hartog (1970). Miki believed Potamogeton to have originated from a different subclass (Arecidae) than is usually considered, i.e., Alismatidae (Cronquist, 1981: Takhtajan, 1969), thereby destroying the “link” to the aquatic members of the Alismatidae. Den Hartog believed that marine representatives of Potamogetonaceae arose from terrestrial ances- tors. If this is the case, then the freshwater Potamogeton are also likely to be of terrestrial ancestry. The more reasonable explanation still appears to be that Pofamogeton originated with floating leaves as an intermediate stage between a terrestrial ancestry and adoption of the hydric habit. Also supporting this argument is the study by Kadono (1982) wherein the epistomatous nature of submersed and seedling leaves of several Potamogeton species suggested to him that submersed leaves arose from floating leaves, and that the genus possibly evolved from ancestors having floating leaves. The appearance and disappearance of floating leaves throughout the genus is no doubt partially due to reversals. Although the genetic basis for heterophylly in Potamogeton is not well under- 320 Rhodora [Vol. 85 stood, some support for this opinion exists. Heterophylly can be regulated in P. nodosus simply by varying the concentration of abscisic acid (Anderson, 1978) thus implying that the genetic mech- anism responsible for controlling the foliar states is relatively sim- ple. Thieret (1971) observed the production of floating leaves in P. richardsonii (a species that normally produces only submersed foliage) as an apparent response to dropping water levels. This changeover indicates that the genetic mechanism may still exist to produce floating leaves in species which were thought to have lost the ability to do so. Hagstrém (1916) apparently recognized conver- gence of floating-leaved species since he included both floating- leaved and submersed-leaved species within subsections on the basis of other overwhelming similarities (e.g., subsections Lucentes and Serrulati). CONCLUSIONS A survey of chromosome numbers reported for species of Potamo- geton has substantiated the view that the genus is characterized by two different polyploid series, one based upon x = 7 and the other upon x = 13. The diploid level of the genus is interpreted to be 2n = 14. The x = 13 lineage is proposed to have arisen from one incidence of aneuploidy from a 2n = 14 progenitor. Chromosome number data support the hypothesis that Potamogeton originated in Asia. A review of chromosome numbers indicates that subsections Pusilli, Oxyphylli, and Amplifolii may not represent entirely natural assemblages and it is suggested that a reconsideration of the classifi- cation of Potamogeton be made in order to accommodate discrepan- cies noted in regard to chromosome numbers and other features. Cytological investigations of large population samples of species are encouraged in order to obtain better insight into evolutionary pro- cesses occurring within the genus. Chromosome numbers cannot be used to directly demonstrate the primitiveness of floating leaves in the genus, although it is believed that additional evidence supports this view. Available evidence suggests a different overview of evolu- tionary patterns in this genus than has been previously made. Hope- fully, a complete synthesis of morphological, anatomical, cyto- genetic, biochemical and other data will eventually enable a more accurate phylogenetic reconstruction of Potamogeton to be made and thus will provide systematists with the best classification to use for this genus. 1983] Les — Potamogeton 321 ACKNOWLEDGMENTS | would like to express my thanks to Dr. D. J. Crawford (The Ohio State University), Dr. R. R. Haynes (The University of Ala- bama) and Dr. R. L. Stuckey (The Ohio State University) for their helpful comments and suggestions for improving this manuscript. LITERATURE CITED AALTo, M. 1970. Potamogetonaceae fruits. I]. Recent and subfossil endocarps of the Fennoscandian species. Acta Bot. Fenn. 88: 1-85. 1974. Potamogetonaceae fruits. Il. Potamogeton robinnsii, a seldom fruiting North American pondweed species. Ann. Bot. Fenn. 11: 29-33. ANDERSON, L. W. J. 1978. Abscisic acid induces formation of floating leaves in the heterophyllous aquatic angiosperm Potamogeton nodosus. Science 201: 1135-1138. Arper, A. 1920. Water Plants: A Study of Aquatic Angiosperms. University Press, Cambridge, 436 pp. ASCHERSON, P., & P. GRAEBNER. 1907. IV. II. Potamogetonaceae /n. A. Engler. Das Pflanzenreich. Leipzig. Verlag von Wilhelm Engelmann. 184 pp. BOLKHOVSKIKH, Z., ET AL. 1969. Chromosome Numbers of Flowering Plants [in Russian]. Acad. Sci. of the USSR, Leningrad. 926 pp. CANDOLLE, AUGUSTE P. de. 1827. Organographie Vegetale. Vol. 1, book 2, chap- ter 3. Paris. Cave, M.S.,ed. 1958-1965. Index to plant chromosome numbers for 1956-1964. Index for the years 1956-57 published by California Botanical Society; Index for 1958-1964 published by the University of North Carolina Press, Chapel Hill. Index to plant chromosome numbers supplement (previous to 1956) pub- lished in 1959 by California Botanical Society. CurysLer. M.A. 1907. The structure and the relationships of the Potamogetona- ceae and allied families. Bot. Gaz. (Crawfordsville) 44: 161-188. Cook, C. D. K., B. J. Gut, E. M. Rix, J. SCHNELLER, & M. Seitz. 1974. Water Plants of the World. W. Junk b.v., publishers, The Hague, The Netherlands. 561 pp. Cronguist, A. 1968. The Evolution and Classification of Flowering Plants. Houghton Mifflin Co., Boston. 396 pp. 1981. An Integrated System of Classification of Flowering Plants. Colum- bia University Press, New York. pp. 1031-1036. DARLINGTON, C. D., & A. P. Wyte. 1956. Chromosome Atlas of Flowering Plants. Macmillan Co., New York. p. 337. EnNRENDOREFER, F., F. KRENDL, E. HABELER, & W. SAUER. 1968. Chromosome numbers and evolution in primitive angiosperms. Taxon 17: 337-468. FERNALD, M.L. 1932. The linear-leaved North American species of Potamogeion section Axillares. Mem. Amer. Acad. Arts. 17: 1-183. Go.psatt, P. 1979. Polyploidy in Angiosperms: Monocotyledons, Jn: W. H. Lewis, ed. Polyploidy: Biological Relevance. Plenum Press, New York. 583 pp. _ed. 1981. Index to plant chromosome numbers 1975-1978. Monographs in Systematic Botany, Vol. §. Missouri Botanical Garden, St. Louis. 553 pp. 322 Rhodora [Vol. 85 GRANT, V. 1963. The Origin of Adaptations. Columbia University Press, New York. 1982. Periodicities in the chromosome numbers of the angiosperms. Bot. Gaz. 143: 379-389. HaGstrom, J.O. 1916. Critical researches on the Potamogetons. Kongl. Svenska Vetenskapsakad. Handl. 55: 1-281. Hara, H. 1972. Corresponding taxa in North America, Japan and the Hima- layas. pp. 61-72, In: D. H. Valentine (ed.), Taxonomy, Phytogeography and Evolution. Academic Press, New York. HARADA, I. 1956. Cytological studies in the Helobiae |. Chromosome idiograms and a list of chromosome numbers in seven families. Cytologia 21: 306-328. HartoG,C. DEN. 1970. The Sea-grasses of the World. North-Holland Publishing Co., Amsterdam. 275 pp. Haynes, R. R. 1974. A revision of North American Potamogeton subsection Pusilli (Potamogetonaceae). Rhodora 76: 564-649. HELiquist, C. B., & G. E. Crow. 1980. Aquatic Vascular Plants of New Eng- land: Part |. Zosteraceae, Potamogetonaceae, Zannichelliaceae, Najadaceae. New Hampshire Agr. Exp. Sta. Bull. 515. 68 pp. Kapono, Y. 1982. On the stoma found in leaves of seedlings of some Potamo- geton species. J. Jap. Bot. 57: 332-336. Komarov, V. L.,ed. 1937. Flora of the USSR (Translated from Russian in 1970 by Israel Program for Scientific Translations, Jerusalem). Vol. 1. National Science Foundation, Washington, D.C. 244 pp. KULESZANKA, J. 1934. Rozwoz ziarn pylk u Potamogeton fluitans. Acta. Soc. Bot. Poloniae 11: 457-462. Love, A.& D. Love. 1974. Cytotaxonomical Atlas of the Slovenian Flora. Verlag von J. Cramer. pp. 32-36. 1975. Cytotaxonomical Atlas of the Arctic Flora. J. Cramer. Inder A. R Ganter Verlag Kommanditgesellschaft. FL-9490 Vaduz. pp. 25-29. Miki, S. The origin of Najas and Potamogeton. Bot. Mag. 51: 472-481. Moore, R. J.,ed. 1970. Index to plant chromosome numbers for 1968. Regnum Veg. 68: 16-17. 1971. Index to plant chromosome numbers for 1969. Regnum Veg. 77: 7. ———. 1972. Index to plant chromosome numbers for 1970. Regnum Veg. 84. 1973. Index to plant chromosome numbers 1967-1971. Oostoek’s Uitgever- smaatschappij B. V., Domstraat 5-13, Utrecht, Netherlands, pp. 47-48. 1974. Index to plant chromosome numbers for 1972. Regnum Veg. 91. 1977. Index to plant chromosome numbers for 1973/74. Regnum Veg. 96: 179. OcpEN, E. C. 1943. The broad-leaved species of Potamogeton. Rhodora 45: 57-105, 119-163, 171-214. 1953. Key to the North American species of Potamogeton. New York State Museum Circular 31, Albany. 11 pp. Ouwti, J. 1965. Flora of Japan. Smithsonian Institution, Washington, D.C. 1067 pp. OrnpburF, R.,ed. 1957. Index to plant chromosome numbers for 1965. Regnum Veg. 50. 1983] Les — Potamogeton 323 1968. Index to plant chromosome numbers for 1966. Regnum Veg. 55: 13. 1969. Index to plant chromosome numbers for 1967. Regnum Veg. 59: 19. PALMGREN, O. 1939. Cytological studies in Potamogeton. Bot. Not. 1939: 246-248. RAUNKIAER, C. 1895-99. De danske blomsterplanters naturhistorie. Bd. 1. Enkim- bladede. Copenhagen. 724 pp. Raven, P. H., & D. I. AXELROD. 1974. Angiosperm biogeography and past con- tinental movements. Ann. Missouri Bot. Gard. 61: 539-673. SCULTHORPE, C. D. 1967. The Biology of Aquatic Vascular Plants. Edward Arnold, London. 610 pp. St. JOHN, H. 1916. A revision of the North American species of Potamogeton of the section Coleophy ili. Rhodora 18: 121-138. 1925. A critical consideration of Hagstrém’s work on Potamogeton. Bull. Torrey Bot. Club 52: 461-471. STEBBINS, G. L. 1966. Chromosomal evolution and variation. Science 153: 1463-1469. 1971. Chromosomal Evolution in Higher Plants. Edward Arnold, Lon- don. 216 pp. 1974. Flowering Plants - Evolution Above the Species Level. Belknap Press of Harvard University Press, Cambridge. 399 pp. STERN, K. R. 1961. Chromosome numbers in nine taxa of Potamogeton. Bull. Torrey Bot. Club 88: 411-414. StuCKEY, R. L. 1975. Submersed aquatic vascular plants as indicators of envir- onmental quality. /n: C. C. King and L. E. Elfner (eds.). Organisms and biolog- ical communities as indicators of environmental quality -a symposium. Ohio Biological Survey Informative Circular No. 8. pp. 27-30. TAKHTAJAN, A. 1969. Flowering Plants -Origin and Dispersal. Oliver and Boyd: Edinburgh. 310 pp. Tuieret, J. W. 1971. Observations on some aquatic plants in northwestern Min- nesota. Michigan Bot. 10: 117-124. ToMLINSON, P. B. 1982. VII. Helobiae (Alismatidae) (including the seagrasses). In: C. R. Metcalfe, ed. Anatomy of the monocotyledons. Clarendon Press, Oxford. 522 pp. WIEGAND, K. M. 1898. Notes on the embryology of Potamogeton. In: W. F. Ganong “Society for Plant Morphology and Physiology” report of the Ithaca meeting, December 28-29, 1897. Bot. Gaz. 25: 116. 1899. The development of the microsporangium and microspores in Convallaria and Potamogeton. Bot. Gaz. 28: 328-359. WISNIEWSKA, E. 1931. Rozwoz ziarn pylku u Potamogeton perfoliatus L. Acta. Soc. Bot. Poloniae 8: 157-174. COLLEGE OF BIOLOGICAL SCIENCES DEPARTMENT OF BOTANY, THE OHIO STATE UNIVERSITY COLUMBUS, OHIO 43210 DISTRIBUTION OF PODOSTEMUM CERATOPHYLLUM MICHX. (PODOSTEMACEAE)! C. THOMAS PHILBRICK? AND GARRETT E. CROW ABSTRACT Podostemum ceratophyllum is an often overlooked freshwater dicot, restricted to areas of swift moving river currents—an apparently ecologically sensitive habitat. Selected specimens representing the entire range of the plant are listed. Podostemum ceratophyllum (Riverweed), a freshwater aquatic dicot, grows attached to rocks in river rapids and waterfalls. The plant tends to be undercollected because of its often inaccessible habitat and inconspicuous appearance (often mistaken for an aquatic moss or freshwater alga) (Philbrick, 1981, 1982). Consequently, its current distribution is unclear. The purpose of this paper is to pro- vide a distributional record of P. ceratophyllum based on specimens from 41 herbaria. Podostemum ceratophyllum is the single North American member of an otherwise tropical family, Podostemaceae. Podostemum cera- tophyllum exhibits the widest distribution of any species in the genus, ranging from central Nova Scotia, central New Brunswick and southern Quebec, southward along the eastern United States to Georgia, and westward to Kentucky and eastern Oklahoma (Fig. 1). Raymond (1950) reports a location in eastern Wisconsin but we have not seen a specimen that documents this site. The species is also represented in the West Indies (Dominican Republic) and in Central America (Honduras, as var. circumvallatum), although very few specimens document this southernmost portion of its range. The remaining 16 or so species in the genus occur in the southern hemi- sphere from southern Brazil south to Argentina and Paraguay (Royen, 1954). There are also species reported from Africa and India, but their position in this genus is uncertain (Graham & Wood, 1975, Royen, 1954; Willis, 1902). \Scientific contribution 1240 from the New Hampshire Agricultural Experiment Station. 2Present address: Biological Sciences Group, Box U-43, Univ. of Connecticut, Storrs, CT 06268 a25 326 Rhodora [Vol. 85 =\ =“ ~. ; : Figure |. Distribution of Podostemum ceratophyllum. 1983] Philbrick & Crow — Podostemum 327 The lack of documentation of Podostemum ceratophyllum was briefly discussed by Muenscher and Maguire (1931) for New York. It has recently been accessed as rare in New England (Crow et al., 1981). However, during the summer of 1981 numerous previously unknown stations for the plant were documented in southern New Hampshire and southern Maine (Philbrick, 1982). This suggests that the plant is much more common in this region than previously believed. On the other hand, while many locations for Podostemum have been documented in New England since the late 1800’s, several of the earlier known stations no longer harbour the plant. The plant has not been collected from the first known New England station (Killingworth, Conn.) since 1925, and Countryman (1978) reported its absence at Jamaica, the earliest known Vermont site. Recent searches of Rhode Island stations were unsuccessful in finding the plant. In many respects the habitat of Podostemum ceratophyllum is eco- logically harsh, as illustrated by the fact that there are virtually no an- giosperms with which Podostemum must compete for growing space. Several species of Potamogeton, however, occasionally grow in the gravel between the rocks on which Podostemum is attached. This habitat also seems ecologically sensitive. Any manipulation of this environment resulting in substratum modification, water flow changes, or deterioration of water quality could result in the death of the plant. Drastic changes in water flow seem to be the most obvious cause of extirpation of the plant. Countryman (1978) reports that the West River population at Jamaica, Vermont, seems to have been ex- tirpated due to river flow manipulation. It is also believed that one of the two documented New Brunswick, Canada, locations is currently inundated to a depth that prevents growth (H. Hinds, pers. comm). Water pollution and siltation may also have drastic negative effects on P. ceratophyllum. Such factors may have resulted in its loss from a location in the Cocheco River at Dover, New Hamp- shire. Although numerous specimens document the site, we could not locate any plants during 1980-1982. Its apparent disappearance may be related to a mass fish-kill in the Cocheco River. This fish-kill is reported to have been linked to industrial waste dumping in July of 1973, several miles upstream from the Podostemum population (Thayer, 1973, unpublished). 328 Rhodora [Vol. 85 Although Podostemum is, no doubt, sensitive to specific pollu- tants, all pollutants are not detrimental to the species. The first author has collected the plant from the Mousam River, Kennebunk, Maine, where the river is noticeably polluted by domestic sewage. The Winooski River in South Burlington, Vermont, is also polluted (W. Countryman, pers. comm.) but the plant grows well. It is there- fore apparent that the plant is not as good an indicator of clean water as Meijer (1976) proposed. ACKNOWLEDGEMENTS We thank Thomas D. Lee and Debra A. Dunlop for their com- ments on the manuscript, and the curators of the herbaria listed below for allowing us to examine their specimens of Podostemum ceratophyllum. We also thank Alex Wilson for providing specimens of the previously unreported Nova Scotia population. LITERATURE CITED COUNTRYMAN, W. D. 1978. Rare and Endangered Vascular Plants Species of Vermont. The New England Botanical Club in cooperation with the U.S. Fish and Wildlife Service [Newton Corner, MA]. Crow, G. E., W. D. Countryman, G. L. CHurcH, L. M. EASTMAN, C. B. HELL- guist, L. J. MEHRHOFF, & 1.M.Storks. 1981. Rare and endangered vascu- lar plant species of New England. Rhodora 83: 259-299, GRAHAM, S. A.,& C. E. Woop, Jk. 1975. The Podostemaceae of the southeastern United States. J. Arn. Arb. 56: 456-465. MEER, W. 1976. A note of Podostemum ceratophyllum Michx. as an indicator of clean streams in and around the Appalachian Mountains. Castanea 41: 319-324. Muenscuer, W. C., & B. MAGuiRE. 1931. Notes on some New York plants. Rhodora 33: 165-167. PHILBRICK, C. T. 1981. Some notes regarding pollination in a New Hampshire population of Podostemum ceratophyllum Michx. (Podostemaceae). Rhodora 83: 319-321. 1982. New locations for Podostemum ceratophyllum Michx. (Podoste- maceae) in New Hampshire and Maine, with some comments on a unique floral form. Rhodora 84: 301-303. RAYMOND, M. 1950. Esquisse phytogeographique du Quebec. Mem. Jarb. Bot. Montreal 5: 1-447. RoyeNn, P. vAN. 1954. The Podostemaceae of the New World. Part III. Act. Bot. Neerl. 3: 215-263. 1983] Philbrick & Crow — Podostemum 329 Tuayer, T. A. 1973. The Cocheco River, its Composition and Effect on the Community. (xeroxed report). University of New Hampshire, Durham. WILLIs, J.C. 1902. A revision of the Podostemaceae of India and Ceylon. Ann. Roy. Bot. Gard. (Peraedeniya) 1: 181-465. DEPARTMENT OF BOTANY AND PLANT PATHOLOGY UNIVERSITY OF NEW HAMPSHIRE, DURHAM, NH 03824. APPENDIX: REPRESENTATIVE SPECIMENS The following locality information is based on specimens from ACAD, APSU, CM, CLEMS, CONN, DUKE, F, FLAS, GA, GH, HNH, LAF, MAINE, MASS, MICH, MIN, MO, NCSC, NCU, NEBC, NHA, NLU, NO, NSPM, NY, OS, OKLA, PH, SMU, TENN, UARK, UNA, US, USF, VDB, VPI, VT, WIS, wva, Tennessee Valley Authority collection (TVA), and Western Kentucky University (wkU). For simplicity most fractional units have been changed to decimal. CANADA. NEW BRUNSWICK. Northumberland Co.: Quarryville, Webster and Fielding 225 (1955) (ACAD, NSPM); near Woodstock, Eel River, Macoun 22593 (1899) (CA, F). NOVA SCOTIA. Lunenburg Co.: LaHave River, Wilson and Wright, s.n. (1982) (NHA, NSPM). ONTARIO. Hastings Co.: Actinolite, Skoutamata River, Catling, s.n. (1977) (0S); vicinity of Ottawa, Brighams Creek, Macoun 5907 (1898) (F). QUEBEC. Argenteuil Co.: opposite Hawkesburg, Dore 18641 (1960) (LAF, NHA). Deuxmontagnes Co.: Ile Yale, Mille Iles River, Marie- Victorin, Gervais and Lavigne, s.n. (1969) (us); Ile Bizard, Marie- Victorin 22081 (1925) (MO, WIS, PH, MIN, F. ACAD); St. Eustache, Victorin 3212 (1916) (MO, NY, US). Gatineau Co.: Hull, Macoun, s.n. (1894) (Mo, MICH, NY, MIN); Hull, Ottawa River, Moyle, s.n. (1959) (MIN). Vaudreuil Co.: Ile Perrot, Cody and Dore 6565 (1952) (PH, MAINE, WIS, MICH, US, NY); Ile Des Pins, Vaudreuil, Marie-Victorin, gs.n. (1960) (Us). UNITED STATES. ALABAMA. Bibb Co.: SW of West Blockton, Little Cahaba River, Haynes 7462 (1979) (UNA). Cherokee Co.: Centre, Terrapin Creek, Kral 31759 (1968) (Os, SMU, VDB); Ellisville, 330 Rhodora [Vol. 85 Terrapin Creek, Haynes 7329 (1979) (UNA). Clay Co.: Lineville, Crooked Creek, Haynes 7372 (1979) (UNA, VDB). Coosa Co.: SW of Rockford, Swamp Creek, Haynes 7904 (1980) (UNA). Covington Co.: “the shoals”, Patsaliga Creek, Harper 106 (1906) (GH, MO, NY, us). De Kalb Co.: Little River, Wiersema 376 (1978) (UNA); NW of Gaylesville, Little River, Haynes 6587 (1978) (UNA); | mi. W of junction with Little River, Hicks Creek, Clark and Landers 5101 (1966) (Ncu); Lake Guntersville State Park, Town Creek, Meigs 930 (1980) (UNA). Geneva Co.: 5 mi. SW of Coffee Springs, Double Bridges Creek, Mc Danie! 8566 (1967) (vpB). Lauderdale Co.: Flor- ence, Cypress Creek, Snoddy, s.n. (1978) (TVA); N of Florence, Little Cypress Creek, Haynes 6784 (1978) (UNA, VDB); Pruitton, Butler Creek, Harper 3870 (1942) (F, GH, PH, US, WIS); Mussell Shoals, Tennessee River, Harper, s.n. (1922) (Ny, PH, US). Lee Co.: near Bean’s Mill, Halawakee Creek, Wiersema 831 (1979) (UNA); Che- wacla State Park, Chewacla Creek, Johnson 5/1 (1959) (NCU); Wright’s Mill, Little Uchee Creek, Haynes 7858 (1980) (UNA). Marshall Co.: 4.5 mi. SSE of Lake Guntersville State Park, Short Creek, Webb 3005 (1980) (TVA). Randolph Co.: Wadly, Tallapoosa River, Haynes 7354 (1979) (UNA). Tallapoosa Co.: N of Reeltown, Songahatchee Creek, Haynes 7879 (1980) (UNA); Samantha, North River, Daven- port 1204 (1979) (UNA, vpB); Alexander City, Elkahatchie Creek, Kral 43308 (1971) (vps). ARKANSAS. Conway Co.: Petit Jean State Park, Cedar Creek, Morrilton 722 (1955) (TENN, UARK). Franklin Co.: Taft, Mulberry Creek, Moore, s.n. (1952) (UARK). Montgomery Co.: Caddo Gap, Caddo River, Moore 54-11] (1954) (F, TENN, UARK, WIS); Pine Ridge, Ouachita River, Harris, Basco and Douglas, s.n. (1977) (NLU). Pope Co.: Ozark National Forest, Big Pingy Creek, Hamilton 41 (1969) (NcU); Dover, Twin Bridges, Tucker 5791] (1967) (NCU); Cossatut Creek, Moore 510673 (1951) (UARK). Van Buren Co.: Shir- ley, Little Red River, Clark 92b (1971) (UARK). CONNECTICUT. Fairfield Co.: Bridgeport, Pequomock River, Eames 9519 (1917) (CONN, NEBC); Fairfield, Mill River, Eames and Godfrey 5296 (1906) (CONN, MIN); Huntington, Farmill River, Eames 5290 (1905) (CONN, MIN); Newtown, Nichols, s.n. (1979) (NEBC); Trumbull, Pequonnock River, Eames 11722 (1935) (CONN, GH). Hartford Co.: Windsor, “Rainbow”, Clark, s.n. (1904) (CONN). Middlesex Co.: Killingworth, [Hammonasset River], Hall, s.n. 1983] Philbrick & Crow — Podostemum 331 (1874) (CONN, NEBC); East Haddam, Devil’s Hopyard, Eames, s.n. (1933) (CONN). New Haven Co.: Oxford, Little River, Harger 2356 (1902) (CONN, GH). Tolland Co.: Mansfield, Mt. Hope River, Mac- gregor and Torrey 4391 (1948) (CONN (2)); Mansfield, Willimantic River, Torrey 4410 (1948) (CONN (3)). Windham Co.: South Chap- lin, Natchaug River, Heisey, s.n. (1974) (CONN); Canterbury, Little River, Goldstein 10] (1974) (CONN). DELAWARE. New Castle Co.: Wilmington, Brandywine Creek, Canby, s.n. (1883) (Mass); Faukland, Red Clay Creek, Commons, s.n. (1886) (CM, USF); Greenbank, Red Clay Creek, Commons, s.n. (1884) (Mo, NY (2), PH US). GEORGIA. Cherokee Co.: SW of Canton near Cherokee Mills, Little River, Duncan 8336 (1948) (GA). Clarke Co.: SW of Athens, Bobbin Mill Creek, Pyron and McVaugh 78 (1935) (No); Tallassee Shoals, Pyron, s.n. (1935) (CLEMS, GA). Dawson Co.: Dawsonville, Amicalola Creek, Pyron and McVaugh 1006 (1936) (GA). Dough- erty Co.: Albany, Flint River, Chapman, s.n. (unknown date) (NY); Muckafoonee Creek, Harper 1950 (1903) (F, GH, MO, NY, US). Elbert Co.: NW of Elberton, Broad River, Duncan 11714 (1950) (GA, GH, ncsc); Ga. Highway 172, Broad River, Coile and Bates 1994 (1978) (FLAS, NCU). Floyd Co.: Rome, Coosa River, Chapman, s.n. (1937) (F, MO, US). Fulton Co.: SE of Roswell, Chattahoochee River, Dun- can and Venard 16522 (1953) (GA, NCSC); NW of Norcross, Chatta- hoochee River, Welch, s.n. (1965) (CM, GA). Gilmer Co.: Doll Mountain public access area, Coosawattee River, Kra/ 50866 (1973) (vps). Gwinnett Co.: | mi. below Old Jones Bridge, Norcross, Huisch, s.n. (1964) (GA). Habersham Co.: Soque River, Nee, Peet and Arnold 1748 (1969) (wis); S of Clayton, Tallulah River Gorge, Michener 1430 (1975) (Ncu). Harris Co.: N of Fortson, Duncan 9248b (1949) (GA); Robert Lewis Scout Camp, Mulberry Creek, Jones 21868 (1972) (NCU). Hart Co.: NW of Hartwell, Shoal Creek, Duncan 7821 (1947) (GA). Jones Co.: E of James, Altamaha River, Suttkus 2518 (1955) (Mo, No); Highway 22, Commissoner Creek, Musselman 4402 (1972) (NCU). Madison Co.: S of Carlton, Andrew’s Shoals, south fork of Broad River, Scott, s.n. (unknown date) (GA, GH, MO); Comer, Pyron, s.n. (1934) (DUKE). Madison- Oglethorpe Co. Line: S of Carlton, south fork of Broad River, Suttkus, s.n. (1953) (DUKE, NCU (2), NO, NLU). McDuffie Co.: Rt. 79, Little River, Leeds, s.n. (1933) (pH). Meriwether Co.: Flat Shoals, 332 Rhodora [Vol. 85 Flint River, W.A./. and B.B.H., s-n. (1938) (GA, LAF, NCU, NLU, WVA). Morgan Co.: Farmington, Double Shoals, Appalachia River, Jones 21012 (1971) (GA). Newton Co.: Covington, Bear Creek, Rogers, s.n. (1963) (GA). Oglethorpe Co.: Lexington, Echols Mill, Montgomery 933 (1967) (DUKE, F. FLAS, GA, LAF, MASS, MO, NCU, NCSC, NLU, NO, NY (2), PH, TENN, UNA (2), USF, VDB, WVA); Comer, south Broad River, Pyron and McVaugh 41 (1935) (DUKE, F, GH, MO, PH (2)); Carlton, Andrew’s Mill, south fork Broad River, Dun- can 11647 (1950) (DUKE, FLAS, GH, MICH, MIN (2), MO, NCSC, SMU, TENN, UNA, US). Pike-Meriwether Co. Line: W of Concord, Flint River, Pyron and McVaugh 1831 (1937) (GA). Rabun Co.: base of Big Mountain, Chattooga River, DuMond 1189 (1968) (Ncsc, WKU); WSW of town of Pine Mountain, Surtkus 2531 (1955) (GA, LAF, NO). Rockdale Co.: Logansville, Haynes Creek, Pyron and McVaugh 552 (1936) (MICH, US). Stevens Co.: | mi. upstream from confluence with Tugaloo River, Panther Creek, Boufford, Massey and Katz 16988 (1975) (NCU); N of Toccoa, Cedar Creek, Boufford and Wood 16799 (1976) (NCU); S of Toccoa, Broad River, Scort 9] (1949) (GA). Talbot Co.: ENE of Manchester, Flint River, Jones 21098 (1971) (GA). Towns Co.: W of Hiawassee, Hiawassee River, Wherry and Pennell 14075 (1927) (pH); Hiawassee, Center Creek, Scott, s.n. (unknown date) (GA). Upson Co.: 3 mi. W of Thomaston, Potatoe Creek, Mc Vaugh 8939 (1948) (MICH, Mo, NLU, NCU, SMU, VDB); SE of Woodbury, Flint River Gorge of Pine Mountain, Jones and Faust 22036 (1972) (DUKE, GA). Walton Co.: Athens, Apalachee River, Pyron and McVaugh 66 (1935) (Ga). White Co.: E of Helen, Duncan 20778 (1958) (Ga). KENTUCKY. Fayette Co.: Kentucky River, Perer, s.n. (1833) (MICH, MASS). Harlan Co.: near Harlan Court House, Kearney 19 (1893) (GH, MIN, US). Jefferson Co.: Falls of Ohio River, Short, s.n. (“1840”) (Ny (3), PH). McCreary Co.: Tennessee-Kentucky Line, big south fork of Cumberland River, Wofford 78-99 (1978) (TENN). Menifee-Wolfe Co. Line: E of Gladie Creek, Red River, Meijer, s.n. (1974) (wKv). Whitley Co.: Cumberland Falls, Cumberland River, McFarland 95 (1940) (DUKE, GA, GH, MICH, MIN (2), MO, NHA, NY, PH, TENN, US, WVA). Whitley-McCreary Co. Line: Cumberland Falls, Braun 2633 (1939) (us, WIS). LOUISIANA. St. Helena Parish: NE of Montpelier, Tickfaw River, Allen 1923 (1972) (su). St. Tammany Parish: NW of Covington, 1983] Philbrick & Crow — Podostemum 333 Bogue Falaya, Thieret 27013 (1967) (DUKE, GH, LAF, VDB). Tangipa- hoa Parish: Amite River, Cooley and Brass 4183 (1955) (usF); E of Independence, Tangipahoa River, Thomas 40377 (1974) (CM, FLAS, GA, MASS, MO, NCU, NLU, PH). Washington Parish: W of Warnerton, Pearl River, Suttkus 2770 (1959) (No); SW of Franklinton, Bogue Chitto River, Thieret 29307 (1968) (DUKE, FLAS, GA, NCU, SMU). MARYLAND. Baltimore Co.: unknown locality, Shreve, s.n. (1902) (MIN). Cecil Co.: ESE of Pilot, Conowingo Creek, Wilkens 6855 (1941) (NcU, PH). Garrett Co.: Deep Creek, Drouet 2346 (1938) (F); Swallow Falls State Forest, Muddy Creek, 8 mi. NNW of Oakland, Herman 14915 (1958) (us); Savage Station, Williamson, S.n. (1905) (PH (3)). Howard Co.: Savage Station, Patuxent River, House 152] (1905) (Mo). Prince Georges Co.: NW of Bowie, Patuxent River, Hermann 9940 (1938) (F, NY). MAINE. Cumberland Co.: Yarmouth, Royal River, Philbrick 1136 (1981) (NHA). Cumberland-Sagadahoc Co. Line: Brunswick-Top- sham line, Androscoggin River, Philbrick 1135 (1981) (NHA). Ken- nebec Co.: Benton Village, Sebasticook River, Rossbach 4965 (1959) (MAINE, NCU); Clinton, Sebasticook River, Rossbach 5029 (1959) (MAINE (4), NCU); West Gardiner, Collin’s Dam, Fassett 18295 (1936) (F, GH, MAINE, MIN, MO, NEBC, NY). Lincoln Co.: Alna, Sheepscott River, Philbrick 1113 (1981) (NHA); Cooper’s Mills, Sheepscott River, Philbrick 1117 (1981) (NHA); Whitefield, Sheep- scott River, Philbrick 1115 (1981) (NHA). Penobscot Co.: Alton, Pushaw Stream, Steinmetz 1188 (1942) (MAINE); Bradley, Chemo Stream, Merrill 861 (1898) (us); Old Town, Gilman Falls, Steinmetz and Ogden 840 (1937) (CONN, F, GH, MAINE, MASS, MICH, MO, NCSC, NCU (2), NEBC, NY, OKLA, PH (2), SMU, US, VT, WIS, WVA); Old Town, Gilman Falls, Bicknell, s.n. (1952) (MAINE); Orono, Pushaw Stream and Stillwater River, Steinmetz, s.n. (1939) (pH, (2)). York Co.: Kennebunk, Mousam River, Philbrick 1137 (1981) (NHA), Sanford, Mousam River, Philbrick 1142 (1981) (NHA). MASSACHUSETTS. Franklin Co.: Deerfield, Poland, s.n. (1959) (mo). Hampshire Co.: Amherst, Connecticut River, Jesup, S.n. (unknown date) (NHA); Belchertown, Jabish Brook, Thies, Thies, Thies and Seymour 5521 (1940) (wis (2)); Mt. Holyoke, Connecticut River, Jesup, s.n. (1874) (MASS, HNH); South Hadley, Stony Brook, Ahles 84604 (1977) (Mass); Southhampton, unknown collector, s.n. (unknown date) (Mo). Middlesex Co.: South Natick, Charles River, 334 Rhodora [Vol. 85 Faxon, s.n. (1883) (GH, MASS, MIN, NEBC (2), US); South Natick, Charles River, Philbrick 1124 (1981) (NHA). MISSISSIPPI. Clarke Co.: 2 mi. N of Enterprise, Chunky River, Dunn’s Fall, McDaniel 23005 (1979) (No, VDB, TENN). Covington Co.: N of Sumrall, Bowie River, Barnes 658 (1976) (vps). Lauder- dale Co.: E of Chunky, Chunky River, McDaniel and Clark 14602 (1970) (GA). Marion Co.: E of Sandy Hook, Pearl River, Suttkus 2537 (1955) (No). Pike Co.: between Pricodale and Hulmerville, Topisaw Creek, Robers 7916 (1972) (NCU, NLU, TENN, USF, VDB). Simpson Co.; 8 mi. SW of Pinola, Strong River, Wallus, s.n. (1971) (FLAS, NLU, USF). Tishomingo Co.: Tishomingo State Park, Bear Creek, Pollen 65351 (1965) (NCU); Bear Creek, Rogers 45866 (1978) (TENN). NEW HAMPSHIRE. Belknap Co.: Barnstead, Suncook River, Krochmal 1496 (1948) (NHA). Carroll Co.: Effingham-Freedom line, Ossippee River, Philbrick 1134 (1981) (NHA). Coos Co.: N of Lan- caster, “Pothole”, Krochmal 1420 (1948) (NHA). Hillsboro Co.: Antrim, North Branch Brook, Krochmal and Dole 1040 (1947) (NEBC, NHA); Hillsboro, Beard’s Brook, Krochmal and Dole 1015 (1947) (NHA); Hillsborough, Beard’s Brook, Philbrick 1099 (1981) (NHA). Merrimack Co.: Boscawen, Contoocook River, Philbrick 1100 (1981) (NHA); Henniker, Contoocook River, Philbrick 1150 (1981) (NHA); Hopkington, Contoocook River, Philbrick 1152 (1981) (NHA). Rockingham Co.: Epping, Lamprey River, Philbrick 1111 (1981) (NHA); Raymond, Lamprey River, Philbrick 1109 (1981) (NHA). Strafford Co.: Barrington, Isinglass River, Philbrick 1119 (1981) (NHA); Dover, Witcher’s Falls, Cocheco River, Hodgdon 19546 (1972) (MASS, NHA); Dover, Witcher’s Falls, Cocheco River, Hellquist 2029 (1973) (wva); Durham, Packer’s Falls, Lamprey River, Philbrick 1102 (1981) (NHA). Sullivan Co.: Plainfield, Con- necticut River, Zika 4433 (1981) (v7). NEW JERSEY. Cape May Co.: Bennett, Stone 15713 (1917) (PH); Stockton, Fisher, s.n. (1895) (pH). Hunterdon Co.: 2 mi. above Wall- pack Bend, Delaware River, Edwards, s.n. (1947) (NY); Princeton, Pretty Brook, unknown collector, s.n. (unknown date) (NY); near Phillipsburg, Delaware River, Porter, s.n. (1870) (us). Sussex Co.: opposite Bushkill, Delaware River, Bartram, s.n. (1917) (PH). Warren Co.: Foul River, Van Pelt, s.n. (1907) (PH); Marble Hill, Delaware River, Porter 1379 (1869) (PH). 1983] Philbrick & Crow — Podostemum 335 NEW YorK. Franklin Co.: Hogansburg, St. Regis River, Muen- scher, Bassett and Maguire 1197 (1930) (F, MIN, MO, US, wIs). St. Lawrence Co.: rapids below Messena, Grasse River, Muenscher, Bassett and Maguire 1198 (1930) (F, GH, HNH, MIN, MO, US, wIs); Hogansburg, St. Regis River, Muenscher and Justice 839 (1938) (CONN, F, FLAS, MAINE, MASS, MICH, MIN, MO, NCSC, NCU (2), NHA, NY, OKLA, PH (2), US, VT). Ulster Co.: Glenerie Falls, Wilson, s.n. (1916) (NY (3)); Shawangunk Kill, Wilson, s.n. (1922) (Ny). NORTH CAROLINA. Alexander Co.: 7 mi. SW of Taylorsville, Mid- dle Little River, Radford 18147 (1956) (NCU); Millersville, Glade Creek, Beal 5870 (1960) (Ncsc). Alleghany Co.: 1.6 mi. N of Amelia, New River, Radford 44194 (1961) (MASS, NCU, UARK, UNA, wis). Ash Co.: 2 mi. SE of Scottsville, Cranberry Creek, Beal 6064 (1961) (Ncsc). Avery Co.: 3.4 mi. E of Grandmother Gap, Wilson Creek, Ahles and Duke 49578 (1958) (NCU). Buncombe Co.: at Batcave, near Rt. 9, Flat Crick Falls, Radford 4867 (1949) (NCU). Burke Co.: High Shoals, Smith 305 (1975) (Ncsc). Caldwell Co.: stream at Col- letsville Radford and Radford 6251 (1952) (GH, NCU); N of Patter- son, Buffalo Creek, Beal 7126 (1961) (NCSC). Chatham Co.: conflu- ence of Deep and Rocky Rivers, Beard 1302 (1956) (NCU); near U.S. 15-501, Rocky River, Radford 44620 (1963) (NcU). Durham Co.: 1.7 mi. N of Weaver, Eno River, Ahles and Bozeman 57977 (1963) (NCU); N of Bragtown, Eno River, Radford 7371 (1953) (GA, GH, NCU, NHA, NLU, NCSC (2), US); 5 mi. N of Durham, near Christian’s Mill, Blomquist, s.n. (1953) (DUKE); Chapel Hill, Ballings Creek, Coker, s.n. (1929) (NCU). Edgecombe Co.: NE of Rocky Mount, Tar River, Fox and Whitford 17403 (1948) (GH, NCSC). Franklin Co.: 2 mi. E of Bunn, Tar River, Beal 5685 (1960) (Ncsc). Graham Co.: 0.7 mi SE of Tapoca, Cheoah River, Radford 11909 (1956) (NCU). Granville Co.; 2 mi. SE of Berea, Tar River, Radford and Ahles 11540 (1956) (Mass, NCU); 3.7 mi. N of Wilton on Dickerson Rd., Tar River, Ahles and Radford 11514 (1956) (GA, NY, vpB). Harnett Co.: SW of Kipling, Hector Creek, Beal 4323 (1958) (Ncsc). Hay- wood Co.: Big Calocoochee River, Bold, s.n. (1939) (NCU); near junction of Pigeon River, Fines Creek, Ahles and Duke 50409 (1958) (NCU); S of Watertown, Pisgah National Forest, Thomas 31762 (1972) (NLU, SMU (2), TENN); Rt. 40, Fines Creek, Boufford and Wood 16469 (1975) (Ncu). Johnston Co.: E of Clayton, Neuse River, Whitford, s.n. (1959) (Ncsc). Lee Co.: N of Comnock, Deep 336 Rhodora [Vol. 85 River, Whitford, s.n. (1958) (Ncsc). Macon Co.: 3 mi. N of Aquore Dam, Nantahala River, Beal 7202 (1961) (Ncsc); SE of Nantahala, Nantahala River, Radford 5252 (1950) (NCU); | mi. S of Gneiss, below falls, Crow Creek, Clausen and Kezer 5620 (1941) (Ny). Mad- ison Co.: 1.3 mi. W of Whiterockon, NC 212, Ahles and Duke 50671 (1958) (FLAS, NCU); 1.0 mi. S of Hot Springs, Spring Creek, Radford and Radford 7409 (1953) (NCU). Mitchell Co.: NW of Toe- cane, Master’s Mill, Beal 2159 (1961) (Ncsc). Nash Co.: | mi. N of Strickland Cross Road, Tar River, Beal 4159 (1958) (Ncsc); 3.1 mi. NE of Red Oak, Swift Creek, Beal 4156 (1958) (Ncsc). Orange Co.: Rt. 70, Eno River, Ahles and Haesloop 53240 (1960) (LAF, MAINE, MIN, NCU, TENN, VPI, WVA); Pleasant Green Road, Eno River, Bouf- ford 15073 (1974) (cM, Mo, NHA); Duke Forest, New Hope Creek, Wilbur 17941 (1974) (DUKE); Carrboro, Morgan’s Creek, Hueske, s.n. (1947) (NCU). Person Co.: fast flowing stream, Hall 15032 (1940) (NY, SMU). Polk Co.: N of junction with Green River, Pulliam Creek, DuMond 1771 (1973) (Ncsc). Richmond Co.: Rt. 73, below bridge, Beal 4250 (1958) (Ncsc); Mountain Creek, Whitford and Schumacker, s.n. (1958) (NCSC). Rockingham Co.: | mi. E of old NC 14, Beal 7090 (1961) (Ncsc). Rutherford Co.: Chimney Rock, Hick- ory Nut Creek, unknown collector, s.n. (1958) (Ncsc). Stanly Co.: NE of Oakboro, Big Bear Creek, Ahles and Leisner 16276 (1956) (FLAS, NCU). Swain Co.: | mi. N of Lauada, Alaska Creek, Beal 7205 (1961) (DUKE, Ncsc). Transylvania Co.: Lower Bearcamp Creek, Rogers and Shake 62129 B (1962) (DUKE); Bearwallow Creek, Pyron and McVaugh 791 (1936) (GA); N of Broad, Davidson River, Beal 5796 (1960) (NCsc); Toxaway Creek, Hardin 2039 (1960) (Ncsc). Wake Co.: near Rolesville, granite outcrops, Deloach and Dukes 36 (1961) (Ncsc); NW of Zebulon, Mitchell’s Mill, Blomquist, Channell and Ebert 16754 (1955) (DUKE); at falls, Neuse River, Massey and Leonard 3260 (1972) (Mo, NCU (2), NHA); NE of Rolesville, Little River, Radford and King 45057 (1966) (APSU, FLAS (2), GA, GH, LAF, MAINE, MICH, MIN, NCU, NY, OKLA, SMU, TENN, US, VDB, VPI, WIS (2), WVvA). Watauga Co.: 2 mi. E of Boone, south fork New River, Beal 5958 (1962) (Ncsc); S of Boone, Watauga River, Whitford and Schumacker, s.n. (1959) (Ncsc); 2.4 mi. SW of Todd, New River, Ahles and Duke 47719 (1958) (NCU). Wilkes Co.: Moravian Falls, Moravian River, Beal 5966 (1962) (Ncsc); 1.5 mi. W of Maple 1983] Philbrick & Crow — Podostemum 337 Springs, Beal 5962 (1961) (Ncsc). Yancey Co.: 2 mi. E of Bald Creek, Cane River, Beal 2/39 (1961) (Ncsc); 2.3 mi. W of Ramsay- ton, Ahles and Duke 50784 (1958) (NCU). OKLAHOMA. McCurtain Co.: 9.0 mi. SE of Bethel, Mountain Fork River, near Ward Creek, Hel/quist 14977 (1981) (NHA); Game Pre- serve, Mountain Fork River, Love 2/2 (1966) (OKLA); 6 mi. S of Smithville, Mountain Fork River, Correll and Mitchell 34387 (1967) (GH, LAF, MICH, NCU, NY, OKLA, US); S of Battiest, Glover River, Taylor and Taylor 9732 (1972) (MIN, OKLA, SMU), 31 mi. E of Antlers, Little River, Waterfall 17217 (1966) (GH, MICH, OKLA (2)). PENNSYLVANIA. Allegheny Co.: Buttermilk Falls, Britton, s.n. (1898) (cM). Bedford Co.: | 1/8 mi. S of Hyndman, rocky stream, Berkheimer 21735A (1962) (CM, PH); | 7/8 mi. SE of Fishertown, rocky stream, Berkheimer 12243 (1951) (CM, PH). Berks Co.: 0.25 mi. SW of Gibraltar, Allegheny Creek, Brumbach 670-33 (1933) (FLAS, GH, PH). Bucks Co.: Norrisville, Burk, s.n. (1872) (PH); Mon- roe, Delaware River, Rutt, s.n. (1885) (PH); 2 mi. NW of Point Pleasant, Tinicum Creek, Proctor 267] (1947) (pH). Chester Co.: Brinton’s Bridge, Brandywine Creek, Pennell 5109 (1913) (PH (2)); near Downington, Brandywine Creek, Burk, s.n. (unknown date) (PH); 1.75 mi. SW of Slonaker, French Creek, Wherry, s.n. (1951) (PH (2)). Cumberland Co.: 0.5 mi. SE of Williams Grove, Yellow Britches Creek, Wahl, Potter and Browning 21566 (1965) (LAF (2), PH); Brandtsville, Yellow Britches Creek, Philbrick 1166 (1982) (NHA). Delaware Co.: Chester Heights, West Branch of Chester Creek, Pennell and Adams 591 (1926) (GH, MO, PH (2)); near Bridge- water Station, Chester Creek, unknown collector, s.n. (unknown date) (PH). Fulton Co.: SW of New Grenada, Sideling Hill Creek, Wherry, s.n. (1941) (pH). Lehigh Co.: Bethlehem, LeHigh River, Wolle, s.n. (1838) (CM). Mercer Co.: 1.5 mi. W of Mercer, Shenango River, Shelar, s.n. (1898) (PH). Monroe Co.: near Stroudsburgh, Buttermilk Falls, Bicknell, s.n. (1898) (PH); | mi. W of Shawnee on Delaware, stream joining with Marshall's Creek, Glowenke, s.n. (1936) (PH); W of Dupue Island, Shawnee on Delaware, Delaware River, Glowenke 4219 (1940) (pH); Water Gap, Delaware River, Bicknell, s.n. (1898) (NY, PH); Slippery rock, Tobykanna Creek, Greene, s.n. (1858) (PH); below Marshall’s Creek, in stream joining Marshall’s Creek, Glowenke 19 (1936) (NY). Montour Co.: 6 mi. SE 338 Rhodora [Vol. 85 of Danville, Roaring Creek, Adams 5897 (1940) (NCU, PH). Northampton Co.: Easton, Delaware River, Burk, s.n. (1872) (PH (2)); Martin’s Creek, Williamson, s.n. (unknown date) (NY); east of island, above Martin’s Creek, Delaware River, Van Pelt, s.n. (1906) (PH). Perry Co.: 1.25 mi. NNW of Newport, Buffalo Creek, Myrtle and Adams 9676 (1970) (PH); 1.0 mi. WNW of Drumgold, Sherman Creek, Myrtle and Adams 41-2 (1944) (NCU, PH). Philadelphia Co.: Byberry, Martindale, s.n. (1865) (PH); Rhawn St. Falls, Pennypack Creek, Hand, s.n. (1939) (PH). Pike Co.: 1.5 mi. S of Dingman’s Ferry, Delaware River, Wahl, O'Neil and Hazzard 19355 (1959) (NCU); High Falls, Dingman’s Creek, Brown and Saunders, s.n. (1899) (pH); 0.75 mi. NNW of Parker’s Glenn, Delaware River, Wahl, O'Neil and Hazzard 19319 (1959) (PH); Sawkill Falls, Nash, s.n. (1909) (NY); near Bushkill, Winona Falls, Fassett and Calvert 19488 (1938) (GH, MO, MSU, wis). Snyder Co.: N of Beavertown, Middle Creek, Hotchkiss 7878 (1963) (us). Tioga Co.: 1.0 mi. WNW of Tiadaghton, Pine Creek, Wah/ 15844 (1955) (GH, MAINE, NCU, PH); 1.25 mi. S of Ansonia, Pine Creek, Wahi 17744 (1956) (LAF, PH). Union Co.: 5 mi. W of White Deer, White Deer Creek, Roberts 3167 (1972) (os, PH). Venango Co.: 2.5 mi. NW of Utica, French Creek, Wahi 18085 (1956) (PH). Warren Co.: Warren, Allegheny Creek, Witz, s.n. (1930) (cM). Wayne Co.: 0.25 mi. below Hancock Bridge, Buckingham Turnpike, west branch of Delaware River, Dix 560 (1942) (PH); 2.5 mi. SE of Balls Eddy, west branch Delaware River, Dix, s.n. (1946) (PH); 1 mi. NE of Equinunk, Delaware River, Wahl, O'Neil and Regan 19276 (1959) (pH); 3 mi. NNE of Damas- cus, Delaware River, Wahl, O'Neil and Regan 19307 (1959) (PH). York Co.: 0.75 mi. SSE of Long Level, Fishing Creek, Moul 10108 (1949) (cM, PH). RHODE ISLAND. Providence Co.: Georgiaville [Smithfield], Woo- nasquatucket River, unknown collector, s.n. (1894) (NEBC); East Providence, Hunt’s Mill, Ten Mile River, Collins, s.n. (1893) (F, CONN (3), MO, NHA, PH, US). SOUTH CAROLINA. Anderson Co.: Big Generostee Creek, Weems Creek, Thomas 501A (1975) (CLEMS); W of Anderson, Tugaloo River, Radford 18001 (1956) (NcU). Lancaster Co.: ENE of Kershaw, S.C. Rt. 265, Lynches River, Ahles and Haesloop 31313 (1957) (cM, MICH, OKLA, SMU). Laurens Co.: state highway 101, Durbin Creek, Freeman 56921 (1956) (NcU). Oconee Co.: S.C. Rt. 59, Beaverdam 1983] Philbrick & Crow — Podostemum 339 Creek, Kral, s.n. (1969) (CLEMS); above Westminster, Brasstown Creek, Taylor 31 (1978) (CLEMS); 3 mi. from S.C. Rt. 11 and 95, behind Station Falls, Ferguson 51 (1975) (cLEMS); NNE of conflu- ence of Brasstown and Little Brasstown Creeks, Boufford, Massey and Katz 16990 (1975) (NCU); 0.8 mi. S of North Carolina state line, Thompson River, Ware and White 3315 ( 1970) (vpB). Richland Co.: 1-126, at Candi Lane, Saluda River, Crow 3274 (1980) (NHA). Union Co.: Rt. 49, Fairforest Creek, Freeman 56631 (1956) (NCU). York Co.: E of Clover, Dan Creek, Ahles 34652 (1957) (NCU). TENNESSEE. Blount Co.: below Cove Creek, Abrams Creek, Hubbs 19721 (1937) (MICH); 0.1 mi. below Ellejoy Creek, Little River, Den- nis, Nakosteen, Farmer, Comiskey, McNeilus and Patrick, s.n. (1979) (TENN); 1.5 mi. N of Rocky Branch, Little River, Webb, s.n. (1975) (NCSC, TENN). Bradley-McMinn Co.: near Patty Branch, Hiawassee River, Dennis and Wofford, s.n. (1978) (TENN, TVA). Cocke Co.: Wolf Creek, Sohmer 32477 (1963) (TENN). Coffee Co.: near Tullahoma, Rutledge Falls, Sharp and Clebrch 4724 (1947) (TENN). Davidson Co.: Couchville Pike, Stones River, Edwards, s.n. (1954) (vps). Greene Co.: Paint Rock, Redfield 7669 (1876) (Mo). Grundy Co.: below Altamount, Falls of Collins River, Svenson 9398 (1938) (GH, PH, SMU, TENN). Johnson Co.: Shady Valley, Bea- verdam Creek, Webb and Webb 47150 (1976) (TENN, VDB). Knox Co.: E of Knoxville, French Broad River, Wofford, Holton and Dennis 79-162 (1979) (TENN). Lauderdale Co.: Cypress Creek, Hall, s.n. (1964) (TVA, UNA). Lewis Co.: 7 mi. SE of Hokenwald, Webb and Price 3091 (1980) (TENN, TVA). Lincoln Co.: 5 mi. S of Mul- berry, Elk River, Webb and Price 3167 (1980) (TENN). Loudon Co.: between Tolivar and Davis Islands, Little Tennessee River, Wof- ford, Dennis and Bates 78-128 (TENN, VDB). Maury Co.: W of con- fluence with Fountain Creek, Duck River, Webb and Dennis 1696 (1978) (TENN, TVA, VDB). Monroe Co.: 6 mi. up stream of Ballplay Park (Sloan Bridge), Tellico River, Webb and Price 3537 (1980) (TENN, TVA); 0.5 mi. past Doublecamp Creek, campground, Citico Creek, Malter, s.n. (1977) (TENN); Rose Island, Little Tennessee River, Bates, s.n. (1976) (TVA). Morgan Co.: Montgomery Bridge, Emory River, Robinson, s.n. (1935) (TENN); NW of Hatfield Moun- tain, Obed River, Patrick and Schmalzer 3125 (1981) (TENN). Perry Co.: Slink Shoals Ford, Buffalo River, Webb 434] (1981) (TENN). 340 Rhodora [Vol. 85 Polk Co.: below Farmer, Rt. 68, Hiawassee River, Rogers and Bowers 43923 (1969) (TENN, vDB). Sevier Co.: West Fork of Little Pigeon River, Hubbs 19720 (1937) (MICH); Sevierville, Pigeon River, Svenson 4007 (1930) (F, PH); W of Gatlinburg, Rt. 73, Little River, Steyermark 65842 (1948) (F). Swain Co.: Greak Smokey Mountains National Park, Ravensford, Barksdale and Tenneson 3109 (1936) (TENN). Van Buren Co.: Rockhouse Falls, Cane Creek, Norris and Sharp 1022 (1951) (DUKE, FLAS, GA, MO, NCSC, NCU, NY, SMU, TENN, UNA, US, WVA). Wayne Co.: 10 mi. N of Waynesboro, Rt. 13, Buffalo River, Webb and Price 1320 (1978) (TENN). VERMONT. Chittenden Co.: South Burlington, Winooski River, Cook and Barrington, s.n. (1977) (vt). Windham Co.: Jamaica, West River, Dobbins, s.n. (1907) (GH). Windsor Co.: Woodstock, “Winslow”, s.n. 1910 (us). VIRGINIA. Bedford Co.: unknown locality, Curtis, s.n. (1872) (F, GH, MASS, MIN, OS, US). Fairfax Co.: 2 mi. SW of Garfield, Accotink Creek, Fosberg 18655 (1942) (No, PH). Giles Co.: Pembroke, at Castle Rock, New River, Wilson, s.n. (1972) (vP1); near Berton, New River, Musselman and Musselman 2885 (1969) (SMU, wis); 1 mi. S of Eggelston, New River, Marx 2640 (1974) (NLU). Loudoun Co.: N of Neersville, Rt. 340, Potomac River, Ahles and James 61739 (1965) (NcU). Montgomery Co.: below Radford Arsenal, New River, Whitehurst, s.n. (1972) (vp1). Prince William Co.: below Bev- erley Mill, Broad River, A//ard 8/52 (1940) (GH, vP!). Stafford Co.: Rt. | near Fredericksburg, Rappahannock River, Webb and Webb 51843 (1976) (TENN, VDB). WEST VIRGINIA. Cabell Co.: 3 mi. above Salt lick, Guyandott River, Berkeley 431 (1930) (Mo). Calhoun Co.: below Annamoriah Ferry, Bartholomew, s.n. (1944) (FLAS, LSU, NCU, UNA). Fayette Co.: 0.25 mi. above mouth of Dowdy Creek, New River, Grafton, s.n. (1976) (MIN); near Wardensville, Cacopon River, Berkeley 1574 (1930) (Mo). Greenbrier Co.: Alderson, Greenbrier River, White 1014 (1976) (wva). Hampshire Co.: New Creek, Smith, s.n. (1880) (Us). Kanawha Co.: Upper Falls, Coal River, Strausbaugh, s.n. (1931) (wva). Lincoln Co.: near Miller School, unknown collector, s.n. (1829) (wva (2)). Pocahontas Co.: Seebert, Greenbrier River, White 1045 (1976) (wva). Raleigh Co.: Sandstone Falls, New River Gorge, Grafton and McGraw, s.n. (1974) (wva). Wirt Co.: shoals below Twelve Pole Creek Bridge, P/ymale 620 (1938) (GH, wva); 3 1983] Philbrick & Crow — Podostemum 341 mi. above Creston, at mouth of Two Run, Kanawha River, Bar- tholomew W1226 (1942) (wva (2)). Wyoming Co.: below Baileys- ville, Guyandott River, Morris 1210 (1900) (F, TENN, US). DOMINICAN REPUBLIC. Prov. Santiago: Distr. San José de las Matas, Rio Magua, Valeur 130 (1931) (F, MO). HONDURAS. Dept. of Comayagua: vicinity of Siguatepeque, Srand- ley 56071 (1928) (F, US (2)); Siguatepeque, Rodriguez 2671 (1945) (F). TRIFOLIUM STOLONIFERUM, RUNNING BUFFALO CLOVER: DESCRIPTION, DISTRIBUTION, AND CURRENT STATUS RALPH E. BROOKS Running buffalo clover, Trifolium stoloniferum Muhl. ex A. Eaton (Fabaceae) is one of three species of this temperate genus indigenous to the eastern United States. Since this little known and perhaps extirpated species was described in 1818, its distribution has been variously and often erroneously reported and its taxonomic status questioned. While the species was collected with some regu- larity during the 1800’s from eastern Kansas to West Virginia, T. stoloniferum has been collected a mere five times since 1900, most recently in 1940 (Webster County, West Virginia). The current interest in rare and endangered species, coupled with the varied literature accounts, led to the current investigation of the character- istics and current status of 7. stoloniferum. METHODS Despite consulting field botanists in the region where Trifolium stoloniferum was known to occur and making field excursions to past collection sites, my efforts to locate extant populations of the plant have proven futile. Thus this investigation is based entirely on herbarium specimens and available literature. Herbaria at more than 50 institutions, as well as several private collections, were consulted during the course of this study. The herbaria consulted are listed according to the acronyms of Holmgren et al. (1981) (an asterisk (*) indicates no specimens found; a double asterisk (**) indicates no response to the loan request): BGSU*, BH**, BHO*, BUT**, CINC, CLM, CM, DS, F, GH, IA, ILL, IND*™, IS, ISC, ISM*, Isu**, JHWU, KANU, KE, KNK*, KSC*, KSP*, KSTC*, KY, LYN*, MCA*, MO, MU, MUR*, MUS*, MWI*, NEB, NDG*, NY, NYS, OC, OKL*, OKLA*, OS, PH, PHIL, PUR, SDSU*, SDU*, SIU*, SMS*, SMU*, SWMT*, TENN™*, UARK, UCHT*, UMO, US, VDB*, VIP*, VT, WAB**, WVA, YUO, Baldwin- Wallace College, Berea, Ohio*, and Urbana College, Urbana, Ohio*. 343 344 Rhodora [Vol. 85 TAXONOMIC HISTORY Henry Muhlenberg originally named Trifolium stoloniferum in 1813, in his Catalogus Plantarum Americae Septentrionalis. Muh- lenberg’s new name was, however, invalid since the name was pub- lished without a description (nomen nudum). Three years after Muhlenberg’s death Amos Eaton published Manual of Botany for the Northern and Middle States (ed. 2) and therein validated Muhlenberg’s name by publishing a description of the plant. While Eaton (1818) alluded to a type specimen, it was neither clearly designated nor has original material ever been located. In view of this a neotype was recently selected (Brooks, in press). DESCRIPTION The majority of botanists describing Trifolium stoloniferum have had few specimens from which to work. Eaton (1818) in preparing the original description and McDermott (1910) in his work with North American Trifolium each had but a single specimen from which to write his description. As a result past descriptions are generally brief and less than informative. The description which follows was prepared after examining all the specimens discovered during this investigation: Trifolium stoloniferum Muhl. ex A. Eaton (Fig. 1) Perennial (probably short-lived), stoloniferous herbs with short subterranean or superficial branching caudex; sometimes with | (2) scapose flowering stems from the base; stolons 0.5-3 (6) dm long, rooting at the nodes, glabrous to rarely sparsely puberulent. Leaves basal or alternate on stolons, trifoliate; leaflets obovate to orbicular or obcordate, (8) 20-30 (40) mm long, nearly as wide, glabrous to rarely sparsely pubescent, apex rounded to usually subtruncate and shallowly or evidently retuse, margins serrulate-denticulate: petioles (1) 3-8 (15) cm long, glabrous or puberulent; stipules usually adnate 1/2 or more of length to petiole, lanceolate to ovate-lanceolate, the free portion narrowed to the apex or somewhat abruptly acuminate or cuspidate, glabrous. Flowering stems scapiform, erect, ebrac- teate, 1-2 from base or solifary from the axils of the stolons, 6-15 (22) cm tall, usually lightly pubescent near the summit, with 2 leaves near the apex; peduncles | (2) from the apex of the stem, 2-6 cm long, glabrous to evidently pubescent: inflorescence subglobose, 1983] Brooks — Running Buffalo Clover 345 ANNOTATION LABEL Der Rolph E. Brooks STEEL EA ALI y ne i : North Scabby North Double Headshot North Scabby South Scabby 4-Farliest published date of discovery, with investigator first listed by Gawler (G.), 1981 or Olday (O.), 1982. b. Population first discovered in 1980. Extinct on island by 1982. 388 Rhodora [Vol. 85 LITERATURE CITED Drury, W.H. 1980. “Rare Species of Plants”. Rhodora 82: 3-48. Gaw_er,S.,ed. 1981. “Rare Vascular Plants in Maine”, A critical areas program report. Maine State Planning Office. 656 pp. Otpay, F.C. 1982. “Occurrence and Distribution of Four Species of Rare Vascu- lar Plants Along the Coast of Eastern Maine”. Maine State Planning Office, Report Number 77 (Draft copy). 49 pp. Pike,R.B. 1963. “Note on Primula laurentiana in Maine.” Rhodora 65: 286-288. UNIVERSITY OF MAINE AT MACHIAS O'BRIEN AVENUE MACHIAS, MAINE 04654 NOTE ON ADIANTUM PEDATUM L. SSP. CALDERI CODY SUSAN C. GAWLER Adiantum pedatum L. ssp. calderi Cody, as described in the Jan- uary 1983 issue of Rhodora (Cody, 1983) occurs at one Maine sta- tion as well as at the stations reported in Quebec, Newfoundland, Vermont, California, and Washington. The Maine station was dis- covered in 1977 by Jonathan Carter, researching plants of serpen- tine outcrops in Maine (Carter, 1979). Following Carter’s directions, H.R. Tyler and | visited the site in September 1978. We found a small population of Adiantum pedatum var. aleuti- cum, as it was called by Carter, growing on a serpentine outcrop on the northeast slope of Whitecap Mountain, Somerset County, Maine. Most of the mountain at this elevation (approximately 3100’) is covered with dense spruce krummholz; the ferns grew in an isolated pocket which, presumably because of the serpentine out- crop, was basically unvegetated. The noticeably glaucous fronds were short and stiffly erect, with contracted pinnules. In stature, morphology, and ecology, these plants correspond to Cody’s Adian- tum pedatum ssp. calderi rather than to the variety to which they were originally referred. Adiantum pedatum var. aleuticum (or, more correctly, ssp. cal- deri) has been listed as “threatened or endangered” in New England because it is known from only two sites in the region (Crow, et al., 1981). The one Maine site for this subspecies has been placed on the Main Register of Critical Areas (Critical Areas Program, 1981). The original specimen from the Maine station (Carter #513, July 1977) was deposited at the herbarium of the University of New Hampshire (NHA). A second specimen (Gawler & Tyler, s.n., 1 Sep- tember 1978) has been deposited at the herbarium of the University of Maine at Orono (MAINE). Cody’s (1983) division of Adiantum pedatum into the subspecies pedatum, calderi, and aleuticum leaves ambiguous the position of the variety subpumilum W. H. Wagner, described in 1978. This extremely dwarfed variety, though known from horticulture, was only recently found in nature, on coastal cliffs of the Brooks Penin- sula of Vancouver Island (Wagner & Boydston, 1978). As Wagner and Boydston point out, further study is needed to understand tho- roughly the Adiantum pedatum complex. 389 390 Rhodora [Vol. 85 LITERATURE CITED Carter, J.K. 1979. A floristic and phytogeographical analysis of selected serpen- tine sites in Maine. Master’s Thesis, Department of Botany, University of New Hampshire, Durham, NH. Copy, W. J. 1983. Adiantum pedatum ssp. calderi, a new subspecies in north- eastern North America. Rhodora 85: 93-96. CRITICAL AREAS PROGRAM. 1981. Rare Vascular Plants of Maine. State Planning Office, Augusta, ME. 656 pp. Crow, G. E., W. D. CountTRYMAN, G. L. Cuurcn, L. M. EASTMAN, C. B. HELLQUIST, L. J. Menruorr, & 1.M.Strorks. 1981. Rare and endangered vascular plant species in New England. Rhodora 83: 259-299. WaGner, W. H., Jr. & K. E. Boypston. 1978. A dwarf coastal variety of Maidenhair fern, Adiantum pedatum. Can. J. Bot. 56: 1726-1729. DEPARTMENT OF BOTANY BIRGE HALL 430 LINCOLN DRIVE MADISON, WISCONSIN 53706 BOOK REVIEW: AN ILLUSTRATED MOSS FLORA OF THE MARITIMES BRENT D. MISHLER Robert Ireland, of the National Museum of Natural Sciences in Ottawa, has produced a most excellent flora of the mosses of the Canadian Maritimes (New Brunswick, Nova Scotia, and Prince Edward Island). This work, the culmination of comprehensive her- barium studies and more than 15 years of field experience, has been carefully prepared with an eye toward maximal usefulness. This flora should certainly serve as a stimulus for increased study of the mosses of the Maritimes; it can equally serve as such for New Eng- land. Due to the geographic proximity and similar habitats in the two regions, almost all New England mosses can be identified with Ireland’s flora, which perhaps surprisingly is the best single work yet available for the identification of New England mosses. Bryophytes are commonly excluded these days from floristic and ecological studies of the “higher” plants. This unhappy circumstance is not merely a historical artifact; consider for example the treat- ments of bryophytes by Sullivant included in the early editions of Gray’s Manual. Rather, knowledge of bryophytes of northeastern North America (especially the effective communication of knowl- edge to general botanists) has not kept pace with increasing knowl- edge of the tracheophytes. Thus the bryophytes have acquired the reputation of being difficult to identify. There really is no other good reason to separate and exclude the bryophytes from general studies of the land flora. Recent morphological and cytological work (some of which is reviewed by Scheirer, 1980, and Paolillo, 1981) has confirmed that the mosses in particular are close phyloge- netic kin to the tracheophytes. Furthermore, bryophytes are subject to factors of distribution similar to those of the tracheophytes (Crum, 1972; Schuster, 1977). As Ireland states in his introduction (p. 9): “The distribution patterns of the Maritime mosses, like the mosses throughout North America, are nearly similar to those of the ITRELAND, ROBERT R. 1982. Mosy Flora of the Maritime Provinces. 738 pp. National Museum of Natural Sciences Publications in Botany, No. 13. National Museums of Canada, Ottawa. Distributed by the University of Chicago Press. (Price: $25.00) 391 392 Rhodora [Vol. 85 flowering plants. However, the small size of mosses allows them to survive in small niches.... Therefore bryophytes are often better indicators of an ancient climate than flowering plants.” Ireland’s new flora, in conjunction with the recently published moss flora of eastern North America by Crum and Anderson (1981), provides not only a means of identification, but more importantly, a concise and useful picture of moss species as biological and ecological entities. These floras provide a foundation which should allow and encour- age general botanists and ecologists to direct their attention to this fascinating group of plants. The Moss Flora of the Maritime Provinces includes introductory material on distribution patterns, the structure and life cycle of mosses, collecting and herbarium techniques, past collectors of Maritime mosses, and hints on identification, methods of study, and the use of keys. There is a single bracketed key to genera, with 198 leads, which is made easier to use by headings restating characters, placed before major sections. This workable generic key is especially important because Crum and Anderson’s flora does not include a generic key, making that flora less useful for non-specialists. Ireland provides descriptions of 135 genera and 381 species; 398 full-page plates, handsome and accurate line drawings grouped at the end of each genus, illustrate most of the species. The species descriptions are short, giving only diagnostic characters, and include habitat, Maritime distribution, general range, chromosome numbers (when known), and miscellaneous remarks. The habitat descriptions and remarks are particularly valuable, reflecting the extensive field studies of the author. Many bryophytes are quite specific in their environmental requirements, making knowledge of habitat an im- portant part of quick and positive identification. The species keys and descriptions work quite well, based on the specimens I ran through as a test. Ireland makes especially good use of pseudopara- phyllia as a character in several pleurocarpous genera (which is not surprising as he has made a special study of these characters: Ire- land, 1971). The flora is completed by a rather long list of excluded taxa (previously reported from the Maritime Provinces but not con- firmed by Ireland), a complete glossary (mostly taken from Crum, 1976, another compact flora that can be of use in New England), and an especially helpful 17 plates illustrating terms from the glossary. 1983] Mishler — Book review 393 Even though New England has been rather well collected, the lack of a modern, focused synthesis of its moss flora makes it impossible to state exactly how many New England species are also found in the Maritimes. It is clear however, that the great majority of species in our flora will satisfactorily key out in the Moss Flora of the Maritime Provinces. The exceptions are a few species of more southerly and westerly distribution in North America that reach the limits of their range in New England, for example Sphagnum henry- ense, Fissidens obtusifolius, Hyvophyla involuta, Desmatodon por- teri, and Bryoandersonia illecebra. Such species are usually uncommon locally, so are unlikely to confuse a Yankee user of Ireland’s flora, but it will be necessary to consult the larger flora of Crum and Anderson (1981) on occasion. The book is printed on good paper, sturdily bound in a sewn, cloth binding. I saw no typographical errors. There is a problem in the reproduction of some plates, really my only criticism of the book. Somewhere in the production process, perhaps in the photo- graphic reduction of the plates, fine lines were lost in some plates, particularly in the habit drawings. Hopefully this problem can be corrected in later printings. In any case, these are only minor imper- fections in what, considering the low price, must be one of the best botanical bargains in years. I strongly recommend this flora for bryologists, general botanists and naturalists, ecologists, and libraries. If you could only own one book on the Maritime and New England mosses, this is the one. Dr. Ireland is to be congratulated upon bringing such an important and useful flora to completion. LITERATURE CITED Crum,H. 1972. The geographic origins of the mosses of North America’s eastern deciduous forest. Journ. Hattori Bot. Lab. 35: 269-298. 1976. Mosses of the Great Lakes forest. Rev. ed. University Herbarium, University of Michigan, Ann Arbor, Michigan. ..& L.E. ANDERSON. 1981. Mosses of Eastern North America. 2 volumes. Columbia University Press, New York. IRELAND, R.R. 1971. Moss pseudoparaphyllia. The Bryologist 74: 312-330. PAoLILLo, D. J., JR. 1981. The swimming sperms of land plants. BioScience 31: 367-373. SCHEIRER, D.C. 1980. Differentiation of bryophyte conducting tissues: structure and histochemistry. Bull. Torrey Bot. Club 107: 298-307. 394 Rhodora [Vol. 85 SCHUSTER, R. M. 1977. Boreal Hepaticae: a manual of the liverworts of Minne- sota and Adjacent Regions. J. Cramer, Vaduz. [reprint of The American Mid- land Naturalist 49 (1953): 257-684; §7 (1957): 203-256, 257-299; and 59 (1958): 257-332.] FARLOW HERBARIUM HARVARD UNIVERSITY CAMBRIDGE, MASSACHUSETTS 02138 LITERATURE FOR NEW ENGLAND BOTANISTS RUSSELL, EMILY W. B. 1983. Indian-set fires in the forests of the northeastern United States. Ecology 64(1): 78-88. Emily Russell presents the most thorough analysis to date of his- torical literature relating to Indian-set fires in the northeast. Her conclusion is that Indians did set fires locally, usually near their dwellings, but there is no evidence to support statements that Indians regularly burned forests over wide areas. WHITNEY, GORDON G. & ROBERT E. MOELLER. 1982. An analysis of the vegetation of Mt. Cardigan, New Hampshire: a rocky, sub- alpine New England summit. Bull. Torrey Bot. Club 109(2): 177-188 The authors present a quantitative analysis, using reciprocal ordi- nation methods, of the vegetation at the summit of Mt. Cardigan. They recognize 2 environmental gradients—exposure and available soil moisture—as most important; and 3 major community types: dwarf evergreen shrub, deciduous shrub, and sub-alpine spruce-fir. They plot the population response patterns for 12 major species on the plot ordination. Despite historical evidence of fires, they question whether the summit of Cardigan (and similar peaks) was ever completely forested at the very top, and feel that these wind-swept areas have served as refugia for arctic-alpine species. PAILLET, FREDERICK L. 1982. The ecological significance of Ameri- can chestnut (Castanea dentata (Marsh.)Borkh.) in the Holocene forests of Connecticut. Bull. Torrey Bot. Club 109(4):457-473. A marked increase in chestnut pollen makes this an excellent indi- cator for the boundary between the C, and C, climate zones (C; = 2000 years ago to present) inferred from New England pollen profile studies. An understanding of the modern ecology of chestnut, especially its reproductive strategy, is necessary in order to interpret correctly this pollen increase in relation to environmental conditions. The author presents working theories as a basis for future research. BROWN, JAMES H. JR., CESAR A. CASTANEDA, & ROBINSON J. HIN- DLE. 1982. Floristic relationships and dynamics of hemlock 395 396 Rhodora [Vol. 85 (Tsuga canadensis) communities in Rhode Island. Bull. Torrey Bot. Club 109(3):385-391. Hemlock stands probably represent some of the oldest and least disturbed forest communities in the state. In the absence of major disturbance, change occurs very slowly, and hemlock reinforces its dominant position in the stands. SICCAMA, THOMAS G., MARGARET BLIss, & H. W. VOGELMAN. 1982. Decline of Red Spruce in the Green Mountains of Vermont. Bull. Torrey Bot. Club 109(2): 162-168. The authors note the distinct decline of red spruce over a 15-year period, and present data to document this. The cause of the decline may be due to air pollution and resultant acid rain and heavy metal accumulation. Insects and fungal diseases also play a role. MARY M. WALKER LIBRARIAN NEW ENGLAND WILD FLOWER SOCIETY FRAMINGHAM, MA 01701 BOOKS RECEIVED MOHLENBROCK, RoBERT H. 1982. The illustrated flora of Illinois. Flowering plants: basswoods to spurges. 234 pp. Carbondale, Illinois, Southern Illinois University Press. (price $22.95) This is the 10th volume in a continuing series. Forty-two genera and 103 species are described, and keys are provided. Additional notes include synonymy, ecology, variations, range. Dot maps show Illinois distribution. A detailed line drawing by Mark William Mohlenbrock is provided for each species. These volumes are use- ful, far beyond the borders of Illinois. MARY M. WALKER LIBRARIAN NEW ENGLAND WILD FLOWER SOCIETY FRAMINGHAM, MA 01701 Vol. 85, No. 842, including pages 127-273, was issued April 21, 1983. INSTRUCTIONS TO CONTRIBUTORS TO RHODORA Manuscripts should be submitted in triplicate (an original and two xerox copies) and must be double-spaced (at least 3/8 of an inch) throughout including footnotes, figure legends, and refer- ences. Please do not use corrasable bond. 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 underscoring, particularly in the citation of specimens. Names of genera and species may be underlined to indi- cate italics in discussions. Specimen citations should be selected critically, especially for common species of broad distributions. Sys- tematic revisions and similar papers should be prepared in the for- mat of “A Monograph of the Genus Malvastrum”, S.R. Hill, Rhodora 84: 1-83, 159-264, 317-409, 1982, particularly with refer- ence to indentation of keys and synonyms. 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, Rhodora 81: 503-548, 1979. For bibliographic cit- ations, 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, 3rd ed. (original title: Sty/e Manual for Biological Jour- nals). In preparing figures (maps, charts, drawings, photos, etc.) please remember that the printed plate will be 4 x 6 inches, be sure that your 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 3 1/2 x 6 in.) Magnifi- cation/reduction values given in text or figure legends should be calculated to reflect the actual printed size. Contrary to prior policy, Rhodora now requests that an abstract and a list of key words be supplied with all papers submitted except for very short articles and notes. RHODORA July 1983 Vol. 85, No. 843 CONTENTS The composition and seasonal periodicity of the marine-estuarine Chloro- phyceae in New Hampshire Arthur C. Mathieson and Edward J. Hehre Note on meeting sites Taxonomic implications of aapaias iad sis oleldy in Potbaninetiai (Potamogetonaceae) Donald H. Les Distribution of Podostemum eaaieshviaia Michx. ‘iclidenmanien C. Thomas Philbrick and Garrett E. Crow : Trifolium stoloniferum, Running Buffalo Clover: description, bisivteakion and current status Ralph E. Brooks Cytological and morphological banvidionat in Calineowe sid ates genera (Asteraceae) Judith M. Canne The type locality of Senecio perm Pura Leonard J. Uttal Range extensions of vascular plants — the oe Pesala: sorthewest Alaska Svivia Kelso Jamaican blue-green algae ithect ind of J. C. Strickland W. John Hayden NEW ENGLAND NOTES: Distribution of three rare a on islands in Machias Bay, Maine Alan J. Lewis Note on Adiantum pedatum 7. ssp. adiari Cc ply Susan C. Gawler Book Review: An illustrated moss — ot the Maritimes Brent D. Mishler Literature for New England ibalanmees Pa Bask aie Mary M. Walker 275 300 300 367 395 Roovdova JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB October 1983 No. 844 Vol. 85 Che New England Hotanical Club, Iuc. Botanical Museum, Oxford Street, Cambridge, Massachusetts 02138 Conducted and published for the Club, by NORTON H. NICKERSON, Editor-in-Chief Associate Editors A. LINN BOGLE GARRETT E. CROW WILLIAM D. COUNTRYMAN RICHARD A. FRALICK GERALD J. GASTONY NORTON G. MILLER ROBERT T. WILCE RHODORA.—Published four times a year, in January, April, July, and October. A quarterly journal of botany, devoted primarily to the flora of North America. Price $20.00 per year, net, postpaid, in funds payable at par in the United States currency at Boston. Some back volumes, and single copies are available. For information and _ prices write RHODORA at address given below. Subscriptions and orders for back issues (making all remittances payable to RHODORA) should be sent to RHODORA, Botanical Museum, Oxford Street, Cambridge, Mass. 02138. In order to receive the next number of RHODORA, changes of address must be received prior to the first day of January, April, July or October. Scientific papers and notes relating to the plants of North America and floristically related areas will be considered by the editorial committee for publication. Articles concerned with systematic botany and cytotaxon- omy in their broader implications are equally acceptable. Brevity is urged whenever possible in all papers. Short items will be published on otherwise blank end pages as soon as possible, even if they appear ahead of longer articles already accepted. All manuscripts should be submitted in TRIPLICATE AND MUST BE DOUBLE (AT LEAST 3) 8 OF AN INCH) OR TRIPLE- SPACED THROUGHOUT. Please conform to the style of recent issues of the journal. Extracted reprints, if ordered in advance, will be furnished at cost. Address manuscripts and proofs to: Russell R. Walton Managing Editor, RHODORA Harvard University Herbaria Building 22 Divinity Avenue Cambridge, Mass. 02138 Second Class Postage Paid at Boston, Mass. PRINTED BY THE LEXINGTON PRESS, INC LEXINGTON, MASSACHUSETTS Cover illustration Trollius laxus Salisb. is very rare as it nears its northeastern limit in Connecticut. Some old records indicate that it has grown in parts of New Hampshire and Maine, so it may yet be found in appropriate habitat. Original artwork by Tess Feltes, Illustrator. TRhodora (ISSN 0035-4902) JOURNAL OF THE NEW ENGLAND BOTANICAL CLUB Vol. 85 October 1983 No. 844 PLASTICITY EXPRESSED BY ROOT GROUND TISSUES OF RHIZOPHORA MANGLE L. (RED MANGROVE) GEORGE S. ELLMORE, SUSAN C. LEE, AND NORTON H. NICKERSON ABSTRACT Roots of Rhizophora undergo striking changes in structure as they grow from the open air into the substrate. To test the hypothesis that darkness and oxygen depriva- tion are responsible for determining root form in the underground environment, above-ground roots were covered with plastic to limit gas exchange, and with foil to block out light. After 9 months growth, roots growing in darkness with restricted gas flow became swollen and branched, as did underground roots. However, in artifi- cially covered roots, axis swelling was due to pith enlargement rather than elabora- tion of cortex lacunae, as occurs in underground roots. Thus, root plasticity 1s expressed via a different mechanism when roots are in darkness and deprived of O, than when they are underground. Other factors, alone or in combination, must govern development of underground Rhizophora roots. Key Words: Mangrove, Rhizophora, root anatomy The Red Mangrove, Rhizophora mangle (Rhizophoraceae) is a widely distributed tropical tree, critical to shore stability along warmer coasts, such as that of south Florida (Craighead, 1971). Its ability to stabilize coastal areas is due to the growth of conspicuous aerial roots which penetrate the substrate. Under ground, a series of striking changes occurs (Gill & Tomlinson, 1977) as roots grow from a single axis into a mass of thin, soil-building lateral roots. The underground root mass of Rhizophora becomes so dense that an anaerobic peat often develops. Gill and Tomlinson (1977) review the typological literature deal- ing with Rhizophra root structure. Their studies greatly improved our understanding of mangrove roots because they stressed the changeable nature of living roots rather than treating the topic 397 398 Rhodora [Vol. 85 solely in terms of phenomenology. Thus, a single growing root tip of Rhizophora is now known to undergo drastic changes in organiza- tion and in the sort of tissue it produces during its life span. How do field conditions induce such fundamental changes in root growth? As a root tip grows from open air into a substrate, ambient conditions important to plant growth change. Among other proper- ties (see discussion), the subterranean environment is moist, dark, and hypoxic relative to open air. The study of root dynamics in Rhizophora by Gill and Tomlinson (1977) suggests that darkness and oxygen supply govern the changes in root growth which occur underground. In this study, we test the notion that darkness and/or oxygen deprivation are responsible for the characteristic growth form of underground mangrove roots. Attempts are made to simulate underground conditions around root tips which are still above ground (aerial). Aerial tips were deprived of oxygen or light, or both, and allowed to grow 9 months before being anatomically examined. We predicted that if low O, or darkness was in fact responsible for the underground growth form, our simulated condi- tions would produce an underground growth habit in spatially aerial roots. On the other hand, 1f no change in root form occurred, or if a form developed which differed from that of underground roots, then some other environmental condition(s) must influence growth of Rhizophora roots in the substrate. MATERIALS AND METHODS Experiments were performed on a natural stand of Rhizophora (2-4 meters tall) growing in oolitic sand on the tropic of Cancer. Hummingbird (Jewfish) Cay, seven miles west of Great Exuma, Bahamas, was the study site. On trees of the study population, aerial roots grew from 1-3 mm per day from July 1981-March 1982. In an attempt to duplicate the dark, moist, subterranean condi- tions, tips of aerial roots were covered. Four sets of roots were studied (3-5 roots/set). Tips (20 cm) of one set were capped by polyethylene bags which were then filled with fresh water. This set of roots grew for 9 months with a restricted gas flow (due to the plastic covering), and with natural illumination (Fig. |). Tips of another set were likewise covered with water-filled bags but in addition, each root tip was wrapped with aluminum foil to block out sunlight. This 1983] Ellmore, et al. — Rhizophora roots 399 set grew with restricted gas flow and in darkness (Fig. |). Two other sets were untreated; one set was of aerial roots in the open air, the other was of underground roots (Fig. 1). At the end of the 9-month growth period, root tips were collected, fixed in FAA, and sectioned for study. Only portions produced by the main root axis during these 9 months were sampled. Anatomy was studied using hand sections, stained in 0.1% toluidine blue and mounted in glycerine. Measurements were made using a calibrated ocular micrometer. stele Pes 20 ¥ diameter cortex Be! = “™. root E diameter ® 10 w = Oo (a) restricted restricted open gas flow gas flow air ® @ (1) darkness _ illuminated Figure |. Cross sectional topography of RAizophora roots grown 3 months under conditions indicated. Each bar represents an average derived from 3 to 5 samples. In all cases. standard error ts less than 10° of the mean. 400 Rhodora [Vol. 85 RESULTS As previously shown by Gill and Tomlinson (1971; 1977) Rhizo- phora root structure changes dramatically as a root grows from the air into solid substrate, which at our study site is oolitic sand. Embedded in oolite, Rhizophora roots become twice as wide as aerial root axes (Fig. 1). Underground roots branch and show the diagnostic anatomy recently described (Gill & Tomlinson, 1977). Our attention is focused on the ground tissues (parenchyma of the cortex and pith). In the wide underground roots, 70% of the diame- ter is occupied by cortex, in contrast to 40% in the cortex of aerial roots (Fig. 1). Both cortex and pith are composed of vacuolate parenchyma cells (Figs. 2 and 3). Interspersed among them are cells filled with tannins and/or oil droplets. Tanniferous cells remain isodiametric, unlike the other parenchyma cells which stretch radially forming a system of intercellular spaces. In the cortex, these spaces average twice as wide (115 um) as those in the pith (57 um). The spaces are formed by radial elongation of cortex cells; cell-to- cell contact is maintained by conspicuous arm-like extensions of the cell wall (Fig. 2). Cells of the pith do not stretch as much as those of the cortex (Fig. 3), with the result that the gas spaces formed in the pith are smaller than those in the cortex. Trichosclereids occur only in the inner cortex. Of the artificial growth conditions used, the combination of res- tricted gas flow and darkness produces roots that most resemble underground roots. Like underground roots, their diameter is over twice that of untreated (control) aerial roots (Fig. 1). However, it is the stele rather than the cortex which takes up over 60% of the root diameter (Fig. 1). In this case, intercellular spaces in the pith are on the average larger (70 um) than those of the cortex (40 um). Extended arms are more developed in pith parenchyma (Fig. 5) than in the cortex (Fig. 4). Trichosclereids occur in both pith and cortex. Untreated aerial roots have a narrow axis (Fig. 1). Small (19 um) gas spaces occur in both the cortex and pith (Figs. 6 and 7). Trichos- clereids are found throughout, as detailed by Gill and Tomlinson (1971). DISCUSSION The aim of our work was to determine the environmental cues responsbile for altering root development as the root tip grows form 1983] Ellmore, et al. — Rhizophora roots 401 g-Llp Fa (a! A Saal . “