SIDA toscan Volume 11 1985 — 1986 SIDA Contributions to Botany volume 11 (in 4 numbers) Copyright 1985, 1986 by Wm. E Mahler SMU Herbarium Dallas, Texas 75275 DATES OF PUBLICATION No. 1, pp. 1— 106: 19 Jun 1985 No. 2, pp. 107 — 253: 12 Dec 1985 No. 3, pp. 255 — 355: 20 May 1986 No. 4, pp. 357-497: 16 Dec 1986 Index, pp. 491 — 497 For contents, see the unnumbered pages forming front cover of the separate issues. Ne se S | DA CONTRIBUTIONS TO BOTANY VOLUME 11 NUMBER 1 JUNE 1985 CONTENTS Forestiera godfreyi (Oleaceae), a new species from Florida and South Carolina. Loran C. Anderson. 1 Petrorhagia (Caryophyllaceae) of North America. Richard K. Rabeler. 6 Portulaca pilosa L., P. mundula J. M. Johnst. and P. parvula Gray in the southwest. James F. Matthews and Patricia A. Levins. 4S Sidus sidarum—V. The North and Central American species of Sida. Paul A. Fryxell. 62 Achyranthes japonica (Miq.) Nakai (Amaranthaceae) in Kentucky and West Virginia: new to North America. Max E. Medley, Hal Bryan, John MacGregor and John W. Thieret. 92 Drymaria viscosa (Caryophyllaceae): correct author citation and range extension to the United States. Bruce D. Parfitt and Wendy Hodgson. 96 des (Cam- NOTES. Themeda quadrivalvis (L.) Kuntze (Poaceae) in Louisiana. 99-—C ! panulaceae) new to Texas. 102—Notes on tw 02g Sep Seritea Vahl var. taccada (Gaertn.) Thieret & Lipscomb, comb. nov. a Sey 10 REVIEWS 104 SIDA Contributions to Botany volume 11 (in 4 numbers) Copyright 1985, 1986 by Wm. E Mahler SMU Herbarium Dallas, Texas 75275 DATES OF PUBLICATION No. 1, pp. 1— 106: 19 Jun 1985 No. 2, pp. 107 — 253: 12 Dec 1985 No. 3, pp. 255-355: 20 May 1986 No. 4, pp. 357—497: 16 Dec 1986 Index, pp. 491—497 For contents, see the unnumbered pages forming front cover of the separate issues. iN le = Avi o~ ee S | DA CONTRIBUTIONS TO BOTANY VOLUME 11 NUMBER 1 JUNE 1985 CONTENTS Forestiera godfreyi (Oleaceae), a new species from Florida and South Carolina. Loran C. Anderson. 1 Petrorhagia (Caryophyllaceae) of North America. Richard K. Rabeler. 6 Portulaca pilosa L., P. mundula I. M. Johnst. and P. parvula Gray in the southwest. James F. Matthews and Patricia A. Levins. 4S Sidus sidarum—V. The North and Central American species of Sida. Paul A. Fryxell, 62 Achyranthes japonica (Miq.) Nakai (Amaranthaceae) in Kentucky and West Virginia: new to North America. Max E. Medley, Hal Bryan, John MacGregor and John W. Thieret. 92 Drymaria viscosa (Caryophyllaceae): correct author citation and range extension to the United States. Bruce D. Parfitt and Wendy Hodgson. 96 NOTES. iin quadrivalvis e : Kuntze (Poaceae) in Louisiana. 99-——Campanula rapunculoides (Cam- panulaceae) new to Texas. 1 plants. 102—Scaevola sericea Vahl var. taccada (Gaertn. . Thierect & toad comb. nov. (Goodeniaceae). 103. REVIEWS 104 US ISSN 0036-1488 SIDA, CONTRIBUTIONS TO BOTANY Founded by Lloyd H. Shinners, 1962 Publisher Wm. F. Mahler SMU Herbarium Dallas, Texas, 75275 Editor Associate Editor Barney L. Lipscomb John W. Thieret SMU Herbarium Northern Kentucky University Dallas, Texas, 75275 Highland Heights, Kentucky, 41076 Guidelines for contributors are available upon request. Subscription: $10.00 (U.S.) per year; numbers issued twice a year. Sida, Contributions to Botany, Volume 11, Number 1, pages 1— 106. Copyright 1985 by Wm. F. Mahler FORESTIERA GODFREYI (OLEACEAE), A NEW SPECIES FROM FLORIDA AND SOUTH CAROLINA LORAN C. ANDERSON Department of Biological Science, Florida State University, Tallahassee, FL 32306, U.S.A ABSTRACT Forestiera godfreyz (Oleaceae) is formally described and illustrated. The new species has previously been confused with F. acuminata and F.. pubescens; comparions amongst the three are given. Several populations of pubescent Forestzera in northern Florida and one in South Carolina have been difficult to place taxomonically. Johnston (1957) included them in F. acuminata (Michx.) Poir. in Lam. forma vestita (Palmer) M. C. Johnst. Typical glabrous forms of F. acuminata occur in our area mostly on river banks, in swampy woodlands, and on pond or lake shores, whereas the pubescent plants occur on bluffs and high hammocks underlain by limestone as well as differing morphologically. In her disserta- tion, Brooks (1977) included these plants in F. pubescens Nutt., a species from limestone hills in Oklahoma, Texas, and New Mexico. She suggested floral characters of our plants were more like those of F. pubescens than like those of F. acuminata. R. K. Godfrey provided me additional collections and field observations incidental to his work on the trees and shrubs of north Florida. He and I believe these pubescent plants are more closely related to F. acuminata than to F. pubescens but that they are distinct from both those species. Therefore, I am describing a new species to accommodate them. ForestiERA godfreyi L. C. Anderson, sp. Frutices vel arbusculae deciduae dioeciae; rami oa rr et folia pubescentia; folia opposita ovata vel elliptica apice obtusa 5 — 8 cm longa 2.3 — 4 cm lata; flores in fasciculi umbellati redactissimi pedunculis et pedicellis pubescentibus dispositae; staminatae apeta- lae staminibus 2—5, 2.1—4 mm longis; pistillatae apetalae staminibus abortivis 3 et pistillis 2.4—3.5 mm longis; drupa matura 8— 12 mm longa 8—9 mm lata glauca caerulea. Deciduous, dioecious shrubs or small trees 2.5 —5 m tall, main stem arching or leaning, branches rigid or divaricate, occasionally a few bran- chlets (5 —7 cm long) developing enlarged bases to become spine-like, young twigs pubescent; leaves opposite, simple, ovate to lance-ovate or SIDA 11(1): 1-5. 1985. US ISSN 0036-1488 SIDA, CONTRIBUTIONS TO BOTANY Founded by Lloyd H. Shinners, 1962 Publisher Wm. F. Mahler SMU Herbarium Dallas, Texas, 75275 Editor Associate Editor Barney L. Lipscomb John W. Thieret SMU Herbarium Northern Kentucky University Dallas, Texas, 75275 Highland Heights, Kentucky, 41076 Guidelines for contributors are available upon request. Subscription: $10.00 (U.S.) per year; numbers issued twice a year. Sida, Contributions to Botany, Volume 11, Number 1, pages 1 — 106. Copyright 1985 by Wm. F. Mahler Fig. 1. Forestiera godfreyi: a., flowering branches, staminate on left, functionally pistillate on right; b., cluster of staminate flowers; c., cluster of functionally pistilate flowers; d., fruiting branch with immature fruits, mature fruit enlarged to left; e., branch of mature leaves (short petiolate), with enlarged portion of abaxial surface to left. West G L. E. Arnold in 1937, fl (FLAS); JACKSON Co.: Florida Caverns State Park, L. E. Arnold & E. West in 1948, veg (FLAS), R. K. Godfrey ae veg (FLAS, FSU, US, U a. R. K. Godfrey 78480a, veg (FSU), R. K. Godfrey 79320, fl (staminate: A, FSU, GA, MO, NY); JEFFERSON on E side Lake Miccosukee, R. ste gel = (FLAS, FSU-2 sheets, MO), 78375, fl ane and pistillate: A, FSU, ; , US; pistullate only: GA, NCU), ee fr (FSU); LEON Co.: Lake eee E. J. ae 384604, veg (MO); Levy Co.: 5 mi NE Inglis, L. E. Arnold s.n., veg (FLAS); Lisertry Co.: Apalachicola River, S of soa) Creek, R. K. Dipak: Gholton 79840, veg (FLAS, ae S. W. Leonard al, 7827, tr (FSU), eee State Park, P. Elfiott 267, veg (FSU), R. K. Godfrey & A. ie 648, fr (FLAS, FSU); Marion Co.: Silver R JRLW. Simoni in 1984, fr (FSU). SoutH CA NA. aren Co.: Blufton, J. H. Millichomp SM, ) Brooks ( 977) lists che following Sepa that I have not seen that ere belong to F, godfreyi: FLORIDA. HERNANDO Co.: 13. m NE arene D. Smith 368 (GH, NY, US); CouNTY UNKNOWN: Tampa ee Sin. NY). SouTH Caan INA. Beaurort Co.: several Millichamp collections from near ae Plants of F. godfreyi are less pubescent than those of F. pubescens but more so than those of F. acuminata forma vestita. They are also intermediate, yet distinctive, in several other features listed in Table 1, in which data from Brooks (1977) were used to supplement personal observations. Structure of the inflorescences in F. godfrey can perhaps best be described as a reduced umbel; it is a very highly reduced umbel or cyme in FP. pubescens and a thyrse in F. acuminata. The new species is separated from the other two fairly well phenologically; ic blooms from mid-January to mid- February, and the other two bloom in February and March. Forestiera godfreyi fruit matures in late April or early May, whereas fruits of the other two mature in May and June (occasionally in late April for F. acuminata). Mature fruits of F. godfreyz are dark blue with smooth surfaces, whereas those of F. acuminata are reddish-purple with wrinkled surfaces (Godfrey, pers. comm.). Nuttall (1837) recorded F. pubescens fruit as black. Fruits of these three taxa also have distinctive sizes and shapes; measurements of dried fruits are given in Table | because those are represented on herbarium TaBLe 1. Comparison of vegetative and female floral features in selected Forestiera taxa. 5 Leaf length, cm (6)7 — 8(9) (5)5.4— 7(8) (2.8)3. a 4.3(5) Leaf width, cm 2—2.8(3.6) - 3)2.7-4 (1.3)1.5— 1.6 Floral brace length, mm pe ee °) 3.8—-4.1 a) Mature peduncle length, mm 8— 11(14) 2. 5-5 O— 1(2) Mature pedicel length, mm 1-2 5-7 4.5 — 6(10) Flower number 5 acadl ) 9 — 23(27) (5)7 — 10 Fruit lengch, dry, mm 11-12 8-9 6- Fruit width, dry, mm 3—4(5) 4—5 3.5-4 5 specimens. Fresh fruit sizes are given by Brooks (1977) for F. acuminata and F. pubescens and here in the species description for F. godfrey7. Calyx development in F. godfreyi flowers appears to be more extensive than in the other taxa. Certainly, additional populations should be ex- amined for enlarged, petaloid sepals as found in Godfrey 79326; staminate flowers of Godfrey 78375, Hall 1270, and Murrill in 1940 lack them. ACKNOWLEDGMENTS R. K. Godfrey kindly made available illustrations prepared by M. Darst; photography was done by K. Womble, and Latin diagnosis by M. Garland. The Arnold Arboretum loaned the type specimen of F. acuminata forma vestita, and the British Museum and Royal Botanical Gardens, Kew, supplied photographs of type specimens of F. pubescens. REFERENCES Brooks, C. J. 1977. A revision of the genus Forestiera (Oleaceae). Dissertation, Universty of Alabama Library. Jounston, M. C. 1957. Synopsis of the United States species of Forestiera (Oleaceae). uthw. Naturalist 2: 140 — Nurratt, T. 1837. Collections towards a flora of the Territory of Arkansas. Trans. Amer. Philos. Soc. n.s. 5:139— PETRORHAGIA (CARYOPHYLLACEAE) OF NORTH AMERICA RICHARD K. RABELER Department of Botany & Plant Pathology and Lyman Briggs Schoo Michigan State University East Lansing, MI 48824 U.S.A ABSTRACT As an outgrowth of the Soro aml ei (L.) P. Ball & Heyw. in — a review is presented of North American ma | of Petrorhagia (Ser. in DC.) Link, a genus revised by Ball and Hees (1964). Four species, all introduced from Europe, are ascribed to North America based on field study and an extensive herbarium survey that included ices aes mavenel The history of introduction (where known), distribution, and current is presented. Petal vein color is described as an additional character that can be used in an distinguishing the three species of the P. his complex. A partial synonymy emphasizing names used in major floristic manuals and additiona | synonyms not listed by Ball and Heywood (1964) is presented to establish a transition to previous treatments INTRODUCTION Introduced plants often receive a very casual treatment in floristic literature, being nme ry such phrases as “sparingly established in waste places in our range,” or “found as a weed here and there in the n. part of our range” (Gleason & tee 1963). The current distribution of an introduced plant may be quite different than what is recorded in the literature, as shown by Shinners (1965) in his study of Holosteum umbellatum L. Identification of aliens can be problematic since recent introductions may not be represented in regional manuals; see Shinners (1969), Pringle (1976), and Rabeler (1980) for examples. An evaluation of the documenta- tion, in the literature and in herbarium collections, of the genus Petrorhagia (Ser. in DC.) Link in North America shows a similar pattern: frequent misidentifications, species with poorly documented distributions, and a complex nomenclature. The account presented here is aimed at dispelling the confusion surrounding Petrorhagia as it exists in North America. An additional problem encountered in dealing with introduced plants is the ambiguous use of terminology employed to describe their status in a given flora. Robbins (1940) defined two of the most frequently used terms, naturalized and adventive, as follows: SIDA 11(1): 6-44. 1985. 7 NATIIRATIZED: ‘1 ] : } 1 } within our borders fora long period, are rather ne distributed, multiply readily, may compete more or less favorably with native species, and behave much as in their own geographical range. ADVENTIVE: ° eigcvely: recent introductions, less widely distributed than naturalized species and not so firmly established.’ Depending on the size of the geographic unit considered, both terms could be applied correctly to three of the four species of Petrorhagia. For example, Petrorhagia prolifera (L.) P. Ball & Heyw. was first collected in North America shortly after 1800 and its current distribution suggests a natural- ized species. Yet, collections from Georgia, Tennessee, and Michigan suggest that “‘adventive”’ is a better term for these populations. Because the probability of reintroduction from cultivation or other means is relatively high at a given site, a local definition of status of introduction is desirable and will be used whenever possible. METHODS AND MATERIALS Many of the data for this study were gathered from herbarium specimens. Five hundred and thirty-one specimens representing North American collections were examined from 82 herbaria (see acknowledgments) {all symbols for herbaria cited follow Holmgren et al. (1981) with the exception of EGV (personal herbarium of Dr. Edward G. Voss)}. Morphological and distributional data presented are based on these specimens. Measurements cited are based on dried materials, using a millimeter rule and an ocular micrometer at 10-30X magnification. For comparative study of Petrorhagia from its native environs, 718 Old World specimens from 26 herbaria, including BH, F, MICH, MO, NA, ND, NY, and US were consulted. Field observations and collection of Petrorhagia was concentrated in Michi- gan (1976 — 1982), with brief visits to California (1980), Texas (1980), and Maryland (1981) added for gathering data on Petrorhagia species occurring there. Forty-seven voucher specimens documenting this work are deposited at MSC The distinctive external morphology of the seeds of these species was examined using two methods. Scanning electron micrographs were taken of the seeds of Petrorhagia prolifera (Figs. | and 2) and P. saxifraga (L.) Link. The seeds were attached to stubs with Tube-Koat adhesive, coated with about 200A of gold under a vacuum in a sputter coater, and photographed at 30X in an ISI Super Mini SEM. Scanning photomacrographs of the seeds of all four species (Figs. 3-6) were taken at 20X by Darwin Dale using the apparatus described in Dale (1982). FIGS. |—2. Scanning electron micrographs : seed surface of pi hie ae ie Michigan, Rabeler 154 (MSC). Scale = 250 wm. 1. Dorsal surface. ral surface HISTORICAL ACCOUNT The most recent revision of Petrorhagia is that of Ball and Heywood (1964), in which 25 species, 4 subspecies, and 4 varieties are recognized. Since the appearance of this work, additional taxa totaling three species' (Phitos, 1966; Greuter & Mouterde, 1970; Brullo & Furnari, 1979), two varieties (Damboldt & Phitos, 1972; Huber-Morath, 1977), and one from (Gamisans, 1974) have been recognized. In addition, Greuter and Burdet have published combinations altering the rank of two Ball and Heywood combinations, raising one variety to the species level (in Greuter & Raus, 1982) and one variety to the subspecific level (in Greuter & Raus, 1984). Most Petrorbagia species are native to the eastern Mediterranean region, with 16 of the 29 species restricted to local areas of Greece, Crete, and/or Turkey. Only three species, each of which has been introduced in North America, have natural distributions that extend northward into Europe and, thus, out of the Mediterranean climatic regime.’ One problem complicating the status of Petrorhagia is the name itself. Most of the species have at some time been placed in the genus Tunica. Ludwig published Twnica in 1757, but his usage of it as a substitute for Dianthus L. (1753) ts illegitimate. In American references, authorship of Tunica is usually attributed to Scopoli (1772) who also used the name as a substitute for Dianthus (Ball & Heywood, 1964). Mertens and Koch (183 1) redefined Tunica, distinguishing the genus from both Gypsophila L. and Dianthus on the basis of seed and petal characters. Even in this form, Tvnica Mertens & Koch remains a later homonym of Tunica Ludwig, which ts a 'Two unnecessary combinations have been made: P. kennedyae (A. K. Jackson & Turrill) Meikle in 1977 (see Rabeler, 1984) and P. vbcordata (Margot & Reuter) S. M. Thomas in 1983 (see Rabeler, in press). FIGS. 3 —6. Scanning photomacrographs of dorsal seed surface of Petrorhagia spp. Scale = 250 pm. 3. P. saxifraga, Michigan, Rabeler 262 (MSC). 4. P. prolifera, Michigan, Rabeler 154 (MSC). 5. P. nanteuilii, California, Bacigalupi, Robbins, & Hoffman 5676 (JEPS). 6. P. velutina, California, Serpa s.n. (MSC). 10 synonym of Dianthus (Ball & Heywood, 1964). Therefore, Tunica must be rejected as illegitimate under Article 64 of the International Code (Voss et al., 1983). Maire (1963) used Tunica and cited “Tunica Boehm. in Ludw. (1760), nom. abort., emend. Mert. et Koch (1831); nom. conserv.”’ This citation reflects the ‘nomen abortivum" approach Sprague (1927) used to define the early misapplications, which, before the Cambridge Congress of 1930, were not considered as sufficient grounds for rejecting a name as a later homonym (Lawrence, 1951), thus allowing an ‘‘emendatum’’, or redefinition, as noted by Maire. Maire (1963) went one step further in noting Tvnica as a nomen conservandum. Janchen (1965) argued for conserva- tion of Tunica as defined by Mertens and Koch, suggesting that Petrorhagia is a superfluous name. I have seen no evidence to indicate that any proposal to conserve Tunica has ever been formally presented. Another name suggested for plants placed in Tunica is Imperatia, pub- lished by Moench (1794) to include one species, Imperatia filiformis (=Gypsophila saxifraga). Degen (1937) noted Imperatia was the correct name for all Tvnica species except those in section Koh/rauschia. However, as Dandy (1957) noted, this name cannot be used since it is a later homonym of Imperata Cirillo, a genus in the Gramineae published in 1792 (Farr et al., 79). Petrorhagia was established as a genus of four species by Link (1831), based implicitly on Gypsophila section Petrorhagia as recognized by Seringe in 1824 (Ball & Heywood, 1964). Dandy (1957) considered this derivation to be a good reason to consider Petrorhagia as a synonym of Gypsophila, and suggests that the proper name for the genus is Koh/rauschia, a name published by Kunth (1838) for separating two species from Dianthus. Dandy’s argument and choice of Gypsophila glomerata Pallas ex M. Bieb. as lectotype of the genus were rejected by Ball and Heywood (1964). They supported Britton’s (1913) choice of P. saxifraga as lectotype of Petrorhagia, citing the bract condition present in P. saxifraga more closely fits Seringe’s description than G. g/omerata and noting that Petrorhagia is the Greek word for “‘saxifraga.”’ agree with the argument of Ball and Heywood and propose the lectotype as P. saxifraga (vide Ball & Heywood, Bull. Brit. Mus. (Nat. Hist.), Bort. 3:130. 1964). Dandy (1957) did not specifically mention Britton’s lectotypification, giving no indication he was intentionally superseding Britton. Ball and Heywood’s action also means that the Britton lectotypification cannot be superseded on the grounds that it is based solely ona largely mechanical method (see Article 8, International Code; Voss et al., 1983). Some authors, including Holub et al. (1972), still consider Kohlrauschia as a distinct genus of five species. Ball and Heywood (1964) indicated the 11 three character states usually used to separate Kohlrauschia from Petrorhagia, namely the annual habit, a capitulate inflorescence, and petals possessing a distinct claw and limb, are found in some Petrorhagia species outside of their section Kohblrauschia. Evidence presented by Schaper (1936) illustrates a great degree of ultrastructural similarity in the seeds of P. prolifera and P. saxifraga, species that would be in different genera if Koh/rauschia is recognized. TAXONOMIC CRITERIA Petrorhagia is a difficult genus to characterize morphologically since variability which can include the predominant states found in both Dz- anthus and Gypsophila is present in some characters (e.g., presence/absence of “epicalyx” bracts, petal structure). Petrorhagia can be defined as having a combination of seed characters found in Dianthus (a straight embryo in the center of a dorsiventrally compressed seed as in Figs. 1 and 2) and calyx characters found in Gypsophila (few veins per sepal and scarious commissures separating adjacent sepals). The separation of Petrorhagia from Gypsophila is a bit more distinct than the boundary between Dianthus and Petrorhagia. As Ball and Heywood (1964) noted, the seed offers a constant feature that clearly aligns Petrorhagia with Dianthus and Velezia. Pollen data presented by Candau (1980) suggest a similar alignment; pollen shape differences existed between species of Dianthus and Petrorhagia, while pollen of Pet- rorhagia and Gypsophila species differed in both grain size and aperture number. On the other hand, calyx characteristics offer a clear separation between most species of Petrorhagia and Dianthus, except for two species of section Dianthella which possess calyces that approach a Dianthus condition, having more veins per sepal and almost lacking scarious commissures between adjacent sepals. A similar situation exists at the species level, with some very evident characters having little if any diagnostic value. The best example of this situation involves Petrorhagia velutina (Guss.) P. Ball & Heyw., a species introduced into California, Oklahoma, and Texas. Most plants of this species, such as those in California, show obvious glandular pubescence on the middle internodes of the stem, illustrating the ‘‘velvety” nature implied by the epithet ve/vtina (Smith, 1972). The stems of plants collected in Oklahoma and Texas are almost always glabrous, a condition Ball (in Tutin et al., 1964) indicated as occurring in some native populations, especially in Italy. Other morphological characters, such as sheath length, petal venation, and seed surface morphology, are very similar on plants from both areas, showing the fallability of pubescence as a diagnostic character. 12 The size and surface morphology of the seeds are often very useful features in distinguishing species of Petrorhagia, including those found in North America (Figs. 3 — 6). Three of our taxa, P. prolifera, P. nanteuilii, and P. velutina are very Closely realted, with P. nanteuilii probably derived from hybridization of P. prolifera and P. velutina sometime in the past. Inspection of the dorsal seed surfaces of these taxa offers a constant character for distinguishing them, with seeds of P. nanteui/1i (Fig. 5) possessing the basic size and shape of P. a (Fig. 4) and a tuberculate surface approaching that of P. velutina (Fig. 6 Other morphological features are correlated with the seed characters to allow positive identification of collections in the absence of seed; these features are noted in the key and descriptions wherever possible. One apparently overlooked character useful in analyzing North American col- lections 1s the presence and pattern of darkened petal veins in three of our four species (Figs. 7 — 10). This character offers another feature that can be used to distinguish members of the P. prolifera ‘‘complex.” Petals of P prolifera (Fig. 8) are uniformly colored; darkened stripes are absent. Petals of P. nanteuili (Fig. 9) that were examined possessed one prominent dark pink stripe on the central vein, while a very faint dark area may be present on the two veins adjacent to the central vein. At least three dark stripes were evident on P. velutina petals (Fig. 10) with five or more stripes occasionally present. A similar pattern of petal vein coloration on European specimens of these taxa was noted. TAXONOMIC TREATMENT The material presented below deals with the delimitation of the genus as it appears in North America. Synonymy present in Ball and Heywood (1964) will not be repeated here except in cases where usage in this paper or North American references warrant it. Several additional binominals noted during the study will be presented to supplement Ball and Heywood’s listings. Abbreviations for major works are taken from Stafleu and Cowan (1976, 1979, 1981, 1983) where possible, with additional abbreviations taken from Tutin et al. (1976, 1980) PETRORHAGIA (Ser. in DC.) Link, Handbuch 2:235.1831. — Moench, Methodus. 60. 1794. Imperata Cirillo, Pl. Rar. Neapol. 2: xxvi, t. 11. 1792. aie sect. Porothacie Ser. in DC., Prodr. 1:354. 1824. Tunica sensu Mert. & Koch in Rohling, ee Fl. . 3. 3:182. 1831. non Tunica bude; Inst. Regn. Veg. ed. 29. 1757, non Tunica Poh in Ludw., Def. Gen. T ed. : 298. 1760, non Tunica Adans., Fam. Pl. 2:255. 1763, non Tyvnica Scop., Fl. Carn. ed. 2. 1:298. 1772. Kohlrauschia Kunth, Fl. Berol. 1:108. 1838. Kolrauschia Kunth ex Fourr., Ann. Soc. Linn. Lyon sér. 2. 16:345. 1868. orth. var. Petroragia Link ex Kuntze, Lex. Gen. Phan. 427. 1903. orth. var. Annual or perennial herbs. Stem internodes pubescent or glabrous, pubescence glandular or not. Leaves simple, opposite, linear to narrowly FIGS. 7 — Flowers sap ilice spp. Scale = 2 _7.P. saxifraga, Canfield Lake, Michigan. 8. P. prolifera, Grand Haven, Michigan oo. P. nantenilii, N of Cazadero, California. 10. a velutina, S of me Texas. 14 lanceolate, usually 1- or 3-veined, margin often basally scabrous or ciliate; sheath of variable length, often 1 — 3 times as long as broad. Inflorescence basically a dichasial cyme, bracteate or not, with flowers solitary, fascicu- late, paniculate, or capitate (ours solitary or capitate). An “epicalyx” of bracts directly subtending the calyx present (ours) or absent; when present, |—3 pairs evident. Sepals 5, fused, each 1 —3 (rarely 5—7) veined, a veinless commissure separating adjacent sepals. Petals 5, clawed (ours) or not; often pink or white; apex entire to bifid; primary veins | — 3 per petal, with or without contrasting color stripes on veins of the limb. Stamens 10, anthers of ours often pink, blue, or white. Styles 2, can be stigmatic along entire length. Capsule oblong, 4-lobed, dehiscing by 4 apical teeth. Seeds dorsiventrally compressed with facial hilum and straight, central embryo; reddish-brown (immature) to blackish-brown (mature); size variable. Pol- len oblate-spheroidal, porate; tectum bearing small spines (Candau, 1980). = 13 and 15 (Favarger, 1966 Type: Petrorhagia i (L.) Link, vide Ball and Heywood (Bull. Brit. Mus. (Nat. Hisc.), Bot. 3:130. 19 KEY TO PETRORHAGIA AND THREE RELATED rE IN NORTH AMERICA | Bracts subtending the calyx present 2 Commissures (veinless scarious areas) present between adjacent sepals, 1— 3 veins per sepa Petrorhagia 2 Commissures absent, 5 or more veins per sepal Dianthus 1 Bracts subtending the calyx absent 3 Commissures present between adjacent sepals, seed laterally com- pressed with curved embryo Gypsophila 3 Commissures absent, seed dorsiventrally compressed with straight Velezia The “epicalyx”, although more evident than calyx commissures, is not completely diagnostic for Petrorhagia since 14 of 29 species do not possess subtending bracts. A key, such as that in Flora Europaea (Walters in Tutin et al., 1964), using calyx commissures as the first character is appropriate for separating all Petrorhagia species from related genera. KEY TO SPECIES OF PETRORHAGIA IN NORTH AMERICA | Flowers solitary, or rarely in fascicles of 2—3 (a few cultivars); subtending racts narrow and short, enclosing to + one-half the calyx ............ . P. saxifraga | Flowers borne in capitate inflorescence (solitary in some very young plants); bracts of the inflorescence very broad and long, enclosing entire calyx of most flowers 2 Leaf sheath about as long as broad, |—2 (rarely 3) mm long; petals truncate or emarginate, no dark colored areas on vein of petal limb 2. P. prolifera 2 Leaf sheath 1.5 — 3 times as long as broad, usually 3 mm or longer; petals obcordate to bifid, 1 —3 (or more) dark areas present on veins of limb 3, Leaf sheath (2)3 —4 mm long; inner inflorescence bracts obtuse or mucronate; 2 of 3 dark veins of petal limb often faint; seeds (1.3) 15-1. — 1.8 mm long, tuberculate 3. P. nanteuilii 3, Leaf sheath variable in length, (3)4 — 6(9) mm long; all inflorescence bracts mucronate; 3 (or more) dark areas on veins of petal limb; seeds 1.0 — 1.3(1.4) mm long, covered with conical papillae ........... 4. P. velutina 1. PETRORHAGIA SAXIFRAGA (L.) Link, Handbuch 2:235. 1831. ee ae Sp. Pl. 1:413. 1753. Gypsophila saxifraga (L.) L., Syst. Nat. ed. 1759. Tunica saxifraga (L.) Scop., Fl. Carn. ed. 2. 1:300. 1772. Silene tunita a a Krause in Sturm, Deutschl. Fl. ed. 2. 5:107. 1901. ice saxifraga (L.) Degen, Fl. Veleb. 2:94. 1937. Kae saxifraga (L.) Dandy, Watsonia 4:42. 1957. Type: not seen. LINN 579.25 (photo at F!; microfiche at US!) is authentic material. See Savage (1945) for explanation of penned names of 579. 24 and Described from Switzerland, France, and Germany. Gypsophila scabra Schultes ex Steudel, Nomencl. Bot. ed. 1. 386. 1821. Tunica meee var. scabra (Steudel) Schur, Oesterr. Bot. Z. 19:16. 1869. Original poaterialp not gl as Guss., Suppl. Fl. Sic. Prodr. 120. 1832. Tunica p ta (Guss. DE. t in Hohen., Bot. ee (Leipzig) 16:295. 1856. Tunica ca var. nee an ) Nicocex: Prodr. Fl. Messan. 122. 1883. Type Locatiry: SICIL original material not seen (NAP? ae is arenicola Dufour, Bull. Sec Bot. France 7:240. 1860. Tunica arenicola (Dufour) Nyman, Consp. Fl. Eur. 100. 1878. Type Locatiry: SPAIN; * ‘frequens in arena maritima valentina (Dehesa)’”’, original material not se Gypsophila rigida Sibth. & Smith, Fl. Graeca lV. p. 75, t. 382. 1823, non L., fide Giirke in Richter (1903) and Degen (1937). Imperatia bithynica Degen, Fl. Veleb. 2: 95. 37. Type Locauiry: TURKEY, Olympus Bithynus, original material not seen. Perennial, sometimes woody at base. Stems much branched near the base, 5 to 40 cm [to 45 cm in Ball and Heywood (1964)} tall; internodes glabrous above, scabrous below. Leaves linear, 5 — 20(30) mm long, 1(2) mm or less wide, 1-veined, margin basally ciliate; leaf sheath 1 mm or less long, about as long as broad. Flowers solitary (fasciculate in some cultivars and var. glomerata), terminal, 10 mm or less long, on long peduncles. Subtending bracts 2(4), ovate, membranous, mucronate, 1-veined, enveloping up to one-half of the calyx. Sepals 1-veined, margin often ciliate. Petals clawed, limb white to pink, claw often white; primary veins 3 per petal, often dark pink near base of limb; apex obcordate. Anthers pink or white. Seeds with tuberculate surface, (0.8)1.0— 1.2(1.3) mm long, 0.5 —0.8(1.0) mm broad. 2n = 30, 60 (Favarger, 1966). Figs. 3, 7, and 11 All North American collections I have seen are referable to P. saxifraga var. saxifraga. Variation within var. saxifraga is formally recognized in a 16 few works [So6 (1970) listed six forms and a “‘variant’’}, although it is dismissed by Ball and Heywood (1964) as being of questionable signifi- cance. Cultivars, for the most part, are not significantly different from naturalized collections, although some ‘modifications’, such as doubled petals, do occur. Schlising and Iltis (1962) cited two common names for this plant; Tunic Flower and Coat Flower, both names referring to the bracts surrounding the flower. Britton (1913) listed Saxifrage Pink and Tunica as common names. A translation of the binomial reveals that both parts are derived from words meaning “rockbreaking”’, Petrorhagia from Greek, saxifraga from Latin, alluding to its prevalence in rock crevices (Smith, 1972). FLORAL BIOLOGY: Meusel and Muhlberg (1979) reported that P. saxifraga is strongly protandrous and suggested that self-pollination is hardly possible. Knuth (1908) noted that the style usually matures late enough to prevent selfing. The flowers produce abundant nectar and are visited mainly by small bees, flies, and butterflies (Meusel and Muhlberg, 1979): ECOLOGY AND DISTRIBUTION: Ball and Heywood (1964) summa- rized the native range as “Central and southern Europe and south-western Asia’, with naturalized introductions noted for Sweden and Great Britain. Within North America, most records represent either obvious cultiva- tions or adventive populations (see Fig. 13). From label data, it is clear that many collections were made along roadsides, often from dry sandy areas, while others represent plants in lawns, waste areas, field edges, sandy forest slopes, a lake edge, and gravelly banks. Flowering reported from June to October, the last flowering possibly limited by first autumn frost. HISTORY OF INTRODUCTION: Petrorhagia saxifraga appeared in North America just over 100 years ago, with initial collections being made along roadsides at College Point and Flushing (Queens Co.), New York (Schrenk in 1876, AC, CU). According to a note on the AC specimen, the populations remained “‘well established” there through 1879. The next collections seen were made at London, Ontario in 1886 and 1887, cited by the collectors (Burgess in 1886, CAN, GH; Dearness in 1886 and 1887, MTMG) as introduced and spreading in gardens, waste places, and road- sides. A look at popular manuals will reveal that these sites are often the only ones listed for P. saxifraga, a practice started by Robinson (1897) and Britton (1897) and continued by Maguire (1950, 1952). Robinson and Fernald (1908) listed these sites in the 7th edition of Gray’s Manual, but Fernald (1950) omitted them from the 8th edition along with any mention of the genus or the species! After studying extant collections, it is readily : ye $5 2 as a& és a8 § b . NY a “, \ \ & ee a Bre se eel AN Re ee = al % a me FIGS. 11— 12. Representative herbarium specimens of Petrorhagia spp. Scale = 11. P. saxifraga, Rabeler 313 (MSC). 12. P. prolifera, Rabeler 314 (MSC). 5 cm. 18 apparent that, by 1950, this expression of range was grossly inadequate; P. saxifraga had been collected in at least 16 additional states (19 including literature references) and 2 provinces (records of six states and one province are obvious cultivations) by that date! CURRENT STATUS: The list of specimens cited below includes the earliest and most recent records seen from a given state or province that are not thought to be cultivated. It appears that the non-cultivated range of the species is much smaller today, with post- 1960 collections seen only from only Michigan, Virginia, Wisconsin, and Ontario. In Michigan, P. saxzfraga has been collected in 10 counties since 1960, almost all specimens representing naturalized pop- ulations. It is extremely abundant as a lawn and roadside weed in the Manistee area, where the first collection dates from 192 Post- 1960 collections of cultivated plants were seen from Massachusetts (Wiegel and Hodgdon 12043, NHA), Michigan (Bourdo 4056 & 5865, MCTFE; Rabeler 308, MOR, MSC, NA), Minnesota (McWz//iams in 1966, NA), and New York (Stites in 1978, BH), indicating the potential for growth and possible escape in other areas. Since P. saxifraga is available commercially as a landscape plant, it is difficult to predict when or where the next escape may take place, and whether or not an adventive population will result. Representative specimens: UNITED STATES. IbaHo. Koorenar Co.: Sandy lake shore, Coeur d’Alene, Oct 1914, Rust 446 (WTU). ILLinois. CHAMPAIGN Co.: Roadside, Champaign, 28 Jun 1950, E/ein 147 (ILL). Cook Co.: South Park is 1887, ee n. (F). MAINE. KNOx Co.: Rockport, 23 Aug 1930, Steyermark 416. . MASSACHU Bristo. Co.: Rockery, North Easton, | Jul 1916, rina N nN. ce age eat pe Dooryard, eae 9 Aug 1943, Clark s.n. (N ). MicHiGANn. Dera Co.: Gravel bank fill, Fishdam River, 3 mi E of Isabella, pee RI8W, , SW "a 11 Sep 1981, panne MIN, MSC, RM, VT); 12 Jul 1982, Rabeler ee UTC, VDB); 17 Aug 1982, Henson 1436 (MICH). Kent Co.: Grand Rapids, 22 Jun 1899, Stevenson s.n. (MICH). Minnesota. St. Louis Co.: Wooded ae at Vermillion Dam, 7 Aug 1940, Lakela 3992 (DUL, MIN, MO, SMU, UC). New Jersey. Cape May Co.: Roadside, Cold Spring, 12 Aug 1915, Brown s.n. (PH); roadside ee row, Cold cae 25 Sep 1920, Brown s.n. (PH). New York. Queens Co.: College Point, 1876, Schrenk s.n. (AC); Flushing, L. I., 1876, Schrenk s.n. (CU). Tompkins Co.: Lawn weed, Cornell Heights, Ithaca, 17 Jul 1940, Eames, Flora of New York 20162 (CU). PENNSYLVANIA. MONTGOMERY Co.: Pennsburg, 28 Jun 1919, Brendles.n. (PH). SourH Dakota. LAWRENCE Co.: Glade on talus slope, alc. 3900 ft, 11 Aug 1942, Bennett 1521 (MO). VirGinia. Pace Co.: Old homesite, Skyland, 4 Jul oe Maze ate ay 7 Sep 1966, au 1781 (NA). WISCONSIN. COLUMBIA Co.: rnfield a oadside, Okee, TION, R8E, S7, 29 Sep 1967, Domres s.n. it SHEBOYGAN Co.: a Shebayean: Aug 1912, Goess/ 5.n. (WIS) CANADA. BritisH CoLuMBIA. Prince George, Sep 1937, Travis s.n. (DAO). ONTAR- 10. MippLesex Co.: Gardens and waste places, London, Sep 1886, Burgess s.n. (CA ‘PEvALIXPS PISPGLOMag JO UOTINGINSIP ULIIIIUTY YYON “€1 “Old 800 RLOMETIAS X yet OSS SCALE Rye ot ; A 2 e pre-1960 a cultivations —*° ® lit.records ®® 1960+ 61 20 : Waste soil near grain elevator, Collingwood, 9 Sep 1978, A. A. and S. A. ce ae re (MICH, [TRT}). Records mapped from literature (specimens not seen): low a. Grinnell(?) (Conard, 1943). MASSACHUSETTS. Worcester (Potter & Woodward, 1935). MicHIGAN. Niles (Beal, 1908). 10. Hocking County in 1930 (Cusick & Silberhorn, 1977). WermMonr. Morrisville (Dole, 1937) Literature records rejected: District Or Co_tumBta. Brookland (McAtee, 1940; Her- mann, 1946). This record is based on a misidentified collection of Gypsophila muralis (Ulke 5.n., without date, US). 2. PETRORHAGIA PROLIFERA (L.) P. Ball & Heyw., Bull. Brit. Mus (Nat. Hist.), Bot. 3:161. 1964. Dianthus prolifer L., Sp: Pl. 1:410. 1753, nec Forsskal, Fl. Aegypt. — Arab. xxv. 1775, Fl. Graec. Prodr. 1:285. 1809, fide Graebner and sien (1921). Tunica proly ioe ., ae Fl. Carn. ed. 2. 1:299. 1772. Kohlrauschia prolifer (L.) Kunth, FI. Bere rt. 1: 109. 1838. Cylic piesa ale (L.) Dulac, Fl. Hautes- 7 1867, nom. a Cans probe ) Arcang., Comp. os Ital. ed. . 303. 1894. Siemiaet.) .H. Krause in sane Deutschl. Fl. ed. 2. 5:107. 190 OL. Lecroryee (T — ye ex Horti Uppsal. (S, microfiche at eee See Thomas for argument rejecting earlier lectotypification of Ball and Heywood (1962 LINN 581.7 eas at BH!, F!, GH!, NY!). Described from Germany a southern Europe Dianthus diminutus L., Sp. Pl. ed. 2. 1:587. 1762. Caryophyllus diminutus (L.) Christm. , Vollst. Pflanzensyst. 6:563. 1780, fide Merrill (1938), nom, illeg. Kohlrauschia diminutus (L.) Reichb., Icon. Fl. Germ. Helv. 6:43, t. 247, f. 5008. 1844. Type: not een. Described from Germany. Dianthus carolinianus sensu Torrey & A. Gray, Fl. N. Amer. 1(2):195. 1838, non Walter. Type: from South Carolina, Wa/ter a (BM. photo at A!). — Annual. Stems simple or branched near the base, (6) 1 1 — 60 (commonly 30) cm tall; internodes mostly glabrous, middle internodes may be slightly scabrous. Leaves linear to linear-oblong, 10 mm long, 2(3) mm or less wide, 3-veined, margin scabrous; leaf sheath 1 - 2 (rarely 3) mm long, as long as broad, or at lower nodes, often broader than long. Inflorescence capitate (occasionally reduced to one flower), 10 mm or more long, usually 5 — 20 mm broad. Subtending bracts broadly ovate, brown-scarious, many veined, enclosing the flowers; tips of bracts obtuse, or outermost may be mucronate. Sepals 3-veined, outer surface not scabrous, margin glabrous. Petals clawed, pink or slightly purplish (rarely white); primary veins | per petal, dark coloration near base of limb absent; apex truncate or emarginate. Anthers pink or blue; pollen 40 fm in diameter, apertures of 4 bm (Candau, 1980). Seeds with fine to coarse reticulate surface, (1.1) .3— 1.6(1.8) mm long, (0.7)0.8 — 1.0(1.1) mm broad . 2n = 30 (Ball & Heywood, 1962; Thomas & Murray, 1983). Figs. 1, 2, 4, 8, and 12. ZA Variance shown in the above characters is, for the most part, distributed throughout the range, although slightly scabrous internodes and finely reticulate seeds are more prevalent in plants from New Jersey, Pennsyl- vania, and Virginia. I agree with Ball and Heywood (1964) in not applying subspecific categories, such as those used by Briquet (1910) and Maire (1963) to describe leaf margin texture variation, or the seven forms and two “variants” listed by Sod (1970), to segregate minor variants within P, prolifera. Britton (1913) gave three common names for P. prolifera; Proliferous Pink, Childing Pink, and Childing Sweet William. Smith (1972) defined the epithet “prolifera” as: “‘Proliferous, i.e. free flowering or producing side shoots or buds in order to increase.” This is an appropriate description of the plant and “proliferous’” can be extended to include the relatively large quantity of seed produced by each plant. FLORAL BIOLOGY: Thomas and Murray (198 1) described two “‘races”’ of P. prolifera after finding selfing and outcrossing populations that were reproductively isolated from each other. Subsequently, Thomas (1983) treated the large-flowered outcrossing plants as a separate “sister” species (P. obcordata), leaving P. prolifera as having small, autogamous flowers that produce little nectar and are homogamous, or as in one of their study populations, protandrous. I have observed distinctly protandrous flowers with protruding stamens and (later) style branches at Grand Haven, Michi- gan, suggesting that some outcrossing may take place from chance insect visits. ECOLOGY AND DISTRIBUTION: Ball and Heywood (1964) stated the native distribution as “Central Europe, mountains of southern Europe, Caucasus, Turkey (northern Anatolia), mountains of western North Afri- ca”, with introductions noted in Great Britain. Nearly all collections of P. prolifera within North America are from the southeastern portion of the continent (see Fig. 14), most being gathered from roadside localities, either in sand, gravel, or shale fill. Dry fields and pastures are mentioned as collection sites from Pennsylvania to North Carolina, while seven Virginia collections are of plants found growing in cinder railroad ballast. The calcareous soils of the Ozark region may be a factor in the spread of P. prolifera away from the roadside and into the glades and pastures in northern Arkansas and southern Missouri. Flowering re- ported from late May through late September. HISTORY OF INTRODUCTION: Petrorhagia prolifera apparently first appeared in North America near Philadelphia, Pennsylvania shortly after 1800. The earliest specimens seen bear the name Dianthus carolinianus, a 22 name published by Walter (1788) in his Flora Caroliniana. A problem develops when one attempts to determine what plant Walter had in mind for D. carolinianus. Index Kewensis (Hooker & Jackson, 1895) lists D. carolinianus as a synonym of Dianthus armeria L. Asa Gray inspected Walter’s herbarium in 1839 (at the time in possession of John Fraser in England) and wrote “his ‘Dianthus carolinianus’ is Frasera! in fruit.” (Brit- ten, 1921). In his monograph of Dianthus, Williams (1893) considered D. carolinianus a synonym of Dodecatheon Meadia L. Britten (1921) reached the same conclusion when he investigated Walter’s herbarium, stating ‘but the Dianthus is not Frasera, but Dodecatheon Meadia.” A translation of Walter's description of D. caroliniana is not of much help: “with flowers clustered on long peduncles, tube scales smaller by one-half.” (Walter, 1788), charac- teristics which could apply to fruiting material of all of the above- mentioned taxa! An inspection of the set of photographs of Walter's herbarium at A (Schubert, 1946-47) revealed a small specimen in the upper left corner of page 40 inserted in a small piece of paper labeled “334 Dianthus carolin.” It is indeed Dodecatheon meadia, consisting of several erect capsules and attached calyces and a 7.5 cm section of the scape. Torrey and Gray (1838) listed D. carolinianus, citing South Carolina for its range and add the following note: “D. prolifer was sometime since cultivated at Bartram’s garden under this name’, connecting the name to early collections from the Philadelphia area. The actual dates of cultivation of P. prolifera at Bartram’s garden remain a mystery. John Bartram started a seed exchange with a number of European botanists in the late 1730's, trading native American plants for those of Europe for cultivation in his garden. This activity continued into the 1830’s under the guidance of Bartram’s children after his death in 1777 (Berkeley & Berkeley, 1982) and would provide a logical explanation of the appearance of P. prolifera. The earliest evidence of its cultivation is the specimen at PH collected by S. W. Conrad labeled “Dianthus carolinianus Walter. At, and in the vicinity of Bartram’s Garden.” This collection would have been made prior to 1815 based on details of Conrad’s life and his use of Linnaean classification (Decandria Digynia) on the label (Harshberger, 1899). The earliest dated collection seen was that of E. M. Durand (Darand in 1837, GH) on which he noted: ‘It seems to be perfectly naturalized on that spot {a rock near the garden] and is not found in the garden itself.”” On both this specimen and an undated collection from South Carolina (Durand, NY), Durand questioned his determination of D. carolinianus, with the South Carolina specimen labeled: “D. prolifer of Europe and believe it to be the same, introduced.” The site of the first naturalized population referred to by Durand is most likely the same one that at least 10 collectors visited in ‘paafijoad PIspgsosjad JO VOLINAINSIP ULIaWY YON “Fl ‘Old ® pre-1960 4 Cultivations — |" ® lit.records —— ®@® 1960+ WES) LonciTuor 24 the next 60 years; a hill, described as “‘dry, micaceous-sand” (C. E. Smith, NY), along the Schuylkill River, near Bartram’s garden and Gray’s Ferry. It is not known how long this population existed, but the most recent collection seen was dated 21 October 1891 (Crawford, PH). Other pre-1900 collections exist for Delaware (1896-97), Maryland (1887), New Jersey (187 1), New York (Britton, 1879, specimens at SIM?), and Ohio (189 1—96). CURRENT STATUS: The specimens cited below include the earliest and most recent records seen from a given state excluding collections of cultivated material. The distribution of post-1960 collections (Fig. 14) is far different from that listed by major manuals. Fernald (1950) stated “sandy fields and roadsides, local, s. N.Y. to Del., Va., Ky. and O.”, which, except for the Ohio reference, corresponds to the distribution of pre- 1949 specimens present in the loan received from GH for this study. Maguire (1952) considered T. prolifera “sparingly introduced in waste places, N.Y. to S.C. and Cal.” This description nearly matches the pre- 1952 specimens seen from NY, the institution Maguire was associated with at that time (the California specimens at NY are P. velatina, although labeled T. prolifera; one Idaho collection (cultivated?) labeled D. armeria, one Ohio collection not mentioned). Both treatments generalize the dis- tribution by inserting “to” and thus connecting widely separated local populations, making the species seem more widespread than it may actually be. The reader is referred to three similar situations recorded elsewhere (Shinners, 1965; Shinners, 1969; Rabeler, 1981) which treat discontinu- ities between current ranges and those given in the literature. In light of the discrepancies noted, afew general conclusions are in order. Several states should be dropped from previously published ranges; Califor- nia (the origin of the Congdon collection (1902) remains a mystery), Ohio (last collected there in 1896), South Carolina (record relies on Durand collection of early 1800's), and Texas, as listed by Shinners (1969) and Correll and Johnston (1970), since all Texas collections are in fact P. velutina. Several areas might be retained, since populations could exist there in spite of available records; Delaware (last collected there in 1897, but present in all neighboring states; see Phillips (1978) for opposing view), Kentucky (single record in Braun (1943) based on a 1941 collection), New York (several collections on Long Island, 1920-48), North Carolina (1949 collection mapped in Radford et al. (1965), but not discussed by the collectors in their report on additions to the flora of North Carolina found in 1949 and early 1950 (Fox, Godfrey, & Blomquist, 1950), or in Radford et al. (1968)), and West Virginia (single report by Core (1941), but Monroe ee) Co. is close to “active” Virginia populations). Post- 1960 records define a pattern of scattered, local populations from New Jersey southwestward into Virginia and then generally westward into Arkansas and Oklahoma, along with disjunct populations in western Michigan [discussed in more detail in Rabeler (1980)}. The southwestward expansion has taken place for the most part since 1930, with an invasion into the westernmost states occurring since 1950. The irregular timing of appearance in adjacent states makes a theory of multiple introduction into the southeast far more plausible than trying to derive all these populations from the initial introduction in Philadelphia. Several collections seen indicated that P. prolifera has been present at some sites for over 30 years, suggesting naturalization has taken place in areas of Maryland, Pennsylvania, and Virginia. The absence of P. prolifera from a recent survey of roadside vegetation at selected sites in southwestern Virginia (Schmaltz, 1981) reinforces the local distribution of P. prolifera, even in the vicinity of naturalized populations, since it has been reported at least once in five of the seven counties included in the study. Other post-1960 collections represent adventive occurrences, especially records in three states (Georgia, Michigan, and Tennessee) where the earliest speci- men seen was collected after 1975. How then does P. prolifera get around? So6 (1970) listed anemochory, endozoochory, and autochory as dispersal mechanisms of P. prolifera in Hungary. Petrorhagia prolifera, as wellas P. nanteuilii and P. velutina, woul be defined by Van der Pijl (1982) as wind-ballists, with seeds falling from the capsules as the long, wiry stems sway in the wind; a combination of anemochory and autochory would be ideal in open areas such as road shoulders. Ridley (1930) reported Dymes has found that one of the common seeds in an ant grainery in Italy was P. saxifraga; similarities between the seeds of P. saxifraga and the above-mentioned species suggest that harvester ants may be responsible for occasional short distance dispersal events. The presence of ‘roadside’ on a large number of labels suggests that occurrences of P. prolifera may be partially “transportation-related’’. Frenkel (1970), ina study of roadside vegetation in California, listed several attributes present in many roadside plants, some of which apply to P. prolifera: annual habit; small, light, non-appendaged seeds produced in abundance; and tolerance of high light intensity. Wofford et al. (1977) suggested that P. prolifera has appeared in Tennessee as a contaminant of grass seed planted along Interstate 40; contaminated seed is likely respons- ible for the appearance of P. velutina in Texas (Shinners, 1969; Correll & Johnston, 1970) and possibly some of the plants in Michigan (Rabeler, 1980). This situation may easily have occurred in other areas along the 26 recently constructed interstate highways of the region, a system primarily built since 1960 Some roadside populations of P. prolifera may be unintentionally en- larged if the shoulder area is mowed in mid-summer, since plants would have open capsules of seed awaiting dispersal. This notion became evident after observing a tremendous increase in both number of plants and the area they occupied along an infrequently mowed Michigan roadside 5 years after discovering the population, an increase that I find hard to explain if only “natural” dispersal ts invoked. Representative specimens: UNITED STATES. ALABAMA. FRANKLIN Co.: Roadside, Co. 79, | mi N of AL 24 jet., Russellville, 28 May 1967, Baskin, Candle, Turner 582 (VDB); Roadside, Co. 83, 0.8 mi N of AL 24 jct., E of Russellville, 5 Jun 1981, Godfrey, Gholson, and Webb 78789 (FSU). MaRION Co.: Sandy clearing gs AL 17, N of se 20 Jul ae Kral ieee (AUA, GH, MICH, MO, OS, SMU, TENN, UNA, US, VDB). 5 N Co.: Old Sasie near creek, Manimoth Spring, 17 jon 1951, Moore See ae a US 63 at Trace Creek, T21N, RSW, S4, 4 Jul 1968, Thomas, Bio. (Bot.) 451 class 10046 (A, CHSC, CM, NCU, NLU(2), NY, SMU, TENN, USF, WTU) WASHINGTON Co.: Along Hwy 71, 6 mi S of Westfork, 8 Jul 1975, Meeks 127 (UARK). CALIFORNIA. Mariposa Co.: Mariposa Creek, 15 Jun 1902, Congdon s.n. (MIN, US). DELAWARE. Sussex Co.: Sandy fields and Lae S. Milford, 16 Jul hae Commons s. n. ne NYS); road to Slaughter Beach, 12 Aug 1897, Commons s.n. (PH). GEC : Pasture with granite outcrops, W of Siltacn. 5 Jul 1983, Allison 1836 GA). ROck- DALE Co.: Roadside granitic flatrock, Conyers, 28 Jul 1983, A//ison 1845 (GA). IDAno. BONNER Co.: Gravel slope, Sandpoint Substation, 11 Sep 1932, Christ 2093 (NY). KENTUCKY. ROBERTSON Co.: Roadside, Kentonville [Kentontown?}, 8 Jul 1941, Braun 4041 (GH, US). MARYLAND. CALverRT Co.: Sandy roadside fill, 1.2 mi S of Bowens, 13 Jun 1981, Rabeler 578 feeeee MARY, MSC). Kent Co.: Sandy fields near Millington, 18 Jun l 87, Brinton s5.n. (PEN PH). MicHiGAN. MuskEGON Co.: Beside paved area, main entrance, P. J. eer State Park, 18 Jul 1983, Wells and ou 83221 ({BLH}, C). Orrawa Co.: Kite une, Grand Haven, 8 Aug 1976, Atwood, Beaman, and Ra 409 (MSC); sandy ae S edge of Kitchel Dune, een C onservancy Preserve} , RIGW, 820, SW 1/4, 9 Aug 1979, Rabeler 314 (CAN, DAO, GA, HSC, MICH Wee NY, SMU, ‘WAT). MissourRt. GREENE Co oe Uinpectone rene Kissick, T28N, R21W, S19, 6 Jun 1982, Summers 1025 (MO). STONE Co.: Rocky places near pasture, 5.5 mi SE of Shell Knob, T22N, R24W, NE sect. 18, NW sect. oF 13 Jul 1956, Steyermark 81924 (F, GH, ILL, MO, UMO). New Jersey. CAMDEN Co.: Roadside, Haddonfield, 3 Jun 1871, Parker s.n. (F, GH, MO, PENN, PH). CumBer.tanp Co.: Fallow fields, NJ 548, near Mauricetown, 31 Jul mr id 98005 (NCU). New York. SUFFOLK Co.: Dry sand-pit, Southampton, 18 Aug 1920, St. Jobn 2878 (CU, GH, NYS, PH, US); Noyack, 16 Sep 1948, Latham 28248 (NYS). oo CAROLINA. ASHE Co.: Dry hillside pasture, near US 221, N of Laurel Knob Gap, 7 Sep 1949, Fox and Godfrey 3368 (GH, NCSC, NY, TENN, US). ForsyrH Co.: Edge of field near Mt. Carmel Church, 19 Jul 1935, Correll 2702 (NA). On10. CuyAHoGa Co.: Cleveland, ae 1891, Beardslee s.n. (MICH, NY(2)); — 8 Aug 1896, Stair s.n. (OS). OKLAHOMA. CHEROKEE Co.: Open roadside, OK mi E of Hulbert, 2 Jun 1951, Wallis 533 (OKLA): Roadside, OK 10, 4.5 mi NE - ooo 30 Jun 1976, Taylor 22206 (DUR, NLU). PENNsYL- VANIA. Berks Co.: Roadside, NE of Virginsville, 16 Jul 1965, Wilkens 11659 (PENN). 27 PHILADELPHIA Co.: At, and in vicinity of Bartram’s Garden, without date, Conrad s.n. (PH); below Bartram’s garden, 18 Jul 1837, Durand s.n. (GH). SouTH eee A. CHARLESTON Co.: Vicinity of Charleston, without date, Durand s.n. (NY). TENNESSEE. CUMBERLAND Co.: See E lane of I-40, near mile 313.8, 28 Jun 1976, eae Wofford, Webb, Rader, and Smith eet (TENN); 13 Jun 1977, D. H. and B. Webb 971 (TENN). VIRGINIA. BATH Co.: Weedy railroad margin at Copeland, 23 June 1982, Wieboldt 4359 ies Fauquier Co.: In quarry lane S of US 55, 8 Jun 1941, A//ard 8889 (CM, F, GH, , US, VPI). MontGomery Co.: Pasture, near Blacksburg, 8 Jun 1930, Morton ea (US) Records mapped from literature (specimens not seen): MARYLAND. Annapolis (Shreve, 1910). New York. Staten Island (Britton, 1879). Vircinia. Augusta, Bedford, Nelson, Pittsylvania, and Roanoke Counties (Harvill et al., 1981). West VirGinia. Monroe Co. (Core, 1941). Literature record rejected: MICHIGAN. Kent Co.: Grand Rapids (Cole, 1901). This record is based on a misidentified collection of Petrorhagia saxifraga (Stevenson in 1899, MICH). 3. PETRORHAGIA NANTEUILI (Burnat) P. Ball & Heyw., Bull. Brit. Mus. (Nat. Hist.), Bot. 3:164. 1964. Dianthus nanteuilii Burnat, Fl. Alpes Marit. 1:221. 1892. Dianthus prolifer var. nanteuil- fii (Bu Hele oincy, J. Bot. ey 12:55. 1898. Tunica nantenilii (Burnat) Giirke in K. Richter, Pl. Eur. 2(3):338. 1903. Tunica prolifera var. nanteuili: (Burnat) Briq., Prodr. FL Corse. 1:569. ce Tunica prolifera subsp. pS eee Tia P. Graebner in Asch. & Graebner, Syn. Mitteleur. FI. 5(2): 264. 1921. Kohlrauschia a fare (Burnat) P. Ball & Heyw., Watsonia 5:115. 1962. Coe prolifera subsp. nanteuilii (Burnat) M. Lainz, Bol. ie ‘stud. pean are Cr lOs ys. 1964. Ces age ea nanteuilit — O. & Vigo, Butl. Inst. Catalana Hist , 38 Bot |: 1974. T en sur 38 éch eee de diverses ole de oe et a Agay, os A a obligeance de M. R. de Nante Original material not seen (G?). Kohlrauschia velutina var. eae Pérez Lara in Willk. Suppl. Prod. Fl. Hisp. 282. . Tun ee » Pérez Lara, Anales Soc. Esp. Hist. Nat. 25:197. 1896. TyPE SALITY: SPAIN, Cadiz province. Original material not seen (MAF). Dianthus prolifer var. atapuercae Coincy, J. Bot. (Morot) 12:54. 1898. pee a a var. eee) See Gurke in K. Richter, Pl. Eur. 2(3):338. 1903. T LITY: N, ‘Les bords de la grotte jurassique d’Atapuerca prés Burgos.” ia Outil ee not seen. Annual. Stems simple or branched near the base, 21 — 52 (often 30) cm tall; internodes glabrous or lower and center internodes somewhat scabrous with eglandular pubescence. Leaves mostly linear, 3-veined, margin scab- rous; leaf sheath mostly (2)3 — 4 mm long, 1.5 to 2 times as long as broad. Inflorescence capitate as in P. prolifera, 10—12 mm long, 9 — 17 mm broad, but with tips of inner inflorescence bracts either obtuse or mucro- nate. Sepals 3-veined, outer surface may be very slightly scabrous, margin glabrous. Petals clawed, pink or slightly purplish (limb rarely white); 28 primary veins 3 per petal, at least the center vein darkly colored at the base of the limb, 2 side veins may show faint darkening; apex obcordate or somewhat bifid. Pollen 46 wm in diameter, apertures of 4.9 km (Candau, 1980). Seeds tuberculate, (1.3)1.5— 1.8 mm long, (0.7)0.9— 1.0 mm broad. 2n = 12, 36 (Borgen, 1974), 60 (Fernandes & Leitao, 1971, Thomas & Murray, 1983). Figs. 5, 9, 15. A note on the inclusion of Tunica pinetorum Pérez Lara as a synonym of P. nanteuili (Burnat) P. Ball & Heyw. is in order. In his description, Pérez Lara (1896) clearly considered the plant intermediate between T. prolifera and 7. velutina and cited leaf sheath, bract, petal, and seed characteristics that are clearly referable to P. nanteniliz, In their catalog of the flora of Cadiz province, Galiano and Silvestre (1977) listed the four collections cited by Pérez Lara after the description of T. pinetorum under P. velutina without any explanation. The specificity of the description of T. pinetorum makes an examination of the specimens crucial before resolution of this discrepancy can be obtained. Much debate has taken place concerning the status of this plant, as shown by the above list of synonyms. Most chromosome counts of this species indicate a tetraploid condition, with 2n = 60. The question then arises as to the origin of the tetraploidy; is it an auto- or allopolyploid? Of hybrid or non-hybrid origin? Bocher et al. (1953) conducted the first investigation of these tetraploids, concluding they represent an autopolyploid race of Koh/rauschia prolifera. In their study, the only clear cut morphological difference cited was seed size, with other characters showing a great deal of overlap between diploid and tetraploid plants. Bocher et al. (1953) used the term “‘polyplotype’” to describe the situation and summarized the problem by stating: “There is greater reason to distinguish the diploid Kohlrauschia velutina from the diploid K. prolifera than to distinguish the polyplotypes within the latter.” This idea roughly parallels the taxonomy most recently expressed by Bolds and Vigo (1974). In considering P. nanteuilii as a subspecies of P. prolifera, they believed P. nantenilis to be a race of P. prolifera that has not attained a degree of difference great enough to warrant treatment as a distinct species. Ball and Heywood (1962) first suggested the alternative view; an allotetraploid with K. prolifera and K. velutina as parents. Their position was based on the presence of characteristics, including flowering time and seed surface texture, that are intermediate between the two species. They noted this heritage would be shown in the karyotype by the presence of a pair of very short chromosomes present only in K. velutina. As part of their study on the breeding systems in Petrorhagia sect. Kohlrauschia, Thomas and Murray (1983) conducted a cytological investiga- FIGS. 15— 16. Representative herbarium specimens of Petrorhagia spp. Scale = 5 cm. 15. P. nanteuilii, Bacigalupi, Robbins, & Hoffman 5676 (DAV). 16. P. velutina, Carter 368 (CHSC). 30 tion of P, nanteuilii. They noted the presence of 29 pairs of metacentric chromosomes and one pair of small telocentric chromosomes, reinforcing the contention of Ball and Heywood (1962), and found that P. nanteuilii behaved as an allotetraploid during meiosis, with strictly bivalent pairing. They suggested one genome was supplied by P. prolifera [based in part on homeologous pairing in artificial sterile hybrids noted earlier by Thomas (1980)}, the other most likely coming from P. velutina, although breeding barriers within P. velutina prevented formation of any P. velutina-nanteuilii hybrids. In addition, Thomas (1980) found that all attempts to cross P. velutina with P. prolifera were unsuccessful, suggesting that any hybridiza- tion that may have led to P. nanteuilii would have taken place prior to the development of the breeding barrier now present in P. velutina. This is confirmed by the lack of transitional forms between P. nanteuilii and the reputed parents, a situation first noted by Nanteuil (Burnat, 1892) and emphasized by Ball and Heywood (1962). On the basis of the above cytological evidence and observed morphologi- cal continuity, I have decided to follow both Ball and Heywood (1964) and Thomas and Murray (1983) in recognizing P. nanteuilti as a separate species. Maire (1963) used stem pubescence to recognize two varieties, a useless distinction considering the instability of this character in section Kohlrauschia. The common name Childing Pink is not used only for P. prolifera, Perring and Farrell (1977) applying it to P. nanteuilii. The epithet ‘“‘nan- teuili” was chosen by Burnat (1892) in commemoration of Edmond Nanteuil, who discovered this species near Cannes, France in 1885. FLORAL BIOLOGY: Thomas and Murray (1981) reported P. nanteuilii to be primarily autogamous, noting that the timing of anthesis and stigma emergence tended to coincide and seed set was high under insect-free conditions. Limited cross-pollination may occur since the stigmas do protrude above the corollas. ECOLOGY AND DISTRIBUTION: Ball and Heywood (1964) reported the species from “western Europe and western North Africa” with col- lections cited from the Channel Islands, Madeira, and the Canary Islands, areas where neither P. prolifera nor P. velutina have been collected. Two specimens have been seen from Australia (Clemens in 1944 and 1949, MICH), indicating an introduction to Queensland. Perring and Farrell (1977) listed it as endangered in England, citing the small number of localities and their accessibility to the public as reasons to be concerned about its status. The single California population (Fig. 17) is located among grasses in dry roadside soil, a habitat similar to several of the sites supporting P. prolifera at and P. velutina in other states. ee is reported in late May, with a few plants continuing into early Augus HISTORY OF INTRODUCTION AND CURRENT STATUS: Un- fortunately, the early history of P. nanteuilit in North America is unknown. Although both species that appear to have contributed to the genome of P. nanteuilii have been collected in California (P. prolifera only once, about 200 miles southeast of Cazadero), evidence presented by Thomas (1980) and Thomas and Murray (1983) suggests that no such hybridization ts currently possible. No evidence of previous cultivation either at or near the roadside site was present, suggesting that P. nanteuilii may have arrived as a seed contaminant. The specimens cited below illustrate the restricted occurrence of P. nantenilii in North America; a single wild population in California known since 1956 and an intentional cultivation at the Bailey Hortorium in 1969. An investigation of seeds and vegetative material from eastern Sonoma County provided by Dr. Charles Quibell revealed no trace of P. nanteuili among the P. velutina plants there, suggesting the population may be spreading slowly if at all. Representative specimens: UNITED STATES. CaLitForNiA. SonoMa Co.: Along secon- dary road following Big Austin Creek, at base of “The Butcher Knife’, elev. c. 600 ft, 31 May 1956, Bacigalupi, Robbins, and Hoffman 5676 (DAV, JEPS); along trail to Big Austin Creek, N side of King Ridge Rd., 4.4 mi N of Cazadero, TON, R11W, “S31”, 8 Aug 1980, Rabeler 507 (CHSC, GH, HSC, MSC, ROPA, UC, US). New York. Tompkins Co.: Hortorium Garden, Ithaca, 2 Aug 1969, Dress 10606 (BH69-192) (BH). 4. PETRORHAGIA VELUTINA (Guss.) P. Ball & Heyw., Bull. Brit. Mus. (Nat. Hist.), Bot. 3:166. 1964. Dianthus velutinus Guss., Ind. Sem. Boccad. 1825:2. 1825; Pl. Rar. 166, t. 32. 1826. ‘unica velutina (Guss.) Fischer & C. Meyer, Index Sem. Hort. Petrop. 6:66. 1840. Koblrauschia velutina (Guss.) Reichb., Icon. Fl. Germ. Helv. 6:43, t. 247, f. 5010. 1844. Dianthus prolifer subsp. velutinus (Guss.) Battand. in Baccand. & Trabur, Fl. ee 1:143. 1888. Dianthus prolifer var. velutinus (Guss.) Coincy, J. Bot. (Morot) _ 1898. Tunica prolifera subsp. velutina (Guss.) Briq., Prodr. Fl. Corse 1:570. i Petrorhagia prolifera ee velutina (Guss.) O. Bolds & Vigo, Butl. Inst. seats Hist. Nat., 38 Bot. _ 1974. Type: SICILY; ‘Val di Mazzara, e Val di , Madonie” (FI, NAP?), o} ‘Meiie (1977). Original material not se Bae ope ses Nicotra, Prodr. Fl. Messan. 123. 1883., nom. illeg., non Salis. Prodr. 303. 1796., non Panéié, Fl. Serbiae 178. 1874, nec Pancic, Nova Elem. Fl. Bulg. 185. 1886. Tyee Locatity: SICILY. Original material not seen Dianthus sartorii Fruehl ex Nyman, Consp. Fl. Eur. 107. 1878. Dianthus velutinus va. sartorit (Fruehl) F. Williams, J. Bot. 23:347. 1885. ig material not seen Dianthus diminutus sensu Desf., Fl. Atlant. 1:345. ts non L Dianthus prolifer sensu Friedr., Reise. 270. 1838, Petrorhagia prolifera sensu Shinn. , Sida 3:345. 1969, 1 Sens eral & M. Johnston, Contr. Texas Res. Found., Bot. Stud. 6:613. 1970, et i Ail Kartesz & Kartesz, Syn. . Vasc. Flora. 153. 1980, non (L.) P. yw Tunica prolifera sensu Munz & Keck, Calif. Fl. 293. een et sent. St. John, Pac. Trop. Bot. Gard. Mem. 1:160. 1973, non (L.) Scop. Annual. Stems often simple or sometimes branched near the base, 9.5 — 60(CA)-9 (TX) [commonly 25 — 40} cm tall; all internodes glabrous or nearly so (most Oklahoma and Texas collections), or middle internodes densely glandular-tomentose (most California collections). Leaves linear to linear-oblong, 10—GO mm long, 1.5(2) mm or less wide, lowermost oblanceolate, often broader (to 5 mm wide), 3-veined, margin scabrous; leaf sheath variable in length, (3)4-6(9) mm long, 2-3 times as long as broad. Inflorescence capitate as in P. prolifera, (10) 14— 20 mm long, 6— 10(23) mm wide, but with tips of all inflorescence bracts mucronate, reddened when young. Sepals 3-veined, outer surface often scabrous, especially along veins; margin glabrous. Petals clawed, pink or purplish (rarely white); primary veins 3 per petal, ‘pencilled crimson at base of limb” (Meikle, 1977), central vein may fork in bifid petals and at least 2 minor veins may also be colored, producing 5—6 dark veins; apex obcordate or (more commonly) bifid. Anthers blue or pink; pollen 32 4m (Candau, 1980). Seeds semipyriform, more angled (concave-convex) than above species, surface covered with conical papillae, 1.0—1.3(1.4) mm long, 0.7 —0.8(1.0) mm broad. 2n = 30 (Bocher et al., 1955; Thomas & Murray, 1983). Figs. 6, 10, and 16. One of the most obvious morphological characters, stem pubescence, proves problematic in P. velutina. The absence of pubescence on many collections from Oklahoma and Texas has led to repeated mis- identifications, as mentioned earlier. Maire (1963) considered glabrous- stemmed plants to represent a distinct variety, a concept not recognized by Ball and Heywood (1964). I agree it is best not to add a formal infraspecific name to the glabrous plants since this character does not correlate with differences in seed testa and, as Briquet (1910) observed in Corsica and I observed in Texas, glabrous- and glandular-stemmed plants may grow together. The consideration of P. velutina as a subspecies of P. prolifera by Bolos and Vigo (1974) is based on the treatment of Briquet (1910) and reflects the ideas of Malinvaud (1893), who suggested placing all members of the “prolifera” group within a single species, limiting “secondary units” to subspecies and varieties. Acceptance of sucha classification would minimize the importance of distinct karyotypic differences between P. prolifera and P. velutina shown by Bocher et al. (1955) and Thomas and Murray (1983) and aes) the breeding barriers isolating P. velwtina from all other taxa in section Kohlrauschia noted by Thomas and Murray (1981). Four common names have been used for P. velutina in North America, although in each case, the epithet prolifera is used in the accompanying scientific name. Howell (1962) used Childing Pink when describing plants from Butte County, California. St. John (1973) cited Tunic Flower in reference to plants in Hawaii, which would be introduced P. velutina according to Ball and Heywood (1964). Niehaus (1974) used Proliferous Pink, while Niehaus and Ripper (1976) preferred Wild Carnation as the common name for plants in the Sierra Nevada foothills. The epithet “velutina’” is an obvious reference to the “dense, glandular-tomentose indumentum” on most middle internodes of the typical P. velutina (Ball & Heywood, 1962). FLORAL BIOLOGY: Thomas and Murray (1981) reported that P. velutina is normally autogamous, finding that the timing of anthesis and stigma receptivity coincide and that the average lengths of filaments and styles are very similar. Some cross-pollination may occur if vectors are present since the stigmas do protrude above the corolla. ECOLOGY AND DISTRIBUTION: P. velutina is native to the Mediterranean region, with introduced populations found in Australia, Hawaii, and South Africa (Ball & Heywood, 1964). Two specimens have been seen from Chile (Junge 2636, US: Looser 4349, GH), indicating an introduction to South America. Except for two cultivated records, the North American distribution of P. velutina is restricted to northern Califor- nia (Fig. 17), southeastern Oklahoma, and eastern Texas (Fig. 18.). Nearly all P. velutina collections from Texas and Oklahoma came from roadside localities, with little evidence of invasion of adjacent communi- ties. P. velutina is also a roadside plant in northern California, although Frenkel (1970) did not list it as a “high presence species’ in his study of roadside vegetation. An inspection of collection labels revealed that P. velutina has left the roadside and has invaded at least two of the Woodland- Savanna communities described by Munz and Keck (1949, 1950); the Northern Oak Woodland and the Foothill Woodland as they occur in the North Coast ranges and the Sierra Nevada foothills. The presence of a Mediterranean climate (hot, dry summer; mild rainy winter) in California (Gleason and Cronquist, 1964) may be a positive factor in the expansion of P. velutina into oak and oak-pine woodlands, valley meadows, and stream banks. Flowering is noted from early April to early June. HISTORY OF INTRODUCTION: It is evident from both time and morphological considerations that P. velutina has been introduced to North 34 America at least twice (excluding known cultivations); once in northern California and once in eastern Texas. For this reason, a separate account will be presented for each region. CatirorNiaA: It appears that P. velwtima was introduced in northern California in the late 1920’s, with six collections noted in a 30 mile X 20 mile X 20 mile triangular area, including parts of NE Butte, SW Nevada, and NE Yuba counties beween 1927 and 1940 (Fig. 17). None of the collection labels indicated cultivated origin; two were gathered along roadsides, three from streamside areas, while the sixth was found on rocky hillocks west of a town. The literature is of little help on this point. Robbins (1940) listed only one location for P. velutina (T. prolifera), indirectly citing Yates’ collection ( Yates 36016 in 1933, UC) in Yuba County. Wolf (1938) summarized the situation when he wrote “we have recently found it in California [Wolf 8632 in 1937} where it apparently has been established for many years, but has been overlooked by collectors.” Texas: Most evidence points to a 1967 — 1968 introduction of P. velutina to eastern Texas. The only contradictory report is the listing of Dianthus prolifer for region 2, the Coastal Prairies, by Cory and Parks (1937). I have seen no specimen to document this statement, no recent collections of P. velutina from this area, and furthermore, Gould (1962) did not list the species as present in Texas. Shinners (1969) noted the “sudden appearance” of P. velutina (his P. prolifera) in May, 1968 along highways in east Texas which ‘I have traveled almost yearly for two decades without finding it.”” He postulated it could have been introduced as a result of state highway department planting of rye grass, Lolium perenne L. var. italicum (A. Braun) Parnell. Correll and Johnston (1970) concurred, noting “its very recent introduction into Texas, probably in contaminated rye grass seed.” On the labels of the earliest Texas collection seen (D. S. and H. B. Correll 35641; CM, LL), the Corrells suggested another vector when they wrote: “probably introduced with Italian clover seed by Texas Highway Dept.” Italian clover, Trifolium incarnatum L. (per Bailey et al., 1976), is listed by Turner (1959) as being “cultivated in the eastern part of the state, but occasionally escaping”’, his map showing its presence in eight counties. The Corrells’ suggestion may be the “correct”? source for several reasons. The Texas State Department of Highways and Public Transportation did spread T. incarnatum seed in the 1960's, the species “generally seeded throughout the eastern one-third of Texas, from Dallas eastward”, although the seeding plans have been revised ‘‘to the extent that Trifolium incarnatum is no longer specified for our current seeding practices.” (B. C. Blaschke, pers. comm.). 2 Trifolium fragments were present at the base of plants in four collections seen spanning four neighboring counties and three growing seasons. Final- ly, personal observations suggest T. incarnatum as the vector, the species being a dominant associate at each collection site, while Lo/zwm perenne was absent from four of the sites visited (Rabeler 351, 352, 353, and 356). N24 120 M16 aunty 38 : © @P. velutina ‘ epre-1960 . @1960+ tows i ® P.nanteuilii . = i jain ere ee RIVER Biot ts gb. ea J ent ag i 25° $0 190 SS ee = eee : ee ae aa | f ik . piEe° \ 2: 100 1 N | | 4 -_ seme o = = 116 FIG. 17. Distribution of Petrorhagia nanteuilii and Petrorhagia velutina in California. 36 CURRENT STATUS: Since the California and Texas-Oklahoma pop- ulations still appear as distinct entities with no range overlap yet documented, the status ot P. velwtina in these areas will be treated sepa- rately. Cauirornia: The list of specimens cited below includes the earliest and most recent collections seen from a given county. Once again, the floristic manuals present a vastly different picture from that of the specimens examined (also see Fig. 17). As noted previously, Robbins (1940) listed P. velutina for Yuba County. Munz and Keck (1959) described Tunica prolifera as “‘occasional as a weed reported from Butte, Nevada, Yuba cos.” Raven (1965) cited the same distribution in a note giving the correct name as P. velutina based on Ball and Heywood (1964) and a confirming identification of the California material by Heywood. Munz (1968) added Sacramento and Shasta Counties and changed the name to Koh/rauschia velutina {Howell (1972) adopted Petrorhagia velutina in his commentary on Munz (1968)}. Niehaus (1974) stated that T. prolifera (P. velutina) was “common as a pink mass under foothill oaks, Nevada Co. and north below 3000 ft."; a very accurate generalized description of the current status of P. velutina in California. He went on to note that this species is “rapidly spreading throughout the foothills”, a statement that is easily verified by arranging the specimens examined in chronological order by date of collection. The following pattern appears: pre-1939, 3 counties; 1940-1949, 1 added; 1950-59, 2 added; 1960-69, 6 added; 1970-on, 5 more counties added. This totals 17 counties, 15 of which are represented by collections dating from after 1960, 10 from 1970 or later. It is evident that this species is fully naturalized and probably still expanding into other areas of northern California | Texas: The earliest collection seen from a given county is included in the list of specimens cited below. Most specimens of P. ve/utina examined were collected between 1968 and 1971. During those years, it apparently spread rapidly, with eight additional counties added to the four counties repre- sented by the 1968 collections (Fig. 18). Correll and Johnston (1970) described it as ‘‘an extremely aggressive plant,” noting that it has spread to much of east Texas. My observations in 1980 confirmed that statement, with a large number of plants seen along the roadsides in eight counties during my brief visit. The stature and density of P. velutina at some sites clearly suggested a naturalized species. Collections made in three countries (Cherokee, Henderson, and Rains) represented new records, indicating that P. velutina was still invading additional sites. Two 1983 collections from southeastern Texas roadsides (Colorado and San Jacinto counties) and a 9/8 9\6 9/4 ° 25 50 75 @ 1968 OK BARK l dm 1969-79 tes >| crane? |i im 1KM o 25 50 1 198) RcumAN © 34 4 oO CLAY T™X ee MONTAGUE RED RIVER RAYSON ore : FANNIN BOWIE DELTA 5 THU E: 2 33 JACK WISE DENTON COLLIN HUNT HOPKINS. z @) = CASS TX ARK m LA 33 CAMP. i RAI O WA wood Neate ‘oo PARKER TARRANT DALLAS ° LO PINTO © © NOT ee Al KAUFMAN | VAN Z: © on ITH Late JOHNSON 7 ca ELUS s S PANOLA ERATH SOMER- HENDERSON © r 0 34 VEU 2 NAVARRO 8 mu © SHELBY coe ANDERSON — \_ CHEROKEE FREESTONE HAMILTON NACOGDOCHES o Fa MCLENNAN LIMESTONE ae HB f SABINE =— CORYELL LEON HOUSTON ANGELINA = FALLS Be TRINITY 18 96 9\4_\Newton FIG. 18. Distribution of Petrorhagia velutina in Oklahoma and Texas, 1968-1980. 1983 collections from Colorado and San Jacinto counties, Texas excluded. 1978 collection from a roadside in southeastern Oklahoma suggest that this species may be more widely distributed than either literature or collections currently indicate. Representative specimens: UNITED STATES. Catirornia. AMADOR Co.: Bank o Consummes River, S of Latrobe, 5.6 mi N of Ca 16 jct., 2 May 1974, ee 1456 hee Burre Co.: Bidwell Bar bridge, 10 mi E of Oroville, 10 May 1940, Cantelow 3078 (RSA); sandy grassland along CA 99, 11 mi N of Chico, 18 Apr 1978, Joslin 11 (CM), Bidwell Bar, 21 Apr 1929, Wilkens 5.n. (UC). CALAvERAS Co.: Wooded slope, 4 mi SE of Milton, 3 May 1963, Breedlove 4768 (SMU). Et Dorapo Co.: Open grassland, Folsom Lake State Park, 18 Apr 1964, Huether 34 (DAV); E shore of Folsom Lake, 5 mi NE of Folsom, 1 May 1960, Simonds 5.n. (CDA). HumBo.pt Co.: Gravel bank along Eel River floodplain, 20 Apr 1973, Anderson s.n. (HSC). Marin Co.: Along road to Black Point, 1.5 mi NE of Ignacio, 20 Apr 1971, True 6477 (CAS, JEPS, NY). Napa Co.: Valley clay, 6 mi N of Aetna Springs, Pope 38 Valley, 8 May 1948, Smiths.n. (DAV). Nevapa Co.: Roadside bank, Hwy 20, 7 May 1938, Heller 15077 (ILL, MO, NY, POM, UC); rocky roadside bank along CA 49, 2 mi N of North San Juan, 4 May 1963, Rice 176 (NCU). Placer Co.: Edge of beach, Folsom Reservoir, 2 mi N of Folsom, 4 May 1964, Clegg 27 (DAV); near old homestead, Orange- vale, 26 Apr 1976, Van Ess 3485 (SACT). PLumas Co.: Rocky slope near Murphy Creek along CA 70, 2.2 mi W of Belden, T25N, RGE, $26, SW1/4, 9 Aug 1980, Rabeler 514 (CHSC, GH, HHH, MSC, ROPA, UC). SacramMENTo Co.: 0.25 mi S of fish spel American River, 20 Apr 1963, Gustafson s.n. (SACT). Siasrs Com Dry grassland, 1.5m of Anderson, 12 Apr 1952, Cutright 32 (JEPS); oak-pine woodland, Co. Rd. AlG, 9.7 a NE of CA 36, T3ON, R8W, S33, 9 Apr 1978, Smith, Sawyer, and ne 9719 (HSC). SONOMA Co.: Grassy area, W slope of Sonoma Mtn, Osborn Nature Conservancy Preserve, E of Rohnert Park, May 1979, Serpa s.n. (MSC, OSH, ROPA). TeHama Co.: Along creek bank at rest area, I-5, 9 mi S of Shasta Co. line, T28N, R4W, $25, 26 May 1977, Halse 1585 (WTU); bank of Reeds Creek at Paskenta Road, 30 Apr 1967, Wheeler 20 (CHSC). Trinity Co.: Along CA 299 at Hayden Flat Campground, T5N, R7E, $24, 17 May 1975, Smith 8118 (HSC); grassy canyon slope, 3 mi SW of Douglas City, 22 Apr 1965, Weber 12284 (DAO, ILL). Yoto Co.: Cache Creek cafion, along CA 16, 2.5 mi NW of Rumsey, 22 Apr 1958, Bacigalupi, Mason, and Mason 6277 (JEPS). Yusa Co.: Oak woodland, Foothill Range Exper. Station, 19 Apr 1966, Carr 160 (MIN); flat W of North San Juan, 11 May 1927, Mason 3735 (JEPS, UC); edge of creek bed, 2 mi N of Brown’s Valley, 13 May 1937, Wolf 8632 ([CAS], {DS}, GH, [LA], NY, POM, RSA, UC, US); Spencerville Rd. at Indian Springs, TISN, R7E, S19, 23 May 1933, Yates 3616 (UC). OKLAHOMA. McCur- TAIN Co.: Along roadside near Yanubbee Creek, US 259, 2.5 mi N of Broken Bow, 30 Apr 1978, Taylor 25906 (DUR, LSU, MO). Texas. ANDERSON Co.: Roadside, TX 19, 12 miS of Athens, 22 May 1969, pag 3054 (NA). Cass Co.: Roadside, 3.2 mi SW of Avinger, 28 Apr 4970, Shinners 33022 {Southern Appalachian Botanical Club 2618} (FLAS, KNK, MASS, MICH, MSC, NLU, SMU, WVA). CHEROKEE Co.: Red — — along US 79, 3.2 mi W of New Suimimerheld. 2 May, 1980, Rabeler 346 (MIN, , NY, SMU). CoLorabo Co.: Along I-10, 1 mi W of FM 949, 15 May 1983, Brown Le (SMU). GREGG Co.: Roadside, US 259, Kilgore, 12 May 1969, Shinners 32631 (SMU). Harrison Co.: Roadside, W side of Hallsville, 13 May 1969, Shinners 32635 (MSC, SMU). HENDERSON Co.: In red sandy clay along E side of TX 19, 10 mi S of Athens, 1 May 1980, Rabeler 330 (GH, MSC, SMU, UC). Marion Co.: Sandy clay road shoulder, TX 49, 2.2 mi W of Jefferson, 13 May 1969, Shinners 32643 (MSC, SMU). Morris Co.: Re-graded road cut, TX 11, Daingerfield, 21 Apr 1969, Shinners 32597 (SMU, TENN). Rains Co.: Roadside, E side of US 69; 6.4 mi NW of Alba, 3 May 1980, Rabeler 353 (MSC, SMU). Rusk Co.: Roadside, 4.2 mi NW of Tatum, 5 May 1968, Shinners 32222 (FLAS, MASS, MSC, SMU, TENN, VDB). San Jacinto Co.: along TX 2025, 2 miS of TX 150, 27 May 1983, Brown 6136 (SMU). SmitrH Co.: Harris Creek Cemetery, W of Winona, 22 May 1971, Thomas 23207 (AUA, DUL, ILL, NLU, TENN). Tirus Co.: Roadside, TX 49, 7 mi E of Mt. Pleasant, 28 Apr 1971, Amerson 389 (SMU). Upsuur Co.: Road shoulder, 5.8 mi ESE of Big Sandy, 7 May 1968, Shinners 32233 (MASS, SMU); grassy roadside, TX 155, 5 mi W of Ore City, | May 1969, Correll 37154 (FSU, GH, LL, MICH, NA, NCU, NY, OKLA, TEX). VAN ZANDT Co.: Roadside, 2.8 mi E of Grand Saline, 7 May 1968, Shinners 32238 (MASS, SMU). Woop Co.: Near pond along Farm Rte 514, 4 mi Eof Yantis, 25 Apr 1968, D. S. and H. B. Correll 35641 (CM, LL). Cultivations: MARYLAND. PRINCE GeorGes Co.: In Glenn Dale Introduction Garden (seed from Turkey), 23 June 1938, Cowgill 808 (BH, GH, MICH, NA). PENNSYLVANIA. PHILADELPHIA Co.: cultivated, Mehans Garden, without date, Burk s.n. (PENN). ACKNOWLEDGEMENTS I thank Drs. John H. Beaman, Susan R. Kephart, and the late William T. Gillis for assistance and guidance provided during this study; Darwin Dale for photographs (Figs. 3 — 6) taken with his scanning optical micros- cope; Roy E. Gereau, Drs. Neil Harriman, Norton Miller, and Edwin Smith for assistance in procuring literature; Dr. Gary R. Hooper, former director of the Electron Optics Center, Michigan State University, for permitting my use of those facilities for production of the seed photographs shown as Figs. 1 and 2; Dr. John McNeill for his helpful comments on the lectotypification of Petrorhagia and other nomenclatural matters; Dr. David Rembert for information on Thomas Walter; Dr. P. Mick Richardson for acquainting me with the work of Dr. Thomas, Dr. John Thieret for reviewing the manuscript; and Dr. Sandra M. Thomas for graciously providing information and reprints from her study of Petrorhagza 1n Europe. Information and/or recent collections of Petrorhagia were provided by Dr. Nancy C. Coile (P. prolifera in Georgia), Barney Lipscomb (P. velutina in Texas), Dr. Peter M. Mazzeo) P. saxifraga in Virginia), Nick Stoynoff (Petrorhagia at MOR), Dr. R. John Taylor (P. velutina in Oklahoma), and Dr. James R. Wells (P. prolifera and P. saxifraga in Michigan). Several individuals deserve recognition for contributing information that enhanced my field efforts: Drs. Charles F. Quibell and Robert A. Schlising in California; Dr. John T. Atwood, Sue Crispin, Stu Ouwinga, and Dr. Anton A. Reznicek in Michigan. Rex Boner of the Nature Conservancy granted permission to collect P. prol#fera at the Kitchel Dune preserve in Michigan. Thanks are also expressed to the curators of the following herbaria for the loan of specimens used in this study: A, AC, AUA, BH, BHSC, CAN, CAS, CDA, CHSC, CM, CU, DAO, DAV, DUKE, DUR, DWC, EGV, F, FARM, FLAS, FSU, GH, HSC, ILL, IND, JEPS, KBSMS, KNK, LAF, LL, LSU, MASS, MIN, MO, MTMG, NA, NCSC, NCU, ND, NDG, NEBC, NHA, NLU, NO, NY, OKL, OKLA, OS, PENN, PH, POM, ROPA, RSA, SACT, SMU, TAES, TENN, TEX, UARK, UC, UMBS, UMO, UNA, US, USF, VDB, VPI, WIS, WMU, WUD, and WVA. I appreciated the hospitality of the curators of A, BH, BLH, CM, DUL, F, GH, HCHM, HHH, MCTF, MICH, MO, NA, NY, NYS, PAC, UBC, US, UWSP, VA, and WTU during my visits there. Library personnel at GH, MICH, MO, MSC, NY, PH, US, the National Agricultural Library, and the Library of Congress were of great assistance while literature was being procured. Partial support of photographic preparation was obtained from the 40 William T. Gillis Memorial Fund, while some travel funds were obtained from Lyman Briggs School, Michigan State University. REFERENCES BAILEY, L. H., E. Z. BAILEY, and STAFF of the L. H. oe Hortorium. 1976. Hortus third. New Vole Macmillan Publ. Co. xv + 1290 p BALL, P. W. and V. H. HEYWOOD. 1962. Taxonomic pain of the cytological races of Kohlrauschia prolifera (L.) Kunth sensu lato. Watsonia 5:113-116. Recereee eet .H. HEYWOOD. i A revision of the genus Pe ee Bull. Bric. Mus. (Nat. Hist.), Bot. 3:119— BEAL, W. J. 1908. i. to the oe flora. Rep. Michigan Acad. Sci. 10:85 — 89. BERKELEY, E. and D. 8. BERKELEY. 1982. The life and travels of John Bartram: from Lake eae to the ae = John. eyes Univ. Presses of Florida. xvi + 376p BOCHER, T. W., K. LARSEN, and K. HN. 1953. Experimental and ceolooical ar on — species. I, ne ot rin ts and Plantago coronopus. Hereditas 201282 = 5 = LARSEN, and K. RAHN. 1955. Experimental and cytological studies on plant ae II. Trifolium arvense and some other pauciennal herbs. Biol. Skr. 8(3):1— 31. BOLOS, O. DE and J. VIGO. 1974. Notes sobre taxonomia i nomenclatura de plantes, I. Bucl. Inst. Catalana Hist. Nat., 38 Bot. 1:61 — BORGEN, L. 1974. Caromocome numbers of some in flowering plants II. Norwegian J. Bot. 21:195 — BRAUN, E. L. 1943. An annotated catalog of Spermatophytes of Kentucky. Cincinnati: John S. Swift Co., Inc. BRIQUET, J. 1910. Biodome de ia Flore Corse. Lyon: Georg & Cie. 1:lvi + 656 pp. BRITTEN, J. 1921. Thomas Walter (1740? — 88) and his grass. J. Bot. 59:69—74. BRITTON, N. L. 1879. Staten Island plants. Bull. Torrey Bot. Club 6:259 — 260. 1897. Caryophyllaceae. IN: N. L. Britton and A. Brown. An Illustrated Flora of the northern United States, Canada and the British Possessions. New York: Charles Scribner’s Sons. 2:6—41 tities A ED, Conaphy lines: IN: N. L. Britton and A. Brown. An illustrated flora of the Motheia United States, Canada and the British Possessions, 2nd ed. New York: Charles Scribner's Sons. 2:61 — BRULLO, S. and F. FURNARI. 1979. Taxonomic and nomenclatural notes on the Flora of Cyrenaica (Libya). Webbia 34:155 — 174. BURNAT, E. 1892. Flore des Alpes Maritimes. Lyon: H. Georg. 1: xii + 302 pp. CANDAU, P. 1980. erence en ae aaa del sur de Espana. -subfamilia Sile- noideae. Lagascalia 9:1 147 2 pl COLE, E, J. 1901. a apd or A catalogue of the flowering plants and ferns. Grand Rapids, MI: A. Van Dort. 170 CONARD, H. S. 1943. ae of Iowa. ee Flora, 6th ed. Grinnell: Grinnell College. >Ps CORE, E. L. 1941. Notes on some West Virginia plants. Castanea 6:86 — CORRELL, D. S. and M. C. JOHNSTON. 1970. Manual of the vascular ae ants a Texas. Contr. Texas Res. Found., Boc. Stud. 6: xv + 1881 41 CORY, V. L. and H. B. PARKS. 1937. Catalogue of the flora of the State of Texas. Texas Agric. Exp. Sta. Bull. 550:1— 130. CUSICK, A. W. and G. M. ae i a vascular plants of unglaciated Ohio. Ohio Biol Surv. Bull. (n.s.) 5 DALE, D. 1982. Scanning ee ar enone 17.) 218' = 21 DAMBOLDT, J. and D. PHITOS. 1972. ae zur Flora Aaa IV. Studien in der Gatctung eas eee Candollea 27:27 — DANDY, J. E. 1957. Some new names in the British flora. eee 4:42. ene A. 1937, Flora Velebitica. patent Akademie der Wissenschaften. 2: i + 667 DOLE, E. J., ed. 1937. The Flora of Vermont, 3rd rev. ed. Burlington: Free Press Printing Co. (Vt. Botanical Club). xiv + 353 p FARR, E. R., J. A. LEUSSINK, and F. A. STABLEU. 1979. Index Nominum Genericor- um (Plantarum). Regnum Veg. 100— 102: xxvi + 1— 1896. FAVARGER, C. 1966. Contribution a ei oe du genre Petrorhagia (= Tunica). Ber. Schweiz. Bot. Ges. 76:270 FERNALD, M. L. 1950. Gray’ s manual - Borday. 8th ed. New York: American Book Co. Ixiv + 1632 pp. FERNANDES, A. and M. T. LEITAO. 1971. Contribution a la connaissance cytotaxino- mique des Spermatophyta du Portugal. III. Caryophyllaceae. Bol. Soc. Brot. 45 (ser. 2):143 — 176. FOX, W. a _K. GODFREY, and H. L. BLOMQUIST. 1950. Notes on distribution of North Carolina plants - II. Rhodora 52:253 — 271. FRENKEL, R. E. 1970. Ruderal vegetation along some California roadsides. Univ. Calif. Publ. Geog. 20: vii + 1-163. GALIANO, E. F. and S. SILVESTRE. 1977. Catalogo de las plantas vasculares de la provincia de Cadiz. III. Centrospermae: Caryophyllaceae. Lagascalia 7:13 — 45. GAMISANS, J. 1974. Contribution a l'étude de la flore de la Corse. VI. Candollea 29s GLEASON, H. A. and A. CRONQUIST. 1963. manual of vascular plants of Northeastern United States and ae Canada. New York: Van Nostrad Reinhold Co. li + 810 pp. and A. CRONQUIST. 1964. The natural geography of plants. New York: Columbia Univ. Press. 420 pp. GOULD, F. W. 1962. Texas ae A checklist and ecological summary. Texas Agric. Ext. Serv. Misc. Pub. 585:1— 11 GRAEBNER, {K. O. R. P.]} P. oe P. GRAEBNER. 1921. Tunica. IN: P. Ascherson and P. Graebner. Synopsis der Mitteleuropaischen Flora. Leipzig: Borntraeger Bros. 5(2):260 — 275. GREUTER, W. and P. MOUTERDE. 1970. Petrorhagia syriaca (Caryophyllaceae): une rehabilitation. Candollea 25:221— 227. es Pe eee, T. RAUS, eds. 1982. Med-checklist notulae, 6. Willdenowia | Pu Kaleem bog and T. RAUS, eds. 1984. Med-checklist notulae, 9. Willdenowia 14:37 — 54. HARSHBERGER. J. W. 1899. The botanists of Philadelphia and their work. Phi- ladelphia: T. C. Davis & ae xii + 457 pp. HARVILL, A. M., JR., T. BRADLEY, and C. E. STEVENS. 1981. Atlas of the Virginia flora, part II. ae Virginia Botanical Associates. v + 61— 148 pp HERMANN, F. J. 1946. A checklist of plants in the Washington — Baltimore Area, 2nd ed. Washington: Conference on District Flora. 130 pp. 42 HOLMGREN, P. K., W. KEUKEN, and E. K. SCHOFIELD. 1981. Index herbariorum, part I, 7th ed. Regnum Veg. 106: vii + 1-452. HOLUB, J.,J.M ESICEK, and V. JAV URKOVA. 1972. Annotated chromosome counts of Czechoslovak plants (3 1 — 60) (Materials for “Flora CSSR” — 3). Folia Geobot. Phytotax. pe 72167 ~ 202 HO R, J. D. and B. D. JACKSON. 1895. Index kewensis. Oxford: Clarendon Press. ne pp; 2:1299 pp. HOWELL, J. T. 1962. Five days to Reno: a botanical mortorlogue. Leafl. W. Bot. 9:233 — 242 972. Miscellaneous notes on Munz’ a California flora and its supplement. Wasmann J. Biol. 30:97 — 107. HUBER- eo A. 1977. Weitere Erginzungen zur Flora der Tirkei. Bauhinia 6:9: cts . 1965. Nomenklatorische Bemerkungen zur Flora Europaea, Vol. 1. Feddes KNUTH, P. E. O. W. 1908. Handbook of a pollination, trans. by J. R. A. Davis. Oxford: Clarendon Press. 2: viii + 703 KUNTH, C. S. 1838. Flora Berolinensis. Berlin: Vendunt Duncker et Humblot. vii + 407 ct Pp. LAWRENCE, G. H. M. 1951. Taxonomy of vascular plants. New York: The Macmillan Co. xtil . 823 pp. LINK, H. F. 1831. Handbuch zur Erkennung der nutzbarsten und am haufigsten vorkom- men ane Gasnichse. Berlin: S. J. Joseephy. 2: iii + 533 pp. LINNAEUS, C. 1753. Species Plantarum. Holmiae. 1: xii + 560 pp. MAGUIRE, B. 1950. Studies in the Caryophyllaceae —1V. Rhodora 52:233 — 245, 1952. Caryophyllaceae. IN: H. A. Gleason. The new Britton and Brown rileoceiced flora of the Northeastern United Scie and adjacent Canada. New York: New York Botanical Garden. 2:118— 145. MAIRE, R. 1963. Flore de I Afrique du Nord, vol. 10. Encycl. Biol. 60:1—300 and 6271336, MALINVAUD, E. 1893. Un Dianthus nouveau pour la flore de I'Herault. Bull. Soc. Bot. Fr, 40:298 — 299. MCATEE, W. L. 1940. Eighth een to the flora of the District of Columbia and A Jos Ge haa be vicinity. Proc. Biol. Soc. Was 1: ; MEIKLE, R. D. 1977. Loh Cays Kew: The Bentham-Moxon Trust, Royal Botanic ardens. 1: xii + 832 MERRILL, E. D. 1938. \ critical consideration of Houttuyn’s new genera and new species Hany 1773 — 1783. J. Arnold ene 1a 20 37S. MERTENS, F. C. and W. D. J. KOCH. 1831. J. C. Rohlings Deutschlands Flora, 3rd ed. PP- MEUSEL, H. and H. MUHLBERG. 1979. Petrorhagia (Ser.) Link. . H. Rechinger. [lusctrierte a von Mitteleuropa, 2nd ed. Berlin- anette Paul Parey. 3(2):978 MOENCH, Ri ot Methodus plantas. Marburg. vii + 780 + 18 pp. MUNZ, P. A. 1968. A supplement to a California flora. Berkeley: Univ. of Calif. Press. 224 pp. and D. D. KECK. 1949. California plant communities. Aliso 2:87 — 105. —______ and D._ D. KECK. 1950. California plant communities - supplement. Aliso 2:199 — 202. 43 and D. D. KECK. 1959. A California flora. Berkeley: Univ. of Calif. Press. 1681 pp. NIEHAUS, T. F. 1974. Sierra wildflowers, Mt. Lassen to Kern eee California Natural History Bae a Berkeley: Univ. of Calif. Press. 223 p dC. L. RIPPER. 1976. A field guide to Pacific state wildflowers. Boston: Houghion ifn ae Peterson Field Guide Series 22: xxxii + 432 pp. PEREZ LA i . 1896. Florula gaditana, part 5. Anales Soc. Esp. Hist. Nat. 24/3 i 3 PERRING, F. and C. FARRELL. 1977. British red data books: 1. Vascular plants. London: The Society for the Promotion of Nature Conservation. xxvi + pp. PHILLIPS, C. E. 1978. Wildflowers of Delaware and the eastern shore. Hockessin: Delaware Nature Education Society. x1 + 303 pp. PHITOS, D. 1966. Drei neue Arten aus Griechenland. Oesterr. Bor. Z. 113:271— 274. PO on R, dN. P. 5. Notes on the flora of Worcester County, Meeahacae Rhodora 37:8 PRINGLE, J. S. 1976. Crp sercmeli perl a a naturalized species in the Great Lakes Region. Mic 2D: RABELER, R. K. 1980. ee din ee species in Michigan. Michigan ot. 19:83 — 88 «1981. Bene: muralis. Is it naturalized in Michigan? Michigan Bot. 20:21 — 26. _________. 1984. Notes on nomenclature within Petrorhagia (Caryophyllaceae). Taxon 33:714— 6. —_____. 1985. Isonyms of Petrorhagia ae (Caryophyllaceae). Taxon 34. in press. RADFORD, A. E., H. E. AHLES, and C. R. BELL. 1965. Atlas of the vascular flora of the Carolinas. ee Carolina Agric. Exp. Sta. aan Bull. 165:1— 208. __________, H. E. AHLES, and C. R. BELL. 1968. Manual of the vascular flora of the mle eee Hill: Univ. of North Carolina Press. Ixi + 1183 p RAVEN, P. H. 1965. On the status of “Tunica prolifera” in California. Teafl W. Bot. 10:179. RICHTER, K. 1903. aie Europaeae, ed. [R. L. A.} M. Gurke. Leipzig: Wilhelm ee 2(3):321 — 480. ee H. 1930. a dispersal of plants throughout the world. London: L. Reeve & O., ies XX ee 744 pp ROS W. W. 1940. Alien Ze ane without cultivation in California. Univ. Calif. Agric. Ae Stat. Bull. 637: N ROBINSON, B. L. 1897. See In: B. L. Robinson, ed. Synoptical Flora of North ee ae ork: American book Co. Haye men ee and M FERNALD. 1908. Gray’s new m tany, 7th ed. New York: pena aie . 926 p ST. nie ‘1973. List 9 summary of the flowering plants in the Hawaiian Islands. . Tro . Gard. Mem. 1:1—519. SAVAGE, a coed A catalogue of the Linnaean Herbarium. London: Linnean Society of ondon. xv + 225 ScuaeER P. 1936. Beitrage zur mikroskopischen ee der wichtigsten Caryophyllaceensamen. Landw. Versuchsstat. 125:1— 1 SCHLISING, R. A. and H. H. ILTIS. ae ier reports on the flora of Wisconsin. No. 46. Caryophyllaceae - Pink F . Trans. Wisconsin Acad. Sci. 50:89 — 139. SCHMALTZ, T. C. 1981. eae cedies: on roadside vegetation in southwestern 44 Virginia. Ph.D. dissertation, Virginia Polytechnic Inst. and State Univ., Blacksburg. vi 145 pp. SCHUBERT, B. C. 1946-47. The herbarium of Thomas Walter at the British Museum. Rea bound at Harvard University. 119 pp. rae J. A. 1772. Flora Carniolica, 2nd ed. 1972 oe Graz, Austria: Akademis- he Dr na -v. hse Seguin [xxxiti}] + Ixviii + 448 San 1965. aed a. (Cayopy ce) in the United Scates: population eal eee and fractionated suici ida 2:119 — 128. ts, 1969, aS pai ( Daan prolifer, Tunica prolifera) (Caryophyl- laceae) in Arkansas and Texas. Sids 346. SHREVE, F. 1910. The nine lie Maryland List of plants collected and observed. Maryland Weather Serv. Spe — 497. SMITH, A. W. 1972. a a a ee of plant names, rev. ed. New York: Sc. Martin’s Press. xii + 39 SOO, R. DE. 1970. A Magyar Plo és Vegetacié Rendszertani Novényfoldrajzi Kézikony- ve. Bucharest: Akademiai Klado. 4:614 pp. SPRAGUE, T. A. 1927. Tunica Mert. & Koch. J. Bot. 65:225 — 227. STAFLEU, F. A. and R.S. COWAN. aie — 1981, 1983. Taxonomic literature, ser ed., vols. 1-4. Regnum Veg. 94: x1 + 1— 1136; 98: xviii + 1-991; 105: xii oa ‘978: 110: ix + 1= THOMAS, S. M. 1980. Population studies and delimitation of species in the genus Petrorhagia, (Section Koblrauschia). Ph.D. thesis, Westfield College, Univ. of London. 281 pp E Dajpavhags ion Kohl, hia(Caranhyl- } | useage ee ee ic clarif laceae). : Linn. Soc., Bot. 87:55 — MURRAY. 1981. Breeding systems and a aa in Petrorhagia sect. Ga (Cochise). Pl. Syst. Evol. 139:7 eee! . G. MURRAY. 1983. Chromosome ce in coe as hybrids of bases sect. a Nake St anaeaee Pl. Syst. Evol. 141: ae Pe TORREY, J. and A. GRAY. . A flora of North America. New ee Wiley and TURNER, B. L. 1959. The legumes of Texas. Austin: Univ. of Texas Press. xi + 284 pp. TUTIN, . G., ET AL., eds. 1964, 1976, 1980. Flora Europaea. Cambridge: Cambridge Univ. Press. 1: xxxiii + 1—464; 4:431—446; 5:375 — 391. VAN DER PIJL, L. es eae of Dispersal in Higher Plants, 3rd ed. New York: Springer- ee x VOSS, E. G., H. M. cana. Ww. G. CHALONER, V. DEMOULIN, P. HIEPKO, J. MCNEILL, es D. MEIKLE, D. H. NICOLSON, R. C. ROLLINS, P. C. SILVA, and W. GRUETER, eds. i International code of botanical nomenclature. Regnum Veo. Tiare t= WALTER, T. 1788 a Flow Caroliniana. Cambridge, MA: photolith of original, Murray Printing Co. vii + 263 pp WILLIAMS, F. N. 1893. A mencorch of the genus Dianthus Linn. J. Linn. Soc., Bot. 29:346 — 478. WOFFORD, B. E., D. H. WEBB, and W. M. DENNIS. 1977. State records and other recent noteworthy collections of Tennessee plants II. Castanea 42:190 — 193. WOLF, C hee California plant notes. II. Occas. Pap. Rancho Santa Ana Bot. Gard. Ser. 1(2) PORTULACA PILOSA L., P. MUNDULA I. M. JOHNST. AND P. PARVULA GRAY IN THE SOUTHWEST JAMES F. MATTHEWS AND PATRICIA A. LEVINS Herbarium, Department of Biology The University of North Carolina Charlotte, NC 28223, U.S.A. ABSTRACT Historically the taxonomic interpretation of three species of Portulaca—P. pilosa L., P. mundula\.M. Johnst., P. ee oy has been unsettled. The basis for each interpretation has been reevaluated using ma collections from both the Southeast and the Southwest United States. oe characters traditionally used in descriptions and keys, along with an examination of the seed surfaces by scanning electron microscopy, have provided evidence that P. mundula is conspecific with P. pilosa. The interpretation of P parvula is still uncertain due to lack of study material. Theoretical patterns of the migration of Portulaca are presented. TAXONOMIC HISTORY In 1753 Linnaeus described Portulaca pilosa as having alternate, subulate leaves, with axillary hairs and sessile flowers. He did not mention flower color, but cited publications by Herman and Commelin as treating the same red-purple flowered species. In 1887 Asa Gray published the name Portulaca parvula, describing the species as having yellow to copper-colored petals. He noted that P. parvula in Texas, New Mexico and Mexico was part of several specimens labeled P. pilosa which were collected by Wright, Fendler, Schaffer (772) and Pringle (543). From Gray’s comments it is apparent that he was treating mixed collections and segregating the yellow-flowered taxon from the red- flowered P. pilosa. He gave the distribution of P. pilosa as Florida to Arizona. Gray’s Manual of Botany, ed. 7 (Robinson and Fernald, 1908) recog- nized P. pilosa as the red-flowered taxon in the Southwest, with no mention of a yellow-flowered taxon. Ivan M. Johnston (1948) examined the mixed specimens that Gray had cited. In his judgment, the red-flowered speci- mens constituted a new species, P. mundula 1:M. Johnst. Johnston com- mented that an examination of illustrations and comments from Commelin (1697) and Herman (1705) left little doubt as to the Linnean concept of P. pilosa. Hence, in his opinion the red-flowered specimens were not P. pilosa. SIDA 11(1): 45—61. 1985. 46 Gray's manual, ed. 8 (Fernald, 1950) recognized P. parvila as the only southwestern oe including in the description both red and yellow flowers, and noting that the treatment in ed. 7 was not P. p/osa of Linnaeus. Shinners (1958) treated P, mundula as a synonym of P. pilosa and commented that the yellow-flowered species, P. parvula, occurred to the west and south of the Dallas-Fort Worth area. Legrand (1962) in his monograph of the American species of Portulaca treated P. mundula as a variety of P. pilosa. He noted that var. mundula differed from the typical variety by a reduction in size of all its organs. The distribution of var. mundula is given by Legrand as from the tropics to northern Mexico (Chihuahua and Coahuila), the southwest United States into Colorado, Texas, Oklahoma, Kansas and Missouri. Legrand commented that there was clearly a transition of morphological traits from var. pilosa to var. mundula trom the tropics to the northern arid habitats. Legrand treated the yellow-flowered P. parvula as part of P. halimoides L. He stated that P. parvila Gray sensu Johnston represented the depauperate form of P. hali- modes resulting from the adaptation to desert regions. Correll and Johnston (1970) modified their treatment of Portulaca from Legrand but recognized P. mundula and P. parvula as species, without recognizing P. pilosa. A NEW EXAMINATION In the course of preparing the treatment of Portulaca for the Vascular Flora of the Southeastern United States it was necessary to decide whose treatment was the most accurate. Additionally, since the Southeast extends into Arkansas and Louisiana, it became necessary to examine specimens from herbaria in the Southwest. Portulaca mundula is reported from Arkan- sas and some specimens from Louisiana have been identified as P. mundula. Several problems emerged: |. Does P. p//osa occur in the southwest? 2. Can P. mundula de differentiated from P. pilosa? 3. What does the name P. parvula represent? To address these problems over 1,800 specimens were borrowed from the following herbaria: ALU, ASTC, DUR, FLAS, FSU, G, GA, JSU, KNK, KY, LAF, LL, MO, NCU, NLU, NO, NY, SMS, SMU, TENN, TEX, UARK, UNA, UNCC, US, USAM, USCH, USF, VDB, VPI, VSC, WILLI. Examination of the specimens both for distribution and morphological variability indicated problems in separating P. mundula from P. pilosa. Since no taxonomic treatments have compared all three taxa, including P. parvida, with analyses of the distinguishing character states separating them, an attempt was made to do so, using the characters provided by I.M. 47 Johnston, Legrand, and Correll and Johnston. Comparison of all of the morphological traits used to define the species showed that the characters of capsule diameter, capsule pedicel length, seed diameter, color and surface texture, and flower color seemed to provide the clearest traits for identifica- tion. These literature data are shown in Table 1. Additionally the Southeastern treatments have been limited to state boundaries without consideration of variation throughout the range. Speci- mens were examined by state, beginning in Florida because P. pilosa is the original taxon to which to compare any later segregates. Portulaca pilosa has been unchallenged as the Florida taxon in treatments by Small (1933), Legrand (1962), Long and Lakela (197 1) and Wunderlin (1982). Legrand, Taste 1. Key morphological features for the three taxa, taken from the literature. Multiple entries are character states given by I.M. Johnston, Legrand, and Correll and Johnston respectively. Lack of a statement is shown by ----. Capsule diameter ---- 2.5-3.5 1.5-2.0 in mm 2.5-4.3 2.0-3.0 ---- ---- 2.5-3.5 1.5-2.0 Pedicel length ---- Short stipitate 1.0-1.5 in mm Up to 1.0 Up to 1.0 ---- ---- Lightly scipitate 1.0-1.5 Seed diameter ---- 0.3-0.5 0.3-0.5 in mm 0.5-0.65 0.5-0.6 ---- ---- 0.3-0.5 0.5 Seed color ---- Black Black Black Black peas Black Black Seed surface ---- Stellate-tuberculate Stellate flattened roughenings Small tubercles Stellate-tuberculate - dorsally, stellate ---- Stellate-cuberculate Stellate flattened roughenings Flower color ---- Purple Yel-orange, bronze Purple — Red-purple Yel-orange, bronze 48 being very familiar with the Central and South American taxa, cited specimens which he considered as P. pilosa from Florida. We have accepted Legrand’s conclusion that P. pz/osa is the correct name for the Florida taxon. Specimens were selected from counties throughout Florida to provide a basis of variability for P. pz/osa. Observations of the same character states, as shown in Table 1, were made. If two specimens from the same county appeared morphologically dissimilar, both were included. Thirty eight specimens were measured or scored, with three observations on each speci- men, for the traits of capsule diameter, capsule pedicel length and seed surface features. Following this, the same observations were made on all specimens from Alabama (n= 6), Mississippi (n= 16), Louisiana (n = 57). Specimens that could not be positively identified as having red flowers were excluded. Since Texas represents a large diversity of regions and habitats, capsule and pedicel measurements were grouped according to distributions in the coastal plain (CP), in the high prairie (HP) and in the Trans-Pecos (TP). Additionally, 23 specimens from Missouri, Oklahoma and Texas, cited by Johnston in his original description of P. mundula were measured and compared. These paratypes, as well as the populations and sub- populations defined above are compared in Table 2. The ANOVA (Table 2) for pedicel length for the seven states, with Texas segregated into three subpopulations, including Jonnston’s paratypes show no significant differences between any two populations (unplanned com- parisons using the Tukey-Kramer procedure, Sokal and Rohlf, 1981). Some interesting relationships are evident. The specimens from the high prairie (HP) of Texas are separate from the other Texas subpopulations. The other Texas subpopulations (CP & TP) are grouped with the Florida population while Johnston’s paratypes are intermediate. For the capsule comparisons, using the same geographical grouping, an ANOVA (Table 2) with unplanned comparisons using the Tukey-Kramer procedure shows the following: Capsule diameters of plants from Arkansas, Oklahoma and Johnston’s paratypes are significantly smaller (p <.05) than the capsule diameters of plants from Texas, Mississippi, Florida and Ala- bama. Plants from Louisiana are intermediate between the two groups. However none of the Texas subpopulations are significantly different from those of Florida. Since there is no significant difference among the nine geographical populations and the paratypes in the pedicel comparisons, and since the two groups in the capsule comparisons are of mixed geographical arrangement, the variability of the taxonomic characters traditionally used cannot be predictably segregated to represent two taxa. A comparison of capsules vs pedicels shows a strong positive correlation (r = 0.310, df = 174, p«.01); 49 Taste 2. Comparison of pedicel lengths and capsule diameters for Alabama (AL), Arkansas (AR), Florida (FL), Louisiana (LA), Mississippi (MS), Oklahoma (OK), Texas coastal plain (TX-CP), Texas high prairie (TX-HP), Texas Trans-Pecos (TX-TP) ‘si Johnston's Paratypes (JP). Means connected by lines are not significantly different at the .05 lev Pop.: AR TX-HP LA JP FL TX-TP TX-CP AL Mean 0.363 0.429 0.441 0. a 0. a 0.539 0.595 0.605 0.613 0.617 SE 0.077 0.053 0.053 0.056 0.053 0.046 0.035 0.050 0.055 0.089 N 8 17 ce i ia 2 38 19 16 6 Pop.: AR OK JP. LA TX-HP TX-TP TX-CP Mean 1.875 1.924 1.965 2.212 2.335 2.426 ae ee 2.600 2.619 SE 0.112 0.077 0.066 0.077. 0.077. 0.073 0.082 0.052 0.130 0.080 N 8 17 2307 17 19 15 38 but these characters do not vary together. Comparison of the averages shown in Table 2 with the ranges listed in Table 1 show that the southeastern United States specimens (Florida, Alabama, Mississippi, Louisiana) fall ac the lower range for capsule dia- meter and pedicel length as given by Legrand for P. pilosa throughout its range. This is not surprising since most of Legrand’s measurements were made on more tropical, hence more robust specimens. His studies centered on Central and South American specimens and Portulaca is primarily a genus of these areas. He did however examine specimens from Florida, Georgia and Mississippi to develop his concept of P. pilosa for the United States. Seed surface texture, Table 1, has been used as a character for species separation. In fact, many specimens lacking petals have been identified using the seed surface texture character. Rep ives of the seed surface patterns for known red-flowered specimens (oni seven states are grouped to illustrate the variability over the entire geographic range, Figs. 1— 22. Original labels were either P. pilosa, P. mundula, or P. parvula. The bar on each figure represents 100 jum. Seeds were sputter coated with gold- palladium ona Hummer V and viewed ona Jeol JSM-35CF SEM. Specimen citations from which seeds were taken are shown elsewhere. Froripa—Thirty eight specimens had stellate-tuberculate surfaces; the tubercles varied from very short to medium, none were flattened (Bigs. — 3): ALABAMA—Six specimens had stellate-tuberculate surfaces; the tuber- cles varied from very short to short (Fig. 4) 50 Mississippi—Nine specimens had stellate-tuberculate surfaces; five had stellate flattened roughenings and two had seeds in which some were tuberculate and some were flattened (Fig. 5). Louis!ANA—Twelve specimens had stellate-tuberculate surfaces (Fig. 6), five had stellate flattened roughenings (Fig. 7). At this location in the geography, the variability of stellate flattened roughenings in northern Louisiana is evident. ARKANSAS—AI] 20 specimens had stellate flatrened roughenings (Fig. 8). This pattern complements that of northern Louisiana. Texas—Twenty eight specimens scattered over the state had stellate- tuberculate surfaces, 23 had stellate flattened roughenings. Texas specimens exhibited the greatest range of diversity. Figures 9 — 12 present the coastal plain variability. Figure 17 shows the same stellate-tuberculate pattern farther inland. Figures 18 — 21 present the range of variability in the Trans-Pecos region, from stellate flattened to highly stellate-tuberculate. OKLAHOMA—Figures 13 — 16 & 22 show the same pattern found in the high prairie of Texas with none to only slight impressions of the stellate-tuberculate pattern. Figure 22 illustrates the extreme flatten- ing in the western populations. To obtain an overall perspective of variability in seed surface texture over the western geographical area, a three X five inch scale map of Louisiana, s 1—24. Scanning electron micrographs of seeds of red-flowered portulacas, originally labeled ee at P. anise or P. parvula.1. Florida. Dade Co.: | Au 1940, Arnold s.n. (FLAS). 2. Florida. Cicru - 13 Aug 1958, Kral 7825 (NY). 3. Florida. Escambia Co.: 8 Aug 1981, paras ae ao 4. Alabama. Mobile Co.: 20 Aug 1968, Kral 32710 (GAD. 5. Mississippi. Harrison Co.: 6 Jul 1952, Demaree 32138 (SMU). 6. Louisiana. Allen Co.: 23 Oct 1964, Thieret 18690 (FSU). 7. era Morehouse Co.: 27 Jul 1977, Thomas & Pias 54274 (NLU). 8. Arkansas. White : 14 Oct 1974, lebe 64262 (GA). 9. Texas. Hardin Co.: 27 Aug 1970, Ammerson [ Water 244 (SM Texas. Refugio Co.: 13 Oct 1956, Shinners 25232 (SMU). 11. Texas. Cameron Co.: > vie 1959, Traverse 1046 (G). 12. Texas. Austin Co.: 12 Oct 1971, Shee eee (NLU). 13. Oklahoma. Bryan Co.: 30 Sep 1974 ee ag (DUR). 14. Oklahoma. Oklahoma Co.: 7 Jul 1976, Taylor aise (NLU). 15. Okla . Harmon Co.: 26 Aug 1948, ee 8716 (G). 16. Oklahoma. Woods Co.: 5 Oct ce Stevens ee 17. Texas. Frio Co.: 24 Jul 1941, Tharp s.n. ees 18. Texas. Brewster Co.: 18 Jul 1936, Warnock 121 (SMU). 19. Texas. Brewster Co.: 23 Aug 1970, ee 408 (MO). 20. Texas. Culberson Co.: 21 Jul 1943, Waterfall 5242 (G). 21. Texas. Ward Co.: 18 Sep 1966, Correll 33652 (NCU). 22. Oklahoma. Cimarron Co.: 11 ee LOTT, ale 25255 (NLU). 23. Oklahoma. Indian Terr.: 21 Sep 1894, Bush 31 (MO). 24. Mexico. Coahuila: 24-26 Aug 1938, Johnston 7088 (G) 54 35 Arkansas, Missouri, Oklahoma and Texas was covered with clear, double- sided tape and individual seeds were placed on the map in the locations of their collections. Examination under a 30 X stereoscopic microscope made possible an analysis of local and broad patterns of variability. The maximum diversity of seed surface texture was noted in the Trans-Pecos (Figs. 17—21). There was an overall trend of flattened roughenings and less tubercles toward the north and west. To see this, compare Figs. 6 — 8 from Louisiana and Arkansas, Figs. 9— 12 from south and central Texas with Figs. 12— 16 from central Oklahoma and Figs. 17 — 21 from the Trans- Pecos with Fig. 22 from Western Oklahoma. From an analysis of these surface patterns (Figs. 1 — 22) it is apparent that variability of seed surface patterns occurs within geographic regions and within the species. Since all these seeds were taken from known red- flowered plants, seed surface texture is not a trait which alone can be used to separate red-flowered P. mundula from P. pilosa. Legrand stated that seed surface texture was not a good character for distinguishing varieties of P. oleracea, so the implication is that it is not a good character for distinguish- ing species, which our data support. THE CONCEPT OF PORTULACA MUNDULA The character states taken from descriptions and keys, listed inTable 1 show very little difference between the taxa called P. pilosa and P. mundula. When Johnston formulated the idea of P. mundula he stated that he had taken his concept of P. pilosa from pictures portraying the Linnean taxon, but there is no indication that he examined any specimens of P. pilosa outside of the Southwest. It would be difficult, in our opinion, to develop a concept of variability of as widespread a taxon as P. pilosa from pictures, particularly those as stylized as Commelin’s and Herman’s. Examination of their descriptions shows only one inconsistency in the typical morphologic- al expression of P. pilosa; this being in the leaf morphology. Herman notes that the tops of the leaves are flat, Commelin makes no such reference. In our observations, the typical terete leaf cross-section may become hemi- spherical (flat on top) in robust plants growing in richer soils. The overall shape, whether terete or hemispherical, is linear and not spatulate or oblanceolate. The general lack of consistency in character states for the specimens from Texas made us wonder about the consistency of the material cited by Johnston. The holotype from Coahuila, Mexico, was measured or scored for the same characters; capsule diameter, pedicel length and seed surface, using 15 capsules from the type instead of three. The pedicel length 56 averaged 0.65 mm, with a range of 0.3— 1.1, and the capsule diameter averaged 1.95 mm, with a range of 1.6— 2.4. Note the ranges are from a single plant. These results also show a smaller set of averages than the ranges given by Johnston in his original description (Table 1). Of the paratypes which were compared in Table 2, only one specimen, Texas, Mill Creek, Aug 1843, Lindheimer s.n., (G) had capsules 2.5 mm in diameter. For the seeds of the paratypes, 11 had seed surfaces that were stellate flattened roughenings, not stellate-tuberculate. These 11 specimens were from west Texas, Oklahoma and Missouri, while the stellate tuberculate seed surfaces were found on specimens from the remainder of Texas. Johnston cited no specimens form Arkansas. Figure 24 is a seed from the holotype, while Fig. 23 is a paratype from Oklahoma, (Indian Territory): Sapulpa, 21 Sep 1894, Bush 31 (G). Note the close resemblance of Fig. 24, the type, to that of Fig. 3 from Florida, and that Fig. 23 does not fit the tuberculate pattern of the type. This tuberculate pattern has been used as a distinguishing morpho- logical feature for the species. In fact, these findings on Johnston’s paratypes show the same trends noted above, with flattened stellate roughenings toward the northern part of the distribution. Portulaca pilosa has long been recognized as a taxon in Florida (Small, 1933). Legrand cited specimens from Florida, Georgia and Mississippi. Our measurements and analysis of capsule diameter, pedicel length, seed size and surface texture show no clearcut separation of red-flowered plants between Florida and Texas. Our data also support the comments of Legrand regarding the decreased size of capsules in plants growing in arid regions in contrast to larger capsules found in the tropics. The specimens with the smallest capsule diameters are from specimens recently collected from Arkansas, Oklahoma and Missouri. Some of the recent Texas specimens had small capsules, but the average of all Texas red-flowered specimens fits easily into the range of P. pilosa. It is possible to select specimens from dry habitats with small capsules and specimens from wet habitats with larger capsules. Unfortunately the habitat data on most of the labels are in- sufficient to permit an extensive analysis. Modern records from Arkansas, Missouri and Oklahoma show that the red-flowered species, which we are calling P. pilosa, is mostly restricted to dry ridges, bluffs and outcrops with sandy soul. THE QUESTION OF PORTULACA PARVULA Concerning the P. prlosa-P. parvula separation, as noted above, Gray treated the yellow-flowered taxon as P. parvila. Johnston redefined P. 57 parvula by adding the traits of small capsules, long pedicels and flattened stellate roughenings on the seed surfaces (Table 1). Measurements of the lectotype designated by Johnston: Mexico, Chihuahua: Sierra Santa Eulalia, fl. yellow 18 Aug 1885, Pringle 543 (G), provided six capsules measuring 1.02 mm in diameter and pedicel lengths averaging 0.6 mm, both less than stated by Johnston, but the seed surfaces had stellate flattened roughenings. We have seen red-flowered specimens from Oklahoma with capsules 1.5 mm in diameter, pedicels 1.0 mm long and seeds with stellate flattened roughenings (Fig. 23). In examining over 700 specimens labeled P. pi/osa, P. mundula, or P. parvula, only two were found with yellow flowers. This low number shows that either yellow-flowered, pilose Portulacas are rarely collected or they are less common than one would think. Taylor (R.J. Taylor, DUR 1984, pers. comm.) reported that yellow-flowered plants occasionally occur with red-flowered plants in Oklahoma. Smith (E.B. Smith, UARK 1984, pers. comm.) and Tucker (G. Tucker, APCR 1984, pers. comm.) have not seen yellow-flowered, pilose Portulacas in Arkansas. Steyermark (1963) does not report a yellow-flowered, pilose taxon in Missouri. Legrand treated the yellow-flowered P. parvula under P. halimoides L. citing the depauperate growth in the deserts and reaching its northern limit in the United States. Portulaca halimoides is a Mexican species, occurring chiefly in the western half of that country which could be invading the Trans-Pecos through or around the Chihuahuan Desert. Portulaca halimoides may not have been known by Gray, hence his describing the yellow- flowered taxon as a new species. Johnston also may have been unfamiliar with it, and accepted Gray’s interpretation. Before any final conclusions can be reached, further studies should be made of yellow-flowered species from a greater southwestern geographical range. Field studies to determine intermixing with red-flowered plants should also be undertaken. There is another possibility for the occurrence of few yellow-flowered plants. More than one species of Portulaca has both red and yellow flowers. Portulaca grandiflora exhibits a wide range of petal colors and Legrand reports that P. amilis Speg. has a yellow-flowered form in South America, although in the United States so far only the red-flowered form has been seen (Judd and Wunderlin, 1981). There have not been any reports of P. pilosa having anything other than red flowers but it is possible that a genetic analysis of western populations may show an occasional yellow-flowered plant. This would account for the low incidence of yellow flowers overall or for the infrequent mention on herbarium labels of both red and yellow flowers in the same population as did Waterfall: Texas: Jeff Davis Co.: 20 mi SSE of Kent, 31 Jul 1943, Waterfall 5415 (G). THE PRESENT STATUS OF P. PILOSA, P. MUNDULA AND P. PARVULA This study points up the problems of limiting the consideration of species concepts to unnatural boundaries and the importance of examining species complexes over a broad geographic range. Even in this case, the final answer will only come with a more extensive look at the Mexican flora. In this study, comparison measurements of the character states of the morphological traits taken from descriptions and used in keys to distin- guish P. mundula from P. pilosa show that there are no consistent characters which can be used, singly or together to separate the taxa. We conclude that P. mundula is conspecific with the more widespread and variable species P. pilosa and the name P.. mundula should be treated as a synonym of P. pilosa. For P. parvula, the only consistent identification trait is yellow petals. Specimens without petals cannot be identified by the seed surface texture as proposed by Johnston. However, without a more extensive examination of known yellow-flowered P. parvula along with a concept of P. halimoides we cannot reach any conclusion on the validity of P. parvula as a species. CYTOLOGY Very little information is known about the cytology of the genus Portula- ca and there are some chromosome counts that are unusual. The cytoplasm stains darkly with aceto-carmine but fortunately the number of chromo- somes is not large. The base numbers have been accepted as X =4 and 9, with polyploid multiples and possible aneuploid sequences. The lowest number, n= 4 (Steiner, 1944), has a direct bearing on the P. pilosa-P. mundula problem. Steiner’s report, unfortunately unvouchered, was taken from a plant collected at Springdale, Arkansas, in the NW corner near Oklahoma and Missouri. Steiner could not identify the plant to species and speculated that it may be a new species. The senior author has verified this count on plants grown from seed collected from a sandstone outcrop in Benton Co., Arkansas near Springdale. This species would now be in- terpreted as P. pilosa, and the specimen appears to be such. This count would represent a new number for this species, giving a sequence of n=4, 8, 9, 18. The n=4 is interesting in that it is the lowest number for the genus. The highest number for P. pilosa, n= 18, was reported by Hsu (1968). Since P. pilosa is geographically widespread, the variation in chromo- some number is not unexpected. However, it appears that P. pilosa exhibits polyploidy from both base numbers of the genus. How widespread the n= 4 number is for the high prairie of the United States and what relationship this number has to the actual numbers and to the potential for gene De) exchange with the southwestern populations is unknown. Chromosome data need to be determined for the western populations. All counts for P. pilosa from the Southeast have been n= 8. A study of the cytology of this taxon in Oklahoma, Texas, Arkansas and Missouri would help to clarify the species concept in that area. If study shows that designation of a new species is warranted, then a new name would be needed, since the description associated with the name P. mundula probably would not encompass this new taxon. Also, the perpetuation of the name P. mundula, redefined to new limits, would only add to the confusion of the species concept. THEORETICAL PATTERNS OF MIGRATION OF PORTULACA N CENTRAL AMERICA AND THE UNITED STATES Portulaca, with a large concentration of species in South America, has spread north into Central and North America, including the Caribbean (Legrand). Portulaca pilosa probably entered Florida from the Caribbean and spread northeast along the Atlantic coastal plain into North Carolina. A greater movement has taken place westward along the Gulf Coast, but there is no indication of an inland movement up the Mississippi embayment into Arkansas and Missourt. Only one isolated population, with measurements similar to those of the coastal plain has been reported from Tennessee, that along Interstate-40 (Wilson Co.). A scattered distribution is noted in northern Louisiana with most of these collections associated with recent human activity, e.g. railroads, road fills and dum Northeast Texas shows the same scattered distribution as northern Louisiana, while the greatest concentration of P. pz/osa is in south Texas along the Gulf Coast and in the Trans-Pecos region. There is a similarity of the plants in the Trans-Pecos area with those of the panhandles of Texas and Oklahoma, having a general reduction in size northeastward. Our label data show the distribution of P. pilosa in Arkansas in the mountains; the measurements show affinities with the gene pool of Oklahoma and not the coastal plain. The dates of collection also show a more recent movement eastward into Arkansas. The habitats in Arkansas are in the highlands and represent affinities with Oklahoma physiographically (G. Tucker, APCR 1984, pers. comm.). Legrand commented on the existence of P. pilosa in the Caribbean Islands, Florida and Mexico, stating there was a morphological change toward smaller plants as one moves toward drier habitats, particularly in Mexico. These statements support the patterns of distribution indicated by the herbarium specimens we have seen. A two-directional pattern of movement of P. pilosa into the United States can be postulated: 1. From the Caribbean into Florida, northeast along the Atlantic coastal plain and also 60 westward along the Gulf Coast to the Mississippi embayment and, 2. From South America into Central America (Mexico) and northeastward into Texas, Oklahoma, Arkansas and Missouri. A similar pattern can be postulated for the movement of P. parvula, if indeed it is really P. halimoides. Portulaca halimoides, according to Legrand, is found in central and western Mexico and “‘appears to be spreading along major highways.”’ Its invasion into northern Mexico and southwest Texas (Trans-Pecos) is probably controlled to some extent by human activity and the availability of habitats as has been shown in other cases, particularly that of P. amilis Judd and Wunderlin, 1981). ACKNOWLEDGEMENTS We wish to thank Josette Gourley Arvey and Alma Amell for assistance in translation, Dr. D.E. Boufford of the Arnold Arboretum of Harvard University for bibliographic assistance, Sandra F. Zane, UNCC, for the scanning electronmicroscopy, Dr. L.S. Barden for statistical analysis and s. J.R. Massey, UNC-Chapel Hill, and T.L. Mellichamp, UNCC, for valuable recommendations in the study and in the preparation of the manuscript. REFERENCES COMMELIN, C. 1697, Hort. Med. Amstelodam, p. CORRELL, D.C. and M.C. JOHNSTON, cn Manual of the vascular plants of Texas. Texas Research Foundation, Renner, Texas. FERNALD, M.L. 1950. Gray’s manual ee 8th ed. American Book Co., New York. GRAY, A. 1887. XV Contributions to American botany. Proc. Amer. Acad. Arts. 2c2 J 2-274) HERMAN, P. 1705. Paradisus batavus, p. 115. HSU, C.C. 1968. se mae chromosome studies on the vascular plants of Taiwan (ID). Taiwania 14:11 — JOHNSTON, I.M. ne Species from Mexico and the United States, IH, J. Arnold Arbor. 4— 196. JUDD, W.S. and R.P. WUNDERLIN, — ‘i First report of Portulaca amilis (Portula- caceae) in the United States. Sida 9(2): LINNAEUS, C. 1753. Species plantarum. a , Holm LEGRAND, C.D. 1962. Las especies Americanas ve tis Anales Mus. Nac. Monte- video 7(3):1— 147. LONG, R.W. and O. LAKELA. 1971. A flora of tropical Florida. University of Miami Press, Coral Gables, Florida. ROBINSON, B.L. and M.L. FERNALD. 1908. Gray's new manual of botany. 7th ed. American Book Co., New York. SHINNERS, L.H. 1958. Spring flora of the Dallas-Fort Worth area Texas. Published by the author, Southern Methodist University, Dallas, Texas. 61 SMALL, J.K. 1933. Manual of the southeastern flora. Published by the author, New York. SMITH, E.B. 1978. An atlas and annotated list of the vascular ote of Arkansas, + Saeplenen Published by the author, Fayetteville, Arkan SOKAL, R.R. and F.J. ROHLF. 1981. Biometry. W.H. peat Company, San Fran- Cisco. STEINER, E. as Cytogenetic studies on Talinum and Portulaca. Bot. Gaz. 105:374 — ee | A. 1963. Flora of Missouri. lowa State University Press, Ames, lowa. WUNDERLIN, R.P. 1982. Guide to the vascular plants of central Florida, Gao Presses of Florida, Gainesville, Florida. SIDUS SIDARUM — V. THE NORTH AND CENTRAL AMERICAN SPECIES OF SIDA. PAUL A. FRYXELL U.S. Department of Agriculture and Texas AGM University College Station, Texas 77843, U.S.A. The genus Sida L. is one of the larger and more difficult genera of the Malvaceae. Over 1000 names have been published in the genus, although recent estimates admit only 150 — 250 species. Even this range may be too high. There is much synonymy. Many species that were originally (or at one time) placed in Sida have been relegated to other genera because the early concept of S7da was as a very inclusive genus, more or less encompassing all mallows that were uniovulate and lacked an involucel. A contemporary revision of Sida is needed. The nearest approach to sucha comprehensive treatment for the New World are the two “tentative keys” published by Kearney (1954, 1958) and the detailed study of Clement (1957). In the subsequent quarter century understandings of Sida have deepened, additional species have been described, others have been segre- gated out of the genus, and a clearer understanding of sections within the genus has emerged. Hence, it is felt that the time ts opportune to begin a new treatment of Sida. The present treatment of the North and Central American species is a beginning. I follow Kearney (1954) and others in taking ‘“North America”’ to include the West Indies and Central America as far as (and including) Panama, but use the more expanded designation in the title for the sake of clarity. Those Caribbean islands that are primarily South American in location and phycogeography (Aruba and Curacao to Trinidad and Tobago) are excluded. Numerous species have been segregated to other genera, e.g. to Al/osz- dastrum (Hochr.) Krapov. Fryx. & Bates (ined.), Bastardiopsis (Schum.) Hassl., Billieturnera Fryx., Dendrosida Fryx., Krapovickasia Fryx., Malvella Jaub. & Spach, Meximalva Fryx., Rhynchosida Fryx., and Sidastrum E. G. Baker. The removal of such species has made the residual genus less heterogeneous and more natural—and thus more readily characterized. earlier (Fryxell, 1978) emphasized the morphology of the mericarps and SIDA I1(1): 62-91. 1985. 63 of the calyx in delimiting Sida. The mericarps are differentiated into a lower, one-seeded, indehiscent cell and an upper, empty, dehiscent portion that is often ornamented with a pair of spines. The upper and lower portions of the mericarp are set apart dorsally by a distinctive “shoulder” that is an extension and joining of two lateral ribs. The presence of this shoulder is distinctive. The lower cell is trigonal in cross section and is often laterally reticulate. The calyces in Sida are typically 10-costate at the base, the ten ribs leading alternately to the sinuses and the apices of the 5-lobed calyx. The former (the commissural nerves) are relatively more prominent (especially in sect. Sida); the latter become the midribs of the calyx lobes. These ribs are often yellowish where they come together at the base of the calyx. Growth habit also distinguishes Sida to a degree. The genus includes annual and perennial herbs and relatively small shrubs, seldom exceeding 1— 1.5 m in height. Characters of the calyx and of the mericarps are useful not only in delimiting the genus but also in subdividing it into sections. Previous subdivisions of the genus (e.g. Gray, 1849; Schumann, 1891, who were followed by Baker, 1892; Rodrigo, 1944; Kearney, 1951, 1954; and Hutchinson, 1967) have not been entirely satisfactory, in part because it was difficule for them to establish natural sections when they accepted the genus as a heterogeneous group. Clement’s (1957) subdivisions were more natural (except his sect. Physalodes A. Gray, which included elements of three or more genera), but Clement dealt with only selected sections and omitted a major portion of Sida. Monteiro’s (1942, 1949) subdivisions were too complex to be coherent. In the following treatment I recognize 11 sections of Sida that I believe represent natural groups. I will comment on species not represented in North America that are included in these sections and will allude to those two sections not represented in North America and include them in the key to the sections. Leaf morphology is a strong supporting character in delimiting the sections of S:da and to a lesser degree in distinguishing species. The accompanying figures (Figs. 1 — G) are presented to facilitate understanding of the sectional descriptions and to demonstrate the characters of leaf form that support the sectional divisions. The figures are not intended (except in certain instances) to be critical in making identifications at the specific level. KEY TO THE SECTIONS OF SIDA a. Leaves entire, narrowly linear to elliptic or broadly lanceolate (Fig. 1, 1—J); inflorescence corymbiform, terminal, essentially leafless... ... sect. ae (p. 65) 64 a. Leaves crenate, dentate, or serrate (rarely subentire), variously shaped; flowers solitary in the axils or variously aggregated into inflorescences, se corymbiform b. Leaves decay palmately lobed. . Mer — jon aom ricarps (and styles and stigmas) commonly 5 (sometimes 7 — 9), strongly differentiated, with two apical spines or aristae; anther olivia, Peru, Ecuador)................... t. Oligandrae (p. 65) c. Mericarps (and styles and stigmas) 8 — 10, relatively npalfferendiaeed. muticous; anthers 10 or more d. Flowers i in al terminal inflorescences; anther numer ous (United States)... ...00020..000...0000... ect. Pseudo-Napaeae (p. 66) d. Flowers axillary, solitary or paired; anthers 10: 20 (Australia, CO) io oh te ae a wae eee ae sect. Hookerianae (p. 66) b. Leaves lobes a Son oped: shallowly so), linear, elliptic, lanceolate, or eee ees ves se or ovate, basally more or less cordate, dentate to the base (Fig. 1, G—H; Figs. 2—4). f. Mericarps (and style and stigmas) ey 5, the spines, if aes a herb vi! so; calyx ribs obscure g. eae es dark-green-margined, lacs oe often scan- OE SO SEDI on res Wie Gating aed 4 oe (p. 68) g. Cay lobes not bi-colored; plants usually erect somes PEOSCLALE) rrsifc 4s ba tie Beadle each sata ee atic einenete ate se Sia (p. 72) f. Wea (and styles and stigmas) 7 or more (sometimes as ae 5); calyx ribs pronouned h. Flowers and fruits commonly subsessile in axillary glomerules and densely aggregated into racemiform or paniculate in- florescences; mericarps muticous, aoe! (if at all) ae plants sometimes prominently setiferous ........ sect. Muticae (p. 76) h. Flowers and fruits variously disposed oe not densely aggre- gated as above (except sometimes in S. cordifolia); meticarps usually spinose and reticulate, the spines often retrorsel barbed; plants usually not setiferous..........s ct. Cordifoliae (p. 77) e. Leaves linear, elliptic, or rhomboid, basally truncate or cuneate, dentate to a base (Fig. 5) or basally entire and distally dentate (Fig. , A—E; Fig. 6). ewe more or less rhomboid (Fig. 6); mericarps 7— 11, usually PV MOG NIE ase has Scag gr niny gah eta ls ees pea oy ec wes sect. Sidae (p. 85) . Leaves linear, oblong, or elliptic j. Mericarps (and styles and eee 5—7, often muricate; flowers and fruits apically congested with leaves and stipules So as to appear involucellate; leaves often dentate only at apex even subentire (Fig. 1, A—E); plants often — Gay whet Aad et elo aoe ane a aigen ease ae a eee . Malachroideae (p. 80) j. Mericarps (and styles and stigmas) ie more than 7 smooth, flowers and fruits solitary in the axils or crowded apically but not appearing involucellate; leaves usually dentate throughout (fig. 5); plants usually erect sect. Ellipticifoliae (p. 82) 65 SIDA section STENINDAE Grisebach, Fl. Brit. W. I. 76. 1859. Tver: Sida lintfolia Cavanilles. Herbs or subshrubs, hirsute or hispidulous to glabrescent. Leaves entire, narrowly linear to elliptic or broadly lanceolate (Fig. 1, 1 —J); petioles 3 —6 mm long; stipules equaling or exceeding the petioles. Flowers borne in a leafless terminal corymbiform inflorescence of 8— 10 flowers; calyx rounded, 5 — 7 mm long; corolla 8— 15 mm long, white or yellowish with dark center. Mericarps 5—9, thin-walled, indehiscent, apically blunt. Sida sect. Stenindae includes two species, one of which, S. /intfolia, occurs in North America. The other, §. hassleri Hochr., is endemic to Paraguay. The two taxa have been distinguished in varietal rank (cf. Clement, 1957). 1. Sipa LIniFOLia Cavanilles, Diss. 1:14. t.2.f.1. 1785. Type: in insula Caienae and in Peru. ae ee L. Richard, Actes Soc. Hist. Nat. Paris 1:111. 1792. Tyee: none Side. viminea Fischer ex Link, Enum. Pl. 2:202. 1822. Type: BRAZIL Sida linearifolia Thonning in Schumacher, Beskr. Guin. Pl. 303. 1827; Dansk. Vid. Selsk. Afh. 4:77. 1829 (non St.-Hilaire, 23 Apr 1827). pte GUINEA. eed 120 (C-4; cf. Junghans, 1961, p. 343; Hepper, 1976, p Sida angustissima Miquel, Stirp. eo Sel. 102. 1850 (non Ee sie 1827). Sida longifolia Brandegee, Zoe 5:212. 1905 Type: MEXICO. Sinaloa, Culiacan, Cerro Colorado, 3 Nov 1904, Sst (HOLOTYPE: UC). Sida flebrigii Wibck: Bot. Jahrb. Syst. 54 (Beibl. 117):72. 1916. Type: PARAGUAY, Cordillera de Altos, Fiebrig 572 (M as photo F-19686). Sida linifolia is usually a savanna plant at elevations below 1200 m. It occurs from Panama north to Mexico (Chiapas to Sonora on the Pacific slope, and eastward to Tabasco and Veracruz); in the West Indies in Cuba, Hispaniola, Jamaica, and Martinique. It is also found throughout much of outh America (Colombia, Venezuela, the Guianas, Brazil, Paraguay, Peru, and Bolivia), in various parts of Africa (Sierra Leone to Tanzania to Angola), and in Fiji. SIDA section OLIGANDRAE Clement, Contr. Gray Herb. 180:64. 1957. Type: Sida palmata Cavanilles. Erect annual herbs, variously pubescent. Leaves palmately lobed, long- petiolate. Inflorescences more or less paniculate; calyx cupuliform, often ecostate, 6 mm or less at anthesis but often accrescent; petals equaling or somewhat exceeding calyx, often red or purplish; stamens 5 — 20. Mert- carps often 5 (-9), the lateral walls adnate to the seed, usually long-aristate (4— 14 mm long) the aristae retrorsely pubescent. 66 Sida sect. Oligandrae includes eight species, none of which occurs in North America (Fries, 1947, pp. 14— 19; Clement, 1957). The section is a distinctive natural group, indigenous to relatively high elevations in Ecuador, Peru, and Bolivia. SIDA section PPEUDO-NAPAEAE A. Gray, Mem. Amer. Acad. Arts 4 (Pl. Fendl.):23. 1849. Type: Sida hermaphrodita (L.) Rusby. Robust perennial herb 1 —5 m tall, glabrous or with sparse pubescence. Leaves large (up to 24 cm long), palmately 5 — 7-lobed, coarsely dentate, petiolate (Fig. 1, F). Inflorescence a terminal panicle composed of umbel- late corymbs; calyx cupuliform, ecostulate; corolla white, 8—9 mm long; stamens numerous. Carpels 8 — 10, indehiscent, with horizontal rostrum, the lateral walls evanescent. The section is monotypic. Clement (1957) points out that it “has no close affinities with any other {section} in the genus, nor with Napaea save in habit.” It is also geographically and ecologically distinct from the remain- der of Sida, being a temperate-zone plant whereas Sida is typically tropical and subtropical. One might plausibly argue that the section be elevated to generic rank. 2. SIDA HERMAPHRODITA (L.) Rusby, Mem. Torrey Bot. Club 5:223. 1894. Basionym: Napaea hermaphrodita L. Sp. Pl. 686. 1753. Type: probably from Hort. Cliff. (BM); the species is not represented in the Linnean Herbarium (LINN) in spite of a assertion of Iltis (1963, _P- 106). Sida napaea Cavanilles, Diss. 5:277. L132 jd. Napaea laevis L. Mant. 2:435. 1771, pro syn. Sida hermaphrodita occurs in the United States, in Maryland, the District of Columbia, Pennsylvania, West Virginia, Tennessee, Ohio, and Michigan—possibly escaped from former cultivation at some localities (cf. Iltis, 1963, pp. 106— 107; Spooner et al. 1985). SIDA section HOOKERIANAE Clement, Contr. Gray Herb. 180:77. 1957. Type: Sida hookeriana Miquel ex Lehmann. Herbs or subshrubs, sometimes decumbent. Leaves 3 — 5-lobed, 3 — 7 cm long, long-petiolate, sparsely pubescent to glabrous. Flowers solitary or paired in the leaf axils on slender peduncles; calyx cupuliform, 5 — 8 mm long; corolla white or pale yellow, slightly exceeding calyx; anthers 10— 20. Mericarps ca. 10, muticous, indehiscent. Sida sect. Hookerianae was described to include two species, one from southwestern Australia (S. Aookeriana Miquel ex Lehmann), the other 67 Figure 1. Leaf outlines of Sida. A-E, Sida sect. Malachroideae; A, S. brittonz; B-D, S. ciliaris; E, S. brachystemon; F, Sida sect. Pseudo-napaea: S. hermaphrodita; G-H, Sida sect. Muticae: S aggregata; \-}, Sida sect. Steninda: S. linifolia. [A, Leén & Roca 7466; B, Hinton 1469; C, Magallanes 3723; D, Lasalle et al. 810629-1; E, Koch et al. 79424; F, Fryxell s.n.; G, Fryxell & Bates 2107; H, Fryxell & Bates 2123 ;1, Ventura 5638; J, Koch & Fryxell 82207.) Scale = 1 cm. 68 African, from Ethiopia to South Africa (S. ternata L. fil.). The section ts not represented in the Americas. SIDA section acne Borssum Waalkes, Blumea 14:180. 1966. Tyee: Sid. ta (Burman fil.) Borssum Waalkes [BASIONYM: Melochia cordata Burman fil. }. Prostrate or decumbent herbs to erect shrubs, variously pubescent with stellate hairs, glandular hairs, and setose hairs. Leaves ovate-cordate, serrate-crenate throughout (Fig. 2). Flowers solitary in the axils, in axillary glomerules, or in diffuse panicles; calyx pentangular and pyramidal, the lobes trullate, the margins and midribs dark green against a light green background; corolla white, yellow, or orange, sometimes with a red center. Mericarps uniformly 5, muticous and glabrous or spinescent and antrorsely pubescent, the dorsal walls rounded and carinate, the lateral walls thin and sometimes evanescent. In addition to the North American species treated here, sect. Ne/avagae includes two additional species from Asia, S. e/ongata Blume and S. mysorensis Wight & Arnott (cf. Borssum Waalkes, 1966) and several species from South America, such as S$. dictyocarpa Grisebach and S. caudata St.-Hilaire (cf. Rodrigo, 1944; Kearney, 1958) Members of sect. Ne/avagae are characterized by a base chromosome number of x = 8 and are distingusihed from the remaining species of the genus, which (where known) have a base number of x = KEY TO THE NORTH AMERICAN SPECIES OF SIDA SECT. NELAVAGA . Plants prostrate, often repent, never viscid. b. Leaves I asymmetr ical, ovate (Fig. 2 C— D); calyx lobes cordate, accrescent becoming sagittate in fruit; staminal column glabro i Maly oop RUE S Bl Deans Bape hs ath. we uadadey sede Gates acays. ates _S. pussteana b. Leaves symmetrical or slightly asymmetrical, orbicular-ovate; calyx Ae triangular, not accrescent; staminal column pubescent. c. Stems repent; mericarps awned.......2..0.00000 00002 eee 7. S. repens . Stems not (or rarely) ae at the nodes, sometimes ascending; mericarps minutely apiculate... ...........0. 000000 ee eee S. cordata a. ba ascending to erect, often s ene viscid or not; leaves symmetrical —B, [—L); calyx not accrescent ic ee subsessile, aguregrated into dense axillary aoe or in axillary p edunculat te “heads”; stems often hispid (hairs 2 — 3 mm long); ABEOUG ah. Bateman ae eae eee e ee ee _S. wrens mericarp d. Flowers se lies a sei longer than the calyx), solitary in the we axils or in open panicles; stems usually viscid. e; eae usually beaked and pubescent; plants viscid or not... .. 4. S. glabra . Mericarps muticous; ee densely viscid. f. eae pide eae blackish at maturity; stems with sim- ple hairs 2—3 mm long in addition to glandular hairs; leaves often ar a aaa blotch along the midvein ........... 9. S. xavieri f. Mericarps pubescent, brownish; stems lacking long hairs; leaves MOG: DIOUG Hed ete 8.248 <2 oo Boe ain ee ee 6. S. nesogena 3. SIDA CORDATA (Burman fil.) Borssum Waalkes, Blumea 14:182. 1966. BasionyM: Melochia cordata Burman fil. Fl. Ind. 143. 1768. Type: ““Melochia cordata’’ s. coll. s.n. (HOLOTYPE: G). Borssum Waalkes (1966, p. 183) comments on the type D A B Cc H G F E J J Figure 2. Leaf outline of Sida sect. Nelavagae. A-B, S. glabra; C-D, S. pussieana; E-G, S. repens; H, S. cordata; 1-J, 8. xaviert; K-L, S. wrens. {A, Koch et al. 79133, B, Ventura 14012; C, Velasco 8874; D, Fryxell & Lott 3312; E, Molina & Williams 31209, FE, Sintenis 2951; G, Sintenis 179; H, Tyson 144; 1, Koch et al. 79470; J, Koch et al. 79502; K, Dorantes et al. 5274; L, Fryxell & Lott 3234.} Scale = 1 cm. Le Sida veronicaefolia Lamarck, Encycl. 1:5. 1783. Type: “Les In ’ Sonnerat s.n des HOLOTYPE: ic LA, MA as ce F-29795). Sida humilis var. ps (mak Masters, Fl . Ind. 1:322. 1875. Sida radicans C ae Diss. 1:8. ¢.9./.3. 1785. Type: Rheede, Hort. Malabar. 1.69 (HOLOTYPE). Sida morifolia Cavanilles, Diss. 1:9. ¢.20.f.2. 1785. Type: MAURITIUS, er. n. (HOLOTYPE: MA). Lamarkia il (Cav.) se Se Phil. Bot. 1: 1789. Sida multicaults Cavanilles, Diss. 1:10. ¢.1./.6. 1785. Type: MALABAR, ae. eet MA). Sida eis var. multicaultis noe ) E. G. Baker 30:293. 1892. Sida humilis Cavanilles, Diss. 5:277. ¢.134.f.2. 1788. Type: Borssum Waalkes (1966) designated as neotype: Luzon, Port Cavite, Née s.7. (MA). Marais (1983) took exception to this choice of neotype, but offered no ine one peoplcm, stating only the opinion that the plant in question did not conform to Sida cordata (Burm. Borss., th ) giving no paneer of the plant or any opinion as to its identity. Both look e fact that Sida humilis Cav. can be satisfactorily lone on the basis of Cavanilles’ ae thus obviating the need to designate a neotype. Sida veronicaefolia var. humilis (Cavanilles) K. Schumann, Mart. FI. Bras. 1303) 320. 1891. Sida unilocularis L'Héritier, Stirp. Nov. 1:117 bis. 4.56 dis. 1789. Type: L’Héritier’s is cited as holotype by Borssum Waalkes (1966); it should be noted that L'Heéritier cites a Commerson collection from Mauritius, which may in fact be type material of the earlier S. morifolia Cav., thus making L’Héritier’s name superfluous. Sida cordata and S. repens (and their several synonyms) have been freely confused in the literature. It is therefore difficult to state their distributions from literature citations without consulting specimens or at least detailed descriptions. In the West Indies 8. cordata apparently occurs in Puerto Rico and Hispaniola, possibly 1n Cuba and elsewhere. It also occurs in Malesia (Borssum Waalkes, 1966). 4, SIDA GLABRA Miller, Gard. Dict. ed. vili, no. 14. 1768 (non Nuttall, 1834). Type, fide Fawcett & Rendle (1926): (BM). Sida eee Cavanilles, Diss. 1:15. ¢.2.f.2. 1785 (non Miller, 1768). Type: INS. SA DOMINGO, Thouin s.n. Sida enna Cavanilles, Diss. 1:16. ¢.2.f.8. 1785. Type: MAURITIUS, Commerson s.n. (LECTOTYPE: P-JU). Borssum Waalkes (1966, p. 90) discusses the choice of lec- totype. Sida arguta Swartz. Prodr. Veg. Ind. Occid. 101. 1788 (non Presl, 1835). Type: JAMAICA, Swartz 5.2. (HOLOTYPE: S; IsoryPEs; G as photo F-7996, B herb. Willd. no. 12692). ?Sida arguta Fischer ex Link, Enum. Pl. 2:206. 1822 (non Swartz, 1788). Typr: BRAZIL. Sida verruculata DC. Prodr. 1:473. 1824. Sida nervosa DC. Prodr. 1:465. 1824. Type: SANTO DOMINGO, oa paniculata,” ertero §$.n. (HOLOTYPE: G-DC; isorypE: MO, W as photo F-333 et) Sida viscidula Blume, Bijdr. 2:76. 1825. Type: ae n. (HOLOTYPE Sida fasciculata W illdenow ex Sprengel, Syst. 3:113. 1826 (non Torrey's & Gray, 1838). Type: Cumana, Humboldt & Bonpland s.n. (B a Willdenow no. 12691). Sida Fl willdenowti D. Dietrich, Syn. Pl. 4:847. 1847. Sida endlicheriana Presl, Reliq. Haenk. 2:111. 1835. Type: MEXICO, Haenke 5.n. OTYPES: , MO). Sida eS Watson ex Rose, Contr. U.S. Natl. Herb. 1:93. 1891; Proc. Amer. Acad. Arts 26:133. 1891. Type: MEXICO. Sonora, Alamos, Palmer 683 (BM GH, ee rapa Hassler, Feddes Repert. Nov. Sp. Regni Veg. 12:264. 1913. Type: GU . Bellavista, in regione flum. Apa, Hassler 10990 (B as photo — ieee kes : Sida cearensis Ulbrich, Notizbl. 6: oe 1915. Tyee: BRAZIL. Ceara, Cerro de Baturité, e 5 (B as photo F-9374, Sida insperata Standley & Sle Cake 3:51. 1952. Type: HONDURAS. Depr. Morazon, vic. of El Zamorano, Standley 24639 (HOLOTYPE: US). Sida glabra is widely distributed from Mexico south to Panama, through much of South America, and in the West Indies. It also occurs in parts of the Old World (cf. Borssum Waalkes, 1966). 5. Srpa JUSSIEANA DC. Prodr. 1:463. 1824. Type: PERU, J. de Jussiew s.n. (P-JU no. 12 Sida pane St-Hilaire & Naudin, Ann. Sci. Nat. Bot. 18:52. 1842. Type: BRAZIL. NAS GERAES, prope Congonhas, Vauthier 21 (P as photo F-35535, and in Rodrigo, ican plate 11). Anoda decumbens (St.-Hil. & Naud.) Hochreutiner, Annuaire Con- serv. Jard. Bot. Genéve 20:56. 19 Sida stolonifera Salzmann ex Turczaninow, Bull. Soc. Nat. Mosc. 31:199. 1858. Type: BRAZIL. Banta, Sa/zmann 5.n. (HOLOTYPE: KW; tsorypes: K, MO, P). Physalias- trum stoloniferum (Salz. ex Turcz.) en Anais XX Congr. Nac. Soc. Bot. Brasil Sida. begontoides Grisebach, Bonplandia 6:3. 1858. Type: PANAMA, Duachassaing 5.n. (HOLOTYPE: GOET). Sida jussieana occurs from southern Mexico (Guerrero, Oaxaca, and Chiapas) through Central America and throughout much of South h America (to Brazil, Paraguay, and Argentina). Most of the above types are discussed by Krapovickas (1969, pp. 20 — 21). 6. Sipa NESOGENA Johnston, Proc. Calif. Acad. Sci. 20:76. 1931. Tyre: MEXICO, Revillagigedo Islands, Socorro Island, Grayson’s Cove, Mason 1613 (HOLOTYPE: CAS; 1sorypeEs: GH, K, MO, US). Sida nesogena is endemic to the Revillagigedo Islands. 7. SIDA REPENS Dombey ex Cavanilles, Diss. 1:7. 1785. Type: PERU. Lima, Dombey s.n. (MA as photo F-29787, P-JU no. 12267, p.p.). Sida dombeyana DC. Prodr. 1:463. 1824 (cf. Krapovickas, 1969, p. 20). Sida chaetodonta Turczaninow, Bull. Soc. Nat. Mosc. 31:199. 1858. Type: Guayaquil, Jameson 392 (HOLOTYPE: KW; isotypes: K, OXF) Sida repens has a scattered distribution in the West Indies and also occurs 72 in Central America (Nicaragua, Panama) and South America (at least in Peru) and in Malesia. Borssum Waalkes (1966) treated this taxon as Sida javensis ssp. expilosa Borss., but the basionym for this name (Szda pilosa Cav.) represents a different species, viz. Sida abutifolia Miller, q.v. In the West Indies Stda repens occurs in Cuba, Hispaniola, Jamaica, the Bahamas, the Virgin Islands, and Grenada. 8. SipA URENS Linnaeus, Syst. Nat. ed. x. L145. 1759. Type: Browne s.n. (LINN-866.20, as photo in Rodrigo, 1944, plate 17). Sida verticillata Cavanilles, Diss. 1:13. ¢./.f.12. 1785. Type: BRAZIL. Rio de Janeiro, Jussten 5.n. (MA as photo F-29796). Sida ait St.-Hilaire, Fl. Bras. Mer. 1:185. 1827. Type; BRAZIL. Minas GERAEs, prope oe a -Hilaire s.n. Baan TYPE: isoTyPE: US). Sida urens var. rufescens St.-Hil. - ker, J. Bot. 30:294, Sida debilis G. ce Hise. 1 os ‘a . pe: GUINEA. Sida oe ” Don, Gen. Hist. 1:499. 1831 (non Hooker, 1828, nec Dietrich, 1847). Type: GUINEA. Sida congensis D. Dietrich, Syn. Pl. 4:859. 1847. os es Steudel ex Triana & Planchon, Ann. Sci. Nat. Bot. 17:177. 1862, nom. Sida Hochreutiner, Annuaire Conserv. Jard. Bot. Genéve 6:40. 1902. Type: ADAGASCAR, Mayotte, Borvin 3331 (HOLOTYPE: G). Sida urens var. aurea Hassler, Feddes Repert. Sp. Nov. Regni Veg. 12:267. 1913. Tye: PARAGU Sida margarttensis Hassler, Feddes Repert. Sp. Nov. Reon Veg. 12:266. 1913. Type: PARAGUAY, prope Cerro Margarita, in regione flum. Apa, Hassler 11065 (BM). Sida urens occurs: in southern Mexico and throughout Central America and the Greater and Lesser Antilles. It also occurs in many parts of South America and in Africa. 9. Sina xaviert Monteiro, Anais XIX Congr. Nac. Soc. Bot. Brasil 47. 1968. Tyre: BRAZIL, Paraisa, Joao Pressoa, 9 Oct 1941, Xavier [342] (Laboratorio de Fibres) (HOLOTYPE: RBR?). Sida xavieri is known from Brazil and has more recently been found to occur also in Mexico, Panama, and Venezuela. SIDA section SPINOSAE Small, Man. Southeast. Fl. 849. 1933. Types: (Article 22): Sida spinosa Linnaeus. Procumbent herbs or erect herbs or subshrubs, variously pubes- cent. Leaves lanceolate, oblong-ovate, or rotund, dentate-crenate throughout, short- to long-petiolate (Fig. 3). Flowers mostly in the leaf axils, che pedicels long or short; calyx angulate, inconspicuous, 10-nerved; corolla yellow or white. Mericarps uniformly 5, usually spinescent, dorsally carinate. Figure 3. Leaf outline of Sida sect. Spinosae. A, S. jamaicensis; B-D, S. spinosa; E, S. viarum; F, S. nummularia; G-H, S. glomerata; |-J, 8. abutifolia. [A, Ekman H2760; B, Rugel s.n.;C, Stevens 9367; D, Vazquez 2165; E, Neil 2972; F, Milne 85; G, Knapp 1905, H, Jiménez 5512; I, Pringle 8440; J, Garcia 250; K, Rzedowski 29547; L, Clarke 17878-1.} Scale = 1 cm KEY TO THE NORTH AMERICAN SPECIES OF SIDA SECT. SPINOSAE a. Plants procumbent (but not repent). b. Leaves short- ara the blades rotund (Fig. 3, F), strongly discolorous, pedicel 1 es lenethy OF CAV Keon ec heehee es Sea Ses 13. 8. nummularia b. Leaves long- ie. the blades oblong (Fig. 3, I—L); pedicel ma times es th of calyx, often with long (1— 2 mm) simple hairs . 10. i abutifolia a. Plants er c. oe on branching pattern) distichous; stipules prominent, lanceolate or falca d. Planes densely stellate-pubescent cite leaves more or . obtuse; mericarps with beaks 1— 1.5 mm long .......... . S. jamaicensis d. Plants hirsute; leaves acute; mericarps inurl apiculate ... a S. glomerata c. Leaves spirally disposed; stipules subulate e. Corolla white (or pale yellow) with red center; stipules commonly SUIVEU teh Lee Loa aad ee RAR eee 15. S. viarum e. Corolla yellow without red center; stipules usually l-nerved ... 14. S$. spinosa 10. Sipa ABUTIFOLIA Miller, lathe Dict. ed. viii. no. 12. 1768. Type: inculk. Miller s.n. (HOLOTYPE: BM). Note: The name is given as “‘abutilifolia’” on the type specimen a in Index Kewensis. Sida ala Swartz, Prodr. Veg. Ind. Occid. 101. 1788. Type: HISPANIOLA, . (LECTOTYPE: S; ISOTYPE: ; Sida den Callie: Diss. 1:9. ¢.1./.8. 1785 (non Miller, yee nec Retzius, 1781, c L’Héritier, 1789, nec Vellozo, 1825). Type: INS. SANTO DOMINGO. Sie se Cavanilles, ee 6 350. £.196.f.2. 1788 (non a 1775, nec G. Don, 31) Paris ex Santo Domingo (P). 7 ae L’Héritier, ae Nov. 109 bis. ¢.52 615. 1789. Type: in Hort. Paris ex Hispaniola (G-DC, MA as photo F-29790). Sida diffusa H.B.K. Nov. Gen. Sp. 5:257. 1822. Type: MEXICO. Zelaya, Humboldt & cae n.(Bherb. Willd. no. 12673 as photo F-9799, P). Sida ramosa W illdenow engel, Syst. 3:120, 1826, pro syn. Sidi Moricand, Pl. Amer. Rar. 1:10. ¢.8. ex Bull. Sci. Nat. Géol. 23: 79. 1830 acquin, 1767). Type: ae eae giedleas 220 (OXF). Sida "Glicaulis 7 & Gray, Fl. N. Amer. 1: 838. Type: TEXAS, 1835, Drummond s. Sida filicaulis var. setosa A. Gray, Smithsonian Contr. Knowl. 5 (Pl. Wright. 2):22. 1853. Types: TEXAS, bottoms of the San Pedro, Wright s.n. (GH?); between Santa Barbara and the copper ae Wright 892 (K). Sida diffusa var. setosa (A. Gray) E. G. Baker, J. Bot. 30:291. 2 editorum Coe a ee Bot. France 71:630. 1924. Type: NEW MEXICO, Wooton 557 Sida abutifolia occurs from the southern United States, through Mexico, Central America, and the West Indies to northern South America. 11. Stipa GLomERATA Cavanilles, Diss. 1:18. ¢.2.f.6. 1785. Type: Jussiew 5.n. (P-JU, n.v ce Sida glomerata occurs widely in South America, and extends northward to the West Indies, Panama, and Costa Rica. 12. Srpa JAMAICENSIS Linnaeus, Syst. Nat. ed. x. 1145. 1759 (non Miller, 1768, nec Vellozo, 1825, nec Dietrich, 1847). Type: “jamaicensis” (LINN- 866. 10). Sida sericea Miller, Gard. Dict. ed. viii. no. 15. 1768 (non Cavanilles, 1802). Tyre: JAMAICA, 1731, Houstoun 5.2. (HOLOTYPE: BM as photo BH, CH). Sida mollis L. Richard, Actes Soc. Hist. Paris 1:111. 1792. Type: not stated, presumably at ae nL B.K. Nov. Gen. Sp. 5:258 [200]. 1822. Type: NOV. GRANAT., Honda, Humboldt & Bonpland s.n. (P as photo F-35538 nee ee Schlechtendal, Linnaea 3:271. 1828. Type: ST. THOMAS, Ehrenberg s.n. (HAL?). cae ee var. antillana Millspaugh, Publ. Field Mus. Hist., Bot. Set. 2:71. e) ida jamaicensis Vellozo (non Linnaeus). [Also cited: Acapulco, Palmer oe near Spot Bay, Grand Cayman, Mi//spaugh 1303. Sida d la Gandoger, Bull. Soc. Bot. France 71:630. 1924. Type: ST. THOMAS, Eggers 195. Sida guadalupensis Gandoger, Bull, Soc. Bot. France 71:630. 1924. Type: GUADE- LOUPE, Duss 3210 (GH). Sida cydonifolia Gandoger, Bull. Soc. Bot. France 71:630. 1924. Type: STO. MI : Sida jamaicensis occurs in Mexico, Central America, the West Indies, and Colombia. 13. SipA NUMMULARIA E. G. J. Bot. 30:290. 1892. Type: CUBA. Iste oF Pines, Milne 85 (HOLOTYPE: K). Sida nummularia is endemic to the Isle of Pines. 14. Sipa sprnosa Linnaeus, Sp. Pl. 683. 1753. Tyee: “spinosa” (LINN-866. 1). A photo of the type is reproduced by Rodrigo (1944, plate 13). Sida alba Linnaeus, Sp. Pl. ed. ii. 960. 1763. Type: Hort. Upsal. “alba” (LINN-866. 2). Sida ulmifolia Miller, Gard. Dict. ed. viii. no. 1. 1768 (non Cavanilles, 1785). Type: cult., Miller herb. (no. 1) (BM)}—mounted with a Wright specimens (no. 2) from Jamaica. Sida ie Miller, Gard. Dict. ed. viii, no. 3. 1768. Type: in cule., Ma/ler s.n. (HO BM). Sida milleri oe Prodr. 1:472. 1824 (as S. miller). Sida Ping Miller, Gard. Dict. ed. viii. no. 4. 1768. Type: (fide Fawcett & , 1926): (BM). Sida reuiei Lamarck, Encycl. 1:4. 1783 (non Miller, 1768, nec Medikus, 1783). es (HOLOTYPE: P-LA). Sida spinosa var. angustifolia Gana ae bach, Fl. Brit. W. I. Sida pusilla re Diss. 1: 6. i L fA. 1785. Types: Jussien 5.n.; Thouin s.n. Sida truncata L’Héritier, Stirp. Nov. 107. ¢.57. 1789 (non Cavanilles, 1785). Type: in horto ex Hispaniola, Saint-Germaine 5.n. (presumably at G). Sida emarginata Will- denow, Sp. Pl. ed. itt. 3:757. 1800. 76 Sida linearis Cavanilles, Icon. 4:6. ¢.312.f.1. 1797. Type: South America ex Cavanilles (C as photo F-21603) or plate 312, fig. 1. The type is not Rimat 1795 (photo F), as stated by Monteiro, because it was not cited by Cavanilles. Sida hyssoptfolta Presl, Reliq. Haenk. 2:109. 1835. Type: MEXICO, Haenke s.n. (PR?). Sida angustifolia var. major Presl, Relig. Haenk. 2:109. 1835. Typr: MEXICO. in portum Acapulco, Haenke s.n. (PR?). oe minor Macfadyen, FI. Fee 1:79. 1837. Type: JAMAICA, Port Royal Moun- Siesta eee Ann. Sci. Nat. Bot. 18:50. 1842. Type: BRAZIL. , Claussen 5.n. (P as photo F-35548). oe ir Hooker, Trans. Linn. Soc. London 20:232. 1847. Sida heterocarpa Englemann ex Gray, Boston J. aa Hist. 6 (PI. Lindh. 2):163. 1850. Type: TEXAS, Houston, Lindheimer s.n. ( Sida affinis Schmide, Beitr. Fl. Cap. Verd. Ins. . 1852 (non Sprengel, 1826). Sida spinosa occurs from the central United States to central Argentina and is also known in the Old World. It is a common, often weedy species. 15. Stipa viarum St.-Hilaire, Fl. Bras. Mer. 1:182. 1827. Type: BRAZIL. Minas Gerags, Comarca do Rio das Mortes, S?t.-Hilaire 5.n. (P). Sida viarum is a South American species that extends northward through Central America as far as southern Mexico. SIDA section MUTICAE bt pay Haenk. 2:104. 1835. Lecroryes (here designated): Sida agegregata Pre Presl’s section ts a ae group of ten species, including species now segregated to genera other than Sida. Presl’s concept is narrowed by the exclusion of most of his species and is stabilized by the selection of a lectotype species. As here understood, the section is monotypic. Shrubs to 1.5 m tall, often setiferous. Leaves petiolate, ovate to weakly 3-lobed, serrate, tomentose. Flowers subsessile in spiciform aggregations that are branched to form paniculate inflorescences; calyx prominently 10-angled, usually hirsute; corolla yellow-orange, sometimes with a red center. Mericarps 5— 8, submuticous, essentially unornamented. 16. SipA AGGREGATA Presl, Relig. Haenk. 2:106. 1835. Type: MEXICO, Haenke 5.n. (PR?). Sida cb oe Presl, saat Haenk. 2:105. 1835. Type: in terris mexicanis occidentalibus, . (PR?, as photo F-32642). as Prest, sat Haenk. 2:106. 1835 (non Swartz, 1788). TypE: MEXICO, ad m Ac oe , Haenke 5.n Sida savannarum Schumann ” Mart. Fl. Bras. 12(3):308. 1891. Type: BRIT. GUIANA, ae 819 (B as photo F-9395). vig Sida aggregata occurs in Mexico, Central America, the West Indies, and northern South America. SIDA section CORDIFOLIAE (DC.) Fryxell, stat. nov. Basionym: Sida {subsect.] Cordifoliae DC. Prodr. 1:463. 1824. Type (Article 22): Sida cordifolia L. Erect shrubs or subshrubs, densely stellate-tomentose. Leaves petiolate, lanceolate to ovate, dentate-crenate throughout (Fig. 4). Flowers solitary in the axils, often aggregated terminally in racemes or panicles; calyx 10- ribbed, tomentose; corolla yellow or yellow-orange, the filaments more or less organized into 5 phalanges. Mericarps usually 7 or more, laterally reticulate, muticous to prominently aristate, the spines often retrorsely barbed In addition to the North American species treated here, the section Cordifoliae occurs in Australia (S. rohlenae Domin and S. atherophora Domin) and includes a number of South American species, such as §. angustissima St.-Hilaire, S. cerradvensis Krapovickas, and several others. Szda cordifolia is pantropical. KEY TO THE NORTH AMERICAN SPECIES OF SIDA SECTION CORDIFOLIAE a. Stems and pedicels viscid; corolla yellow-orange, fading rose; mericarps plabrous:Onespatsely puvescent sacs heel au Tes awe tedaeey es ein aes a. Stems and pedicels stellate- on not viscid; corolla usually yellow o yellow-orange, sometimes with a red center; mericarps usually Seed b. Calyx 7— 1 lon c. Calyx irregularly 6 Sebel .tu cu Neue assess eoeoe ss c. Calyx regulary 5-lobed. Ae Mericarps MMT COUSs 41 eon Wade eke he eee aoe ee 22. S. tragiifolia d. Mericarps Abies! apically, the dehiscent portion ca. half the length of the mericarp, the wall hyaline...........-.... 19. S. hyalina b. Calyx 5 — e. Corolla twice length of calyx, with red center..............20. S. maculata e. Corolla barely exceeding calyx, without red center ers and fruits crowded in terminal Hanicalate nee 23. S. xanti 17. S. barclayi yx prominently 102tibbed <4.4.ccdsaweds eee tein wens ._ S. cordifolia f ee axillary; calyx not prominently ribbed.......... oe S. salviifolia 17. Sia Barctayi E. G. Baker, J. Bot. 30:236. 1892. Type: SAN SALVA- DOR, Gulf of Fonseco, Sierra de Conchagua, Barclay s.n. (HOLOTYPE: BM). Sida anomalocalyx Fryxell, Syst. Bot. 4:255. 1979. Type: MEXICO. Cuiapas, Mpio. de La Trinitaria, 18 km SW of La Trinitaria, Breedlove 42230 (HOLOTYPE: DS; IsOTY PES: , pf). Sida barclayi is known from southern Mexico (Oaxaca and Chiapas) and Central America (El Salvador, Nicaragua, and Costa Rica). i ( Figure 4. Leaf outlines of Sida sect. Cordifoliae. A, 8. barclayi; B-C, S. xanti; D, S. — E,S. hyalina; F,S. cordifolia; G-I, 8. salviifolia. (A, Stevens 11108; B-C, Sanders et ,D, Palmer 103; E, Gentry 14301; F, — 725, G, Fryxell 1155a; H, Koch & ses ey I, Lott et al. 1552.} Scale = ie, 18. Siva corpirouia Linnaeus, Sp. Pl. 684. 1753. Tyee: “cordifolia 4” (LINN-866. 12). Sida micans Cavanilles, Diss. 1:19. ¢.3.f.1. 1785. Type: SANTO DOMINGO, Jussieu Sida althacifolia Swartz, Prodr. Veg. Ind. Occid. 101. 1788. Type: JAMAICA, Swartz SM, oe OTYPE: S). ee var. ae (Swartz) Grisebach, Fl. Brit. W. I. 76. 1859. Sida one Link, Enum. Pl. 2:207. 1822 (non Salzmann ex Triana & Planchon, 1862, & Mocino, 1894). eer oo Sida cordifolia var. conferta (Link) Coe re W. I. 76. Sida pellita H. . Nov. Gen. — a an {205}. 1822. Type: ad Orinocum prope Angostura, pies G a sun. (P) Sida pungens H.B.K. Nov. Gen. 5:263 [204]. 1822. Type: ad Orinocum prope An Ae es Humboldt & Bland s.n. (B herb. Willd. no. 12176). Sida artstata Willdenow ex Sprengel, Syst. 3:116. 1826, pro syn. Sida Nee W illdenow ex enieneel: Syst. 3:112. 1826. Type: Cumana, Humboldt & Bonpland s.n. (B. herb. Willd. no. 12693). Sida eg Schumacher & Thonning ex ean aaagee ee ve - ee 1827; ongel Danske Vidensk. Selsk. Naturvidensk. Afh. 4:81. 1829. Type: ae 119 - “no type specimen traced” ee ee p. Sy Sida ampla M. E. Jones, Contr. W. Bot. 15:146. 1929. Type: MEXICO. Nayarit, Acaponeta, Jones 22858 (HOLOTYPE: POM; tsoTyPeE: CAS) Numerous other synonyms are given by Schumann (1891), Kearney (1954), and Borssum Waalkes (1966). Sida cordifolia is pantropical and subtropical in distribution. 19. Sipa HYALINA Fryxell, Sida 7:227. 1978. Type: MEXICO. Sinatoa, Cerros de Navachiste about Bahia Topolobampo, Gentry 14301 (HoLoTyPE: LL; soryPes: DE US) Sida hyalina is known from northwestern Mexico (Sonora and Sinaloa) from the vicinity of Guaymas to the vicinity of Cultacan. 20. Sioa MACULATA Cavanilles, Diss. 1:20. ¢.3.f.7. 1785. Tyrer: SANTO DOMINGO, ae n. oa -JU 12266). Sida cordifolia ssp. maculata (Cavanilles) Marais, Kew Bull. 38: 198: Sida suberosa LW Héritier, Stirp. Nov. 5:113. 1.54 1789 (non Dietrich, 1847). Type: Hort. Paris ex Hispaniola (P?). Sida maculata occurs in Cuba and Hispaniola. It is questionably distinct from S. cordifolia. Marais (1983) cites it from the Mascarenes. 21. Stipa saLvuro.ia Presl, Relig. Haenk. 2:110. 1835. Type: MEXICO, prope Acapulco, Haenke 5.n. (HOLOTYPE: PR; isoTyPE: BM). Sida spinosa var. salviaefolia (Presl) E. G. Baker, J. Bot. 30:237. 1892. Sida erecta Macfadyen, Fl. Jamaica 1:80. 1837. Type: JAMAICA, near Half-Way Tree (lectotype: K) Sida campestris Bentham, Pl. Hartw. 113. 1843. Type: ECUADOR. Guayagut., Hartweg 034 (OXF). Sida sea gee var. moritziana K. Schumann, Mart. FI.Bras. 12(3):336. 1891. Type: RANAT., prope Maracaybo, _ — pie E. G. Baker & Rose, Cc * Natl. Herb. 5:176. 1899. Type: iXICO. Moreos, Cuautla, Ho/way ep ee US; isoryprs: BM, GH). Sida salviifolia occurs in Mexico, the West Indies, and parts of South America 22. SipA TRAGUFOLIA A. Gray, Boston J. Nat. Hist. 6:164. 1850. Tyee: Cambridge (Mass.) Bot. Gard. ex southern Texas, Gray s.n. (HOLOTYPE: GH). Sida tragtifolia occurs in northeastern Mexico and southern Texas. eB ten xANTI A. Gray, Proc. Amer. Acad. Arts 22:296. 1887. Tver: EXICO. Baja Catirornia, Cape San Lucas, Xantus 8 (HoLoTYPE: GH; isorypers: K, se US). Sida xanti occurs on the Baja California peninsula and on some of the adjacent islands. SIDA section MALACHROIDEAE G. Don, Gen. Hist. 1:498. 1831. Lectotype: Sida anomatla St.-Hilaire (lectotype designated by Fryxell, 1975). Sida sect. Psendomalachra K. Schumann in Engler & Prantl, Nat. Pflanzenfam. 3(6):43. 1890. Type: Sida ciliaris L. In generic rank: Pseadomalachra (Schumann) Monteiro, Portugal. Acta Biol. B, 12 (1-4): 141. 1974 Sida [sect.} Ciliares Small, Man. Southeast. Fl. 849. 1933. Type: (Article 22): Sida iaris L CHELaYIS Perennial herbs or subshrubs, the stems prostrate to ascending, stellate- pubescent or scabrous to glabrescent. Leaves short- agra ae lanceolate to linear, basally entire, distally dentate (Fig. 1, A—E), extreme cases entire throughout (Fig. 1, E). Flowers congested at the apices of the branches through abrupt shortening of internodes, the pedicels short, adnate to petioles, crowded with the stipules so as to seem involucellate; petals yellow to rose. Mericarps 5— 8, smooth to strongly muricate. Section Malachroideae includes the South American species S. centuriata Clements, S. surumuensis Ulbrich, S. paradoxa Rodrigo, and S. plumosa Cavanilles and the Old World species S. cuneifolia Roxb. , in addition to the North American species treated here. KEY TO THE NORTH AMERICAN SPECIES OF SIDA SECT. MALACHROIDEAE a. Petals 16-23 mm long; leaves 4— 10 times as long as broad, entire obscurely tridentate at apex (Fig. 1, E).....................2 es ain a. Petals 5 — 13 mm long; leaves generally broader, manifestly dentate at apex (Fig. 1, A—D). b. Leaves broadly oblong or obovate, long-hirsute on both surfaces (Fig. A) epetalssyellow srl? ls smi lohtiers s vx cy a@ een anedoneen ce S. brittonit b. Leaves variable in shape, appressed-pubescent to glabrate (Fig. 1, B-D); petals: commonly rose; 5:= 11-13) mm long «2.0.4 sees eee ee 267 8. aris 24. SIDA BRACHYSTEMON DC. Prodr. 1:459. 1824. Type: Icon. Fl. Mex. s.n. (Torner Collection acc. no. 6331.1803, Hunt Institute). Sida wendtii Fryxell, Phytologia 46: 393. 1980. Type: MEXICO. Oaxaca, Mpio. de menta, terraceria a Santo Domingo, 29 km al oeste de Pinotepa Nacional, Koch, Fryxell & Wendt 79424 (HoLorypr: ENCB; isorypes : CAS, CHAPA, CTES, F, K, MARY, MEXU, MICH, MO, NA, NO, NY, TEX. WIS, XAL, pf. Sida brachystemon is at present known from isolated collections from Belize, Costa Rica, and Mexico. 25. SIDA BRITTONI Leon, Torreya 19:172. 1919. Type: CUBA. Pinar be Rio, Chirigota, Leon & Roca 7466 (GH). Sida brittonit ts endemic to Cuba. 26. Supa citiaris Linnaeus, Syst. Nat. ed. x. 1145. 1759. Type: JAMAICA, Browne s.n. (LINN-866.8). [See comment on type by Clement (1957, p. 24); the type is not 866.6 as stated by Borssum ioe (1966).} Pseadomalachra ciliaris (L.) Monteiro, Portugal. Acta Biol. B, 12 (1-4):1 1974. Sida tridentata Sag Icon, fa t,312.f.2. 1797. Tyee: in insul. Sancti Dominici, mae n.(P-JU). i a trider (Cav.) Monteiro, Portugal, Acta Biol. B, : 1974. Pe muricata Cavanilles, Icon. 6:78. £.597./.2. 1801. Type: Nova Hispania [Vera- ruz?}, prope Chalma, Nées.2. (MA as photo F- oo ‘inseaced by Rodrigo, 1944, late 3, mislabeled as type of S. ciliaris L.). Sida fulva St.-Hilaire, Fl. Bras. Mer. 1:176. 1827. Type: Manguinhos, prov. Rio De Hae St.-Hilaires.n. (MA as rc F- re P). Sida ciliaris var. fulva (St.-Hil.) Schumann, Mart. Fl. Bras. 12(3):2 Sida anomala St. iii. Fl. Bras. Mer ] - . Type: in prov. Cisplatina, pres le ill ando, S¢t.-Hilaire e 2/ 2470 ae as oe F-35529 and in Rodrigo, 1944, plate 2). Sida aos var. anomala (St.-Hil.) Schum. in Engler & Prantl, Nat. Pflanzenfam. 3(6):43. 1890. Sida anomala vat. mexicana 1 Morican d, - Nouv. Amér. 36. ¢ 1837. Type: MEXICO, ea , Berlandier 66 (BM, G, OXF). Sida ae var. mexicana (Moricand) See Field & Lab 21: 94. 1953. Bel pe & Gray, Fl. N. Amer. 1: oe = 1838 (non Willdenow ex Sprengel, . Type: TEXAS, Drummond 47 (BM F). Sid. moat Richard, Hist. Phys. Cuba, PL. 8 63 (p. 162, French ed.). 1845. CUBA, de la Sagra s.n. (F, P). 82 Sida erosa Salzmann ex Triana & Planchon, Ann. Sci. Nat. Bot. 17:176. 1862 (non Link, 2), pro syn oe linearifolium Buckley, Proc. Acad. Nat. Sci. Philadelphia 13:449. 1862. DN S, May 1861, Buckley 5.n. (HOLOTYPE: PH). rn aes Gandoger, Bull. Soc. Bot. France 71:629. 1924. Type: MEXICO. JALISCO, near Guadalajara, Pringle 4497 (MEXU, MICH, PH, US, VT, pf). Sida ciliaris occurs in the United States, the West Indies, Central and South America, Africa, southeastern Asia, and Fiji. I¢ is common, weedy, and highly variable. Consequently it has been handled differently by different authors, sometimes being broken up into several varieties. For example, Clement (1957) recognized five varieties within S7da ciliaris. In my experience, the variability in this species is continuous, not discrete, and I am unable co justify or distinguish more than a single taxon, recognizing that it is a highly variable one. SIDA section ELLIPTICIFOLIAE Fryxell, sect. nov. Type: Sida rzedowskii Fryx. Suffrutices ascendentes vel erecti, uaa vel glabri. Folia breve petiolata, late vel anguste elliptica vel linearia, ubique dentata (Fig. 5). Pedicelli solitarit in axillis foliorum (longissimi vel ; revi), secus caulem dispersi vel apicem versus tc ob er? abbreviata; corolla flava vel lilacina (centro flava). Mericarpia (5S-) 8— 11, mutica vel acuta vix spinifera, lateralicer reticulata vel laevia. Ascending to erect subshrubs, pubescent to glabrate. Leaves short- petiolate, broadly or narrowly elliptic to linear, dentate throughout (Fig. 5). Pedicels solitary in the axils (long or short), scattered along the stem or congested apically through shortened internodes; corolla yellow to rose- lavender with a yellow center. Mericarps (5-) 8 — 11, muticous to acute but scarcely spined, laterally reticulate to smooth. KEY TO THE SPECIES OF SIDA SECT. ELLIPTICIFOLIAE a. Pedicels up to 15 cm long, usually more than twice the length of the subtending leaves. b. Leaves sea lanceolate, 7 — 10 (-15) times as long as wide (Fig. 5, H); Gale Oi ONG panes on aan ee aA ee oe EY 30. S. longipes b. oe broadly elliptic, 1.2 —2 times as long as wide (Fig. 5, 1); calyx 9= Th inm long i) enka cid beats ended’ Pelanaeh he dias 32. S. potosina a. Pedicels no more than 6 cm long, usually shorter than the subtending leaves. c. Calyx 7— 10 mm long d. Pedicels 2—6 cm long, often equaling the subtending leaf. 29. S. lindhermeri d. Pedicels less than 2 cm long, shorter than subtending leaf .... 28. S. inflexa c. Calyx 5— 7 mm long. e. Leaves elliptic, 1.5 —6 times as long as broad (Fig. 5, L—P). f Corolla r rose or purple (with yellow center); mericarps 8 — 11; leaves as long as broad; flowers and fruits saatltedly con- gested apically ign de Aa e oa dee 4 eee epee 33. S. rzedowskit Dol ie ! Figure 5. Leaf outlines of Sida sect. Ellipticifoliae. A-C, S. lindheimeri; D-F, S. elliottit; G, S. inflexa; H, S. longipes; 1, S. potosina; J-K, S. neomexicana; L-M, S. rzedowskii; N-P, S. turneroides. [A, Hill 5353; B, Wilkinson 285, C, Hill 10609; D, Hernandez Magana 5127; E Dorantes & Acosta 2156; F, Hill 5310; G, Fernald & Long 11373; H, Wright 50; 1, Purpus 4906; J, Sanders et al. 2970; K, Spellenberg & Soreng 6845; L, Rzedowski 32004; M, Ventura 245; N, Fryxell & Anderson 3613; O, Fryxell 3823; P, Fryxell & Anderson 3644.] Scale = 1 cm. P poe f. Corolla yellow; mericarps 5—8; leaves 1.5 —3 times as long as broad, sharply serrate; Toe “ind fruits slightly congested oe call . 3. turneroides Jnee narrowly lanceolate to linear, 10 — 20 times as long as an Fig. 5, D K). g. Plants oy branching from the base; flowers and fruits apically LOT ON 126 ane ao a an aed ae 1. S. neomexicana g. Plants few stemmed; flowers and fruits little if at all congested 27. S. elliottit Ay 252 ob ae bY ue ian fue ede eee ee 2 27. Sipa ELLiIoTTu Torrey & wink Fl. N. Amer. 1:231. 1838. pale M: Sida gracilis Elliott, Sketch Bot. 2:159. 1822 (non Richard, ae Typr: SOU CAROLINA, near Peaiene Elliott Sm. (HOLOTYPE: CHARL). 7 ee Nash, Bull. Torrey Bot. Club 23: 102. 1896. Type: Florida, Tampa, 472 s AIS), i; Sida siasabel Saul. Bull. osiey Bot. Club 25:468. 1898. Type: ex Torrey herb. 5.co/d. Sida elliottii occurs in the southeastern United States from North Carolina south to Florida and west to southernmost Missouri, Arkansas, and Texas; it occurs in Mexico from Nuevo Le6n and Tamaulipas south to Veracruz and Chiapas, and extends to Guatemala. 28. SIDA INFLEXA Fernald, Rhodora 42:463. 1940. Typr: VIRGINIA. South- hampton Co., near Three oe northwest of Carey Bridge, Fernald & Long 11373 (HOLOTYPE: GH: ISOTYPES: M , US). Sida inflexa is confined to the southeast corner of the state of Virginia. 29. SIDA LINDHEIMERI Engelmann & Gray, Boston J. Nat. Hist. 5:21 1845. Tyee: TEXAS, prairies east of the Brazos [River], fascicle 1, 1843, Lindheimer 24 (HOLOTYPE: GH, Isorypes: K, OXF). Note: A second collection by Lindheimer, also numbered 24, is part of fascicle H, 1844, and is not type material. Sida ellittit var. texana Torrey & Gray, Fl. N. Amer. 1:681. 1840. Type: TEXAS, Drummond 14 (BM, OXE). Sida texana (Torrey & Gray) Small, Fl. Southeast. U.S. 772. 1903 jos Sida lindheimeri occurs principally in central Texas and in Louisiana and sporadically in Mexico. 30. Stipa LONGipes A. Gray, Smithsonian Contr. Knowl. 3 (art. 5, Pl. Wright. 1):19. 1852 (non Meyer ex Harvey & Sonder, 1860). Type: TEXAS, prairies of Live Oak Creek, Wright 50 (HOLOTYPE: GH; tsorypes: OXF, US). Sida longipes occurs in Coahuila and western Texas. 1. SIDA NEOMEXICANA A. Gray, Proc. Amer. Acad. Arts 22:296. 1887 85 (non Gandoger, 1924). Type: NEW MEXICO, on mountains at the Copper Mines, Wright s.n. (PH, Sida elliottii var ?humilis A. Gray, Smithsonian Contr. Knowl. 5 (art. 6, Pl. Wright. 2):21. 1853. Type: [the same as for S. neomexicana}. Sida neomexicana occurs in western Texas, southern New Mexico, Chi- huahua, Coahuila, and Durango. 32. eae POTOSINA Brandegee, Univ. Calif. Publ. Bot. 4:184. 1911. Tyee: XICO. San Luis Potosi, Minas San Rafael, Purpus 4906 (HOLOTYPE: UC; IsOTYPES: No. US) Sida potosina occurs in the Mexican states of San Luis Potosi and Tamaulipas. 33. Sipa RZEDOWSKI Fryxell, Sida 8:125. 1979. Type: MEXICO. Hipaico, Cerro Ventoso, entre Pachuca y Real del Monte, Rzedowski 20500 (HOLOTYPE: ENCB). Sida rzedowskii occurs from Jalisco to Chiapas, at elevations of 2000 to 2700 m, being relatively common in the Valley of Mexico. 34. SIDA TURNEROIDES Standley, Publ. Field Mus. Nat. Hist., Bot. Ser. 22:90. 1940. Type: MEXICO. Tamautipas, Jaumave, Sierra near San Lucas, von Rozynski 514 (HOLOTYPE: F). Sida turneroides occurs in the Mexican states of Hidalgo, San Luis Potosi, and Tamaulipas. SIDA section SIDAE. Lectotype: Sida alnifolia Linnaeus. Erect subshrubs, pubescent or puberulent to glabrate. Leaves short- petiolate, more or less rhombic to lanceolate or elliptic, basally entire, distally crenate-serrate, acute or obtuse (Fig. 6). Pedicels usually solitary in the leaf axils, sometimes aggregated apically, sometimes in axillary pedunculate clusters; calyx prominently 10-ribbed, the ribs often yellowish at base; corolla yellow, with or without a red center. Mericarps 6— 13, usually apically 2-spined, laterally reticulate. Sida sect. Sidae includes species from Africa (e.g. 8. dreger Burtt Davy), South America (e.g. S. g/aziovii Schumann), Asia (e.g. S. szechuensis Matsu- da), and Polynesia (e.g. S. fallax Walpers), in addition to the North American species treated here. KEY TO THE NORTH AMERICAN SPECIES F SIDA SECT. SIDAE a. Leaves and ace oe pattern distichous; stipules usually ae SEVELAL-Vel Mears s pce 8 Means A ee pga hod. nn ne es acne memes. ye Sk meee _ 3. acuta a. Leaves eae disposed; stipules linear or subulate, usually 1 (-3)- Sa: ore v0f 1 mile Figure 6. Leaf outlines in Sida sect. Sidae. A, S. santaremensis; B-D, S. setosa; E-G, S. collina; H-J, 7 ba geen K-L, S. rbombifolie, M-N, S. haenkeana; O-Q, S. acuta. [A, Crewz si 2271; B, Gentry et al. 29193; C, Dwyer 10280; D, ee 22877; E, Fryxell et al. 20; F. a & Strother 40577; G, Koch et al. 79469; H, Hill 11074; 1, Correll & Proctor We 9; J, Avery 2020; K, Fryxell 2091; L, Koch et al. 79329; M, Koch & Fryxell 77331;N, Koch & Fryxell 77384; O, Chavez s.n.; Q, Breedlove 20929A, R, Gandara & Dorantes 113.} Scale = 1 cm. b. Calyx 7— 10 mm long; flowers often aggregated apically more or less above the leaves; leaves often loosely pubescent beneath; mericarps muticous. c. Corolla yellow with purple center; pedicels twice length of calyx or less; mericar Bs Oe Di ssala i acy ok SES ig RS Ch cen ee ae S. collina ee is Ss BG ease aia Bla gis tat ah acd teed Oo ae 38. i . haenkeana b. Calyx 3 —7 mm long; flowers commonly scattered along the stem; leaves minutely puberulent beneath; mericarps muticous or beaked or aristate. . Mericarps aristate, the 2 aristae capillary, curled, subequal to body of merce powers sometimes borne in axillary, pedunculate, sub- eaves ee large (up to 10 cm long or more), lanceolate and hey acute (Fi2.-0, BD) d.cceor geile eee 41. S. setosa . Mericarps muticous to spinescent, the spines (when present) less than half che length of ee flowers 1 or more in the leaf axils; leaves seldom more than 6 cm e. Leaves 0.5 — fou Qo : f. Flowers subsessile in the axils; calyx 5—6 mm long; leaves Fotund t6/Elliptic- ioc. enedeensie he ke ene eee t 36. S. antillensis f. Pedicels | — 2 cm long; calyx 3 —4 mm long; leaves oblong or OblanceG late. 2u4 52% shew ees bed eee ccees sine eed 42. 8. troyana e. Leaves generally 2—6 cm long, more or less rhomboid. olla cream-colored with red center; leaves dentate almost to the base; pubescence (e.g. on petioles) up to 0.5 mm long Rassias Ghee tle tee oe ena ada es oe a 40. S. santaremensis g- Corolla yellow without dark center; leaves generally entire in basal third (or more); pubescence never more than 0.1 mm WONG 35258 5 ha teh eee ae teee ek ek see as 39. S$. rhombifolia 35. Supa acuta Burman fil. Fl. Indica 147. 1768. Type: PERU (G-Del as photo F-7558); Java, s.coll. s.n. (Lectotype: G). Lectotype chosen by Borssum Waalkes (1966). Sida carpinifolia var. acuta (Burm. f.) Kurz, J. Asiat. Soc. Bengal, Pt. 2, Nat. Hist. 45:119. 1876 Sida carpinifolia Linnaeus fil., Suppl. 307. 1781 (non Miller, 1768). Type: MADEIRA, Masson s.n. (BM?). Sida acuta var. carpinifolia (L. F.) Schumann, Mart. Fl. Bras. 12(3):326. 1891. Sida frutescens Cavanilles, Diss. 1:12. ¢.10.f.1. 1785. Type: ex R. Hort. Paris (P). Sida sacle Link, Enum. Pl. 2:203. 1822. Based on: Sida ulmifolia Willdenow, 1. Hort. Berol. Suppl. 49. 1814, nom. nud. Type: herb. Willd. no. 12654 © Sid apple var. spiraeifolia (Link) Millspaugh, Field Mus. Bot. Ser. 2:72. ei (aos ee Bull. Soc. Nat. Mosc. 31:197. 1858. Type: MEXICO, Berlandier 49 (HOLOTYPE: KW aoe aces ne fil. var. eemcuserg Grisebach, Fl. Brit. W. I. 73. 1859. CUBA, Wright 1565 (PH). Note: Grisebach did not explicitly cite a type, ee ne (Grisebach, 1866, p. 24) cites the Wright collection as representing var. brevicuspidata, which constitutes a lectotypification. Note: Borssum Waalkes (1966) cites numerous additional synonyms. 88 Sida acuta 1s pantropical in distribution, generally below 1500 m eleva- tion, where it is a prominent component of the weed flora. 36. SIDA ANTILLENSIS Urban, Symb. at 5:418. 1908. Type: 10 specimens are cited by Urban; lectotypification is need Sida antillensis occurs in the West Indies, southern Florida, and parts of Central America. Many authors have submerged this species in S. acuta, but Correll & Correll (1982) maintain it as distinct. 37. Sipa COLLINA Schlechtendal, Linnaea 11:364. 1837. Type: MEXICO {Veracruz}, prope Hacienda de la Laguna, Jul 1829, Schtede s5.n. (HOLOTYPE: HAL; as photo F-9377) Sida costata Schlechtendal, Linnaea 11:365. 1837. Type: MEXICO pinta: pee Hacienda de la Laguna, Jul 1829, Schiede 5.n. (HOLOTYPE: HAL; isorypes: GOET, AL; as photo F-9377). Sida pes R.E. Fries, Bull. Herb. Boissier 7:988. 1907. Type: MEXICO [Vera- cruz}, region d’Orizaba, Bourgean 2863 (G, Sida collina occurs in the Mexican states of Sinaloa, Nayarit, Jalisco, Veracruz, and Oaxaca and extends into Central America 38. SIDA HAENKEANA Presl, Relig. Haenk. 2:104. 1835. Type: in parte occidentali Mexici, Haenke 5.2. (HOLOTYPE: PR; IsoTyPE: MO). Sida woronowit Ulbrich, Notizbl. Bor. Gart. Berlin 11:536. 1932. Type: MEXICO, Michoacan, prope Uruapan, Woronow 2890 (B, LE). Sida haenkeana occurs at elevations of 1000 to 2800 m in Mexico VJalisco to Chiapas), Nicaragua, and Costa Rica. 39. SIDA RHOMBIFOLIA Linnaeus, Sp. Pl. 684. 1753. Type: “2 rhombifolia” (LINN-866.3); as photo in Rodrigo (1944, plate 28). Sida rhombordea Roxburgh ex Flem. As. Res. 11:178. 1810; Hort. Beng. 501. 1813. s. loc., 5.coll, 2228 (LECTOTYPE: BR). ey en var. rhomboidea (Roxb). Masters in Bose Fl. Brit. India 1:323. 18 Sida ae. H.B.K. Nov. Gen. Sp. 5:261 205) "1922. Tyee: NOV. GRANAT., rope Honda, Humboldt & Bonpland 1709 ( Side ruderata Macfadyen, Fl. Jamaica 1:81. 18 7 . Type: JAMAICA gaa ae ae ae Bull. Soc. Bot. France 71: 631. 1924. Type: MEXICO. JALISCO, io Hondo, Pringle 4095. Sida ae Marais, Kew Bull. 38:42. 1983. Type: MAURITIUS, Moka, Ayres 5.2. (HOLOTYPE: ). Sida unicornis Marais, Kew Bull. 38:42. 1983. Type: MAURITIUS, Bouton 5.n. (HOLOTYPE: K). Sida rhombifolia is virtually pantropical in distribution and reaches the temperate zones to a limited extent as an annual. 89 40. SIDA SANTAREMENSIS Monteiro, Monogr. Malv. Bras. Fasc. I. Gen. Sida 44. 1936. Type: BRAZIL. ParA, Santarem. Sida santaremensis is a South American species (Brazil, Argentina, Boli- via) recently discovered in the vicinity of Tampa, Florida (Fryxell et al., 4l. ee SETOSA Martius ex Colla, Herb. Pedem. 1:416. 1833. Type: AZIL, Rio Belmonte, Martius 5.2. (HOLOTYPE: TO) Sida kohautiana Presl, Reliq. Haenk. 2:108. 1835. Type: IND. OCCID., Martinique, Kohaut s.n. ; Sida surinamensis Miquel, Linnaea 22:469. 1849. Type: SURINAM, Hostman 1079 (Bas photo F-9396, K, pf) Sida setosa is a South American species that extends northward into Panama. 42. SIDA TROYANA Urban, ae Antill. 5:419. 1908. Type: JAMAICA, prope Troy, Harris 8805 (BM?, Urban (loc. cit.) suggested that this species has its affinity with S. rhombifolia, and Fawcett & Rendle (1926) suggested that “it may be perhaps a depauperate form of S. rhombifolia."’ Adams (197 2) reduced it to synonym. However, it is sufficiently distinct (cf. key) to be recognized taxonomically, especially in the small calyces, which are this small in only a few other species (e.g. the South American S. serrata Willd.). ACKNOWLEDGMENTS Iam grateful to A. Krapovickas for sharing ideas and clarifying several points concerning species identities and other matters. The responsibility for the interpretations presented here, however, is entirely my own. APPENDIX I Species frequently included in Sidg (Kearney, 1954; Clement, 1957; various floras) that are here excluded S. acuminata DC. = Sidastrum multiflorum Jacq. — Ss eae A. Gray = Billieturnera helleri (Rose) F S. eggersii E. G. Baker = Abutilon virginianum Kiapov. S. filipes A. Gra = Meximalva filipes (A. Gray) Fryx. S. gr m. ex Kearn. = Billieturnera helleri (Rose) Fryx. S. hederacea (Doug!.) Torr. = Malvella leprosa (Ortega) Krapov. S. helleri Rose = Billieturnera helleri (Rose) Fryx. S. hilariana Pres] = Allosidastrum as oo Krap., Fryx. & Bates S. integrifolia Sessé & Mocifio = Anoda pentaschist, S. interrupta DC. = Allosidastrum sia oe ) Krap., Fryx. & Bates So - — S. lepidota A. Gray = Malvella lepidota (A. Gray) Fryx. S. lodiegensis E. G. Baker = Sidastrum lodiegense (Baker) Fryx. S. mexicana Scopoli = / (L. hleche S. micrantha St.-Hil. = Sidastrum micranthum (Sc.-Hil.) Fryx. S. multiflora oo = Sidastrum multiflorum Jacq.) Fryx. S. oxyphylla DC. = sy saad sesser (Lag.) Bates S. palmeri E. Baker = Meximalva venusta (Schlecht.) Fryx. S. physocalyx A. Gray = Aiea da physocalyx (A. ng Fryx. S. paniculata = Sidastrum paniculatum (L.) F S. pyramidata Cav. = Allosidastrum pyramidatum jee ) Krap., Fryx. & Bates S. guinquenervinm Duchass. = Sidastrum quinquenervium (Duchass.) Baker S. sabeana Buckley = Melochia pyramidata L. S. sagittifolia A. Gray = Malvella sagittifolia (Gray) F S. standleyi Clem. = Krapovickasia physaloides (Presi) Fryx. S. stricta Standley = Sidastrum strictum (Standley) Fryx. S. tehuacana Brandegee = Sidastrum tehuacanum (Brandegee) Fryx. S. triloba Sessé & Mocino = Allowtssadula sesset (Lagasca) Bates APPENDIX I Species of North America Sida doubtful or inadequately known >». amatlensis Sesse & Mocino, Pl. Nov. Hisp. 110. 1887. [Mexico] anoda Sessé & Mocino, Pl. Nov. Hisp. 109. 1887. [Mexico - Anoda cristata ?} . bicallosa sn FI. badow. 91. IS 17 . [USA - Sida spinosa ?} bicolor Cav. Icones 4:6. t.311. 1797. [Mexico - Anoda pentaschista ?} 7 —— ea Rafin. Fl. Ludov. 90. 1817. [USA - Sida efftottis fide Ewan, but flowers too lar cali ey ie — ed. vill. no. 2. 1768 [non Linn. f., 1781} deflexa C ienc. Nat. 6:337. 1803. [Cuba] gracilis . i i a 5 Actt. 311. 1797. — - Anoda pentaschtsta ?) astifoll é & Mocino, Fl. Mex. ed. ti. 155. 1894. [Mexico] eomaumtaelie Willd. Se. 3:117. 1800. [Mexico - herb. Willd. 12650} ats Bertol. es Guatim. 28. 1840. [Guatemala] suta Miller, Gard. Dict. ed. viit. no. 9. 1768 . ae Sessé & Mocino (non Willd.), FI. Mex. ed. if. 156. 1894. [Cuba] AnaAnn > REFERENCES ADAMS, C. D. 1972. Flowering plants of Jamaica. University of the West Indies, Mona, Jamaica. 848 pp. BAKER, E. G. 1892. Synopsis of genera and species of Malveae, XVII. Sida. J. Bot. 30:138 — 142, 235 — 240, 290 — 296, 324 — 332 BORSSUM Meee J. van. 1966. Malesian Malyaceie revised. Blumea 14:1 — 213. CLEMENT, I. 1957. Studies in Sida. Contr. Gray Herb. 180:1—91. CORRELL, D. . and H. B. CORRELL. 1982. Flora of the Bahama Archipelago, J. 692. FAWCETT, W. and A. B. RENDLE. 1926. Sida, in: Flora of Jamaica, 5:107 — 120. gl sor R. E. 1947. Zur Kenntniss der siid und zentralamerikanischen Malvaceenflora. ungl. Svensk. Vet.-Akad. Handl. 24(2):1—37 + 9 plates meta P. A. 1975. Sidus sidarum. Sida 6:1 -_________.. 1978. Neotropical segregates fers: Sida L. (Malvaceae). Brittonia 30:447 — 462. FRYXELL, P.A., A. KRAPOVICKAS, and D. CREWZ. eee ae Se IV. Anew record of ee in North America, S. santaremensis. Sida 10:319 — GRAY, A. 1849. Plantae Fendlerianae Novi-Mexicanae. Men ee Acad. Arts 4:1—116 [Malvaceae on pp. 15 — 25}. ereaGal A. 1866. Catalogus Plantarum Cubensium. Leipzig, pp. 1— 301. {Sida on pp. 24— 26}. HEPPER, F. N. ae The West African herbaria of Isert & Thonning. Royal Botanic Gardens, Kew HUTCHINSON, jy. ys ” The genera oe plants, family 119. Malvaceae. vol. 2, pp. 536-567. Oxford University Pre ILTIS, H. H. 1963. a dioica ie: Whence came the type? Amer. Midl. Naturalist 70:90 — JUNGHANS, J. mn = ao Thonning and Isert’s collections ee cae Guinea” (Ghana) in West Tropical Africa. Bot. Tidsskr. 57:310 — 355; 58 KEARNEY, T. H. 1951. The American genera of Malvaceae. je ca Sais A693 = 131, —______. 1954. A tentative key to the North American species of Sida L. Leafl. W. Bor. ae 138—- 150. 1958. A tentative key to the South American species of Sida L. Eeans W. Bot. 8:249 — 270 KRAPOVICKAS, A. oe Notas citotaxonémicas sobre Malvaceas. Sytem 24. RAIS, 1983. Notes on Mascarene Malvaceae. Kew Bull. 38:41 — MONTEIRO FILHO, H. daC. 1942. Sida sul-riograndenses. I-paree. Chaves ie especies. . Inform. Agric., Min. Agric., Rio de Janeiro. pp. 1— 10. 9. As especies ai besiicies e uruguayas da seccaoa Ma/vinda do genero Sida. Lilloa 17:501—52 RODRIGO, A. del P. 1944. Las a argentinas y uruguayas del género Sida (Mal- vaceae). Revista Mus. La Plata, Secc. Bot. 6:81—212 + 31 plates. SCHUMANN, K. 1891. Sida, in: Martius, Flora Bras. 12(3):279 — 347. SPOONER, D.M., A.W. CUSICK, G.F. HALL and J.M. BASKIN. 1985. Observations on the distribution and ecology of Sida hermaphrodita (L.) Rusby (Malvaceae). Sida {in press}. STAFLEU, F. A. 1967. Taxonomic literature. (Regnum Veg. vol. 52). Utrecht, pp. xx + 56 ACHYRANTHES JAPONICA (MIQ.) NAKAI (AMARANTHACEAE) IN KENTUCKY AND WEST VIRGINIA: NEW TO NORTH AMERICA MAX E. MEDLEY Department of Biology, U seal of Louisville Louisville, KY 40292, As HAL BRYAN Kentucky Transportation Cabinet Division of Environmental Analysis Frankfort, KY 40601, U.S.A. JOHN MacGREGOR Department of Fish and Wildlife Nongame Wildlife Program Frankfort, KY 40601, U.S.A. JOHN W. THIERET Department of Biological Sciences, Northern Kentucky University Highland Heights, KY 41076, U.S.A. On 26 August 1981 a plant unknown to them was collected by HB and JM on the banks of Tug Fork of the Big Sandy River at Warfield, Martin County, Kentucky (Bryan & MacGregor s.n., DHL). Although identifica- tion to family—Amaranthaceae—and genus—Achyranthes—posed no problem, the specimens were not convincingly identifiable to species in pertinent eastern North American floristic works (Fernald 1950, Gleason 1952, Gleason and Cronquist 1963, Radford et al. 1968, Robertson 1981, Small 1933, Standley 1917, Steyermark 1963, Strausbaugh and Core 1D 7 8i) According to Robertson, in his account of A haceae for Vascular Flora of the Southeastern United States (unpublished), Achyranthes in the southeast is represented by a single species, A. aspera, L., with two subspecies (recognized as two species, under Centrostachys, by Standley {1917}). Our specimens did not fit the descriptions of either of these subspecies. We finally concluded, however, that they did fit descriptions of SIDA die 92 95, 198); 22 an Asiatic species, A. saponica (Miq.) Nakai (Liu and Kao 1976, Ohwi 1965); they matched the illustration of this species in Liu and Kao (1976). Our tentative identification was verified by comparison with authentic specimens of A. japonica from Honshu, Japan, that we borrowed from US. Our specimens (Fig. la, b, c) appear to represent var. hachijoensis Honda and are, we believe, the first from North America. According to Liu and Kao (1976), A. japonica, in addition to its occurrence on Taiwan, is ‘widely distributed in China, the Ryukyus, Japan and Korea and the temperate and subtropical regions of southeastern Asia.”’ We have since collected the species in Lawrence County (3 mi S of Louisa, 7 Nov 1982, Medley, Hotchkiss, & Woodward 7160-82, DHL) and Pike County (along Tug Fork in river birch woods, ca 3 miles NW of South Williamson, 9 Sep 1984, Medley s.n., DHL), Kentucky, and in Mingo County, West Virginia (wooded banks of Tug Fork, ca 3 miles S of Kermit, 13 Aug 1982, Medley 6082-82, DHL). It occurs as open colonies of individual plants up to 1.5 m tall on wooded river banks in areas that have an incomplete or light canopy and that are annually flooded. Dominant species in the community at the Martin County site included Betula nigra, Microstegium vimineum, Pilea pumila, and Boehmeria cylindrica. This shady habitat of A. japonica is in contrast to that of A. aspera, which grows in open waste places and on roadsides (Walker 1976). Dispersal of A. japonica in Kentucky and West Virginia appears to be accomplished largely by water. However, at the Lawrence County site in early November, when the infructescences were fully elongated (quite reminiscent of those of Phryma leptostachya) and the seeds were mature, the fruiting calyces—each with its accompanying subulate-spinose bracteoles—detached from the plants and clung to clothing, indicating an adaptation to dispersal by animals. (Zoochory, for A. aspera, was described by Bullock and Primack [1977].) The origin of A. japonica on Tug Fork is unknown. It was possibly via the major railroad that parallels the watercourse at the site where the species was first found. This species will probably be found eventually in all Kentucky and West Virginia counties bordering Tug Fork and the lower Big Sandy River; it may ultimately be found along the banks of the Ohio River downstream from the mouth of the Big Sandy. In time it may also move up the tributary valleys of Tug Fork and the Big Sandy via transport by small mammals and fall migrant birds. The two species of Achyranthes now known to occur in the conterminous United States can best be separated on the basis of characteristics of their staminodes, as follows: A. aspera—staminodes fimbriate at apex (Fig. 1d); 94 A. japontta—staminodes entire to denticulate or slightly notched at apex (Fig. 1c). As an atd to other workers who find A. japonica we present the following description of the species, which is based on our Kentucky and West Virginia material. Herb (reported as perennial; Liu and Kao 1976, Ohwi 1965, Walker 1976). Stem erect or ascending, sometimes becoming decumbent late in the season, 75-150 cm tall, glabrous to lightly pubescent, 4-angled, vertically 12-lined (2 lines per angle, 1 line per face). Leaves simple, opposite, blades oblong-elliptic, 2.5 — 13.5 cm long, 1.2 — 6.8 cm wide, Figure 1. Achyranthes japonica. a, upper part of plant, X 0.4; b, infructescence, X 0.4; c¢, staminodes and stamens, X 14. A. aspera. d, staminodes and stamens, X 14. 95 pinnately veined (veins opposite to alternate), short pubescent above, pubescent on veins below, apex acute to acuminate, margin entire; petioles 0.4— 3.5 cm long. Inflorescence spicate, terminal on main stem and upper branches, erect, many flowered; spikes 2 — 4 cm long and compact in early flower, elongating to 21 cm and becoming more open, especially prox- imally, in mature fruit. Flowers perfect, regular, hypogynous, apetalous, sessile, divergent at right angles im anthesis, sharply deflexed in mature fruit, then becoming as much as 1.5 cm apart in lower part of spike; each flower subtended by a membranous bract ca 2 mm long and by 2 rigid, subulate-spinose bracteoles 3-4 mm long, each bracteole with 2 basal, suborbicular, membranous auricles, the bracteoles and the flower falling as a unit; sepals 5, linear-lanceolate, 4-5 mm long, acuminate; stamens 5, alternating and connate below with 5 entire, denticulate, or slightly notched staminodes. Fruit a 1-seeded utricle, oblong, 2.5 mm long, 1 mm wide, tipped by the slender style, this 1 mm long. REFERENCES BULLOCK, S. H., and R.B. career 1977. Comparative experimental study of seed dispersal on aerials Ecology 58:68 CHUNG-KUO, K. 1979. a. Flora Reipublicae Popularis Sinicae 25(2):226- 231. {In Chinese; illustration of A. japonica (A. bidentata).| FERNALD, M. L. 1950. Gray’s manual of botany, 8th ed. American Book Company, New York. GLEASON, H. A. 1952. The new Britton and Brown illustrated flora of the CO aes United States and adjacent Canada. New York Botanical Garden, New York. GLEASON, H. A., and A. CRONQUIST. 1963. Manual of the ee plants of northeastern United States and adjacent Canada. Van Nostrand, New York. LIU, T.-S., and M.-T. 1976. Amaranthaceae. In H.-L. Li et al., editors. Flora of ean. a Il. Aagtocpenmde: Epoch Publishing Company, ai Taiw OHW/I, ae 1965. Flora eres Smithsonian ry Washin a os C. RADFORD, A. E., H. HLES, and C. R. Bell. 1968. Manual of ia a. flora of the Gane ee a ae Carolina tee Chapel Hill. sa hae SON, K. R. 1981. The ae of Amaranthaceae in the southeastern United ates. - Arnold ae oS i J. K. 1933. Manual of the en flora. University of North Carolina Press, Chapel Hi IL STANDLEY, P. C. 1917. Amaranthaceae. N. Amer. Flora 21(2):95 — 169. STEYERMARK, J. A. 1963. Flora of Missouri, Iowa State University Press, Ames. STRAUSBAUGH, P. ‘ee , and E. L. CORE. 1978. Flora of West Virginia, 2nd ed. Seneca Books, Grantsville, W. Va. WALKER, E. H. 1976. ee of Okinawa and the southern Ryukyu Islands. Smithsonian fnseieutiony: Washington, D.C. DRYMARIA VISCOSA (CARYOPHYLLACEAE): CORRECT AUTHOR CITATION AND RANGE EXTENSION TO THE UNITED STATES BRUCE D. PARFITT AND WENDY HODGSON Desert Botanical Garden, 1201 N. Galvin Parkway Phoenix, AZ 85008, U.S.A. Although cited as Drymaria viscosa 8. Wats. ex Orc. in both monographs of the genus (Wiggins, 1944; Duke, 1961) and Index Kewensis (Durand & Jackson, 1906), the species was neither described by Orcutt nor ascribed to Watson by him. Ina narrative of his trip to Baja California, Mexico, Orcutt (1886) merely mentioned the name, creating a nomen nudum: At Socorro we found a few lichens, shells, a young palm tree, and various nice plants of which I will mention Dalea Seemannii, Nemacaulis nuttallii and a variety, Aesculus Parryi, Euphorbia micromera, drymaria (sic) viscosa, n. sp., Astragalus hornii, A. menziesii, . . . etc. The specimen collected by Orcutt at “Socono” (Socorro, the hand- written label is easily misread), and cited by Watson (1887:469) with his description of the species, represents the holotype (“C. R. Orcutt, April, 1886" = Orcutt 1330, GH!). With regard to D. viscosa, Watson makes no other reference to Orcutt. While it is likely that Watson knowingly used and legitimized Orcutt’s name for the new plant, he did not give Orcutt credit for the name. Furthermore, because the plant bears an indument of glandular trichomes to which grains of sand adhere, it is equally possible that Watson arrived independently at the epithet viscosa. Articles 46 of the International Code of Botanical Nomenclature (Voss et 1983) states that’. . . it is necessary to cite the name of the author(s) who first validly published the name concerned . . .”’ This is followed by Recommendation 46E. 1. which explains that where an author has validly published a name and ascribed it to another person, “the name of the other person, followed by the word ex, may be inserted before the name of the publishing author, if desired” (italics ours). Thus Watson 1s the author who validly published Drymaria viscosa and must be cited. Had Watson ascribed the name to Orcutt, D. viscosa Orc. ex S. Wats. would have been the correct option, though opposite the order of authors cited in the monographs (Wiggins, 1944; Duke, 1961) and Index Kewens1s (Durand & Jackson, 1906. However, because neither Watson nor SIDA 11(1): 96—98. 1985. 27 Orcutt ascribed D. viscosa to the other author, the use of the connector “ex” between their names in either combination is clearly incorrect. As cited in Shreve and Wiggins (1964) and Wiggins (1980) and as suggested in the Gray Herbarium Card Index (Harvard University, 1968), but contrary to the monographs, Drymaria viscosa S. Wats. is correct and Drymaria viscosa Orc. remains a nomen nudum. DISTRIBUTION Wiggins (1980) considered D. viscosa as endemic to Baja California, occurring in sandy areas from San Quintin to the Cape region. However, Felger (1980) reported the species from the Gran Desierto of northwestern Sonora and there is in CAS a specimen from Sonora collected in 1966 (Ripley 14226). These specimens and others represent a mainland distribution from the region of the Pinacate Lava Flow in Sonora southward along the coast of Sonora nearly to Guaymas. MEXICO. Sonora: Dunes ca 5 mi NE of Sierra del Rosario, 32°08’N 114°09'W, Felger Pa et er (ARIZ); Sand desert N of Sierra Pinacate, 0.5 mi S of the San Luis—Sonoita road, Mason 1832 et al. (ARIZ); Moon Crater (Cratero Chichi), SW part of the Pinacate ee Felger 19239 (ARIZ); ca 1 mi S of Moon Crater, 31.7°N 113.6°W, Felger 19095 & Hanson (ARIZ); ca 6 mi S of Moon Crater, 31.6°N 113.6°W, Felger 19034 & deRosa (ARIZ); pune 0.2 km S of Hwy 8, ca 29 km SW of Sonoita, Burgess 4761 (ARIZ); 7.2 mi from cabin 245 at Choya, Burch s.n. (ASU); La Mancha Blanca, = side of Cerro Tepopa 29°24'N 112°24°W, Felger 20850 et al. (ARIZ); Sand dunes, Puerto Kino, 15 Feb me pes 14226 (CAS); High beach dunes ca 1 mi NW of village . Tastiata, 28°20%'N 111 Felger 20881 et al. (ARIZ, GH); Coastal sand dunes S of Moro (Morro) Colorado, 5 mi oe of Estero Tastiota, Copp 70-3 (CAS) Y Thus although not previously reported for the United States, the occur- rence of Drymaria viscosa from the United States portion of the Pinacate region is not as unusual as Wiggins’ (1980) report of endemism in Baja California would suggest. One specimen is now known from the U.S.A.: Arizona, Yuma Co., eastern edge of the Pinacate Lava Flow, along the Camino del Diablo, Cabeza Prieta Game Range, occasional on sand dunes with Triteliopsis palmeri, ca 900 ft elev., 17 Apr 1983, Hodgson 2080 & Engard (DES). This species may be distinguished from other Arizona species of Drymar- ia by its occurrence below 1000 ft elev.; the others occur above 4000 ft (Kearney & Peebles, 1960). Because the iene are several per node it most strongly resembles, and keys with, D. molluginea (Lagasca) Didr. but ts heavily glandular pubescent with smooth, tan seeds. REFERENCES DUKE, J. A. 1961. Preliminary revision of the genus Drymaria. Ann. Missouri Bot. Gard. 48:173 — 98 DURAND, T. & B. D. JACKSON. 1906. Index kewensis. Supplement 1. Clarendon Press, Oxford. FELGER, R. S. 1980. Vegetation and flora of the Gran Desierto. Desert Plants 2:87 — 114. HARVARD UNIVERSITY. 1968. Gray herbarium index. vol. 4. G. K. Hall and Co., Boston oo KEARNEY, T. H. & R. H. PEEBLES. 1960. Arizona flora. 2nd edition with supplement by J. T. Pat & E. McClintock. University of California Press, Berkeley. ORCUTT, C. R. 1886. A botanical trip. West Amer. Sci. 2:53 —58. SHREVE, F. & 1. L. WIGGINS. 1964. anny and flora of the Sonoran Desert, vol. Stanford University Press, Stanford, Californ VOSS, E. G., ET AL., eds. 1983. International wer of botanical nomenclature. Regnum Veg. 1 l:i-xv + a WATSON, S. 1887. Dscripudns of some new species of plants. Proc. Amer. Acad. Arts. 22:466— 481. WIGGINS, I. L. 1944. The genus Drymaria in and adjacent to the Sonoran Desert. Proc. Calif. Acad. Sci. 25:189—214 1980. Flora of Baja California. Stanford University Press, Stanford, California. NOTES THEMEDA QUADRIVALVIS (L.) KUNTZE (POACEAE) IN LOUISIANA.—Themeda quadrivalvis (L.) Kuntze was first reported from the United States from St. Landry Parish, Louisiana, based on several collections from populations established on agricultural lands (Brown 1945). Since the time of Brown’s report Themeda has apparently not been recollected in Louisiana, and publications citing the name with reference to the Louisiana flora contain no new information on the status of Themeda in Louisiana (Allen 1980; Thieret 1972; Thomas & Allen 1984). We report here that populations of T. guvadrivalvis still thrive in Louisiana in St. Landry Parish based on our field observations of it in September 1984 (Figs. 1 — 4). Anthesis had begun on 6 September, when we first located stands of Themeda in the field, and grains had begun to develop a week later when we discovered additional stands. Dr. Brown’s original description of Themeda’s habitat in Louisiana is still accurate for the stands we recently observed: ‘“‘Dense stands were found on the headlands of cultivated fields, along fence rows, and along the ridges of cultivated fields outside the influence of the last cultivation. In places the stand of this grass was so thick that the usual weeds of these sites were excluded.” From our observations Themeda appears to be in no danger of dying out in Louisiana in spite of intensive cultivation of soybeans in the areas in which the grass grows. In fact, because T. guadriva/vis is an annual, it probably could not persist without maintenance of favorable sites for it through agriculture. The most robust plants we saw were ina large fallow field dominated by Ambrosia trifida and Setaria sp. where there were several small colonies of Themeda, widely separated from one another, and includ- ing some very tall culms to 2.62 m. The plants in this fallow field were much larger on average, and more mature, than those growing around fields actually under current cultivation. Presumably the plants in the fallow field grew so well because they were able to get an early start and were not knocked down or otherwise disturbed by agricultural practices. In well- tended soybean fields, in which the headlands and field edges were kept mowed, Themeda was lacking. It is apparently not able to grow right out in the field among the soybean plants as some other grasses—Sorghum halepense for example—do so successfully. In the United States T. guadrivalvis 1s also known from Manatee County, Florida, where it is said to be an escape from cultivation (Wunderlin 1982). Specimens of our recent colletions of T. SIDA 11(1): 99, 1985. : Ss x S = Ry x Re CR . . x wen . Figure |. Typical habitat of Themeda quadrivalvis in St. Landry Parish, Louisiana; fencerow indicated by trees at right, soybean field at left, Themeda growing between the fencerow and the soybeans. Figure 2. Site similar to that shown in Fig. 1 but lacking Themeda because field edges are § kept mowed. SIDA 11(1): 100, 1985. Figure 3. Stand of Themeda quadrivalvis in a fallow field with Setaria. sp. and Ambrosia trifida. Tallest culms of Themeda here were over 2.6 m. Figure 4. Inflorescenses of Themeda quadrivalvis. SIDA 11(1): 101. 1985. 102 quadrivalvis, in addition to those at LAF, are being distributed to GA, GH, IBE, KNK, LSU, LTU, MICH, MO, NATC, NCU, NLU, NO, NY, SMU, TAES, US, and VD We thank John W. Triceet for suggesting that an attempt be made to relocate T. quadrivalvis in Louisiana, and Debra Waters for participating in the field work.—Wi/liam D. Reese and Garrie P. Landry, Biology Department, University of Southwestern Louisiana, Lafayette, LA 70504, U.S.A REFERENCES ALLEN, C. M. 1980. Grasses of Louisiana. University of Southwestern Louisiana, Lafay- CCre, BROWN, C. A 1945. Notes on additions to the flora of Louisiana. Proc. Louisiana Acad. Sci. 9:4 — THIERET, S Ww. 1972. Savane of a vascular flora of Louisiana. Part I. Techn. Bull. Lafayette Natural History Mus THOMAS, R. and C. M. ALLEN. er A preliminary checklist of the pteridosperms, gymnosperms, and monocotyledons of Louisiana. Contr. Herb. Northeast Louisiana Univ. 4:1—55. WUNDERLIN, R. P. 1982. Guide to the vascular plants of Central Florida. Univ. Presses of Florida, Tampa, St. Petersburg, Fort Meyers, Sarasota. CAMPANULA RAPUNCULOIDES (CAMPANULACEAE) NEW TO TEXAS—On June 9, 1984 I collected a single plant of Campanula rapuncu- lordes L. (Brown 7507, SMU) in a woodland of Quercus glaucoides on a rocky limestone slope above the Sabinal River in Lost Maples State Park north of Vanderpool in Bandera County. This native of Europe is reported by Fernald (FERNALD, Gray’s manual. 1950) to be naturalized in eastern orth America south to Indiana, Illinois, and Missouri.—Larry E. Brown, Houston Community College, 726 Horncastle St. Channelview, TX 77530, NOTES ON TWO TEXAS PLANTS—Juncus caprrarus Weigel Juncaceae) was first reported for Texas by Gould in 1962 [1963] from Walker County. This collection by S. R. Warner was annotated as a new species (CoryPE: SMU) by Tharp & Barkley. A literature search indicates the name was never published. Since then the species has been determined as J. capitatus—an introduction from the Old World—now scattered over the southern United States. Two new locations are here reported for the state. Nees data: Walker Co.: Rock Springs Church, 3 May 1944, S. R. ee iM, (SMU). Bandera Co.: first low water crossing on FM 187 N of Utopia ca 5 mi ae. 2 May 1984, T. M. Keeney 3876 (SMU, UVST); same locality, 6 May oat SIDA I1(1): 102. 1985. 103 3878 (UVST); same locality, 8 May 1984, 3879 (UVST). Uvalde Co.: sandy soil, Frio River bed NW of Knippa, 11 May 1985, Keeney 4493 (SMU, UVST). VERONICA AMERICANA (Raf.) Schwein. (Scrophulariaceae) was first re- ported for Texas based on a single collection from Kendall County in the Edwards Plateau (Correll & Johnston, 1970; Correll & Correll, 1972). A recent collection of this aquatic is apparently the first for the state since 1961. Collection data for the two known locations are: Kendall Co.: gravelly stream along ae Creek, ca 4 mi NW of Boerne, 11 Jul 1961, K. 1. G&L. W. Miller 975 (SMU). Kerr : in Verde Creek ca 300 yds E of bridge over Verde Creek at Camp Verde on Hwy 173, 6 a 1984, T. M. Keeney 4099 (SMU, UVST). —Toney M. Keeney, Herbarium, Southwest Texas Junior College, Uvalde, TX 78801, U.S.A. and Barney L. Lipscomb, Herbarium, Southern Methodist University, Dallas, TX 75275, U.S.A REFERENCES CORRELL, D. S. & H. B. CORRELL. 1975. Aquatic and wetland plants of southwestern United States. Stanford Univ. Press, Stanford, Califor CORRELL, D. S. & M. C. JOHNSTON. 1970. Manual a the vascular plants of Texas. Texas Research Foundation, Renner. GOULD, F. W. 1962 {1963}. Texas plants—A checklist and ecological summary. Texas Agri. Exp. Sta. MP-585. SCAEVOLA SERICEA VAHL VAR. TACCADA (GAERTN.) THIERET & LIPSCOMB, COMB. NOV. (GOODENIACEAE)—Based on Lobelia taccada Gaertn., Fruct. Sem. Pl. 1:119, t. 25, fig. 5. 1788. Scaevola sericea, a widely distributed Indo-Pacific strand plant, is grown as an ornamental and has escaped in southern Florida. It has been called S. taccada (Gaertn.) Roxb. in some recent floras, but C. Jeffrey has shown (Kew Bull. 34:537 — 545. 1979) that S. sericea Vahl is the correct name. Two variants of the species occur in Florida—as “‘on many other tropical shores” (H. St. John, Proc. Biol. Soc. Washington $3275 = 15.1975), one with sericeous leaves, S. sericea var. sericea, and one with glabrous leaves, for which the above new combination is made.—John W. Thieret, Dept. of Biological Sciences, Northern Kentucky University, Highland Heights, KY 41076, U.S.A.; Bar Lipscomb, Southern Methodist University Herbarium, Dallas, TX 75275, U.S.A. SIDA 11(1): 103. 1985. REVIEWS INDICES TO THE MICROFICHE OF THE TYPES AND SPECIAL COL- LECTIONS (FLOWERING PLANTS AND FERNS) OF THE HERBARIUM OF THE ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA. James A. Mears. 274 pp. 1984. Meckler Publishing, 11 Ferry Lane West, Westport, Connecticut 06880. $350.00 The thorough title tells it all. Two copies of this clothbound set of indices are included in the $3700 purchase price for the complete microfiche collection. The volume is available separately, however, for anyone (or, more likely, any institu- tion) desiring it. This work alone will give some clue as to whether a type may be in the Philadelphia Academy, and thus can be useful even without the microfiche (or herbarium itself)—and it contains some useful historical information besides. But the phrase ‘‘special collections” is an important part of the title. This is not an index to verified types. Many of the PH collections date from before the type method. The major portion of this volume (185 pp.) covers the “Collection of Types and Early Authenticating Specimens,” and the latter phrase includes a multitude of specimens possessed or annotated by authors of species names or simply by important early botanists. Even the verified types are not designated as such in this index. Collectors’ names but, strangely, not their numbers are given and so are geographic sources, presumably whenever known (which is sometimes not often); but the index does not reveal, and makes it difficult to determine without examination of the specimen or photograph, whether any specimen is in fact a type. Perhaps this is good—it encourages checking the original data—but it does reduce the usefulness of the index. Shorter indices cover the G. H. E. Muhlenberg Herbarium, the B. S. Barton Herbarium Fragment, the A. B. Lambert Herbarium Fragment, the Lewis & Clark Collection, and specimens of various other notables. The verified types from some of these, however, are listed in the initial index to types and early authenticating specimens and not with the collection in which they are in fact filed (e.g., Barton, Muhlenberg). Some specimens which borrowers did not return for this project did not get photographed at all (but may be indexed), and “many type specimens (particularly unlabeled isotypes)” are “still in the general and local herbaria of the Academy.” An introduction to each of the indices explains the origin of the particular collection and the sequence of names (Muhlenberg following his Cata- logue, Barton with families after the Bessey system, Lambert alphabetical by species as named in Pursh, the initial index after Dalla Torre & Harms, etc.). All this apparent chaos 1s partly alleviated by a concluding index to all pages on which specimens of each family are listed. The broad scope of this work greatly enhances its usefulness—not alone the generous definition of ‘‘special collections” besides types, but also the inclusion of specimens from collections on indefinite loan to the Academy, such as those of the SIDA 11(1): 104. 1985. 105 American Philosophical Society and the University of Pennsylvania. Many botan- ists may be pleasantly surprised to learn that the PH collections—well known as an extraordinarily rich source of early North American specimens—include so much Old World material (with some even from Linnaeus himself). I am unable to say how many specimens are indexed. An advertisement states that there are 454 microfiche each with a maximum of 60 plants. Slightly less than 300 of the fiche cover the herbarium of “types and early authenticating specimens’ —which are said to number about 40,000 vascular plants. Perhaps less than half of these are photographed—or does the apparent contradiction in num- bers merely mean there is an average of two collections per photo?—Edward G. Voss, Herbarium, University of Michigan, Ann Arbor, MI, 48109, U.S.A METHODS IN PLANT VIROLOGY. 1984. Hill, Stephen A. Methods in Plant Pathology Vol. 1. Blackwell Scientific Publications, Ltd., Oxford, London, Edinburg, Boston, Palo Alto, Melbourne. 167 pp. + viii. Paper, $24.00. The volume concentrates on the methods for virus diagnosis. The beginning student of plant virology should learn the basic techniques and gradually evolve into the more sophisticated techniques. The techniques are sequential and may provide a route to virus characterization. The Chapters are D) Introduction, 2) Histological and other Basic Methods, 3) Basic Virus Charac and Storage, 4) Transmission Tests (Sap, graft, and vector transmission), 5) Serological Tech- niques (Precipitation and agglutination tests, labelled antibody techniques), 6) Electron Microscopy (Quick methods for sample preparation and immuno electron microscopy). By grouping the basic methods of characterization into one volume, the author has produced a valuable handbook that not only describes the peonee ne of seal method but also lists the materials required and procedes in a step by step cookb fashion. The text was written for senior undergraduates and researchers in ce pathology and plant virology and is recommended not only for its total contents but for the brief, concise individual nature of each recipe.—Wm. F. Mahler, Southern Methodist University Herbarium, Dallas, TX 75275, U.S.A INTRODUCTION TO MODERN MYCOLOGY. 1984. Deacon, J. W. Basic Microbiology Series Vol. 7. Blackwell Scientific Publications, Ltd., Oxford, London, Edinburg, Boston, Palo Alto, Melbourne. 167 pp. + viii. Paper, $24.00. The text is an introduction to the biology of the fungi and deals with their structure, function, and some aspects of their life history. The Chapters include topics on structure and fine structure, growth, differentiation, nutrition, metabol- ism, environmental conditions for growth, genetic systems, spore dispersal, the role of fungi as saprophytes and parasites (plant and animal), and on prevention and control of fungal growth. SIDA 11(1): 105. 1985. 106 From an educational viewpoint, this text is recommended after a solid founda- tion in the “modern” taxonomic and life cycle concepts has been achieved in order to permit the student to attain the proper perspective of the whole while studying various detailed aspects. This text on the biology of the fungi is an excellent one for the second course in a two course sequence of an “Introduction to Mycology." — Wm. F. Mahler, Southern Methodist University Herbarium, Dallas, TX 75275, Can PLANT PATHOLOGY & PLANT PATHOGENS. 1982. Dickinson, C. H. and J. A. Lucas. Basic Microbiology Series Vol. 6. Blackwell Scientific Publica- tions, Ltd., Oxford, London, Edinburg, Boston, Palo Alto, Melbourne. 2nd ed. 229 pp. + viii. Paper, $19.75. In the Preface, the authors state that the “aim in this book is to provide a balanced treatment of all aspects of disease caused by microbial agents” with an emphasis on the host-pathogen complex and the development of general principles. The Chapters cover the concept of disease, the microbial pathogens, pathogen structure and function, infection and colonization, host-pathogen interaction at the population, whole plant, cellular, and molecular levels, host-pathogen specific- ity and disease control. An annotated list of pathogens and the diseases they cause comprise the Appendix. The text achieves the aims of the authors and covers the basic concepts concerned with plant diseases caused by microbial agents and the appendix is especially useful for quick reference.—Wm. F. Mahler, Southern Methodist University Herbarium, Dallas, TX 75275, U.S.A. SIDA 11(1): 106. 1985. S | DA CONTRIBUTIONS TO BOTANY VOLUME 11 NUMBER 2 DECEMBER 1985 CONTENTS Systematics of Leucophyllum and Eremogeton (Scrophulariaceae). James Henrickson and L. David Flyr. 107 A revision of the clasping-leaved Potamogeton (Potamogetonaceae). Robert R. Haynes. lee) Taxonomy, distribution and rarity status of Leavenworthia and Lesquerella (Brassicaceae) in Kentucky. Ray Cranfill, Jerry M. Baskin and Max E. Medley 189 Dioecy in North American Cactaceae: A review. Brace D. Parfitt. 200 Amphibromus scabrivalvis (Gramineae) in Louisiana. M. Lynn Calaway and John W. Thieret. 207 Observations on the distribution and ecology of Sida hermaphrodita (L.) Rusby (Malvaceae). David M. Spooner, Allison W. Cusick, George E. Hall and Jerry M. Baskin. 219 — Additions and noteworthy vascular plants from Arkansas, with some ecological notes. Steve L. Orzell, Edwin L. Bridges and S. Lance Peacock. 226 The vascular flora of Central Florida: taxonomic and nomenclatural changes, additional taxa. R. P. Wunderlin, B. E. Hansen and D. W. Hall. 252 NOTES. Dactyloctenium aegyptinm (Gramineae) new to Belize. 245—Additional notes on Coreopsis. 2 var. neomexicana (Gray) Cronq. (Asteraceae) new to Texas. 246—Nestronia umbellila Rat. (Santalaceae) new to Mississippi. 246—Aletris farinosa, Cyperus difformis, and Cyperus pilosus new for Florida. 247—A White-flowered form of Spigelia marilandica L. (Loganiaceae) new to Tennessee. 248—New combinations in Erigeron (Asteraceae). 249. Documented plant chromosome numbers 1985: 4h, 251 US ISSN 0036-1488 SIDA, CONTRIBUTIONS TO BOTANY Founded by Lloyd H. Shinners, 1962 Publisher Wm. F. Mahler SMU Herbarium Dallas, Texas, 75275 Editor Associate Editor Barney L. Lipscomb John W. Thieret SMU Herbarium Northern Kentucky University Dallas, Texas, 75275 Highland Heights, Kentucky, 41076 Guidelines for contributors are available upon request. Subscription: $10.00 (U.S.) per year; numbers issued twice a year. Sida, Contributions to Botany, Volume 11, Number 2, pages 107 — 253. Copyright 1985 by Wm. F. Mahler SYSTEMATICS OF LEUCOPHYLLUM AND EREMOGETON (SCROPHULARIACEAE) JAMES HENRICKSON Department of Biology, oo nia State University, Los Angeles, CA 90032, U.S.A. L. DAVID FLYR Department of Botany, University of Texas, AU, LM POT Ley Ud ptt (Deceased) ABSTRACT ea de and Eremogeton were placed in Leucophylleae series Pseudosolaneae by Ben- tham and in Verbasceae by Wettstein at the base of Scrophulariaceae. The two genera ha been ponte somewhat anomalous in Scrophulariaceae because of their or habits od other features. Recent comparisons have shown aces with woody Myoporaceae but the woody habit of Lemp appears to be derived: this and comparisons of ovary structure show relationships are clearly with ee Data from comparative anatomy and a eee gy of leaves, stems, flowers, and trichomes, from chromosome number, ane from eee hy suggest recognition of 12 species of eens includ- ing F nonorypic subgenus. The species range from Puebla rough arid aaa of Mexico into on Texas. Lencophyllum langmaniae and L. nia) are scribed as new. Eremogeton, retained as a monotypic genus related to Lexcophyllum, is ieee to Central America. INTRODUCTION Since its inception, Lewcophyllum has been placed in Scrophulariaceae, however, its position within the family has varied. Bentham (1846), in de Candolle’s Prodromus, placed the genus in his suborder (subfamily) Anti- rrhinideae, tribe Gratioleae, subtribe Aptosimeae, next to the Old World genera Aptostmum Burchell, Pelzostomum E. Mey., Anticharis Endl., and Doratanthera Benth. ex Endl. (= Anticharis). Later, Bentham (1876) in his Genera Plantarum treatment, placed Leucophyllum along with Ghiesbrechtia Gray (= Eremogeton Standl. & Wil- liams) and Heteranthia Nees & Mart. (a monotypic Brazilian genus now placed in Solanaceae) in his series Pseudosolaneae (with alternate leaves, simple centripetal inflorescences, corolla with posterior lobes external in bud), tribe Leucophylleae (corolla tubes campanulate), at the base of the family next to the South African Aptosimeae (with Aptosmum, Plisotomum, and Anticharis) and the Old World Verbasceae (with Verbascum L., Celsia L., and Staurophragma Fisch. & E. Mey.). SIDA 11(2): 107-172. 1985. 108 Wettstein (1891), in Engler and Prantl’s Natérlichen Pflanzenfamilien, placed Leaucophyllum and Ghiesbrechtia in the series Pseudosolaneae, tribe Verbasceae (again next to Aptosimeae) and emphasized the multicellular, branched trichomes, mostly axillary flowers with actinomorphic or zygomorphic corollas with short, campanulate tubes and 5 stamens with anther thecae united at the tip or throughout. Within the tribe the New World genera were separated from the Old World Verbascum, Celsia, and Staurophragma on the basis of campanulate corolla tubes, anther sacs united at tips only, and solitary, axillary flowers. Flyr (1970) suggested that Lewcophy/lum may best be placed in Myopor- aceae. In an analysis of pollen, Niezgoda and Tomb (1975) showed that, unlike most Scrophulariaceae, Lewcophyllum, Faxonanthus (here treated as a subgenus of Lecophyllum), and Eremogeton have distinctive, prolate or oblate—spheroidal, 3-colpate, diorate pollen, characterized by two ellip- tical apertures on each colpus, one at each side of the equator, a condition also found in the three genera of the Myoporaceae: Bontia L., Eremophila R.Br., Myoporum Soland. ex Forst.f. As noted by Niezgoda and Tomb (1975), however, the same pollen type is present in Capraria L. (Scrophular- taceae). Argue (1980) noted similar pollen in sections Mimulus and Erythranthe of Mzmulus L., in several genera of the Gratioleae (Lancea Hook.f. & Thoms., Artanema D. Don. and Conobea Aubl.), in Penstemon Mitch. of the tribe Cheloneae, and in Celesza of Verbasceae. On the basis of pollen data, Niezgoda and Tomb (1975) suggested the tribe Leucophylleae be transferred to the Myoporaceae as a subfamily. However the paper by Argue (1980) weakens their argument. Tomb (pers. comm. 1984) has recently found similar 3-colpate, diorate pollen in Bignoniaceae. Karrfalt and Tomb (1983) suggested ontogenetic similarities between the epithelium-lined secretory cavities of Bontia (Myoporaceae) and the larger, non-lined air cavities in leaves of selected leuacophyllums noting that scattered cells surrounding the air cavities in Lewcophyllum have epithelial- like expansions. They also cited the occurrence of branched multicellular trichomes on vegetative buds on Bontia, the co-occurrence of isobilateral leaf structure in Lencophyllum and some species of Eremophila and Myoporum, and noted the similarity in habit between the shrubby, gray-leaved Leucophyllum and some Eremophila. In spite of these and other similarities, there exist strong differences between Leacophyllum and Myoporaceae, particularly in gynoecial and fruit characters. As noted by Karrfalt and Tomb (1983), the 2-carpelled, 2- loculed, superior ovaries with axile placentas, many ovules, and septicidal capsules that also open loculicidally at the tip that characterize Lewcophyllum and Eremogeton are typical of Scrophulariaceae but unlike the conditions 109 found in Myoporaceae. To this we may add that the campylotropous ovules and seed structure of Lewcophyllum and Eremogeton are also identical to that of Scrophulartaceae. nile Myoporaceae also tend to have 2-carpelled ovaries with placenta oriented in a manner similar to that of Scrophulariaceae (i.e. with enlarged placenta intruding into the locules perpendicularly from the ovary septum) the placentae in Myoporaceae often extend all the way to the outer ovary wall forming false and incomplete septa that effectively divide each carpel into 2 locules. In some species placentae are not so well developed and carpels are untlocular. Ovaries of Myoporum, in contrast, typically are divided into 2—4 (-up to 12 in Myoporum sandwicense A. Gray) (Webster 195 1) untovulate compartments and this has been interpreted as consisting of 2 carpels divided by supernumerary partitions (Cronquist 1981) or it could possibly be a multicarpellate ovary with up to 6 carpels. Ovules are few in number, typically 1—3 (very rarely 4) pair per carpel, and when consisting of 2 or more pair per carpel, the ovules are superimposed in the narrow locules bordering the intruded placentae. Unlike the Scrophular- 1aceae or Leucophylleae ovules are pendulous, anatropous with micropyles superior (Wertstein 1895; Chinnock pers. comm. 1984). Differences also occur in stigma structure. In the Leucophylleae style tips are expanded, flattened, rounded to acute and stigmatic along the some- what thickened distal margins. In Myoporaceae style tips are either capitate (Myoporum) or more often slender with stigmatic surfaces restricted to a notch at the slender tips, with this rarely expanded into a slightly bifid tip (Chinnock idem.). Unlike the Scrophulariaceae and Leucophylleae mature fruit of Myopor- aceae are indehiscent and can be dry or drupaceous. The endocarp of the fruit walls can be thickened and sclerified or thin and rather cartilaginous (Chinnock idem.). The exocarp can be dry and papery and separable from the endocarp by breakdown of the mesocarp. The mesocarp is often dry, pithy or firmly pithy. These dry fruit are indehiscent but there may be some disintergration of tissue near the top of the ovary and carpels may separate slightly (Chinnock idem.). In Myoporum and some species of Evemophila the mesocarp is thickened and fleshy and the fruit are considered drupaceous though there may be several seeds from more than one carpel inside the hardened endocarp (Chinnock idem.). As ovule number in Myoporaceae typically ranges from | — 3 and these are much larger than in the Leucophylleae measuring 2 — 3.5 mm in length with testa surfaces smooth or faintly reticulate (Chinnock idem. ). Unlike the Scrophulariaceae and Leucophylleae seeds have scanty or no endosperm. pair per carpel, fruit have only | — 8(-12) seeds 110 While Myoporaceae appear to be uniformly woody, data presented here indicate that Lewcophyllum may be secondarily woody, i.e., derived from herbaceous ancestors. Woody growth habits, of course, also occur in other tribes of the Scrophulariaceae. Species of Leacophyllum are very similar vegetatively to some species of Eremophila. As they both occur in semiarid and arid habitats, this to some degree may be convergence. They both have dorsi-ventral to tsobilateral leaves, and can be pubescent to densely canescent often with dendritically- branched hairs. In both, corollas can be weakly zygomorphic, though Eremophila, with 140 species (Chinnock idem. ), exhibits considerably more diversity with corollas ranging from neary actinomorphic to strongly zygomorphic, with some species having 4 posterior corolla lobes and solitary anterior lobes. The corolla and androecium of Leacophyllum strongly resemble those of some species of Myoporum (e.g. M. laetum) even down the surface texture of the corolla trichomes and in the arrangement of the anther sacs. However, many of these characteristics occur throughout both famil- ies. There, however, remain significant differences between the families in gynoecial features, and chromosome numbers etc. While Lewcophyllum and Eremogeton may constitute a somewhat discordant element within Scrophulariaceae, they are strongly discordant in Myopor- aceae and their inclusion in that family would completely go against the few (mostly gynoecial) characters that distinguish Myoporaceae from Scrophu- lariaceae. On the other hand, there are so many characteristics 1n Common between Myoporaceae and Scrophulariaceae that recognition of Myopor- aceae as a distinct family could be questione In the paragraphs below we present data on morphology, anatomy, and cytology of Lewcophyllum and Eremogeton and wherever possible we comment on characteristics found in other Scrophulartaceae and/or Myoporaceae. The picture that emerges is that in many features Lewcophyllum and Eremogeton are similar to both families and in other features (some critical) they are more similar to Scrophulariaceae. These data plus phytographic een ee discussed below, support retention of Leacophyllum and Eremogeton in the tribe Leucophylleae, series Pseudosolaneae, at the base of the Scrophular- iaceae along with other Old World tribes as initially suggested by Bentham (1376). MORPHOLOGY AND ANATOMY HABIT: Leacophyllum species are mostly small- to medium-sized, rounded shrubs 0.5 — 2.5 m tall with divaricately branching stems. Branching may be ascending or divergent. In some species, dead branches persist and give the plants a thorny aspect. The habit of L. pringle: differs from others in its 111 basal woody burl from which arise several, slender, erect stems that branch only in the distal portion. Eremogeton, which neither of us has seen, is apparently a taller, branched shrub to small tree 1.5 — 8 m high (fide labels). YOUNG $ : Young stems of Leacophyllum are terete and vestitured. The nodes develop distinct, persistent protuberances, and long-shoot leaves abscise just above the swollen bases. Anatomically, young stem pith initially consists of large parenchyma cells that develop into lignified brachysclereids within the first year. Vascular tissue develops in a con- tinuous ring, 1.e., not broken by medullary rays (see xylem description below). A continuous ring of primary phloem fibers to 0.06 mm thick in Leucophyllum, to 0.2 mm thick in Eremogeton, occur outside the non-lignified his cylinder of fibers is broken apart as stems increase in dia- meter. In some species the areas between the primary phloem fibers develop brachysclereids. The cortex is parenchymatous, the outer portion consisting of chlorenchyma and later storing starch. A phellogen develops in the outermost cortex layer immediately beneath the epidermis. Periderm (phel- lem) cells do not collapse radially, and the cork is often soft and thickened. In several species (e.g., L. revolutum, L. pruinosum, Eremogeton, and probably others), the tangential pockets of radially thickened, elongate secondary phloem fibers eventually are included in the periderm. In some species certain phellem cells develop into brachysclereids. XYLEM: Data on xylem anatomy of Leacophyllum kindly have been pro- vided in part by David Michener, whose interest in Lewcophyllum was sparked by previous studies of woods of shrubby Scrophulariaceae, namely Keckiella (Michener 1981). A full report on comparative xylem anatomy of Leacophyllum will be presented elsewhere by Michener. His data provide evidence that the woody habit of Leacophyllum may be secondarily derived from herbaceous ancestors. The data on xylem anatomy were obtained from standard transverse, tangential, and radial sections and macerations from samples of each of the eight Leucophyllum species native to Coahuila, Chihuahua, Nuevo Leén, Tamaulipas, and Texas. The following summary constitutes a generic sce of the xylem of Lewcophyllum. Woods of Eremogeton have not been stu ae of Leacophyllum has both distinct and indistinct growth rings even in one stem reflecting periodic growth flushes probably tied to rainfall. Vessel elements are aggregated and widest in the initial portion of a growth ring; they continue across the growth increment as narrow elements, mostly loosely aggregated in meandering to radial chains associated with para- tracheal axial parenchyma. The growth ring terminates in a narrow band of very narrow vessel elements (+ 12— 15 ttm in diameter) that can be Pie distinguished from fibers in cross section by their pitting. Occasional elements in this zone do not develop perforations and thus constitute vascular tracheids. Vessel elements have simple perforations. The wider elements frequently have narrow tails at one or both ends and perforations are oblique to transverse. Narrower vessel elements tend to have more oblique to nearly lateral perforations. Intervascular pitting 1s of alternate, circular-bordered pits. Prominent tertiary helical thickenings are frequent both in wide and narrow vessel elements. Vessel elements are short (mean lengths range from 158 + 41 wm to 238 + 51 wm) and narrow to very narrow (mean diameters range from 40 + 17 fxm to 24 + 9 pm) in species studied. Axial parenchyma ts paratracheal, confluent, associated with the radial chains and initial aggregations of vessel elements at the beginning of the growth ring. Successive growth rings and minor growth flushes are sepa- rated by lI-celled bands of axial parenchyma. Axial parenchyma is com- monly once-divided tranversely and constitutes parenchyma strands. The ground matrix of the xylem consists of masses of narrow, moderately thick-walled fiber-tracheids with distinct outer pit apertures. In the one specimen of L. frutescens in which they were measured they had an average length of 427 wm with a maximum length of 550 wm, minimal length of 220 wm with average length 2.25 times that of the average vessel element length for the same collection. Both untseriate and multiseriate rays are present. Uniseriate rays are usually short, 2 — 3 cells tall (to 10 cells tallin L. zygophyl/lum). Multiseriate rays are 2—3 cells wide and usually less than 12 (rarely 18) cells tall. Ray cells are primarily procumbent; erect to square cells occur at the zone between successive growth increments. Analysis of ontogenetic development of the vascular cambium of a collection of L. frutescens (Michener 4308) showed a drop in vessel-element length from metaxylem into the first four increments of secondary xylem. The mean vessel element length in late metaxylem was 396 wm, for the metaxylem-secondary xylem transition area, 307 fm, and for the first through fourth years of secondary growth, 187, 134, 163, and 137 wm respectively. The drop in vessel-element length reflects subdivision of cambial initials and the failure of the initials to elongate during secondary growth. This pattern is considered by Carlquist (1962, 1975) to be paedomorphic, and this pattern is considered characteristic of plants that are secondarily woody and derived from herbaceous ancestors. If this 1s the case in Leacophyllum the shrubby habit of Lewcophyl/am need not be consid- ered such an anomaly in Scrophulariaceae nor indicative of relationship to the woody Myoporaceae. Lis Figure 1. Leacophyllum frutescens. a. Habit. b. Face view of flower showing orientation of stamens, style, and spot pattern on floor of corolla throat. Note trichomes on corolla lobes, throat, and orientation of lobes in upper buds (Henrickson 19094). LEAVES: Leaves are alternate, occasionally subopposite (when crowded), in a 2/5 — 5/13 phyllotaxy in all species except L. zygophyllum, in which they are opposite, and L. candidum, in which they are mostly alternate but tend to be subopposite to opposite on uppermost stems. Leaves are borne only on long shoots, axillary short-shoot spurs do not form. However, axillary shoots with crowded leaves may develop in L. minus and L. laevigatum var. griseum, Leaves are simple, oblanceolate, obovate to orbicular, acute to rounded, emarginate at tip, narrowly to broadly cuneate or rounded at base, entire, sometimes undulate, revolute in L. revolutum, (fig. 2 a), toothed in Eremogeton, (fig. 21), flat, or conduplicately folded along the midrib in L. zygophyllum and L. frutescens, soft, pliable, mostly equally, often densely, tomentose on both sides with dendritic trichomes or less strongly vestitured on the upper surface (L. revolutum, L. frutescens) to nearly glabrous (L laevigatum var. laevigatum, L. langmaniae). Stipules are absent. Leaves of most species are isobilateral, as noted by Karrfalc and Tomb 1983), with palisade layers 2—5 cell layers thick on both surfaces (fig. 2 c-d), though often less well developed on the lower surface. Leaves of the bicolored L. revolutum and L. frutescens (ig. 2.a—c), as well as Eremogeton, Figure 2. Leaf ae of ene aes a. Cross section of leaf of L. revolutum showing revolute margins and lon sstiture on adaxial surface. b. Same leaf showing short oe on upper surface a dorsal-ventral mesophyll structure. (Henrickson and Hess 19074). c. L. frutescens. Leaf cross section showing unequal development of palisade and dorsi- se leat structure (Henrickson and Hess 19054). d. L. praimosun. Leaf cross section . of palisade in an tsobilateral leaf; note also base of trichome Cleared leaf with venation Se Note venation showing equal d 3.6 mm wide (Gentry (Henrickson and Hess 19117 pattern and concentration of sclerified elements near the tip. Leaf is - .-g. L. pringlei. . Cleared leaf with venation stained. g. Terminal portion of same leaf showing marginal and terminal sclerified bundle-sheath elements - of leaves of all species (Prigge 3239). Scales inb, c,d, g = 0.1 mm; )oe. Le minus. 605). as in f oe } ina,e, f = | mm. lea) have a dorsi-ventral structure. The mesophyll of many species has poorly to well developed air spaces (Karrfalt and Tomb 1983, figs. 16d, 10a). These are lacking in L. ambiguum, L. pruniosum, and L. pringler, Stomata have anisocytic subsidiary cells; some anomocytic arrangements were observed along with the anisocytic ones in Eremogeton. Leaves receive a single trace rom a unilacunar, l-trace node. Cleared leaves show venation to be pinnate, camptodromous, and brochidodromous (Dilcher 1974; fig. 2 e, f, g). Secondary veins are few (2—3), abruptly curved upward near margins adjoining superadjacent secondary veins at right angles, sometimes the lowermost pair, however, do not join. Tertiary veins form orthogonal reticulae with veins diverging at right angles from secondary and midveins. Areoles are well developed, small to medium sized (Dilcher 1974), mostly (3-)4(-5)-angled, without or with simple, linear, rarely curved veinlets (fig. 2e—g). Leaf venation 1s distinguished by development of groups of enlarged sclerified xylem tracheids located at tips of veinlet endings in areoles of some species and the development of larger, more conspicuous, sometimes continuous, aggrega- tions of similar sclerified elements along the distal leaf margin, with greatest development at the leaf tip in all species. These more massive groups of sclerified cells consist of xylem tracheids and adjacent sclerified bundle-sheath elements with somewhat elongated, fully bordered pits. In some leaves layers of leaf mesophyll cells are also sclerified and exhibit small non-bordered pits. Development of marginal and terminal sclerified ele- ments appears identical to that found in leaves of Fouquieriaceae (Henrick- son 1972), where they were designated water-storage tracheids. Lersten and Carvey (1974) questioned their function in water storage in Fouquteriaceae and referred to them as sclerified veinlet elements. Their role in water transport-storage is unknown. Vv URE: Vestiture provides important and useful taxonomic charac- teristics in Lewcophyllum and Eremogeton. Trichomes cover the leaves, young stems, petioles, calyces, in some species the ovary and fruit apex, style base, and various portions of the corolla. Several types of trichomes occur together in any one species. All species have short, stipitate glands in the understory of the longer trichomes on both leaf surfaces, on calyces, and occasionally, on the stems (fig. 3c, f, i). The trichome stalks are uniseriate, consisting of one slightly elongated or short, thick-walled cell with a thin-walled short cell immediately beneath the gland (fig. 3c, i). The gland in most species of Leucophyllum and Eremogeton consists of 2 or 4 to 7 vertical cells. In L. pringler the trichome gland is considerably larger and is divided vertically into 10-15 cells; the glands appear to be “‘sessile’”’ but are actually short-stalked, L116 sunken into the leaf surface. Cronquist (1981) noted that glands divided by vertical walls are characteristic of Myoporaceae. However, nearly identical stipitate glands with glands vertically divided into 2 cells occur in Ver- bascum and other genera in Scrophulariaceae. In Eremogeton trichomes on vegetative portions of the plants are unt- serrate, multicellular, mostly unbranched, and typically antrorsely curved (figs. 3. g —h, 20d). Occasional trichomes are branched and one branch may terminate with a gland (fig. 3h). In dried specimens the thin, transparent walls of individual cells are often collapsed. In Leacophyllum trichomes are once dendritically branched and consist of a uniseriate, multicellular central axis bearing |—4 lateral, divergent, single-cell radii or “branches” at each “node” or cell junction of the central axis of trichomes in all species (fig. 3 b) except L. pringlei, which has multicellular lateral radu (fig. 3 d, e). Interspecific trichome variation involves the relative elongation and numbers of central-axis and branch cells. The trichomes may be very short, with short, tapering radii (L. langmaniae, L. minus, L. laevigatum, and some L. frutescens) or tall, with slender radii to 3(-5) mm long (L. ambiguum, L. pruinosum, L. ultra- monticola) or they may have an elongate central axis with short radii (L. candidum), Constituent cells have clear, transparent, smooth walls that may be relatively firm and remain terete at maturity or they may be thin and collapse upon drying. The central axis may be straight or may zig-zag at the junction of each cell. Commonly tall, much-branched trichomes have no lateral branches on the lower portion of the axis (fig. 16 d). Trichome radii typically are of equal length along the axis as in the “bottle brush” trichomes on the young stems of L. candidum (fig. 16 ac). Occasionally, however, in L. candidum and others, radit on the distal portion of the trichome as distinctly shorter. In many species young leaves are densely woolly-tomentose but vestiture is reduced in stature and density 1n mature leaves. This occurs partly due to the increases size of mature leaves but it is mostly because of weathering of terminal portions of individual trichomes. In instances where the more distal radii are shorter and proximal radii longer, the change in vestiture pattern, from one high stature with short radii to one of low stature with long radit on older leaves, can be considerable (fig. 16 c). The most dramatic change in vestiture in young to old leaves is seen in L. laevigatum ina taxon described as L. virescens (= L. laevigatum var. griseum) in which young leaves are white with densely-crowded trichomes with short radii. Older leaves have a sparse vestiture of trichomes with much longer radii to 0.3(0.5) mm long (fig. 14a, b). This apparently is not due to loss of the distal trichome radii, but, it appears that trichome radit development 117 Figure 3. Trichome structure in Leacophyllam and Eremogeton. a-b. ee eats a. Unicellular trichome from inner floor of corolla tube. Note somewhat c : sculptured outer surface indicated in section at night. Note undulate pattern 2 aa inner corolla suena 8 ells. b. Long dendritically branched trichome with 2 radii per “node” showing multicellular nature of central axis. Other trichomes may have 3-several aon peer node pees and Hess 19694). c. L. revolutum. Glandular trichomes from leaf (I (right). Note each consists of 2 stalk cells, the lower with thickened walls, the Can eiehee twin eile and glandular contents. Head of glands throughout the two genera ve only vertically-arranged cells, 2— 15 in number. Note top view of gland with 4 cells on upper left (Henrickson and Hess 19074). d—f. L. pringler. d. Branched trichome from stem showing multicellular structure. Note multicellular radii. ¢. Trichome with lateral pro- tuberence accounting for branching f. Leaves have shorter, le haped glands with 2 stalk cells and multicellular heads with 10— 15 vertical cells arranged as shown in circular drawings above (R. Cruz C. 2098). g-i. Eremogeton grandiflorus. g. Stem, leaves have multicellular trichomes that typically curve distally and have ae walls (See fig. 20 d). h. Variation with branched trichome, one branch gland-tipped. 1. Glandular trichome. Glands have 4—7 cells. Scale = 0.1 mm _ 118 may be dependent on environment with leaves developing after rains having longer radii than those produced later in the season when conditions are drier. Trichomes of L. minus (fig. 14 e, f) appear stellate but actually are only compressed dendritic with several broad, thin-walled radii extending from the tip of the central axis. Other trichomes have radii extending from the top few “nodes” of the central axis. Often in this species the terminal cells of the axis may die before differentiating and thus create a darkened, gland- appearing poince at the trichome tip. In L. pring/e?, nonglandular trichomes are reduced and occur primarily on the stem, petioles, and occasionally along basal portions of the leaf midrib. The trichomes are uniseriate and multicellular and may be either simple or distally forked or branched (fig. 3 d, e). Trichome branching occurs either through multicellular radii or from lateral protuberances of individual cells (figs. 3d, E; 20 c). Cell walls in this species are firm and do not collapse. These trichome differences provide useful taxonomic characters. Trichomes of most a. are illustrated with scanning electron micro- graphs (figs. 8, 9, 10, 13, 14, 16, 20). INFLORESCENCES: ner are solitary in axils of upper leaves on terete, ascending, slender pedicels | — 5(-9) mm long in Lewcophyllum and 2— 3.5 cm long in Eremogeton. Pedicels are usually vestitured as the stems. In Eremogeton, pedicels are strongly accrescent and 4—6 cm long in fruit. Bracts and bracteoles are absent. Leucophyllum is noted for its showy, though brief, display of flowers after rains in late summer but species occasionally flower at any time of the year. The brief display of flowers in some instances may effectively tsolate sympatric species, however, simultaneous, syntopic flowering of some species can occur when the first summer rains are very late. CALYCES: Calyces are divided into 5 lobes to or almost to the base. The tube, when present, is broadly campanulate. Calyx lobes are oblong, oblong-lanceolate to -oblanceolate, acute to obtuse at the tip, entire, slightly accrescent in Leacophyllum, with sessile or stipitate glands and various other vestiture (sometimes less than the subtending pedicels) abax- tally and to some extend on the distal adaxial surface. Sepals of Eremogeton are large, oblong-oblanceolate, green, leafy, vestitures as the leaves, an accrescent. COROLLAS. Corollas in Leacophyllum are sympetalous, horizontally oriented, slightly zygomorphic, (4-)5(-6)-lobed. Corolla lobes are orbicular or Groadet than long, mostly emarginate to rounded, entire, occasionally erose. The posterior 2 lobes are external in bud (fig. | a) and reflexed at anthesis. The anterior 3 lobes are spreading to reflexed-recurved with the 119 medial lobe larger than the lateral two and all are slightly larger than the posterior two. Externally corollas may be glabrous to sparsely stellate or stipitate-glandular. Inner corolla surfaces may have short to long, tangled, unicellular trichomes 0.2 — 3 mm long with blunt, rounded tips and warty surfaces (fig. 3a) located on the throat floor, sometimes also on the throat roof, and sometimes extending to lobes; in some specimens they are restricted to the lobe margins. Corollas range from lavender, pinkish, violet to white (albino), and most have a white patch on the floor of the throat beset with irregular rows of yellow to yellow-brown spots (fig. 1 b), or they may be of a solid color or with a white patch with dark purple-violet spots. The spots apparently serve as nectar guides. Corollas of Eremogeton are very large (6 — 7 cm long), 5-lobed (fig. 2 1): the lobes are oblong with the posterior 2 lobes united nearly to the obtuse to acute tips. The posterior 3 lobes are reflexed to spreading. Corollas are stipitate-glanduar outside and weakly so inside. The lobes are ciliate with long, crinkled trichomes. The corollas are thick and whitish; according to label data, they open tn the evening. ANDROECIA: Leacophyllum typically has 4 didynamous (rarely 3 or 5) stamens. Rarely a medial, posterior staminode is present. Filaments are adnate to the corolla tube for one-fourth to one-third their total length. They may be glabrous or pilose at the base but are glabrous and whitish where free above. In the species decriptions the filament lengths recorded are measured from the base of the corolla to the anthers because insertion ts often variable even in an individual flower. Filaments of the posterior stamens extend along the upper margin of the tube-throat and abruptly turn inward just below the anther; the anthers are positioned along the roof of the mouth (fig. 1b). The relatively shorter filaments of the anterior stamens extend along the margin of the corolla tube floor and turn inward just below the anthers, which are then situated at the floor of the corolla mouth (figs. 1 b, 19 d). Anthers are white to yellowish, glabrous, bithecal but 3-locular (fig. 5 q), with the inner 2 locules shorter and distinct while the outer anther sac is longer, confluent across the anther tip. Dehiscence occurs between the inner and outer anther locules, and after anthesis the anther sacs are explanate, divergent 120 — 180 degrees (fig. 5 q). Eremogeton, 1n contrast, has only the two anterior stamens, (the two posterior stamens are sometimes represented as filamentous staminodia) with glabrous, whitish filaments inserted at the base of the corolla tube and anther sacs exserted and situated below the two posterior corolla lobes. The whitish anthers sacs are similar to those of Lexcophyllum but are divaricate only 30 —40 degrees after pollen release. 120 Pollen in all species is distinctive, 3-colporate, diorate, with mesocolpia reticulate, tectate, and exhibits some variation in sculpturing (Niezgoda & Tomb 1975). Pollen grains of Lewcophyllum range from 21—26 wm in equatorial diameter, 19—30 wm in polar diameter (Niezgoda & Tomb 1975). Those in Eremogeton are similar but slightly larger (30.6— 31.6 jum). GYNOECIA: Ovaries are superior, 2 (rarely 3)-carpelled, 2(-3)-loculed, with expanded, axile, medially-furrowed placentae (figs. 4 c, d; 5 q). Ovules are numerous in each locule, borne on the expanded placentae, campylotropous (fig. 5 n), tenuinucellular, and unitegmic. The styles are terminal, cylindrical, and glabrous or variously vestitured at the base. The style tip expands into a slightly flatcened, acute to rounded, mostly rhomboid, thickened tip that is papillate and stigmatic across the distal margin (or margins when acute). There is some interspecific variation in stigma structure; some species have blunt tips, others have more elongate, acute tips that sometimes fold back upon drying. FRUITS: Fruits are woody capsules that dehise septicidally to the base and part way to the base loculicidally. The outer 3—5 layers of the fruit wall consist of soft cells; the inner 3—4 layers consist of radially oriented lignified sclereitds with the innermost sclerified layer parallel to the inner carpel wall surface (fig. 4 d, e). Fruits are contained within the persistent, slightly accrescent calyx and may be glabrous or glabrate or persistently beset with dendritic trichomes near the tip. Capsule structure is basically identical to that found in Penstemon (Scrophulariaceae). S : Seeds are small, often somewhat flattened, angular, in Lewcophylluam frutescens 1— 1.3 mm long, 0.5 -0.7 mm wide, mostly 0.2—0.5 mm thick with shape affected by the close packing of the seeds between the expanded axile placentae and the ovary wall (fig. 4 f). Testa are brown to gray with a reticulate pattern formed as tangential walls collapse between erect radial walls (fig. 4 g). The embryo occupies about 80 percent of the seed, is 0.8— 1.1 mm long, flattened, and has two oblong cotyledons that are rounded at the tip and are about equal in length to the hypocotyl. Endosperm is about 0.1 —0.15 mm thick and consists of 3—5 layers of cells with unevenly thickened walls. Both endosperm and embryo contain oil droplets that stain with Sudan IV. Seeds are identical to those sampled in Penstemon and Verbascum except for the convoluted external sculpturing in the latter. Seeds of Eremogeton are similar in structure but slightly larger. R VASCULARIZATION: Flower vascularization of Leacophyllum was studied from serial sections and clearings of whole flowers (figs. 4 b — d, 5a—p). Pedicels contain a continuous cylinder of vascular tissue (fig. 5 a). Ten vascular traces emerge in one series in the receptacle: five medial sepal noce elements at tip of epi H wie 1905 »dD.L, pruniosum, Cross section of anther below attac ene of filament showing 4 aie oe (Henrickson and Hess 19117). ¢. L. revolutum. Cross section of ovary showing axial placenta- tion. Ovary is 1.26 mm long (Henvichson ee Hess 19074). d—e. L. frutescens. d. Cross section of fruic note locatin of sclerenchyma in fruit wall and axial placentae with few developing seeds. Ovary ts 4mm long. e. Englargement of fruit wall showing epidermis, (on right), few seakean layers and (on left) inner layer of 3 radial and innermost tangentially- elongated s nchyma c ells. Structure is identical to that found in Scrophulariaceae (Henrickson ae 19054). f—g. L. fratescens. t. SEM of seed, note surface sculpturin Enlarged view of seed surface showing reticulate pattern developed by collapse of cangential epidermal walls (Henrickson and Hess 19301). Scales ina, b, d = 1mm; inc, e, f, g = 0.1 mm. oh, LZ? traces alternate with five petal-sepal traces. The petal-sepal traces soon branch off two lateral traces, one going to each adjacent sepal as lateral sepal traces. The remaining adaxial vascular tissue then continues as the petal trace (fig. 5b —d, 1). Each sepal then receives three separate traces (fig. 5 p) that branch and anastomose distally in the lobes. They are associated with sclerified bundle-sheath cells similar to those in the leaves (figs. 4 a, 5 p)- Similar thickenings also occur along the thickened basal portions of the sepal traces (figs. 4a, 5 p). The remaining five petal traces each branch into three traces at the base of the corolla tube and continue co branch and anastomose further up the corolla tube and into the lobes (fig. 5 0). After the sepal and petal traces diverge, the remaining receptacular vascular tissue forms into a cylinder giving rise to the four stamen traces (shown in black in fig. 5 g—j). These merge into the corolla, eventually becoming free some distance above the corolla base. The remaining vascular tissue organizes into the ovary traces. That portion of the vascular tissue that would have gone to the posterior or fifth stamen develops into the dorsal trace of the posterior carpel thus affecting the vertical orientation of the ovary. The remaining vascular tissue forms into the other dorsal, vente and lateral traces, with the dorsal traces continuing into the style (fig. 4a). The pattern illustrated in fig. 5 a— | was found in all flowers of Lewcophyl/um studied, except in one flower of L. fratescens in which the lateral sepal traces of two sepals developed from the medial sepal trace rather than the alternate petal traces. The vascularization pattern was identical to that found in both Bere Flower vascularization and structure of Lewcophyllam. a— |. Diagrams of vascular as seen in serial cross sections of L. Meee at levels indicated in k. Stamen traces are indicated by black circles. Note sepals receive 3 traces, medial traces directly from recepta- cle, lateral pes branch from oe petal traces. Also note origin of stamen traces ( Henrichs 19/11 . Cutaway longitudinal eee of flower showing patterns of vascular bundles in pees and es ate levels of sections a— j. 1. Two-dimensional diagram of vascular system as seen from inside flower cut between anterior lobes, with lowest, innermost series of traces going to ovary (D = dorsal traces, V = ventral traces), second series to sepals (SE), note medial trace develops directly, lateral sepal traces develop with petal traces. Petal traces branch at corolla base, stamen traces terminate with black circles. m-r. L. frutescens. m. Cross section of ovay showing 2 carpels, dorsal traces (D), expanded axile placentae, ovules. n. Camplyotropous ovule (diagramatic). o. Vascular system of corolla as seen from adaxial surface cut between 2 posterior lobes. Note basal branching of initial 5 petal traces and basal portion . 4 stamens. p. Vascular system of : ified mature calyx showing 3 traces, distal trace branching an at f sclerified elements near sepal tip. q. Anther structure as seen from adaxial (left), ~— abaxial side views and after anthesis (right). Note outer thecum is continuous around 2 inner locules. r. Style tip showing marginal papillate thickened stigmatic = (m-r from Henrickson and Hess 19074). Scale above a = | mm holds for a-j; ino = | em; ing = | mm 15 No Ds Penstemon and Myoporum laetum except for the 3-carpelled gynoecium and reduced ovule number in the latter. Vascularization of flowers of Eremogeton was not studied as no material was available. CYTOLOGY Flyer (1970) reported one unvouchered chromosome count for Lewcophy!- lum(n = 15, for L. minus). He noted that meiosis apparently is very rapid in pollen parent cells, for, despite numerous attempts, no stages between prophase and the tetrad stage could be seen with this one exception. The number n = 16 may reflect x = 8; x = 8 has been reported for other Scrophulariaceae (e.g., some species of Verbascum and Mimulus and through- out the genera Antirrhinum, Pedicularis, and Penstemon). Myoporaceae has x = 18 in Eremophila, (Barlow 1971) and x = 27 in Myoporum (Hair and Beuzenberg 1959). SPECIES RELATIONSHIPS Various phenetic and cladistic analyses were performed on character data obtained from Leacophyllum species. Problems arose with characters involv- ing vestiture. Plants with dense stem-leaf vestiture had an uneven and apparently meaningless, though species-specific, continuation of vestiture onto various floral features. Certain species, however, grouped together in most analyses, including: (1) L. ambiguum, L. ultramonticola, L. pruinosum; (2) L. laevigatum, L. griseum, L minus; and (3) L. zygophyllum, L. candidum. Leucophyllum pringlei usually did not group closely with other species. The positions of L. fratescens, L. revolutum, L. langmaniae, and particularly L. flyrtt were very variable. Use of different or reduced data-sets gave variable ow results but from this came a synthesis of data resulting in a tentative tree (fg. 6), derived by applying the method of grouping of species by shared derived character states (synapomorphies). This is presented in the form of a cladogram to faciliatate representation of character states in relation to proposed relationships. The basic structure of the tree was derived from a reduce data-set using only || characters but these characters are considered to be significant. Relationships at some upper points of the tree were resolved through other data-sets. Attempts to apply the outgroup method (Stevens 1980) of character polarization were largely unsatisfactory. Rela- tionships undoubtly lie within Scrophulariaceae but exactly where is not known. However, as all species of Lewcophyllum are woody, have some form of branched, dendritic vestiture, have glands with 2 stalk cells and multi- cellular gland heads with vertically-oriented cells, have bractless, ebrac- teolate flowers with 5 subequal corolla lobes, (the posterior 2 lobes external in bud), and have 4 anthers with continuous outer thecae and bicarpellate [25 ovaries with axial placentae, etc., these characters are considered to be primitive (pliesomorphic) for the genus. Factors such as thorny, reduce habit and degrees of vestiture development are considered to be highly variable and poor characters for estimating (guessing at) phylogeny. The pattern that emerges from phytogeographic evidence (see below) ts that Lencophyllum may be an old group that has been in existance in the altiplano of Mexico in arid and semiarid habitats since perhaps the late Eocene (Axelrod 1979) and its history may be marked by hybridization (as ts occurring today) with derivatives recombining characteristics and thus its phylogenetic history may include reticulations between major lines. In figure 6, trichomes of moderate stature with single-celled, moderate- lengthed radii, cuneate, alternate leaves, and yellow corolla-throat spots are considered plesiomorphic. From these states, opposite, linear or orbicular leaves with either very long or reduced vestiture with very long and slender or short radii and development of purple or no corolla throat spots are considered derived. Vestiture types appear to have evolved more than once. Vestiture varies throughout the range of L. frutescens from very tall to highly reduced and obscure. In its reduced state the vestiture is very similiar to that found in L. dangmaniae, which we do not, however, consider to be closely related. Corolla throat spotting varies throughout the range of L. ambiguum. Most populations have no corolla spots but collections near Meztitquitlan in Hidalgo have yellow corolla-tube spots—is this an atavistic or apomorphic feature? Data summarized in fig. 6 must be considered tentative but reflective of possible relationships of the taxa of Leacophyllum. Our biggest questions concern relationships of L. f/yriz. It is similar to L. revolutum in its cuneate leaves and purple corolla spots but, unlike L. revolutum, its leaves are isobilateral not dorsi-ventral. It also shares characteristics with members of the L. ambiguum-pruinosum complex, which can also have purple spots on the corolla floor, isobilateral (but orbicular) leaves, and long trichome radii. Eremogeton, with its large, toothed leaves and large white, more strongly zygomorphic flowers with only 2 stamens, is considered to be distinct from the hypothetical ancestor of Lewcophyl/am. It is not included in this discus- sion. Because of the large number of apomorphic features that distinguish L. pringlet, it is here considered as a distinct subgenus. PHY TOGEOGRAPHY Leucophyllum is one of several genera endemic to the arid and semiarid regions of Mexico and adjacent United States (Rzedowski 1962, 1973). Axelrod (1979) considered Leacophyllum among those genera that evolved authochthonously in this region, perhaps from ancestors present during late Cretaceous-Paleocene times (50 — 60 mybp) that initially adapted to local Ls 9c L. PRUINOSUM I i var. LAEVIGATUM Short, th y h ; tyl ry pil ; A lla lob lab 1 ? : . 1 , y un 1 L wes. L. AMBIGUUM . des ULTRAMONTICOLA L. var. GRISEUM Tall : Pp) t th 1 , inner lower corolla lobes een Plants thorny; itrich j ded 1 La FLYRII 1] ith } i D os fa bad th on throat. o a cay long stipitate ae x and corollas; kame. L. LANGMANIAE ae om lob ! ] ou Corolla tubes broadly REVOLUTUM campanulate. ich dii t 1 Serene, Vestit d ig with age. Leaves dorsiventral, margins revolute L. FRUTESCENS Leaves Sear ieny p a L. CANDIDUM Corolla lobes ciliate. Leaves cuneate; Leaves all opposit i Corollas pinkish; 1 Na fl conduplicate, ichomes tall, leaves conduplicate, titure low. radi long at base. dorsiventral, lateral veins raised. llas: Glands on corollas Trichome radii broad. inner calyx glabrous. Leaves partly opposite; blades orbicular. di L. PRINGLEI Leaves cuneate. pos SS ow on 49 Trichomes simple, forked, radii multicellular; Glands with 10-15 cells; Leaves mnt oor nae vestitured; at throat. Trichomes dendritic, radii simple; Glands sine 2-4 cells; Leaves cuneate to orbicular; bruptly expand YW L27 semi-arid sites and became adapted to the more arid climate that arose in late Eocene and later times. A possible scenario for development of the genus involves the development of arid zones in central Mexico from Tertiary onward. In early Teritary local dry sites were scattered all across central Mexico in the lee of mountains allowing plants to gradually adapt to these sites (Axelrod 1979). During Eocene, the uplift of the Sierra Madre Oriental and other parallel ranges in central Mexico (the Hidalgoan Orogeny of de Cserna; see de Cserna 1960; Guzman, and de Cserna 1963) and additional volcanism resulted ina more widespread development of arid sites throughout Mexico. Much later in Miocene-Pleistocene, development of the igneous-rock Sierra Madre Occidental caused still further expansion of these arid and semi-arid zones throughout central Mexico and allowed taxa to adapt to arid- and semi-arid habitats to spread throughout the expanded dry zone that extended from Puebla to northern Mexico. Vegeta- tion in the Miocene and early Pliocene in central Mexico, which according to Axelrod (1979) probably was warmer and had higher precipitation than today, may have consisted of dry tropical forests and woodland with thorn scrub and drier edaphic and lee-slope-habitats supported semi-arid taxa. During late Pliocene to Recent time extensive volcanic activity formed the trans-volcanic belt (Guzman and de Cserna 1963) which isolates the Pueb- lan arid zone (present home of L. pring/ez) from those in the north; later uplifting and volcanic activity isolated the Hidalgoan arid region from that of the Chihuahuan Desert Region in the north (Axelrod 1979). While the southern Puebla arid regions were protected from cold winter northern fronts by the trans-volcanic belt, species diversity in the northern Chihua- huan Desert was impoverished by increasingly colder climates particularly during Quaternary glacial periods. During the Quaternary these semi-arid and arid zones were alternately reduced and expanded during pluvial and interpluvial periods. Van Devender (1977) noted that during the latest pulvial period pinyon-juniper-oak woodland vegetation extended into the areas now occupied by the northern Chihuahuan Desert, which expanded to its present conformation only within the last 8000 (- 4000) years. With the absence of fossil evidence it is not possible to know exactly how Leucophyllum and Eremogeton fit into such a scenario. Consideration that the nearest realtives of these genera of Scrophulariaceae have always been considered to be Old World herbacecus and suffrutescent genera implies that Lewcophyllum and Eremogeton may be New World vicariants or at least Figure 6. Diagram of an intuitive phylogeny of the species of Lewcophy/um. See text for explanation, 128 New World survivors. Michener’s xylem ontogeny data indicate that the shrubby habits of our taxa may be a derived condition. Leacophyllum particularly appears to be well adapted to arid and semi-arid habitats in central Mexico exhibiting a number of xeromorphic features and, un- doubtedly, a number of physiological features that allow it to survive in these environments. The present distribution of taxa appears to fit well into a vicariance model. The distinctive Eremogeton is geographically well isolated from the more northern species of Leacophyllum. Lencophyllum pringle:, which here ts considered to be the sole member of a separate subgenus, occurs in Puebla and Oaxaca, south of the late-Pliocene-Recent Trans-volcanic belt. The other species of Leacophyllum occur north of this volcanic axis, one in Hidalgo-Querétero, the others in or around the Chihuahuan Desert. These eleven species show various distribution patterns in related lines. The three species with long trichome radii (L. ambiguum, L. pruinosum, and L. ultramonticola) seem to form a related group and are widely disyunce: L. ambiguum occurs in the southern disjunction of the Chihuahuan Desert in Hidalgo and Querétaro, L. pruinosum in basins in southern Nuevo Leon and adjacent Tamaulipas and San Luis Potosi, and L. «/tramonticola in an arid pocket along the Rio Atengo in Southeast Zacatecas (fig. 16). Their present ranges probably reflect their respective refugia during the past pluvial period. It 1s of course attractive to propose that they may have been one widespread species during the previous interpluvial period and that subse- quent isolation of refugial populations resulted in the differentiation of these three closely related vicariant spect Many of the present day species of Lem len occur in a Chihuahuan Desert scrub, mostly in a mixed desert scrub association above the Larrea zone up to the Yuacca-Dasylirion-dominated zones; their ranges were prob- ably restricted during the latest pluvial period, or at least they were restricted to arid habitats within the presumed dominant woodland com- munities. The presumed bottlenecks in population size could well provide for more rapid character differentiation that is reflected in today’s more expanded populations. At present, populations of L. candidum, L. minus, and, particularly, L. fratescens seem to be expanding their ranges. The latter species appears to be expanding northward from a potential refugium east of the Sierra Madre Oriental into southern Texas and spilling over into the Chihuahuan Desert. The occurrence of three species of Lewcophyllum in the basins around southern Nuevo Leon separated from the main mass of the Chihuahuan Desert by series of low mountains is of considerable interest. This area may well have served as a refugium during the past pluvial period: 129 itis the present-day home of L. prainosum, L. revolutum, and L. zygophyllum. Many other endemics also occur in this area Three species of Lewcophyllum appear to be relictual and occur only in reduced populations. Lewcophylum langmaniae occurs in isolated arid pock- ets in the woodlands in canyons between Monterrey, Nuevo Leon, and Saltillo, Coahuila. The geographical extent of this species is not known. Lencophyllum flyri7 is known only from two populations, one near Laguna Seca (Gral. Candido Navarro) and one in the adjacent sierras in central San Luis Potosi. The extent of L. “/tramonticola in southwestern Zacatecas is unknown, ECONOMIC. USE Lencophyllum frutescens ts commonly cultivated and is known as cenizo, ceniza, Texas silverleaf, Texas ranger, purple sage, and, less frequently, barometer bush, liar bush, ash bush, Texas rain sage (Texas Agricultural Experiment Station Tract L-2058). The species has been widely cultivated in south-central Texas and is becoming popular throughout the Southwest in relatively frost-free, arid regions. The plants can, however, tolerate light to moderate frost. The species is popular because of its white-gray foliage, rounded habit, and great show of pink-purple flowers after rains in late summer-fall. According to the Texas Agricultural Experiment Station, the plants prefer open sun and alkaline soils with good drainage. Recently the Texas Agricultural Experiment Station has made available a white- flowered, gray-leaf cultivar ‘White Cloud,’ and a green-leaved, purple- flowered cultivar ‘Green Cloud’ of L. /rwtescens, and a deep violet-flowered, gray-leaved cultivar of L. candidum under the name ‘Silver Cloud.’ Plants are propagated by cuttings. Flowering plants may be produced from seeds in about two years. TAXONOMIC. TREATMENT A. Corollas 60-70 mm long, white, posterior 2 lobes united to near tip, erect, anterior 3 lobes oblong- ovate, ees 25—33 mm long, Me — 15 mm wide; calyx lobes 2— 3.6 cm long, 4 — 7 mm wide; stamens 2; trichomes multicellular, tapering, pete. curv os pot 4—12cm long; México (Chiapas) and Guatemala ........... . Evemogeton. E. erandiflorus AA. Corollas 8.5 — 28 mm long, lavender, violet, pinkish, | white (albi- e no), often with white on floor of tube marked with yellow or purple dots, lobes all orbicular to broadly oblong, ae ul, 3— 10 mm long and wide; ca ae lobes 5— 10 mm long, 0.7— 1.5 mm wide; stamens 4, didynamous; trichomes multicellular, erect, stellate or dendritic or ee dist ae ice 6—35 mm long; Mexico (Oaxaca) to sw. U.S... .. 1. Leacophyllum Leaves linear-lanceolate, 6— 12 mm long, 1.2—2.2 mm wide, ed, glabrous except for sessile glands; young stems with erect, Crowc tapering or distally forked trichomes 0. 1 — 0.25 (-0.4 Puebla doa OaXaCde we ad ees Ak eee & eae eee Dee one od BB. I lat i) mm long; L. pringles , obovate to orbicular, mostly 2.5 — 12(-22) mm wide, tomentose or with scattered stellate or dendritic ane : young stems canescent to tomentose with stellate to dendritic trichomes; Hidalgo and northward. CC. Leav ives bicolored, upper leaf surfaces more green with slightly to much-reduced vestiture (be sure to look at both surfaces o one leaf). £ D. Leaf blades oblanceolate, 2 — 4(-5) mm wide, at mar- gins distinctly revolute, often inrolled towards leaf base; corollas violet with dark violet spots inside on floor of tube; sw. Tamaulipas to adjacent n. San Luis Potosi. ~~ L. revolutum DD. Leaf blades obovate to oblong-obovate, (4-)6 — 16 mm wide, margins flat or variously undulacte, noc revolute; corollas pinkish to lavender, with yellow spots inside on floor of tube; w. central Texas s. to e. central Coahuila, Nuevo Leon, s. Tamaulipas. .................. | Leaves concolorous, upper and lower surfaces equal in vestiture and color (although sometimes more strongly vestitured along midrib beneath). E. Vestiture of leaves and young stems densely and closely ilver-gray canescent, trichomes crowded, overlapping, eee stellate, radii broad, translucent to whitish, to 0.1 mm long, radiating from a central, often gland- tipped, axis (use 30 X magnification); leaves small, crowded — (aw, at nodes; ee) shrubs from sw. New Mexico through Trans-Pecos Texas to e. Chihuahua, s. to about SO km ne. of Saltillo, Coahuila EE. Vestiture of at least young stems and often of lower leaf margins of dendritic trichomes, this often of an uneven stature, or if not, then either older leaves greenish or trichome arms slender, longer. Leaf-blades mostly ovate-orbicular, abruptly narrowed at base; petioles usually conspicuous; leaves silver-gray. G. Leaves all opposite, usually conduplicately folded along midrib; leat-blades commonly orbicular or nearly so, closely vestitured; Nuevo Leon, sw. Tamaulipas and adjacent San Luis Pocost. GG. Leaves mostly alternate, occasionally some leaves opposite or sub- opposite, seldom conduplicately folded; leaf-blades orbicular or not, with dense, thick vestiture Corollas unformly violet to purple throughout, occasionally with white along very base of tube but without colored dots in lower tube; Hidalgo to Querétero. 2... ...00....0.-. Bierey 2 HH. Corollas lavender to violet but ae lighter or white patch on floor of tube marked with yellow or deeper violet spots; plants of more northern distribution. i frutescens L. minus zygophyllum . ambiguum F Lag} _— Dendritic trichomes of young stems and leaf blades long and slender, mostly 0.1 —O0.5 mm in diameter with radii 0. 1 — 0.25 mm long, only a few times longer than the thickness of the trichome axis; compact shrubs of Brewster Co., Texas, s. through central Coahuila to Zacatecas, ¢. Durango, central Chihuahua....... 5S. L. candidum 5 — _ Dendritic trichomes of stems and leaf-blades 0.3 — 1. mm in diameter, with radi (O. 1-) 0.2 —0.5(-8) mm long, many times longer than the thickness of the axis; shrubs from s. Nuevo Leon to Zacatecas. Style, ovary, and capsule tip glabrous or with few glands; corolla lobes pilose inside; sw. Zacatecas. go cocina ae ba as Os 3 Gy Be ARO eco een ay aE 10. L. awltramonticola Style, ovary, and capsule tip pilose to densely — — pilose; corolla lobes glabrous inside; s. Nuevo Leon, adjacent Tamaulipas, San Luis Potost. BS GSAS ks Wee ae ete ts Sk eb dee ose QD. L. prumosum Leaf-blades obovate-oblanceolate, cuneate, gradually narrowed to base; petiole i" Sa not conspicuous; leaves green or not K. Cal ae x lobes with scattered long-stipitate glands extending well e the vesticure; corolla with dark purple (not yellow) dots on floor of tube inside; leaves permanently gray-canescent; ne. of city of Saisie: POLS st at daa week eau thee yh Wee yan ee eee 8. L. flyrii A if Calyx lobes lacking long-stipitate glands; corollas with yellow dots on floor of tube inside; mature leaves often greenish LL. Leaves appearing glabrous but uniformly covered with re- latively dense, but minute trichomes to 0.05 mm wide on both surfaces; radit about as long as central trichome - corolla lobes not ciliate; local between Monterrey, Nuevc Le6n, and Saltillo, Coahuila... .....00.0.........2. ' langmaniae Leaves with scattered to dense, stellate trichomes 0.1 — 0.2 —~ -0.8) mm broad; radit mostly 2 times longer than trichome axis; leaves either green or the youngest leaves (occasionally gray; corolla lobes distinctly ciliate; plants from Chi- sot] — huahua and Coahuila, to San Luis Potost, Durango, and WHACATECAS 2... dep dah fh dod esd daarae ayers piney pS EE 4. L. laevigatum I]. LEUCOPHYLLUM Bonpl. in Humb. & Bonpl., Pl. Aequinoct. 2:95. 1812. Type: Leacophyllum ambiguum Bonpl. in Humb. & Bonpl. » Terania Berlandier, Mem. Com. Limit. Mier y Teran 4. 1823. Type: Terania frutescens Berlandier Rounded to rhomboid, moderately to strongly branched, evergreen to semi-deciduous, silvery-gray to greenish shrubs; young stems subterete, uniformly to irregularly tomentose to canescent with stellate to dendritically-branched, rarely forked, spreading trichomes, 1n age more uniformly tomentose-canescent due to weathering, eventually glabrate; [oz older stems with dark to light gray or reddish-brown, _ or vertically fissured bark; older dead stems sometimes persisting as thorns. Leaves alternate, subopposite to opposite, simple, often crowded or acne also in reduced subfasciculate, axillary shoots, oblanceolate, obovate, elliptical, orbicular to ovate-orbicular, obtuse, rounded to acute or emarginate, with midrib continuing as a blunt, sometimes reflexed apiculation at tip, cuneate, sessile or subsessile to abruptly cuneate-rounded and distinctly petiolate at base, entire to revolute, often somewhat conduplicately folded along midrib or otherwise undulate, soft, pliable but usually chickish due to thick indumentum, silvery-gray to greenish, densely tomentose to canescent or sparsely vestitured with dendritically-branched or stellate trichomes, rarely glabrous. Trichomes uniform on both surfaces or shorter or more sparse above, consisting of a series of long, slender or short, tapering, straight or wavy simple radii extending from a short to elongated multicellular central axis that may fragment causing a reduction in vesti- ture stature through a season, mostly with sessile to stipitate glands in understory, with midrib and sometimes secondary veins raised beneath, abscissing above prominent, usually persistent leaf bases. Flowers 1(-2) in axils of leaves, mostly produced in abundance after rains; bracts, bracteoles absent; pedicels ascending, vestitured as young stems; calyces divided to near base into 5 lobes, lobes lanceolate to oblong-ovate, acute to attenuate, valvate in bud, tomentose, canescent as leaves or young stems or with a diminished vestiture, obscurely glandular, rarely long stipitate-glandular outside, glabrous, glandular or sericeous or sparsely stellate except a tip inside; corollas showy, lavender, light violet to pink-lavender, blue, violet, rarely white, mostly with one or more white patches on floor of tube marked with rows of yellow-brown to orange dots, or of a solid color marked with dark purple-violet spots, zygomorphic, funnelform to campanulate, tube gradually to abruptly ampliate, usually slightly to moderately dorsi- ventrally compressed, lobes (4-)5(-7), shorter than tube, imbricate in bud, spreading, posterior two sometimes more reflexed than anterior three, often with long, slender, tangled unicellular trichomes on floor of tube and on lower throat, with straight, erect, slightly clavate, colored hairs on throat and often on inner surface and margins of lobes, glabrous to sparsely glandular-pilose, rarely sparsely stellate outside; stamens (3-)4(-5), di- dynamous, included or the longer, posterior pair slightly exserted; fila- ments adnate to base of corolla tube for one-fourth to one-third length, glabrous or pilose near base, cylindrical to slightly compressed, whitish above, posterior pair spreading, inwardly curved at tip, anterior pair upward] y curved at tip; anthers white to yellowish, glabrous, anther sacs becoming widely divaricate (120 — 180°), outer locules confluent, inner locules distinct, dehiscent across continuous end, promixal anthers oriented parallel co tube axis, anterior pair oriented nearly perpendicular to floral axis; Ovary superior, ovoid, glabrous or pilose or loosely stellate at tip, 2-loculed; placentae oblong, attached medially along septum; ovules many; [32 styles terminal, cylindrical, glabrous to sparsely pilose, straight, extending to posterior corolla tube- throat, sometimes slighty exserted, decurved at tip, flattened, stigmatic along distal margin of terminal, obtuse to some- ae acute tip. Fruit of dark brown, woody, ovoid, apiculate capsules, these dehiscing first septicidally to near base, then loculicidally half way to base, enclosed in slightly accrescent calyx; seeds 15 — 25 per locule, small, irregulary ovoid, yellowish-brown, minutely reticulate, smooth to some- what angled. Chromosome number 7 = 16 (one species). LEUCOPHYLLUM Bonpl. in Humb. & Bonpl. subgen. LEUCO- HYLLUM Trichomes dendritic, with unicellular radii; glandular trichimes with 2-7 cells in head; leaves oblanceolate, obovate to orbicular, mostly covered with trichomes on both surfaces. Distribution: Texas, adjacent New Mexico, south from Chihuahua and Tamaulipas to San Luis Potosi and Zacatecas, also in Hidalgo and Quereé- taro. (Species No. 1— 11). LEUCOPHYLLUM Bonpl. in Humb. & Bonpl. ees Faxonanthus (Greenman in Sargent) Henrickson & Flyr, sg et s nov. faxonan- t Greenman in Sargent, Trees & Shrubs 1:23. pl. 12. ao ee ne pringlei Greenman in Sargent Trichomes simple or distally forked, with multicellular radi1; glandular trichomes with 10—15 cells in head; leaves linear-lanceolate, glabrous except for sessile glands. Distribution: Southern Puebla and adjacent Oaxaca. (Species No. 12) 1. LeEUCOPHYLLUM FRUTESCENS (Berlandier) I. M. Johnston, Contr. Gray Herb. 70:89. 1924. Terania frutescens Berlandier Mem. Com. Limit. Mier Y Teran 4, 1832. Typr: MEXICO. Nuevo Leon: Monterrey, Berlandier 1406 (Lectotype: BM!; ISOTYPE: OXF! Lencophyllum texanun Benth. in DC., Prodromus 10:344. 1846. Tyvek: TEXAS. Wess Co.: “Laredo,” Berlandier ae 0 (LECTOTYPE: here designate Eesha (Berlandier) 1. M. Johnston forma a/biflor unc lover, Madrono 4:97. TYPE: XAS. Srarr Co.: Roma, Clover 492 (HoLotryPE: MICH!) Lop nese (Berlandier) I. M. Johnston forma a/binewm Lundell, Contr. Univ. Mi 1942. Type: TEXAS. CAMERON Co.: 8 mi W of Boca Chica, 12 ae 1942. C.L. Lundell and A. A. Lundell 10699 ee pe: MICH!; tsoryees: LL!, TEX!). ee ) 1. M. Johnston var. ee R. A. Vines, nom. nud. T ind woody vines of the southwest p. 920. 1960).] anon pn (Berlandier) I. M. Johnston var. glaucum R. A. Vines, nom. nud. bs, and woody vines of the southwest p. 920. 1960.] \ a jon Erect, rounded, alternately-branched shrubs 0.5 — 2(-3) m tall; young stems densel ly tomentose with conical to cylindrical, dendritic trichomes 134 0.1 —0.3 mm long with short tapering radit 0.05 — 0. | mm long, tardily glabrate; older stems with reddish-brown to light-gray bark. Leaves alter- nate, rarely opposite, obovate, oblong-obovate, to obovate-orbicular, 10 — 25(-35) mm long, (4-)6— 16 mm wide, obtuse, rounded, often bluntly apiculate at tip, cuneate to petiole 1—2 mm long at base, at margins entire, sometimes slightly revolute, usually conduplicateley folded along midrib, mostly silvery-gray, tomentose on both surfaces but with shorter, more open vestiture and more gray-green to green above, upper surface with dendritic trichomes 0. | — 0.3 mm high or mixed dendritic and shorter stellate trichomes or rarely of reduced, scattered stellate trichomes to 0.1 mm long, sometimes appearing nearly glabrous except for glands, lower surface mostly densely, irregularly tomentose with dendritic trichomes 0.2—0.3(-1.0) mm high, radit mostly straight, slender, 0.1—0.2 mm long, midvein and secondary veins raised beneath. Flowers with tomentose pedicels |—4 mm long; calyces 5—7 mm long, lobes oblong-lanceolate, 3—5 mm long, 1.3—2.1 mm wide, acute, densely tomentose with dendritic trichomes as on stem outside, more glabrous, strigose slightly glandular in lower half inside, to6 mm long, 2.5 mm wide in fruit; corollas rose-lavender, light violet, rose-pink to reddish-pink (rarely white), with white patch marked with gold-brown dots on floor of tube, 18-26 mm long, tube ampliate, to 7-12 mm wide at throat (pressed), lobes 5(-7), oblong to reniform, reflexed-spreading, 8 — 10 mm long, 6-9 mm wide, emarginate, sparsely pilose on floor of tube, more densely pilose at throat with straight trichomes 0.5 — 2 mm long, lower lobes pilose and ciliate, upper lobes ciliate only, corolla glabrous through- out outside; stamens (3-)4(-5), anther glabrous, posterior filaments 5 — 11 mm long, anterior filaments 4—7 mm long, glabrous; styles 9- 14 mm long, glabrous; ovaries g ee Capsules 3.5—4.5 mm long, glabrous. Leucophyllum frutescens (figs. 1, 7 — 10) is characterized by its relatively large shrub habit (1 — 3 m an moderately large, obovate, cuneate-based, bicolored, dorsiventral leaves with both mid and major lateral veins raised eneath, and the relatively large, mostly pinkish to lavender (rarely white) corollas. The species exhibits considerable variation in leaf vestiture over its range. In all instances, vestiture is less well developed on the upper surface than on the lower surface. At one extreme, generally in plants in the northern portion of the range, both upper and lower leaf surfaces are covered with erect, dendritic trichomes generally 0. 1 —0.3 mm tall on the upper surface and 0.2 —0.5(-1.0) mm tall on the lower surface (fig. 8 c-d). On both surfaces the horizonal radii are 0.06 —0.15(-0.2) mm long. Upper leaf surfaces may appear grayish-white or greenish when vestiture is diminished through time. In other plants (fig. 9 a—d) throughout this northern range the upper leaf surface vestiture is somewhat reduced and consists of a mixture of dendritic trichomes, sometimes with the more distal radii shorter than the Figure 7. Line drawings of Lewcophyllum frutescens. a. Stem showing orientation of leaves and flowers. b. Lateral, frontal, cutaway side views of flowers. Note orientation of longer, posterior and shorter, anterior stamens. c. Mature fruit showing characteristic septicidal and loculicidal dehiscence. (From cultivated material in Austin, Texas). Scales = | cm basal radii on a trichome, and shorter stalked or sessile stellate trichomes (basically reduced dendritic trichomes), or only of stellate trichomes mostly with radit 0.06 —0.15(-0.2) mm long. In contrast, the lower leaf surface has a taller, more dense vestiture of dendritic trichomes 0.2 —0.5 mm tall again with radii 0.06 —0.2 mm long. In these plants upper leaf surfaces often appear green in living plants while the lower surface is distinctly whitish. The diminution of vestiture continues in the southern portion of the range in Nuevo Leén and Tamaulipas with upper leaf surface vestiture consisting of often more widely scattered, reduced stellate trichomes some- times mixed with widely scattered dendritic trichomes but with radii mostly only 0.02 —0.06 mm long and occasional radii to 0.1 mm long. Lower leaf surface may consist of dendritic trichomes to 0.2 mm tall with Figure 8. Leaf vesticure of Lewcophyllum lanemaniae and L. fritescens. a—b. L. Species characterized by uniformly short oo with very reduced radit on both a surfaces. a. Mature leaf, abaxial surface. b. * frutescens. ¢. Strongly vescitured langmaniae. ichomes enlarged (LeSweur 435). c-d. leat ae longer trichomes on abaxial surface. : Expanded view of lower surface vesticure (shown here downside up). Note each “node” of central axis has 2 or 3 radii. (Compare with fig. 3b for internal struccure). (Clark et al sen. of Sabinas, Coahuila). Scales = 0.1 mm Figure 9. Leaf vesticure of Lewcophyllum fritescens. a. Cross section of leaf showing shorter vestiture than in fig. 8c, note also trichomes of lower surface are longer than those of upper surface. b. Surface view of adaxial vestiture. Note moderately spaced trichomes (Henrickson and Hess 19052; 20 km sse of Montemorelos, Nuevo Leon). c. Cross section of leaf showing unbranched base of trichomes and long radii. d. Surface view of adaxial vestiture, note long radii (Lewsy s..; near Amistad Dam, Pecos County, Texas). Scale ina = 0.1 mm; ho Dard ds for 138 radit to 0.2 mm long or only of stellate trichomes with reduced radu similar to those present on the upper surface but still more dense than on the upper surface. This series culminates in plants with leaves that appear nearly glabrous on the upper surface (fig. 10 a-b, d) but have small, well-spaced, highly reduced stellate and short dendritic trichomes to 0.05 mm tall with radii to 0.02 mm long on the upper surface, and to 0.07 mm tall with radit to 0.04 mm long on the lower surface (fig. 10.c, e). As the branched trichomes are reduced in density, the underlying gland-tipped trichomes, (which are present in leaves of all the species), become more conspicuous. In Flyr (1970), plants with such reduced vestiture from southern Nuevo Leon and Tamaulipas were recognized as a distinct variety but the pattern of variation 1s c/7na/ witha general reduction in stature and density of trichome observed in populations from the north into southern Tamaulipas. Howev- er, even in these southern populations there is a variation in leaf vestiture: some plants have reduced stellate trichomes with short radii and others have a mixture of stellate and larger dendritic trichomes with longer radii similar to those found in northern populations. Also, occasional plants from Texas such as McKsnney 039 (LL) from native habitats in Austin (fig. 10d, e) and Johnston 53257,6 (TEX) from Star County have leaf vestiture comparable to that of populations in Tamaulpas and Nuevo Leon. The only effective way to separate these northern and southern population series would be on the basis of upper leaf surface trichome size with the northern populations having radi mostly 0.06—0.15(-02) mm long, those in the southern populations having shorter radit mostly 0.02 —0.06 mm long. Although the trend of diminished vestiture is generally recognizable, the interpopulational variation, the minute nature of the vestiture differences (effectively observed only with magnifications of 30 power or more), the lack of correlated characteristics, makes recognition of varieties along this clinal variation gradient both arbritrary and difficult. Therefore, no in- fraspecific taxa are recognize From a viewpoint in Texas one tends to consider that the species is typically rather strongly vestitured and that the reduced vestiture in southern populations is a derived feature. But paleobiogeographic con- siderations tend to support the idea that populations with reduced or diminished vestiture in the south may be relictural from pluvial times and that the taxon has been spreading northward and spilling over into the Chithuahuan Desert with more densely vestitured populations. On the other hand tt 1s hardly detensible to say that the highly reduced vestiture found in some southern populations (fig. 10 a—c) is pletsomor- phic because the vestiture is so obviously reduced form a more dendritic Ley Figure 10. Leaf vestiture 2 ee hee frutescens. a. Cross section of leaf with very reduced vestiture on both sur . Compare with fig. 8 c, 9 a, c. Note air lacunae in mesophyll. b. Surface view of a vestiture. ¢. Surface view of abaxial vestiture, both surfaces are green in color (Graham and Johnston 440. 17 mi S of Victoria, Tamaulipas). d. Surface view of adaxial vestiture. e. Abaxial vestiture of plant from Barton ae Austin, Texas, showing similar ace vestiture (McKinney 039). Scale ine = 0.1 mm; holds for a—d. 140 type. It is more probable that selection has been occurring in both direc- tions over time. Leucophyllum frutescens is the widest ranging species in the genus and occurs from Trans-Pecos Texas (Brewster Co.) east through the southern Edwards Plateau to the south Texas plains south through Coahuila, north central Nuevo Leon to southern Tamaulipas (fig. 11) where it occurs mostly in limestone, calcareous, sandy to clay plains and hills from Chihuahuan Desert to Tamaulipan Thorn Scrub habitats from 5 to 1200 m. The species is commonly cultivated in Texas and over much of the southwestern United States. It is quite cold hardy and can withstand moderate frosts. 2. LeucopHyttum langmaniae Flyr, sp. nov. A speciebus alliis foliis concoloribus, pubescentits densis, ctrichomatibus stellato- dendriticis, radits 0.01 —0.04 mm longis differt. Erect, alcernately-branched shrubs 0.6 — 2.5 m tall; young stems close- ly, sometimes irregularly tomentose-canescent with dendritic trichomes 0.04—0.2 mm long, 0.04—0.1 mm wide with blunt, short radii 0.2 —0.05 mm long, tardily glabrate; older stems light brown; internodes 1—7 mm long. Leaves alternate, crowded near tips of branches, oblan- ceolate to spatulate-obovate, 10 —8 mm long, 4—9 wide, obtuse, round- ed, rarely acute, usually bluntly apiculate to retuse at tip, narrowly cuneate at base, obscurely petiolate, greenish, concolorous, both faces with fairly dense, reduced stellate to stellate-dendritic trichomes 0.03 — 0.1 mm long and wide, radii very short, sometimes scarcely developed 0.01 — 0.04 mm long, mixed with scattered glands, trichomes more dense along raised midvein beneath and on petioles. Flowers on slender pedicels 3 — 8 mm long, calyces 3.5 —4.5 mm long, lobes linear-lanceolate to obovate, 3 — 4 mm long, | — 1.2 mm wide, greenish, closely vestitured as leaves, slightly less then pedicels, greenish, sparsely pilose inside except near tip; corollas lavender-blue, with yellowish dots in floor of tube, 12 — 26 mm long, tube ampliate to 4—9 mm wide at throat (pressed), lobes obovate, reniform, 4 —8(-10) mm long, lower medial lobe emarginate, 5— 12.5 mm long, others slightly smaller, all undulate at margin, tube moderately pilose inside with tangled trichomes |— 2 mm long, pilose with shorter hairs at base of lobes near throat, otherwise glabrous, not ciliate, corollas glabrous outside; stamens 4, anthers glabrous, posterior filaments 5 — 10 mm long, anterior filaments 3 — 7.5 mm long, glabrous; style 5.6 — 10.5 mm long, sparsely pilose; ovary sparsely pilose. Capsules 4.5 —5 mm long, 3—3.5 mm wide, pilose at tip. Type: MEXICO. Nuevo Leon: hills near Monterrey, 1700 ft, 31 Aug 1903, C. G. Pringle 11056. (HOLOTYPE: SMU!; isotypes: F!, GH!, LL!, MEXU (2 sheets)!, MICH!, PH!, US!). 141 Additional collections: MEXICO. Nuevo Leon: Obispado near bay ae Feb 1909, Ablurs, n. (PH); Huasteca Canyon near Monterrey, Aug 1938, LeSvevr 435 (TEX); 17.2 mi w of Santa Catarina, 3420 ft, 8 Aug 1959, Youngpeter and Cohn 02 (MICH) and ‘ Aug 1959, Youngpeter and Cohn 80 (ENCB, MICH); ca 1 m ee ym entrance of Huasteca Canyon, 2 Aug Sls McGill, Brown, and Pinkava 9734. (ASU, ENCB), Huasteca Canyon, 0.1 mi below » No. 17, ca 850 m, 12 Aug 1983, Prigze - Michener 5311 (RSA, TEX). Leucophyllum langmaniae is characterized by its oblanceolate, greenish, isolateral leaves covered with a close vestiture of closely-spaced, stellate- dendritic trichomes with very short radit 0.01 — 0.04 mm long (fig. 8 a, b). In its uniformly short leaf vestiture it differs form all other species in the genus though a similar vestiture occurs on the upper leaf surfaces of some L. frutescens from Nuevo Leon and Tamaulipas, however, in the latter species, leaves are bicolored, with dorsiventral structure and the lower leaf surfaces have denser dendritic trichomes. This similarity in vestiture caused Flyr (1970) to consider relationships with L. fratescens. Relationships appear to lie with L. /aevigatum var. laevigatum, which also has greenish leaves equally vestitured on both sides, but with fewer, larger trichomes. In both L. langmaniae and L. laevigatum var. laevigatum newly formed leaves are also green, 1.e., the leaves are not covered by a dense layer of trichomes that eventually fall away giving rise to a more open vestiture. They differ, however, ina number of floral features. Their similarity may be entirely due to their shared reduction in vestiture. Lewcophyllum langmaniae appears to be restricted to the canyons of the Sierra Madre Occidental in Nuevo Leon between Monterrey and Saltillo (fig. 18); an area of considerable endemism. The new species 1s named for Ida K. Langman, author of the monumental guide to the literature of the flowering plants of Mexico. 3. Leucoptyttum minus A. Gray in Torrey, Bot. Boundary Surv. 115. 1859. Type: TEXAS. Pecos Co.: rocky hills of the Pecos, 4 June 1851, Wright 1481 (field number 345) (lectotype: here designated GH!; isotypes: GH!, MO!, I): — ee minus A. Gray forma uae Pennell, Proc. Acad. Nat. Sci. Philadelphia 295. 1940 [8 Apr 194 41}. Type: TE REWSTER Co.: Persimmon Gap area, 21 me 1939, O.E. Sperry aie (HOLOTYPE: “PHI: isorypE: TAES!). Intricately, alternately-, divaricately-branched shrubs 2—8(-15) dm tall, often rather thorny due to persistence of young dead branches; young stems closely canescent-tomentose with appressed stellate trichomes Q.1—0.2 mm in diameter with rather thick, tapering radii, these tardily glabrescent; older stems reddish-brown to light, rarely dark, gray: in- ternodes 0.5 — 10 mm long. Leaves alternate, crowded in axillary fascicles or on compressed lateral shoots, oblanceolate or spatulate, to obovate- orbicular, typically small, (2-)3— 10(-16) mm long, 1.8—5(-10) mm wide, obtuse to acute, often apiculate at tip, gradually cuneate (sometimes in small or broader leaves abruptly cuneate) to a petiole (0.5-)1 — 3(-4) mm 142 long, both surfaces equally silvery-gray, uniformly canescent-tomentose with short dendritic to “stellate” trichomes 0. | — 0.2 mm in diameter with radii often appearing to originate from a gland-like umbo, radit broad, tapering, thick- or thin-walled, variously twisted, 0.05 —0.1 mm long, occasionally some marginal trichomes more dendritic, midrib slightly raised beneath. Flowers on slender tomentose pedicels 1 — 3(-5) mm long; calyces 2. ) > — 4.5 mm long, lobes oblong, oblong-lanceolate, (2.2-)3 — 4.4 mm long, | —4(-1.9) mm wide, acute, strongly canescent-tomentose as stem outside, glabrous, sparsely glandular inside except at tip; corollas light purple, lavender, rarely pink, white, with white patch marked with yellow-brown dots in floor of throat, 12 — 18(-24) mm long, tube rather abruptly ampliate |—3 mm above base, to 6—7 mm wide at throat (pressed), slightly compressed, lobes subequal, orbicular-reniform, obo- vate, reflexed, 3 — 7.5 mm long and wide, often emarginate, sparsely pilose in tube with tangled hairs to 2 mm long, more pilose in throat, lobes mostly glabrous to sparsely pilose, distinctly ciliate with hairs 0.2 — 0.5 mm long, corolla glabrous outside; stamens 4, anthers with short hairs at tip, post- erior filaments 5— 11.5 mm long, anterior filaments 3— 8.5 mm long, glabrous to pilose; style (6-)8 — 12 mm long, glabrous to sparsely pilose. Capsule usually glabrous, rarely pilose, 4-5 mm long, 2.3 — 3 mm wide. Chromosome number 7 = 16. Leucophyllum minus (figs. 15 c, 13 e, f) is characterized by its low, intricately-branched habit with many old stems persisting as thorns, by its small, concolorous, silver-gray leaves crowded on short, lateral shoots and particulary by its distinctive ‘‘stellate” vestiture with thick, tapering, broad-based, but thin-walled radii extending froma central stalk (fig. 14 ¢, f) best seen under 30 X magnification. The vestiture has been described best by Johnston (1941:120): “in L. minus the indument is very dense and close and almost suggests a coating of aluminum paint. The very numerous small white trichomes are flat and stellate. The primary axis of the trichomes is extremely shortened and its top appears as a small dot or knob at the center of the radially arranged arms.” Actually, the trichomes are not stellate but compressed-dendritic with radii extending from several levels but typically with one series of radii at the top. The terminal, sometimes darkened, trichome tip ts not always conspicuous. In young leaves scattered marginal trichomes are clearly dendritic but they weather away and shorten with age. Corollas are typically light purple, usually small, and the lobes are dis- tinctly ciliate. Leucophyllum minus may occur sympatrically with L. fratescens and L. candidum in Trans-Pecos Texas and adjacent México (figs. 11, 17, 18). Flyr 143 Lae T r T T T 0 : ‘ ; ° 106 Lod Loz‘ 100° 98° 96° 32° EUCOPHYLLUM REVOLUTUM A LEUCOPHYLLUM LAEVIGATUM ° oe a JAR. GRISE a Oo LEUCOPHYLLUM FRUTESCENS OQ 30° 30° 4 106° 104° l Figure 11. Distribution of Leacophyllam revolutum, L. pe var. faevigatum, L. laevigatum var. grisenm, and L. frutescens in Texas and northern Méxic (1970) presented evidence of hybridization between L. minus and L. frutescens in a population in Big Bend National Park. Lencophyllum minus occurs in rocky limestone to igneous-rock hills, plains, and canyons in Larrea, Mixed Desert Scrub, Izotal, and Chaparral in the Chihuahuan Desert region from southeastern New Mexico through Trans-Pecos Texas into northeastern Chihuahua and Coahuila (fig. 18) from 700 to 1700 m. 4, LeEUCOPHYLLUM LAEVIGATUM Standley Erect to strongly, alternately-branched shrubs (3-)5 — 15(-20) dm tall, stems erect or not, sometimes thorny from persistence of old stems; young stems closely tomentose-canescent with irregular stellate and dendritic 144 trichomes 0.1 —0.2 mm high, radii very short, tapering, 0.02 —0.1 mm long, vestiture weathering in time, tardily glabrate; old stems with dark gray to black bark; internodes | — 8(-11) mm long. Leaves alternate, borne on erect long shoots or sometimes in compressed axillary clusters, oblan- ceolate, obovate (3-)5 — 18(-27) mm long, (1.3-)3 —8(-10) mm wide, obtuse, rounded to truncate, often bluntly apiculate to emarginate at tip, narrowly cuneate to a petiole | — 3.5 mm long as base, at margins entire or variously folded, greenish to canescent-tomentose when young, sparsely to densely vestitured equally on each surface with low stellate to dendritic trichomes 0.1—0.3 mm high, 0.1 —0.2(-0.8) mm in diameter with tapering, thin-walled radii 0.02 — 0. 15(-0.2, rarely to 0.5) mm long, but more concentrated along midveins, often gradually glabrate and scattered with interlying glands visible on each surface in mature leaves, mature leaves then green (drying dark brown) but petioles vestitured as stems. Flowers with pedicels (1-) 2—5(-9) mm long, vestitured as stems; calyces 2—5 mm long, lobes linear-lanceolate, (1.5-)2.5 — 4 mm long, to 5.2 mm long in fruit, 0.6— 1.2 mm wide at base, acute, sparsely to moderately vestitured with stellate to dendritic trichomes, often with very short radii, with underlying glands often visible, glabrous to glandular inside; corollas lavender-purple, violet, light lavender-violet, rarely blue, white, or purple with whitish patch marked with yellow-brown spots on floor of tube, (10-)14 — 20(-28) mm long, tube gradually ampliate, slightly compressed, lobes subequal, broadly obovate-orbicular, 4—6.5(-11) mm long and wide, emarginate; tube and throat with tangled hairs 0.5 — 2 mm long, lobes strongly ciliate, often pilose with shorter trichomes 0.2 —0.7 mm long at least on lower lobes inside, corolla glabrous to sparsely stellate, rarely with stipitate glands outside; stamens 4, anthers glabrous or pilose below; styles 4— 10 mm long, mostly glabrous to sparsely pilose below; ovaries glabrous or pilose at tip. Capsules oblong in outline, 4— 7 mm long, 2.5—3.7 mm wide, glabrous or pilose at tip. Leucophyllum laevigatum (fig.s 12, 13, 14) is characterized by its small shrub habit with either erect, straight stems or of a more gnarled habit with some old stems persisting as naked thorns, by its alternate, often fascicu- late, concolorous, obovate-oblanceolate, cuneate-based, typically green or grayish leaves with a vestiture of open or crowded stellate or mixed stellate-dendritic trichomes 0. | — 0.3(-0.8) mm in diameter. Flowers have lavender to purple-violet or blue corollas 10 — 24 mm long witha yellow-to brownish-spotted whitish patch on the bottom of the throat. Corolla lobes are distinctly ciliate and calyx lobes are linear, usually sparsely vescitured. Two varieties are recognized, separable by the following key 145 Figure 12. Line drawings of Leacophyllum laevigatum. a. L. laevigatum vat. laevigatum stem with characteristic scattered leaves and (in this specimen) large flowers. Note long pecicels: enlarged immature fruit and calyx (upper right) (Chiang, Wendt and Johnston 904 1). L. laevigatum var. griseum. Stem with characteristic thorny branches, more crowded, smaller leaves. Circular inserts indicate variation of vestiture in young, grayish leaves (c) and mature, more green leaves (d). Scales CITY, d circles measure + |] mm in diameter. A. Young and older leaves both green, thichomes in both typically well- spaced, not or slightly overlapping, radii short, 0.02 — 0.07(-O. 15) rarely to 0.3 mm long; leaves typically (S-) 10 — 18(-27) mm ea borne along erect stems; stems seldom persisting as thorns... 4a. L. laevigatum var. laevigatum AA. Young leaves gray, canescent-tomentose mre cen ued mature leaves either gray, with a dense mat of trichomes with radii 0.03 — 0. 16 mm long, or more green with trichome radu overlapping, 0. 1 — 0.5 mm long; leaves 3 — 10(-15) mm long, crowded into me planc thorny with old leafless stems persistng as thorns. . .4b. ee var. grisenm 4a. Lh ere LAEVIGATUM Standley var. LAEVIGATUM, Contr. U.S. Natl. Herb. 23:1305. 1924. Tyrer: MEXICO. Durango: between Ramos and Inde, l e =. 14 Aug 1898, E. W. Nelson 4089 (HOLOTY PE: US!. ISOTY PES: GH!, K!, PH!). Leucophyllum laevigatum Standley var. coabuilensis Kiger, Rhodora 74:347. 1972. Type: MEXICO. Coanuita: Caneros Pass area, along México Hwy 54, about 23 mi S of Salrillo, 5 Aug 1971, /).L. Reveal, W. H. Hess, and R. W. Kiger 2617 (HoLoty er: US!; isorypes: LL!, and elsewhere) Erect-stemmed shrubs 5 — 15(-20) dm tall, seldom with old stems persisting as thorns. Leaves alternate, sometimes also in axillary fascicles, (5-) 10 — 18(-27) mm long, (2.5-)4 — 8(-10) mm wide, green when young and at maturity, trichomes stellate to dendritic-stellate, radii 0.2 —0.8 (-0.15, rarely to 0.3) mm long, trichomes mostly well-spaced on mature leaves except along midvein; pedicels 3—9 mm long; ovary and style mostly glabrous. In LeucophyHum laevigatum var. laevigatum (fig. 12 a) both young and mature leaves are green with moderately- to well-spaced, stellate to stellate- dendritic trichomes with short, rather thickish, tapering, translucent radii 0.2 —0.8(-1.5) mm long (fig. 13 a—d). Similar but much more crowded trichomes cover young stems and this denser vestiture extends onto petioles and often along the midribs of both leaf surfaces. This vestiture pattern is found throughout the range of the taxon, except in eastern Durango and west-central Coahuila where plants tend to have some leaves with larger trichomes with radi to 0.3 mm long (fig. 13 a, b). The variety also tends to have a more erect habit with stems bearing alternate leaves. Corolla color varies and Kiger (1972) described a new variety from the Caneros Pass area in southeastern Coahuila with strong blue corollas. This variety occurs in limestone, caliche hillsides and alluvial fans in the Chihuahuan Desert from southeastern Chihuahua, western and southeast- ern Coahuila, and eastern Durango to Zacatecas, and San Luis Potosi (fig. 11) from Larrea, Mixed Desert Scrub to Izoral zones from 1200 to 2200 m. 4b. LeucopHytium LAzviGatum Standley var. griseum (1. M. Johnston) Henrickson comb. et stat. nov. aie ante M. Johnston, J. Arnold. XICO. Arbor. 22:119. 1941. Type: ME AHUILA: foothills of the Sierra Planchada, 6 mi N of Esmaralda, 16 Aug 1940, 1. M. Syne H. Mueller 341 (HOLOTYPE: GH!; ISOTYPE: LL > oa virescens 1. M. Johnston, J. Arnold Arbor. 21:253. 1940. Type: MEXICO. URANGO: near La Loma valley of the Rio Nazas, 4900 ft, 22 Aug 1939, F. Shreve 91I9T (HOLOTYPE: GH!) Low, rounded shrubs 3— 10(-18) dm tall; typically with old naked branches persisting as thorns. Leaves alternate and crowded in axillary fascicles in distal 2—5(-12) cm of branches, 3— 10(-15) mm long, 1.4—5(-7) mm wide; young leaves gray, densely covered with a close vestiture of stellate, stellace-dendritic to dendritic trichomes, older leaves gray, densely vestitured or green with a sparse vestiture, trichome radii Figure 13. Leaf vesticure of Lewcophyllum laevigatum. a-b. L. laevigatum var. laevigatum. In old (b) leaves are green, with scattered, well-spaced ou typical variety, both young (a) anc trichomes. Note sessile glands (adaxial surfaces) (F/yr 672a; near Pedricena, Durango). c—d. Laevigatum var. laevigatum. Phase with smaller well-spaced tric qually distrib- uted in young (c) and mature (d) leaves (abaxial surtaces) (Correll and Johnston 20238; 11 mi E La Zarca, N of Durango). e— f. L. /aevigata var. grisenm. In this variety young leaves are gray with dense vestiture (e), mature leaves may be gray or green (f) with sparse vestiture, (abaxial surface) (Johnston and Mueller 341; isotype of L. grisewm from N of Esmaralda, central Coahuila). Scale in a = 0.1 mm, holds for b— f. 148 (0.03-)0. 1-0. L6(-0.3, rarely to 0.5) mm long; peduncles 1.5—5 mm long; ovary and styles mostly pilose. Leucophyllum laevigatum var. griseum occurs in west-central and southern Coahuila and adjacent northern Zacatecas and San Luis Potosi (fig. 11). In several areas it is sympatric and appears to intergrade with L. /. var. laevigatum. In Flyr (1970) L. daevigatum and L. griseum were recognized as distinct species distinguished in the key by habic with L. grisevm having persistent, leafless, short, lateral branches giving the plants a thorny appearance while L. /aevigatum characteristically had longer, leafy shoots and specimens lacked such short, lateral branches unless the plants had been browsed (Flyr 1970). While such habit differences can be recognized in most specimens, they are not consistent and certainly can be influenced by environmental conditions. Vestiture differences tend to correlate with habit. In the more openly- branched variety /aevigatum both young and mature leaves are green, with an open, typically non-overlapping vestiture of trichomes typically with short radit. In the more tightly-branched variety gr/sevm with shorter, more crowded leaves, the young leaves are gray with a dense, low vestiture and mature leaves may either retain this dense gray vestiture or the vestiture may thin; che mature leaves are then green. In many southern Coahuila, Zacatecas, and San Luis Potosi populations of variety griseam both young and mature leaves are gray, covered with a dense to moderately dense vestiture of stellate or a mixture of stellate and short-dendritic trichomes with short radit 0.02 —0.08(-0.15) mm long, basically with trichomes with short radii similar to those of variety /aevigatum but very crowded. In specimens of variety grisewm from west-central Coahuila (including the type of L. grisewm), young leaves tend to be gray, densely vestitured with stellate and dendritic trichomes, however, as the leaves mature, dendritic trichomes with short radii tend to fall away and the remaining stellate, stellate-dendritic trichomes are more dispersed, though typically with overlapping radu (fig. 13 e, f). They also tend to have longer radii, 0.1 —O0.2 mm long, similar to the longer-trichome radii found on leaves of variety /aevigatum in that region. In some specimens trichomes on young leaves have long radii at the base and short radii in discal portions and as leaves mature the distal portions of the trichomes fall away leaving the larger basal radii. A similar pattern is seen in some specimens of variety griseum in northern Zacatecas. This trend reaches its extreme in southwestern Coahuila and adjacent Durango, where in some specimens, young leaves are gray with a dense, close veestiture of trichomes with short radii (fig. 14 a) and older leaves are greenish, with moderately dense, typically overlapping, large trichomes 149 Figure 14. Leaf vestiture of Leacophyllum laevigatum var. griseum and L. minus. a—b. L oc, var. eae Vestiture similar to that of type of L. vrrescens: OUnes leaf has dense, vestiture (a); mature leaves have very sparse vestiture with radii to 0.4 mm long (b), (a ul Se (Chiang, Wendt iden i‘ 50; near Sierra Jimulco, Coahuila). c— d Ss oe var. hire In S Coahuila, Zacatecas, vestiture of both young (c) and mature (d) se and both eis and mature leaves are gray (adaxial surfaces) (Sanford, ee and Nees 15 of Parras, Coahuila). minus. Both young and old leaves have dense vestiture of a ‘Hate’, actually short “dendritic trichomes. ec. Mature leaf, abaxial surface (McGill and Keil 7626). £. Enlarged view showing broad, flat radii. Base at radii at trichome tip often darken giving appearance of a terminal gland (Engard and Gentry 605). Scale ina = 0.1 mm, holds fora-e, f = O.1 1 mm. 150 with radii 0.13 —0.5 mm long (fig. 14 b). In some specimens trichome radii vary among adjacent leaves on a stem; some leaves have trichome radii 0.143 —0.2 mm long, others have radii 0.2 — 0.5 mm long, indicating that trichome radii development may be influenced by environmental con- ditions. A specimen with this type of long-rayed trichomes similar to that shown in fig. 14 b was designated the type of L. virescens by 1. M. Johnston (1940), who emphasized the short trichome radit on young leaves in contrast to the long radu on trichomes of old leaves. However, the specimen appears to be completely referable to variety grisewm and differs only in its very long trichome radit. Additional specimens bridge the gap between this specimen and typical grisewm. Flyr (1970) considered L. virescens to be a hybrid beeween L. /aevigatum and L. candidum because specimens referable to these two taxa were the only plants found during two searches at the type locality of L. virescens. The occurrence of trichomes with long-basal radi and short distal radii and others with only short radii is reminiscent of those of L. candidum but the specimen exhibits no other features of L. candidum. Whether L. /. var. griseam merely represents a more xeromorphic derivative of L. 1. var. daevigatum, i.e., with a reduced, more thorny habit, more strongly vestitured leaves, or presents intergradation with L. candi- dum, or other more densely vestitured species, is not known. Specimens available indicate a continuum of variation between the two varieties. The line between the two taxa must be drawn arbitrarily. Distinction on the basis of habit (sensu Flyr, 1970) versus vestiture (emphasized here) gives a slightly different assignment of specimens that have long, erect stems (as in variety /aevigatum) but dense vestiture on young leaves (as in variety griseum). The younger epithet grisewm Johnston 194 1) is recognized at the varietal level over the older virescens (1940) because the type of grisevm is more representative of the taxon and grisewm has been more widely used than VIFESCENS. Leucophyllum 1. var. griseum grows on limestone and calcareous hillsides from Larrea- to Yucca-dominated zones to chaparral from 1400 to 2400 m (fig. 3. LeucopHyLLUM CANDIDUM I. M. ean J. Arnold Arbor. 22:120. 1941 {15 Jan 1941]. Type: MEXICO. Coanuta: between Carrizo and Carricito on (gypsaceous’) ridge, 11 Aug 1940, 7. M. en aie H. Mueller 160 (HOLOTYPE: GH!; tsorypee: LL!). aa eh violaceum Pennell, Proc. Acad. Nat. Sci. Philadelphia 92:295. 1940 [8 Apr 41} XAS. BrewsTeR Co.: frequent on flats from Lone Mountain to ies Nugent Mt., Chisos Mt. area, 2 Aug 1937, B. H. Warnock 1124 (HOLOTYPE: US!; IsoTYPES: GH!, PH!, SRSC, TAES!, TEX!). Compact, divaricately, alternately-branched, rounded to erect shrubs 3 — 10(-15) dm tall; young stems densely, unevenly tomentose with spread- ing dendritic trichomes, the longer 0.2 — 0.7 mm long, 0.15 — 0.3 mm tn diameter, these eventually weathering to a more uniform stature and eventually glabrate; older stems red-brown to light or dark gray. Leaves alternate to subopposite, often opposite near tip of stem, often with reduced NS IN <1 ae Neal A SS= ts Nine 1s i pe 3 ws: Bb Nes Sie EQVINITS SON: Seer) ek FZ. eee = ie} ~_ Pa eon oO Og oe ie} wt rc a a 20 aos m2 a A ° ore o oes orn = N O == siete) Hx yu Od im i a & 090 SERS) 2 2 yw 1°) fe} co o = fo} ro) fo ro < Fe Ee: “a Si] l l l Figure 17. Distribution of Le eae zygophyllum and L. candidum in southern trans- Pecos Texas and northern Méx on rocky hillsides from Larrea scrub to chaparral or submontane scrub in southwestern Tamaulipas near Miquihuana and Bustamente and adjacent northern San Luis Potosi east of El Huizache Junction (fig. 11) over an elevational range of 1600 to 2200 m. 158 8. LeucopHytL_uM etyrut B. L. Turner, Sida 5:54. 1972. Typr: MEXICO. SAN Luts Porost: 4.4 mi NE of Laguna Seca (Gral. Candido Navarro), 29 Jul 1966, D. Flyr 1113 (HoLoryee: TEX!). Strongly, alternately-branched, rounded shrubs 5— 16 dm tall, often rather thorny due to persistence of young dead branches; young stems densely, unevenly, silvery-gray tomentose with dendritic trichomes 0.2 —0.4 mm long, 0.2 — 0.3 in diameter, radii wavy, slender, 0.1 —0.2 mm long, trichomes weathering, vestiture more uniform in time, tardily glabrate, older stems with dark gray bark; internodes 2—8 mm long. Leaves alternate, oblanceolate to obovate-spathulate, broadest in distal one-fourth, 10—22(-27) mm long, 4—9(-11) mm wide, obtuse, sub- rounded, bluntly apiculate at tip, tapering to a cuneate base, true petiole not discernible, at margins entire, densely unevenly gray (slightly tan in herbarium specimens) tomentose on both faces with dendritic trichomes 0.1—0.4 mm long, radii wavy, slender, 0.1 —0.2 mm long, trichomes diminishing in stature through weathering; midrib raised beneath. Flowers with tomentose pedicels 1— 3 mm long; calyces 5 — 6.5 mm long, lobes lanceolate, 4.5—5.5 mm long, 1.1—1.5 mm wide, acute-attenuate, sparsely beset with dendritic hairs and slender, spreading, stipitate- glandular trichomes 0.1—0.3 mm long, glabrous inside except for stipi- tate glands; corollas purple to light violet with small dark purple spots throughout lower tube, yellow only at very base, (16-)19 — 21(-25) mm long, tube broadly ampliate, slightly compressed, to 8 mm wide at throat (pressed), lobes obovate, subequal, 5 — 7 mm long and wide, emarginate, tube sparsely pilose on floor inside with crinkled hairs | — 2 mm long, lobes nearly glabrous inside, not ciliate, corolla sparsely stipitate glandular outside; stamens 4, anthers glabrous, posterior filaments 8— 10.5 mm long, anterior flaments 6 — 9 mm long, glabrous; styles 10 — 12 mm long, sparsely pilose near base; ovaries sparsely pilose, stipitate-glandular at tip. Capsules dark brown, 5-6 mm long, 3-4 mm wide, pilose near tip. Leucophyllum flyrii is distinguished by its somewhat thorny habit, oblan- ceolate to obovate, cuneate-based leaves that are equally tomentose on both sides with dendritic trichomes with moderately long wavy radii. In addi- tion the calyces, unlike any other species in the genus, have an overstory of slender stipitate glands on the outer surface and are less strongly vestitured than the pedicels. The corollas have broadly ampliate tubes, with dark purple dots to 1 mm wide on the floor of the tube. It is known from a few localities northeast of Ciudad San Luis Potosi in the southernmost margin of the Chithuahuan Desert and in the adjacent Sierra San Pedro (fig. 18) from 1800 to 2200 m. I 9. LeucopHyLtuMm pRuINOsUM I. M. Johnston, J. Arnold Arbor. 22:119. 1941. Type: MEXICO. San Luts Porosi: 11 mi S$ of Matehuala, 10-11 Sep 1938, 1. M. Johnston 7569 (HOLOTYPE: GH!). Strongly, alternately-branched shrubs sometimes somewhat thorny due to persistence of old stems, (3-)8 — 15(-26) dm tall; young stems loosely tomentose with dendritic trichomes 0.3 —0.8 mm long, with long, slen- er, wavy radii 0.1 —0.5 mm long, vestiture tardily glabrescent; old stems with gray to brownish bark. Leaves alternate, orbicular to broadly ovate, rarely broadly elliptical, 8— 16(-27) mm long, 6— 13(-17) mm wide, rounded, obtuse, often obscurely apiculate at tip, rounded to abruptly cuneate above petiole 1 —3.5(-G) mm long at base, at margin entire to crisped-undulate or variously folded, densely but loosely gray tomentose on both faces (sometimes slightly tan in herbarium specimens) with elongate dendritic trichomes 0.4—0.8 mm long with slender, wavy radii 0.1 —0.3(-0.5) mm long, trichomes often more dense along raised midvein beneath. Flowers on tomentose pedicels 1.5 — 3 mm long; calyces 4— 6.5 mm long, lobes oblong-lanceolate, 3-5 mm long, |— 1.7 mm wide, acute to attenuate, often unequal, densely tomentose outside and near tip inside as young stems, sparsely to moderately pilose, glandular inside; corollas dark purple to violet with a reddish tinge, with a large white patch with gold dots in floor of tube inside, with grape odor, 8.5 — 11(-14) mm long, tube broadly campanulate, abruptly expanded above base, (5-)7 —9 mm broad at throat (pressed), lobes orbicular, obovate, subequal, 3—5 (-6.5) mm long and wide, sometimes crisped, often emarginate, tube long pilose only in the cube base and lower throat, lobes mostly glabrous, ciliate with shorter hairs, corolla stipitate-glandular outside; stamens 4, anthers pilose at tip, posterior filaments 5 — 9 mm long, anterior filaments 3 — 4S mm long; styles 5 — 9(-11) mm long, sparsely pilose; ovaries densely pilose and stipitate-glandular at tip. Capsules 3.5 —5 mm long, 2.3—3 mm wide, pilose at tip. Leucophyllum pruinosum (fig. 19 a) can be distinguished by its distinctive vestiture of large dendritic trichomes with slender radii 0. | — 0.5 mm long; the consistently longest trichome radii found in the genus (fig. 20 b), the orbicular to suborbicular, petiolate leaves usually with finely crisped to undulate-folded margins, and the relatively short corollas with the dis- tinctive campanulate tube that abruptly increases in diameter above the base. It is most similar to L. w/tramonticola as discussed under that species. It differs from L. ambiguum in the conspicuous yellow-spotted white patch on the corolla-tube floor. Niezgoda and Tomb (1975) noted this was the only species of Lewcophyllum with rugulate rather than reticulate pollen sculptur- 160 ing. Tomb (pers. comm. 1984) noted differences also occur in seed sculpturing. The species can occur sympatrically with L. revolutum and L. zygophyllum but no hybrids have been observed. Leacophyllum pruinosum occurs on rocky limestone slopes and alluvial fans in Larrea to Mixed Desert Scrub from southern Nuevo Leon, southwestern Tamaulipas, and eastern San Luis Potosi (fig. 18) from 1000 to 1600 m. 10. LeucorHytitum ultramonticola Flyr sp. nov. AL. prummosum ovarus et stylis glaberis non pilosis, corollas lobis inferis pilosis non glabris, foliis orbicularioribus, distributione in Zacatecas meridio-occidental differt. Alternately- rather openly-branched, erect shrubs 10—17 dm tall; young stems irregularly, densely gray-tomentose (turning brown in herbar- tum specimens) with cylindric, dendritic trichomes (0. 1-)0.3 —0.7 mm long with slender, wavy radii 0.1 —0.2 mm long, tardily glabrate, older stems light gray, often remaining as coarse thorns; internodes | — 10 mm long. Leaves alternate, crowded, broadly ovate-orbicular to broadly ellip- tical, 1O— 25 mm long, 8 — 25(-30) mm wide, obtuse to rounded, occa- sionally retuse or bluntly apiculate at tip, abruptly cuneate-rounded at base above a tapering petiole 2—5S mm long, at margins entire, undulate, densely gray to rather greenish (turning brownish in herbarium specimens), tomentose on both faces with dendritic trichomes (0. 1-)0.2 —0.4(-0.7) mm long, radi slender, straight or wavy (0.1-)0.2—0.4 mm long, trichomes weathering, becoming more uniform in age, midvein and in larger leaves basal, lateral veins prominent beneath. Flowers with tomen- tose pedicels 3— 3.5 mm long; calyces 5.5 —8(-9.2) mm long, lobes oblong to oblanceolate, 4.2 —7 mm long, |1.4— 2.1 mm wide, obtuse- acute, densely tomentose as stems outside and inside at tip, moderately sericeous-pilose inside, glabrous, sparsely glandular near base inside; corol- las purple-violet, with a large yellow patch marked with red-brown spots on tube floor inside, 12-16 mm long, tube broadly campanulate, to 6.5 —8.5 mm broad at throat (pressed), lobes obovate, 4 — 5 mm long and 3.5 —4.5 mm wide, truncate to emarginate at tip, undulate, short ciliate, tube with long tangled hairs 1.5 — 2 mm long on basal floor, lobes pilose with shorter, wavy hairs 0.3 —0.5 mm long, to 1.0 mm long near throat; corolla glabrous to very sparsely stipitate-glandular outside with hairs 0.1—0.2 mm long; stamens 4, posterior filaments 8.5 — 10 mm long, anterior filaments 5—6.5 mm long, glabrous; styles 9— 10 mm long, glabrous; ovaries glabrous. Capsules 4.5 mm long, 2.5 mm wide, glabrous except for some stipitate glands near base. Type: MEXICO. Zacatecas: San Juan Capistrano, 21 Aug 1897, J. N. Rose 2452 (HOLOTYPE: US!). 161 Additional collections: MEXICO. Zacatecas: 9 mi W of Huajaquilla El Alto, 8.8 mi E of Rio Atengo on sandy tuffs, 17 July 1984, Michener, Prigge and Meyer 4405 (A, MEXU, TEX) Leucophyllum ultramonticola is clearly related to L. pruznosum with which tt shares the Spun broadly campanulate corolla tube, leaf Shae vesti- 102° oe 98° cya LEUCOPHYLLUM FLYRI & LE HYLLUM M LEUCOPHYLLUM GMANIAE @ LEUCOPHYLLUM AMBIGUUM @ LEUCOPHYLLUM PRUINOSUM A LEUCOPHYLLUM ULTRAMONTICOLA fil 30° if 28° 26°=4 ( ~ eae CN 24° 4 106° 104° 102° jroo? 98° i at i ure 18. Distribution of Leacophyllam flyrii, L. minus, L. langmaniae, L. ambiguum, L. ee and L. Cas in Trans-Pecos Texas, adjacent New Mexico to Querétaro and Hidalgo in central Meéxi 162 ture, and texture, and openly branched habit. The new species differs from L. pruinosum in its nearly glabrous (not densely pilose) ovary and style base, in its more strongly vestitured (not glabrous) inner lower corolla lobe surfaces, the orange-brown rather than yellow spots on the corolla floor, its slightly larger flowers, generally larger stature with larger, more often orbicular leaves, and its distribution on the Pacific drainage in extreme southwestern Zacatecas (fig. 16). All evidence indicates that it is a rather recent vicariant of L. prvinosum of southern Nuevo Leén, adjacent Tamauli- pas, and San Luis Potosi (fig. 18). The new species also shares many characters with L. ambiguum of Hidalgo and Querétaro, but the latter has narrower corolla tube-throat, typically lacks yellow-brown spots on the corolla tube floor, and has densely pilose ovary and style bases. Rose's type specimen collected in 1897, was for a long time the only known collection of L. a/tramonticola. It has been recently recollected by Michener, Prigge, and Meyer near the type locality where it is locally common on xeric, well-drained, sandy, whitish and reddish volcanic tufts in association with Fouquieria splendens, Agave, Opuntia, Jatropha, H yptis, Acacta, Prosopis and other leguminaceous shrubs and trees along the road between Huejaquilla El Alto and the Rio Atengo from 1000 to 1500 m elevation just east of San Juan Capistrano in an area that apparently has been very poorly collected. IL. LeucopHyitumM ampBiGuum Bonpl. in Humb. & Bonpl. Pl. Ae- quinoct. 2:95, pl. 109. 1812. Type: MEXICO. Hipatco. prope Actopan, 1050 hex., A. one Bonpland s.n HOLOTYPE: {Microfiche of Humboldt and Bonpland Herbariu teiaii Tio campanulatum Miers. Ann. _ Nat. Hist. 5:254. 1850. Type: Coulter 1271 (HOLOTYPE: BM!; isorypes: GH!, K!, NY!, PH!). Leucophyllum altamiranti oe eos inst: MSc -Nac. México. 8:275. 1906. Type: “O. QUERETARO: D a Mesa, 20 Aug 1905, Altamirano 1557 (HOLOTYPE: unknown). Strongly, alternately-branched, rounded shrubs 0.6 — 1.5(-20) dm tall, sometimes somewhat thorny due to persistence of old stems; young stems irregularly, densely gray-tomentose (or brown in herbarium specimens) with cylindric, dendritic trichomes 0.2 —0.6 mm long, (0.1-)0.2—0.4 mm in diameter with slender, wavy radii (0.06-)0.1—0.25 mm long, tardily glabrate; old stems light gray to reddish-brown; internodes 1 — 10(- 14) mm long. Leaves alternate, rarely some opposite, sometimes crowded on compressed lateral shoots, broadly ovate to orbicular, rarely broadly elliptical, 8 — 22(-40) mm long, 5 — 15(-27) mm wide, often variable on same plant, obtuse-rounded, bluntly apiculate at tip, abruptly cuneate to rounded at base above a broad, 1.5 — 7 mm long petiole, entire to undulate 163 <2 —< ZZ FT os —— ain = Li AG fon ip i Figure 19. Line drawings of Lewcophyllum pruinosum and L. pringlet. a. L. prurnosum. Stem with leaves and flowers. Corolla in side view (lower right) shows characteristic expanded tube (F/yr 564e). b—e. L. pringler. b. Stem with crowded, overlapping leaves and flowers. c. Enlargement of stem and leaves. Note expanded leaf bases and branched trichomes at lower right. d. Transparent view of corolla as seen from above, note relative position of stamens. e. Immature fruit enclosed in persistent calyx and copped with style (R. Craz C, 2089). Scales at margins, densely, irregularly gray (or brownish in herbarium specimens) tomentose on both faces with dendritic trichomes ().2 — 0.6 mm long, radit slender, straight or wavy (0. 1-)0.2— 0.4 mm long, trichomes weathering, becoming more uniform in age, midvein prominent beneath. Flowers with tomentose pedicels 1 — 2.5(-4) mm long; calyces 4.5 — 6 mm long, lobes oblong-lanceolate, oblong-ovate, 3—5 mm long, 1 — 1.5(-1) mm wide, acute, densely tomentose as young stems outside and at inside tip, mod- erately sericeous-pilose, glandular inside; corollas violet to purple through- out, white only at very base of tube, or with purple spots on floor of tube, rarely white with yellow spots at floor of tube, 12— 18 mm long, tube 164 dorsi-ventrally compressed, three times wider than high, cylindrical to ampliate, lobes obovate to orbicular-reniform, 3 — 4.5(-6) mm long and wide, undulate, slightly emarginate, ciliate, tube densely pilose through- out inside with long, tangled trichomes to 2 mm long, lobes densely pilose inside with straight trichomes 0.3 — 0.8 mm long, corolla sparsely stipitate glandular outside; stamens 4(-5), posterior filaments 6—8 mm long, anterior filaments 4—5 mm long, glabrous to pilose; styles 7— 13 mm long, pilose; ovaries densely pilose. Capsules 3 — 4 mm long, densely pilose above, glandular below. Lencophyllum ambiguum is characterized by its concolorous, broadly ovate, orbicular, petiolate leaves, by its distinctive dendritic trichomes (fig. 20 a) that usually have long, slender, wavy radii 0.2 — 0.4 mm long on stems and leaves, by its distinctive violet to purple corollas that typically lack a whitish patch on the tube floor, and the dense tangled and straight trichomes on the inside surfaces of the corolla tube, throat, and lobes. It also is a disjunct species occuring in the southern extension of the Chihuahuan Desert in arid portions of Hidalgo and Querétaro (fig. 18). Throughout its range it exhibits some notable variation. Occasional plants have shorter vestiture with trichome radii only 0. 1 mm long on both stems and leaves. A population northwest of Metzquititlan, Hidalgo, has corollas with a white patch on the floor of the corolla tube beset with yellow spots (Rzedowski 32513, ENCB, MEXU; F. Gonzalez M. 8486, MEXU) and in this character approaches L. w/tramonticola and L. pruinosum. Occa- sional plants have some opposite leaves. Leucophyllum ambiguum occurs on limestone and calcareous hillsides with Larrea and other desert shrubs up to submontane scrub from 1200 to 2500 m 12. LeucopHyLLUM PRINGLEI (Greenman) Standley, Contr. U.S. Natl. Herb. 23:305. 1923. Faxonanthus pringler Greenman in Sargent, Trees & a 1:23, pl. 12. 1902. Type: MEXICO. Puesta: limestone hills near Tehuacan, “6000 fe. Aug 1901, C. G. Pringle 8594 (HOLOTYPE: GH! isoryprs: A!, ENCB!, ae sheets)!, NY!, PH!, US!). Erect shrubs (1.5-)3 —6 dm tall with several erect stems from a thick woody base, irregularly branched above, old branches persistent: young stems terete, 1— 1.5 mm in diameter, hispidulous with erect, multi- cellular, tapering or distally forked or branched, white trichomes 0.1 —0.25(-0.4) mm long, and with sessile oo older stems marked with raised, persistent leaf bases; largest stems 4 — 6 mm in diameter: bark gray, vertically fissured; internodes 0.5 — 1(-2.5) mm long. Leaves alter- nate, crowded in terminal 2 — 3 cm of branches, linear-lanceolate to linear- VAN a Zee” _ Leaf and stem vestiture of Lecophyllum ambiguum, L. pruinosum, L. pringler and : ion showing equal vestiture on both Figure 20 Eremogeton grandiflorus. a. L. ambiguum. Leat cross section basal simple portion of trichomes. Note also long radii (Flyr 5876). b. L. pruinosum. Surface view of adaxial leaf vestiture showing long, slender radit (Henrickson and Hess 19117). c. L. pringle’. Vesciture of upper stem with weakly branched trichomes. No Cruz C. 2098). d. Ereomegeton grandiflorus. ger unb surfaces an large glands. (Compare with fg. 3 d—f). (RK Cross section of leaf with adaxial surface on right, abaxial surface with longer t ranched, curved trichomes on left. Note collapsed trichome cells (Stevens, Donoghue. and Scott 2344). Scale in a = 0.1 mm and holds for b~d. 166 elliptical, 6 — 12 mm long, 1.2 — 2.2 mm wide, ca 0.4 mm thick, acute to obtuse at tip, tapering in lower half toa broad petiole-like base about | mm long, 0.5 —0.7 mm wide but expanded at very base where jointed with stem, entire, typically glabrous except for a few trichomes as in stem on basal | —4 mm or with trichomes scattered throughout, viscid, strongly glandular co glandular-punctate, midrib obscure. Flowers with ascending pedicels 3 —6 mm long with scattered erect to distally forked trichomes, sometimes with short, stipitate glands above; calyces 6.5 —8 mm long, tube ampliate, | — 2.2 mm long, lobes 5, linear-oblanceolate, subequal, 5.5 —7(-10) mm long, 0.5 — 1.5(-2) mm wide, green, leafy, acute at tip, entire, viscid, with sessile (to stipitate) glands on both sides; corollas light purple to blue, with a white patch marked with golded dots on floor of throat, (20-)22 — 25 mm long, tube 1-2 mm long, throat broadly cam- panulate to 12— 15 mm wide (pressed), slightly ventricose at floor, lobes orbicular to obovate, medial anterior lobe largest 9 — 12(-14) mm long and wide, emarginate, 2 lateral anterior lobes slightly smaller, 8 — 10 mm wide and long, rounded at tip, posterior lobes 8 — 9 mm long and wide, rounded at tip, tube with slender, unicellular trichomes | — 3 mm long on floor of distal tube and near throat inside, lobes glabrous inside; corolla glabrous outside; stamens 4, included, posterior filaments 9 — 10 mm long, anterior filaments 6 —7 mm long; anthers whitish, glabrous except for tuft of hairs near tip; proximal anthers 3 — 3.5 mm long, anterior anthers 2— 2.5 mm long; ovaries glandular near tip; styles 10 — 14 mm long, sparsely glandular to stipitate-gandular, expanded, rhomboid, acute at tip. Capsules 5 — 6.5 mm long, 2.5—3 mm wide; seeds ellipsoidal, 0.5 —0.8 mm long, 0.4—0.5 mm wide, angular, muricate in vertical lines. Leucophyllum pringler (figs. 19 b, 20 c) is one of the more distinctive species in the genus and is distinguished by its multicellular, uniseriate, tapering or distally-forked, occasionally distally-branched trichomes with multicellular rays, the tack-shaped, subsessile glands with 10 — 15 cells in the heads (figs. 3 d-f, 20 ¢), its crowded, linear-oblanceolate, -elliptical, viscid, glanduliferous, leaves and its habit consiting of a series erect stems developing from a woody knot several centimeters in diameter. Corollas are also broadly ampliate and ventricose along the floor. The stigmas are longer, more rhomboid, acute, and usually recurve at the tip when dried. In Flyr (1970), the taxon constituted the monotypic genus Faxonanthus, However, the species fits well within Lewcophy/lum in flower and fruit characters and differs only in its more open corolla throat (a feature also found in the smaller-flowered L. pruinosum), larger-headed glands, trichomes with multicellular rays, and more rhomboid, acute style tips. Its branched or forked trichomes (figs. 3 d, e, 20 c) appear derived from a 167 broader, dendritic trichome type but with smaller multicellular radit. Equally sparse trichomes also occur in L. /aevigatum var. laevigatum. The conspicuous glands are not unique; leaves of all species of Leacophyllum have glands with multicellular heads, those of L. pringler are just larger with a greater number of cells. In other species glands are typically obscured by the crowded, nonglandlar trichomes. Phenetically the taxon is quite distinct and it is here recognized as a separate subgenus. Recognition at the generic level is, in the senior author's opinion, not consistent with other generic distinctions in Scrophulartaceae. The two subgenera share a large number of basic charactersitics, particular- ly those associated with flowers and fruits. Differences lie mostly with vestiture and vegetative morphology. The species is also well-isolated geographically from other speices in the genus south of Mexican trans- volcanic axis. This isolation has undoubtedly contributed to its morpho- logical distinction. Leacophyllum pringlei is restricted to limestone slopes in south-central Puebla and adjacent Oaxaca from 1600 to 2350 m (fig. 22) and flowers from May through October depending on rainfall. Il. EREMOGETON Standley & L. O. Williams, Ceiba 3:172. 1953. Ghiesbreghtia A. Gray, Proc. Amer. Acad. Arts. 8:630. 1873 (non A. Richard & Galeotti 1845). Monotypic. 1. EREMOGETON GRANDIELORUS (A. Gray) Standley & L. O. Williams, Ceiba 4172: pier Ghiesbreghtia grandiflora A. Gray, Proc. Amer. Acad. Arts. 8:630. 1873. Type: MEXICO. Criapas: 1864-70, Ghiesbreght 723 (HOLOTYPE: GH!). Large suffrucescent shrubs to small trees 3 — 6(-8) m tall; young stems subterete, uniformly villous-sericeous to tomentose with mostly soft, an- trorsely curved to wavy, multicellular trichomes 0.2 — 0.7(-1.0) mm long, older stems (3-)4 — 6 mm in diameter, prominently marked with raised leaf and pedicel bases; bark furrowed; internodes (1-)6 — 15 mm long. Leaves alternate to subopposite, oblong-obovate, oblong-ovate, elliptical, ovate, (4-)5 — 9-12) cm long, (1.5-)2 — 5(-6) cm wide, obtuse to acute, apiculate at tip, broadly cuneate at base with margins extending down a8 — 15 mm long, 2—4 mm wide, winged petiole, at margins coarsely dentate, serrate to crenate in distal part, teeth obtuse to acute, apiculate, revolute, entire and revolute in lower portion, rarely entire throughout, close vestitured on both surfaces, upper (adaxial) surface with more scattered, soft to slightly rigid (then scabrous) tapering, ancrorsely curved, multicellular trichomes (.2—0.5 mm long, often mixed with stipitate glands, lower (abaxial) surface more densely vestitured with similar, tapering, antrorsely curved, 168 sometimes longer, more wavy trichomes 0.2 —0.6(-1.0) mm long, vesti- ture more tomentose, longer along prominently raised mid and secondary veins. Flowers solitary in leaf axils, pedicels ascending, 2— 3.5 cm long at anthesis, to 4—6 cm long in fruit, vestitured as young stems; calyces 5-lobed, lobes green, leafy, oblong, oblong-oblanceolate, ascending, 2—3.6 cm long, 4—7 mm wide, separate to within 2—3 mm of base, obtuse to acute at tip, entire, vestitured as lower leaf surface outside, with scattered stipitate glands inside; corollas showy, white, turning cream- white or yellowish, opening in late afternoon, zygomorphic, 6—7.2 cm long, tube abruptly expanded above base, cylindrical (25-)35 —40 mm long, 15 —20 mm in diameter, (25 — 30 mm wide pressed), posterior 2 lobes united, 25 — 30 mm long, together 22 — 28 mm wide, terminal teeth L[1— 15 mm long, obtuse to acute, erect, anterior 3 lobes spreading to descending, oblong-ovate, 25 — 33 mm long, 13 — 15 mm wide, acute to obtuse, corolla chick, glabrous to stipitate-glandular inside, stipitate- glandular outside where exposed in bud, ciliate with longer, crinkled hairs; fertile stamens 2, filaments 55 — 60 mm long, | — 1.5 mm thick, whitish, glabrous, adnate to expanded corolla tube base; anthers situated near tip of posterior corolla lobes, anther lobes 2, oblong-linear, slightly divaricate, cream-white, glabrous, longitudinally dehiscent across continuous apex, glabrous; sterile stamens 2 or absent, with filaments 3.3 —7 mm long; ovary superior glabrous, grooved along septum, 2-loculed, placentae attached medially along septa; ovules many; style accrescent, 6 — 7.2 mm long, slightly expanded at obtuse tip, stigmatic along broad terminal band, glabrous to sparsely pilose. Fruit of dark brown ovoid, grooved, apiculate capsules, these dehiscing septicidally to near base, then loculicidally half way to base, subtended by persistent calyx; seeds 50 — 100 per locule, ellipsoid, 1.7 — 2.2 mm long, 0.6 —0.8 mm wide, dark brown, angulate to flattened by compression, muricate in vertical rows; endosperm oily; embryo small. Evemogeton (tig. 21) 1s a very distinct monotypic genus characterized by its uniseriate, multicellular, tapering, unbranched, antrorse trichomes (fig. 20 d), its large, ovate leaves, and its very large flowers with regular, deeply- parted calyces and conspicuous, zygomorphic, thickish corollas, 6 — 7.2 cm long, with 3 anterior, reflexed to spreading lobes and a 2-toothed, erect, posterior lobe. According to label-data of W. D. Stevens et al. 2344, (TEX), the corollas are “white at anthesis, just opening in late afternoon, later cream white.’ Stamens are typically 2, anterior, however, occasional speci- mens (including the type) may have two additional abortive, antherless staminodes with short filaments 3 — 7 mm long (Flyr 1970). The genus also has a conspicuous, probably nectariferous gland on the lower portion of the 169 Figure 21. Line drawing of Eremogeton grandiflorus. Stem with leaves and large flowers; developing fruit with long pedicel at right. (Stevens, Donoghue, and Scott 2344). Scale = | cm. ovary. Capsules are also much larger than in Leucophyllum and contain many more seeds. Eremogeton grandiflorus occurs most slopes in oak-pine forests of south-central Chiapas, Mexico, and Guatemala from 1200 to 2200 m (fig. 22). — y along limestone bluffs and steep ACKNOWLEDGEMENTS In January 1970 Lowell David Flyr compl sleted his dissertation in the Department of Botany at The Univeristy of Texas at Austin entitled, “A systematic study of the tribe Leucophylleac (Scrophulariaceae).’’ He recog- nized 13 species plus two varieties of Lewcophyllum and the monotypic genera Faxonanthus and Eremogeton. Unfortunately for the botanical community, Flyr ended his life on 2 November 197 1 without publishing his dissertation 170 (| er 87° = ° co ca } A EREMOGETON GRANDIFLORUS @ LEUCOPHYLLUM PRINGLEI fe] wo a { Figure 22. Distribution of Leacophyllum pringlei and Eremogeton grandiflorus in southern México and Guatemala. 171] (see obituary by Turner, 1972). In the spring of 1982 Henrickson began a study of the group, using Flyr’s dissertation as a starting point, working with the extensive collections at TEX-LL and borrowing or visiting the collections of ASU, ENCB, GH, MEXU, NY, RSA-POM, and US reev- aluated the taxonomy and reduced the number of taxa recognized two genera, Lencophyllum (including Faxonanthus), with 12 species and 2 vari- eties, and the monotypic Eremogeton. In this treatment the basic systematics of Flyr are followed but descriptions and discussions are oo ae Specimens were visited or borrowed by Flyr from A, BM, Woy Fy GH, K, LA, LL, MICH, MO, MSC, NMC, NY, OXF, PH, ic TAES, UC, and US, thanks are extended to curators of these herbaria for loans and courtesies extended. Dr. B. L. Turner, J. Strother, A. M. Powell, R. Irving, J. Rzedowski, P. Echlin, M. C. Johnston, L. H. Shinners, and others were acknowledged by Flyer (1970) for their help with his disserta- tion. John Strother, A. S. Tomb, L. Dorr, B. Prigge, D. Michener, : Mauseth, and K. Nixon as well as B. L. Turner and M. C. Johnston have, various ways, greatly aided in the preparation of this paper. Robert I. Chinnock of the State Herbarium in Adelaide, South Australia, who ts kindly provided data on charac- ae monographing Eremophila (Myoporaceae teristics of the Myoporaceac. REFERENCES ARGUE, CG. L. 1980. Pollen morphology in the genus Mimalus (Scrophulariaceae) and its taxonomic ae Amer. J. Bot. 67:68 — 87. LROD, D. I. oS Age and origin of Sonoran Desert Vegetation. Occas. pap. Calif. Acad. Say 132: = TOs RLOW, B. A. ok Cytogeography of the genus Eremophila. Austral. J. Boc. BENTHAM, G. 1846. pene tay in A. P. De Candolle, Prodomus Systematis Naturalis Beep Vegetabilis 86-598. bs eco ec ae in G. Bentham and J. D. Hooker, Genera Plantarum 11(2):913 - 980 CARLOQUIST, S. 1962. A theory of paedomorphosis in dicotyledonous woods. Phytomorphology 11:30 —45. _, - 1975 Ecological strategies of xylem evolution. Univ. Calif. Press, Berkeley. CRONQUIST, A 1981. An integrated system of classification of dewerne plants. Colum- k DE CSERNA, Z. 1960. Orogenesis in time and space in Mexico. Geol. Rundschau. DILCHER, D. L. 1974. Approaches to the identification of angiosperm leaf remains. Bot. Rev. (Lancaster) 40:1 — 157 FLYR, L. D. 1970. A systematic oa of the tribe Leucophylleae (Scrophulariaceae). Ph.D. Dissertation, University of Texas, Austin GUZMAN, E. and Z. DE CSERNA. 1963. ‘Teeionie history of Mexico. Amer. Assoc. Pet. Geol. Mem. 2:113— 129. 172 HAIR, J. B. and E. J. BEUZENBERG. 1959. Contribution to a chromosome atlas of the New Zealand Flora. New Zealand J. Sci. (Wellington) 2:148 — 156 HENRICKSON, J. 1972. A taxonomic revision of the Fouquieriaceae. Aliso 7:439 — 537 JOHNSTON, I. M. 1941. New phanerogams from Mexico, J. Arnold Arbor. 22:110— 124. KARRFLAT, E. E. and A. S. TOMB. 1983. Air spaces, secretory cavities, and the relationship between Leucophylleae (Scrophulariaceae) and Myoporaceae. Syst. Bot. a 4? 8:2 Zi KIGER, R. W. 1972. A new of Leacophyllum laevigatum (Scrophulariaceae) from Mexico. Aa 74:347 — 349. LERSTEN, N. R. and K. - CARVEY. 1974. Leaf anatomy of Ocotillo (Fouquieria splendens; Fouerec especially vein endings and associated veinlet elements. Canad. , Ot. 222017 = 2021, MIC HENER, D. C. “1981. and leaf anatomy of (Scrophulariaceae): Ecological considerations. Aliso 10:39 — NIEZGODA, C. J. and A. S. es 1975. Systematic palynology of a Leucophylleae Cac and selected Myoporaceae. Pollen & Spores 17:4 — 351 16. oe DOWSKI, J. 1962. Contribuciones a la fitogeogrofia floristica e cand ca de México. 1. Algunas consideraciones acerca del elementa endémico en la flora Mecicura: Bol. Soc. Bot. México. 27:52—65. . 1973. Geographical relationships of the flora of the Mexican dry regions. pp. 61-72. in A. Graham ed., Vegetation and vegatational history of northern Latin America, Elsevier Sci. Publ., Amsterdam STEVENS, P. 1980. Evolutional polarity of character states. Ann. Rev. Ecol. Syst. 1:333-—358. TURNER, B. L. ae ane David Flyr, 1937-1971. Sida 5:54 VAN DEVENDER, T. 1977. Holocene woodlands in the eee deserts. Science arleo = 192: WEBSTER, G. L. 1951. The pales species of Myoporum. Pacific Sci. 5:52 —77. WETTSTEIN, R. von. 1891. gar angen In A. Engler and E. K. Prancl, Die pence Arcam 4(3b):39 Ty | yoporaceae. In A. Engler and E. K. Pranel, Die Natiirlichen Pflanzen- fenilien . 354 — 360 CQ A REVISION OF THE CLASPING-LEAVED POTAMOGETON (POTAMOGETONACEAE) ROBERT R. HAYNES Aquatic Biology Program, Department of Biology 927, University of Alabama University, AL 35486, U.S.A. The genus Potamogeton Linnaeus has long been considered to be one of the more taxonomically difficult and more ecologically important genera of all the aquatic vascular plants. These difficulties and importances have been summarized earlier (Haynes 1974, 1978) and will not here be dis- cussed. Because of the importance and taxonomic confusion, I have begun a taxonomic revision of the genus. Potamogeton is a cosmopolitan genus with approximately LOO species of submersed and floating-leaved aquatic plants. The genus has been separat- ed into two subgenera (Raunkiaer 1896) and numerous sections and subsections (Ascherson and Graebner 1907; Hagstrom 1916). My ap- proach has been to prepare treatments of one or a few subsections at a time, without considering whether the subsection should be recognized, and eventually to combine all these treatments into one volume for the entire genus. At that time the infrageneric classification will be re-evaluated and a treatment presented. This paper concentrates on two quite small subsections, subsection Perfoliati Graebner and subsection Praelong: Hagstrom. Ret atives of these two subsections are the only species of the genus with arias leaves, these leaves all submersed. As a result, Ascherson and Graebner (1907) combined them in subsection Perfoliati, although Raunkiaer (1903) had earlier separated them into two groups, these groups without formal rank. Hagstr6m (1916) followed Raunkiaer in separating the groups, naming the P. praelongus-group of Raunkiaer subsection Praelongi. Fernald (1932) and Ogden (1943) accepted Hagstrom’s taxono- my without comment. The morphological features, as wel dence for combining the two subsections into one. However, the stem anat- — as the flavonoid chemistry, are evi- omy is quite different between representatives of the two subsections. Al- though my inclinations are to combine the two subsections, I do not understand the within and between subsectional variability of the above mentioned characteristics, as well as the other characteristics. SIDA 11f2); 173 — 188. 1985. 174 The nomenclature and morphology of the genus, including descrip- tions, have been published earlier (Haynes 1974, 1978) and will not be restated here. The treatment that follows is based on extensive field study, growth of plants in similar and varied conditions, cytological examination, chemical examination, and an examination of over 3000 herbarium specimens from the following 26 herbaria: AAU, , B, BM, BR, C, CAN, DAO, EG, GH, LE, K, M, MEL, MO, NSW, NY, P, S, TI, UNA, US, V, W, Z, (abbreviations according to Holmgren et al. 1981). Keys are based upon materials containing fruits; dimensions of leaves are taken from the fully expanded, longest leaves of a specimen; measurements of widths are taken approximately at the widest point of the leaf; and descriptions of fruits are taken strictly from mature structures. KEY TO THE SUBSECTIONS OF CLASPING-LEAVED POTAMOGETON |. Leaf apex cucullate; fruits with dorsal keel, mm long ........... Praelongi I. Leaf apex non-cucullate (flattened); fruits without dorsal keels, 1.6 —4.2 Stas Se Shania Soe Pa fag Gan eg he, Sede a eo ah ee we Geeta Rake eae wey ee Perfoliati mm long POTAMOGETON subsection PRAELONGI Hagstrém, Kongl. Svenska Vetenskapsakad. Hand]. 55(5):250. 1916. Type: Potamogeton praelongus Wulfen. Characteristics of the species. POTAMOGETON PRAELONGUS Wulfen in Roemer, Arch. 3:331. 1805S. Type: AUSTRIA. Labuch, 8 Jun 1763, Walfen s.n. Gsorypr: MEL!). P. flexuosum Wredow, Mecklenb. FI. P. ess Dethard. in Sterlitzer Anzeig., P. perfoliatus L. var. lacustris Wallman in S. Liljebl. : acuminatum Wahlenb. , Fl. Upsal. 116. 1820. g. eg. Scand. 1:213. 1845. Naturalist 3:17. 1913. pete praelongus (Wulfen) Nieuwland, Amer. Mid. 1807. no. SO. 1809. Urkast Sv. Fl. 706. I81G, Stems very pale green, simple or branched near apex, terete without nodal glands, to 210 cm long, to 3 mm diam. Leaves usually pale green, rarely olive-green, delicate, mostly alternate, rarely opposite, 11 —33-nerved, 8—28 cm long, 1.1—4.6 cm wide; apex cucullate; lacunae absent: stipules white, fibrous, convolute, free from leaf blade, shredding at ape persistent, 3—8.1 cm long, 3-8 mm wide. Winter buds (turions) ab- sent. Peduncles cylindric, terminal or axillary, erect to spreading, 9.5 — cm long, 1-4 mm diam. Spikes cylindric, 3.4—7.5 cm long, | — diam; verticels 7 — 10, separated by 3 —8 mm. Perianth segments 2 — a6. 5 mm long, 1.5— 2.5 mm wide; anthers 1.1— 2.5 mm long. Fruit olive- green, with dorsal keel, occasionally with lateral keels, widest at or above 2cm Fig. 1. Potamogeton praelongus. A. Habit sketch with axillary inflorescence. B. Fruit. C Enlargement of node with stipule. D. Enlargement of leaf apex with cucullate tip. middle, 4—5.7 mm long, 3.2—4 mm wide; beak erect, marginal, 0.6— 1mm long, 0.6—0.9 mm diam; sides rounded or occasionally with slight central bulge; wall texture mostly wrinkled, rarely smooth. Chro- mosome number, 2n= 52. STEM ANATOMY: The stem anatomy of Potamogeton praelongus 1s char- acterized by a central stele with seven to nine seperate vascular bundles. Ogden (1943) considered this type of stele to be the most primitive and thus labelled it “prototype.” The endodermis is composed of U-cells. These 176 are cells that are thickened on the inner and lateral faces and thin on the outer face. The stem also has interlacunae bundles (vascular bundles at the junctures of walls separating the lacunae), subepidermal bundles (vascular bundles directly underneath the epidermis), and a pseudo-hypodermis of several layers of cells immediately beneath and adjacent to the epidermis. FLAVONOID CHEMISTRY: The leaf flavonoid chemistry of Potamogeton praelongus is the most diverse of any species of the genus reported to date. Roberts and Haynes (submitted) isolated nine flavonoid compounds from the species. These include luteolin aglycone and its glycosides, 7-0-glucoside and 7-Q-glucuronide, isoorientin, apigenin aglycone, and its glycosides, 7-O-monoglucoside and 7-0-diglucoside, and chrysoeriol aglycone and its 7-0-glucoside. DISTRI N: Apparently circumpolar in the Northern Hemi- sphere. In the Western Hemisphere, from north-central Alaska to southeastern Labrador, south to southern Newfoundland, Maryland, southeastern Colorado, and the Aleutian Islands; also central Mexico and the east-central coast of Greenland. The species, in the Western Hemi- sphere, reaches its northern limit about latitude 73°N and southern limit about latitude 38°N, disregarding the one Mexican locality. In the East- ern Hemisphere, from northwestern Norway to western Russia, south to west Yugoslvia, west to east-central France and northern Ireland; also Kamchatka Peninsula, southwest to east-central Japan. In the Eastern Hemisphere, the species reaches its northern limit about latitude 68°N and southern limit about latitude 35°N Although I have seen no material from the vast majority of Russia, the species undoubtedly occurs there. Juzepczuk (1934) lists the species from all parts of the country. Potamogeton praelongus, with its zig-zagging stem, semi-clasping leaves, and cucullate leaf tips, is one of the most easily recognizable species of pondweeds. The species has been known to hybridize with P. perfoliatus on occasion and to produce a sterile off-spring which can perpetuate itself by vegetative methods. The most commonly collected locality for the hybrid is Varming Lake on the Ribe River near Ribe, Denmark. That lake has now been altered considerably and probably does not exist anymore. With the destruction of the lake, the hybrid apparently was extirpated. POTAMOGETON subsection PERFOLIATI Graebner in Ascherson and Graebner, Pflanzenr. 4(11):92. 1907. Tver: Potamogeton perfoliatus L. Plants submersed in fresh or brackish waters, perennial. Stems branched or unbranched, terete, without nodal glands. Leaves all submersed, pellu- cid, sessile, without lacunae, lanceolate to orbicular, obtuse to acute at Te ‘uOLINqIISIP paqusuNs0q ‘s#Suopavad uojasoumrjog “7 “B14 178 apex, clasping at base, entire, 3 —35-nerved. Stipulates convolute, free from base of blade. Winter buds (turions) absent. Inflorescence emersed, a cylindrical spike with 3— 11 whorls of flowers, compact, mostly with 4 flowers at each whorl. Fruit dorsally rounded or keeled, to 4.2 mm long. Chromosome number, 2” = 52 KEY TO THE SPECIES . Stipules fibrous, persisting as fibers; leaf apex mostly acute; fruits 2.2 — 4.2 BTVENY: LOMO 3 esis cscs nae tiaceh aha nade ace Woke ds ee Genie ts P.orichardsonit Stipules delicate, without fibers, mostly decaying early; leaf apex mostly obtuse, rarely acute; fruit 1.6—3 mm long ....................-. P. perfoliatus 2. POTAMOGETON PERFOLIATUS L., Sp. Pl. 1:126. 1753. Tver: SWEDEN/ FINLAND. Lappland. Linnaeus (tecrorypr here designated: Institute de France, Paris, B. Delessert Library; photo of lectotype, BM!). Potamogeton amplexicanlis Kar. Bull. Soc. Nat. Mosc. 173. 1839. P. bupleuroides Fernald in Gray, Manual ed. 7. 75. 1908. P. loeseltt Roem, & Schultes, Syst. Veg. 3:508. 1818. P. perfoliatus L. var. buplenrordes (Fern: ay Farwell, Amer. Midl. Naturalist 8:264. ee P. perfoliatus L. var. caudiformis Aschers. & Graebn., Syn. Mitteleur. Fl. 1:314. 1897. P. perfoliatus, L. var. eee K. Mert. & W. Koch. in Rohl., cane Fl. ed, 3. 14852, 182 P. perfoliatus L. var. deni G. Mey., i Han. 523. 1836. P. perfoliatus L. var. gracilis Fr., Novit. Fl. . Alc. 42. P perfoliatus L. var. /oeselrs ( Roem: & ae io. & Graebn., Syn. Mitteleur. Fl. 1:314. 1836. P perfoliatus L. var. mandichuriensis A. Benn., Annuaire Conserv. Jard. Bot. Genéve OS P. perfoliatus L. var. prolixus Hagstrom, Knogl. Svenska Vetensk. Acad. Hand). 55(5):254. 1916. P. perfoltatus L. var. pseudodensus Aschers. & Graebn., Syn. Mitteleur. Fl. 1:314. 1897. P. perfoliatus L. var. rotundifolius Sonder, Fl. Hamb. 98. 1851. Spirillus perfoliatus (L.) Nieuwl., Amer. Midl. Naturalist 3:17. 1913. Stem pale green, simple or branched near apex, terete, to 2.5 m long, 0.2— 1.9 mm diam. Leaves usually olive-green, dees 3 — 25-nerved, 0.9 —7.6(-9.7) cm long, 0.7—-4 cm wide; apex mostly round, rarely acute; lateral nerves joining midrib at apex. Stipules translucent, delicate, not shredding at apex, convolute, fugacious, 3.5 — 6.5 cm lone, T==15 mm diam. Peduncles cylindrical, terminal or axillary, erect to rarely recurved, 1—7.3 cm long, 0.5—2.2 mm diam. Spike cylindrical, 0.4—4.8 cm long, 4.5 —8 mm diam; verticels 3-9, separated by 1.5 — 3 mm. Perianth segments 1.3 — 2.1 mm long, 0.7 — 1.9 mm wide. Anthers 1— 1.2 mm long, 0.2 —0.7 mm wide. Fruit light green to brown, with- out dorsal keel, widest at or above middle, 1.6 —3 mm lone. Lio 232 ee) C Fig. 3. Potamogeton perfoliatus. A, Habit sketch with axillary and terminal inflorescences. B. Fruit. C. Enlargement of node after stipule has decayed. D. Enlargement of leaf apex with flattened (non-cucullate) tip. mm wide; sides mostly depressed, occasionally rounded; beak central, 0.1—0.9 mm long, 0.1—0.9 mm diam.; wall texture smooth. Chromo- some number, 27 = 52. STEM ANATOMY: The stem anatomy of Potamogeton perfolratus 1s chat- acterized by a “trio-type” stele (see discussion of P. richardsonni for explana- tion). The endodermis is composed of O-cells, cells that are thickened on 180 all—inner, lateral, and outer—faces, giving an appearance of an O. The stems also possess a pseudo-hypodermis, but lack interlacunae bundles and subepidermal bundles (see discussion under P. praelongus for an explanation of these tissues). FLAVONOID CHEMISTRY: The leaf flavonoid chemistry of Potamogeton perfoliatus is more complex than P. richardsonii. Roberts and aynes (submitted) isolated seven compounds from the species. These in- clude luteolin aglycone and its glycosides, 7-O-glucoside and 7-0- glucuronide, apigenin aglycone and its 7-O-monoglucoside, and chryso- eriol aglycone and its 7-O0-glucoside. Harbourne and Williams (1976) isolated a C-glycoflavone from the species. Roberts and Haynes (sub- mitted) did isolate the C-glycoflavone, isoorientin, from the related species P. praelongus and this very possibly is the C-glycoflavone which Harbourne and Williams isolated from P_ perfoliatus. DISTRIBUTION: In the Western Hemisphere, from Labrador to Newfoundland and SW to southern Ontario and North Carolina; also Lake Autlan, Guatemala, east-central coast of Greenland, northeastern Iceland, and the Gulf of Mexico outer coastal plain from western Florida to eastern Loutsiana. In the Eastern Hemisphere, widespread from extreme northern Norway to western Kamchatka Peninsula, Russia, south to southern Ja- pan, southern India, and northern Spain; also, southeastern Australia, eastern Sudan, and southeastern Algeria. Potamogeton perfoliatus, morphologically, is extremely variable, especially in Europe. Asa result, the taxon has been divided into two specfic and 10 or more subspecific categories. These segregates were based exclusively on vegetative variability. After examining hundreds of specimens, I have been unable to divide the morphology into ranges which warrant taxonomic rec- ognition. This variability is continuous, with no geographical dis- tinctions. | have, therefore, decided not to accept any subspecific catego- ries: Potamogeton perfoliatus var. mueller’ Bennett was published based upon several collections by Mr. Mueller from southeastern Australia. I have visited the type locality and have examined the type specimen, as well as several others taken from the type locality. In my opinion, the taxon is not P. perfoliatus and Lam, therefore, excluding it from further consideration at this time. 3. POTAMOGETON RICHARDSONI (A. Benn.) Rydberg, Bull. Torrey Bot. Club 32:599. 1905. Type: U.S.A. Michigan. Robbins 5.n. (ecroryer: GH!). P. perfoliatus L. var. lanceolatus J. W. Robbins in A. Gray, Man. ed. 5. 488. 1867, non 1861 P perfoliatus L. var. richardson A. Benn., J. Bot. 27:25, 889. 181 ‘uorInqiasIp paruauns0q ‘s#ripofsed uojadourjog “y “3tg 182 P. perfoliatus L. ssp. richardsonti Hultén, Fl. Alaska & Yukon 102. 1940. Spirillus perfoliatus (L.) Nieuwl. var. richardsonii (A. Benn.) Nieuwl. Amer. Midl. Naturalist 3:17. 1913 Stem pale green, simple or occasionally branched near apex, terete, ca 97cm long, |— 2.8 mm diam. Leaves lanceolate, usually olive-green, del- icate, 3— 35-nerved, 1.6— 13cm long, 0.5 — 2.8 cm wide €; apex acute to obtuse; lateral nerves joined midrib at apex. Stipules white, fibrous, shred- ding at apex, mostly persistent as fibers, 1.2 — 1.7 mm long, 1-—3.2 mm diam. Peduncles clavate, terminal or axillary, erect to rarely recurved, 1.5 — 14.8 cm long, |—3.9 mm diam. Spike cylindrical, 1.3—3.7 em long, 5— 11 mm diam.; verticels 5 — 11, separated by 2—5 mm. Perianth segments 1.1 —2.5 mm long, |— 2.4 mm wide. Anthers 1.5— 1.6 mm long, 0.5 —0.6 mm wide. Fruit light green to brown, mostly without dorsal keel, widest at or above middle, 2.2 —4.2 mm long, 1.7 —2.9 mm wide; beak central, 0.1— 1.1 mm long, O.1-0.9 mm diam: sides rounded, rarely centrally depressed; wall texture smooth. Chromosome number, 27=52. STEM ANATOMY: The stem anatomy of Potamogeton richardsonii 1s characterized by a central stele in which three of the four median bundles have united to form a “trio” bundle. There are, therefore, only two bundles in the median region of the stele, one of them with two patches of phloe m on the inner face. Ogden (1943) designated this stele as the “trio- type” and considered it an advancement over the proto- type but scill noc the most advanced type. The endodermis is composed of O-cells. These are cells that are thickened on all-inner, lateral, and outer—faces, giving an appearance ofan OQ. The stems also possess a pseudo-hypodermis, but lack interlacunae bundles and subepidermal bundles (see discussion under P praclongus for an explanation of these tissues). FLAVONOID CHEMISTRY: The leaf flavonoid chemistry of Potamogeton richardsonti is less diverse than P perfoliatus. Roberts and Haynes (submitted) isolated five compounds from the species. These in- clude luteolin aglycone, and its glucosides, 7-0- glucoside and 7-0- glucoronide, and chrysoeriol aglycone and its 7-0- glucoside. Harbourne and Williams (1976) isolated a C-glycoflavone from the species. Roberts and Haynes (submitted) did isolate the C-glycoflavone, isoorientin, from the related species P. praefongus, and this very possib! y 1s the C-glycoflavone which Harbourne and Williams isolated from P. richardsonii. DISTRIBUTION: From Aleutian Islands, central Alaska, and Mackenzie Delta to central Quebec, south to New York, Ohio, Colorado, and northern California. Potamogeton richardsoni has been included with P. perfoliatus (Ascherson 183 and Graebner 1907) and accepted at the specific level by Ogden (1943) and Hagstrom (1916). The two taxa are quite easily separated over the vast majority of their ranges. In areas of sympatry, however, the two taxa are difficult to separate. This apparent integradation is evidence for recogni- Fig. 5. Potamogeton richardsonti. A. Habit sketch with axillary inflorescences. B. Fruit. C. Flower. D. Enlargement of node with persistent stipule. 184 tion at a subspecific level. These intermediate forms, however, are almost always sterile. If flowers develop, these flowers rarely produce fruit. Should some subspecific category be warranted, one would expect the inter- mediates produced in an area of sympatry to be fertile, at least fairly commonly. To test the hypothesis that the two taxa are distinct and should be ac- cepted at the specific rank, individuals of the two taxa were grown in similar conditions (both flowing and non-flowing) in the aquatic biology facility at the University of Alabama. The individuals of the two taxa remained morphologically distinct over the duration of the growth period, Also, 77 specimens of the complex were examined for 16 morphological characteristics (see Table 1). These specimens were taken at random from a group of specimens, each of which possessed all the characteristics meas- ured, including both mature flowers and mature fruits. These specimens were tentatively identified to species, 32 as P. perfoliatus and 45 as P. richar- dsonit. The data were analyzed by the N'I/SYS statistical package (Rolf, et al. 1977). The specimens were sorted into two groups by the statistical program, the 32 originally determined as P. perfoliatus in one group and the 45 originally determined as P. richardsonii in the other group. TABLE | P perrouatrus (N = 32) P. RICHARDSONH (N = 45) CHARACTER Stem diameter (mm)* 0.2—1.9 0.92 0.78-1.05 1.0-2.8 1.73 1.58—1.88 Leaf width (cm) 0.7-2.6 1 1.05-1.35 0.7-2.8 1.24 1.1-1, Leaf length (cm)* 0.9-3.8 1.9 1.6-2.2 2.2-10.2 4.4 3.9-4.9 Number veins 3-2] 9.9 8.1-11.7 4—35 12.5 10.1— 14.9 Peduncle length (cm)* 1O-7.3 3.6 3.2-4.0 1.5-11.6 5.4 1.5-6.3 Peduncle diameter (mm)* 0.5-2.2 1.4 Lede 10-3.9 2.2 2.0-2.4 Spike length (cm) 0.4-4.8 1.4 1O- 1.8 13-3.7 2.5 2.3-2.7 Spike diameter (mm) !5—8.0 64 5.9-6.9 5.0-11.0 8.0 7.2—8.8 Number inflorescence verticels* 3-9 5.0 4.4—5.6 sl 3 7.5-8.5 Perianth length (mm) 13—2.1 1.52 1.41-1.63 1.1-2.5 1.71 1.61-1.81 Perianth width (mm) O.7-1.9 1.27 1.17—1.37 1O0—-2.4 1.61 1.36— 1.86 Fruit lengch (mm)* 16-2.8 2.26 2.15-2.37 2.2-4.2 2.9 2.79-3.01 Fruit width (mm)* Lo=22 1.8 1.7—1.9 17-2.9 2.3 2.2-2.4 Fruit beak length (mm) O.1-0.9 0.45 0.39-0.51 0.1-0.9 0.55 0.50—0.60 Fruit beak width (mm)* O.1-0.9 0.32 0.27 —0.37 0O.1—-0.9 0.44 0.40—0.48 Number fruics per spike* 3— 35 10.8 7.6— 14.0 1-65 21.3 16.54— 26.1 X—Mean for all measurements of that characteristic. S E R ig ] | } Z 9 sft *— Indicates characteristics for which there is no overlap of the two species of the two standard errors the mean of tk above and two standard errors below the mean. 185 “UOLINGINSIP palUaWMI0G] “uospavgts uojasourjog “9 “31J 186 The means and standard error were calculated for each group for each characteristic. Table I lists the range for each characteristic, the mean for that characteristic, and the range of two standard errors above and two standard errors below the mean. Whenever there is no overlap between the standard error ranges of a particular characteristic for the two taxa, then that characteristic 1s considered to be statistically valid in separating the two taxa. The standard error ranges do not overlap for 11 of the characteris- tics examined. These characteristics are indicated with an asterisk in Table The growth studies and the statistical analysis, along with the phytoge- ography of the species in North America, lead me to the conclusion that the taxa should be accepted at the species level. The chemical data can be used to help understand the origin of Potamogeton richardsonit. Hagstrom (1916) proposed that P. richardsonii evolved by hybridization of P praelongus and P. perfoliatus. Ogden (1943) did not accept that theory since either one of the putative parents is or both are absent over much of the range of P richardsonii. Neither luteolin aglycone and its two glycosides nor chrysoeriol aglycone and its glycoside may be used as evidence for or against the hybrid origin theory. However, apigenin and its monoglucoside are quite helpful as both these compounds are shared by the two putative parents, although in trace amounts in P perfoliatus. Neither compound has been detected in P. richardsonii, although large quantities of plants have been extracted. The chemical profile of hy- brid taxa is normally additive of the two putative parents (Haynes and Williams 1975), although novel compounds may be found in the hybrid (Ordnuff et al. 1973). The chemical data, with apigenin aglycone and its monoglucoside being present in both putative parental species and being absent in the putative hybrid, support Ogden’s (1943) theory that P richardson arose by means other than hybridization of P. praelongus and P perfoliatus. Hulten (1937) proposed that there were two main refugia in North America—the Rocky Mountains and the continental shelf outside eastern North America—where plants survived during continental glaciation. From these refugia, he stated, the plants probably spread in an easterly and westerly direction, respectively, toward the center of the continent. Some taxa probably survived in one area, while other taxa possibly survived in both areas. Some of the taxa which survived in both refugia migrated until their ranges overlapped; thus, they now have a continuous range across North America. For others, however, migration ceased before the ranges overlapped. Hultén conceded that species surviving in other areas south of the glactal boundary did migrate north and, to some extent, into the once 187 glaciated lands. However, he suggested that a much smaller proportion of the species now found in the glaciated areas survived 1n the vast areas south of maximum glaciation than in the two other refugia. I propose that Potamogeton perfoliatus was, prior to the glacial period, a circumpolar species, with a more or less continuous distribution, includ- ing North America. Glaciation would have separated this North American population into two smaller allopatric ones, one to the east and one to the west. During glaciation, the two populations differentiated, resulting in each population evolving into a different species. Once the period of glacia- tion was complete and the distributions overlapped, the reproductive barriers that had been established during glaciation evidently now prevent gene flow from one population to the other. The western population represents P. richardsonii and the eastern P. perfoliatus. Utis (1965) indicated that Gentianopsis procera and G. crinita possibly had a similar history. Apparently species or varieties could have evolved in a matter of 10,000 years. One need only examine a few of the Great Lakes endemics, e.g. [ris lacustria (Guire and Voss 1963), Cirsium pitcher Johnson and Itis 1963), and Calamovilfa longifolia var. magna (Thieret 1960) for confirmation. Hagstrom (1916), when proposing that P. richardsonii arose as a hybrid between P. praelongus and P. perfoliatus, was correct in pointing out that P. richardsonii is intermediate in size and the fibrous nature of the stipules between the two putative parents. However, the scem anatomy being completely unlike P praelongus, the flavonoid chemistry being non- additive, and the phytogeography all individually and collectively contra- dict Hagstrém’s theory. ACKNOWLEDGEMENTS I am grateful to the curators of the above mentioned herbarta for loan of their specimens. I also thank Mrs. Jan Lay for preparing the illustrations and Mrs. Anni Sloth and Dr. David Lentz for help with the photography. A portion of the manuscript was prepared while | was a Visiting Scientist at the Botanical Institute, University of Aarhus, Denmark. This research was supported by NSF grants DEB 80-21387 and DEB 78-18402. This ts con- tribution No. 67 from the Aquatic Biology Program, The University of Alabama. REFERENCES ASCHERSON, P, and P GRAEBNER. 1907. Potamogetonaccae. In: A. Eng gler. Das Pflanzenreich Regni Vegetabilis Conspectus 4(11):1— 183. Wilhelm Engelmann, Le FERNALD, M. L. 1932. The linear-leaved North American ee of ee section Axillares. Mem. Amer. Acad. Arts 17:1— 183 (Also, Mem. Gray Herb. 188 GUIRE, K. E., and E. G. VOSS. 1963. Distributions of distinctive shoreline plants in the Great Lakes oe Michigan Bot. 2:99 — 114. HAGSTOM, J.O. 1916. Critical researches on the Potamogetons. Kongl. Svenska Veten- skapsakad. Handi: De eel. HARBOURNE, J. B., and C. A. WILLIAMS. 1976. Occurrence of sulphated flavones and caffeic acid esters in members of the Fluviales. Biochem. Syst. & Ecol. 4:37—41. HAYNES, R. R. 1974. A revision 2 fe ae ee eee subsection Pasilli (Potamogetonaceae). Rhodora 76:5 HAYNES, R. R. 1978. The ea in the southeastern United States. J. Arnold Arb. 59:170— 191. HAYNES, R. R., and D. C. WILLIAMS. 1975. Evidence for the hybrid origin of Potamogeton longiligulatus Fern (Potamogetonaceae) Mic oo Bot. 14:94 — 100. HOLMGREN, P K., et al. 1981. Index Herbariorum. Pr. I. The Herbaria of the World. 7th ed, Dr. W. Junk B. V., Publ. The Hague. 452 pp. HULTEN, E. 1937. Outline of the history of arctic and boreal a during the quaternary ate Bokforlags Aktiebolaget Thule, Stockholm. 168 pp + 43 pl. ILTIS, H. H. 1965. The Sa Gentianopsts (Gentianaceae): transfers and phytogeographic comments. Sida 2:129 — JOHNSON, M. EF, and H. ILTIS. 1963. Preliminary a on the flora of Wisconsin No. 48. oS I—Composite family 1. Trans. Wisconsin Acad. Sct. 22054 — ZUK, s V. 1934. Potamogetonaceae. In: V. L. Komarov, Flora of the U.S.S.R. 1:229 — 265 OGDEN, E. C. 1943. The broad-leaved species of Potamogeton of North America north of Mexico. Rhodora 45:57 — 105, 119— 163, 171-214. ORNDUFE R., et al. mc nae of artificial interspecific hybrids in Lasthenia. Biochem. Syst. 1:14 RAUNKIAER, C. oe = one Blomsterplanters Naturhistorie I. Helobieae. Copen- hagen. RAUNKIAER, C. 1903. Antomical Potamogeton-studies and Potamogeton fluitans. Bot. Tidsskr. 25:253 — 380, ROBERTS, M. L., and R. R. HAYNES. (submitted) Leaf flavonoid chemistry of Potamogeton dee Praelongi and Perfoliati. Nordic J. Bot. ILE E Y..; ea Ne ws eta imerical taxonomy system of Multivariate statistical programs. Program printou THIERET, J. W. 660: Calamovilfa longifola and its variety magna. Amer. Mid]. Naturalist 63:169— 176. TAXONOMY, DISTRIBUTION AND RARITY STATUS OF LEAVENWORTHIA AND LESQUERELLA (BRASSICACEAE) IN KENTUCKY RAY CRANFILL Department of Botany, University of California, Berkeley, CA 94720, U.S.A. JERRY M. BASKIN School of Biological Sctences, University of Kentucky, Lexington, KY 40506, U.S.A. MAX E. MEDLEY Department of Biology, University of Louisville, Louisville, KY 40292, U.S.A. ABSTRACT A taxonomic treatment and illustrations are provided for the three taxa of Leavenworthta and two of Lesquerella in Kentucky, and the distribution of each taxon in Kentucky ts shown on a dot distribution map. The ecology and rarity status of each taxon at the state and federal level are discussed. Leavenworthia exigua var. laciniata and Lesquerella globosa Ccurrent- ly are under review by the Fish and Wildlife Service for listing as endangered and threatened, respectively. The vascular flora of Kentucky is interesting and diverse, but it has never been comprehensively studied. In her bibliography of Kentucky’s botanical literature, Fuller (1979) lists many floristic studies but few taxonomic treatments of individual genera or families. With the ever accel- erating destruction of Kentucky’s natural vegetation, such treatments be- come vital, not only as a future document of what is lost but to help save what remains. There is currently a great need for reliable information on the extent and status of all of the state’s flora, and this is especially true for the rare taxa. Since all five taxa of Leavenworthia and Lesquerella are rare 1n Kentucky (Branson et al. 1981; Chester 1982), and since at the time this study was begun Leavenworthia exigna var. laciniata, L. torulosa and Lesqu- erella globosa were under review by the Fish and Wildlife Service for listing as endangered or threatened {Federal Register 45(242): 82517, 82518, 15 December 1980}, we undertook a study of the taxonomy and distribution of these two genera in Kentucky. Leavenworthia torulosa since has been re- moved from the list [Federal Register 48(229):53666, 28 November 1983}. SIDA 11(2): 189— 199, 1985, 90 METHODS This treatment is based on a study of 82 herbarium specimens of Leaven- worthtia and Lesquerella collected in Kentucky and located at ALU, APSC, DHL, E GH, KY, Ky. Agri. Exp. Sta. Herbarium, MICH, MEM, MO, PH, PUL, US, VDB and in several private collections and on extensive field work in Kentucky over the past several growing seasons (herbarium abbreviatons follow Holmgren et al. 1981). Determinations of taxa were made using keys in Rollins (1963) for Leavenworthia and those in Rollins and Shaw (1973) for Lesquerel/a, Measurements of various plant parts cited in the paper were made using a standard dissecting microscope and are based soley on material from Kentucky. Each symbol on the distribution maps is based on an herbarium specimen. RESULTS AND DISCUSSION Leavenworthta Herbaceous, rosette-forming winter annuals; flowers borne singly in erect peduncules, later flowers (if present) borne in lateral, loose racemes: stliques parallel to septum, gynophore short, funiculus free; seeds in a single row in silique, orbicular and flattened (Fig. 1). The systematics (Rollins 1963), evolution (e.g., Lloyd 1965, 1967, 1969; Rollins 1963; Solbrig 1972; Solbrig and Rollins 1977) and ecology (e.g., Baskin and Baskin 1971, 1972, 1976) of Leavenworthia have been studied in considerable detail. All taxa are winter annuals that in preset- tlement times were restricted to cedar (limestone or dolomite) glades (Rollins 1963). However, several of the species, including those that occur in Kentucky, also now occur in disturbed situations such as rocky pastures and plowed fields (Rollins 1981) KEY TO THE SPECIES OF LEAVENWORTHIA IN KENTUCKY I. Siliques markedly corulose; wing of seed very narrow or absent...... 2. L. toralosa I. Siliques not torulose; wing of the seed well developed. 2 2. Petals entire, 5 ~ 7 mm long; terminal leaflet not markedly larger than the lateral ones; style short and stout, 1.0— 2.5 mm LONE noes aed lL. L. uniflora 2. Petals emarginate, 6 — 10 mm long; terminal leaflet much larger than lateral ones; style slender, 2.0 — 3.0 mm long ....... 3. L. exteua var. laciniata I. L. uNirrora (Michx.) Britton, Mem. Torrey Bot. Club 5:171. 1894 Leavenworthia uniflora is the most abundant and widespread species of the genus. In Kentucky, it has been collected in 12 countries (Fig. 2). tt grows on limestone or dolomite glades and outcrops in northwest Georgia, northern Alabama, eastern and central Tennessee, southern Ohio, Figure 1. Morphological features of Leavenw Me L. one a. habit, b. flower, e. leaves from basal rosette, f. silique; L. wiflora, c. flower, d. leaves, g. silique; L exigua vat. laciniata, hh. silique. Bar equals | cm. (Flowers aa from “Rellins 1963.) southeastern Indiana and the Ozark Region of southern Missouri and northern Arkansas (Rollins 1963). In Kentucky, L. wniflora grows in cedar glades, on disturbed rocky ledges and outcrops and in rocky pastures. Population size varies greatly; often only a few individuals are found at a site, although in favorable situations there may be several thousand plants. e know of only about a dozen extant populations, but this may be due to lack of thorough botanical collecting in the parts of Kentucky where it grows rather than to its real rarity. In addition, this species ts inconspicu- ous during most of the year, especially during the summer collecting sea- son. Given the large size of several populations and their scattered occurrence over a large area of the state, we do not feel that L. wniflora is threatened in Kentucky. It is not considered threatened nationally and ts not currently under review. However, L. wniflora is listed as endangered in Indiana (Bacone and Hedge 1980) and threatened in Ohio (Roberts and Cooperrider 1982; McCance and Burns 1984). OS > wee, fee ee y eerasiyy’ “ge, as aaa wi CHRISTIAN Figure 2. Distribution of Leavenworthia in Kentucky. L. exigua var. laciniata (triangles); L. torulosa (squares), L. uniflora (circles). Symbols represent exact localities, except for the Oldham County site for L. wniflora. 2. L. rorurosa Gray, Bot. Gaz. 5:26. 1880. Leavenworthia torulosa is primarily a species of the Nashville Basin with outlying populations on the Mississippian Plateau of Kentucky (Fig. 2) and in the Ridge and Valley Physiographic Province of east Tennessee (Rollins 1963). It is found on glades and in disturbed rocky pastures, where it often grows in and around seasonal pools and wet depressions. Although not listed from Alabama by Rollins (1963), there is a specimen of L. torulosa in the Mohr Herbarium at the University of Alabama in Tuscaloosa (Baskin and Baskin 1984). The label on the specimen indicates that 1t was collected by Charles Mohr in 1880 in Madison County. The specimen has not been annotated by Rollins, and thus he apparently was 193 unaware that the species had been collected in Alabama. Mohr (1901) in- cludes the species in his Plant Life of Alabama and refers to its rarity in that state. Leavenworthia torulosa potentially may be of economic value. Its seeds contain a high level of the fatty acid eicosenoic acid (Miller et al. 1965) that may prove to be of industrial value. Appelquist (197 1) states that further studies of L. torulosa should be undertaken if a specific industrial need for eicosenoic acid arises. Baskin and Baskin (1977) have discussed the status of this species in Kentucky and proposed that it was threatened with extinction in the state. In 1980, this species was under review for listing by the Fish and Wildlife Service as threatened [Federal Register 45(242): 82517, 15 December 1980}, but it no longer is under consideration [Federal Register 48(229): 53666, 28 November 1983]. Leavenworthia torulosa is listed as endangered in Kentucky (Branson et al. 1981) and threatened in Tennessee (Collins et al. 1978; Tenn. Dept. Conserv. 1982) and Alabama (Freeman et al. 1979). Given the paucity of specimens of this taxon collected in Alabama and the apparent lack of any recent collections, we suggest that it should be listed as endangered in that state. It probably is extirpated in Alabama. 3. L. exicua Rollins var. raciniata Rollins, Contr. Gray Herb. 192: 75. 1963. Leavenworthia exigua vat. laciniata apparently is endemic to Silurian dolomite and limestone outcrops (cedar glades) of Bullitt and Jefferson counties, Kentucky (Fig. 2). It grows in dry sites on glades similar to those of L. uniflora, although it never has been found with this species. Leaven- worthia exiqua vat. laciniata frequently is found on glades that have been disturbed by pasturing or that occur along roadsides, and it has invaded a plowed field adjacent to a glade at one Bullitt County site. Baskin and Baskin (1981) have studied the ecology and distribution of this species in Kentucky. Leavenworthia exigua vat. laciniata is currently under review for listing by the Fish and Wildlife Service as endangered [Federal Register 45(242): 82517, 15 December 1980]. Because of its lim- ited geographical range and narrow habitat requirements, we believe that it warrants protection at both the state and federal levels. This taxon is listed as endangered in Kentucky (Branson et al. 1981). Leavenworthia ex- igua Rollins var. exigua is listed as threatened in Georgia (McCollum and Ettman 1977) and endangered in Tennessee (Collins et al. 1978), and L. exigua Rollins var. dwtea Rollins is listed as endangered in Tennessee (Collins et al. 1978) and Alabama (Freeman et al. 1979). Leavenworthia exigna vat. exigua and L. exigua var. lutea currently are under review for 194 listing as threatened by the Fish and Wildlife Service [Federal Register 48(229): 53652, 28 November 1983}. EXCLUDED SPECIES Leavenworthia aurea Vorrey was reported from Kentucky by Short (1840). However, material at PH collected by Short and labeled by him as such ts L. torulosa. One of the specimens ts dated 1840, and the other has no date. In 1840, only two species of Leavenworthia (L. uniflora and L. aurea) had been described. Short knew L. wniflora, and he must have assumed that any Leavenworthia that was not L. wniflora was L. aurea. Leavenworthia aurea is restricted to calcareous glades and outcrops in southeastern Oklahoma and eastern Texas (Rollins 1963). A specimen of Leavenworthia stylosa Gray in the herbarium of the Field Museum (sheet #790163) bears the label “Kentucky,” but it is without name of collector or date. Dr. R. L. Stuckey, of The Ohio State University, kindly analyzed the handwriting and confirmed our suspicion that it is not that of Dr. Short. Since the source of the specimen is in doubt and the collector ts unknown, this species should not be included in the Kentucky flora until more convincing evidence is found that it does occur in the state. LESQUERELLA Herbaceous, stellate-pubescent annuals or perennials; flowers orange- yellow (in ours), borne in dense, nearly flat-topped racemes (elongating as fruit matures); siliques globose or somewhat compressed, wall turgid or thin and papery, funiculus attached to septum towards the base, septum entire with prominent central nerve above; seeds globose (Fig. 3). This genus of about 70 species recently was monographed by Rollins and Shaw (1973). [t has been the subject of some interesting studies of inter- specific hybridization (Rollins 1957; Rollins and Solbrig 1973). Seeds of a number of species of Lesquerel/a have been investigated as a source of indus- trial oils (Hinman 1984). Both species of Lesgwere/la that occur in Ken- tucky have been investigated for their oil content (Barclay et al. 1962; Mikolajczak et al. 1962). Very little information is available in the litera- ture on the species’ ecology. KEY TO THE SPECIES OF LESOQUERELLA IN KENTUCKY 1. Plancs of alluvial situations; most leaves auriculace; siliques 4 — 7 mm long, SECON COMPIESSEU 6 fede ween dae eee RRS bees bee pane lL. Ly lesearit L. Plants of dry, rocky or gravelly situations; leaves never auriculate; siliques 2208 Mt Lang, IODC 6.4» 29 s.e ected eee enn eesya daw eaueu a os lo Dee igure 3. Morphological features of Lesquerella. L. lescuri, a. habit, c. silique, e. trichomes; L. ag b. silique, d. trichomes. Bar equals 1 cm, except for trichomes where it equals O. 1 1 1. L. rescurn (Gray) Wats., Proc. Amer. Acad. Arts 23:250. 1888. Lesquerella lescurii recently was collected in Kentucky for the first time, from Trigg County (Fig. 4) just north of the Tennessee state line, by Dr. E. W. Chester of Austin Peay State University in Tennessee (Chester 1982). According to Rollins (1981), it is a species with “weedy tendencies” that grows in old fields, river bottoms and roadsides. In Kentucky, 1t 1s restricted to disturbed bottomlands along the Cumberland River. The im- poundment of the Cumberland River by Kentucky Dam probably de- stroyed much of the suitable habitat for this species in Kentucky. This species was shown to be more abundant in Tennessee than previous- ly thought and thus is no longer under consideration for protection by fed- eral authorities {Federal Register 45(242): 82553, 15 December 1980}. Lesquerella lescurii was listed as threatened in Tennessee by Collins et al. (1978), but it no longer is considered to be rare in Tennessee (Tenn. Dept. Conserv. 1982). Apparently, the species has a very narrow geographical distribution in Kentucky and thus ts a rare plant in that state. 196 2. L. GLososa (Desv.) Wats., Proc. Amer. Acad. Arts 23:252. 1888. Lesquerella globosa is a taxonomically isolated member of the genus con- fined to Kentucky (Fig. 4) and Tennessee, except for its occurrence at one outlying site in southwestern Indiana (Rollins and Shaw 1973). It was collected once in Indiana by Deam et al. (1941), but Deam thought that it may have been introduced into that state (Rollins and Shaw 1973). Accord- ing to Rollins and Shaw (1973), L. globosa is a perennial; Payson (1921) says that it is biennial or perennial. Essentially nothing is known about the ecological life cycle of this species, and its autecology is in need of study. We have collected it in a variety of situations, from gravelly roadsides and rights-of-way to (calcareous) shaley, exposed slopes and rocky ledges. e BOURBON Ss, WOODFORD ® ® LS JESSAMINE MD Saws: tig Figure 4, Distribution of Lesquerella in Kentucky. L. globosa (circles); L. lescurii & (triangle), Symbols represent exact localities, except for the sites in western Jessamine and southwestern Fayette counties where they are approximate. Ly Rollins and Shaw (1973) give the distribution of L. globosa in Kentucky as Franklin, Mercer and Powell counties, and they cite only one specimen from each county. Interestingly, L. g/obosa has been reported from Ohio (Jones 1940) and Kansas (Rydberg 1932), but apparently there are no specimens from either of these two states. Lesquerella globosa currently is under review by the Fish and Wildlife Service for listing as threatened [Federal Register 45(242):82518, 15 De- cember 1980}. The species is listed as endangered in Indiana (Bacone and Hedge 1980) and as threatened in Tennessee (Collins et al. 1978; Tenn. Dept. Conserv. 1982) and Kentucky (Branson et al. 1981). Although sev- eral extant populations in Kentucky are known, the continued existence of this species is precarious. Most of the sites are immediately adjacent to roads, which, if widened, could destroy them. According to Rollins (1952), L. globosa is the only species of Leavenworthia or Lesquerella in Ken- tucky that does not have weedy tendencies. However, in Kentucky, L. globosa has been found along gravelly roadsides and on recent roadcuts REFERENCES 1971. Lipids in Cruciferae: VILL. The fatty acid ee of sceds . Soc. 48:74 Amer. O11 Che : 740 — 744. and APPELQUIST, L.-A. of some wild or pa ally sees ated species. J. BAC : A. and C. L. IGE. 1980. A_ preliminary lise of end engcres Shree vi ee plants i in i ina. Proc. Sia BARCLAY Aj as biotse GEN and Q. JO crops Il. Lesquerella ( es as a source of new oil s BASKIN, J. M. and C. C. BASKIN. 1971. Germination ecology and adi sed to ha in Leavenworthia iS cena ae Amer. Midl. Naturalist 85:22 The ecological life cycle of the cece ur gli ae endemic Leaven- Boc. SO: 1711 — 172 Indiana Acad. Sci. 8$ NES. 1962. T ine search for new oer eeds. Econ. Bot. 16:95 0. yitat oo _~ and worthta Ce var. a & anad. J. peter 2A incl eee ok as tor metabolic ape to flooding in Leaventwor- thia onli J.C ene Ter 141 —447 1977 : Eiaeeas orb nee Gray: An endangered plant species in aera Castanea “42: 15 = Ne and «1981. Geographical distribution and notes on the ecology of the rare endemic Leavenu ee exigua vat. laviniata. Castanea 46:243 — 247 1984. On the historical occurrence of two cedar gl mae endemics in als bee Ree a a discussion of Mohr's yellow-flowered Leavenworthia. Castanea 49: 167 — es BRANSON, , D. E HARKER, J. M. BASKIN, M. E. MEDLEY, D. L. BA re M . WARREN, JR., W. H. DAVIS, W. C. HOUTCOOPER, B. MONROE, JR., L. R. PHILLIPPE si P CUPP. 1981. Endangered, threatened, and — 89. rare animals and plants of Kentucky. Trans. Kentucky Acad. Sct. 42:77 CHESTER 1982. Some new ‘auibucional records for ee ue eee Waco, es including - first report for Kentucky. S COLLINS, J. L., H. R. DESELM, A. EVANS, R. KRAL and B. BE. WOFFORD. as The rare vascular sas of Teese J. Tennessee Acad: Sct. 957128— 13 DEAM, C. C., R. BEL, T. G. YUNCKER and R. C. FRIESNER. on Indiana 198 hop distribution records. Proc. Indiana Acad. Sci. $1:120 — 129. FREEMAN, J. D., A. S. CAUSEY, J. W. SHORT and R. R. HAYNES. 1979. End angered, threatened and special concern plants of Alabama. J. Alabama Acad. Sci. 50:1 — 26. FULLER, M. J. 1979. Field botany in Kentucky: a reference list. Trans. Kentucky Acad. Sci. 40:43 — 51. HINMAN, C. - _ New crops for arid ae Science 225:1445 — 1448. HOLMGREN, , W. KEUKEN and I SCHOFIELD (Compilers). 1981. Index Tene - 1, edition 7. The i saria of the world. Bohn, Scheltema & Holkema, Utrecht/Antwerpen. Dr. W. Junk B. V., Publishers. The Hague/Boston. 452 JONES, C. H. 1940. es to the revised catalogue of Ohio vascular plants. VIII. Ohio J. Sci. 40:200 — LLOYD. DG. 1965, an of self- com partpiliny and racial differentiation in Leaven- worthia (Cruciferae). Contr. Gray Herb. 195:3 — 134. ————. 1967. The genetics of i. incompatibility in Leavenworthia crassa Rollins (Cruciferae). Genetica 38:227 — ; —_. 1969. Petal color cineca ae in Leavenworthia (Cruciferae). Contr. Gray Herb: 198:9 — 40. MCCANCE, R. M., JR. and J. F BURNS, eds. 1984. Ohio endangered and threatened vascular plants: Abstracts of state-listed taxa. Division of Natural Areas and Preserves, Department of Natural eae Columbus, Ohio. 635 p MCCOLLUM, J. L. and D. R. ETTMAN. 1977. ya protected plants. Georgia Dept. of Neral ees Atlanta, Georgia. 64 p. MIKOLAJCZAK, K. L., EF R. EARLE and I. A. WOLFE 1962. Search for new industrial oils. VI. Seed of - genus Lesquerella. J. Amer. Oil Chem. Soc. 39:78 — 80 MILLER, R. W., . EARLE and I. A. WOLFE 1965. Search for new industrial oils from 102 species of pe rac. J. Amer. Oil Chem. Soc. 42:817 — 821 MOHR. C. 1901. Plane life of Alabama. Contr. U.S. Natl. Herb. a 1-921. are E. - 1921. A monograph of the genus Lesqvere//a. Ann. Missouri Bot. Gard. ee M. L. and T. S. COOPERRIDER. 1982. Dicotyledons. IN: Cooperrider, T. S., ed. Endangered and threatened plants of Ohio. Pp. 48-81. Ohio Biol. Surv. Biol. Notes No. 16, 92 p. ROLLINS, R. C. 1952. Some Cruciferae of the Nashville Basin of Tennessee. Rhodora 54:182— 192 imate hybridization in Lesguerella (Cruciferae). Contr. Gray Herb. I81:1— ; _ evolution and systematics of Leavenworthia (Cruciferae). Contr. Gray Herb. 192:3 eae OO Wee of the Cruciferae (Brassicaceae) in North America. J. Arnold Arbor. 62:5 17 — 54 nd . “SHAW. 1973. The genus Lesguerella (Cruciferae) in North America. Harvard Univ. Press, Cambridge, Massachusetts. 288 p and O. T. SOLBRIG. 1973. Interspecific hybridization in Lesquerella, Contr. Gray L RYDBERG, P. A. 1932 . Flora of the prairies and plains of North America. Published by the New York Beenie Garden, New York. 969 p. SHORT, C. W. 1840. A fourth ieee catalogue of the plants of Kentucky. 192 Western J. of Medicine and Surgery 11:283 — 288. SOLBRIG, O. T. 1972. Breeding system and genetic variation in Leavenworthia. Evolution 2621957 160: and R. C. ROLLINS. 1977. a ee of autogamy in species of the mustard a recat orthia. Evolution 31:265 — 28 TENNESSEE DEPARTMENT OF oe ie. 1982. Official rare plant list of a 8 p. (unpubl.). DIOECY IN NORTH AMERICAN CACTACEAE: A REVIEW BRUCE D. PARFITT Desert Botantcal Garden 1201 N. Galvin Parkway Phoenix, AZ 85008, U.S.A. ABSTRACT Of the six species of Cactaceae described as dioecious, only Opuntia stenopetala Engelm.., O. grandis Pfeiffer and O. glaucescens Salm-Dyck are dioecious. Mammillaria dioica K. Brandegee and M. neopalmers Craig are gynodioecious or possibly trioecious, differing from one population to another. Se/enicerens innesii Kimnach is gynodioecious or probably hermaphroditic with sterile, abortive flowers that appear pistillate. Inadequate data and careless word usages have obscured the true sexual condition of the latter three spec The sexual condition of the Cactaceae is generally regarded as being hermaphroditic, or monoclinous (Core 1955; Porter 1959): that is, all plants bearing perfece flowers (Swartz 1971). That there are exceptions is indicated by the Britton and Rose (1937) description of the family: “Flowers usually perfect " Likewise, Bravo-Hollis (1978) in her description of the order Cactales hints that exceptions exist: “Flores casi siempre hermafroditas . . .” Benson (1969a, 1969b, 1969c, 1982), in his description of the family avoids the issue. ‘To pursue the statement of Britton and Rose (1937), one must scan their 1235 species descriptions, for no mention of imperfect flowers is made at the generic level. One finds that three Mexican species of Opuntia, series Stenopetalae, and one primarily Mexican species of Mammillaria (as Neomam- millaria) are considered dioecious. Recently a new species from St. Vincent Island, West Indies, Se/enicerens innesit Kimnach, was described as “the only confirmed example of complete dioecity (sic) in the Cactaceae” (Kimnach 1982). TERMINOLOGY Before examining these claims of dioecy (= dioecism, cf. Bawa & Opler L975) in the Cactaceae one must first establish an understanding of the terminology. The usual sexual condition in cacti is hermaphroditic or monoclinous. This means that all plants of a given taxon have perfect (bi- sexual) flowers (Usher 1966; Swartz 1971, Radford et al. 1974). In con- trast to hermaphroditic is monoecious: plants with flowers not perfect, the SIDA 112). 200 — 206. 1985, 201 staminate and pistillate flowers on the same individual. Dzsecrous plants also have all flowers imperfect (unisexual) but with the staminate and pistillate flowers on separate individuals (Radford et al. 1974). Gynodtve- cious seems to be transitional between hermaphroditic and dioecious (Ross 1970) with some plants bearing perfect flowers and others pistillate ones. The uncommon term, ¢rioecious, refers to a species with some plants stamt- nate, some pistillate, and some perfect Jackson 1928, p. 392; Usher 1966; Swartz 1971; Radford et al. 1974, p. 144). though some authors describe individual flowers (rather than whole plants) as dioecious (Britton & Rose 1937; Kimnach 1982) or monoecious (Standley 1920 — 1926), in modern usage these words correctly may be used only to describe the arrangement of reproductive parts on whole plants (Lawrence 1951). Hence one may call a single flower staminate, pistillate or perfect, but before one may use the words defined in the preceeding paragraph, one must know the sexual condition of other indi- vidual plants of the species (Lawrence 1951). DISCUSSION The pistillate flowers of Se/enicereus innesii are described as lacking stamens and staminal nectaries, and as often having a reduced number of ovules (Kimnach 1982). The stamen-bearing flowers have nectaries, more numerous ovules, and a “style ca 4—5 cm long, ca | mm thick, the apical portion magenta, white below, the stigma lobes 7 — 8, lorate, obstuse, slightly expanding, 1— 1.5 mm long and to 1 mm wide near apex.’ (Kimnach 1982). Kimnach refers to these as “perfect flowers.” It is clear from the description and illustrations (Kimnach 1982) that S. innesti is gynodioecious with pistillate and perfect flowers but no staminate ones. Therefore, this cannot be considered a species with “complete dioecy.” Of interest is the description of the ovule chamber in the flowers of the pistillate plant: “much of the cavity being occupied by one or more rudi- mentary styles terminating in stigmatic papillae.” This, combined with the fact that the fruits and seeds are unknown, suggests that the structural- ly pistillate flowers may actually be malformed to the point of being totally sterile. If this proves to be the case and only the perfect flowers are func- tional, the species is neither dioecious nor gyndioecious but functionally hermaphroditic. Mammillaria divica K. Brandegee is one of the four species considered by Britton and Rose (1937) as dioecious. However, they say it 1s “incompletely dioecious.” When the species was originally described, 202 Brandegee (1897) stated that “Both the type and the variety (zsw/aris) are nearly dioecious, many plants male, with imperfect, less-divided style- branches, which rarely bear fruit, and the few which occasionally appear (are) very slender and few-seeded; many female, with entirely abortive anthers and very small flowers, which usually produce a row of thick oval or clavate, coral berries; others hermaphordite or imperfectly dioecious in all degrees.” Lindsay (1967) commented that the M. diocia segregate, M. angelensis Craig, also has “occasional pseudo-dioecious flowers.” Of M. dioica, Benson (1969b) states “plant with a strong tendency to be dioe- cious, 1.¢., for the flowers of some plants to have small, sterile anthers and cee stigmas and those of other plants the opposite.” In their discussion of M. dtoica, Lindsay and Dawson (1952) state that dioecy represents “an ex- ceptional rather than a usual character of the plant. The dioecius condition is not frequently observed, and moreover, is not confined to M. dioica but occurs occasionally in other species such as M. neopalmeri.” Brandegee (1897) discussed M. neopalmeri (as M. dioica var. insularis K. Brandegee) with M. droia, stating that flower parts are the same in both taxa (see above quotation from Brandegee). A population of M. dioica was examined by Ganders and Kennedy (1978). They found some plants with perfect flowers and others with pistil- late flowers. Both set fruit with apparently normal seed. No “male” flowers were seen. The pistillate flowers bore “stamens with indehiscent anthers that contain no pollen” (Ganders & Kennedy 1978). In a microscopic ex- amination of the flowers of M. dioica and its segregates, M. estebanensis Lindsay, and M. mu/tidigttata Lindsay, Bemis et al. (1972) determined that the functionally pistillate, “male sterile,” flowers have indehiscent anthers with malformed pollen. Ganders and Kennedy correctly state that a (func- tionally) gynodioecious condition is indicated. However, they point out that they (Ganders and Kennedy) studied only one inland population, and that coastal plants observed by Brandegee (1897) may have been misinter- preted or may actually have had a different sexual condition. If the plants are as described by Brandegee (1897) and as indicated by Lindsay and Dawson (1952), the term trioecious would most accurately describe M dioica and M. neopalmeri. The description of Opuntia series Stenopetalae (Britton & Rose 1937) states “This 1s an anomalous group in Opuntia since the flowers are dioe- cious and the petals are linear and more or less erect.” Opuntia stenopetala Engelm. ts described as having “male flowers with an abortive, pointed style, but female flowers with 8 or 9 yellow stigma lobes on style . Opuntia grandis Pfeiffer and O. glaucescens Salm-Dyck, the other two species of the series, are not described in comparable detail; no further mention is « 203 made of dioecy in Opuntia in Britton and Rose (1937). Brave-Hollis (1978) supports the observations of Britton and Rose (1937) in her description of the genus Opuntia: “Flores generalmente hermafroditas . . .” Her key separates subgenus Stenopuntia from subgenus Opuntia on the basis of plants “unisexual” versus hermaphroditic, respec- tively. The “unisexual” character is repeated in the descriptions of the sub- genus Stenopuntia, of Opuntia stenopetala and of var. stenopetala. The latter description is the most detailed: “estilo abortado en las flores mascultnas, en las femeninas es muy grueso en la parte media; lobulos del estigma 8 a 9, amarillos . . .” A population of this variety from el Cardonal, Hidalgo, Mexico, is described thus: “En las flores masculinas el gineceo esta parci- almente atrofiado y el estilo es claviforme, abajo rosa y arriba amarillento con los I6bulos del estigma atrofiados pues terminan en una punta aguda, rigida; el ovulario también atrofiado” (Bravo-Hollis 1978). The pistillate flowers are not described. Opuntia stenopetala var. inerme Bravo has simular flowers and O. g/awcescens is “generalmente dioicas” (Bravo-Hollis 1978). Although Bravo-Hollis’ description of 0. grandis does not mention dioecy, the species is in the subgenus characterized as “unisexual.” I examined the flowers of O. stenopetala on herbarium specimens at ASU. Seven of the eight sheets from different localities had only staminate flowers, a disproportionate number resulting from collection for metotic chromosome studies (Pinkava, pers. comm.). Present with the stamens was a pointed style which lacked a stigma (Fig. 1).' Ovules were apparently lacking in the reduced ovule chamber in the stiminate flowers of all but one specimen. The eighth sheet had flowers with style, stigma, and stamens (Fig. 2), but when the mature stamens were examined at 400x magnifica- tion, they were found to be indehiscent and completely lacking pollen. Thus, as in Mammillaria dioica, the flowers of the specimen appear perfect but are functionally pistillate. Because this small sample supports the previously published descriptions of the species, 0. sfenopetala is to be con- sidered functionally dioecious. Opuntia grandis and O. glaucescens were not available for me to study. However, we might cautiously assume that they are also dioecious because according to both Bravo-Hollis (1978) and Britton and Rose (1937) the three species constitute a series or subgenus characterized by unisexual flowers. CONCLUSIONS Only six species of Cactaceae have been described as dioecious. Of these only Opuntia stenopetala is almost certainly dioecious. Opuntia grandis and 0. ‘For illustrated longitudinal sections of the flowers of M. dioica and S$. mmnesi, see Ganders & Kennedy (1978) and Kimnach (1982), respectively. Figure |. Staminate flower of Opuntia stenopetala in longitudinal section. Scale line equal to | cm. glaucescens are probably dioecious also. Mammillaria divica and M. neopalmert are gynodioecious or possibly trioecious, apparently differing from one population to the next. Se/enicerens innesii is gynodioecious or possibly hermaphroditic with sterile, abortive flowers that appear pistil- late, The sexual conditions of these cacti have been misunderstood for two reasons. First, inadequate data for the populations make it difficult to accu- rately assess the sexual condition of the species. Second, careless usage of the word dioecious has usually obscured the true sexual condition even where populations or species were studied adequately. According to Brandegee (1897), other species, presumably of Mammil- larta, are “completely unisexual.” Doubtless there are species, in addition to those discussed here, in which the arrangements of reproductive parts need to be carefully observed and accurately reported. 205 Figure 2. Functionally pistillate flower of Opuntia stenopetala in longitudinal section; the stamens produce no pollen. Scale line equal to 1 cm. ACKNOWLEDGEMENTS I chank D. J. Pinkava, T. F Daniel, and N. H. Harriman for critically reading the manuscript, and Mark W. Mohlenbrock for comments on the Mammillavia divica complex which he is now studying. REFERENCES BAWA, K. S. & P A. OPLER. 1975. Dioecism in tropical forest trees. Evolution 34:467 — 474. 206 BEMIS, W. PB, J. W. BERRY & A. J. DEUTSCHMAN. 1972. Observations on male sterile ee Cact. Succ. J. (Los Angeles) 44:256. BENSO 1969a. The cacti of Arizona. Ed. 3. University of Arizona Press, Tucson. ea ae ne a native cacti of California. Stanford University Press, Palo Alto, Gina L, 1969c. Cactaceae. In: C. L. Lundell & collaborators. Flora of Texas 2:221— 317. ‘Texas ae Foundation, ate Texas The cacti of the United States and Canada. Stanford University Press, Stant we aa ea E ( K. 1897. Notes on cacteae. | I-12 ors HOL : IS. H. 1978. Las cactaceas de México. Ed. 2. Vol. | nae de México, México D. F JN J R . Cacteae of Baja California. Erythea Universidad Nacional BRIT _L. & J. N. ROSE. 1937. The Cactaceae. Ed. 2. Vol. | — 4. Dover Publica- tions, se (reprint, 1963), New York. RE, E. L. 1955. Plant taxonomy. Prentice-Hall, , Engelwood Cliffs, New Jersey. GANDE RS, ER. aH. KENNEDY. Madrono 25:2 JACKSON, B. D. a A glossary of botanic terms. Ed. 4 1960), London. sealer M. 1982. Selenierens mnesit, an aberrant new species from the West Indies. “act. Succ. 2 (Los Angeles) 54:3 —7. ee NCE, G. H. M. 1951. Taxonomy of vascular plants. The MacMillan Co., New 1978, ae in Mammutlaria diowa (Cactaceae). . Phototype Ltd. (reprint, Yo LINDSAY. - 39:31 = & E. 'y. DAWSON. 1952. Mammillarias of the islands off northwestern Baja California, Mexico. Cact. Succ. J. (Los Angeles) 24 FER, C. 1 967. A new species of Mammillaria. Cact. Succ. J. (Los Angeles) es R, . L. 1959. ‘Taxonomy of flowering plants. W. H. Freeman and Co., San oe He E., W. C. DICKISON, J. R. MASSEY, & C. R. BELL. 1974 plant systematics. Pia se and Row Publishers, Inc., New York ROSS, M 1970. Evolution of dtoecy from gynodioecy. Evolution 24:827 — 828. STANDLEY, PC. 1920-1926. Trees and shrubs of Mexico. Contr. U.S. Natl. Herb. . Vascular 23:1-— 1721. SWARTZ, D. 1971. Collegiate dictionary of botany. The Ronald Press Co., New York. USHER, G. 1966. A dictionary of botany. D. Van Nostrand Co. , Inc., Princeton, New Jersey. AMPHIBROMUS SCABRIVALVIS (GRAMINEAE) IN LOUISIANA M. LYNN CALAWAY and JOHN W. THIERET Department of Biological Sciences, Northern Kentucky University Highland Heights, KY 41076, U.S.A The South American grass Amphibromus scabrivalvis (Trin.) Swallen was reported in 1967 (Flinchum & Baker 1967) as an introduced weed in Lout- siana strawberry fields (Tangipahoa Parish). Since then, the name of the species has appeared in some pertinent floristic accounts (Allen 1980; Kartesz & Kartesz 1980; Thieret 1972) but not in others that should have included it (Gould & Shaw 1983; Shetler & Skog 1978; Soil Conservation Service 1982). The purpose of our paper is to call attention to the contin- ued occurrence of this grass in the United States and to present descriptive data on the species. AMPHIBROMUS IN LOUISIANA The date of arrival and the method of introduction of A. scabrivalvis into Louisiana are uncertain, although the species is said to have been dis- covered in Tangipahoa Parish “in the late 1950's” (Flinchum & Baker 1967). One strawberry grower told us that he first noted Amphibromus the year after he had set out strawberry plants imported from Argentina. Whatever the source, the new weed was obviously “established and actively growing,” increasing “the cost of production [of strawberries]... because of the extra labor and time needed to control it” (Flinchum 1966). In 1984, about a quarter of a century later, attempts are still being made to eliminate the grass, which apparently has not spread far from the original point of introduction. In Louisiana, A. scabrivalvis is well adapted to the cultural practices used in the production of strawberries, which are grown there as annuals and are generally mulched with black plastic. When the soil is prepared in fall (late October or early November) to receive the strawberry transplants, the grass makes its appearance. The infestation may be from perennating buds left in the soil or from caryopses produced during the preceding season. The plant is more or less dormant during winter, but vigorous growth and tllering are resumed as soon as spring weather becomes favorable. Terminal panicles appear in April-June. Maturation of Amphibromus caryopses occurs just before the strawberry season is completed. After harvest, a non-selective, SIDA 11(2): 207-214. 1985. 208 top-kill herbicide (e.g., Paraquat) is sprayed over the fields to eliminate strawberry plants and weeds (the grass, if not sprayed, may continue vege- tative growth during the summer). A second crop (e.g., of peppers) is then set into the soil through existing holes in the plastic mulch. CLEISTOGAMY According to Campbell et al. (1983), cleistogamy has been reported in 83 genera of grasses—about 19% of the total number of genera in the fami- ly. Amphibromus is one of these (Burkart 1969; Nicora 1978: Rosengurtt & Arrillaga de Maffei 1961; Rosengurtt et al. 1970; Stopp 1958; Torres 1970). In A. scabrivalvis, cleistogamous spikelets (Fig. Id, e, f, g) are produced within the leaf sheaths at all nodes of the stem (the number of nodes may be as many as 10). Up to three or four of these nodes may, as often as not, be underground. The spikelets (“cleistogenes”; see Chase ID 18) at the lowermost one or two nodes (Fig. Id, e) are strikingly differ- ent from those of the terminal panicle (Fig. la, b, c), so different, indeed, that if their source were not known they would hardly be considered to belong to A. scabrivalvis at all. Floret number is one to three (spikelets of the terminal panicle produce three to nine); the apex of the lemma is but slightly notched, erose, or mucronulate (lemmas of terminal spikelets are deeply 2-lobed to or slightly below the middle and bear a long, dorsal awn); and the caryopses are 3—4.5 mm long and 1.5—2 mm wide (caryopses of terminal spikelets are 2—3 mm long and 1.2—1.3 mm wide). At successively higher nodes the cleistogamous spikelets (Fig. If, ¢) be- come progressively more like those of the terminal panicle: floret number increases, lemma lobing is initiated and becomes deeper, awns appear and grow longer (alchough not more than about half the length of the awns of terminal spikelets), and caryopsis size decreases. That A. scabrivalvis commonly produces its most reduced cleistogamous spikelets underground is a phenomenon matched, we believe, by few other grasses. Indeed, Campbell et al. (1983) listed only four genera— Amphicarpum, Chloris, Eremitis, and Paspalum—that have subterranean spikelets. The spikelets in these are “borne on specialized rhizomes” rather than at the base of the culms as in A. scabrivalvis. Such burial, resulting in the complete loss of dispersal from the parent plant, would seem to be an Figure 1. Amphibromus scabrivalvis. a. panicle; b. chasmogamous spikelet from panicle; c. floret from chasmogamous s sptkelec; d, e. cleistogamous spikelets from lowest leaf sheath (underground), f, g. cleistcogamous spikelets from Sth and 6th leaf sheaths. The vertical lines = 5 mm 209 SSS. SS ee ——> ae Le PGS BE ZS ee * — a= a Ohio ae floodplain, US Re 52 at Pond Run Rd, 3 Oct 1978, AWC 18914 (OS). - 4 Co.: mouth of Big Creek, Sect 8, Guyan Twp, 13 Oct 1979, DMS 215 (OS); ey of Ohio River, SW of Chickamauga Creek, Gallipolis, DMS 2158, 11 Aug 1982 (MICH, NY, OS, US). Fa AWRENCE Co.: Ohio River bank, S of Proctorville, 17 Sep 1979, DMS 189 (MU), Sect 7, Rome Ts DMS 2159, 11 Aug 1982 (NY, OS, US). Meics Co.: weedy thickets along Bie River 0.5 mi S of Middleport, 15 Jul 1974, AWC 13509 (KE). Scioro Co.: Moores Lane, Washington ee 25 Sep 1981, DMS & M. Silagy 1026 (OS). Wittiams Co.: Sect. 31, Superior Twp., 18 Aug 1981, AWC 21116 (OS); Sect 5 & 6, Center Twp, 18 Aug PENNSYLVANIA: Berks Co.: Oley Furnace, Oley Twp, 27 July 1969, W. C. Brumbach 6880 (BH). Mireuin Co.: ca 4 mi E of Lewsteu n, 12 Aug 1921, E. M. Gress 5.x. (OKL) 224 VIRGINIA: ALBEMARLE Co.: James River at Warren Ferry, 25 Jun 1977, C. Stevens s.n, (herbarium of C. E. Stevens). Fairrax Co.: bank of Potomac River N of Dead Run, Bradley 69206, 16 Sep 1974 (GMUBF); same (eee oh. 12 Aug 1982, DMS, Thomas & Aber- crombie ae (OS, US) Snare EvOuCAEE) ST VIRGINIA: Cape, Co.: along Hwy 2, 7 mi E of Lesage, 30 Sep 1975, D. Evans i (MUHW),; onal of Ohio River near Guyandotte, | 1 Aug 1982, DMS 2160 (NY, OS). Faverre Co.: New er, 2.5 mi below McCreery, 11 Aug 1941, J. Tosh 1158 pene KANAWHA Co.: ane on Kanawha State Forest, 15 Aug 1966, M. . n, > N side of carol River at Re I-64 bridge, Charleston, 13 Aug 1984, DMS ye 7 3 (E MICH, OS, NCU, WVA, US). Mason Co.: St Re 2, - : Gis pe 17 hs 1970, We : 246 (OS); just N of Cabell - line on St Re 2, g 1982, DMS 2161 (OS, US) ( piacuuee voucher); along RR, Rt 62, 2 mi SE 7 jct “ot St Re 2, . l Aug 1984, DMS & AWC 2368 (MICH, OS, NCU, a. US). Purnam Co.: roadside, jct of 1-64 & US Re 35, Saucheide, 26 Sep 1975, D. Kirk 25 (WVA); RR embankments, 1.25 mi S$ of Robertsburg, 11 Aug 1984, AWC & DMS 23793, (E MICH, OS, NCU, VA, US). RALEIGH Co.: W side of a eee ¥2 mi E of Terry PO., 12 Aug 1984, AWC i DMS 23809 (MICH, OS, NCU, , US); W side of New er at Soe Falls N o Pans 12 Aug 1984, DMS & ie. 2381 (KE MICH, OS, NCU, WVA, US). coe - New River, Hinton, 13 Aug 1963, E. McNeil s.n. (WVA) REFERENCES ALLISON, L. E., W. B. BOLLEN, & C. D. MOODLE. 1965. Total i In: C. A, BLACK a ). Met hods of soil analysis. Agron. Monogr. 9:1346— ARNETT, G. R. 1983. Endangered and threatened wildlife and ie sul aa to review of Gee taxa for listing; proposed rule. Fed. Reg. 48:53640 — 5367 BASKIN, J. M. & C. C. BASKIN. 1984. Environmental conditions required germina- tion of sea sida (Sida pinta’: Weed Sci. 32:786—-791. ATES, D. . Chromosome numbers in the Malvales. II. Miscellaneous counts from ab Dye and Malvaceae. Genres Herb. 11:143— 150. HARD, JR. 1970. Chromosome numbers in the Malvales. II. New or ae, Reet counts relevant to classification in the Malvaceae. Amer. J. Bot. 57:927 — 934, BRADNER, E. 1892. A partial a of the flora of Steuben County. Ann. Rept. Indiana Geol. Surv. 17:135 — 159. BROOME, C. R., J. L. one A. O. TUCKER, & N. H. DILL. 1979. Rare and — viseulee plant species in Maryland. US Fish & Wildlife Service, Newton orner ae Cc ME NT. I. 1957. Studies in Sida (Malvaceae). |. A review of the genus and mono- graph of He sections Malachroideae, Physalodes, Pndinarsn Incanifolia, Oligan- drae, gue Hookerta, and Steninda. Contr. Gray Herb. 180:5—91. CRANFILL, R. & M. E. MEDLEY. 1981. nan on the flora of Kentucky. New and nceestng plants in Kentucky. Rhodora 83:125 — 131. CUSICK, A. & G. M. SILBERH ies 1977. The vascular plants of unglaciated Ohio. Ohio Biol. Sin Bull. N.S. 5(4): 1 3 DEAM, C. C. 1940. Flora of Indiana, ne Dept. Consery., Indianapolis. DMIT RASHKO, P [ 1970. Effect of Cobalt-60 X-rays on Sida bewiabhradite {(L.)} Rusby seeds. Ukrajins’k. Bot. Zhurn. 27:795 — 796. (in Russian, English summary). Zo) 1972. Some Ss concerning the quality of Sida hermaphrodita {0L.)} Rusby seeds. Ukrajins’k. Bot. Zhurn. 29:235 — 236. (in Russian, English summary). oe LO ae Hardness of Sida hermaphrodita seeds. Bjull. Glavn. Bor. Sada 87: ae O09. (in Russian). V. G. NIKOLAYENSKLY, & L. D. NIKOLAYEVSKAYA. 1971. On the influ- ¢ cultivation conditions on the growth and anatomical structure of the Szda ee stem. Rastit. Rusur. 7:227 — 234. (in Russian). EGLEY, G. H. & R. N. PAUL, JR. 1981. Morphological observations on the ae imbibi- tion of water by Sida spinosa (Malvaceae) seed. Amer. J. Bot. 68:1056 — & 982. Development, structure and function of subpalisade ce in water impermeable Sida spinosa seeds. Amer. J. Bot. 69:1402 — 1409. FRYXELL, P A. 1985. Sidus sidarum—V. The North and Central American species of a1. HARVILL, ae ,JR., T. R. BRADLEY, & C. E. STEVENS. 1981. Atlas of the Virginia flora, Pt. 2. Dicotyledons. Virginia Bot. Associates, Farmville, Virginia. ILTIS, TL. H. 1963. beg divica (Malvaceae): Whence came the type? Amer. Midl. Naturalist 70:90 — & S. KAW. ANO. 1964, Cytotaxonomy of Napaea dioica (Malvaceae). Amer. Midl. Naturalist 72:76 — KEIL, D. J. & T. EST oe 1975. Chromosome counts of Compositae from the United States, Mexico, and Guatemala. Rhodora 77:171— 195. KILMER, V. J. & L. T. oe 1949. Methods of making mechanical analysis of soils. Soil Sci. 68:15 MEDVEDE ee E 7 AQ, — fibrous crops in the USSR. Sel’khozgiz. Moscow, Lenin- grad. (in NATIONAL cleerte SOIL SURVEY. 1979. SOILS-5 Form, Sloan Series; Whee- ling Series. Soil Conservation Service, Columbus, Ohio > ROLSTON, M. P. 1978. Water ne seed domaaney Bot. Rev. 44:36 SAVCHENKO, M. I. & P. I. DMITRASHKO. 1973. Seed structure BRS beanie {(L.)} Rusby. Bot. Zhurn. 58:570 — 576. (in Russian). are VSTED, A. 1935. Chromosome numbers in the Malvaceae. I. J. Genet. B1e263=7 296. coi CONSERVATION SERVICE. 1972. Soil survey methods and procedures for collect- ing soil samples. Soil Surv. Inv. Rept. No. 1, U.S. Dept. Agric., U.S. Gove. Printing Office, Washington, D.C. SPOONER, D. M. &G. E HALL. 1983. Sida hermaphrodita (Malvaceae), Virginia mallow, a common eles Ohio J. Sci. 83:8 (abstr. ) THOMAS, , JR. ae Distribution and ecology of Sida hermaphrodita: a rare plant species. Baron 46:51—59. he ee and extinction of a rare plant species, Virginia ma ay ee pate (L.) Rusby) on National Park service areas. Proc. Second Conf. . Res. Natl. Parks. U.S.D.I. 8:60—75 TUTHILL, E H. 1876. Some notes on the flora near Kalamazoo, Mich. Bot. Gaz. A Telit hae WIEGMAN, P G. 1979. Rare and endangered vascular plant species in Pennsylvania. Prepared by the Western Pennsylvania Conservancy in cooperation with U.S. Fish and Wildlife Service. ADDITIONS AND NOTEWORTHY VASCULAR PLANTS FROM ARKANSAS, WITH SOME ECOLOGICAL NOTES STEVE L. ORZELL Arkansas Natural Heritage Commission Little Rock, AR 72201, U.S.A EDWIN L. BRIDGES and $8. LANCE PEACOCK Arkansas Nature Conservancy Little Rock, AR 72201, U.S.A. ABSTRACT Six vascular plant taxa are reported as new to the Arkansas flora (Cirsinm muticum, Liparis loeselit, Pedtcularis lanceolata, Rhynchospora capillacea, Solidago patula var. strictula, nage riddellit) and two noteworthy collections (Buchnera floridana, Scleria verticillata) ¢ presented with brief ecological notes. This paper presents eight new or otherwise noteworthy records of vascu- lar plants collected during 1984 field studies. These collections provide evidence that much remains to be learned about the Arkansas flora. Many counties have not been systematically searched by collectors and offer potential for significant discoveries. Nomenclature follows Kartesz & Kartesz (1980) except for Buchnera floridana and Cypripedium kentuckiense. BUCHNERA FLORIDANA Gandog. (Scrophulariaceae). Ashley Co.: NW of section 1, T18S, R8W, Crossett North 7.5’ Quad, Crossett Prairie, 2 Sep 1984, Orzell 1420 & L. Peacock (APCR, UARK, B). en in a 10 acre remnant coastal plain prairie. Previously reported for Arkansas by Buchholz & Palmer (1926) and Demaree (1941, 1943) but considered unsubstantiated by Smith (1978). Similar to Buchnera america- na, but distinguished by the obscurely 3-veined, lanceolate to oblanceolate leaves, shorter corolla tube (6— 10 mm) and corolla lobes (2—5 mm) (Godfrey & Wooten 1981, Radford et al. 1968, Pennell 1935, Correll & Johnston 1970). Vouchers were verified by Dr. Robert Kral at Vanderbilt University. Cirsium MutTicuM Michx. (Asteraceae). Garland Co.: SW, SW, NW 4 of section 16, T3S, R22W, Pearcy 7.5’ Quad, along Meyers Creek, Ouachita National Forest, 5 Sep 1984, Orze// 1424 (UARK), 20 Sep 1984, Orzell 1427, G. Tucker & L. Peacock (MO). Rare, in a wooded acid seep on SIDA 112): 226-231. 19895. Za saturated muck underlain by gravelly substrate, shaded by Acer rubrum L., Carpinus caroliniana Walt., and Magnolia tripetala L. Associates were Cypripedium kentuckiense C. E Reed, Juncus coriaceus Mackenzie, Liparis loeselti (L.) L. C. Rich., Onoclea sensibilis L., Pedtcularis canadensis L., Rudbeckia fulgida Ait. var. umbrosa (C. L. Boynt. & Beadle) Crong., Senecio aureus L., Smilax bona-nox L., and Thelypteris palustris Schott. Although Ciérsiam muticum is rather wide ranging (Cronquist 1980, Correll & Correll 1975, Godfrey & Wooten 198 1) it is local and sporadic in the southern extension of its range, particularly in Louisiana, Texas, Oklahoma, Missouri and Tennessee. The Arkansas record is the first from the state and from the Ouachita Province (Fenneman 1938) for this northern plant. Nearest collections are isolated occurrences in east Texas, where it is very rare (Correll & Correll 1975), a single collection from southeastern Oklahoma (Taylor & Taylor 1978), scattered records in southeastern Missouri (Steyermark 1963, Orzell 1984), reported from Louisiana (R. Dale Thomas, pers. comm.), and recent collections by E. Bridges in the Western Highland Rim of west-central Tennessee. The in- frequent occurrence of C. muticum in the southern extent of the range seems to be correlated to its fidelity to rather undisturbed seepage wetlands. Liparis LtoeseLu (L.) L. C. Rich. (Orchidaceae). Garland Co.: SW, SW 4, NW % of section 16, T3S, R22W, Pearcy 7.5’ Quad, along Meyers Creek, Ouachita National Forest, 26 Jul 1984, Orze// 1391 (NYS, VDB), 20 Sep 1984, Orzell 1428, G. Tucker & L. Peacock (MO, UARK). Scattered in wooded acid seeps usually covered with the moss, Thu:dium delicatulum (Hedw.) B.S.G. over a seepy gravel substrate. Associates include Acer rubrum L., Athyrium filix-femina (L.) Roth var. asplentotdes (Michx.) Farw., Carex bromoides Willd., Cirsium muticum Michx., Exonymus americanus L., Lindera benzoin (L.) Blume, Magnolia tripetala L., Platanthera clavellata (Michx.) Luer, Ranunculus recurvatus Poir., and Senecio aureus L. This the the first record of Liparts loeselii from Arkansas, disjunct approx- imately 200 miles from populations isolated in the Ozark Plateaus of southeastern Missouri (Orzell 1983, 1984). The Garland County station tn the Ouachita Province of Arkansas represents a significantly disjunct population for this northeastern species, and is the southernmost known occurrence of L. doeseli7, Specimens were verified by Dr. Charles Sheviak at the New York State Museum. PEDICULARIS LANCEOLATA Michx. (Scrophulariaceae). Fulton Co.: NW, SEY, NEM of section 7, T20N, R8W, Salem 15 ’ Quad, 13 July 1984, Orzell 1373 (APCR), 2 Oct 1984, Orzell 1437 (APCR, MO, UARK, VDB). Occasional on quaking sphagnous peat, saturated by cold minerotrophic seepage with Carex lurida Wahlenb., Duichanthelinm 278 scopartum (Lam.) Gould, Expatorinm perfoliatum L., Linum striatum Walt., Oxypolis rigidior (L.) Rat., Parnassia grandifolia DC., Rhynchospora capitel- lata (Michx.) Vahl, Rudbeckia fulgida Ait. var. umbrosa (C. L. Boynt. & Beadle) Crong., and Senecio aureus L. Pediularis lanceolata is primarily a northeastern calcicole with isolated localities in the Ozark Plateaus of southeastern Missouri. The Fulton County collection is the first in Arkansas and a range extension into the state from adjacent southeastern Missouri. RHYNCHOSPORA CAPILLACEA Torr. (Cyperaceae). Baxter Co.: Sections 13, 14, & 15, TI8N, R12W, Norfork Dam South 7.5’ Quad, along streamside of Otter Creek, 10 Jul 1984, Orze// 1353 (APCR, MO, UARK, VDB). Seepage margins of streamsides along calcareous seep-fed streams, with Calamintha arkansana (Nutt.) Shinners, Heleninm sp., Lysimachia quadriflora Sims, and Rudbeckia fulgtda Ait. var. umbrosa (C. L. Boynt. & Beadle) Crong., and on damp weathered dolomite bedrock with Eguisetum hyemale L., Juncus spp. and Vernonia sp. The senior author also has observed Rhynchospora capillacea in Arkansas from streamside seep-fens on the Salem Plateau in Sharp, Stone, and Marion counties, and froma sedge-shrub fen in Marion County. Rhynchospora capillacea is an obligate calcicole in the Ozark Plateaus of southeastern Missouri where it occurs in calcareous seep fens and on moist calcareous (dolomite) outcrops. The Arkansas collections represent the first in the state and extension of the known range southward from Missouri. Vouchers were verified by Dr. Robert Kral at Vanderbilt University. SCLERIA VERTICILLATA Muhl. ex Willd. (Cyperaceae). Sharp Co.: Sec- tion 7, TI8N, R4-5W, Sitka 7.5’ Quad, along Rock Creek, Harold Alexander Wildlife Management Area, 23 Oct 1984, Orzel/ 1557 G E. Bridges (APCR, MO, UARK). Dolomite bedrock along streamside seep fens with other calcicoles, such as Lysimachia quadriflora Sims, Parnassia grandifolia DC., Pycnanthemum virginianum (L.) Durand & Jackson, Rhynchospora capillacea Torr., and Solidago riddellii Frank. In the midwest, where its distribution is local and sporadic, Sclerza verti- cillata requires a constant supply of cold, calcareous, minerotrophic water and a mildly disturbed substrate (Smith 1983). Both ecological requirements are met along Rock Creek, which is permanently fed by cold springs moderating the microclimate, and where scouring flash floods produce a suitable disturbed substrate. Nearest records are from the Ozark Plateaus of southeastern Missouri (Steyermark 1963), where Scleria verticillata is a rare disjunct restricted to fens (Orzell 1984). There is a historical collection from northwestern Ar- kansas, Benton County, Plank s.., undated specimen at MO (Smith 1978, Fairey 1967). 229 Rediscovery of Scleria verticillata in Arkansas from a calcareous seep-fed stream is further evidence that such streams provide a refugium for disjunct taxa in the Ozark Plateaus. SOLIDAGO PATULA Muhl. var. srricruta Torr. & Gray (Asteraceae). Un- ion Co.: Corner of sections 17, 19, & 20, T16S, R14W, Calion 7.5’ Quad, 2 Sep 1984, Orzell 1415, C. Amason & L. Peacock (APCR, UARK). Un- common in artificially open acid seepage slope under powerline with A/nas serrulata (Ait.) Willd., Aster umbellatus PR Mill., scattered Myrica cerifera L., Panicum sp., and Xyris torta Sm. Solidago patula var. strictula is reported as occurring mostly on the coastal plain from Virginia to Florida and west to Texas (Cronquist 1980, Wilhelm 1984). Although widely distributed, the variety is considered infrequent in the Carolinas (Radford et al. 1968), and Louisiana (R. Dale Thomas, pers. comm.). It is rather frequent in southeastern Texas but has only been collected twice in southeastern Oklahoma (Solidago salicina El. in Taylor & Taylor 1978, 1984). The Union County collection is the first from Arkansas and a range extension from adjacent northern Louisiana parishes. SOLIDAGO RIDDELLI Frank (Asteraceae). Sharp Co.: Section 7, TI8N, R4-5W, Sitka 7.5’ Quad, along Rock Creek, Harold Alexander Wildlife Management Area, 23 Oct 1984, Orzell 1555 & E. Bridges (APCR, MO, UARK, VDB). Abundant along a narrow streamside of Rock Creek, a cal- careous spring-fed stream, and in calcareous seep fens surrounding springs. Associates include Lysimachia quadriflora Sims, Parnassia grandifolia DC., Pycnanthemum virginianum (L.) Durand & ih Rhynchospora capillacea Torr., and Scleria verticillata Muhl. ex W1 es riddellii is a rare disjunct, nee to fens in the Ozark Plateaus of southeastern Missouri (Orzell 1984). Populations of S. ridde/liz in southeastern Missouri and the Sharp County location in northeastern Arkansas are several hundred miles disjunct from the main range in the north-central states from Ohio to Minnesota. The Arkansas station is the southernmost record. SUMMARY Buchnera floridana, although previously reported from Arkansas, had not been generally recognized as occurring in the state and is here documented from the Arkansas coastal plain. Three Arkansas state records (Pedicularis lanceolata, Rhynchospora capillacea, Solidago viddellii) with fidelity to fens are generally northern in distribution. All have a disjunct population center in the Ozark Plateaus of southeastern Missouri which ts now known to extend into adjacent northeastern Arkansas. Liparis /oese/i7, with a northeastern 20) distribution, represents a significant disjunct new to Arkansas and to the Ouachita Province, by far the southernmost locality for this orchid. Three additional Arkansas state record plants (Cérsiam muticum, Scleria verticil- lata, Solidago patula var. strictula) although wide ranging are restricted to seepage wetlands with local and sporadic distribution particularly in Ar kansas and several surrounding states. ACKNOWLEDGEMENTS We thank Dr. Robert Kral of Vanderbile University and Dr. Charles Sheviak of the New York State Museum for verification of some vouchers. Dr. R. Dale Thomas of Northeast Louisiana University provided distribu- tion information on Cirsiwm muticum and Solidago patula var. strictula in Lou- isiana, Special thanks to Carl Amason of Calion, Arkansas for sharing with us his botanical expertise on the Arkansas coastal plain. REFERENCES BUCHHOLZ, J. T. and 7 _ PALMER. 1926. . ae to the catalogue of Arkansas cer Trans. Acad. . St. Louts 25:9] CORRELL, D. S. and - - CORRELL. eg ak and wetland plants of south- western he States. Vol. II. Stanford Univ. Press, California. 1777 pp. — D.S. and M. C. JOHNSTON. 1970. Manual of the vascular plants of Texas. xas Research Foundation, Renner. 1881] C ee IST, A. 1980. Vascular flora of the suchen United States. Vol. L. Asteraceae. Univ. “ North Carolina Press, Chapel Hill. 261 pp. DEMARE D. 1941. Norewosthy Arkansas eee I. Proc. Arkansas Acad. Sci. Ee ‘ 9. ———. 1943. A catalogue of the vascular plants of Arkansas. Taxodium 1:1 — 88. FAIREY, J. E. 1967. The genus Sc/eria in the southeastern United States. Castanea FENNEMAN, N. M. 1938. Physiography of the eastern United States. McGraw-Hill es Se aag New York. 714 pp. GODF R. K. and J. W. WOOTEN. 1981. Aquatic and wetland plants of ac United States, Dicoryledons. Univ. of Georgia, Athens. 933 pp. KARTESZ, J. T. and R. KARTESZ. 1980. A sy a mized checklist of che vascular flora of the United States, Canada, and Greenland. Vol. I. The Biota of North America. Univ. of North Carolina Press, Chapel Hill. 498 pp. ORZELL, S. L. 1983 pe on rare and endangered Missouri fen plants. Trans. Missouri Acad. Set. 17: om 1984. dona notes on rare, endangered and unusual Missouri fen plants. Trans. Missouri Aci id. . (in pr PENNELL, EF W. 1935 the oe ae arlaccae al eastern temperate North America. Acad. . Sci. Philadelphia Monogr. 1:1 —65 RADE OR D, ALE, AHLES, and C. . er 1968. Manual of the vascular flora of the C sae tins. of North Carolina Press, Chapel Hill. 1183 pp. SMITH, E. . Anatlas and annotated list of che vascular ue of Arkansas. Student Union cea Univ. of Arkansas, Fayetteville. 592 pp. (Supplements I-IV, 1979 — 1982), Poa | SMITH, W. R. 1983. A range extension of Scleria verticillata in Minnesota, Michigan Bot. 22:27 — 30. TEYERMARK, J. A. 1963. Flora of Missouri. lowa State Univ. Press, Ames. 1728 pp. TAYLOR, R. J. and C. E. TAYLOR. 1978. Additions to the vascular flora of Oklahoma. Sida 7:361 — 368. __.- 1984. Solidago (Asteraceae) in Oklahoma and Texas. Sida 10:223— 251. WILHELM, G. 1984. Vascular flora of the Pensacola region. Ph.D. diss., Southern [linois Univ., Carbondale. 213 pp. + biblio. THE VASCULAR FLORA OF CENTRAL FLORIDA: TAXONOMIC AND NOMENCLATURAL CHANGES, ADDITIONAL TAXA' R. RP WUNDERLIN, B. E HANSEN Department of Biology, ee of South Florida Tampa, FL 33620, D. W. HALL The Herbarium, Department of Natural Sciences Florida State Museum, University of Florida Gainesville, FL 3361 ABSTRACT Fifty-one taxa new to the vascular flora of central Florida, 30 of which are exotics, and 65 nomenclatural or taxonomic changes are reported. When a flora is published, it is only a statement of present knowledge and nota finite product; there are invariably changes, some even before the ink is dry. An excellent example is the report of over 30 additions by Anderson (1984) to the vascular flora of the Florida panhandle (Clewell, in press). The publication of a flora usually has a stimulatory effect resulting in the urge to discover taxa overlooked by the author(s) and to refine certain treatments. This 1s, or should be, one of the intentions of the author(s) of a flora. The recently published vascular flora of central Florida (Wunderlin 1982) has had this desired effect, and some of the results are presented here. We hope this paper will further stimulate others to bring forth their findings. The following includes 51 taxa reported as new to the region. Of these, 30 are exotic species, the introduction of which carries strong implications concerning possibly detrimental changes in the native flora of Florida. Specimens examined or representive specimens and the herbaria in which they repose are cited. Also reported are 65 nomenclatural or taxonomic changes deemed necessary because of recent taxonomic findings. The fami- lies are arranged according to the Englerian sequence. PTERIDACEAE PreRIS MULTIFIDA Poir. This Old World species is pee ane occasionally natural- ized. Citrus Co.: Diddell s.n. (FLAS); E. St. Jobn s.n. (FLAS): R. St. John s.n. (FLAS). ‘This paper is Florida Agricultural Experiment Station Journal Series No. 6192. SIDA 112): 232-244. 1985, 259 TRISMERIA TRIFOLIATA (L.) Diels. This tropical species native to south Florida is probably a chance introduction into Hernando County. Hernando Co.: Mosura s.n. (FLAS). Palm Beach Co.: Farnsworth s.n. (FLAS BLECHNACEAE STENOCHLAENA TENUIFOLIA Moore. A native of Asia and occasionally cultivated, this species was found as an escape at one site in central Florida in the 1930's where it still occurs. It was fs by Lakela and Long (1976) but overlooked a Wunderlin (1982). It 1984 at a ‘ila site. Hillsborough Co.: E. St. John s.n. (FLAS), wee G re 9824 (USF ASPIDIACEAE THELYPTERIS RETICULATA (L.) Proctor. This is the northernmost station for this tropical species in Florida. Lee Co.: Craighead s.n. (FTG). ZANICHELLIACEAE ZANICHELLIA PALUSTRIS L. The inclusion of this species adds the family Zanichelliaceae to the flora. Citrus Co.: Hartman 51 (FLAS); Swindale 1156 (FLAS). CY MODOCEACEAE RINGODIUM FILIFORME Kuetz. = Cymodocea filtformis (Kuetz.) Correll. Leaf form and Ae differences provide oe evidence that Syrivgodinm should be treated as distinct from Cymodocea. Our species is retained in the former genus by den Hartog (1970). POACEAE ERIOCHLOA MICHAUXH (Poir.) Hitchc. var. siIMPSONI Hitchce. This rare, distinc variety is endemic to Collier and Lee counties. Lee Co.: Brumbach 5583 (USF), ea 5788 (FLAS). LEERSIA VIRGINICA Willd. This species ts eae frequent in north Florida and so expected in our area. Marion Co.: Ha// 1354 (FLAS). Osceola Co.: Shuey & Poppleton 152: (USF). Sumter Co.: Wanderlin et al. 6590 (USF). LEPTOCHLOA FILIFORMIS (Lam.) Beauv. This South American species occurs in our area as a weed in a sugar cane field. Palm Beach Co.: Bays s.n, (FLAS). ZIOLA FLUITANS (Michx.) Terrell & H. Robins. ydrochloa caroliniensis Beauv. in is reduced to synonymy under the older name fie the oldest valid epithet ts ae of Michaux (Terrell & Robinson 1974 LUM NICORAE Parodi. This ineteductiod from South America is found locally in Poe and oe roadsides in our area. Hendry Co.: Ha// 618 (FLAS, USF). Orange Co.: Grais s.n. (FL PHARUS LAPPULACEUS Aubl. Study by E. Judziewicz (pers. comm.) reveals that this is the correct name for the Florida material and must replace P. parvifolins Nash which has been misapplied. Acceptance of Piptochaetium as a segregate of Stipa (M. Barkworth, pers. comm.), necessi- = tates the following two changes. PIPTOCHAETIUM AVENACEUM (L.) Parodi = Stipa avenacea L. PIPTOCHAETIUM AVENACIOIDES (Nash) Valencias & Costas = Stipa avenacioides Nash. ROTTBOELLIA EXALTATA L. f. An introduction from tropical Asia, this troublesome weedy grass is becoming increasingly frequent in Florida. Palm Beach Co.: Johnson s.n. 234 (FLAS). Martin Co.: Orsenigo s.n. (FLAS); Bregger 5.n. (FLAS). Hillsborough Co.: Wanderlin 9426 (USF). SeTARIA IrALica (L.) Beauv. This native of Eurasia is cultivated and sparingly natural- ized in Florida. Sarasota Co.: Shuey 2586 USP). CYPERACEAE Carex stipAta Muhl. This common species of eastern North America is now known to extend into the northwestern part of central Florida. Citrus Co.: Burdett s.n. (USF). Dichromena is not readily separated from Rhynchospora and is best treated as a section of the latter (Thomas 1984); the following two nomenclatural changes are necessary. RHYNCHOSPORA COLORATA (L.) Pfeiffer = Dichromena colorata (L.) Hitche. RHYNCHOSPORA LATIFOLIA (Baldw.) Thomas = Dichromena latifolia Baldw RiyNcHospoRA ELLIOTTH A. Dietr. This northern species has been found well into central Florida. Polk Co.: Wheeler s.n. (FLAS) RiYNCHOSPORA FLORIDENSIS (Britt. ex Small) Pfeiffer. Thomas (1984) cites the follow- ing specimen of this Caribbean and south Florida species. Polk Co.: Jennings s.n. (CM, n.v.). ARECACEAE LivistONA CHINENSIS R. Br. This Old World palm is commonly cultivated in Florida and sparingly naturalized. Manatee Co.: Shuey 2589 (USF) BROMELIACEAE = CKIA BREVIFOLIA Bak. A native of Brazil, this species 1s occasionally cultivated in central florida and is locally escaped. Two patches of plants and scattered seedlings were found in a dry disturbed area that was formerly sand pine scrub. Pinellas Co.: Beckner 2002A (USF) COMMELINACEAE COMMELINA CAROLINIANA Walt. Although reported from Florida by Small (1933) Radford et al. (1968), and Ward (1968), its rareness and resemblance to C. erecta obscured its pe = study by R. Faden (pers. comm.) confirmed its presence in central florida, Lee : Hoffman 12 (FLAS). ae Co.: Genelle & Fleming 2207 (USF); Garber s.n. (US). = ‘Beach Co.: Felix s.n. (FLAS, US) COMMELINA NIGRITIANA Bet se var. GAMBIAE (C. B. Clark) Brenan. Robert Faden (pers. comm.) has determined that our plants are best referred to var. gambiae Plants from Highlands County previously sot ain to Tradescantia hir oneal Small are dest considered as variants of T: roseolens (R. Faden, pers. comm.). Tradescantia hirsuticaulis is found to the north of our area and is exc hicléel from our flora. pay AGAVACEAE — AGAVE DESMETTIANA Jacobi. Sea originally native to Mexico, this species is culti- vated in Florida and rarely escaped. The collections listed are cited by Gentry (1982). Lee Co.: Brumbach 7798 (FLAS): Sea 8459 (FLAS). AMARYLLIDACEAE Reexamination of our naturalized Crinum species reveals specimens previously identified as C, amabile Donn should be provisionally determined as follows, pending further studies of this difficult genus (A. Meerow, pers. comm.). 235 Crinum asiaticuM L. A native of tropical Asia and cultivated in Florida where locally escaped. Lee C ate jive van Todd 126 (FLAS, USF). cRINUM LATIFOLIUM L. var. ZEYLANICUM (L.) Hook. f. ex Trimen. A native of Asia, this species has been found as an escape from cultivation along roadsides and in pastures. ae Co.: Wanderlin et al 9494 (USF). Pasco Co.: Hansen & Hansen 9948 (USF) Sumter Wunderlin et al. 9813 (USF). ee TUBISPATHA aes Native to the West Indies, this species is occasion- ally cultivated in Florida and rarely encountered as an escape. Pinellas Co.: Beckner 2620 (USF) —~ ORCHIDACEAE ANACHEILIUM COCHLEATUM (L.) Hoffsgg. var. TRIANDUM (Ames) Sauleda, Wunderlin & Hansen = Eneyclia cochleata (L.) Dressler. A baba a segregate of Encyclia, 1s consid- ered a distinct genus following Pabst et al. (1981). The Florida plants are triandrous and should be recognized at the varietal level. MESADENUS POLYANTHUS (Reichenb. f.) Schlechter =Spéranthes polyantha Reichenb. f Garay (1982) recognizes Mesadenus, a segregate of Spiranthes. ACOILA LANCEOLATA (Aubl.) Garay = Stenorrhynchos lanceolatus (Aubl.) L. C. Rich ex Spreng. Garay (19820 recognizes Sacoila, a segregate of Spiranthes and Stenorrhy- nchos. SACOILA LANCEOLATA (Aubl.) Garay) var. LUTEOALBA (Reichenb. f.) Sauleda, Wunderlin, & Hansen. Materials of this tropical American taxon were previously seen from collections only to the south of our area in Florida. DeSoto Co.: Beckner 2255 (FLAS). Okeechobee Co.: Sawleda & Sauleda 8646 (USF). CASUARINACEAE SUARINA EQUISETIFOLIA L. =C. /etorea L. Casuarina litorea was published in the dissertation of Linnaeus’ student Stickman in 175 . The names published in this work are now specifically rendered invalid according to rene 34, ICBN (Voss 1983). Therefore, the more familiar combination C. eguisetifolia L. (Amoen. Acad. 4:143. 1759) is reinstated. There is still controversy concerning the author citation for both the genus and species Bullock (1960) considers Linnaeus’ description of Casuarina equisetifolia inadequate. Rogers (1982) concurs with this view. However, Friis (1980) considers the description of Linnacus sufficient for valid publication. If the arguments of Bullock are accepted, the correct citations are: Casvarina L. ex Adans. and Casuarina equisetifolia L. ex J. R. & G. Forse. Qu JUGLANDACEAE Carya ALBA (L.) Nutt. ex Ell. This common tree of eastern - S. has been ere in the northern counties. Sumter Co.: Ofer G6 (FLAS). Volusi : Brichard s.n. (FL The mockernut hickory has long been known as cons tomentosa (Poir.) Nutt., ee to confusion in the protologue of Jaglans alba L.. However, the typification of the later by Crantz (Inst. Rei Herb. 1.157. 1766) and subsequent clarifications by Rehder (1945) and Howard and Staples (1983) reveal that Carya alba is the correct name for this taxon. The earliest valid transfer of the epithet into Cayra was that of Elliot (Sketch. Bot. 8. Carol. aN 2:624. 24). Although the description accompanying Elliott's transfer of Linnaeus’ Juglans alba clearly refers to the s i yark hickory (C . ovata (Mill.) K. Koch), the transfer s valid according to article 55.2, ICBN (Voss 1983 ae FAGACEAE Quercus LYRATA Walt. This oak of bottomlands in the southeastern United 236 States extends into our area along the Suwannee River. Levy Co.: Skean 945 (FLAS, USF URTICACEAE Urtica ureNS L. A native of Europe, this species is adventive in our area in vegetable helds. Orange Co.: Riffle s.n. (FLAS). POLY GONACEAE ANTENORON VIRGINIANUM (L.) Roberty & Vautier = Tovara virginiana (L.) Raf. (FI. Ludov. 28. 1817) must replace Tovara Adans. (Fam. Pl. 2:276. 1763) since the latter is rejected in favor of the conserved later ae nee & ae (Prodr. 49. 1794) of the Tovariaceae (Graham & Wood 1965; Vos 3) ERIOGONUM LONGIFOLIUM Nutt var. GNAPHALIFOLIUM Sade = Ertogonum floridanum Small. This taxon is best treated as a variety of the widespread lower great plains species E. longifolium (Reveal 1968). Reevaluation by Nesom & Bates (1984) provides evid best treated at the specific leve POLYGONELLA BASIRAMIA (Small) Nesom & Bates = Polygonella ciliata Meisn. var. basiramta (Small) Horton. a that the following two taxa are LYGONELLA ROBUSTA (Small) Nesom & Bates = Polygonella fimbriata var. robusta (Small) Horton. CARYOPHYLLACEAE STIPULICIDA SETACEA Michx. var. LACERATA James. A study of the variability of Sr/puli- cida in Florida results in the recognition of two varieties (var. sefacea and var. lacerata) for our area (Judd 1983). CERATOPHYLLACEAE aaa itas LUM MURICATUM Cham. = Ceratophyllum echinatum A. Gray—Based on a study of the neotropical species of the genus, the correct name for this species 1s muricatum (Lowden 1978). Purther study by Donald Les (pers. comm.) supports this inter- pretation BRASSICACEAE Coronopus bipyMus (L.) J. E. Smith. A native of Europe, this widespread weed of North America has been collected in Florida to ie north and south of our range and was to be expected. Palm Beach Co.: Correll et al. 51528 (USF). [aie ROSACEAE DucitesNea INDICA (Andrz.) Focke. This native of Asia is widely naturalized in Europe and eastern North America; the following is the southernmost in Florida. Marion Co.: Norman & Buckner 5.n. (USF FABACEAE ALYSICARPUS OVALIFOLIUS (Schum. & Thonn.) J. Léonard. This is the correct name for the common weedy plant in Florida and must replace A/ysicarpus vaginalis (L.) DC. which has been misapplied. A/ysicarpus vaginalis is restricted in Florida to the southernmost coun- ties, outside our range ba ARENICOLUM (Small) Hermann =C, floridanum (Brice.) Lakela. Small's name (FI. SE U.S. 651. 1903) predates Britton’s (Torreya 4:142. 1904) by one year. 2 Adoption of Irwin and Barneby’s (1982) treatment of the New World Cassiinae necessi- tates changes for the central Florida species previously placed in Cassia, now redistributed in Chamaecrista and Senna CHAMAECRISTA FASCICULATA (Michx.) Greene = Cassia chamaecrista L, CHAMAECRISTA NICTITANS (L.) Moench = Cassia nictitans L. CHAMAECRISTA NICTITANS (L.) Moench var. ASPERA (Muhl. ex EIL.) Irwin & Barneby = Cassia niuctitans L. vat. eae (Muhl. ex Ell.) Greene. CHAMAECRISTA PIL( oo RISTA ROTUNDIFOLIA (Pers.) Greene = Cassia rotundifolia Pers. CHAMAECRISTA SERPENS (L.) Greene = Cassza serpens L. SENNA ALATA (L.) Roxb. =Cassta alata I SENNA LIGUSTRINA (L.) Irwin & Barnaby = Cassa ee ce SENNA MARILANDICA (L.) Tae = Cassia marilandica SENNA OBTUSIFOLIA (L.) Irwin & Barneby = Cassra ae L. SENNA OCCIDENTALIS (L.) Link = Cassra occidentalis L. uta (Willd.) Irwin & Barneby var. GLABRATA (Vog.) Irwin & Barneby = Cassia bern Coll. CORONILLA VARIA L. A native of Eurasia, this ground cover is infrequently planted in Florida and rarely escapes. Lee Co.: Brumbach 8449 (FLAS), Brumbach 8623 (FLAS A revision of Crotalaria for Africa and Madagascar (Polhill 1983) necessitates follow- ing two nomenclatural changes Rails CROTALARIA PALLIDA Ait. =C. muconata Desv. CROTALARIA BREVIDENS Benth. var. INTERMEDIA (Kotschy) Polhill =C. sntermedia Kotschy. CORTALARIA ROTUNDIFOLIA (Walt.) Gmel. There is little justification for maintaining var. eee W indler KUMMEROWIA STRATA (Thunb.) Schindler = Lespedeza striata (Thunb.) Hook. & Arn. The acceptance of the genus Kummerowia as a segregate of Lespedeza (Ohashi et al. 1981) necessitates this change. RUTACEAE Cirrus X parapisi Macf. =C. paradisi (L.) Macf. Recent studies by Scora et al. (1982) ube the propasal that this is a hybrid between the sweet orange, C. sinensis (L.) Osbeck, ind the Pummelo, C. maxima (Burm.) Merr. (=C. grandis (L.) Osbeck, fide Scora). EUPHORBIACEAE JATROPHA GOSSYPHFOLIA L. This species of tropical America 1s infrequently cultivated in Florida and locally escaped. The Pinellas County collection was made along the edge of a mangrove stand, conditions similar to the usual habirat for che species in tropical America, Pinellas Co.: Beckner 2627 (USF) SAPINDACEAE SAPINDUS SAPONARIA L. Sapindus marginata Willd. is distinguished from S$. saponarta by its unwinged leaf rachis. We feel the Florida material is best treated as a single species, following Little (1979). RHAMNACEAE ZizipHus CELATA Judd & Hall. This species, possibly extinct, is known from only two collections from Highlands County, the type collected in 1948 (Judd and Hall 1984) and 238 the following made in 1955. Highlands Co.: Brass 25333 (Archbold Biological Station Herbarium). VITACEAE Cissus VERTICILLATA (L.) Nicols. & Jarvis = Cissus sicyotdes L. The deletion of Article 1, ICBN (Voss et al. 1983) rejecting names based on monstrosities necessitates this change (Nicolson and Jarvis 1984). Viris ROTUNDIFOLIA Michx. The northern scuppernong grape is now known to extend into the northwestern part of our area. Hernando Co.: Beckner 2670 (USF) MALVACEAE Hipiscus CANNABINUS L. Native to Africa, this species is occasionally planted and found as an escape. Okeechobee Co.: Beckner 1970 (FLAS). SIDA SANTAREMENSIS Monteiro. A native of Brazil, Argentina, and Bolivia, this species is adventive in our area. This was first reported for North America from central Florida by ee etal. (1984). The following collections were made from a second site in addition to the one from Pee County cited by Fryxell et al. Hillsborough Co.: Wunderlin & Van Hoek 9819; Van Hoek s.n. (USF); Moffler 236 (USE). CISTACEAE LeCHEA PULCHELLA Raf. A species of the Atlantic coastal plain, this was previously known in Florida to the north of our area. Volusia Co.: Hansen & Richardson 6235 (USE). BEGONIACEAE BeEGONIA CUCULLATA Willd. var. HOOKERI (A. DC.) L. B. Smith & Schub. = Begonia semperflorens Link & Otto. According to Hortus Third (Bailey Hortorium 1976), this is the correct name for this taxon. CACTACEAE CEREUS GRANDIFLORUS (L.) Mill. var. ARMATUS (K. Schum.) L. Bens. =Cerens con- iflorus Weingart. This is the correct name for this taxon according to Benson (1982). MYRTACEAE MELALEUCA LINARHFOLIA Sm. this is the second species of this predominantly Australian genus to become naturalized in Florida. Its potential as a noxious weed tree like M. quingu- enervia is unknown. Osceola Co.: Sauleda & Sanleda 8006 (USF). ERICACEAE LYONIA LIGUSTRINA (L.) DC. var. rouiosirLora (Michx.) Fern. Judd’s (1981) mono- graph of the genus indicates that the material in central Florida should be placed in this variety rather than the typical. RHODODENDRON CANESCENS (Michx.) Sweet. This is the southernmost station for this species in Florida. Marion Co.: Judd 3228 (FLAS, USF). PRIMULACEAE SALLIS MINIMA (L.) Krause = Centunculus minimus L. The differences between Centunculus and Anagallis are not sufficent in our opinion to maintain the former as a distinct genus. We are following Godfrey and Wooten (1981) and Ferguson (1972). 259 SAPOTACEAE BUMELIA SALICIFOLIA (L.) Sw. = Dipholis salicifolia (L.) A. Rich. Dipholis ts best united with Bumelia as discussed by Stearn (1968) OLEACEAE LiGUSTRUM SINENSE Lour. Native to China, this commonly cultivated shrub is occasion- ally found as persistent and less commonly as an escape. Hillsborough Co.: Wanderlin et al. 9514 (USF) GENTIANACEAE NyYMPHOIDES CORDATA (Ell.) Fern. This species enters our area from north Florida. Lake Co.: Easterday 279 (FLAS). APOCYNACEAE TABERNAEMONTANA DIVARICATA (L.) R. Br. = Ervatamia coronaria (L.) Stapf. Accord- ing to Leeuwenberg (1976), there is little justification for the splitting of the pantropical Tabernaemontana into segregate genera. ASCLEPIADACEAE MATELEA GONOCARPA (Walt.) Shinners. Drapelik (1970) reports that M. suberosa (L.) Shinners is misapplied to the North American plants. CONVOLVULACEAE IPOMOEA CORDATOTRILOBA Dennst. = lpomoea trichocarpa Ell. This is the correct name for this common Florida species (Manitz 1983). Although the epit thet was originally phenated (“cordato-triloba”) by Dennstedt (1810), according to Article 73.9, ICBN ee et al. 1983) the hyphen should be deleted. IPOMOEA IMPERATI (Vahl) Griseb. =I. sto/onifera (Cyrillo) J. Fl Gmel. La Valva and Sabato (1983) show that I. smperatr is the correct name for this speci POMOEA VIOLACEA L. =I. macranthera Roem. & Schult. Manitz oa shows I. violacea is the correct name for this well known species. (See also Powell et al. 1978; Powell 1979). POLEMONIACEAE PuHiox pitosa L. A highly variable species for which subsp. detonsa (A. Gray) Wherry can not be maintained. VERBENACEAE GLANDULARIA CANADENSIS (L.) Nutt. Common north of here, this species is adventive in our area. Citrus Co.: Arnold et al. s.n. (FLAS); Baltzell 4713 (FLAS); Schmidt A-165 tae Wunderlin 9403 (USF). Highlands Co.: Brass 15267. Seminole Co.: Schallert s.n (FLAS). The aceptance of Phy/a as distinct from Lippia necessitates the following two changes. PHYLA NopIFLORA (L.) Greene = Lippia nodiftora (L.) Michx PHYLA STOECHADIFOLIA (L.) HBK. = Lippia ae , ) Small. LAMIACEAE LEONURUs sIBIRICUS L. This native of central Asia, introduced into North America for its medicinal properties, is occasional in north Florida and was to be expected in our range. Hillsborough Co.: Lindsey & Arcuri s.n. (USF) 240 MENTHA spicata L. Known from only two sites in central Florida and from sterile material, these collections are provisionally place -d. Native to Europe, it is also naturalized in north Florida. Hillsborough Co.: Beckner 2602 (USF), Wanderlin & ule 9826 (USF). SOLANACEAE BRUGMANSIA SUAVEOLENS (Humb. & Bonpl. ex Willd.) Brecht. & J. Presl. This native of Brazil is occasionally cultivated in central Florida and has become locally naturalized Hernando Co.: Wanderlin & Beckner Rae Okeechobee, 1924, Small et al. : AS). PETUNIA X HYBRIDA Vilm County unknown: Eastern shore of Lake = ~ yeiee (Lam.) BSP. The cultivated petunia, locally escaped in Florida, is believed to be a hybrid of P. axi/laris and violacea Lindl. Evidence of ee segregation back to the parental types is frequent in our materials. SALIS WALTERI Nutt. Recent study of our collections by J. Sullivan, University of es indicaces this is che correct name for Florida material previously identified as Physalis viscosa L. In addition, the previously recognized var. e//ioftii (Kunze) Waterfall and var, maritima (Curtis) Rydb. are reduced to synonyr PHYSALIS ANGUSTIFOLIA Nutt. = Physalis viscosa var. elliottit £. glabra Waterfall. This taxon 1s secon as a distinct species by J. Sullivan. a between P. angustifolia and walter! are Common where the two species are sympa SALPICHROA ORIGANIFOLIA (Lam.) Baill. native of southern South America, this species 1s sparingly naturalized in our ares. Citrus Co.: Keating s.n. (FLAS); Weber s.n (FLAS). Volusia Co.: Evans et al. 45513 (FSU) SCROPHULARIACEAE LiMNopHILA INDICA (L.) Druce. This is the first report of this Old World species in Florida. Both this and the cee species L. sessiliflora are grown as aquarium plants (Godfrey and Wooten 1981) which may account for their occasional occurrence in North America. Pinellas Co.: eee 891 (USF). RUBIACEAE GENIPA CLUSHFOLIA (Jacq.) Griseb. = Casasia clusiifolia Jacq.) Urban. If Casasia is submerged into Genipa as is advocated by most workers (e.g. Little 1979), Genipa clustifolia is the correct name for this taxon. ASTERACEAE ACMELLA REPENS (Walt.) L. C. Rich. =Spilanthes americana (Mutis ex L.f.) Hieron. R. Jansen (1981) segregates Acmella trom Spilanthes. Recent study of our collections by Jansen shows that our plants are best referred to Acmella repens (= Spilanthes americana var. repens ae =) . H. Moore). “ONTINALIS Alex. This taxon, previously placed in ane es under A. dumosus, is ee specifically distinct Jones 1984; J. a pers. ¢ . Additional study is needed to determine if this species is distinct from 984). Lee . leonts Brit itton fror ce (Jones l . Lee Co.: Brumbach 7049 (FLAS); Brumbach ae (USF); Brumbach 9131 (FLAS, USE). ASTER SIMMONDst Small. This taxon, Baton is considered eames distinct (Jones | ; J. Semple pers. comm.). The following representative spec een sated . J. Semple: Brevard Co.: Shuey & Poppleton 1554 (USF). Broward pee ‘McC art 11256 (USF). Citrus Co.: Genelle & Fleming 1157 (USE). DeSoto Co.: Fulton 296 (USF). ae Co.: Shuey 2026 (USF). Hendry Co.: Brass 33407 placed in synonymy under A. damosus, tas 24] (USF). Hernando Co.: Cooley et al. 8099 (USF). Hillsborough Co.: Lakela 25607 (USF). Indian River Co.: Wanderlin & Beckner 6414 (USF). Lee Co.: Wunderlin et al. 5383 (USF). Levy Co.: Ray 9690 (USF). Manatee Co.: Lakela 24883 (USF). Martin Co.: Popenoe 1032 oe Palm Beach Co.: McCart 10398 (USF). Polk Co.: Lakela 23726 (USF) ATHERUM PUNCTATUM Cass. A native of tropical America, this species 1s occa- ets ieeaea in Florida and rarely found as an excape. Volusia Co.: Harmon 5.n. (FLAS). DyssoDIA TENUILOBA (DC.) Robins. This western U.S. Plant in cultivation in St Peter- sburg has escaped locally and is spreading along roadsides. Pinellas Co.: Burdett 5.n. (USF), Chayet 148 (FLAS). EcLIPTA PROSTRATA (L.) L. = Eclipta alba (L.) Hassk. Roxburgh (Fl. Ind. 3: 438. 1832) united E. prostrata meer E, ale under E. prostrata predating Hasskarl (Pl. Jav. Rar. 528. 1848) who united the taxa under E. a/ba. (Koyama and Boufford 1981; Voss et al. 1983). Following the publications of Sieren (198 1) and Taylor and Taylor (1983), reexamination of our materials of Exthamia results in recognition of two rather than three species and the following two nomenclatural changes EUTHAMIA GRAMINIFOLIA (L.) Nutt. var. HirTIPES (Fern.) C. & J. Taylor. Materials ee determined as E. /eptocephala (Torr. & Gray) Greene and E. tenuifolia are best placed here. bene ie is excluded from the flora. an TENUIFOLIA (Pursh) Greene = Ewthamia minor (Michx.) Greene. SOLIDAGO ODORA Ait. var. CHAPMANII (A. Gray) Crong. = Solidago chapmanit A. Gra The slight differences between S. odora and S. chapmanii and the number of intermediates in the area of sympatry indicate the latter is best treated as a variety of the former (Cronquist een Sotiva Mutisii HBK. Materials of South American species were priviously misiden- tified as Soliva anthemifolia Juss.) R. x Less., a species not known to occur in central Florida (See Cabrera 1949; Correll & 7 anston 1970; Gandhi and Thomas 1984). SOLIVA PTEROSPERMA (Juss.) Less. This native of South America is a common turf weed in north and west Florida. Lake Co.: Daubenmire G Daubenmire s.n. (USF) ACKNOWLEDGMENTS The authors gratefully acknowledge the contributions of: Loran Anderson, Florida State University; Anthony Arcuri, Environmental Sciences and Engineering; Daniel Austin, Florida Atlantic University, Mary Barkworth, Utah State University; John Beckner, St. Petersburg; Allen Burdett, Florida Department of Environmental Regulation; Davi Crewz, Florida Department of Natural Resources; Donovant and Helen Correll, Fairchild Tropical Garden; Rexford and Jean Daubenmire, Sorrento; Robert Godfrey, Florida State University; Robert Faden, Smith- sonian Institution; Duane Isely, lowa State University; Walter Judd, Unt- versity of Florida; Emmet Judziewicz, Univeristy of Wisconsin-Madison; Donald Les, Ohio State University; David Lindsey, Environmental Sciences and Engineering; Eliane Norman, Stetson University; John Popenoe, Fairchild Tropical Gardens; Donald Richardson, University of South Florida; Ruben Sauleda, Fairchild Tropical Gardens; John Semple, 242 University of Waterloo; Allen Shuey, Bradenton; Gerould Wilhelm, Morton Arboretum. REFERENCES ANDERSON, L. C. 1984. Noteworthy plants from north Florida. Sida 10:295 — 297. BAILEY HORTORIUM. 1976. Hortus third. Macmillan Publishing Co., New York. BENSON, L. 1982. The cacti of the United States and Canada. aa University, BULLOCK, A. A. 1960. The types of some generic names. Kew Bull. 14:40 — 45. CABRERA, A. L. 1949. Sinopsis del genero So/tva (Compositae). Notas Mus. La Plata, Bot. 14(70):123 — 139. erase A. E In press. Guide to the vascular plants of the Florida panhandle. Florida State ape Press/University Presses of Florida, Gainesville. C ORR ,D.S., and M. C. JOHNSTON. 1970. Manual of he vascular plants of Texas. Texas 7. Foundation, Renner. CRONQUIST, A. nase Notes on sie Asteraceae of the southeastern United States. Brittonia 29:217 2): DENNSTEDT, A. W. ‘1810. Nomenclator botanicus. Eisenberg. DRAPALIK, D. J. 1970. A biosystematic study of the genus Mate/ea in southeastern United cee Ph.D. Dissertation. University of North Carolina, Chapel Hill. = US SON, L. E re In T. G. Tutin et al., eds. Flora Europea. 3:28 — 29. Cambridge ersity ne Cambridge. sere " 1980. The ston and date of publication of the genus Casuarina and its type species. Taxon 29:499— 501 FRYXELL, P A., A. KRAPOVICKAS, and D. CREWZ. 1984. Sidus sidarum—lIV. A new oe for Sida in North America, S. santaremensis (Malvaceae). Sida 10:319 — 320. GAND ) oa Pi Anthemideae and Senecioneae ae of Louisiana. Phytologia 56:199 — GARAY, L. A. 1982. A> generic revision of ae “Guiness, Bot. Mus. Leafl. aes GENTRY, H. S. 1982. Agaves of continental North America. University of Arizona Press, Tucson. GODFREY, R. K., and J. W. WOOTEN. 1981. Aquatic and wetland plants of the southeastern er States. Dicotyledons. The University of Georgia Press, Athens. GRAHAM, S. A., and C. E. WOOD. 1965. The . of Po ijsondcens in the southeastern United Sanes J. Arnold Arbor. 46: 91 — HARTOG, C. DEN. 1970. The seagrasses of the world. oa Holland Pu London. HOWARD, R.A., and G. W. STAPLES. 1983. The modern names for Catesby’s plants. J. Arnold Arbor. 64:511— 546. IRWIN, H. S., and R. C. BARNEBY. 1982. The American Cassiinae, a synoptical revi- sion of Leguminosae, tribe Cassieae, subtribe Cassiinae in the New World. Mem. New York Bot. Gard. 35:1—918. JANSEN, R. K. 1981. Systematics of Spilanthes (Compositae: Helianthae). Syst. Bot. a) blishing Co., JONES, A. G. 1980. A classification of the New World species of Aster (Asteraceae). I 239. JONES, A. G. 1984. Nomenclatural notes on a ane (Asteraceae)—II. New combinations and some ieee Phytologia 55:37 8 243 ae W. S. 1981. Monograph of Lyonia (Ericaceae). J. Arnold Arbor. 62:63 — 209, — 436. Pra W. S. 1983. The taxonomic status of Strpulicida filiformis (Caryophyllaceae). Sida JUDD, W. S., and D. W. HALL. 1984. A new species of Ziziphus (Rhamnaceae) from ee oe 86: a =a87 OYAM ., and D BOUFFORD. 1981. Proposal to change one of the examples in ee 57. Taxon “ 504 0D) LAKELA, O., and R LONG. 1976. Ferns of Florida. Banyan Books, Miami. LA VALVA, V., a S: Barto. 1983. Nomenclature and typification of Ipomoea imperati (Coneiniee Taxon 32:110— 132. LEEUWENBERG, A. J. M. 1976. The Apocynaceae of Africa. I: Tabernaemontana L. 1. le eae) remarks to a revision of the species represented in Africa. Adansonia, ser. 392. Sear E. L. 1979. Checklist of the United States tr ( { naturalized). Agricul- ture Handbook No. 541. Forest Service, USDA, Washington. LOWDEN, . 1978. Studies on the submerged genus Ceratophyllum L. in the Neotro- pics. ee Bor. 4:127 — 142. MANITZ, H. 1977. Was ist Ipomoea violacea L.? Feddes Repert. 88:265 MANITZ, H. 1983. Zur nomenklatur einiger Convolvulaceae und pee L Feddes Repert. 94(3-4):173 — 182. E , G. L., and V. M. BATES. 1984. Recvalvaiign of infraspecific taxonomy in Polygonella ee Brittonia 36:37 — 44. NICOLSON, , and C. JARVIS. 1984. Cissus verticillata, a new combination for C. ee Vice Taxon 33:726 — HASHI, R., R. M. POLHILL, and B. G. SHUBERT. 1981. Desmodieae. Polhill and P. H. fae eds. Advances in legume systematics 1: 292 — 300. Press, London and New York. PABST, G. E, J. L. MOUTINHO, and A. V. PINTO. 1981. An attempt to establish the correct statement for genus oe Hoffmgg. and revision of the genus Hormidium Lindl. ex Heynh. Bradea 3:173 — | POLHILL, R. M. 1982. Crotalaria in vee and Madagascar. A. A. Balkema, Rotterdam POWELL, D. A., D. H. HICHOLSON, and D. F AUSTIN. 1978. C ailualaa cate es Jacq. (Convolvulaceae) re-examined. Brittonia a 199 — 202 POWELL, 2 1979. The Convolvulaceae of the Lesser Anois: J. Arnold Arbor. C19 RADFORD, a - , H. E. AHLES, and C. R. BELL. 1968. eee of the vascular flora of the Carolinas. Daivetsiey of North Carolina Press, Chapel REHDER, y 1945. Carya alba proposed as nomen ees J. Arnold) Arbor. 26:482 — 483. REVEAL, J. 1968. Notes on the Texas eriogonums. Sida 3:195 — 205. ROGERS, G. K. ie The Casuarinaceae in the southeastern United States. J. Arnold Arbor 69135 7= 3 SCORA, R. W., et - 1982. ial to the origin of the grapefruit, C. paradisi (Rutaceae). Syst. Bot. 7:170— 17 SIEREN, D. J. 1981. The taxonomy i the genus Euthamia. Rhodora 83:551—579. SMALL, J. K. 1933. Manual of the southeastern flora. University of North Carolina Press, Chapel Hill. STEARN, W. T. 1968. Jamaican and other species of Bumelia (Sapotaceae). J. Arnold Arbor. 49:280 — 289. n ree 244 TAYLOR, C. E. S., and R. J. TAYLOR. 1983. New species, new ean and notes on the elena Galen and Solidago—Asteraceae). Sida | 83. : E., and H. ROBINSON. se) Luziolinae, a new ee of oryzoid grasses. Bail: Torrey Bou Club 101:235—- THOMAS, W. W. gate T . systematics of te section Dichromena. Mem. New York Bot. Gard. 16. — E.G. etal., — is International code of botanical nomenclature. Regnum Veg. 1, Utreche. oe D. B. 1968. Checklist of the vascular flora of Florida, Part 1. Agriculcure Experi- ment Stations Bulletin 726, Gainesville. WUNDERLIN, R. P. 1982. Guide to the vascular plants of central Florida. University of South Florida Press/University Presses of Florida, Gainesville. NOTES DACTYLOCTENIUM AEGYPTIUM (GRAMINEAE) NEW TO BELIZE.—Dactyloctenium aegyptium (L.) Willd. has apparently not been recorded from Belize, as it is not listed among the monocots of Belize by Spellman et al. (Rhodora 77:105 — 140. 1975), nor in the plants of the Belize cays by Fosberg et al. (Atoll Res. Bull. No. 258. 1982). The discov- ery of this widespread, weedy species in Belize is, however, hardly surpris- ing. [recently found D. aegyptivm plentiful in disturbed areas in the central part of South Water Cay, ca 22.4 km SE of Dangriga, Stann Creek District, where it had apparently become established since my floristic survey of the site in 1979-1980. Voucher: Pring/e 2536, 10 Jan 1985 (HAM) (Contribu- tion No. 57 from the Royal Botanical Gardens).—James S. Pringle, Royal Bo- tanical Gardens, Box 399, Hamilton, Ontario, Canada L8N 3H8. ADDITIONAL NOTES ON COREOPSIS—In my recent paper on the Californian Coreopsis (Sida 10:276 — 289), several errors unfortunately were made. These errors were kindly called to my attention by John L. Strother (UC), and I want to thank him for his advice and attention to detail in these matters, for his review of a preliminary version of this note, and for the loan of selected microfiche (932 and 944) of the Candolle Prodromus herbarium. The errors were mostly in typification matters and require the following clarifications and emendations (page numbers refer to my Sida article pagination): p. 282:—The Tyre for Coreopsis calliopsidea (DC.) A. Gray, Bot. Mex. Bound. 90. 1859.—Agartsta calltopsidea DC., Prodr. 5:569. 1836, should be: California, Douglas 49 (HoLoryPE: G-DC, microfiche!; Isoryprs: BM!, K two sheets!). With the G-DC microfiche, lectotypification was unnecessary. — The Tyre for Leptosyne calliopsidea var. nana A. Gray, Synop. Fl. N. Amer., 2nd ed. 1(2):300. 1886 {1884}, should be: California. San Bernardino Co.: “At Mohave Station etc.,” May 1882, Pringle s.n. (LECTOTYPE, here a en GH!; tsoLecrorypes, here desienaed CAS!, F two sheets!, MO!, , UC!). It is more likely that Gray saw the GH specimen than the CAS specimen. p. 283 — The Tver for Coreopsis bigelovit (A. Gray) H. M. Pale Univ. Calif. Publ. Bot. 3:141. 1907.—Pugiopappus bigelovit A. Gray, Pacif. Railr. 4:104. 1857, should be: California. San Bernardino Co. oo the Mohave Creele. in the desert east [west] of Colorado [River], Mar 1854, /. M. Bigelow 5.n. (HOLOTYPE: GH); isotype: NY!). Lectotypification was unnecessary. — The Tyre for Pugiopappus brewert A. Gray, Proc. Amer. Acad. Arts SIDA 11(2): 245. 1985. 246 8:660. 1873, should be: California. Ventura Co.: On dry hills at San Buenaventura, Brewer 241 (HOLOTYPE: GH}; isorypes: K!, UC!). Lectotypification was unnecessary. p. 285 — The section “Euleptosyne” should be section Leptosyne as follows: Section LEPTOSYNE (DC.) O. Hoffm., Nat. Pflanzenfam. 4(5):243. 1894 {1889}. : PE: Coreopsis douglasit (DC.) H. M. Hall. Leptosyne DC., Prodr. 5:531. (as genus). 1836. Leptosyne sect. Gilepiajaed. ce Synop. Fl. N. Amer., 2nd ed. 1(2):299. 1886 84}. Corespsis sect. Exleptosyne (A. Gray) Blake, Proc. Amer. Acad. Arts 49:341. — The oT ay for Bee stillmanit (A. Gray) Blake, Proc. Amer. son Arts 49:342. 3..—Leptosyne stillmanii A. Gray in E. Dur d, J. Acad. Sci. Phil., 2nd Ser. a 1855, should be: California. Hi the oat of the ie Sacramento {River}, Sti//man s.n. (HOLOTYPE: GH!; isotypes: NY two sheets!, F photograph and fragment!). Lectotypification was unnecessary. p. 286—The Tver for Coreopsis douglasii (DC.) H. M. Hall, Univ. Calif. Publ. Bot. 3:140. 1907.—Leptosyne douglasti DC. , Prodr. 5:531. 1836, should be: California, 1833, Douglas 8 eae ype: G-DC, microfiche!: isorypes: B plant C!, K two sheets!, GH plant I!, UC fragment!). With the G-DC microfiche, lectotypification was unnecessar —Edwin B. Smith, eke a Botany & Microbiology, University of Arkan- sas, Fayetteville, AR 7270 SOLIDAGO SPATHULATA DC. VAR. NEOMEXICANA (GRAY) CRONQ. (ASTERACEAE) NEW TO TEXAS—A collection of Solidago saat DC. var. neomexicana (Gray) Crong. (Worthington 3366, SMU, UTEP) from the Davis Mts., Jeff Davis Co., Texas, represents the first collection of this essentially Rocky Mountain oe of goldenrod from the state. The nearest populations occur approximately 275 km to the north- west in the Sacramento Mts. of Otero and Lincoln cos., NM. The locality is on the forested crest of the mountain 1.6 km ESE of the top of Mt. Liver- more (30°37'40"N-104°09'30"W), at 2285 m elev. Ponderosa pine, Pinus ponderosa Laws., is common on this ridge, but the forest is complex due to edaphic peter and includes Pinus strobiformis Engelm., P. cembroides Zucc., and oaks often growing in close proximity to one another. I would like co thank Mr. Darrell E. Ward for making the preliminary determina- tion.—Richard D. Worthington, Department of Biological Sciences, The Univer- sity of Texas at El Paso, El Paso, TX 79968, U.S.A NESTRONIA UMBELLULA RAE (SANTALACEAE) NEW TO MISSISSIPPI—Nestronia umbellula Raf. is a deciduous, dioecious, clonal shrub which parasitizes the roots of pines and hardwoods. Previous records show a scattered distribution in the Piedmont, Coastal Plain, and Cumber- land Plateau from southern Virginia to Alabama and Tennessee (Kral 1983: Horn & Kral 1984). During the course of field work to determine woody plant distribution in Mississippi’s upper Pearl River Basin, I located a population of N. wmbellula resulting in the first report of this monotypic genus from the state. The Mississippi location extends the western range of the species by ca 230 Four clones of N. wmbellula were found along a 0.4 km stretch of MS Hwy 19. The largest consisted of hundreds of shoots occupying a thinly wooded upland area of ca 1800 m*. Two clones were found on each side of the highway with each extending from the edge of the ROW into the adja- cent upland woods. All shoots were sterile with the exception of a few per- sistent peduncles from male inflorescences on a few shoots. Souls at the sites are sandy and consist of the Smithdale-Ruston and Lauderdale-Arundel associations (Galberry 1981). Future work will be directed toward deter- mining the sex of each clone as well as toward looking for additional sites in the area. Collection — are as follows: MISSISSIPPI. Neshoba Co.: ca 9.0 mi NW of Philadel- phia (TI2N RIOE S10 NE4) side of Hwy - ca 2.7 mi SE of Winston Co. line sandpit on sandy upland ridge. Up to 1.0 m shrubs, most covered with whitish powdery mildew, no fertile plants found, locally nae under Overcus incana, Q. margaretta, ae Pinus nee 1 Jul 1985, Sth 1408 (BE); same locality 8 Aug 1985, ao 1484 IBE —Timothy E. Smith, Institute for Botanical Exploration and Department of Bio- gs Sciences, Mississippi State University, Muassissippi State, MS 397062 REFERENCES GALBERRY, H. S. 1981. Soil survey of Neshoba County, Mississippi. Soil Conservation and Forestry Experiment Sta- — Service in cooperation with the Mississippi Agricultura tion. HORN, D. D. and R. KRAL. 1984. Nestronia umbellula Rat. (Santalaceae), a new state record for Tennessee. Castanea 49(2):69 — 73. KRAL, R. 1983. A report on some rare, threatened, or endangered forest-related vascular plants of the south. Vol. 1: Isoetaceae through Euphorbiaceae. USDA Forest Service Technical Publication R8-TP2, Atlanta. ALETRIS FARINOSA, CYPERUS DIFFORMIS, AND CYPERUS PILOSUS NEW FOR FLORIDA.—Since recent reports by Anderson (1984) and Burkhalter (1984), three more plants collected in Escambia SIDA 11(2): 247. 1985. 248 County, Florida, have, upon examination of exsiccatae at FLAS and FSU, been determined to apparently be new additions to the vascular flora of Florida. These are noted below. ALETRIS FARINOSA L. Escambia Co.: near Barth, E of RR tracks at cross- ing near abandoned Bickerstaff Brickyard, 6 May 1978, Burkhalter 5835 (FLAS, UWEFP); S of McDavid, E of U.S. Hwy 29 ca 200 ft S of jct with Pine Barren Rd, roadside, 18 Apr 1982, Burkhalter 8163 (UWEFP); Barrineau Park, N side of Co Rd 196 ca “% mi E of jct with Co Rd 99, 28 Apr 1984, Burkhalter 9332 (UWFP); N of Barth, ca 0.2 mi N of Cotton Lake Rd along E side of L & N RR tracks, 19 May 1984, Burkhalter 9377 (UWEFP). CyPERUS DIFFORMIS L. Escambia Co.: near Pensacola, W side of Pensa- cola Blvd (U.S. Hwy 29) just N of Hill-Kelly Dodge, in water-filled roadside ditch, 18 Nov 1983, Burkhalter 9166 (FLAS, FSU, UWEFEP) {De- termined by R. K. Godfrey, FSU}; near Pensacola, S side of Nine Mile Rd, W of jct with Holsberry Ln, roadside ditch, 17 Nov 1984, Burkhalter 9813 (FLAS, FSU, MOR, UWFP),. Cyperus pitosus Vahl. Escambia Co.: near Pensacola, E side of Fairfield Dr just N of jct with Hestia Pl, roadside ditch, 12 Oct 1983, Burkhalter 9099 (FLAS, FSU, UWFP) [Determined by R. Kral, VDB].—James R. eee Herbarium, University of West Florida, Pensa- cola, FL 32514, U.S REFERENCES ANDERSON, L. C. 1984. Noteworthy plants from north Florida. Sida 10:295 — 297. BURKHALTER, J. R. 1984. Additions to the vascular flora of Florida. Castanea 49: 180 — 186. A WHITE-FLOWERED FORM OF SPIGELIA MARILANDICA L. (LOGANIACEAE) NEW TO TENNESSEE—A unique color form of Spigelta marilandica L. (Indian Pink) was discovered in 1968 growing in a suburban woodlot in Chattanooga, Hamilton Co., Tennessee. This new form has a corolla that is white on the outside, which contrasts with the scarlet outside of the typical form. Both forms have a yellow throat. The new form also differs from the typical Indian Pink by having a whorl of three leaves on some stems, but opposite leaves on other stems. Observa- tions in the summer of 1984 of four stems producing white flowers revealed that all four stems grew from the same rhizome, which had 24 stem scars from previous seasons. Three of these stems had whorled leaves and one stem had opposite leaves. Pollen samples of the white form averaged about SIDA 11(2): 248. 1985, 249 95% fertility as determined by aniline blue in lactophenol solution. A description of this form is: SPIGELIA MARILANDICA L., forma eburnea Van Horn and Freeman, forma nov. Corolla alba extra, luteola intra; folia insolenter verticillaria, interdum opposita. Spigelia marilandica forma eburnea differs from forma marilandica by its corollas being white on the outside and by most stems having leaves in whorls of threes. Tyre: TENNESSEE. Hamilton Co.: growing among oaks, hickories, and dogwoods in lot adjacent to 3116 Lockwood Drive, Chattanooga, elev. 260 m, 19 May 1982, John Freeman |. (HOLOTYPE: NCU) —Gene §. Van Horn and John R. Freeman, Department of Biology, The Univer- sity of Tennessee at Chattanooga, TN 37403, U.S.A NEW COMBINATIONS IN ERIGERON (ASTERACEAE)—In his studies of Aster, the second author recently examined types of A. procumbens Houstoun ex P Miller and A. scaberrimus and found them to be the same species as Erigeron myrionactis Small and E. /ongipes DC., respectively; the names in Aster also are the earliest for these taxa. New combinations are proposed below, and the remainder of the nomenclature for species 1s presented. The first species occurs along the Gulf Coast from Mississippi through Louisiana and Texas and into Mexico as far south as the vicinity of the city of Veracruz. The second species is highly variable and occurs in Mexico from Veracruz westward to Durango and Nayarit and southward through Chiapas into Central America as far as Nicaragua. ERIGERON procumbens (Houstoun ex P. Miller) Nesom, comb. nov. Aster procumbens pera ex P Miller, Gard. Dict. ed. 8, Aster no. 32, plat fig. 176 pe: MEXICO. [Veracruz], “growing in plenty in the sandy a seat e€ ee 1729, W. 0 toun s.n. (HOLOTYPE: BM!). Erigeron scaposus DC. “var. ? cunetfolins’ A. Gray, Proc. Amer. Acad. Arts 16:94. 1881. Erigeron repens A. Gray, nom. nov., Syn. Fl. N. Amer. I(2):217. 1884; non Weddell, 1855. Erigeron myrtonactis Small, nom. nov., Fl. SE U.S. 1229. 1903. Tyee: UNIT- ED STATES. Texas, [Galveston Co.], coast of Galveston, Nov 1842, F ee J.n. (LECTOTYPE, designated here: GH!; lsorecrotypr: MO!). In the ae ription of Es oe var. cuneifolius, Gray referred to “Texas, sandy seashore” but simply cited rlier publications for reference to specimens. Two collections at GH Lindi sn. and Wright s.n., were labeled by Gray as L. waposus var. cunetfolins and ater annorated by him as E. repens as vouchers for the Synoptic Flora. In the latter treatment, Gray Neer to var. cimetfolins as a synonym of E repens and cited SIDA 112): 249, 1985 250 collections by Berlandier, Drummond, Wright, and Lindheimer. Because the label data on the Lindheimer collection specifically refers to the Texas coast, it is chosen as the lectotype. Erigeron repens A, aks var, oir aces: Proc. Amer. Acad. Arts 41: ee 1905. Type: MEXIC ere , Lizardo, Dr. Wawra 276 (Greenman cited sheets of this ac GH and B; fs G : ee oe not be located and the B sheet _ not been examine ERIGERON Sey (Less.) Nesom, comb. nov. Aster scaberrimus Less., Linnaea 5:14 1830. Tyee: MEXICO, [Veracruz], “in sylvaticus a ve Jalapam,” 28 Aug 1828, C.J. W. Schiede 308 (HOLOTYPE: 2; ISOTYPES: HAL!, fragments and tracing of B -GH!). = is longipes DC., Prodr. 5:285. 1836. Type: MEXICO, W. FE Karvir aSRY SP (HOLOTYPE: M, fide McVaugh, ei fragment of type at G-DC, microfiche!) Erigeron scaposus DC., Prod. 5:287. 1836. Type: MEXICO, “circa Tolucam a 1834, G. Andrienx 277 (HoLtoryer: G-DC, microfiche!, G H-phoro!; tsorypr: GH!). Erigeron oe DC. 8B fatifolium DC., Prodr. 5:287. 1836. {Erigeron macranthus : Pas ip. ex Klatt, pro syn. sub. E. scaposus B latifolium (see comment following 7 below), Leopoldina 20:91. 1884; non Nuttall, 1840.} Typr: MEXICO, “environs de [Cd.} Mexico,” 20 Jun 1827, J. L. Berlandier 375 (HOLOTYPE: G-DC, mucrofiche!, GH-photo!; tsorypr: US!). Erigeron as DC., Prodr. 5:289. 1836; non Badillo, 1946. [Erigeron miradoris Schultz- Bip. ex Klatt, pro syn. sub E. affine, Leopoldina 20:91, 1884.] Type: MEXICO, “environs de [{Cd.] Mexico,” Jun 1827, J. L. Berlandier 522 (HOLOTYPE: G-DC, fide McVaugh, 1974, microfiche!; isorype: B, pencil drawing ex herb. Klatt-GH!). Klace clearly recognized the synonymy of Schultz’s names with the earlier ones of De Candolle. We believe his publication of these epithets represented the use of herbarium names aoe lied by Schultz to collections made by Schiede. It is possible, however, that Schultz already had va ie published them in one of Linden’s diffc cult-to-find ae catalogs (see Stafleu and Cowan 1981). Erigeron ortzabensis Schultz-Bip. ex Klatt, Le ypoldina 20:91. 1884. Type: MEXICO, Veracruz}, “Pic Orizaba,” 11,000 ft, Leebmann 503 (HOLOTYPE?, drawing ex herb. Klatt at GH!). ao —Guy Nesom, Dept. of Biology, Memphis State University, Memphis TN 38152, U.S.A. and Dept. of Botany, University of Texas, Austin TX 78712, U.S.A. and Scott Sundberg, Dept. of Botany, University of Texas, Austin TX PoFil, TS A REFERENCES McVAUGH, R. 1974. ia and taxonomic notes on Mexican Compositae. Rhodora 74:495 — STAPFLEU, FE A. and R. 7 OWAN. 1981. Taxonomic literature. Volume HI: Lh-O. Junk Publishers, The ran SIDA D2)! 250). 1985, DOCUMENTED PLANT CHROMOSOME NUMBERS 1985: 1. MISCELLANEOUS COUNTS FROM ONTARIO AND QUEBEC JERRY G. CHMIELEWSKI Department of Biology, University of Waterloo, Ontario, CANADA, N2L 3G1. Chromosome number determinations for individuals predominantly from Ontario and representing 11 genera are reported: all support previ- ously published counts for the respective taxa. The need for documentation of such reports has been stated many times (e.g., Strother 1972; Semple et al. 1983). I hope that these contributions will assist others in accomplish- ing their goals. MATERIALS AND METHODS Chromosome counts were made from freshly prepared material follow- ing the procedures of Chmielewski and Semple (1983). Root-tip cells were taken from transplanted material for plants collected in the wild. Mitotic metaphase-stage cells were examined. Voucher specimens were deposited in the Herbarium of the Department of Biology, University of Waterloo (WAT). In citation, Chmielewski is abbreviated to C. RESULTS BALSAMINACEAE IMPATIENS CAPENSIS Meerb. 27 = 20. CANADA. ONrTario. Waterloo Reg. Mun.: Univ. Waterloo, Laurel Creek, C. 1207. BERBERIDACEAE PoODOPHYLLUM PELTATUM L. 27 = 12. CANADA. Onrario. Waterloo Reg. Mun.: Univ. Waterloo, North Campus, C. 1/159. CAMPANULACEAE LOBELIA SPICATA Lam. 27 = 14. CANADA. Onrario. Oxford Co.: East Zorra Township, 4.9 km W of Cassel, C. 1645. SIDA jot. ToS). COMPOSITAE SOLIDAGO ALTISSIMA L. 27 = 54. CANADA. Onrario. Huron Co.: 5 km NE of Ethel, C. 1659, SOLIDAGO BICOLOR L. 27 = 18. CANADA. Onrario. Haldimand- Norfolk Reg. Mun.: Seneca Township, Abbey Rd., 3.4 km N of Hwy 54, ee es SOLIDAGO CAESIA L. 27 = 18. CANADA. Onrario Frontenac Co.: Olden Township, Sharbot Lake, C. 2/65, Lambton Co.: Bosanquet Town- ship, Lake Rd., S of Grand Bend, C. & Ringivs 1982. Waterloo Reg. Mun.: North Dumfries Township, River Rd., 0.6 km N of Dumfries North 15. Ce 2729. SOLIDAGO GIGANTEA Ait. 27 = 36. CANADA. Onrario. Lennox- Addington Co.: Kaladar Township, 6 km E of Kaladar, C. 2182. Northumberland Co.: Seymour Township, E of Cambellford, C, 2145, SOLIDAGO HispIDA Muhl. 27 = 18. CANADA. Onrario. Frontenac Co.: Hinchinbrooke Township, 0.8 km S of Parham, C. 2164. SOLIDAGO JUNCEA Ait. 27 = 18. CANADA. Onrario. Perth Co.: Blanshard Township, N of Townline Rd., C. 1893. SOLIDAGO PTARMICOIDES (Nees) Boivin. 27 = 18. CANADA. Quesec. Timiskaming District: Ile du College, 9.6 km NW of Ville Marie, C. & C. 1287. SOLIDAGO RIDDELLU Frank. 27 = 18. CANADA. Onrario. Kent Co.: Dover Township, Ferry crossing south of Wallaceburg, C. & Leeder 2110. SOLIDAGO RUGOSA Mill. 27 = 18. CANADA. Onrario. Haldimand- Norfolk Reg. Mun.: Townsend Township, 3.7 km S of Waterford, C. 1927. Huron Co.: Morris Township, 3.8 km NW of Brussels, C. 1664. Kent Co.: Orford Township, Orford Con. 12, E of Moravian Indian Re- serve, C. & Leeder 2087. Kent Co.: Zone Twp. Zone Sd. Rd. 5, 0.5 km N of Zone Con. 7, C. & Leeder 2095, Lennox-Addington Co.: Kaladar Town- ship, 6 km E of Kaladar, C. 2/81. Middlesex Co.: Caradoc Township, intersection of Caradoc Con. 7 and Caradoc Sd. Rd. 4, C. 2278. Northumberland Co.: Percy Township, Warkworth Hill, C. 2/42. Northumberland Co.: Seymour Township, 2.6 km E of Petherwick Corners, ©. 2150. Simcoe Co.: Tecumseh Township, intersection of Simcoe Rd. 10 and Hwy 9, C. & Ringins 2012. Wellington Co.: Puslinch Township, Puslinch Con. 4, | km E of Rozell Rd., C. 2047. SOLIDAGO RuGosa Mill. 27 = 36. Kent Co.: Camden Township, 5.6 km SE of Wabash, C. & Leeder 2102. SIDA T1(2). 252. 1985, 290 SOLIDAGO SQUARROSA Muhl. 27 = 18. CANADA. Onrario. Nipiss- ing District: 24 km E of Whitney, C. G C.1581. SOLIDAGO ULIGINOSA Nutt. 27 = 18. CANADA. Ontario. Nipissing District: Hunstville, C. & C. 1578. Peterborough Co.: Smith Township, NW of Peterborough, C. 2177. TANACETUM VULGARE L. 27 = 18. CANADA. Onrario. Brant Co.: Onondaga Township, intersection of Hwy 54 and Brant Co. 22, C. 1566. Waterloo Reg. Mun.: St. Jacobs, C, 1569. LILIACEAE CLINTONIA BOREALIS (Ait.) Raf. 27 = 28. CANADA. ONrario. Waterloo Reg. Mun.: Univ. Waterloo, Main Campus, C. 1/1/60. ERYTHRONIUM AMERICANUM Ker. 2” = 24. CANADA. Ontario. Waterloo Reg. Mun.: Univ. Waterloo, Main Campus, C. 1158. POoLYGONATUM BIFLORUM (Walt.) ELL. 27 = 30. CANADA. Onrario. Waterloo Reg. Mun.: Conservation Drive, 0.5 km W of Erbsville, C. 1174 TRILLIUM, GRANDIFLORUM (Michx.) Salisb. 27 = 10. CANADA. Ontario. Waterloo Reg. Mun.: Univ. Waterloo, Main Campus, C. 1/6] PLANTAGINACEAE PLANTAGO MAJOR L. 27 = 12. CANADA. Ontario. Waterloo Reg. Mun.: Untv. Waterloo, North Campus, C. /209 RANUNCULACEAE RANUNCULUS REPENS L. 27 = 32. CANADA. Onrario. Waterloo Reg. Mun.: Hwy 24, 6.0 km SW of Cambridge, C. 1162 ACKNOWLEDGEMENTS This work was funded by University of Waterloo and Ontario Graduate scholarships to JGC and Natural Sciences and Engineering Research Coun- cil of Canada operating grants to Dr. J. C. Semple. REFERENCES CHMIELEWSKI, J. G. and J. C. SEMPLE. 1983. The cytogeography of Aster lanceolatus. II. Cycoecology in southern Ontario. Canad. - Bot Glele7o— PLE, a C.J. SEM G. CHMIELEWSKI, and C. CHINNAPPA. 1983. Chromosome r determinations in Aster L. Compost with comments on pi a See and chromosome morphology. Amer. J. Bot. 70:1432— 14 Soy a J. L. 1972. Chromosome studies in western North pane eee mer. J. Bor. 59:242 — 247. SIDA 11(2). 253. 1985. _ re S DA CONTRIBUTIONS TO BOTANY ~ VOLUME 11 NUMBER 3 MAY 1986 CONTENTS Lectotypification of Lupinus subcarnosus and L. texensis (Fabaceae). B.L. Turner and Jean Andrews, 25) Observations on the vegetation of the Mosquitia in Honduras. Andre F. Clewell. 258 Systematics of Nama (Hydrophyllaceae): comparison of seed morphology of sects. Arachnoidea and Cinerascentia with five species of Eriodictyon and Turricula parryi. John D. Bacon, Louis H. Bragg, and Gary L. Hannan. 271 Vegetative apomixis in Muhlenbergia repens (Poaceae: Eragrostideae). Clifford W. Morden and Stephan L. Hatch. 282 Anisacanthus quadrifidus sens /ato (Acanthaceae). James Henrickson. 286 Taxonomy of Flyriella (Asteraceae- Secs. M.M. Baker and B.L. Turner. 300 Distribution of Portulaca oleracea L. (Portulacaceae) subspecies in Florida. Avinoam Danin and Loran C. Anderson. S18 Some observations on Pinus glabra Walter (Pinaceae). Richard Stalter and Steve Dial. 52) A new species of Wallenia (Myrsinaceae) from Haiti. Walter 8. Judd. a2 Solidago (Asteraceae) of limited distribution in the central United States. Constance E.S. Taylor and R. John Taylor. 334 The rediscovery of Potamogeton floridanus Small (Potamogetonaceae). Gerould S. Wilhelm and Robert H. Mohlenbrock. 340 Recent collections and status of Lesquerella lyrata Rollins (Cruciferae). David H. Webb and Robert Kral. 347 NOTES. A recombination in ee L. (Saxifragaceae). 352.—Monerma cylindrica (Poaceae: Monermeae) new to Texas. REVIEWS 353 US ISSN 0036-1488 SIDA, CONTRIBUTIONS TO BOTANY Founded by Lloyd H. Shinners, 1962 Publisher Wm. F. Mahler SMU Herbarium Dallas, Texas, 75275 Editor Associate Editor Barney L. Lipscomb John W. Thieret SMU Herbarium Northern Kentucky University Dallas, Texas, 75275 Highland Heights, Kentucky, 41076 Guidelines for contributors are available upon request. Subscription: $10.00 (U.S.) per year, numbers issued twice a year. © Sida, Contributions to Botany, Volume 11, Number 3, pages 255 — 355. Copyright 1986 by Wm. E Mahler LECTOTYPIFICATION OF LUPINUS SUBCARNOSUS AND L. TEXENSIS (FABACEAE) B. L. TURNER and JEAN ANDREWS The University of Texas, Austin, TX 78713, U.S.A. Shinners (1953) correctly surmized that the name of our then and only state flower of Texas, Lupinus subcarnosus Hook., was perhaps typified by two or more discordant elements, noting that “It is possible that Hooker had more than one species among the collections...” upon which he based his name. In connection with a forthcoming book on the bluebonnets of Texas by the junior author (Andrews 1986), all collections of the Texas bluebonnets housed at The Royal Botanic Gardens, Kew, England (where Hooker worked), were borrowed so as to resolve any possible nomencla- tural ambiguities with respect to the application of the correct scientific names of both Lupinus subcarnosus and L. texensis, the two most abundant bluebonnets of central Texas. Lupinus subcarnosus was first described by Hooker in the Botanical Maga- zine in 1835. His description was accompanied by a colored plate (t. 3467), this being a fairly accurate drawing of what has long been accepted as the common clay-land bluebonnet of the more western portions of central Tex- as. Shinners (1953) notes that “despite the inaccurate illustration...his description leaves no doubt as to the application of the name sabcarnosus.” Nevertheless, examination of the possible type material available to Hooker shows that the only two specimens which he cited in his protologue of this species are, indeed, discordant elements. The first cited collection in the protologue of Lupinus subcarnosus, a specimen from Bexar County, Texas made by Berlandier in 1828, is actually a specimen of what has long been called L. texensis Hook. The second and only other cited specimen is that of Drummond collected “between Brazoria and San Felipe” in 1835. This latter collection is what most workers have long called L. swbcarnosus. The original description, as noted above, 1s ac- companied by a hand-colored plate. The plate itself does not match the description. Apparently Hooker described L. sabcarnosus largely from the pressed material at his disposal, the drawing having been rendered by a staff artist from plants of what Hooker subsequently described as L. texensis. When he described the latter species in the same year he was clearly befuddled by the very similar illustrations for he comments “Much SIDA 3) 2952297 1 980: 256 and closely as this plant resembles Lupinus subcarnosus figured at tab. 3467, it nevertheless appears to me to be really distinct.” Which it is! The problem ts simply that Joth of the illustrations rendered are of the same species, L. fexensis. Under Article 7 of the International Code of Botanical Nomenclature, where a holotype has not been designated for a given taxon, a lectotype must be selected from among the syntypes cited or examined by the origi- nal author. This is especially critical where discordant elements make up the syntypes. Happily, in this instance, we are obliged to lectotypify Lupinus subcarnosus by the aforementioned Drummond collection, which is the only such specimen in the Hooker Herbarium that is annotated by Hooker himself. In his description of Lupinus texensis, which was also published in the Botanical Magazine of 1835, this too accompanied by a colored plate (t.3492), Hooker does not cite a specimen but rather merely notes the species to occur in Texas, citing specifically the city of San Felipe, which is in Austin County and about which both L. swbcarnosus and L. texensis may be found to this day. Three herbarium sheets of L. texensis are found in the Hooker Herbarium housed at Kew. Two of these bear Drummond numbers 143 of his third collection made in 1835. The other sheet also bears this Drummond collec- tion, along with a Lindheimer collection made in 1847. From among these we have selected the Royal Botanic Gardens specimen number 2 (penciled loan number) as the lectotype. This sheet has both flowering and fruiting material of the species concerned and, in addi- tion, upon this is a handwritten notation, presumably by Hooker, which reads, “similar to 142, but different.” We selected Drummond 142, as noted above, as the lectotype of Lupinus subcarnosus. In summary, Hooker's descriptions of Lupinus subcarnosus and L. texensts were accompanied by colored illustrations of the same taxon. This has caused some confusion with respect to the correct application of the names concerned. Study of the types and protologues of both species reveals that L. subcarnosus is correctly applied to the more eastern sandy-land bluebonnet and that L. ftexensis applies to the more widespread, more western, Clay-land bluebonnet. It is altogether fitting that in this, our sesquicentennial year, the scien- tific names of our two most common official state flowers (Lupinus spp., cf. Andrews 1986), both described in the year 1835, can now be said to rest upon solid typifications. Za7 ACKNOWLEDGEMENTS We are grateful to Dr. Arthur Bell, Director, The Royal Botanic Gardens, Kew for the expeditious loan of the Hooker materials concerned REFERENCES ane re ae The Texas bluebonnets. Univ. of Texas Press, Austin. (In press). SHINNER 1953. The bluebonnets (Lupinus) of Texas. Field and Lab. eae OBSERVATIONS ON THE VEGETATION OF THE MOSQUITIA IN HONDURAS ANDRE FE CLEWELL A. FE Clewell, Inc., 1345 University Parkway Sarasota, Fl 34243 ABSTRACT Species lists and vegetational descriptions are given for sedge savannahs (primarily), colluvial creek swamps, and pine-oak-nance woods of the Mosquitia region along the east- ern Caribbean coast of Honduras. This vegetation is contrasted with that of adjacent tropi- cal wet forests. INTRODUCTION La Mosquitia is a lowland extending for about 480 km along the Carib- bean coast from Cape Cameron, Honduras, nearly to Bluefields, Nicaragua. Approximately three-fourths of this area is characterized by savannahs. Savannah vegetation consists of open grassy stands of Pinus caribaea Morelet on the better drained sites and by low sedge prairies on the wetter sites. The vegetation of these Miskito pine savannahs, as they are sometimes called, contrasts markedly with the broadleaved tropical wet forests of all other regions of the Caribbean coast of Central America, ex- cept Beliz, where savannahs also occur. The region is sparsely inhabited, mainly by Miskito Indians. The purpose of this paper is to describe the vegetation of several plant communities, primarily the sedge savannahs, the colluvial creek swamps traversing these savannahs, and pine-oak-nance woods that are transitional between pine savannahs and tropical wet forest. The flora of the tropical wet forests is described briefly as it occurs at the inner edge of the Mosqu- itia and along an alluvial river within the Mosquitia. PREVIOUS STUDIES Carr (1950) briefly characterized this region, noting its foundation of marine Pleistocene deposits. Arnold (1954) described the geology, soils, and physiography of the Honduran Mosquitia and related his observations to recent changes in sea level. He gave few climatic data and noted the gross aspects of the vegetation. Parsons (1955) described the Mosquitia of both Honduras and Nicaragua with emphasis on the history and potential of land use. He included a map of the entire region and gave information on SIDA 11(3):258— 270. 1986. 259 climate, soils, fire, and vegetation. Radley (1960) described the climate, soils, geology, and physiography of Mosquita of Nicaragua. Holdridge (1962) mapped the Mosquitia as a wet tropical forest region. Taylor (1963) briefly described the pine savannahs in his survey of the vegetation of Nicaragua. Wagner (1964) compared the Mosquitia to the savannahs of Beliz and Yucatan. Munro (1966) discussed the vegetation of the Mosqu- itia of Nicaragua in relation to fire. The United Nations Development Program (UNDP, 1968) mapped and characterized the pine resources in the Honduran Mosquitia. Systematic plant collecting began in the Honduran Mosquitia with sev- eral expeditions, including those reported herein, by personnel from the National University between 1970 and 1976. Nelson (1976, 1978) published check lists of most species collected on these expeditons. Proctor (1983) mentioned his having made subsequent collections in 1981 and described two species. REGIONAL DESCRIPTION The Mosquitia of Honduras is shown in Figure 1. The entire region ts within the Department of Gracias a Dios, and the principal town is Puerto Lempira. The 100 meter contour, shown by the dotted line, 1s approxi- mately the interior boundary of the Mosquitia. The actual boundary is at a slightly lower elevation, where steep mountains clothed with tropical wet forests abruptly meet the flat coastal plain. The savannahs of the Mosquitia occupy flat or gently rolling terrain. Several rivers, notably the Rio Patuca and the Rio Platano, extend from the mountains through the savannahs. Fertile alluvium flanks each of these rivers and supports a corridor of tropt- cal wet forest that extends from the mountains to the coast. These forests, with canopies 30 —40 m tall, stand in sharp contrast and often without ecotone to the low, grassy savannahs. Annual rainfall varies from about 200 to 400 cm. Parsons (1955) said that the region, “is probably the rainiest area of its size in the New World with a savanna-type vegetation.” A 2— 3 month dry season from March to May is mild in some areas but severe in others, sometimes with less than 3 cm of rain a month. Munro (1966) said that high temperatures (mean an- nual >24C) and low spring rainfall stressed the vegetation. Upper soil horizons are highly leached and infertile, gravelly quartz sands and sandy loams lacking in organic matter. Subsoils are poorly drained, inhibiting root growth and thereby limiting that volume of soil from which nutrients can be exploited and in which anchorage can be attained. Pines lack tap roots and often topple in hurricanes (Munro 1966). 260 x 1] ro) ° e 2 ° a « ‘| 2 g % 2 ‘ , fi Q o 6} A = ceere Yn Po > . Be : : eo : a e i : j) 3 * « » 2 e 3 . 2 xe e pr] *3 Hi 7 Tak : . ze ° e ° “we, o; S Ps ‘ : % : : ; e . eeete e s * i. o 5 : : a DEY ‘ a) tee % : a8 ‘ Be fet se e . “ *e NM 2 versa aN o> 8 2 “A; \ ig é , . 4) e ° va Sooo, - ow wee, ., = we — ° e © pe OW bly La “TN ig. 1. Mosquitia region of Honduras, showing collecting locations. A - Ahuas, B - Brus Laguna, P - Puerto Lempira, W - Wampusirpe. 261 Percolation is rapid in dry seasons, causing stress to the shallowly rooted vegetation. Fires burn the savannahs frequently, often annually. The UNDP (1968) reported that 28% of the pine stands in the Mosquitia had been burned within the year of inventory. Fires are ignited at any season but particularly the dry season. At least some fires are set intentionally to provide fresh grass for livestock, to rid ticks and snakes, to facilitate hunting, or simply to provide entertainment (Arnold 1954, Munro 1966). Flames reach to the edges of the tropical wet forests but do not enter them for lack of grami- naceous vegetation or other flammable materials. Frequent fires and infer- tile soils combine to prevent plants of tropical wet forests from colonizing the savannahs. The predominant vegetation towards the coast is sedge prairie, dominated by species of RAynchospora. Pinelands occur primarily in the in- terior of the Mosquitia, where soils are better drained and finer textured. The UNDP (1968) estimated that 76% of the Mosquitia was pineland but that much of this area had been logged. This report said that tree density was low but tree size was large, with 24% of the trees being over 50 cm in diameter. Most trees were reported as having defects, which were thought to have resulted from storms. Thickets of shrubs and small trees occassionally interrupt the savannahs. These are usually dominated by palms (Pawrotis wrightii) or by coppicing hardwoods (Davilla kunthii, Quercus oleoides). Nearer the alluvial rivers, thickets occur that are dominated by species of such genera as Miconia, Isertia, Psychotria, and Helicteres. Parsons (1955, citing R. L. Pendleton) said that these hardwood islands may have developed on former camp sites of Miskito Indians, where there had been protection from fire and where the soil had been enriched by refuse and ashes. . The Mosquitia contains many small creeks with headwaters within the flat coastal plain; some are shown in Figure 1. These minor drainages show little topographic relief and do not have enough flow for much alluvial transport of sediments. They are flanked by narrow ribbons of creek swamp, consisting of short, broadleaved trees. These creek swamps are floristically depauperate, compared to the much taller tropical wet forests along the large, alluvial rivers. Near the base of the mountains, the pine savannahs contain a greater mixture of woody species than elsewhere. Although Pznus caribaea is genet- ally the sole occupant of the overstory, the understory is distinctive for its open growth of broadleaved trees and shrubs. The most conspicuous under- story species are an oak (Quercus olevides) and nance (Byrsonima crassifolia). This pine-oak-nance woods appears to suffer less fire than the open 262 pinelands. Munro (1966) described similar vegetation in Nicaragua and said that it had a fire frequency of about every five years. Vegetation at the coast is distinct from that of the savannahs and consists of beach strand vegetation, mangrove forests, tidal marshes, and scrubby woodlands, similar to that described by Sauer (1967). Ruderal and other obviously disturbed vegetation is restricted to the vicinity of the very few towns and settlements. METHODS Observations in this paper were made during three collecting trips total- ing 15 days in 1972 —73. The first trip was made on December 12 — 14, 1972, near Ahuas (Figure 1) in sedge prairies and creek swamps. The sec- ond trip was made May 17— 23, 1973, with Gustavo Cruz and A. H. Gentry in the vicinity of the Rio Platano. Collecting locations included sedge prairies, alluvial river swamps, tropical wet forests near the river 10 and 18 km inland, and tropical wet forests in low mountains 25 km inland. The third trip was taken August 30-September 3, 1973, to Ahuas, Brus Laguna, and Wampusirpe. Collecting locations included sedge prairies, colluvial creek swamps, pine savannahs, pine-oak-nance woods, and tropi- cal wet forests, the latter near Wampusirpe at the foot of mountains border- ing the savannah region. Sedge savannahs were collected intensively on all three trips, allowing a rather thorough, systematic inventory of the flora. Creek swamps and pine- oak-nance woods were collected less intensively. Although tropical wet forests were visited on two trips, only a fraction of this rich vegetation could be collected. The Mosquitia was reached by twice weekly scheduled flights from Tegucigalpa, weather permitting. Accomodations, electricity, and most logistical ammenities were scarce. Transportation within the region was made difficult from the lack of roads and vehicles. River transportation was by piragua, including one called a tuk-tuk. This craft was a 40 foot long mahogany dugout and was named for the sound of its one-cylinder inboard engine. Several landing strips for light aircraft were scattered throughout the region. Rides were available from the Missionary Aviation Fellowship (Alas de Socorro) which served the Mosquitia. The most complete sets of specimens were deposited at the National University (TEFH) and at the Escuela Agricola Panamericana (EAP). A partial set containing specimens primarily from the savannahs was deposited at Florida State University (FSU). Another nearly complete set and all remaining duplicates were given to the Missouri Botanical Garden (MO) in exchange for a travel grant for one of the field trips. Plant names 263 follow Standley et al. (1958-1975) and Molina (1975) for most groups, Ames and Correll (1953) for orchids, and Clewell (1975) for composites. RESULTS Table 1 lists the flora of the grass-sedge savannahs. Table 2 lists the vas- cular flora of creek swamps along the colluvial streams with headwaters within the Mosquitia. Table 3 lists characteristic plants of the pine-oak- nance woods. Table 4 lists characteristic trees of tropical wet forests. DISCUSSION SEDGE SAVANNAHS. One hundred sixty-five species were recorded in the sedge savannahs (Table 1). The most abundant species was Rhynchospora globosa, and perhaps the next most abundant was Paspalum pulchellum. Other abundant species were Rhynchospora chapmant and R. barbata. Partic- ularly wet areas lacked grasses, but plants of Tonina fluviatilis and Utricu- laria subulata grew abundantly between the cespitose tussocks of Rhynchospora globosa. The relatively dry sedge savannahs contained much Bulbostylis paradoxa and Declieuxia fruticosa. Most herbaceous vegetation was 1— 3 dm all. Thickets of Paurotis wrightii were Common 1n wet areas, and shrubs of Davilla kunthii and Quercus oleoides preferred higher ground. Pine savannahs neighboring the sedge savannahs were characterized by much Pteridium aquilinum, Rhynchospora rugosa, and Setaria geniculata. Species denoted with an asterisk in Table 1 were those that occurred largely or entirely in disturbed horse pastures within the savannahs. These species may not be characteristic of the savannah flora but rather of ruderal habitats. Parsons (1955) noted how, “the open, park-like Miskito savanna bears an extraordinary resemblance to the pine flats of Louisiana or Florida.” Wagner (1964) reiterated Parson’s observation. Twenty-two species (13.3%) listed in Table 1 are indigenous to northwestern Florida (Clewell 1985). Notable floristic di ilarities between the two regions include the prominance of the Compositae and the unimportance of the Melasto- mataceae and Rubiaceae along the northern Gulf coast, as compared to the Mosquitia. CREEK SWAMPS. The canopy of the creek swamp was dense, generally less than 10 m tall, and consisted mainly of broadleaved, evergreen species. Table 2 lists 52 species for the creek swamps. The Guttiferae provided the most important elements of the overstory, and Symphonia globulifera may have been the most abundant tree. Most species were trees and shrubs; most of the rest were epiphytes and woody vines. Terrestrial herbs were 264 Taste 1. Plants of Sedge Savannahs and Their Life Forms. - tree, S - shrub, V - woody vine, H - terrestrial herb, E - epiphytic herb, P - parasice. An asterisk (*) denotes species occurring largely or entirely in savannahs used as pasture POLY PODIACEAE Rhynchospora barbata (Vahl) a H Blechnum indicum Burm.—H Rhynchospora cephalotes (L.) Vahl— Cochlidium rostratum Hook. ) Maxon—E Rhynchospora chapmanti Coe Preridium aquilinum (L.) Kuhn.—H Rhynchospora cyperoides (Sw.) Mart.—H CYATH EACEAE Rhynchospora divergens Curtis—H Alsophila microdonta Desv.—S Rhynchospora globosa (HBK.) R. & S.—H Trichopteris microdonta coe ) Tryon—S Rhynchospora hirsuta Vahl—H LYCOPODIACEAE Rhynchospora marisculus Nees Lycopodium carolinianum L.—H Rhynchospora rugosa (Vahl) Gale—H Lycopodium cernuum V..— Rhynchospora setacea (Berg) Boeckel—H ACEAE Scleria cyperina Kunth—H Pinus caribaea Morelet—T PALMAE GRAMINEAE Panrotis wrightii (Griseb. & Wendl.) Britt. —S Andropogon bicornis L.—H ARACEAE pee leu ostachyus HBK.—H Anthurium turrialbense Engler—H Androp rgatus Desv.—H XYRIDACEAE at pale (HBK.) Nees—H Nyris ambigua saa ex Kunth—H Eragrostis eltrottit S. Wats.—H ERIOCAULACEAE Eragrostis cas steies ae ) Steud. — Tonina fluviatilis Aubl.—H Homolepis aturensis (HBK.) Chase—H COMMELINAC Isachne polygonoides (Lam.) Doell.—H Anetlema geniculata (Jacq.) se ae H Leptocoryphium lanatum (HBK.) Nees—H HYPOXIDACEAE Mesosetum blake: Swallen—H Curculigo scorzoneraefolia (Lam.) Baker—H Panicum blakei Swallen—H IRIDACEAE Panicum cyanescens Nees ex Trin. —H Cipura paludosa Aubl.— Panicum hians Ell.—H Nemastylis tenuis (Herb.) Baker—H Panicum pilosum Sw.—H BURMANNIACEAE Panicum aes Schrad.—H Burmannia _ (Walt.) Mart. —H Paspalum clavuliferum C. Wright—H RCHIDACEAE Paspalum minus Fourn.—H Habenaria “a (Lind1.) ene f.—H Paspalum plicatulum Michx.—H Laelia tibicinis (Batem. ex Lindl.) L. Wms.—E Paspalum pulchellam Kunth—H Scaphyglottis cuneata Schltr.—E Sacctolepis myuros (Lam.) Chase—H Schomburgkia tibicints Batem.—E Setarta geniculata (Lam.) Beauy.—H Spiranthes sp.— Thrasya paspatloides HBK.—H MYRICACE AE — angustifolius (HBK.) Nees ex Myrica certfera L. k.—H FAGACEAE Ti a um sp.—H Quercus oleoides 8. & C.—T RACE AMA reeked ane ae sale He pene " nae Alternanthera sessilis (L.) R. Br.—H diffusus Vahl—H CAPPARAC we ps ae flavus ( cs Nees—H Cleome serrata els —H Cyperus haspan L.—H a AE Cyperus odoratus L.*—H Drosera le Por. Eleocharis filiculmis Kunth— CHRY uae EAE Eleocharts retroflexa (Poir.) Urban Fimbristylis antumnalis (L.) R. & S.—H Hartella racemosa Lam.—S H Chrysobalanus tcaco L.— TasLe | (continued) LEGUMINOSAE Aeschynomene histrix Poit.— Cassia diphyla L.— Cassia flexuosa L.—H Cassia tagera L.—H Clitoria rubiginosa Juss.—V Crotalaria purshii DC.—H Desmanthus virgatus (L. ) Willd. — Desmodium barbatum ee ) Benth. & ie —H Eriosema diffusum (HBK.) G. Don—H Ertosema peer canner Eriosema violaceum (Aubl.) G. Don—S Galactia sp Mimosa pudica L.—H Phaseolus linearis HBK.— Stylosanthes guyanensis (Aubl.) Sw.—H Tephrosia nitens Benth.— Zornia reticulata Sw.—H MALPIGHIACEAE Byrosonima crassifolia (L Polygala adenophora ae —H Polygala hygrophyla < Polygala salviniana eer Polygala timouton Aubl.—H EUP HORBIACEAE Acalypha arve E.*—H Caperonia ple 1 St. Hil.— Croton trinitatis Millsp.*—H Euphorbia anychioides _ —H Euphorbia hyssopifolia L.—H Phyllanthus niruri L.— ALVACEAE Abutilon eat Rose—H Urena lobata L.* STERCULIACEAE Melochia villosa (Mill.) Fawc. & Rendl.—S DILLEN [ACEAE Davilla kunthit St. Hil.—S OCHNACEAE Sauvagesia erecta L.— TURNERACEAE Bee cistoides (L.) Mey. ex Steud.—H Turnera sp. PASSIFLORACEAE Passiflora foetida L.—H MYRTACEAE Psidium guianeense Sw.—S he ete Gis ) Breg.—S TOMATACEAE Aciotis es hae Triana*—H Acisanthera bivalvis (Aubl.) Re —H Acisanthera quadrata Pers.— Clidemia capitellata Bae : D Don—S Clidemia densiflora (Standl. Miconia albicans ee ane Muconia ciliata (L. Rich.) DC.—S Miconta hondurensis D. Sm.—S Miconia lundelliana L. Wms. —S Tibouchina aspera aa NAGRACEAE Jussiaea nervosa Poir.—S GENTIANACEAE Centaurium pringleanum (Wittr.) Rob.—H Schultesta se m.— OCY NACEAE ee ee Blake—S CONVOLVULACEAE Cuscuta ae ane Citharexylum cau oe Stachytarpheta inti (Mill, ) VahlI—H Tamonea spicata Au ABIATAE ee atrorubens Poit.— ee micranthum WJ illd.*— SCROP ives pais ncelonia ciliaris Rob.— Gerardia as Engelm.—H NTIBUL ARIACEAE Utricularia ae Lam.—H Utricularia subulata L.—H UBIACEAE Alibertia edulis (L. Rich.) A. Rich. ex DC.—H Anisomerts protracta (Bartl.) Stand!.—S Borreria laevis (Lam.) Griseb.—H Borreria ocymioides (Burm.) DC.—H Borrerta suaveolens Mey.—H C iia thas hirsutum Bartl.— ae ) Kuntze—S Isertia haenkeana DC.— 266 Taste | (continued) Palicourea galeottiana Mart.—S Spilanthes americana (Mutis) Hieron.—H Palicourea triphylla DC.—SV Sptlanthes merrillii Standl. & Wms.—H Psychotria capitata R. & P—S Spilanthes poliolepidica Moore—H Psychotria sar (Aub bs S$ Vernonia cine erea (L.) Less.—H MPOSI Wedelia trilobata (L.) Hitche.— Erechtites ee L.) re a Zexmenta pinetorum Standl. & Steyerm.—S Orthopappus oe (Sw.) Gl.— represented mainly by a few ferns and dense patches of the clambering Scleria secans. Many shrubs of the sedge savannahs (Table 1) were increasingly common with proximity to creek swamps. Some of these shrubs rightfully could be included as components of both conmmunities, although they were restricted to the sunny edges of creek swamps. Calliandra houstoniana commonly flanked creek swamps. In physiognomy, these swamps resembled creek swamps along the northern Gulf coast that are dominated by species of C/iftonia, Cyrilla, Ilex, Lyonia, and Magnolia. These Gulf coastal swamps share little in common floristically with Honduran creek swamps. AK-NANCE WOODS. As noted above, this ecotonal commu- nity 1s nee for its three most conspicuous woody species, Pinus caribaea, Quercus oleoides, and Byrsonima crassifolia (nance). The pines comprised an open overstory 20— 25 m tall. One pine stump 56 cm in diameter con- tained more than 100 annual rings. Oaks were upwards to 12 m tall and typically formed an open understory along with small trees and large shrubs of other species. Nance was the most common of these other species and grew toa height of 5 meters. Calliandra houstoniana was common, and a tree fern, Al/sophtla myosuroides, was conspicuous. Grasses and forbs gener- ally comprised a continuous ground cover. This ground cover became sparse or absent in the occasional dense thickets of understory trees. Table 3 lists 28 species characteristic of the community. Many other species belong to this community that also occur in the pine savannahs. Unfortunately, the site that was examined had been recently burned, mak- ing a thorough inventory impossible. The most common herb was Paspalum pectinatum. Grasses, rather than sedges, predominated in the ground cover. Notable by its absence was Rhynchospora globosa, the most abundant species of the sedge savannahs. The soil was loamy and probably more fertile than in the savannahs. The community resembled the pine- oak-hickory woods of the Tertiary highlands near the northern Gulf coast, both in physiognomy and with respect to soil type and fire frequency. 267 Taste 2. Plants of the Creek Swamps and Their Life Forms T - tree, S - shrub, V - woody vine, H - terrestrial herb, E - epiphytic herb, P - parasite. POLY PODIACEAE Lindsaea stricta (Sw.) Dryand. Vittaria lineata (L Scleria secans (L.) Urban—H ARACEA A E Syngonium ae Schott—H LIACEAE Aechmea paiae re : Griseb.—E Bromelta pinguin L.—E Tillandsia ee Schult. a Tillandsia bulbosa Hook.— PIP SACRE Piper aduncum L.— THACEAE Phoradendron quadrangulare (HBK.) Krug. & Urban—PS Struthanthus Benes (HBK.) Blume—PV ANNONACEA Xylopia Se dam ) Mart. rae MINOSA Calliandra eee (Mill. ) en —S Cassia bacillaris L.f.—V Pithecellobium donnell-smithii (B. & R.) Standl.— MALPIGHIACEAE Byrsonima Beet (L.) HBK.—T ORBIACEAE Pera arborea ae —T AQUIFOLIACEAE Ilex guianensis (Aubl.) Kuntze—T VITACEAE Cissus erosa L. Rich. =e Cissus salutaris HBK.— RCH Hibiscus furcellatus Lam.—S TERCULIACEAE Helicteres acai HBK.—S NIACEAE Curatella americana ne GUTTIFERAE Clusi ia flava Jacq a Clusia salvinit D. a ae L. ria —T Vismia oe Sprague & cove aces OCHLOSPERMACEA aia eke Willd. ex ae Le Eugenta a me fe ere Clidemia ae (Sw.) DC.—S Conostegia tcosandra (Sw.) DC.—S Henriettea succosa Sour ) DC.—T Miconta borealis Gl.— Muiconta thaguensis eine Triana—T Miconia schlechtendalit Cogn.— Tococa aa oe —S AGRACEAE Jussiaea eae See OTACEAE Chrysophyllum re ee CY NACEAE Mandevilla ait (R. & PB.) Woodson—V VOLVULACEAE Cuscuta pe IRBEN Negeee Sg Mike a L.— Lantana canary SOLANACEAE Solanum jamaicense Mill.—S RUB IAC Chiococca pachyphylla Wernham—V Guettarda combsit Urban— TROPICAL WET FOREST. As mentioned earlier, the flora of the tropi- cal wet forests contrasted markedly with that of the savannahs, including the creek swamps and the pine-oak-nance woods. Since a systematic in- ventory of tropical wet forest was not possible, only a few trees have been Taste 3. Plants Characteristic of the Pine-Oak-Nance Woods and Their Life Forms. T - tree, S - shrub, V - woody vine, H - terrestrial herb, E - epiphytic herb, P - parasite. POLY PODIACEAE GUTTIFERAE Smt ee Vismta Bae) Bea he & Riley—TV Polypodt isertale § i MELAS a iAE Vittaria ee (L.) J. Sra —E Bellucta costaricensts i. ogn CYATHEACEAE Clidemia capitellata onl ) > Don—S Alsophila ales Liebm.—S Henrtettea fascicularis (Sw.) Gomez—T PINACEAE Miéconta borealis Gleason— Pinus caribaea Morelet-—T Miconia thaguensis (Bonpl.) Triana—T GRAMINEAE Miconta lacera (Bonpl.) Naud.—S Axonopus purpusit ( ae Chase—H Miconta prasina (Sw.) DC.—T EAE ARALIACEAE Anthurium tee a ) Engler—H Didymopanax morototoni (Aubl.) Dene. ORCHIDACEAE Planch.—T Epidendrum nocturnum Jacq.—E LETHRACEAE FAGACE Clethra macrophylla Mart. & Gal.—T Quercus oleoides S. & C.—T RUBIACEAE LEG NOSAE Altbertia edulis (L. Rich.) 7 Rich.—S Calliandra kes a ) Standl.—S Palicourea triphylla DC Cassia bactl ae f.— Psychotria cuspidata a. ex R. & S.—S KeriG aAGiae Sabicea panamensis Wernham— Byrsonima ee L.) HBK.—T COMPOSITAE AQUIFOLIACEAE Zexmenia pinetorum Standl. & Steyerm.—S Ilex gutanensis (Aubl.) Kuntze—T we Tasie 4. Trees Characteristic of che Tropical Wet Forest. SWAMP FOREST, NEAR MOUTH RIVERINE FOREST 10-25 KM FROM or Rio PLATANO MOUTH OF Rio PLATANO Montrichardia arborescens “s ) Schott Pourouma aspera Trecul. Coccoloba rae 5 Jac Jochysta bondurensis oe nnona aes Cupania auriculata Stand. Inga spur ex Willd. Apeiba membranacea Spruce ex Benth. ps aia ine (L.) Benth. Protium glabrum (Rose) Engler s rapa guianensis Aubl. Annona murjcata L. achira aquatica Aubl. Sloanea zuliaensis Pitt. : ‘ymphonia globulifera Lf. ixa orellana Pleuranthodendron mexicana (Gray) L. Wms. Aspidosperma mecalarton | Muell.-Arg. Grias integrifolta (Stand. ) Kunth Miconia hondurensis D. ri hes bicolor ramea stenura Sigil, Hamelta rovirosae Wernham Palicourea fastigata Benth. 269 listed to introduce the tropical wet forest and to indicate the considerable floristic differences with savannas. Table 4 lists trees collected in a swamp near the mouth of the Rio Platano and in the floodplain and associated slopes of this river at collecting stations 10, 18, and 25 km inland from the coast. Symphonia globulifera was the only species common to the tropical wet forest and savannah flora listed in Tables 1-3. The same degree of floristic contrast is evident along the northern Gulf coast, where hardwood forests of floodplains and ravines differ substantial- ly from adjacent pinelands and savannahs (Clewell 1977, Clewell et al. 1982). ACKNOWLEDGEMENTS I am indebted to Antonio Molina R. of EAP for help with plant identifi- cations. Iam grateful to Dr. and Mrs. Samuel Marks for their hospitality at the Moravian Mission Hospital at Ahuas, to John and Gretchen Eoff for outfitting me with guides and piraguas on the Rio Platano, and to John Watson for a flight to Wampusirpe. REFERENCES AMES, O., and D. S. CORRELL. 1953. Orchids of Guatemala. Fieldiana, Boc. 26:1—72 ARNOLD, B. A. 1954. Notes on the geography of the Brus Lagoon area of northeastern Honduras. Unpubl. report for Office Naval Res. Contract 222(11) NR 388 067 to C. O. Sauer, Univ. Calif., Dept. Geogr. 15 pp CARR, A. i , Jr 1950. Ourline for a classification of animal habitats in Honduras. Bull. Amer. Mus. Nat. Hist. 94:563 — ; ae A. E 1975. Las compuestas de Honduras. Ceiba 19: 119 — 244. 1977. Geobotany of the Apal: eee River region. Florida Dept. Nat. Resources, Marine Res. Publ. No. 26, pp. >. «1985. Guide to the vascular plants of the Florida panhandle. Tallahassee: Florida State University Press/Univ. Presses Floric : . GOOLSBY, and A. G. SHUEY. 382. Riverine forests of the Souch Prong Alafia River system, Florida. Wetlands 2:21 — rence: L. R. 1962. Mapa ecologico de ee ‘Ousnieaeien de los Estados Americ MOLINA R., un 1975. Enumeracién de las plantas de Honduras. Ceiba 19:1 — 118. MUNRO, N. 1966. The fire ecology of C ee an pine in Nicaragua. Proc. 5th Ann. Tall Timbers Fire Ecol. Confr. pp. 67 —8 NELSON, C. 1976. Plantas nuevas par 7 flora de Honduras. Ceiba 20:58-68. «1978. Contribucidnes a la flora de la Mosquitia, Honduras. Ceiba 22:4 1-64. see: i 2 1955. The Miskito pine savanna of Nicaragua and Honduras. Ann. Assoc. er. Geog. 45 Soren G. R. 1983. New plant records from the Mosquitia region of Honduras. Moscosoa 2:19 — 22 27 seosal a bear The physical geography of the east coast of Nicaragua. Office Naval I. Rpt. for Contract 222(11) NR 388 067 to C. O. Sauer, Univ. Calif., SAUER, - “1967, Geographic cacy of seashore vegetation ey the Mexican Gulf coast. ae State Univ. Press, Coastal Studies Ser. No. 21. 59 pp. STANDLEY, P. C., et al. 1958— Flora of Guatemala. ne Bot. 24, Parts b= 2: TAYLOR, B. W. 1963. An outline of the vegetation of Nicaragua. J. Ecology 51:27 — 54. UNITED NATIONS DEVELOPMENT PROGRAM. 1968. Survey of pine forests, Honduras, final report. Food & Agric. Organization, FAO/SF: 26 - HON 50. 79 pp. + maps. WAGNER, PL. 1964. Natural vegetation of Middle America. Pp. 216— 264 in R. C West, ed., Handbook of Middle American Indians. Vol. 1. Natural Environments and early culcures. Austin: Univ. Texas Press. SYSTEMATICS OF NAMA (HYDROPHYLLACEAE): COMPARISON OF SEED MORPHOLOGY OF SECTS. ARACHNOIDEA AND CINERASCENTIA WITH FIVE SPECIES OF ERIODICTYON AND TURRICULA PARRYI JOHN D. BACON and LOUIS H. BRAGG Department of Biology, The University of Texas at Arlington, Arlington, TX 76019 GARY L. HANNAN Department of Biology, Eastern Michigan University, Ypsilanti, MI 48197 ABSTRACT Nama lobbii and N. rothrockii, members of the monotypic sects. Arachnoidea and Cinerascentia, respectively, are divergent elements in Nama and their relationships with the genus have been questioned. Alternative genera with which the two might be allied are Eriodictyon, traditionally allied with Nama, or possibly, Turricula, historically associated with both genera. Utilizing scanning electron microscopy, seeds from five distinctive species of Ertodictyon, E. crassifolium var. denudatum, E. tomentosum, E. traskiae subsp. smithit, E. angustifolium and E. californicum, Turricula parryi and the two namas were examined for implications resident in seed morphology that might bear on placement of N, /obdii and N. rothrockit. Seeds of the namas are distinct from both Eriodictyon and Turricula, while seeds of the latter taxa are quite similar. Seed morphology combined with other available evidence suggests that N. /obbii and N. rothrockii are distinct phyletic elements within the Hydrophyllaceae. INTRODUCTION Nama is the second largest genus in the Hydrophyllaceae, housing about 50 species ranging from short-lived annuals to robust subshrubs. Distribu- tionally, species occur predominantly in western and southwestern United States and Mexico but three venture into South America and one is endemic to Hawaii. They occupy a variety of habitats ranging from arid, alkaline- gypseous flats in the central Chihuahuan Desert to more mesic, but seaso- nally dry slopes and higher elevations in the sierras of western North America and eastern Mexico. As treated by Hitchcock (1933), the genus is structured of five sections; Arachnoidea, Cinerascentia, Conanthus, Zonolacus and Nama (as Eunama). Much the largest, sect. Nama comprises an extremely diverse group of SIDA 11(3):271 — 281. 1986, 272 species unified by their entire leaves, styles that are free or connate less than half their length, a superior ovary, membranous, loculicidally dehiscent capsules and axillary, extra-axillary or terminal, one, two or few-flowered cymose inflorescences. Each of the smaller sections is separated from the largest by clearly distinctive features; sect. Conanthus, (3 sp.), houses dichotomously branched annuals with styles connate over half their length; sect. Zonolacus, (1 sp.), exhibits a half inferior ovary; sect. Arachnoidea, (1 sp.), produces cartilaginous, loculicidally and septicidally dehiscent capsules; sect. Cymerascentia, (1 sp.), displays crenate leaves and capitate inflorescences. Species of sects. Arachnoidea and Cinerascentia, Nama lobbii Gray and N. rothrockit Gray, respectively, are critical elements in Nama. They are the most divergent namas, morphologically, and thus, set the limits of Nama. Moreover, it is through N. /obbii that the nearest relative of Nama has tradi- tionally been identified as Eriodictyon, so the former taxon serves in positioning Nama, as well. However, a growing body of evidence portrays these taxa as more distant from Nama proper than was reckoned by Hitchcock (1933) and their inclusion in Nama lately has been questioned (Bacon 1984, 1974; Chance and Bacon 1984; Raven and Axelrod 1978). For, not only do they diverge from the bulk of Nama in gross morphology, they also diverge in seed morphology (Chance and Bacon 1984), chromo- some number (Bacon 1984, Cave and Constance 1947, 1959; see Con- stance 1963, for summary) and flavonoid chemistry (Bacon, Fang and Mabry, in review). Thus, placement of N. /obbii and N. rothrockii appears seriously challenged and their positioning elsewhere must be considered. Since N. /obbii has been closely associated with Eriodictyon, traditionally, the latter taxon ts a logical alternative genus with which the former species and, perhaps, N. rothrockit might be more closely allied. A second alterna- tive is the monotypic Turricula, historically associated with both Nama and Eriodityon. Since seed morphology has proven informative in suggesting relationships not previously evident, morphologically, in Nama (and other groups, see Chance and Bacon 1984, and references therein), it follows that comparison of seed morphology among these four taxa might provide in- sight relevant to positioning of N. /obbii and N. rothrockii. Therefore, we have examined seed morphology of the two species of Nama, five distinc- tive species of Eriodictyon, E. crassifolium Benth. var. denudatum Abrams, E. tomentosum Benth., E. traskiae Eastw. subsp. smithii Munz, E. angustifolinm utt., E. californicum (H. & A.) Torr., and Turricula parryi (Gray) Macbr. We report herein results of that survey and implications as they bear on placement of N. lobbiz, N. rothrockii and systematics of Nama proper. 213 MATERIALS AND METHODS Mature, whole seeds removed from herbarium specimens were mounted on brass stubs with double-stick carpet tape. For examination of internal features of the testa, seeds were sectioned free-hand with a razor blade and mounted as for whole seeds. Specimens were coated to a thickness of ap- proximately 25 —30 nm in a Polaron E5100 sputter coater using gold- palladium target and examined with a JEOL JSM 35-C SEM at an acceler- ating voltage of 15 kV. A minimum of five seeds per collection was exam- ined RESULTS Seeds of examined species of Eriodictyon and Turricula are remarkably ho- mogeneous. While there are variations in precise shape and size, as noted for seeds in other plant groups (see Chance and Bacon 1984, and references therein), generally, seeds of both taxa are irregularly oblong (Figs. 1,2,3,7,8,9) in outline, more or less angulate-wedge shaped in cross sec- tion and average over 1 mm in length (Table 1). The outer testa in all is clearly cellular in composition and surface patterning in all is shallowly reticulate, following the terminology of Chance and Bacon (1984). Reticulum cells are arranged into definite rows, with each cell elongated at right angles to the long axis of the seed. The longer radial walls in each cell are conspicuously concave. This organization accounts for their “transverse corrugations” as seen under the light microscope. In section (Figs. 4,5,6,10,11,12), the lower tangential wall of each reticulum cell appears weakly to strongly undulate suggesting thickenings or, perhaps, compres- Taste 1. Seed characteristics of N. lobbir, N. rothrocktt, Evtodictyon and Turricula. Length Width Coat Thickness Taxon (X,mm) (X,mm) (min-max, [L) N. lobbit 1.26 0.82 12—40 N. rothrockit 1.52 1.01 30-58 E. angustifolium 1.02 0.63 13—24 E. crassifolinm var. datum t222 0.78 4-8 E. tomentosum 1.09 0.70 3-6 E. traski bsp. Smithti Hannan 65 1.11 0.80 6-12 H 60 1.29 0.71 10—12 E. californicum 1.03 0.62 6-8 T. parryi 274 Figs. 1—6. Seeds of Eriodictyon. (Bar in whole seed figures = 100p; bar in section figures = 10Op; bars apply to all figures) 1. E. crassifolium var. denudatum, whole seed. 2. E. tomentosum, whole seed. 3. E. i : ASS Le 9 Hi. ASST OE aay smithit subsp. traskiae, whole seed. 4. E. cr um vat. denudatum, seed section (T = seed coat, E endosperm, applies to all sections). 5. E. tomentosum, seed section. 6. E. smithii subsp. traskiae, seed ection. at) sion ridges. Radial walls, however, lack thickenings, with the possible ex- ception of E. tomentosum (Fig. 5, upper left corner); however, additional examination is required to confirm or deny this possibility since potential thickenings do not occur in each reticulum cell in this species. The seed coat in these taxa varies in thickness among species and at different points within a single seed section (Table 1) and ranges from a low of 3p in E. tomentosum to a high of 24 in E. angustifolium. Seeds of N. rothrockiz and N. lobbii (Figs. 13,14) essentially conform with those as reported by Chance and Bacon (1984) except that seed coat thick- ness and average seed length of both is slightly less for seeds examined in this study. Seeds of the former taxon exhibit a foveolate-alveolate surface patterning and a solid outer testa 30-58 in thickness (Fig. 15; Table 1); those of the latter exhibit a papillose surface (Fig. 14) and a solid outer testa 12-40 in thickness (Fig. 16; Table 1). Seeds of both N. /obbii and N. rothrockii are distinctive when compared with Eriodictyon ot Turricula. Neither of the former species exhibits the chambered, ordered reticulum of the latter and the seed coat is generally much thicker in the former. In this last respect, however, the thinner areas in the testa of N. /obbii are within the range of variation in thickness of the testa in Eriodictyon and Turricula. Nevertheless, in overall morphology there is little similarity among seeds of N. /obbii or N. rothrockii and those of the latter genera. DISCUSSION The lack of similarity in seed features of N. /obbi, N. rothrockit and Eriodictyon is accentuated by the striking similarity among seeds of the lat- ter and Turricula; seed features clearly countenance an alliance of these two genera. However, neither of the namas can be closely allied to Erodictyon or Turricula based on seed characteristics, nor can they be closely allied one to another on these features. Indeed, seed morphology implies that both N. lobbii and N. rothrockii are elements phyletically distinct from Ertodictyon, Turricula and Nama (see Chance and Bacon 1984) and, therefore, distinct elements within the Hydrophyllaceae, as suggested by Bacon, Fang and Mabry (in review). Moreover, such a view is consistent with other available evidence and reconciles the lack of agreement in past dispositions of N. lobbit. In structuring Nama into five sections, Hitchcock (1933) was recogniz- ing the unique features exhibited by species forming the four smaller sections. Indeed, features of these species had long been appreciated, but there had been little agreement as to their systematic significance. Various workers had subjectively appraised one or another of these taxa as distinct 276 a; & y californicum seed Se 12. Seeds of Eriodictyon and Turricula. 7. E. angustifolium, whole seed. 8. E 9. T. parryi, whole seed. 10. E. angustifolium, seed section. 11. E. californicum, 7 eed. harryt, seed section. Figs. whole s 12.7 ’ & 3 é x “ gee aes : a: € & i¢< a bs . a # * re cee : Brg S ® oe ae & i ‘ x s oe SS ak ee ‘ % fice Rees a a i : & = x Figs. 13 — 16. Seeds of N. rothrockit and N. lobbii. 13. N. rothrockii, whole seed. 14. N. lobbit, whole seed. 15. N. rothrocki1, seed section. 16. N. fobbiz, seed section. genera (e.g., Peter 1897, Gray 1875), as subgenera of Nama (Brand 1913, Jepson 1925) or as members of other recognized genera (Greene 1885, Hall 1906). In positioning N. /obbiz, Hitchcock (1933) was aware that the taxon had been transferred to Ertodictyon by Greene (1885), accepted there by Hall (1906), but excluded from that genus in its then most recent treat- ment by Abrams and Smiley (1915). Entering into Hitchcock's (1933) deliberations, too, was the positioning of the problematical Turricula; orig- inally described as a Nama (Gray 1880), the taxon was transferred to Eriodictyon by Greene (1889), where it was accepted by Hall (1902), but was rejected there by Abrams and Smiley (1915) and subsequently established as the monotypic Turricula by Macbride (1917). Hitchcock 278 concluded, nevertheless, that 7) parryi was more appropriately housed in Ertodictyon; he retained N. /obbii as a section of Nama and then allied Nama and Eriodictyon through these two taxa. ubsequent to Hitchcock’s work, Cave and Constance (1942, 1944, 1947, 1950, 1959) and Constance (1963) have extensively surveyed chro- mosome numbers within Hydrophyllaceae. All species of Eriodictyon exhib- it n= 14; T. parryi has n= 13, supporting its recognition as a monotypic genus. In contrast, Nama, including sects. Conanthus and Zonolacus, 1s strikingly diploid with x=7 (only three species of Nama proper are con- sistently, tetraploid; all are herbaceous and clearly to be related to other species in Nama proper; see Bacon 1984). However, N. /obbii exhibits n=14, while N. rothrockii has n= 17 (Bacon 1974, 1984; Constance 1963), marking them as distinct within Nama. Chromosome number of N. fobbiz supports its potential alliance with Eriodictyon; however, the number in N. rothrockii is known elsewhere in the family only in the South African Codon. Chance and Bacon (1984) have found that seed morphology identifies six species groups within Nama. Species of sect. Nama compose five groups; importantly, species of sects. Conanthus and Zonolacus are separately accomodated in two of these five groupings. Species of sects. Arachnoidea and Cinerascentia, in contrast, form the sixth seed group; their seeds are much larger, exhibit a thicker seed coat and possess distinctive sculpturing patterns when compared with other groups. However, as evident herein, seeds of each taxon are distinctively organized, and Chance and Bacon (1984) noted that any relationship between the two should be viewed as distant with relationships to the bulk of Nama even more remote. Results presented in this report substantiate the views of Chance and Bacon (1984) and, furthermore, suggest that the two namas should be positioned some distance from both Eriodictyon and Turricula Flavonoid chemistry of N. /obbii and N. rothrockii has been determined (Bacon, Fang and Mabry, in review), and it is unique within Nama, as well; both taxa accumulate 6-oxygenated flavone aglycones, whereas other namas accumulate flavonoids with usual oxygenation patterns. However. flavonoids of three species of Eriodictyon, E. tomentosum, E. californicum and E. angustifolium, also have been resolved (Bacon et al., in review), and al accumulate 6-oxygenated flavone aglycones; chemistry of the two namas and three eriodictyons is quite similar, arguing that they are to be allied, nevertheless Morphologically, Erzodictyon is one of the more homogeneous genera of Hydrophyllaceae. All species are rhizomatous shrubs with scorpioid in- florescences, except for E. capitatum Eastw., and loculicidally and septici- 272 dally dehiscent capsules (Munz and Keck 1959, Carlquist et al. 1983). As well, their uniform chromosome number and the uniformity in flavonoid chemistry and seed morphology among examined species argues convinc- ingly that the genus represents a single phyletic element. To position N. lobbii, as suggested by Greene (1885), or N. rothrockiz within Ertodictyon would destroy the homogeneity of the latter; although N. /obbi7 is woody and both namas are rhizomatous (Bacon, pers. obs.; Jepson 1943), neither is a shrub, both tend to be spreading and mat-forming, and neither ex- hibits scorpioid inflorescences. Moreover, seed morphology of these taxa does not contenance such a union. Considering Turricula, its seed morphol- ogy clearly suggests a close alliance with Erzodictyon; however, while it possesses scorpioid inflorescences and its capsules dehisce in the same man- ner as do those of Eriodictyon, this perennial’s upright, herbaceous habit and its chromosome number argue that it is correctly recognized. Seed mor- phology clearly distinguishes N. /ob6i7, in particular, and Nama proper, in general (see Chance and Bacon 1984), from Turricula. Therefore, any rela- tionship between Nama and Eriodictyon is portrayed as less direct and more remote than surmised by Hitchcock (1933); in fact, how Nama proper relates to N. lobbii, N. rothrockii or Eriodictyon is not clear. Indeed, accumu- lated evidence suggests that N. /obbii and N. rothrockii are distinct phyletic elements within the Hydrophyllaceae. Their affinities, as suggested by their flavonoid chemistry and rhizomatous habit, lie nearer to Erzodictyon than to Nama, although N. /obbiz, with its capsular dehiscence pattern, woody habit and chromosome number is nearer Ertodictyon than 1s N. rothrockii. Such a view renders Nama, although still diverse, more homoge- neous and acknowledges the distinctive characteristics of N. lobbu and N. rothrockit. A treatment formalizing this view is forthcoming. ACKNOWLEDGEMENTS This work was supported by NSF Grant DEB 8108513. APPENDIX: Source of Seed Samples Vouchers for Eriodictyon are deposited at EMC; those for Nama will be deposited at TEX. Nama lobbti: CALIFORNIA: Eldorado Co.: 14 mi N of South Lake Tahoe on Hwy 89, near Emerald Bay, Bacon 1700. N. rothrockii: CALIFORNIA: Inyo Co.: 1.9 mi below Onion Creek Campground, 10.1 mi W of Independence, Bacon 1699. Eriodictyon angustifolium: UTAH: Washington Co.: 1 mi S of Pintura, Hannan 503. E. crasstfolium var. denudatum: CALIFOR- NIA: Ventura Co.: California Hwy 33, 20.4 mi N of jct with California Hwy 150, Hannan 280 64. E. tomentosum: CatirorNia: Kern Co.: Park Hill Rd, 3.7 mi S of California Hwy 58, ss 58. E. smithii ssp. traskiae: CALIFORNIA: Santa eee Co.: E Camino Cielo Road, of Gibralear Road, Hannan 60; California Hw 2.8 mi N of Burton Mesa Rd., pie Hills near Lompoc. Hannan 05. E. Bie CALIFORNIA: Calaveras Co.: Old Priest Grade (California Hwy 120) Ww of Chinese Camp, Hannan 57. Turricula parryi: CH) CALIFORNIA: San Diego Co.: Pine Valley, Rose 35306 (MICH). REFERENCES ABRAMS, L. R. and E J. SMILEY. 1915. Taxonomy and distribution of Eriodictyon. Bot. Gaz, 60s 15 — 153. BACON, J. D. 1984. aes eee and taxonomic notes in the genus Nama ee ee ss 197 ai mosome rumbes and taxonomic notes in the genus Nama (Hy deoshyliacess). Brittonia 26:1] BRAND, A. 1913. gl ee re Pavlen nc IV, 25:1— 210. CARLQUIST, S., V. M. ECKHART and D. C. MICHENER. 1983. Wood anatomy of Hydeophyllaceae I. Erin Aliso 10: 2 412. LAVE, M. and L. CONSTA ; ane Chromosome numbers in the Hydrophyllaceae. Univ. Calif Publ. Bot. - ae . 1944. Chromosome one in the Hydrophyllaceae. H. Univ. Calif. Publ. Bot. 18: 293 = 298. 1947. Chromosome numbers in the Hydrophyllaceae. II]. Univ. Calif. Publ. Bot. 18: 449 465 1950. Chismigsoine numbers in the Hydrophyllaceae. IV. Univ. Calif. Publ. Bot. 23: 363 = Oe. 1959. Chromosome numbers in the Hydrophyllaceae. V. Calif. Publ. Bot. 301234 258. CHANCE, G. D. and J. D. BACON. 1 ogy in Nama (Hydrophyllaceae). Amer. J. Bor. CONSTANCE, L. — Chromosome number classification in Hydrophyllaceae. Brittonia 15:27 GRAY, A. 1875. i‘ conspectus of North American Hydrophyllaceae. Proc. Amer. Acad. 332: Arts 10:3 e 984. aden implications of seed coat morphol- eae eee B80, ao! of California. In: Geological survey of California, 1:1 — 628. GREENE, E. : Studies in the botany of California and parts adjacent. Bull. Calif. Acad. Sci. ; 201 = 30 02. aoe ts . New or noteworthy species. VI. Pittonia 2:13 — — H. M. ae A botanical survey of San Jacinto Mountains. an Cale Publ. Bot. — 140. Seeenemene ees he 0 57 Sa on Californian plants. Il. Zoe 5:263 — 266. HITCHCOCK, Cc. 1933. A taxonomic study of the genus Nama. Amer. J. Bot. 26:415 — 430, eee JEPSON, W. L. 1925. Hydrophyllaceae . A manual of the flowering plants of Califor- nia. Univ. California Pre 835. . 1943. 5 Gu ronrill ce ae. In: A flora of California. Univ. California Press. pp. . 809 — 223> MACBRIDE, i. FE 1917. Notes on the feces and a few other North American spermatophytes. Contr. Gray Herb. 49:23 —59. 281 MUNZ, P A. and D. A. KECK. 1959. Hydrophyllaceae. In: A California flora. pp. D0: PETER, A. 1897. Hydrophyllaceae. In: Engler and Prantl, Naturl. Pflanzenf. IV, 3a:54—71 RAVEN, P. H. and D. I. AXELROD. 1978. Origin and relationships of the California flora. Univ. Calif. Publ. Bot. 72:1— 134. VEGETATIVE APOMIXIS IN MUHLENBERGIA REPENS (POACEAE: ERAGROSTIDEAE) CLIFFORD W. MORDEN and STEPHAN L. HATCH S. M. Tracy Herbarium, Department of Range Science Texas AGM University, College Station, TX 77843, U.S.A. Vegetative apomixis in grasses refers to the production of bulbils (a small bulb or bulblike structure produced on above ground parts, Radford et al. 1974, p. 92) in flowers by modification of spikelet structures (lemma, palea, androecium, and/or gynoecium). The term “vivipary”, which is the germination of a seed within the flower (Stebbins 1941), has often been misapplied to describe this phenomenon (Nygren 1954, Gould 1968). Vegetative apomixis 1s known to occur in the families Agavaceae, Liliaceae, Poaceae, Polygonaceae, and Saxifragaceae. In grasses, it is most commonly associated with the tribes Poeae (Poa and Festuca) and Aveneae (Deschampsia and Agrostis). However, it has also been described in the Chlorideae for Bouteloua (Hill 1982), the Eragrostideae for E/eusine (Nygren 1954), and in the Paniceae for Setaria (Nygren 1954). This condition has been previously mentioned in the Eragrostideae for Muhlenbergia richardsonis (Trin.) Rydb. (Cronquist et al. 1977) although not described. In the course of field collections, three populations of M. repens (Presl) Hitchc. were collected that had spikelets exhibiting vegetative apomixis. Two of these popula- tions were separated by approximately 50 kilometers and disjunct from the third by approximately 600 kilometers. Collection data for the specimens are given in Table 1. Following is a description of the spikelet parts of the three populations. Spikelet structures of non-apomictic spikelets have glumes 1.6— 2.4 mm long, lemmas 2.6—3.2 mm long, and paleas 2.1—2.7 mm long (Fig. 1). The vegetative apomictic specimens of Morden et al. 518 have spikelets with dimensions that are average for the species (Figs. 2 and 3). The rachillas of these spikelets are not elongated. The androecia and gynoecia are modified into bulbils and the leaf of the bulbil either extends out the top of the spikelet between the lemma and palea, or grows through the palea and curves upward, usually twisting. The production of bulbils in Morden et al. 521 and Morden 684 is similar to that of population 578. However, the appearance of these spikelets are greatly modified. The glumes are greater than 3 mm long (longer than many sexual spikelets). The rachilla is also elongated, in most cases 1 — 2 SIDA 11(3):282 — 285. 1986. 283 mm long, resulting in the floret being exposed well above the glumes. The lemma is elongated to 5 mm or more in length and the palea appears to be absent or modified forming the outer leaf of the developing bulbil. The appearance of the bulbil in other respects is similar to that previously de- scribed. Vegetative apomixis in these populations was not present in all spikelets of the inflorescence. Apomictic conditions are usually limited to several spikelets within an inflorescence and only a few individuals from each population. There are several postulations as to why vegetative apomixis develops in populations. There is a strong correlation between the presence of this con- dition and the ploidy level of the species involved such that as ploidy level 4 Figs. 1—5. Vegetative apomixis in Muhlenbergia repens. 1. : non-vegetative apomictic spikelet of M. repen san average spikelet dimensions (Morden ef al. 521). 2. Vegetative apomictic spikelet with che leaf of the bulbil extending out the top of the spikelet Condes et al, 518). 4. Vegetative geanens spikelet with the leaf of the bulbil extending through the palea and upwards (Morden et al. 518). 4. Vegetative apomictic spikelec with elongated glumes, rachilla, and lemma. Bulbil ts not evident (Morden 684). 5. Vegetative apomictic spikelet with ee glumes, rachilla, and lemma. Bulbil ts extending through the palea and upward (Morden et al. 5 284 : |. Collection data for vegetative apomictic populations in Muhlenbergia repens. Specimens are reo at TAES. XICO. Nuevo Leon. Scattered patches along hillside, 8 km W of “Y” intersection in Galeana, ani m elev., 27 Aug 1981, Morden, Hatch, & Valdes R. 518; in low areas along Hwy 57, 32 km N of San Roberto, 18 km § of San Rafael, small cae associated with M. atilis and M. williflora, ee m elev., 27 Aug 1981, Morden, Hatch, & Valdes R . UNITED STATES. Texas. Jeff Davis Co.: 17 miS of Kent in canyon along Hwy 118, 1730 m ee 13 Oct 1982, Morden 684. increases, vegetative apomixis becomes more prevalent (Stebbins 1941, Nygrene 1954). This is found either within a single species with various ploidy levels or within a genus. In his work on Deschampsia, Nygren (1954) found vegetative apomixis to be caused by a single recessive gene. Nygren also reported that for Poa alpina this condition was either gene related or environmentally stimulated. Because of the presence of two such distinct morphological forms, we feel that vegetative apomixis in Muahlenbergia repens 1s the result of two independent evolutionary occurrences. Since it is seldom found in natural populations, it is possibly the result of a recessive gene that 1s prompted by environmental stimulation resulting in a low frequency of occurence in the population. Another means by which this may have become incorporated into M. repens is through hybridization. Muhlenbergia richardsonis is closely related to M. repens and independent studies by the first author have shown there to be intergradation of morphological characteristics in eastern Arizona and western New Mexico where the two species overlap in distribution. It is possible that genes carrying traits for vegetative apomixis were passed from one species to the other and introgressed into the other populations. This process could have happened in either direction. However, it does not seem likely because of the two differing morphological types. ACKNOWLEDGEMENTS This is a technical bulletin T.A. No. 20969 of the Texas Agricultural Experiment Station. Thanks are extended to Sigma Xi, the Scientific Re- search Society for partial financial support for this study. Appreciation is expressed for the constructive remarks of Kevin Jensen, Chuck Coffey, K N. Gandhi and two anonymous reviewers of the manuscript. REFERENCES ar A., A. H. HOLMGREN, N. H. HOLMGREN, J. L. REVEAL, and P. K. HOLMGREN. 1977. Intermountain flora: vascular plants of the Intermountain West, vol. : Columbia University Press, New York 285 GOULD, E W. 1968. Grass systematics. McGraw-Hill Book Company, New York. HILL, S. R. 1982. Vegetative apomixis (‘vivipary’) in Bouteloua hirsuta Lag. (Poaceae). Sida O(4)399 3357. NYGREN, A. 1954. oo in the angiosperms. II. Bot. Review 20:577 — 649. RADFORD, A. E., W. C. DICKISON, J. R. MASSEY, and C. R. BELL. 1974. Vascular plant systematics. Harper & Row Publishers, New York. STEBBINS, G. L. 1941. Apomixis in the angiosperms. Bot. Rev. 7:507— 542. ANISACANTHUS QUADRIFIDUS SENSU LATO (ACANTHACEAE) JAMES HENRICKSON Department of Biology California State University, Los Angeles, CA 90032, U.S.A. Studies of a third linear-leaved taxon of Anisacanthus from the Chihu- ahuan Desert Region {see Henrickson and Lott, 1982 for discussion of Antsacanthus linearis (Hagen) Henrickson and Lott, and Daniel and Henric- kson, 1982 for discussion of A. junceus (Torr.) Hemsl.} have lead to a re- evaluation of two wide-spread species: A. wrightii (Torr.) Gray and A. quadrifidus (Vahl) Nees. Anisacanthus wrightii and A. quadrifidus were dis- tinguished by Hagen in his 1941 monograph of the North American species of the genus on the basis of calyx size and calyx lobing. Anisacanthus quadrifidus, which ranges from San Luis Potosi south to Hidalgo, Puebla and Oaxaca, has calyces 7— 10 mm long with attenuate lobes 5-6 mm long while A. wrightii of Texas, Coahuila, Nuevo Leon, Tamaulipas has shorter calyces under 5 mm in length with acute lobes | — 3 mm long. In his monograph, Hagen named a new variety of A. wrightii from near Saltillo, Coahuila with narrow leaves and very short calyx lobes as A. w. var. brevilobus Hagen. similar, and even more distinct taxon with shorter calyces and even narrower leaves is here separated from Hagen’'s A. quadrifidus. The new taxon, from the states of San Luis Potosi and Zacatecas, was hidden within Hagen’s A. quadrifidus. A total of six specimens referable to the taxon were cited by Hagen (1941) in his specimens examined, but their measurements were neither included tn the species description nor in the key, and until his cited specimens were seen, recent collections of the taxon were consid- ered novelties. The taxon is here referred to by its manuscript name “potosinus.” Hagen (1941) considered vestiture, corolla size and particularly calyx characters as important features in distinguishing taxa of Anisacanthus. While A. quadrifidus and A. wrightii differ in some quantitative leaf and calyx features, they are similar in a large number of both qualitative and quantitative features including growth habit, stem and leaf vestiture, leaf texture, inflorescence and bract size, orientation, and structure, calyx vestiture and corolla size, structure and orientation etc. (see descriptions SIDA 11(3):286— 299, 1986. 287 below). In contrast, other North American species in the genus have some complex of character states that set them apart from these and other species (see Hagen 1941). In addition to the calyx characters indicated by Hagen, A. quadrifidus and A. wrightii also differ in leaf size and petiole length. Anssacanthus wrightii tends to have broader leaves [(7-)11— 20 mm wide} and longer petioles {(2.5-)5 — 10 mm long} than A. guadrifidus which typically has mature leaves only 3— 11 mm wide and petioles only |—4 mm long. The quantitative differences between these taxa are illustrated by means of bivariate plots in Fig. la, b, and by line drawings in Fig. 2. Figure la is a simple bivariate plot showing petiole length (in mm) as y-axis and maxi- mum leaf width (in mm) as x-axis for each of the four recognized taxa and an intermediate population. Leaf width-petiole length measurements were taken from 1 leaf per herbarium specimen (from 1 per collection when duplicates were present) from “typical” large leaves. While it is recognized that this neither represents a statistical sample nor necessarily represents the largest leaves present on the plant, it does reflect the characters availa- ble to a taxonomist working with herbarium material. Figure lb is a similar bivariate plot for calyx features with the y-axis showing calyx-tube length, the x-axis showing calyx-lobe length for the same specimens and taxa. Calyx measurements were taken from flowers at or near anthesis as calyx tubes are slightly accrescent. There always is some variation among these measurements on any specimen and the data presented here are intended to show only the trends in variation. Data in Fig. la, depicting leaf width vs. petiole length, show a continu- ous range of variation from the broad-leaved, long-petioled A. wrightir to the progressively narrower-leaved, shorter-petioled A. quadrifidus, and taxa brevilobus and “potosinus” with some overlap between each taxon. Figure 1b, showing calyx tube vs. calyx lobe lengths, illustrates a similar pattern. However, in these characters, A. quadrifidus has the longest calyx lobes, 2 to 4 times longer than the calyx tubes, and A. wrightii and taxon “potosinus” both have shorter calyx lobes, 1 to 2 times longer than the tubes. Calyx lobes of taxon brevilobus, in contrast, are as long as or shorter than their calyx tubes (Fig. 1b). When both data sets are considered, one sees that each taxon has its own unique quantitative characters. Anisacanthus wrighti has broader leaves and longer petioles; A. quadrifidus has the longest sepal lobes, relatively short petioles; taxon brevilobus has moderately narrow leaves, and very short calyx lobes; while taxon “potosinus” has very narrow leaves, small calyces but with moderately long lobes. These characters, however, are not discrete from those of the other taxa. Rather the character states are contiguous or ct S — oiler ay 8 5 \ ee ~~. & \ = mr z aN 5 | le e ° ou moO ff og va = - e 7 8 ba ~eo @ | = e 8 +, ‘o ON o x e? q|eso 2 s ™ ae A See ™ \ nae = — % re T T T T T T T oo] ~ Ne} wo Boul mn N onl ma NI HLONAT AHOT XATVO < 2 a e ra = 8 ef Le > w — oO p nn r = 8 BS me Lo 7 PERE : , anes a "8 s5ee2 0 [" z | On Fama Mm Lica: [ocms cms cme mec) i aqtaqade fe} Py PP PPPS ra . Le a eomo8 ioe 1” Lee =| E ee PoE ry ee e eX & e =. ofs Do a ® ®o0'00 T T T T T T T T T T T T { T T T T T T T o a 0 Lng re) w a) a] N La =) a 2 ~ ~ wo i foe) N re a = — cal = et a ~ - - = aa NI HLGIM Ava‘ 1g. 1. Bivariate plots of leaf and calyx characters of Amisacanthus quadrifidus. a. — Standard plot of petiole length (in mm) vs. maximum leaf width (in mm). Data obtained from largest leaf present in herbarium specimens of collections available for study. Designa- tion of infraspecific taxa are indicated as shown. Plants near Galeana, Nuevo Leon with leaves as in A. g. var. wrightti but calyces as in A. q. var. brevilobus are indicated as interme- diate between these taxa. b. — Standard plot of calyx cube length vs. calyx lobe length tn mm of calyces of flowers at or very near anthesis. Dashed lines indicate ratios of lobe vs. tube lengths where lobes are equal to tube length (1/1), two times tube length (2/1), and four times tube length (4/1). 289 slightly overlapping with those of related taxa (Fig. la, b). The fact that these four taxa share a large number of vegetative and floral characteristics and that other North American taxa in the genus differ to some degree from these taxa in various quantitative or qualitative characteristics supports the contention, presented here, that the four taxa are best ranked at the in- fraspecific level rather than as distinct species. Each taxon also has its own geographical range. Anisacanthus quadrifidus is known from arid deciduous scrub in southern Mexico from Oaxaca, Puebla north to the states of Mexico, Hidalgo and Queretaro from 1000 — 2400 m elevation (Fig. 3). In contrast, A. wrightii occurs in the Tamaulipan scrub from southern Texas through Coahuila, Nuevo Leon south into Tamaulipas mostly from 320— 900 m elevation. Although these two allopatric taxa are sometimes quantitatively similar, identifica- tion of herbarium specimens is never a problem as the taxa easily can be distinguished geographically. Hagen’s A. wrightii var. brevilobus occurs lo- cally in canyons and drainages east of Saltillo, Coahuila in the transition between semi-arid plains and chaparral scrub vegetation on the slopes from 1500 — 1800 m elevation (Fig. 3). There are populations combining the characters of A. wrightii and taxon brevilobus in the Sierra Madre Oriental about 110 km southeast of this area near Galeana and Pablillo, Nuevo Leon that have the short calyx lobes of taxon brevilobus and the long-petioled, broader leaves typical of A. wrightii. They also occur at an elevation more characteristic of taxon brevilobus —1700—2000 m. These intermediate plants are designated in Figs. 1, 3 by intermediate (half-darkened) symbols. Not all collections from this area have short calyx lobes however. The origin of the short calyx lobes may be related to past introgression between A. wrightii and the taxon brevilobus. The connecting link between A. quadrifidus and A. wrightii is provided by the narrow-leaved taxon “potosinus” which occurs along the southern border of the Chihuahuan Desert Region and in more montane areas within the southern portion of the Chihuahuan Desert proper from San Luis Potosi to northern Zacatecas (Fig. 3). The taxon occurs in an arid scrub mostly from 1800-2100 m elevation. With its very narrow, linear-lanceolate leaves and relatively short calyces, it appears very much like a more xeric- adapted derivitive of A. quadrifidus. While taxon “potosinus” is disjunct from A. quadrifidus in the south, its northern range shows a geographical link with A. wrightii. The question arises whether the short calyx-lobe character of A. wrighti var. brevilobus could have arisen from past introgression with northern populations of short-calyced “potosinus.” As noted above, populations of “hotosinus” currently occur at relatively high elevations along the southern 291 border of, and in scattered montane islands within the southern Chihu- ahuan Desert. It is probable that during the Holocene “potosinus” was more yeas and may have come into contact with other taxa of Anisacan- ee (1982) reports that while natural hybrids are not known in Anzsa- canthus, he has successfully made artificial crosses between A. wrightii and the longer-flowered A. /inearis and A. thurberi. The progeny of these crosses exhibited pollen stainability of 99 and 91 percent respectively. If one finds such high cytological compatibility between these diverse species, may we expect the same from the four taxa dealt with in this paper? The striking vegetative similarity between “potosinus” and the linear- leaved A. /inearis (Henrickson and Lott 1982) and A. junceus (Daniel and Henrickson 1982) of the northern Chihuahuan Desert is very apparent. Vegetatively these three taxa are difficult to distinguish, however, A. linearis and A. junceus belong to a separate lineage within the genus that has much longer corollas (40 — 56 mm long) with longer lobes and filaments (18—35 mm long) than those of the A. quadrifidus-wrightii taxa whose corollas and filaments measure 30 — 38 mm (rarely to 45) and 9— 11 mm respectively. They also differ in a number of other characteristics involving vestiture, pedicels, calyx structure etc., characters that are not in evidence in “potosinus.” Whether the linear-leaved characteristics of these taxa and our “potosinus” is due to convergent adapatation to xeric habitats or is due to past introgression with these taxa is unknown. But the lack of A. /mmearis and A. junceus characteristics in Dotosinus” suggests the independent devel- opment of linear leaves by “potosznus.” As noted above, the occurrence of “potosinus” in the Chihuahuan Desert was unknown to me until recently. A collection from northern Zacatecas (Henrickson 6222) was confused with A. janceus and illustrated as A. junceus by Henrickson in Daniel and Henrickson (1982, Fig. 1c). As noted in the paper it differed from true A. junceus in its more distinctly glandular calyx. Fig. 2. Line drawings of calyces and leaves of varieties of Anisacanthus quadrifidus, a- Bye Aseg var Ces a. — Calyx showing area: bract. Note position of line abscission above persisting bract base and long calyx lobes. b. — Leaves. Note relacively short petiole. Both Lott & Wendt P35 (TEX). c-d. — A. q. var. potosinus . c. — Calyx showing moderate long lobes. d. — Leaves are agate! narrow in this variety. Both Wendt et al. 2235 (TEX). e-f. — A. q. var. wrightii. e. — Calyx with moderate long lobes. Note decurrent lines of decurved hairs. f. — Leaves. ae large size, long petioles. Both Smith 229 (TEX). g-h. — A. q. var. brevilobus. g. — Calyx showing characteristic short lobes. h. — Leaves which tend to be narrow. Both Waterfall & Wallis 13246 (F). All calyces and ae drawn to scale. Magnification indicated in a and b. Drawing by K. Cook. 292 While some questions remain as to the role of past introgression in northern Mexican Anisacanthus, the data presented here indicates that A. quadrifidus and A. wrightii: (1) share a large number of characteristics; (2) as a unit they are well distinguished by these characteristics from other species in the genus (see Hagen 1941); (3) their distinguishing characteris- tics tend to grade into those of other taxa (Figs. 1, 2); (4) while each taxon has its own geographical range, there is some indication of introgression in the short calyx-lobed populations of A. wrightit near Galeana in Nuevo Leon. To reflect this pattern, the taxa in Anisacanthus quadrifidus and A. wrightii are here combined into a single species which takes the name of the oldest specific epithet and the type of the genus Anisacanthus quadrifidus (Vahl) Nees and the other taxa are reduced to varieties of this species. ANISACANTHUS QUADRIFIDUS (Vahl) Nees Moderately to strongly branched, erect to spreading shrubs 0.5 — 1.5 (-2.6) m tall; young stems with internodes (5-)10—45(-65) mm long, 0.7—1.5 mm wide, terete, non to weakly striate, yellow-green, with decurved to more or less straight hairs 0. 1 — 0.4 (-0.6) mm long in 2 broad decurrent lines extending from leaf-petiole margins across a stipular ridge and down the complete internode, this well developed or sparse, otherwise glabrous or sometimes sparsely hirtellous, rarely sparsely glandular but soon to tardily glabrate; older stems tan to grayish, with bark not peeling in sheets. Leaves opposite, lanceolate, linear-lanceolate to linear, (8-) 15 — 35(-65) mm _ long, (0.8)1.0— 16(-20)[-30] mm wide, (reduced above), tapering to an acute tip, rounded to cuneate, sometimes oblique, subsessile or with petioles (0.5-)1 — 7(-10) mm long at base, unevenly en- tire at yellowish, sometimes revolute (when dry) margins, glabrous to glabrate on both surfaces except for few to many antrorsely curved hairs along lower impressed midrib above and along lower leaf and petiole margins; leaf-blades yellow-green, more or less glandular-punctate, with midvein raised, yellowish beneath; leaves abscissing at base above a yellow- ish, broad, persisting leaf-base 0.3 —0.5 mm high, 0.8— 1.6 mm wide, with abscission line often appearing black. Flowers | (-3) at upper nodes borne on elongate, slender, secund, upwardly arching, spicate racemes along terminal (2-)5 — 14 cm of seasons shoots; inflorescence internodes 2— 14(-18) mm long, vestitured as stems or more hirtellous; bracts (re- duced leaves) lanceolate-deltate, 2—7 mm long to 0.5 — 1.3 mm wide at base, acute to attenuate; paired bracteoles similar, 2—3 mm long, both rather soon caducous leaving distinct protruding bract-bases topped with conspicuous tan, corky-rimmed abscission scars; peduncles broad, (0.5-) 1 — 2(-3) mm long; calyces (3-)4 — 8(-10) mm long, tube 0.9 —2.5(-3) 295 mm long, lobes lanceolate, oblong-lanceolate to deltate, (1.2-)2 — 6.5 mm long, 0.9 — 1.2 mm wide at base, lobes 0.8 — 4.8 times as long as tube at anthesis; pedicels, calyces hirtellous with erect hairs more or less 0.03 mm long particularly at tube base, and with more or less conspicuous yellow- headed, stipitate glands 0.05 — 0.1 mm long with heads 0.03 — 0.07 mm wide outside, margins and tips of lobes weakly pilose with hairs 0.1 — 0.2 mm long, these often tufted at tip, inner lobe surfaces weakly to densely sessile glandular and weakly strigose with scattered antrorse hairs near tube; corollas red to orange-red, often yellowish where shaded in bud, (29-) 32 — 38(-45) mm long, tube expanded at base around ovary, slightly ampliate distally, 19— 25(-28) mm long, upper (posterior) lobe 14 — 17 mm long, separating 3-7 mm proximal to lower 3-lobes, lower 3 lobes 9 — 13 mm long, all lobes flaring, oblong to strap-shaped, 2 — 3 mm wide, acute; corollas pilose-pubescent externally with multicellular, tapering, retrorsely bent-spreading hairs 0.2 —0.3 mm long; stamens 2, inserted in lower lobe distal to separation of upper lobe; filaments glabrous (7-)9 — 11 mm long, orange-yellow, glabrous, straight; anthers (2.3-)2.7— 3.4 (-3.7) mm long, thecae subequal to equal, parallel, separate for 1 — 1.5 mm at base, sporangia reddish, connective dull yellowish; style 26 — 35 mm long, glabrous, exserted; stigmatic lobes ca 0.2 mm long, acute. Capsules 13 — 16 mm long, glabrous, tan, basal flattened stipe 5 —8 mm long, head ovoid, 6-8 mm long, apiculate, retinacula 2.5 — 3 mm long; seeds (2-)3 — 4, notched basally, 4.5 — 5.6 mm long, 4— 4.5 mm wide, to 1 mm thick, at maturity brown, more or less bullate and tuberculate on both faces, or mainly on inner face, with thickened margins. Four varieties are separable by the following key: A. Calyx lobes (1-)1.5 — 4.3 times as long as calyx tubes at anthesis; leaf- blades linear to lanceolate B. Petioles of larger leaves 0.5 — 3(-4) mm long; larger leaf-blades linear- lanceolate, to 1 — 9(-13) mm wide. Leaf-blades lanceolate, to (3-)4.5—9(-13) mm _ wide; petioles (1-)2 — 3(-4) mm long; calyx lobes (1.6-)3 — 4.3 times as long as calyx cubes at anthesis; Hidalgo, Queretaro, south to Oaxaca Sentai eked i te Ss wish Seah a Mech eens Bt eke . q. vat. quadrifidus CC. Leaf-blades linear-lanceolate, to (0.9-)1.5—3(-5) mm_ wide; petioles 0.5 — 1.2 mm long; ~~ lobes (1.1-)1.5 — 2(-3) times longer than tubes at hesis; San Luis Potosi to A. q. vat. potosinus BB. Petioles of ee ae (3.5-)4—7(-11) mm long; larger leaf- sisaees lanceolate to lance-ovate, 7 — 15(-20) mm wide; calyx lobes 2.3 times as long as calyx tubes at anthesis; Texas, Coahuila, vie Leon tomlamaulipas nek youd header eunas doboeeser . q. vat. wrightii 294 AA. Calyx lobes 0.5 — 0.8(-1.2) times as as C a Awe at anthesis; leaf- blades lanceolate to linear-lanceolate, 2.3 — (-5.5) mm_ wide; Coahuila (southeast of Saltillo, and a ee with A. q. var. wrightii near Galeana, Nuevo Leon) ................0005. A. q. var. brevilobus 1. ANISACANTHUS QUADRIfiDUS (Vahl) Nees var. QUADRIfipUS, Linnaea - our 1842. ees cai Vahl, ry Pl. 1:124. 1804, based on Justicia Cav 199. 1793, non Aublet 1775. Anssacanthus sere (Vahl) a ae U.S. Nat. Herb. 23:1343. 1926. Typr: Nuevo Hispanica. Lecrorype here ain MA (photo!); IsoLECTOTYPE: F! Antsacanthus ede (Salisb.) Nees in DC. Prod. 11:445. 1847. Justicia ae Salisb., Parad. London. pl. 50. 1806.Type: HoLoTyPe: plate 50 in Parad. on. (see Hagen a p. 405, a.). Correspondence with K and BM ee no specimens referable to this collection exist. It was probably described and illustrated al. from garden materia Justicia a Hort. ex Nees in DC. Prod. 11:445. 1847. Type: (in herb. Hort. Berol. fide Nee). Justicta eee ae ex Nees in DC. Prod. 11:445. 1847, non L. 1753, pro syn. (in r fide Nee). Sploaise arses a. Bull. Herb. Boiss. 2:546. 1894. Type: MEXICO. . Tlacolula, prope Zoquitlan, Jun 1888, Se/er 76 (HOLOTYPE: B, oan ae PHOTOTYPES: GH, MICH, NY fide Hilsenbeck, pers. comm.). Plants strongly branched; stems and inflorescence with decurrent lines of erect to decurved hairs (0.03-)0.1—0.2(-0.4) mm long, otherwise glabrous; petioles (1-)2—3(-4) mm long; leaves lanceolate, (1.4-) 2.0 — 4.0(-5.0) cm long, (3.0-)4.5 — 9.0(-13.0) mm wide, margins more or less revolute when dry; calyces at anthesis (3.4-)4 — 7(-8) mm long, lobes (3.5-)4 —5.5(-6.5) mm long, tube 1.5 —2 mm long, lobes (1.6-) 3 — 4.3 times as long as tubes, sparsely to densely hirtellous and with stipi- tate glands 0.06 — 0.12 mm long, their glands 0.04 — 0.06 mm wide out- side, lobes densely glandular inside; corollas 33 —40 mm long; anthers 2 3.4 mm long. Central Oaxaca to Puebla, Mexico, Hidalgo, Queretaro; 1000 — 2400 m elevation (Figs. 2a-b, 3). Anmsacanthus quadrifidus var. quadrifidus is readily recognizable by its long, lanceolate sepal lobes and relatively narrow leaves with short petioles. In describing Justicia quadrifida, Vahl (1804, page 124) cites Cavanille’s (1793) Justicia coccinea as a synonym and appears to make a su- perfluous name for the taxon in a manner contrary to ICBN Article 63. However, on page 120, Vahl recognizes Aublet’s (1775) Justicia coccinea as a species indicating that he considered the Cavanille name to be a later tautonym. Representative specimens: MEXICO. Oaxaca: 1.4 mi E of Tlacolula, 1670 m, 13 Nov 1980, Fryxell and Lott 3415 (TEX); road to Yagul, between Oaxaca and Mitla, | Mar 1960, 295 T 105 102 99 96 O + 24 @) "8 + 21 214 +18 7 ra ANISACANTHUS QUADRIFIDUS @ VAR. QUADRIFIDUS O VAR. POTOSINUS See VAR. WRIGHTII O VAR. BREVILOBUS i VAR. BREVILOBUS VAR. WRIGHTII ve 102 99 96 Fig. 3. Distribution of varieties of Anisacanthus quadrifidus in Mexico and Texas. 296 Carlson 3698 (F — 2 sheets); Oaxaca Valley, 5000 ft, 7 Nov 1894, Smith 713 (F). PUEBLA: San Bartolo, 15 km al NW de Tehuacan, 1700 m, 9 Jan 1959, Rzedowski 9501 (TEX); 7.7 km al N de Tehuacan, 1900 m, 13 Sep 1980, Lort and Wendt P35 (TEX). Mexico: Cerro de Santa Cruz, N de Sierra de Guadalupe, 2300 m, 2 Nov 1963, Matuda 37329 (LL—2 sheets). HiDALGo: Cerros al N de Pachuca, 2450 m, 18 Sep 1966, Rzedowski 23177 (TEX). 2. ANISACANTHUS QUADRIfiIDUs var. potosinus Henrickson var. nov. A A. quadrifida vat. quadrifida foliis lineario-lanceolatis (non lanceolatis), (0.9-)1 — 3(-5) [non (3-)4.5 — 9(-11)} mm latis, petiolis 0.5 — 1.2 [non (1-) 2—4] mm longis differt; a A. linearea et A. juncea corollis 29 — 36(-43) (non 40 — 56) mm longis, filamentis brevioribus id est 7— 11 (non 18— 35) mm longis differt. Plants strongly branched, twiggy, erect to widely spreading, 0.5 — 1.5 m tall, often wider than tall; stems, inflorescence glabrous or with decurrent lines of decurved hairs 0.03 —0.1 mm long, sometimes more striate; petioles 0.5 — 1.2 mm long; leaves linear-lanceolate to linear, (1.5-) 2—4(-5) cm long, (0.9-) 1.5 — 3(-5) mm wide, attenuate, at margins of- ten more or less revolute with sparse decurved hairs, otherwise glabrous; leaf-blade typically conduplicate along midrib; calyces at anthesis 3.5 — 5(-5.5) mm long; lobes 2 — 3(-4) mm long, tube 1.3 — 2 mm long, lobes (1. 1-)1.5 — 2 (-3) times as long as tube, weakly to strongly hirtellous and with stipitate glands 0.06 — 0.12 mm long, their glands 0.05 — 0.07 mm wide, mostly glandular and strigose inside lobes, marginal hairs 0.2— 0.3 mm long; corollas 29 — 36 (-38) mm long; anthers 2.8 — 3.2 mm long. San Luts Potosi to northern Zacatecas. 1800 — 2100 m. (Figs. 2c-d, Tyre: MEXICO. San Luis Porosi: Mpio. Villa de Reyes 3.9 km. al O. de la Carretera San Luis Potosi-Queretaro, por Carretera que va a Villa de Reyes. Km 20 de carretera S.L.P.-Queretaro, 1950 m, 8 Oct 1979, J. Garcia, T. Wendt and E. J. Lott 1297 (HOLOTYPE: MEXU; tsorypes: TEX, CHAP and to be distributed. d: MEXICO. San Luis Poros: Chiefly in region of San Luis Potosi, 6000 — 8000 ft, 1878, Parry and Palmer 706 (F); San Luis Potosi, 1879, egies 367 (F); gravel washes, Bocas, 17 Aug 1891, Pringle 3820 (F); Zaragoza, 2050 m, 7 Jul 1954, Rzedowski 3516 (F); 13 mi NE of San Luis Potosi, 20 Aug 1954, eas 15679 (F); same location as type, 8 Oct 1979, Garcia P. et al. 1297 (TEX, CHAP); 11 Oct 1979, Wendt et al. 2235 (TEX, CHAP); on road between Real de Catorce and Doc de aes ca 2100 m, near 23° 43’N lat, 100° 52’°W long, 18 Sep 1980, Henrickson & Bekey seas 18616, 18617 (TEX). Zacarecas: sin loc., Lloyd 10 (FE); 0.4 mi S of Zacatecas-Coahuila border along Hwy 54, near 24° 43’N lat, 101° 13’W long, 6100 ft, 30 Aug 1971, eee 6222 (TEX). 3. ANISACANTHUS QUADRIfIDUS var. wrightii (Torr.) Henrickson comb. NOV. Drejera wrightit Torr. in U.S. and Mex. Bound. Surv. Bot. 123. 1859. Anisa- canthus wrighti (Torr.) Gray, Syn. Fl. N. Am. 2(1):238. 1878. Type: UNITED STATES. New Mexico: May — Oct, 1848. C, Wright 435 (HOLOTYPE: NY; ISOTYPE: X!). Fig. 4. Anisacanthus quadrifidus vat. potosinus Henrickson. a.— Stem showing linear leaves and flowers. This was drawn from a dried specimen and several flowers have fallen off distorting the secund nature of the inflorescence. b. — Node with leaves showing line of abscission at base and distinctive stipular ridge connecting opposite leaf margins. c. — Calyx with associated bractlets. Calyx shown in dried condition with lobes flaring. d. — Corolla showing expanded base chat surrounds ovary, strap-shaped lobes and position of anthers and style. e. — Anther, adaxial view (left) and abaxial view. f. —Open capsule showing retinacula. All from Garcia P. et al. 1297 (TEX). Scale bars = 1 cm except in e where bar = 1 mm. Drawing by B. Angell. 298 Plants moderately branched, erect to spreading; stems mostly with dense decurrent lines of decurved hairs (0. 1-)0.2 — 0.4(-0.6) mm long; in- florescences rather strongly hirtellous with erect hairs 0.03 mm long with or without decurrent longer hairs; petioles (3-)4 — 7(-11) mm long; leaf- blades lanceolate to broadly lanceolate (25-)30—50(-62) mm_ long, 7 — 15(-20){-30] mm wide, margins mostly revolute when dried; calyces at anthesis (2.6-)3.5 — 4.5 (-5) mm long, lobes (1.3-)1.8 — 3.2(-4) mm long, tubes (1.2-)1.4 — 2(-2.0) mm long, lobes (0.9-)1.1 — 2.3 times as long as tubes, both hirtellous and with stipitate glands 0.03 —0.08 mm long, their glands 0.03 —0.06 mm wide, marginal hairs 0.1 —0.2 mm long; corollas 31—40 mm long; anthers 2.5 — 3.2 mm long. Texas, Coahuila, Nuevo Leon and Tamaulipas. 300-900 m. (Figs. 2e-f, 3). Representative specimens: UNITED STATES. Texas. Uvalde Co.: Con Can, along Rio Frio, 22 Jun 1963, Correll and Wasshausen ca (LL). MEXICO. Nuevo Leon: Lampazos, 21 Jun 1937, Edwards 282 (E TEX); 12 mi W of Linares, 1200 ft, 26 Jul 1956, Fearing and Thompson 25 (TEX); behind Cola de ate Monterrey, 29 May 1960, Smith M229 (EF TEX); Alamar to Puerto Blanco, 15 mi SW of Galeana, 23 Jul 1934, Maller and Maller 1197 (TEX); Rayones, 900 m, 9 Jun 1980, Hinton 17824 (TEX). CoAnulta: uzquiz, Sabinas, 7 Nov 1936, Marsh 408 (TEX); Rancho Babia, 18 May 1938, Marsh 1209 (FE, TEX). TAMAULIPAS: vicinity of San Jose, La Vegonia, 20 Jul 1930, Bartlett 10526 (E TEX); vicinity of Victoria, 320 m, 1 Feb-9 Apr 1930, Palmer 120 (F). 4. ANISACANTHUS QUADRIfiDUS var. brevilobus (Hagen) Henrickson comb. nov. salisihiae wrightu var. brevilobus Hagen, Ann. Missouri Bot. Gard. 28:400. 1941. Type: MEXICO. Coanuita: Chojo ae 27 mi SE of Saltillo, 16 Jul 1905, Palmer 719 (HOLOTYPE: NY!; tsorypes: CB, O). Plants moderately branched, spreading to erect; stems more or less striate, mostly with decurrent lines of decurved hairs 0.1 — 0.3 mm long; inflorescences more or less decurrently, sulcately grooved, with short decurved hairs in sulci, otherwise glabrous; petioles (1-)1.5—2.1 mm long; leaf-blades narrowly lanceolate to linear-lanceolate, (17-)20— 35(50-) mm long, 2.3—4.5(-5.5) mm wide, margins more or less revolute; calyces at anthesis 2.5 — 3.6 (-4.0) mm long, lobes 1— 1.7 mm long, tubes (1.3-)1.5 — 2 mm long, lobes 0.5 — 0.8(-1.2) times as long as tubes, hirtellous and with mostly many stipitate glands 0.03 — 0.06 mm long, their heads 0.04 — 0.05 mm wide, strongly glandular and strigose inside, marginal hairs few, limited to a terminal tuft of hairs 0.1 —0.2 mm long; corollas 33 — 37 mm long; anthers 2.2 — 2.9 mm long. Coahuila, in canyons southeast of Saltillo, 1700 — 2000 m. (Figs. 2g-h, 3). Representative specimens: MEXICO. Coanuita: Paso del Aguila, E of Saltillo, 12 Jul 1946, Barkley et al. 16M496 (TEX); same location, 2 Jul 1947, Barkley et al. 7244 (TEX — 2 sheets); 12 air mi E of Saltillo, lower Canon de Chorro, 1900 m, 27 Sep 1980, 29) Henrickson and Bekey ie 18644 (TEX); 18 mi NE of Saltillo, 6 Aug 1957, Waterfall and Wallis 13246 (F— 2 shee Collections from near Galeana, Nuevo Leon have relatively broad leaves as in A. q. var. wrightii and short sepals as in A. q. var. brevilobus and are considered intermediate between the two taxa. While these characters tend to be uniform through this area, some specimens from the area have longer ia (see Muller and Muller 1197). ecimens examined: XICO. Nuevo Leon: Taray to Alamar, about 15 mi SW of ee 20 Jul 1934, ee and Muller 1108 (E TEX); Hacienda Pabillo, Galeana, 27 Aug 1936, Taylor 262 (KE TEX); Mcpo. Galeana, 5400 ft, 2 Aug 1939, Chase 7731 (F—2 sheets); 1 mi W of Galeana on road to Cerro Potosi, 5400 m, 16 Sep 1980, Henrickson and Bekey 18540 (TEX). ACKNOWLEDGMENTS I thank Tom Daniel, Tom Wendt, Emily Lott, Kevin Nixon for dis- cussions regarding these taxa, M.C. Johnston for Latin diagnosis, the Plant Resources Center at the University of Texas for use of facilities, K. Cook for delination of Fig. 2, Bobbi Angell for Fig. 4, and TEX, LL, and F for use of herbarium specimens. REFERENCES AUBLET, J. 1775. Historie des plants de la Guiane Francoise. Vol. DANIEL, T. E 1982. Artificial interspecine hybridization of three species of Anisacanthus (Acanthaceae). J. Ariz. : DANIEL, T. FE and J. HENRICKSON aes On the recognition of Anisacanthus junceus (Acanthaceae). Brittonia 34:17 HAGEN, S. H. 1941. A revision na a ak American species of the genus Amisacanthus . Ann. Missouri Bot. Gard. 04. HENRICKSON J. and E. J. fa 1982. New combinations in Chihuahuan Desert Anisacanthus (Acanthaceae). Brittonia 34:170— 176. VAHL, M. 1804. Enumeratio Plantarium. Vol. 1. TAXONOMY OF FLYRIELLA (ASTERACEAE-EUPATORIEAE) M. M. BAKER 714 Churchill Drive, Chapel Hill, NC 27514, U.S.A. B. L. TURNER Department of Botany, University of Texas, Austin, TX 78713-7640, U.S.A. ABSTRACT A taxonomic treatment of the genus F/yrie//a is rendered fclnding information relating to its generic relationship, phyletic arrangement of its four species, chromosomal, an chemical data. Illustrations of the species are presented along with a map showing dis- tributions. INTRODUCTION Flyriella King & Robinson is a North American genus of four species which is largely confined to north central and northeastern México, but extends into the United States in the border regions of Trans-Pecos Texas. The genus was proposed by King & Robinson (1972b). It is named in honor of Dr. David Flyr, plant systematist from The University of Texas (Turner 1972), who placed the generitype (Eupatorium parryi) in the genus Brickellia (as B. shineri) after removing it from its original position in Eupatorium. This placement by Flyr (1968) was based upon a number of morphological features which he took to be brickellioid. Flyr noted that B. shineri did, however, differ from other species of Brickellia (with the excep- tion of B. fendleri Gray) in having only five-ribbed rather than ten-ribbed achenes as is typical of Brickellia. The dissociation of Expatorium parryi from Brickellia by King & Robinson left B. fendleri as anomalous in Brickellia in possessing five- ribbed achenes. Gray, as noted by Flyr, was uncertain as to the proper placement of B. fendleri; it remained for King & Robinson (197 2a) to place the latter in a newly erected monotypic genus, Brickelliastrum. In their es- tablishment of Flyriella, King & Robinson recognized five species: F. chrysostyla, EF. sphenopoda, F. leonensis, FE. parryi, and F. stanfordii. In 1982, they added an additional species, FE Aarrimanit. In the present treatment we have recognized only four species: 1) F parryi, a wide-ranging, variable, taxon that includes F. chrysostyla and F SIDA 11(3):300— 317. 1986. 301 sphenopoda, 2) FE. leonensis, a weakly differentiated taxon from near Monterrey, México; 3) F. stanfordii, a species along the Sierra Madre Orten- tale; an FE. harrimaniit, a very distinct local endemic in the lower montane regions about Gomez Farias, Tamaulipas. CHROMOSOMAL STUDIES Chromosome counts are available for only two species, F /eonenszs and F parryi, both diploid, as indicated in Table 1. The base number x = 10, 1s not known co occur in Brickellia (King et al. 1976). The latter genus has a base chromosome number of x = 9. Expatorium has a base number of x = 10, as does Alomia (Mabry et al. 1981), near which we would position Flyriella. FLAVONOID STUDIES Mabry et al. (1981) have made the only chemical study of F/yriella. They examined three of the four species, (FE parry, F. leonensis, and F. stanford:1) as shown in Figures 1 and 2. Unfortunately, only two populations of F parry: were examined and only one population each of EF /eonensis and EF. stanfordit. All the species examined produce glycylated quercetins. Flyriella stanford differs significantly aul E. parryi and E_ leonensis in producing only monoglycosidic form Mabry et al. (1981) treated F. sphenopoda as a distinct taxon, and the chromatographic profiles of plants referable to this name lacked the monoglucosides (Fig. 1). However, in view of the considerable morpholo- gical variability in its flavonoid components, no significant taxonomic im- port is given the chromatographic profile concerned. SPECIES RLATIONSHIPS WITHIN FLYRIELLA Flyriella parryi, a wide-ranging variable species, appears to be most closely related to F /eonensis and is sympatric with it in the vicinity of Taste 1. Chromosome Numbers in F/yrie/la Species Chromosome Number Location and Voucher (2n) F. leonensis 20 MEXICO: Nuevo Leon. 17 mi by oad W of Horsetail Falls. Turner 10037 (LL). EF. parryi 20 MEXICO: Coahuila. Above Las Delicias. Powell 2699 (TEX) F. parry! 20 MEXICO: Nuevo Leon. 5 mi W of El Alamo. Turner 10007 (LL). 302 Monterrey, Mexico. Additional work in this area is needed to ascertain if they grow in close proximity which might occasion hybridization. Chemi- cal analysis reveals that F /eonensis is almost identical with F. parryi, sharing three of the four flavonoids concerned: quercetin 3-arabinoside, ombuoside, and tamarixetin 3-rhamoglucoside (Mabry et al. 1981, Figs. Flyriella stanfordi is believed to be more closely related to FE harrimanii than are F. parry: and FE. leonensis. Both FE. stanfordii and F. harrimanii possess short glandular trichomes and winged petioles and both occur in the more mesic habitats of southern Tamaulipas. Nevertheless, FE stanfordii is probably more closely related to E parryi and EF. leonensis than it is to F harrimanti to judge from its involucral characters. Relationships among the four species are shown in Figures 3 and 4. This is largely derived from consideration of hypothetical polarized character states as shown in Table 2. That is, we have erected an imaginary primitive state for the characters concerned, based on a wide experience with what appears to be specialized characters in the tribe Eupatorieae generally. We do not believe that arbitrary selection of an out-group for cladistic analysis OC Flyriella parryi F. leonensis F. sphenopoda (=F. parryi) F. stanfordii ig. 1. Two-dimensional chromatographic flavonoid patterns for Flyrtella (modified from Timmerman 1980). 303 — 2. Hypothetical es versus advanced character states in Flyriella (P = FE parryi, L = F ee S = EF stanfordit, H FE. harrimanit). PRIMITIVE (Q) ADVANCED (1) SPECIES P LSH 1. Leaves with glandular trichomes 1. Leaves without glandular crichomes 0 100 2. Petioles alate (clearly winged) 2. Petioles not alate (poorly winge 1 100 3. Involucral bracts not scarious 3. Involucral bracts scarious 1 110 4. Outer involucral bracts leafy 4. Outer involucral bracts reduced 1 110 5. Involucral bracts numerous 5. Involucral bracts re 1 110 6. Numerous florets/head 6. Fewer florets/head 2210 5 = 0 30 =49 =" 1 = = 2 7. Corolla flared 7. Corolla tubular or constricted 1 100 8. Carpopodia mostly not contorted 8. Sal elaae mostly contorted 1 100 9. Habitat mesic 9. Habitat xeric 2 100 Pe OY) Totals 10 10 4 0 at this time would be a meaningful phyletic exercise, although it might reduce the circularity of our premise. Based upon the characters in Table 2, F Aarrimanii appears to be the most ptimitive species in the genus. Its leafy involucral bracts as well as other “primitive” or less advanced features suggest that the remaining taxa evolved out of similar ancestral prototypes, culminating in the more xeric FE. parryi (Fig. 4) whose involucre and florets evolved so as to "mimic’” species of Brickellia. As can be seen from the cladogram (Fig. 4), character states 2, 6, 7, 8, and 9 are synapomorphies linking F /eonensis and F. parry:. Character states 9? and 1 are automorphies for FE parryi and F. leonensis respectively. Charac- ter states 3, 4, 5, and 6 are synapomorphies for the three species FE. leonensis, E. parryi, and F. stanfordii. Flyriella harrimanii appears to be, as noted above, the most primitive species. It will be interesting to obtain additional chemical data for F. stanfordi and F. harrimanii, for it appears from current analyses that the more highly evolved diglycosidic flavonoids might be missing from the latter species, substantiating the suggestions made here, which are largely based upon morphological data. GENERIC RELATIONSHIPS OF FLYRIELLA In his transfer of Expatorium parryi into the genus Brickellia, Flyr (1968) largely emphasized its narrow, erect-lobed, constricted corollas and FLYRIELLA PARRYI Quercetin 3-glucoside ine) ul Quercetin 3-arabinodise 15 % HOAc 3 = Tamarixetin 3-rhamnoglucoside 4 = Ombuoside TBA . 2. Two-dimensional i profile of F/yriella parryi: TBA = t-BuOH-HOAc-H;0O, 3: : i "(modified from Mabry et al. 1981). narrow, green and white, seriate phyllaries. Indeed, in these and in certain microfeatures, chiefly the hirsute stylar node, it is very similar to Brickellia. Flyriella, however, lacks fringed pappus setae and, of course, possesses fewer ribs on its achenes and has a base chromosome number of x = 10. Nevertheless, F/yrie//a shares a close relationship with Brickellia and both genera are placed in the subtribe Alomiinae by Robinson & King (1977) Mabry ect al. (1981), upon completion of a series of chemical analyses, found no support for a close link between F/yriella and Brickellia. They note that the flavonoid chemistry of Flyriella parryi and Brickellia laciniata (Timmerman et al. 1979) reveal very different patterns. Brickellia species contain 6-methoxylated glycosides, sulfates, and aglycones, none of which is found in F/yriella (Figs. 2, 3). Turner, in Mabry et al. (1981), suggest that Flyriella is more closely related to Alomia rather than Brickellia. There is much evidence to support this hypothesis, and it comes from several lines of investigation, as noted elow 305 ig. 3. Wagnerian diagram showing eae sclacionshiie among the four species of F/yriella. Character states for the construction shown in Table In addition to the chemical data, two of the four species of F/yrzella have been shown to be diploid with 27 =20 or x = 10 (Table 1). This base number also characterizes Alomia (Mabry et al. 1981). Brickellia, on the other hand, has a base chromosome number of x =9. Both Flyriella and Alomia are spring-flowering. Brickellia is predomi- nately a fall-flowering genus as are the majority of its close relatives. Be- cause most Eupatorioid genera of the Chihuahuan Desert region are fall- flowering, it can be suggested that F/yriella harrimanu, which appears to be the most primitive member of the genus on morphological grounds, 1s the species which lies geographically nearest the hypothetical center-of- origin of the group. The more advanced features of F. parry: are perhaps adaptations to drier habitats. Even so, label data and observations in the field by the junior author show the species to be largely confined to more mesic sites in the desert regions (along perennial streams mostly in cool, moist Canyons). In short, since Flyriella appears not to be closely related, phyletically speaking, to Brickellia, and since it has no extant desert relatives from which it might have evolved, the most likely ancestral candidate at this time appears to be A/omia or a close relative of the latter (Table 3). Alomia may have had a double origin, partly Ageratoid and partly Trichonioid, as noted by both B. L. Robinson (1913) and Robinson and King (1977). Indeed, the junior author has suggested that F/yrie//a and Alomia might be better positioned in the subtribe Ageratinae. This implies that the 306 Taste 3. Comparison of selected characters of Alomia and Flyriella. Brickellia i 1 here since it is exceedingly variable and such a litany would be beyond the scope of this study. If compared, i would differ from A/omia and F/yriella largely by its base chromosome number, x =9, fall- flowering. and more xeric habitat ALOMIA FLYRIELLA Distribution Seeane regions of Subtropical to drier temperate xico & South America regions of north centra and northeastern México Habitat Mesic esic to semi-xeric Habit Perennial herbs Perennial herbs Stem vestiture Long non-glandular Long non-glandular trichomes or glandular trichomes or glandular trichomes trichom Leaves Mostly opposite Mol Opposite Peciole Alate or non-alate Alate or non-alate Phyllary shape Mostly acute to Mostly acute to acuminate acuminate Corolla Flaring or tubular Flaring, or tubular with or without apical constriction Stylar node None Present Carpopodium Contorted Contorted or “turbinate” Achene 4 —5-ribbed 4—5-rib Pappus bristles None or reduced Well- “developed hea sone no. x =10 x =10 nchesis Spring Spring morphological characters which relate it to Brickellia (discussed above) are convergent. Additional study, especially chemical, is needed to confirm or refute the close relationship of F/yriedla and Alomia. TAXONOMY FLYRIELLA K. & R., Phytologia 24:69. 1972. Perennial herbs, 0.5 — 2.0 m tall; stems erect, puberulent to viscid from a tap root or short rhizome. Leaves opposite below, often becoming alter- nate above. Capitulescence a terminal corymboid panicle. Heads turbinate to hemispheric, the florets yellowish-white. Involucral bracts in 3 — 5 im- bricated series, striated green and white, occasionally tinged with purple. Receptacle plane, naked. Corollas tubular, gradually flaring upward but often constricted just below the lobes. Lobes 5, short, narrowly triangular. Anthers included, appendages ovate. Style branches linear, yellowish- white, sometimes turning rusty colored upon drying. Achenes 4—5- ribbed, sparsely pubescent, carpopodium short to elongate, often con- torted. Pappus of 20 — 40, white, ciliate setae. Base chromosome number, x =10 Type species: Expatorium parryi A. Gray ry | \ A A ie stanfordii ake harrimanii Hypothetical ancestor Fig. 4. Cladogram constructed from results obtained from computerized method (Wagner 78, Version 25/8/79, James Farris, State University of ‘ York, Stony Brook, IBM 370/178 Computer). Data utilized based on character states from Table *Change in characters number 6 & 9 from @incacner states | & 2. KEY TO SPECIES A. Outer involucral bracts somewhat leaf-like, as long or longer (10 — 15 mm) the innermost bracts; florets 70— 100 per head ..........-- L. FE harrvimani A. Outermost involucral bracts not leafy, much reduced (3-6 mm long); florets 10 — 49 per head (B) B. Corolla throat narrowly funnelform, not constricted below the lobes; petioles conspicuously winged see montane subtropical areas of southern Tamaulipas and Nue OTM 5 8 sth ech eck gees . E stanfordii B. Corolla throat tubular (ca 0.5 mm ee not conspicuously pee below the lobes; petioles only partially winged, if at all; montane mostly desert areas of north central México (C) 308 C. Stems and branches of capitulescence with glandular trichomes or these intermixed with long, straight mostly stiff hairs; widespread variable species of north central México and adjacent areas of Texas C. Stems puberulenc, without glandular trichomes, the hairs crisped and variously recumbent, restricted species about Monterrey, México . 1. Eo leonensis I. FryrigttA HARRIMAN K. & R., Phytologia 50:380. 1982. Ter: MEXICO. Tamaulipas: on rout B-5, 3 a mi E of Gomez Farias, across from gravel pit, 600 ft, woods on pitted eae 31 Mar 1975, Harriman et al. 10698 (HOLOTYPE: US!). Erect perennial herb to 200 cm tall. Leaves opposite; blades ovate- deltoid, 4.0— 13.5 cm wide, 5— 17 cm long, irregularly serrate, both surfaces glandular; petioles 3— 11 cm long, glandular pubescent, winged. Capitulescence in ultimate pedunculate units of | — 3 heads, each head on a bracteolate pedicel 1 — 3 cm long. Involucre campanulate, 1.5 — 2.0 cm long, ca 2 cm wide; bracts herbaceous, imbricate in 3 — 5 series, glandular, 10—15 mm long, ca 1.4 mm wide. Florets 70— 100; corolla white, funnelform, ca 4 mm long, ca | mm wide; lobes narrowly triangular, acute, 0.5 —0.7 mm long, 0.2 —0.4 mm wide. Anthers ca 2.2 mm long. Style branches (when dried) amber, eas ape L153: iam long, sometimes unequal. Achenes 4 —‘S-ribbed, ca 4 mm long, brown- black, very sparsely pubescent i aeicoe pappus 5.0—5.5 mm long, composed of 30—35, filiform, ciliate setae. Distribution and Habitat (Fig. 5): Subtropical, lower montane regions in the area of Gomez Farias in Tamaulipas. Flowering Mar. and Nov. Additional specimen examined: MEXICO. Tamautipas: Gomez Farias area: Rancho Del Cielo below Aguacates turnoff, 24 Nov 1968, Richardson 958 (TEX) Flyriella harrimani ts readily distinguished from other species in the ge- nus by its larger, fewer heads that are composed of 70 — 100 florets and by its larger leafy outer involucral bracts (Fig. 6). Nevertheless, it is presu- mably most closely related to E. stanfordii, as noted in the section on Species Relationships. 2. FLYRIELLA STANFORDII K. & R. Phytologia 24:69. 1972. Type: MEXICO. Tamautipas: 4 km W of Miquihuana in canyon with luxuriant vegetation, 4 Aug 1941, Stanford et al. 675 (HoLoryrE: GH!). Erect “rhizomatous” herb, 50 — 100 cm tall. Leaves becoming alternate above; blades deltoid, 2.0 — 8.5 cm wide, 3— 11 cm long, irregularly ser- rate, short-glandular on both surfaces; petioles 2—5 cm long, winged throughout; capitulescence in ultimate pedunculate units of 10—25 heads, each head on a bracteolate, glandular, pedicel, 3.0 —6.5 mm long. 309 Involucre campanulate, 6.0—8.0 mm long, 5.0-7.0 mm wide; bracts imbricate in 3 —5 series, glandular, tinged with purple, lanceolate-ovate, 2.5—6.5 mm long, ca 1 mm wide. Florets 20 — 30; corolla white, tinged with purple, tubular to narrowly funnelform, 4—5 mm long, 0.4 — 106 100 T { Flyriella parryi Flyriella leonensis Flyriella stanfordii Flyriella harrimanii rPO*e Fig. 5. Distribution of the four species of F/yrie/la. 310 mm wide, not constricted near the apex, glabrous; lobes narrowly tri- angular, acute, ca 0.3 mm long, minutely atomiferous-glandular. Anthers ca 1.2 mm long. Style branches linear-oblanceolate, brownish-yellow, ca 0.4 mm long. Achenes 4—5-ribbed, ca 2.3 mm long, black, sparsely pubescent throughout. Pappus 4-5 mm long, composed of 25 — 35 filiform, ciliate setae. Distribution and Habitat (Fig. 5): Subtropical montane areas along the Sierra Madre Oriental from 1850 — 2320 m. In limestone along crags, in oak-pine forest clearings and along stream beds. Flowering: Jun — Jul. Additional specimens examined: MEXICO. Nuevo Leon/TamMautipas: Nuevo Leén and just E into border of Tamaulipas, 28 Jun 1948, Meyer G Rogers 2687 (US); Dulces Nombres, 20 Jul 1948, Meyer & Rogers 2831 (MO, US); Zarogoza, Encantada, 17 Jun 1979, Hinton 17544 (TEX); ca 10 km SSW of Zaragoza, Jul 1977, Wells & Nesom 554 (LL) Flyriella stanfordii is distinguished from other species in the genus large- ly by its narrowly funnelform corolla throat that is not abruptly constricted and by its conspicuously winged petioles (Fig. 7). Additional comments on its relationship are given in the section on Species Relationships. 3. FryrigtLa PARRYI (A. Gray) K. & R. ee = 69. 1972. Eupato- rium parry! A. Gray, In Torr. Bot. Mex. Bound. 859. Type: MEXICO. Cuinua- HUA: “Sierra de Carmel”, sone an error for — a Case in Coahuila, 16 Nov 1852, Parry 3 (HOLOTYPE: GH!). Eupatorium chrysostylum Robinson, Proc. Amer. Acad. Arts. 41:274. 1905. Tyee: MEXICO. Cuimuanua: dry ledges, rocky hills near Chihuahua City, 17 Apr 1885, Pringle 135 (HOLOTYPE: GH!; isorypes: DS!, MICH!, NY!, US!). Flyriella chrysostyla (Robinson) K. & R. Phycologia 24:69. 1972. mek sphenopodum Robinson, Proc. Amer. Acad. Arts. 43:35. 1907. Type: EXICO. CoanulLa: above ica on ees i of limerock, 16 Jul 1906, Shee 10259 (HOLOTYPE: GH!; isorypes: CAS!, MO!" US!). Flyriella spheno- poda (Robinson) K. & R. Phyrcloeie 24:69. io Brickellia fee M. E. Jones ex Flyr, Sida 3:254. 1968. Type: MEXICO. Nuevo Leon: Ojo de Agua, ca 2.5 mi W Sabinas Hidalgo, [26] 25 Mar 1932, M. E. Jones 2941 es nas POM; pHOTOHOLOTY PE: US!; isorypE: UC!). (date and specific locality from Blake 1945). Brickellia shineri M. E. Jones, Contr. W. Bot. 18:22. 1933. Nom. illegic., with out Latin diagnosis. Erect herb from perennial tap root, 0.5 — 1.0 m tall. Leaves mostly op- posite, alternate and reduced just below the capitulescence; blades predo- minately deltoid to cordate, 3— 10 cm wide, 5 — 11 cm long, irregularly serrate, dentate to lobed, glandular on both surfaces, especially along edges and veins; petioles 3—7 cm long, glandular, abruptly winged, just below blade. Capitulescence in ultimate pedunculate units of 3— 16 heads, each head on bracteolate pedicel 3-7 mm long. Involucre turbi- nate, 8— 10 mm long, 4—5 mm wide; bracts in 3—5 series, ovate to il Wh iy ii WW) \ Fl SIL Ack : es q ; } J. YY £] 1, longitudinal section; o d) neues branch; e) portion of stem “showirik manddlae trichomes. linear lanceolate, 2.5 -—9.5 mm long, ca 1 mm wide, the apices acute to acuminate. Florets 10 — 25; corolla yellowish-white or tinged with purple, tubular with constriction ca 1.0 mm long, 0.2 mm wide at top of tube; lobes narrowly triangular-acute, ca 0.2 mm long, ca 0.1 mm wide. Anthers ca 1 mm long. Style branches yellowish-white, linear oblanceo- late, 5—7 mm long. Achenes 4—5-ribbed, ca 3.5 mm long, black, sparsely pubescent throughout; pappus 4—5 mm long, composed of 20 — 30 ciliate setae. Chromosome number, ” = 10 pairs. ot Distribution and Habitat (Fig. 5): wide ranging, highly variable species occurring in mesic habitats of the Chihuahuan desert regions of north cen- tral Mexico and adjacent Trans-Pecos areas of Texas, mostly in calcareous soils at elevations from 900 — 2200 m. Flowering: Apr — Jul. Additional specimens examined: UNITED STATES. Texas. Brewster Co.: end of Pine Canyon Trail, 12 Nov 1976, Corre// 35402 (LL, MO); Cattail Falls, 9 Nov 1964, Correll & Correll 30592 (LL); Juniper Canyon, 16 Jun 1964, Correll et al. 29721 (LL); Big Bend ae Park, Chisos Mts., Pine Canyon, 17 Jun 1963, Correll & W. —— on (CAS, MO); Chisos Mts. Aug 1935, Marsh 149 (F); Cattail ee 7 Jul 1931, Moore & Samar 3408 (CAS, GH, LL, MO, US); Chisos Mts., 5 Aug 1931, ue ee (GH, T US); Pine Canyon, near waterfall under oaks and ee 24 Sep 1977, Powell et al. ee (LL, US); Lower Pulliam Canyon, 4 Jul 1936, Sperry 193 (LL, US); Lower Pulliam Canyon, 26 Aug 1936, Sperry 293 (LL, US); Cattail Falls, 19 Apr 1955, Wallmo 7 (LL); waterfalls in Pine Canyon, | Jul 1937, Warnock 760 (GH, TEX, US); mid Green re . Jul 1950, Warnock 9146 (LL); upper Green Gulch, 28 Jun 1941, Warnock 20975 (GH MO, TEX); Green Gulch, 22 Aug 1947, Warnock & Tharp 4005 (US). Val Verde Gos = 2 mi S of Pandale, 9 May 1967, Correll 34150(GH, LL); Eagle Cave Canyon E of Langtry, 21 Apr 1966, Correll & Rollins 32010 (GH, LL, UC); 5 mi N of Langtry on Ozona road, 8 May 1947, McVaugh 8232 (EF GH) MEXICO. Cruimuanua: Mouth “i ae alca Canyon, 11 May 1959, Correll & Johnston 21758 (LL); near Chihuahua, 22 Jun 1936, LeSueur 963 (F); vicinity of Chihuahua, 8 — 27 Apr 1908, Palmer 40 (E GH, NY, Mapula Mts., 27 Apr 1887, Pringle 1595 (MICH, MSC, UC). comune Sierra Mojada, just $ of Esmeralda, above Sociedad Cooperative Minera, | Sep 1972, ee et al. 9070f (LL); middle of upper reaches of Canon de la Hacienda, almost due S$ of Rancho Cerro de la Madera, N slope of Sierra de la Madera, 21 pa aes Chiang et al. 9449 (LL); ca 32 air mi NE of San Pedro, 1 mi SW of Las Delicias, 27 Au 1971, Henrickson 6072 (LL); ca 64 air mi SE of Big Bend National Park basin jin highway 22 towards Muz am 4 Aug 1976, Henrickson & Prigge 14910 (LL); ca 35 air mi W of Cuatro Ciénegas, in mid-canyon de la Hacienda of Sierra de la Madera, 6 Aug 1973, Henrickson . Wendt 11988 (LL); SW end . Sierra de la Fragua, 1-2 km N of Puerto Colorado, 2 Sep 1941, Johnston 8784 (LL); Canon de Jara, 30-40 km W of Cuatro Ciénegas, 4 —5 Sep 1941, Johnston 8857 ae Rio Grande, side canyons at upper Madison Falls, 10 Apr 1973, Johnston et al. 10611 (LL); Sierra San Marcos, N part jutting into Cienegas basin, higher slopes, just reaching the lowest ponderosa pines, 9 May 1973, Johnston et al, 10943 (LL); ca 5 km SW of Mina El Popo, dissected E slope of Sierra del Carmen, 28 Jul 1973 yas ies 11891 (LL); El Popo, ca 2 km S$ of Cafion El Diablo, 29 Jul 1973, Johnston et al. 119297 (LL); southern part of Sierra de los Organos, 8 Aug 1973, Johnston et al, 12132 (LL), Mears Spring 1935, Marsh 203 (TEX); a 8 Jul 1936, Marsh, Jr. 340 (TEX), Muzquiz, 1939, Marsh s.n. (F); Muzquiz, 12— 13 Apr 1936, Marsh 2136 (GH, TEX); 92.5 mi NW of Mazguiz on highway 53 to Boquillas del Carmen, 29 Jun 1982, Poole & Watson 2531 (TEX); near the spring (pool) above Las Delicias at base of bluffs, 20 May 1974, Powell & Turner 2699 (TEX); canyons in the Sierra del Sobaco, a few km W of Las Delicias, 1 Oct 1942, Santos 2804 (GH), Sierra Mojada, above San Salvador Mine near Esmeralda, 4 Aug 1941, Stewart 1078 (E GH, LL); Cafion del Milagro, ca 12 Km W of Hacienda de la Encantada, 10 — 16 Sep 1941, Sai 1717 (GH, LL); Sierra del Carmen, Pico de Cerda, 12 Aug 1974, Wendt 578 (LL); Canon de la Barrica at base of 313 ee cliffs of Picacho El Pajarito, 28 Aug 1975, Wendt & Lott 1366 (LL); Sierra de la a, 6 Jan 1976, Wendt et al. 1428 (LL); Sierra de la Gloria, Canon Obscuro Chiquillo, 7 ot 1976, Wendt & Riskind 1720 (LL); Muzquiz, Hacienda Mariposa, E slope of the Sierra de Puerto Santa Ana, 23 June 1936, Wynd G Mueller 254 (GH, MICH, MO, MSC, NY, US). Nuevo Leon: Sabinas Hidalgo, Ojo de Agua, 16 Jun 1939, Chase 7020 (GH, NY); Sabinal, 26 Mar 1932, Jones 29411 (CAS); Sierra Madre above Monterrey, 25 May 1908, Pringle 15615 (LL, US); 5 mi W of El Alamo, | Jun 1978, Turner 10007 (E LL). Flyriella parryi is readily distinguished from other species by its glandu- lar trichomes which are often interspersed with long straight eglandular . 7. Flyriella stanfordit: a) flowering head, b) ect section of head; ¢) portion of stem seats glandular trichomes; d) floret; e) flowering branc 314 hairs. In addition, as noted by King and Robinson (197 2b), it is marked by its conspicuously constricted corolla throat and relatively few-flowered heads (Fig. 8 Flyriella parryi is quite similar to FE /eonensis , but the latter does not possess glandular trichomes. No doubt the two species are closely related and perhaps hybridize in the area of Monterrey since both species have been collected on Chipinque Mesa to the west of the city. F/yriella parryi occurs at lower, more xeric, habitats in this region, while F /eonensis occurs at more elevated mesic habitats (e.g., the Quvercus -dominated forest at ca 2000 meters above Chipinque Mesa). 4. FLYRIELLA LEONENSIS (Robinson) K. & R. Phytologia 24:69. 1972. Eupatorinm leonensis Robinson, Proc. Amer. Acad. Arts. 36:479. 1901. Type: Nuevo Lr6n: on the Sierra Madre near Monterrey, 16 Jun 1887, C. G Pringle 2277 (HOLOTYPE: GH! Expatorium chrysostyloides Robinson, Proc. Amer. Acad. Arts. 43:30. 1907. Type: MEXICO. Nuevo Leon: Sierra Madre above Monterrey, limerock, 915 m, 27 Apr 1906, C. G. Pringle 10231 (HoLotyeE: GH!; tsoryre: US!). Erect perennial herb 25 — 50 cm tall. Leaves mostly opposite, alternate and reduced just below the capitulescence; blades predominately deltoid, 3—6 cm wide, 3.5 —7.0 cm long; irregularly serrate to lobate, minutely puberulent on both surfaces, often glabrate; petioles 3—8 cm long, puberulent, abruptly winged just below the blade. Capitulescence of ulti- mate pedunculate units of 10—25 heads, each head on a bracteolate pedicel, 2—8 mm long. Involucre narrowly campanulate, 5 — 7 mm long, ca 5 mm wide; bracts imbricate in 3 — 5 series, linear-lanceolate, 3 — 7 mm long, | mm wide or less. Florets 30— 40: corolla “yellowish-white”, tubular throughout, ca 5 mm long, 0.5 mm wide, glabrous; lobes narrow- ly triangular, acute, ca 0.3 mm long, minutely atomiferous glandular. Anthers ca 1 mm long. Style branches ee yellowish- white, smooth, 5—8 mm long. Achenes 4 —5-ribbed, 2 long, black, sparsely pubescent mostly near the apex; pappus ae mm long, composed of 40 — 50 filiform, ciliate setae. Chromosome number, » = 10 pairs. Distribution and Habitat (Fig. 5): Mostly eastern Nuevo Leon along the front range of the Sierra Madre Oriental from 600— 1700 m where it occurs in limestone soils mostly in mesophytic habitats, along streams and cool moist areas. Flowering: Apr —Ju Additional specimens examined: MEXICO. Nurvo Lron: Chipinque Park, N facing slopes in pine-oak forest, 11 Jun 1978, Poole & Watson 1394 (TEX); Horsetails Falls, ca 35 km SSE of Monterrey, 26 Apr 1976, Sanders & Harborne 76060 (TEX); Cola de Caballo, 6 Apr 1971, Svegler & Becker Ds3133 (MSC); Cola de Caballo, 31 May 1970, Siegler et al. pulp) DS-2570 (MSC); Monterrey, 22 May 1960, Smith M219 (TEX); Chipinque Mesa, slopes just above motel, ca 6000 ft, 2 Jun 1978, Turner 10019 (LL); 17 mi by road W of Horsetail Falls, 2 Jun 1978, Turner 10037 (CAS, LL). Flyriella leonensis is readily distinguished from other species in the genus by its puberulent, non-glandular, foliage. In addition, it can be dis- tinguished by its sparsely ciliate achenes (mostly near apex) and generally smaller and broader heads (Fig. 9). C \ g a ZZ > SSS == fal . } 1 a | : Af | ‘ yoy ] ; £ h Fig. 8. Flyriella parryi 2 : glandular trichomes; d) floret; e) flowering branch. 316 Eupatorium chrysostyloides Robinson (not to be confused with E. chrysos- tylum Robinson, which is a synonym of F/yriella parryi ), clearly belongs to this species. King and Robinson (1972) also treated these as synonymous. According to B. L. Robinson, in his original description, E. chrysostyloides is distinguished by its “mixed pubescence.” Actually the pubescence is puberulous, much as in F/yriella leonensis. As noted above, FE /eonensis appears to occur at higher elevations, up to 1700 m, and in more mesic habitats than F. parryi, at least in the Monterrey area, ACKNOWLEDGMENTS The present contribution is part of a Master's Program completed at The University of Texas by the senior author. We are grateful to Ms. Molly Conner-Ogorzaly for the illustrations and to Dr. Kevin Nixon for generous si% | Br", : y \W TW Fig. 9. Flyrtella leonensis: a) floret; b) portion of stem showing eglandular trichomes; c) flowering head; d) longitudinal section of head; e) flowering branch. 317 help with the cladistic analysis. Dr. Guy Nesom, as reviewer, added assort- ed pungent criticisms which proved helpful. This study is based upon approximately 130 specimens deposited in the following herbaria: CAS (2); DS(1); F (10); GH (19); LL (38); MICH (6); MO (6); MSC (5); NY (7); TEX (20); US (16). REFERENCES ee S. FE 1945. Asteraceae, described from ae and the southwestern United States by M. E. Jones, 1908 — 1935. Contr. U. S. Natl. Herb. 29:117— 137 FLYR, a 1968. New names and records in oe. (Compositae). Sida 3. 252-256. KING, R. M., D. W. KYHOS, A. M. POWELL, P H. RAVEN, and H ROBINSON. 1976. ee eee mumberes in Compositae, XIII. Eupatorieae. Ann. Mis- souri Bot. Gard. 63:862 — 888. KING, R. M. and H. Ron NSON. 1972a. Studies in the Eupatorieae (Asteraceae) LXXVIII. A new genus, Brickelliastrum. Phytologia 24:63 — 64. KING, R. M. and H. ROBINSON. 1972B. Studies in the Eupatorieae (Asteraceae) LXXX. A new genus, Flyriella. Phytologia 24:67 — 69. KING, R. M. and H. ROBINSON. 1982. Studies in the Eupatorieae (Asteraceae) CCXII. Additions to a Elaie Flyriella, and Teixeiranthus, Phycologia 50:379 — 381. MABRY, T. J., MERMAN, N. HEIL, and A. M. POWELL. 1981. Systematic implications i i pen and chromosomes of F/yrie/la (Compositae — Eupatorieae). Pl. Syst. Evol. 137:275 — 280. ROBINSON, B. L. 1901. New species and newly noted synonymy among the ais tophytes of Mexico and Central America. Proc. Amer. Acad. Arts 36:471— 488. ROBINSON, B. L. 1907. New or es nts spermatophytes, nae from Mexico. Proc. Amer. Acad. Arts 43:21 — 48. ROBINSON, B. L. 1913. Revisions ee Ageratum and Oxylobus . Proc. Amer. Acad. Arts 49:439 — 454. ROBINSON, H. and R. M. KING. 1977. Eupatorieae — systematic review. In Heywood, V. H., J. B. Harborne, B. L. Turner, (Eds.): The biology and the chemistry of the Compositae. Academic Press, New York. TIMMERMAN, B. N. 1980. Phytochemical investigations of the genus Brickellia (Compositae) emphasizing flavonoids. Ph.D. Thesis, The Univeristy of Texas, Austin. TIMMERMAN, B. N., K. MUES, T. J. MABRY, and A. M. POWELL. 1979. 6- Methoxy-flavonoids from Brickellia laciniata. (Compositae). Phytochemistry 18:1855 — 1858. TURNER, B. L. 1972. Lowell David Flyr, 1937 — 1971. Sida 5:54 —58. DISTRIBUTION OF PORTULACA OLERACEA L. (PORTULACACEAE) SUBSPECIES IN FLORIDA AVINOAM DANIN Department of Botany, The Hebrew University of Jerusalem Jerusalem, ISRAEL 91904 LORAN C. ANDERSON Department of Biological Science, F a State University Tallahassee, FL 32306, U.S.A. ABSTRACT Range maps for the subspecies of P. o/eracea (Portulacaceae) in Florida are provided. The collections are enumerated. Unusual specimens are noted and their seeds illustrated with scanning electron micrographs. INTRODUCTION The cosmopolitan Portulaca oleracea L. is an aggressive weedy colonizer. Chromosomally, the species is x =9, and diploids, tetraploids, and hexaploids are known. The different cytotypes have distinctive seed size and seed coat texture, which makes subspecies determinations relatively easy (Danin et al. 1978). Matthews and Levins (1985a) stated: “Seed surface markings must be correlated with other morphology features when used in delimiting taxa,” and they did not recognize subspecies or varieties of P. oleracea (1985b). Apparently they were not aware that seed size, shape, and markings were correlated with different chromosome numbers in this species. The subspecies are not evenly distributed throughout the world (Danin tal. 1978; Danin 1983, 1985), but sympatric populations are frequently encountered. Matthews and Levins (1985a) noted Portulaca has a large concentration of species in South America. They suggested P. pilosa entered Florida from the Carribean and spread northeast into North Carolina and westward along the Gulf Coast with evidence of recent migration into Arkansas (probably from Texan populations rather than coastal populations). Study of the distribution of P. o/eracea in Florida is warranted because the state is strategically placed in possible migration zones between tropical and tem- perate regions. In this study, only a few specimens were found in primary habitats, such as Godfrey 72166 from mangrove flats in Monroe County. Most specimens reported here were collected in secondary synanthropic habitats such as gardens, patches in lawns, and along roadsides. SIDA 11(3):318 — 324. 1986. 319 METHODS Ripe seeds were obtained from living plants or from herbartum specimens (principally from FLAS, FSU, and USF). Identifications were made from seed size and seed coat texture as determined with the dissect- ing ae under diffused light. A key to the subspecies is in Danin et al. (1978). The subspecific epithets granulatostellulata and ane lulata were originally hyphenated, but according to Article 73.9, (Voss 1983), the hyphen should be deleted. Seeds were mounted on stubs, vacuum dried, then sputter coated with 100 — 200 A of AuPd (60/40). SEM micrographs were taken on a Cam- bridge Stereoscan S$4-10 at 20 keV. Seeds are illustrated to show diagnostic surface features (Figs. 1—9). The stellulae (plate-like epidermal cells) may have stellate radiating arms (Figs. 2, 5) or have irregularly lobed margins (Fig. 7). Stellulae may have prominently raised centers (tubercles) and papillae on the stellular arms (Fig. 9), but these features do not always occur together. The cell surfaces are usually smooth, but they may be covered with wax in some specimens (Figs. 2, 3). DISTRIBUTION The known distribution of Portulaca oleracea in Florida as determined from specimens from selected herbaria is mapped by subspecies in Figure 10 and enumerated below. It probably occurs in every county in the state, but fleshy, weedy plants (such as this species) are often neglected by collec- tors. Chromosome numbers are noted for each subspecies (Danin et al. 1978); no counts were made in the present study. P OLERACEA subsp. ge os Sa LLULATA (Poelln.) Danin & Baker (x 18). Broward Co.: 10 mi W of Deerfield, Aug 1969, Cart 10866 (FLAS). Hendry Co.: Clewiston (mixed with subsp. eee nn 29 Dec 1984, Danin s.n. (FSU). Hernando Co.: Weeki Wachee, 25 Oct 1971, Genella & Floning 1062 (USF). Hillsbo- ee Co.: Tampa (mixed with subsp. nicaraguensis), 9 Jul 1977, Crewz 1097 (USF). Lake Co. mi W of Yalaha (mixed with subsp. papillatostellulata), 12 Oct 1980, Baltzell oes (ELAS). Lee Co.: Fort Myers, 26 Dec 1984, Danin s.n. (FSU); Florida Forests Foundation, 20 Aug 1963, Hoffman 11 (FLAS). Leon Co.: N of Tallahassee, 30 Aug 1984, Danin s.n. (FSU); Tallahassee (mixed with subsp. #tida), 11 Sep 1984, Danin s.n. (FSU); Tallahassee, 22 Mar 1985, Danin s.n. (FSU); Tallahassee, 21 Jul 1942, ee s.n. (FLAS). Manatee Co.: Bradenton, 18 Nov 1947, Bargis s.n. (FLAS). Marion Co.; 2 mi NW of Weirsdale, 18 Nov 1973, Baltzell 5619 (FLAS). Monroe Co.: Cape Sable, : May 1965, Lakela & Long 28566 (USF). Okaloosa Co.: ee Air Force Base, 21 Nov 1983, W/belm 11915 (USF). Pinellas Co.: Clearwater, 25 Jul 1984, Danin s.n. (FSU); Belleair Bluffs to Belleair Shores, 24 Jul 1975, Scale et al. Be (USF). Wakulla Co.: Panacea, 27 May 1985, Danin s.n. (FSU). 320 521 P OLERACEA subsp. NICARAGUENSIS Danin & Baker (vy = 9). Brevard Co.: Malabar Cape, 31 Dec 1974, Lakela & Long 28072 (USF); North Merritt Island, 12 Dec 1972, Shuey M0684 (USF). Clay Co.: 2.5 mi E of Middleburg, 31 May 1981, Sauleda & Ragan 5380 (USF). Collier Co.: NW of Naples, 29 May 1965, Lakela 28681A (USF). Dade Co.: old field, 1 Nov 1979, Alexander s.n. (USF); Coral Gables, 26 Dec 1984, Danin s.n. (FSU); Miami Beach, 29 Dec 1984, Danin s.n. (FSU); Miami, 19 Sep 1980, Sauleda 4631 (USF); Key Biscayne, 2 Nov 1965, Cees: n, ae Flagler Co.: Marineland, 18 Nov 1961, Godfrey 61693 (FSU). Franklin Co.: Apalachicola, 30 Aug 1984, Anderson 7545 (FSU). Hendry Co.: Clewiston, 30 Jun 1967, ner wee (FLAS). Hillsborough Co.: Tampa (mixed with subsp. granulatostellulata), 9 Jul 1977, Crewz 1097 (USF); B mi S oe River View, 1 Apr 1976, Wunderlin et al. 5007 (USF). Lake Co.: Mt. Plymouth, | Aug 1983, Daubenmire s.n. (USF). Lee Co.: S tip Sanibel Island, 13 Mar 1954, Cooley 2568 (FLAS, USE); Sanibel Island, 30 Mar 1968, Brumbach 6190 (FLAS), 28 Oct 1978, Wunderlin et al. 6188 (USF); Fort Myers, Correl/ 30330 (GH); Mound Key, 13 Jul 1974, es 127 (FLAS, USF). Levy Co.: 3.3 mi NE of Cedar Key, 11 June 1976, Baltzell 8551 (FLAS). Martin Co.: 11.5 mi N of Port Mayaca, 25 ve De, Blarzell 91 30 ees ee Co.: Flam a es 1956, ee n. (FLAS); ngo (mixed with subsp. papi/latostellulata), 28 1984, Danin s.n. (FSU); Cudjoe Key 16 Seb 1972, Godfrey 72166 (FSU); Key West, re oe 1974, Godfrey 74027 (FSU), 14 Sep 1979, Hansen 6265 (USF); Long Key 14 Nov 1964, Lakela 27929 (USF); Big Pine Key, 7 Aug 1966, Long et al. 2241 (USF); Content Keys, 4 Jun 1967, Long 2675 (FSU); Spanish Harbor Key, 29 Nov 1969, Long 3026 (USF), Marathon, 27 Aug 1961, Rossbach 2887 (FLAS). Palm Beach Co.: Palm Beach, 29 Jun eke Cassen 99 (USF). Pinellas Co.: Clearwater, 25 Jul 1984, Danin 5.n. (FSU); NE of ater, 28 Sep 1970, Genella & Fleming 349 (USF); Dunedin, 3 Oct 1976, Genella & Fleming 2490 (USE). Putnam Co.: Welaka, 26 Jun 1940, Laessle s.n. — Volusia Co.: mi S of New Smyrna Beach, 27 Apr 1961, Ray 10787 (FSU, USF). P OLERACEA subsp. NiTIDA Danin & Hees = 18). Alachua ce Gai ISM ae Fleck 48 (FLAS). Collier Co.: Marco Island, 20 Aug 1965, Sie we a Leo : Tallahassee (mixed with subsp. granulatostellulata), 11 Sep 1984, Danin s.n. ae) P OLERACEA subsp. PAPILLATOSTELLULATA Danin & Baker (v7 = 27). Duval Co.: Holly Oaks F 17 May 1965, ee 424 (FLAS). Escambia Co.: Pensacola, 16 Aug 1983, Wilbelm ao: (USF). Hendry Co.: Clewiston (mixed with subsp. renronaes 29 Dec 1984, Danins.n. (FSU). Indian River Co.: 4 mi S$ of Vero Beach, 1 1962, Godfrey & Reinert 61480 (FSU). Lake Co.: 0.5 mi W of Yalaha (mixed with su — granulate lulata), 12 Oct ue ae 11235 ean cae Co.: Clearwater, 25 Dec 198: Danin s.n. (FSU). M e Co.: Flar th subsp. nicaraguensis), 28 Dec ee Danin s.n. (FSU). P OLERACEA subsp. STELLATA Danin & Baker (v7 = 27). Hillsborough Co.: Egmont Key, | Sep 1978, Crewz 1480 (USF). Figures 1-9. Scanning electron micrographs of selected Portulaca oleracea seeds. Lakela 27 ee (subsp. stellata X subsp. nicaraguensis?). 2. Lakela 27374, with typical stellulae of Bee stellata. 3. Lakela 27374, surface detail showing waxy covering characteristic of many subsp. AG Ges specimens. 4. D'Arcy 2942 (affin. subsp. nitida?). 5. D'Arcy 2942, with non-waxy stellulae that are individually much larger than chose in ae: 2. o Godfrey en (subsp. nana ne non-waxy form). 7. Godfrey 61693, 8. Danins.n.in 1985 (subsp. only. 9. Danin s.n. in 1985, with tubercles and papillae on the epidermal 322 PROBLEMATIC SPECIMENS ra The following specimens are not comfortably assigned to any subspecies. They are not mapped in Fig. 10, but are illustrated in Figs. 1—5. They possibly represent new subspecific taxa or hybrids. Lakela 27374 (FLAS, USF) from Fort de Soto Park in Pinellas County has rather diverse seed size. A few seeds have size and shape like those of subsp. sve//ata but are covered with wax as in subsp. nicaraguensis. Figures 1—3 are of an unusually small seed of Lakela 27374 with the stellate epidermal cells of subsp. ste//ata and the waxy covering (and smaller seed size) of subsp. micaraguensis (Fig. 3). These specimens probably represent hybrids between the two subspecies. D'Arcy 2942 (FLAS) from Indian River Island in Indian River County has seeds similar to those of subsp. nétida in size and shape, but the specimen has linear rather than obovate leaves. The epidermal cells of the seeds are stellate as in subsp. witida (Figs. 4, 5), but the individual cells are much larger in D’Arcy’s specimen (i.e. there are fewer cells across the face of the seed) than they are in typical subsp. nitida. DISCUSSION Legrand, in a study of American species of Portulaca (1962), noted varia- tion in seed surface features in P oleracea, but he did not recognize any varieties or subspecies taxonomically. Matthews and Levins (1985a) found that seed surface markings were not helpful in distinguishing some other species of Portulaca. They did not cite the study of Danin et al. (1978) in either of their papers, and even though they noted some variability in seed coat features in P. oleracea, they followed the taxonomy of Legrand. We have found that seed size and surface texture are of considerable taxonomic utili- ty in distinguishing subspecies of P. oleracea. All the Florida collections were easily assigned subspecies with the exception of the two mentioned as problematic specimens. Unfortunately, the different subspecies (cyto- types) are not distinguishable vegetatively, and chromosome numbers and seed ultrastructure are not useful for identification of subspecies in the The cytotypes of P. oleracea are not evenly distributed on a world-wide basis, and they show an uneven distribution in Florida as well. The subtropical subsp. nicaraguensis is the most frequently collected in Florida (Fig. 10). Its range in Florida is apparently part of the original distribution of the subspecies rather than due to human interference. Migration routes for this subspecies from central America and the Carribean into Florida appear to be similar to those noted for P. pilosa (Matthews and Levins 1985a). Subspecies ste//ata is generally found at higher latitudes; its 229 occurrence in Florida is surprising and does not reflect migration from the tropics. Putative hybrids are known from Yucatan (Danin et al. 1978) and Flort- da (interspecific and intraspecific, respectively). This suggests the Gulf o Mexico region is an area of active speciation for Portulaca oleracea. This species is well suited for detailed cytogenetic studies at the population level because sympatric subspecies occur. subsp. granu/atoste//ulata * subsp. n/ide f subsp. sfe//ata @ | \ ae subsp. n/caraguensis @ subsp. pap//latoste/lulata * é 0 eee Figure 10. Distribution of Portulaca oleracea subspecies in Florida. 324 REFERENCES DANIN, A. 1983. Portulaca oleracea L. subsp. cg In: Med-Checklist Notulae 8, W. Greuter and T. Raus, eds. Willdenowia 13:2 —_______.. 1985. Portulaca L. 2 Flora Tne Made (in press). DANIN, A., 1. BAKER, and H. G. BAKER. 1978. iia ea and taxonomy of the Portulaca oleracea L. polyploid ae Israel J. Bot. 27:177— 211. ee Las especies americanas de ye Anal. Mus. Hist. Nat. Montevideo 7:1— 1 MATTHEWS, J. E, ae i A. LEVINS. 1985a. Portulaca pilosa L., P. mundula\. M. Johnst. and P. parvula Gray in the southwest. Sida 11:45—61 1985b. The genus Portu/aca in the southeastern United States. Castanea VOSS, E. .G. 1983. International code of botanical nomenclature. Regnum Veg. = 472 SOME OBSERVATIONS ON PINUS GLABRA WALTER (PINACEAE) RICHARD STALTER St. John's University, Jamaica, NY 11439, U.S.A. STEVE DIAL Pfeiffer College, Misenhezmer, NC 28109, U.S.A. ABSTRACT Pinus glabra Walter is the least common of the southeastern pines. It rarely if ever grows in pure stands. Quadrats were established at 14 sites in the southeastern United States wherever P. glabra was abundant throughout its range. Importance values of trees found in sample plots indicate that P glabra had a higher importance value in Mississippi than had been previously reported. INTRODUCTION Pinus glabra Walter, spruce pine, the least common of the southeastern pines, rarely if ever grows in pure stands. It occupies soils that are loamy, low in organic content, and acid in reaction. Pinus glabra may be found in all stages of its life history in mature forests within its range, an unusual occurrence for a pine. Its seedlings and saplings have been reported grow- ing where light intensity was as low as 250 foot candles (Dial et al. 1976). Dial et al. (1976) have studied P. glabra on sites in the lower coastal plain of South Carolina where P glabra is locally abundant. They sampled six areas in Colleton and Dorchester counties, South Carolina and found P. glabra to be associated with Liguidambar styraciflua, Pinus taeda, and Quercus virginiana. Harrar (1964) reported that P. glabra reaches its maximum development within an area in northwest Florida between the Choctawhatchee and Chattahoochee rivers. Observations by the present authors in this area indi- cate that Pinus clausa, sand pine, not P. glabra, occurred in pure stands in this area. The objective of the present study was to determine the relative abundance of P. glabra within its range in the southeastern United States. METHODS To determine the relative abundance of P. g/abra within portions of its range in the southeastern United States, vegetation was sampled by the quadrat method at 14 stations. Six areas in Dorchester and Colleton coun- SIDA 11(3):325 — 328. 1986. 326 ties, South Carolina had been sampled and reported previously by the present researchers in an earlier paper (Dial et al. 1976). Eight additional sites in Georgia, Florida, Mississippi, Alabama and Louisiana were selected to sample vegetation in the area of maximum development of P glabra. Site 2 was selected in Seminole County, Georgia; Site 3 in Jackson County, Florida; Site 4 in Gadsden County, Florida; Site 5 in Washington County, Florida; Site 6 in Okaloosa County, Florida; Site 7 in Pearl River County, Mississippi; Site 8 in Geneva County, Alabama; and Site 9 in Washington Parish, Louisiana. In each of the aforementioned areas, twenty 10 x 10 m quadrats were selected to sample the arborescent vegetation. Importance values of the arborescent species were calculated by the follow- ing formula: Importance Value = the sum of the relative density, relative frequency and relative dominance (percent basal area). Nomenclature follows that of Radford et al. (1968). A smaller 2 m by 4 m plot located near the center of the larger plot was used to sample seedlings and saplings. A 1 m? plot located near the center of the 2 m by 4 m plot was used to sample herbs and grasses. Sample plots were randomly selected within the specific areas sampled; no two plots were ever adjacent to each other. The fact that the sampling was not random must be emphasized; the locations for the 14 sample areas were chosen with discrimination for the purpose of sampling P. glabra in areas of high density. RESULTS & DISCUSSION Pinus glabra reached its best development in a stand located in Pearl County, Mississippi, just east of the Pearl River. Here P. g/abra attained an importance value of 157, while Carpinus caroliniana, the most important associate attained an importance value of 58 (Table 1). If Carpimus, an un- derstory tree, were not included in the importance value calculations, P glabra would have had an even higher importance value. The average im- portance value of P glabra on the Florida sites was slightly higher (78.5) than the importance values for P. glabra on the South Carolina sites where the average importance value was 69. Importance values for P. g/abra were higher tn Mississippi and Georgia, but represent data from single study sites sampled by 20 quadrats. In Florida, P. g/abra attained an importance value of 127 at the Okaloosa site, which was located west of the Choctawhatchee River. Pinus glabra was usually associated with Liguidambar styraciflua and Pinus taeda in South Carolina. Carpinus caroliniana, Quercus virginiana and Carya glabra might also be associated with P. glabra in South Carolina but generally were not as common as Liquidambar and P. taeda. In Florida, Q. 527 _ ‘aBLE 1. Importance values of tree species associated with Pinus glabra at nine study sites in the southeastern United Stares. Sire 1, Sourh Carolina; Site 2, Georgia; Sites 3 — 6, Florida; Site 7, Missis- sippi;, Site 8, Alabama; Site 9, een See the methods section for site (eeation, ImporTANCE VALUES FOR 9 STUDY SITES SPECIES ] 2 3 4 5 6 7 8 9 Pinus glabra 69 130 43 53 91 127 17 45 118 Satan nigra 15 59 5 42 17 21 Pinus taeda 35 22 1] 4 Teens styraciflua 38 13 31 16 22 26 7 19 Ilex opac 9 7 4 13 Acer rubrum 17 7. 13 3 6 13 Salix nigre 2 3 Carya glab 20 3 uercus virginiana 26 22 48 75 25 74 Fagus grandifolia 6 54 34 55 Carpinus caroliniana 15 13 9 20 58 31 11 Persea palustris 1 Cornus flori 1 Nyssa biflora 8 4 5 6 3] 29 19 Quercus michauxii 10 ) 9 Ulmus alata 8 Fraxinus americana 3 Quercus laurifolia 6 17 32 82 S| 13 Quercus stellata 3 3 Ostrya virginiana 3 3 Quercus alba 4 5 Pinus elliotcit 76 13 15 Ulmus americana 5 5 5 spp. 4 20 Wepncle grandiflora 130 23 8 10 Quercus falcata 14 6 arya aquati 9 6 12 6 Taxodium distichum 4 5 3 7 Fraxinus caroliniana S, Betula 3 20 Prunu ae 5 virginiana was the usual associate of P glabra, while Quercus nigra, Liquidambar, Carpinus, Nyssa biflora, Magnolia grandiflora and Quercus laurifolia were less frequently associated with the species (Table 1). Gener- ally, the aforementioned trees were commonly found with Pinus glabra in the southeastern United States. Additional trees associated with P. glabra included Acer rubrum, Pinus elliottii, Carya aquatica, and Taxodium d1s- tichum, yet none of these trees with the exception of P e//zotti attained high importance values in the sites sampled in the present study Pinus glabra often forms a part of the understory and occasionally the overstory in mixed hardwood pine forests in the lower coastal plain of the 328 southeastern United States. Pinus glabra has the capacity for survival in deep shade and specimens of all sizes and ages from seedlings to mature seed trees, thrive in deep shade. Green (1938) calls P glabra the most shade tolerant of the eastern pines. Pinus glabra was conspicuously present in all stages of its life history, and was one of the most commonly encountered seedlings and saplings in the study plots. Other seedlings and saplings associated with P. g/abra, listed in decreasing order of density include: Liguidambar styraciflua, Quercus nigra, Acer rubrum, Carya glabra, Ilex opaca, Persea palustris, and Carpinus caroliniana. Others with lower density were also present. Shrubs and lianas were represented by Rubus spp., Myrica cerifera, Sebas- tiana ligustrina, Sabal minor, Vaccinium spp., Aralia spinosa, Smilax spp., Rhus radicans, Gelsemium sempervirens, Lonicera japonica, Vitis rotundifolta, Parthenocissus quinquefolia, Campsis radicans, Bignonia capreolata, and many others. Common grasses and herbs included Arundinaria tecta, Panicum spp., Uniola laxa, Elephantopus tomentosus, Scleria spp., and others. Polypo- dium polypodiordes was usually present on the limbs of large live oaks, if Quercus virginiana occupied the study sites. Harrar (1964) reported that Pinus glabra rarely, if ever, grows in so called pure stands “with the exception of an area in northwest Florida between the Choctawhatchee and Chattahoochee rivers where it reaches its maximum development.” Data in the present study indicate that the development of P. glabra in northwestern Florida was no more pronounced than in other areas sampled. In fact, development was not as great in northwestern Flori- da as in Pearl River County, Mississippi or Washington Parish, Louisiana (Table 1), No pure stands of P. glabra were observed in this work in north- western Florida as reported by Harrar (1964) although sand pine, Pinus clausa was observed in nearly pure stands on higher drier sites in this area. REFERENCES DIAL, S. C., W. T. BATSON and R. STALTER. Be ue ecological and morphological observations of Pinus glabra Walter. Castanea 41: —377 GREEN, G 1938. Trees of North America fae of Mexico), Edwards Bros,, Ann Arbor, hate HARRAR, E. S, 1964, Hough's os of American woods, Vol. IV. Robert Speller and Sons, Publ., a York, N. RADFORD, A, E., H, E, AHLES, ie C. R. BELL, 1968, Manual of the vascular flora of the Carolinas, The fe of North Carolina Press, Chapel Hill, North Carolina, A NEW SPECIES OF WALLENIA (MYRSINACEAE) FROM HAITI WALTER S. JUDD Department of ce University of Florida Gainesville, FL 32611, U.S.A ABSTRACT A new species, Wallenia formonensis, is described from a disturbed cloud forest at 1520 m alt. on the southern slope of Morne Formon in the Massif de la oe of southern Haiti. The ioe Wallenia, of some 25 s s, is, itselt, endemic to the t Indies. The cloud forests and moist pine ane of ie Wipes elevations of Morne ee and pas M. aya are diverse and contain many local endemics. The flora of this area is known mainly through the field work of Erik L. Ekman, but is in need of additional botanical es tion. The species described in this paper was discovered in the high elevations of Morne Formon in the Massif de la Hotte, southern Haiti. Field wor formed part of an inventory of the flora and fauna of the recently established Pic Macaya National Park. The discovery of this species brings to eight the number of species of Wallenia recorded for Hispaniola (Barker and Dardeau 1930, Moscoso 1943, Liogier 1971). Two other species of Wa//enta occur in Pic Macaya National Park: W. aquifolia Urb. & Ekm. and W. ekmanii Urb. Both species are endemic to the mountains of southern Haiti. Other Myrsi- naceae occurring in the park include Ardisia fuertesi7 Urb., Myrsine coriacea (Sw.) R. Br. ex Roem. & Schult., and Myrsine magnoliifolta (Urb. & Ekm.) Alain. Although the high elevation cloud forests and moist pinelands of the Massif de la Hotte were explored by Erik L. Ekman (Ekman 1928, Moscoso 1943, unpublished field notes of E. L. Ekman), the diverse and highly endemic flora of the region is in need of additional botanical exploration, as evidenced by the discovery of several “new” species, one of which is de- scribed herein. WaLLeNIA formonensis W. Judd, sp. nov. (Figure 1) ayes haec ab Wallenia jacquinioides (Griseb.) Mez diftert foltis parviortbus, 1.¢., ca —4.7 cm vs. 5—9.5 cm longis, et inflorescentiis plerumque brevioribus. Tree to ca LO m tall. Indumentum of multicellular, peltate, gland- 'This paper is Florida Agricultural Experiment Station Journal Series No. 6779. SIDA ENG 329 = 33:35 1930. 330 headed hairs. Twigs roughened, glandular-punctate. Leaves alternate and pseudoverticillate; petiole 4— 13 mm long, adaxially grooved, glandular- punctate; blade obovate, (1.6)2.5 —4.7 cm long, (0.5)1—2.2 cm wide, coriaceous with ovoidal, brownish, resin-containing, secretory cavities in mesophyll, especially near margin, the apex truncate to rounded, the base attenuate, the margin entire, plane to revolute, especially along proximal portion of blade, the adaxial surface dull to slightly lustrous and obscurely glandular-punctate, the abaxial surface pale and conspicuously glandular- punctate, the venation brochidodromous, adaxially obscure, abaxially with secondary veins slightly raised and visible, the tertiary veins obscure, not raised-reticulate. Inflorescences axillary racemes or rarely scarsely branched panicles, the primary axis 2—4 cm long, glandular-punctate, bearing ca 13 to 30 flowers. Flowers imperfect (plants dioecious) but appearing perfect, 5-merous, each subtended by a more or less linear caducous bract to 3 mm long; only staminate flowers seen. Pedicels 2 — 4.5 mm long, glandular-punctate. Sepals 5, imbricate, widely ovate with rounded to obtuse apices and slightly erose, glandular-fimbriate margins, 1.3—1.7 mm long, 1.3— 1.5 mm wide, sparsely glandular-punctate and with conspicuous red to brown ovoidal secretory cavities. Corolla broadly campanulate, white with conspicuous brownish ovoidal secretory cavities, the tube ca | mm long, the lobes 5, triangular/ovate with rounded apices, 1.3—2 mm long, 1.3 — 1.9 mm wide. Stamens 5, the filaments narrow, 3—4 mm long, the anthers ca | mm long with conspicuous dorsal cluster of reddish ovoidal secretory cavities. Pistillodium ca 1.7 mm long with ovoid, glabrous ovary tapering to short style; secretory cavities present. Pistillate flowers and drupes not seen. TYPE: HAITI. Departement pu Sup: Massif de la Hotte, Pic Macaya National Park, disturbed cloud forest on southern slope of Morne Formon, ca 1520 m alt., north of com- munity of Formon, occasional, 11 Jun 1984, James D. Skean, Jr. 1524 (HoLtorype: FLAS; IsOTYPES: EHH, NY). Wallenta formonensis is a member of Wallenia subgenus Homowallenia Mez, a group characterized by scarsely heteromorphic flowers that are borne on axillary racemose inflorescences (Mez 190 1, 1902). The species is quite similar to the Oriente/Cuban taxon, Wallenia jacquinioides (Griseb.) Mez, and the two species are likely closely related. Wallenia formonensis differs from W. jacquinioides in its consistently smaller leaves (i.e., (1.6)2.5 — 4.7 by (0.5)1 — 2.2 cm vs. (3)5 — 9.5 by (0.8)1.5 — 3.2 cm in W. jacquinioides), shorter inflorescences (i.e., 2—4 cm vs. 3-6 cm in W. jacquinioides), and a tendency toward producing more flowers per in- florescence. The new species is also easily distinguished from the similar Dominican species, W. apiculata Urb. and W. urbaniana Mez, by its less a9:| coriaceous leaves with obscurely reticulate higher-order venation, t.e., the tertiary and higher-order veins are not raised-reticulate. The calyx and co- rolla of the latter two species lack prominent red to brown ovoidal secretory cavities. Wallenia formonensis differs from the recently described W. gracilis Alain (Liogier 1971) in its blunt-tipped and only obscurely reticulate- veined leaves, inflorescences with several to many flowers, and corolla with rounded lobes. The cloud forests and moist forests of Pinus occidentalis Sw. of the higher elevations (i.e., chiefly above 1300 m alt.) of the Massif de la Hotte are Figure 1. Wallenta formonensis W. Judd: A, habit; B, leaf; C, staminate flower. Drawn from the holotype. oP floristically diverse and contain numerous endemics, e.g., ca 34% of the flowering plants of Pic Macaya National Park are endemic to Hispaniola (Judd unpublished data). Such forests in the Massif de la Hotte are now essentially limited to the upper slopes of Morne Formon and M. Macaya. Other trees and shrubs endemic to the Massif de la Hotte that recently have been collected by the author (or J. D. Skean) in Macaya National Park include: Calyptranthes hotteana Urb. & Ekm., Cestrum filipes Urb. & Ekm., Cleyera ternstroemioides (O. E. Schmidt) Kobuski, Dendrophthora carnosa Urb. & Ekm., Eugenia formonica Urb. & Ekm., Expatorinm Hlavidulum Urb. & Ekm., E. pihenis Ekm.., E. Nanda a Urb. & Ekm., Haenianthus ines Urb., Mecranium microdictyum Urb. & Ekm., M. ie ostatum Urb. Ekm., Meliosma abbreviata Urb., Meriania squamulosa Urb. & Ekm., Miconta apiculata Urb. & Ekm., M. barkeri Urb. & Ekm., M. bypiodes Urb. & Ekm., M. ossaeifolia Urb. & Ekm.., Myrsine magnoltifolia (Urb. & Ekm.) Alain, Pachyanthus blancheanus (Utb.) Urb., Psychotria alpestris Urb. & Ekm., Rondeletia formonia Urb. & Ekm., Sapium haitiense Urb., Solanum formonense O. E. Schulz, S. hotteanum Urb. & Ekm., Stevensia hotteana Utb. Ekm., Symplocos hotteana Urb. & Ekm., Ternstroemia barkeri Ekm. & Schmidt, and Wallenia aquifolia Urb. & Ekm. The region also supports many endemics at lower elevations in the moist forest on limestone in the vicinity of Formon. Woody taxa occurring with this newly described species include Besleria lutea L., Brunellia comocladiifolia Humb. & Bonpl. subsp. domingensis Cuatr. . Didmibanas tremulum Krug & Urb., Gomidesia lindentana Berg, Gyrotaenia myriocarpa Griseb., Heterotrichum ne Urb., Mecraninm microdictyum Urb. & E cm., Miconia subcompressa Urb., Monive coriacea (Sw.) R. Br. ex ene & Schule , [urpinia picardae Urb., and Vernonia saepium Ekm. There is hope that she establishment of the Pic Macaya National Park will lead co the preservation of these interest- ing forests. ACKNOWLEDGEMENTS I thank Dr. Charles Woods, Florida State Museum, coordinator and principal investigator of the U.S.A.1.D. sponsored project, Brogeophysical Inventory of the National Parks of Haiti, who organized the field trips to the Massif de la Hotte. Thanks are also due to James D. Skean who assisted in the collection of the plants of this region. I am greatful to Dr. Dana G. Griffin, II for his helpful suggestions concerning the manuscript, and to the New York Botanical Garden for the loan of comparative material of Wallenia. Finally, | wish to thank Wendy Zomlefer for preparing the illus- tration. 359 REFERENCES ea HNO. 1957. Flora de Cuba. Vol. 4. Ocas. Mus. Hist. Nat. Col. La Salle — 44]. enn H. D. and W. 8. DARDEAU. 1930. Flore d’Haiti. 456 pp. Service anes du Departement de LAgriculature et de LEnseignement Professionel, Port-au-Prince. EKMAN, E. L. 1928. A botanical excursion in La Hotte, Haiti. Svensk. Bot. Tidskr. 22:200 — 219. LIOGIER, A. H. (= HNO. ALAIN). 1971. Novitates antillanae IV. Mem. N. Y. Bot. Gard. 21:107— 157. MEZ, C. 1901. Myrsinaceae. Symbolae Antillanae 2:389 — 43: mec R. M. 1943. Catalogus florae Domingensis. 732 pp. une Santo Domingo, New York. SOLIDAGO (ASTERACEAE) OF LIMITED DISTRIBUTION IN THE CENTRAL UNITED STATES CONSTANCE E. 8. TAYLOR and R. JOHN TAYLOR Department of Biology Southeastern Oklahoma State University, Durant, OK 74701, U.S.A. ABSTRACT Solid. achitensis C. & J. Taylor, a rare endemic of the Ouachita Mountains of Arkansas and Gldshora is descilhed new to science. The range of Solidago speciosa var. pallida 1s extended southward into New Niece sat Oni INTRODUCTION Continuing work on Solidago has resulted in the location of two addi- tional taxa of So/zdago for the south central United States (Taylor & Taylor 1983, 1984). One of these taxa is a new species from the Ouachita Mountains of Arkansas and Oklahoma. The other taxon, Solidago spectosa var. pallida, is sporadic along the front range of the Rocky Mountains and is being reported new to Oklahoma and New Mexico. Both taxa are examples of schizoendemics, their isolation probably occurring since the glaciers receeded. 1. SoLIDAGO ouachitensis C. & J. Taylor, sp. nov. Solidago caulibus simplicibus glabris, e caude ramoso; foliis lanceolatis glabris acutis, usque ad 16 cm longis et 6 cm latis, grosse serratis, in caulibus decurrentibus; inflorescentiis pallens bracteis involucrorum glabris; pedunculis pubescentibus vel glabratis; floribus 4—8, ligulatis 1 (O—2) per capitulum, acheniis 18 seriis glabris. Plants 7 — 12 dm tall, with few to several mostly unbranched glabrous stems from a branched rootstock. Leaves alternate, lanceolate to obovate, acute, largest leaves at midstem up to 10 cm long, 6 om wide, with prominent midvein and numerous smaller lateral veins, glabrous, edges coarsely toothed, serrations up to 5 mm in size. Base of leaf decurrent on stem caus- ing striations. Leaf internodes mostly 5 — 7 cm long. Inflorescence axillary, the upper 1/3 — 1/2 of stem leaves normally with 4—8 heads per axil, pedicels glabrous to puberulent, involucral bracts 4.5—5 mm by 1 mm, inner obtuse, outer acute, glabrous, margins ciliate, l-nerved. Total flowers 4 —8, ray flowers 1 (O — 2) per head, ligule 3 mm x 0.5mm; disk corollas 4 mm long, (tube 2 mm, lobes 2 mm). Mature achenes 4 mm long, with about 18 striations, glabrous. SIDA 11(3):334— 339. 1986. pe) Tape 1. Comparison of selected characters between Solidago taxa: onachitensis, Caesia var. caesta, ¢. vat. curtisit, and flaccidifolia. Character OQUACHITENSIS CAESIA CURTISII FLACCIDIFOLIA Longest leaf (cm) 12— 16 10— 12.5 11.5-—17 10-15 Widest leaf (cm) 4-6 1.5-—2.5 2-2.8 e533) 5 Internode Igth (cm) 5-7 L524 255 ledie 355 2 Sian Floral axis glabrous glabrous pubescent pubescent Heads/leaf axil 4-8 4—10 Dal) 4—10 Phyllary Igch (mm) 4.5—5 3.5-4.5 4—4.5 3.5—4 Phyllary width (mm) l 0.7 0.7 - . Phyllary surface glabrous glabrous + pubescent pubescent Flower number 4— 8-9 = Dir?) Ray flower number 1(0 — 2) 3-4 3-4 4-4 Achene size (mm) 4 2.5 225 3 Achene surface glabrous pubescent pubescent pubescent Endemic to very mesic forests on north-facing slopes in the Ouachita Mountains of Arkansas and Oklahoma. (Figs. 1, 2). Type: U.S.A. OktAHOMA. LeFlore Co.: mesic forest on north- facing is of Rich Mountain, 5.3 mi N and 7 mi E of Big Cedar, 7 Oct 1984, J. & C. Taylor 32788. (HOLOTYPE: DUR; isoryprs: GH, MO, NLU, NY, OCLA, OKL, OKLA, ra UARK). Solidago ouachitensis is a member of the caesia-flexicaulis complex of the eastern United States. It was first collected by G. W. Stevens in 1913 from near Page, Oklahoma. Until last year, it had been collected only four times from a total of three locations. The distribution shown for S. cwrtisi7 in Arkansas (Smith 1978) is based on material of this species as 1s apparently the report of S. flaccidifolia Small for Oklahoma (Uttal 1984). Uttal’s report is based on G. W. Stevens, 5.n., Oklahoma, LeFlore Co., near Page 1n open woods on mountain side (US). Our specimen (DUR) with the same collect- ing data is Stevens 2761 and is in early bud. Table 1 gives a comparison of selected characters of sympatric Solidago ouachitensis and S. caesia var. caesia as they occur in the Ouachita Mountains of Arkansas and Oklahoma. Also included for comparison are two other closely related allopatric members of the caesia-flextcaulis complex. Measurments for S. flaccidifolia (S. caesia var. paniculata Gray) and S. eaesta var. curtisii Wood (S. curtisit Gray) are based on the original descriptions and measurements from specimens collected by the authors in the Appala- chian Mountains of North carolina and Tennessee. Examination of data in Table | shows several morphological differences between S. ovachitensis and S. caesia var. caesia: leaf length/width ratio, leaf width, internode length, size of head, ray number, and achene size and surface. In the last seven years since our discovery of S. ovachitensis, no in- 336 Figure 1. Holotype of Solidago onachitensis C. & J. Taylor. 507 termediates have been found by the authors despite yearly field trips in the area it 1S growing. In the Appalachian Mountains, the closest related taxa are S. caesta var. curtisii and S. flacctdifolia. As the caesia-flexicaulis complex is variously treated by taxonomists (Cronquist 1980; Gleason 1968; Mackenzie 1933; Small 1903; Radford et al. 1968; Taylor & Taylor 1983; Uttall 1984a, 1984b, and pers. comm.) the measurements of Table | are based on the strict sense of the taxa. The disagreement as to number of species to be VET , Ug co cog ye CIN IH NIN pees rer Ane ena, aoe: Set on Poe ReaMigee seen fa =_ Pee Hh Crs Pherae LE Pape SH re eee inst pe PO LES Peis | ARRAS A ace: la PERC et at CAS CRE yr EECA FA fe aera FEEL ng ~ yy LEH pata os De A PPT ami AEE BRST Bats APT siesbEseEeeesgan a fueieigs oO Solidago ouachitensis ae @ Solidago speciosa var, pallida ee Loa Figure 2. Distribution of Solidago ouachitensis and Solidago speciosa var. pallida. (R — additional locations reported by McGregor et al. 1977) 338 recognized is due to the large number of intermediates which do not clearly key to any single taxon. The characters which separate S. owachitensis from these Appalachian taxa, are the glabrous achene, larger achene size, the single ray flower, greater internode length, and very wide leaf. The vegetation of the Ouachita Mountains is closely allied with the Ap- palachian Mountains, and the dominants in the forest are similar or the same. During a cooler moister time, the goldenrod populations of these mountains occupied much larger ranges and were probably sympatric. Solidago ouachitensis is now a relict, and appears to be a holoschizoendemic as described by Keener (1983) Solidago ouachitensis is restricted to very mesic, north-facing slopes, and its limited distribution makes it one of the rarest endemics of the Ouachita Mountains and a candidate for listing as endangered under the Endangered Species Act. Populations located in Mt. Nebo State Park in Arkansas are currently protected under Arkansas State Park regulations. The type loca- tion on the north slope of Rich Mountain is de facto wilderness, but the National Forest Service has not recommended this botanically rich area be included in the national wilderness system. The third site is in private ownership. A similar holoschizoendemic, Solidago spithamaea, is also known from only three locations in the Appalachian Mountains, and is currently being processed for designation as Endangered. Other specimens examined (Fig. 2) are ARKANSAS: Polk Co.: north-facing slope | cent to Big Fork Creek, 0.6 mi NW on Hwy 8 then 0.6 mi N of the town of Big Fork Creek, J. & C. Taylor 32786 (DUR, SMU); and N facing slope and small spring on Big Fork Creek, about 18 mi S of Mena, elev. 1000— 1100 ft., McWilliams, s.n., 30 Oct 1954 (UARK); rocky dry area, top of mountain, PO. Mena, cee 2600, ea 59493 (SMU). Yell Co.: bench Byerieok: on Mt. Nebo, 8 mi E of Dardanelle, J. & C. Taylor 32770 (DUR, SMU), and rocky woods, north side of Mt. Nebo, Pa/mer 26487 es OKLAHOMA: LeFlore Co.: open woods on mountain side near Page, Stevens 2761 (DUR, US?), and top north slope of Rich Mountain, Taylor 25502 (DUR). 2. SOLIDAGO spECcIOSA Nutt. var. PALLIDA Porter. Bull. Torrey Bot. Club 19:130. 1892. S. pallida (Porter) Rydb. Bull. Torrey Bot. Club 33:153. 1906. New to NEW MEXICO: Colfax Co.: Raton Pass, 17 Oct 1932, A. GR. Nelson s.n. (RM); and OKLAHOMA: Cimarron Co.: upper slickrock edge of Fern Canyon, 5 mi E of Kenton, J. & C. Taylor 32592 (DUR, SMU). These populations at the southern end of the range of the variety (Fig. 2) appear to be relict, part of a spotty and limited distribution along the canyons and foothills of the front range of the Rocky Mountains from Oklahoma to South Dakota. McGregor et al. (1977) shows this variety maintains a thread of contact with Solidago speciosa Nutt. proper along the Niobrara River in Nebraska. 339 However, the morphological affinities of var. pallida are not with the var. rigidiuscula Torr. & Gray, the common variety in the prairies, but with var. jejunifolia (Steele) Cronq. found in sandy areas adjacent to the Great Lakes. Cronquist (1947) indicated the closeness of var. jejunifolia with pallida and based his description and recognition of jejunifolia on its separate geo- graphical distribution. The affinity of our plants with those of more northern areas indicates again the relict nature of these populations, persisting and surviving a cli- mate changing to warmer and drier conditions. ACKNOWLEDGEMENTS The following herbaria have assisted the authors by loan of specimens: COLO, CS, MO, RM, SMS, UARK. We extend our sincere gratitude to the curators of these institutions. Special thanks is given to Paul Fryxell for help with the Latin description. REFERENCES CRONQUIST, A. 1947. Notes on Compositae of northeastern United States IV. Solidago. Rhodora 49:69 — 79. _______. 1980. Vascular flora of the southeastern United States. University of North Coline Press, Chapel H GLEASON, H. A. 1968. The: new Britton & Brown illustrated flora. Hafner Publishing Company, Inc., New York. KEENER, C. S. 1983. Discribution and biohistory of the endemic flora of the mid- appalachian shale barrens. Bot. Rev. 49(1):65 — 115. MACKENZIE, K. K. 1933. Seen in Small, J. K., Manual of the southeastern flora. Published by the author, New MARTIN, W. C. and C. R. eke 1981. A flora of New Mexico. Strauss and Cramer Publishers, Piischibeer Germany. MCGREGOR, R. L., T. M. BARKLEY, et al. 1977. Atlas of the flora of the Great Plains. Iowa State eae Press, Ames RADFORD, A. E., H. E. AHLES ae C. R. BELL. 1968. Manual of the vascular flora of the ing Univesity of North Carolina Press, Chapel Hill. SMALL, J. K. 1898. Studies in the botany of the southeastern United States. Bull. Torrey Bot. Club 26:465 — 484. 1903. Manual of the southeast flora. Published by the author, New York. SMITH, E. B. 1978. An atlas and annotated list of the vascular plants of Arkansas. Student Union ee University of tere Saat ille. TAYLOR, _§. and R. J. TAYLOR. 1983. New species, new combinations, and notes on the plier ene and Ae oo — Asteraceae). Sida 10: = 1 ee ee 0 (Asteraceae) in Oklahoma and Texas. Sida 10: oe a “251. UTTAL, io i 4a. An overlooked Solidago in Virginia. Jeffersonia 15(4):72 —74. eee eer oe Solidago aap Small (Asteraceae) new to Oklahoma and other ae notes. Sida 10:324 — THE REDISCOVERY OF POTAMOGETON FLORIDANUS SMALL (POTAMOGETONACEAE) GEROULD S. WILHELM The Morton Arboretum, Lisle, IL 60532, U.S.A, ROBERT H. MOHLENBROCK Department of Botany, Southern Illinois University Carbondale, IL 62901, U.S.A. ABSTRACT Potamogeton floridanus Small, first collected in Milton, Florida in 1886 and represented by wo specimens, was named in 1903. Its taxonomic status has fallen since into ambiguity among students of the genus, owing largely to the fact that its existence has continued to be represented only by the two original collections. It has not been recognized as a distinct species by anyone other than Small since 1933. Recently the plant was rediscovered and is known now from four disparate populations. Specimens from these populations match exactly the type and the type description. Observations in the field suggest strongly chat these specimens represent a valid species endemic to the lower Blackwater River drainage in Santa Rosa County, Florida. In May, 1886, A. H. Curtiss sent to the Torrey Herbarium a pondweed from “. . . the Blackwater River, northwestern Florida.” It was regarded simply as a“. . . peculiar form of Potamogeton natans” (Morong 1886). Morong, noting this collection, described it as having “. . . small, acute, elliptical leaves, 4-6 cm long by 5— 15 mm wide, and erect peduncles about 6 cm long.” He noted further that it looked “. . . exactly like specimens in che Torrey Herbarium from India which are labeled Potamogeton natans var.” Curtiss subsequently sent a specimen collected from the same Blackwater River site in late June of the same year. Neither of these collections included fruiting material. Small (1903) included in his manual a new pondweed from the “Blackwater River, W. Fla.,” and named it Potamogeton floridanus. He com- pared it with and treated it next to P natans L., giving the overall dimen- sions of P. floridanus as smaller than those of P. natans, and with narrower floating leaves. In his description of the species, Small did not indicate that he had seen fruiting material, although in his 1913 edition he made an inexplicable reference to the drupelet. Bennett (1907) expressed little doubt that the specimen upon which Small based P. floridanus was one and the same as that upon which Morong commented in 1886; he was, however, of the opinion that the specimen SIDA 11(3):340— 346. 1986. 341 was actually P. tepperi Benn., a species which has much the appearance of a “small natans, and is often so named.” Bennett determined that the “P natans vat.” specimens from India to which Morong referred were indeed P teppert. The binomial Potamogeton floridanus Small was relegated to synonymy, with equivocations, under P. tepperi by Ascherson and Graebner (1907), apparently on the advice of Bennett. Taylor (1909) included P. floridanus under P. natans on the basis of the speculation that the former was . an immature form ...” of the latter. He noted also, however, the “. . . slender stem and leaves acute at both ends .” and that mature fruit was unknown. Potamogeton tepperi was not mentioned, either as a synonym or even as a species attributed to North America. Small (1913 & 1933), nevertheless, continued to recognize the two Blackwater River specimens as representing an indigenous, albeit rare, Florida species. Ogden (1943), apparently reluctant to determine this plant as the Asian species P. tepperi, acknowledged that it might be “. . . a pronounced ecolog- ical form of P. oakesianus or P. natans . . .” but pointed out that neither species “. . . has been otherwise found within 600 miles of Florida.” He preferred to regard the two Curtiss specimens as representing hybrids be- tween some linear-leaved species and P i//inoensts Morong, although his own detailed studies of the stem anatomy caused him to cast considerable doubt upon this hypothesis. As a result of the passage of the Endangered Species Act of 1973, the Smithsonian Institution was directed to review the status of the nearly 25,000 kinds of plants which are native to the United States. Potamogeton floridanus was among the plants that were nominated for additional consid- eration, but since it had not been seen alive since 1886, it was subsequently listed in the Federal Register as “possibly extinct.’ Haynes (1978), nearly one hundred years after Curtiss collected his ma- terial, could add nothing more to our understanding of this plant. He retired the problem by noting that “. . . the exact nature of P /loridanus Small, based on two collections made by Curtiss in 1886 (NY), is uncer- tain.” The absence of additional collections, along with equivocations by the major students of the group, left the floristic botanist with little choice other than to relegate P floridanus to synonymy or to ignore it altogether. The only recent authors whom one might have expected to treat P. flort- danus chose not to do so (Ward 1968, 1979; Godfrey and Wooten 1979). In 1980, when an update of the status of United States plants appeared in the Federal Register, P floridanus had been dropped from the list because of its “ ie 342 reputed hybrid origin, since it had been decided by personnel charged to operate the endangered species program that hybrids no longer would be considered. Three recent collections (Wilhelm & Blackmon 9706, 23 November 1981, NY; Wilhelm & Blackmon 11522, 10 Jul 1983, USF; and Burkhalter 9398, 3 Jun 1984, UWFP), with attendant field observations, have added more information regarding the taxonomic status of P floridanus. Sadly, however, Small’s (1913) inexplicable reference to the drupelet notwith- standing, fruits for this species have yet to be collected. The idea that P. floridanus may represent an early introduction is, on the face of it, not altogether spurious. The old port of Pensacola long has been a place of entry into this country for ballast weeds and other species native to areas remote from the central Gulf coast region (Mohr 1878). Nayjas ancis- trocarpus Magnus, for example, was collected “. . . in tidal creeks near Milton at the head of Pensacola Bay” and reported under the name N. con- ferta A. Br. (Fernald 1902). Specimens, however, of the Asian species Potamogeton teppert (Litvinov 3352 MO, determined by Bennett, and Merri// 1723 MO) are coarse broad-leaved plants with rounded leaf bases and stout petioles and pedun- cles, resembling P natans much more closely than the Florida material. The recent collections of P. floridanus match exactly the description of Small’s species and the type specimens (Fig. 1). Examination of all the herbarium specimens and plants in the field reveals a degree of uniformity in floating-leaf morphology which is typical of other Potamogeton species. The floating leaves are lanceolate, tapering at both ends, and with long slender petioles. The length/width ratio of P floridanus is 5.1 + 0.8; that of P. tepperi is 1.8 + 0.3. Taylor's implication that the Curtiss specimens are likely to be nothing more than immature forms of P. natans is to us unacceptable. Obvious morphological differences notwithstanding, P natans remains unknown from the southeastern United States (Godfrey and Wooten 1979). Potamogeton oakestanus Robb., the other possible species of which Ogden speculated P. floridanus might be a “pronounced ecological form,” is even more remote from the Gulf coastal plain than P. natans (Fernald 1950). Ogden’s equivocal speculation that Curtiss’ collections represent hybrids between a broad-leaved and narrow-leaved species is not only inconsistent with his own anatomical studies, but field observations on the habitat and associates of the plant further frustrate the hybrid hypothesis. There currently are four small disparate populations of P floridanus known, all in the vicinity of Milton. Two are in Pond Creek, a clear- flowing tributary of the Blackwater River; one just south of the U.S. Route 343 90 bridge, the other about 4 mile southwest of there in the NEY% SW Sec. 9, TIN, R23W. A third population is in the tidal channel connecting Bob’s Bayou with the Blackwater River north of East Milton, and the other is at the entrance to a tidal channel of an unnamed bayou off the west side of the Blackwater River in the SW Sec. 25, T2N, R28W. In each case the plants were found growing in two to four feet of water and in the vicinity of SOLYPE f Absbams 1902. Kesbert K Revision Broad-icaved N. Am. Species of Potamagetou pe Sheets Py LLLIN ~loaved species E.COgden 1944 : at ke suatos one AIF HERTAMIGM DE OTTG KURT IE , 0 “ : ete: De oF ry “a> PD - . ad i Frosented by MR, W OAT e i geese i a OUTER SERTLE, tTaRomabart ‘e es 4 a Assent; sue ge tir at £ ro coe 4 kiwwakes 4 pee FE tee nev tl “ant te; faced fe Ye fps ide i a MEME YORK i> en Or, he beck, 49h % wp tN Yao ay, ao é : a Ay oe 7 - ee Cth, Al, , ¥ fo ho) pe ee é L BR a ee rere ia f es Figure |. Isotype of Potamogeton floridanus Small (NY). 344 Potamogeton diversifolius Raf. No other species of Potamogeton were observed. The narrow-leaved species P. curtissii# Morong (=P. foltosus var. macellus Fern.) was described by Morong (1886) based upon material collected by Curtiss from the same location he collected P. floridanus. We have been unable to relocate these plants. The reluctance by students of the genus to recognize P. floridanus as a species seems to have originated not so much from a lack of morphological distinctness as from the fact of its very restricted range, and from the fact that it remained unrepresented by additional collections. Endemism in the central Gulf coast area, however, is not uncommon. There are quite literally dozens of species with very restricted ranges in the region. Baptista calycosa var. villosa Canby, Chrysopsis godfreyi Semple, C. gossypina ssp. crutseana (Dress) Semple, Conradina glabra Shinn., Crataegus lacrimata Small, Eriocaulon lineare var. gigas Mold., Hymenocallis choctawensis Traub, H. henryae Traub, Hypericum chapmanii P. Adams, H. J/issophloeus P. Adams, Lilium iridollae Henry, and Verbesina chapmanti J. R. Colem. are only a few of many examples. It is likely that Potamogeton floridanus was somewhat more widespread in the Milton area in the O’s when Curtiss was able simply to visit reaches of the river in the vicinity of the railroad whistle-stops, but the water in such areas today apparently is no longer suitable for the plant. It is certain, with the chronic degradation of our rivers, streams, and lakes, that the presettlement populations of most of our native aquatic plants have been decimated several times over. Potamogeton floridanus is clearly not an immature form of P. natans, nor is it a “pronounced ecological form” either of P natans, P. oakesianus, nor any other Potamogeton. It 1s not the Asian P. fepperi. It is our opinion that P floridanus is a valid native North American species, unlikely to be of hybrid origin. It is still extant in at least four locations near Milton, Florida. Be- Tilano et Giada ia 1S Vila phiacaler gen e Cvhicu. unbf loc. ff. Flaw Ja Cael AiCiute Ga (396° (lard. tas a fee fee fried ug frct 7 Figure 2. Label with original handwriting from isotype of Potamogeton floridanus Small, collected by . H. Curtiss, May, 1886. Figure 3. Specimen of Potamogeton floridanus Small, collected by Wilhelm & Blackmon (9706) NY, 23 Nov 1981. 346 cause of its extreme rareness, narrow distribution, and obvious vulnerabili- ty, we feel 1t would be appropriate to reconsider it as a federally endangered species. REFERENCES ASCHERSON, P. and i GRAEBNER. 1907. Potamogeton. In: A. Engler, ed., Das Pflanzenreich 4(IL):6 BENNETT, A. 1907. ne on Potamogeton. J. of Bot. 45:373. FERNALD, M. 1902. Some little-known plants from Florida and Georgia. Bot. Gaz. 33:154— 157 pee DO Cais manual of botany. American Book Co. 8th ed. Ixiv + 1632 pp. GODFREY, R. and J. WOOTEN. 1979. Aquatic and wetland plants of the southeastern United States. Univ. of Georgia Press, Athens. HAYNES, R. 1978. The Potamogetonaceae in the southeastern United States. J. Arnold Arbor. 59:170— 191. MOHR, C. 1878. ees plants introduced into the Gulf States. Bor. Gaz. 3:42 — 46. MORONG, T. 1886. A new species of Potamogeton. Bull. Torrey Bot. Club 13:145. OGDEN, E. 1943. ae broad-leaved species of Pods aaueion of North America north of Mexico. Rhodora 45: 192. SMALL, J. K. 1903. Flora of the southeastern United States. Publ. by the author, N.Y. p. a7. 1913. Flora of the southeastern United States. Publ. by the author, N.Y. p. 40. ——_—_—.. 1933. Manual of the southeastern ae Publ. by the author, N.Y. p. 16. TAYLOR, N. eae Zannichelliaceae. In: N. L. Britton and L. M. ana ed, Bi. Amer. Fl. 17:16. AR D. 'D. _ Le Checklist of the vascular flora of Florida. Part 1. Florida Agri. Exp. Sta. Bull. 726. ee, 1979, Rare and endangered biota of Florida: Plants, Vol. : Ligies Committee on Rare and Endangered Plants and Animals. Univ. of Fla. Pre RECENT COLLECTIONS AND STATUS OF LESQUERELLA LYRATA ROLLINS (CRUCIFERAE) DAVID H. WEBB 1729 Beckman Drive, Florence, AL 35630, U.S.A. ROBERT KRAL Department of Biology, sii University Nashville, TN 37235, U.S.A. ABSTRACT he narrow endemic Lesquerella lyrata Rollins (Cruciferae), considered by Department of Interior specialists as possibly extinct, was found in several localities, including the type locality, in April 1984. Ideas as to its sporadic yet abundant occurrence are proposed, and positive management practices suggested. Lesquerella lyrata Rollins is a narrow endemic reported to be restricted to a few localities (Fig. 1) in the eastern portion of Franklin County, Alabama (Rollins 1955, Rollins and Shaw 1973). It isa component of the glade flora and is generally found in association with limestone outcroppings. The rarity and restricted distribution of L. /yrata have resulted in its listing as an endangered species in Alabama (Freeman et al. 1979a, 1979b). Lesque- rella lyrata also is currently under review by the U. S. Department of the Interior (1980) as an endangered or threatened species and is included in the rare plant inventory prepared by Kral (1983) for the USDA Forest Serv- ice. Recent collections of L. /yrata reported in this paper are of significance since the Department of the Interior lists L. /yrata as possibly extinct. HABITAT AND POPULATION SIZE During early April of 1984, L. /yrata was rediscovered near the type locality described by Rollins (1955) as 7 miles east of Russellville near Richardson's Crossing. Even earlier that month a large population of L. lyrata was found on limestone outcroppings and in adjacent fields in the vicinity of Spring Valley in Colbert County, Alabama (Fig. 1). The popula- tion at Richardson’s Crossing consisted of several hundred plants scattered in a field that was plowed during the fall of 1983. An adjacent field that was not plowed in the fall of 1983 but apparently cropped during 1983 was found to harbor more than a thousand plants of bladderpod. An additional SIDA 11(3):347 — 351. 1986. 348 population of a few dozen individuals (Fig. 1) was observed along the roadside of County 83 approximately 0.8 mile north of Richardson's Cross- ing. The population of L. /yrata just north of Spring Valley consisted of sever- al thousand plants scattered along the roadside, in adjacent fields, and around small limestone outcroppings. Hundreds of plants occurred in un- planted cotton fields that were plowed and disked in the fall of 1983. An additional population of a couple of hundred plants was observed in the same glade system in a pasture approximately one-half mile west of the large population. The presence of large populations numbering in the thousands was con- firmed again in late March and early April of 1985 just north of Spring Valley and at Richardson's Crossing. A large population of L. /yrata also was present at Richardson's Crossing in the early spring of 1983 (Reed C. Rollins pers. comm.). A population of several thousand plants was found in 1985 ina pasture 0.8 mile north of Richardson's Crossing where only a few dozen plants were observed in April 1984. Although the two other localities cited in the original description by Rollins (1955) were visited during April 1985 as well as other suitable habitat in the eastern portion of Franklin County, no additional populations of L. Lyrata were located. The populations of L. /yrata from Franklin and Colbert counties repre- sent two separate and distinct glade systems. While both are in the Interior Low Plateau described by Fenneman (1958), the Spring Valley populations are in the Tennessee Valley and those at Richardson's Crossing in the Moulton Valley. These two subdivisions of the Interior Low Plateau are separated by Little Mountain which is characterized by sandstone outcrop- pings described by Harper (1942). The collection sites in the Tennessee Valley and Moulton Valley vary by about 200 ft. elevation. COMMENDED MANAGEMENT PRACTICES Lesquerella ee likely evolved on the glade systems that are now highly disturbed and occur as isolated pockets surrounded by agricultural lands. Acquisition and protection of the glade systems containing L. /yrata are paramount in protecting the species. The glade system near Richardson’s Crossing also has other rare plant species listed at the State and Federal level such as Leavenworthia alabamica Rollins var. alabamica, Delphinium alabamicum Kral, Psoralea subacaulis T. & G., Isoetes butleri Engelm., a calcaricum Ware. Petalostemum taemneee Heller, and P foltosum . The latter species was reported by Baskin and Baskin (1973) just oni of Richardson's Crossing but has not been relocated by the authors. Muscle Shoals Tuscumbia ( Tennessee Valley ( Colbert Py County Little Mountain _—-—_— - aw Lawrence A | County ! Moulton Valley : ! gw Richardson's Franklin Crossing al County | | | | Cumberland Plateau | ee | 5 Miles A Collections cited by Rollins (1955) @ Type locality (Rollins 1955) @ 1984 and 1985 collections Figure 1. Distribution of Lesquerella lyrata in northwestern Alabama. 350 The large numbers of L. /yrata in cultivated fields is a phenomenon that also has been observed for Leavenworthia alabamica vat. alabamica, L. crassa Rollins, and Lesquerella densipila Rollins at scattered localities in Colbert, Franklin, Morgan, and Lawrence counties, Alabama. These taxa apparent- ly move from the thin soil of glades into fields that are planted tn cotton or soybeans on an annual basis. The typical mid-May to June planting of soybeans allows these early flowering annuals to set seed prior to soil tillage and planting. The presence of Lesquerella and Leavenworthia in unplanted cotton fields may be sporadic since the mid-April to May planting and the earlier spring application of pre-emergent herbicides may prevent or preclude the establishment of large populations. However, wet springs such as that in 1984 that delay soil tillage may allow germination and completion of the life cycle of Lesquerella and Leavenworthia. The impor- tance of seed banks in cultivated fields in relation to survival of these species should not be ignored. Conservation easements with private land owners should be pursued. As evidenced by the populations numbering in the thousands, agricultural use and survival of these species are not incom- patable. Preservation of selected glade systems along with easements that harmonize tillage patterns and herbicide applications with the life history of these rare annuals should insure their survival. Further research is re- quired relating to the use of herbicides and germination of species of Lesquerella and Leavenworthia. CITATION OF RECENT COLLECTIONS ALABAMA. Colbert Co.,: sandy silty clay of plowed field by Co. 77 just N (0.5 mi) of Spring Valley, 2.2 miS of jct AL 157 and Co. 77, 2 Apr 1984, Kral & Webb 71171 (VDB); thin dark soils over limestone and in small limestone glade, horse pasture by schoolhouse, Spring Valley School, W side of Spring Valley, 2 Apr 1984, Kral & Webb 71175 (VDB). Franklin Co.: fallow field on N side of AL 24, 6.0 mi E of ject AL 24 and US 43 (in pane along AL 24 near jct with Co. 83, 7 Apr 1984, Webb & Pardue 4943 (TENN, B); same locality, 22 Mar 1983, Rollins & Rollins 8317 (GH); pasture on E side of Co. - 0.8 mi N of jct with AL 24 (Richardson's Crossing), ca 6 mi E of Russellville, 6 Apr 1985, Webd 5077 (TENN, VDB). ACKNOWLEDGEMENTS We thank Professor Reed C. Rollins for permission to cite his recent collection of L. /yrata, REFERENCES BASKIN, J. M. and C, C. BASKIN. 1973. The past and present distribution of Petalostemum Car and notes on its ecology. Rhodora 75:132— 140. NNEMAN, N. M. 1938. Physiography of eastern United States. McGraw-Hill Book Co., Inc., i York. 714 pp 351 FREEMAN, J. D., A. S. CAUSEY, J. W. SHORT, and R. R. HAYNES. 1979a. Endangered, threatened, and special concern plants of Alabama. J. Alabama Acad. Sci. 50:1 — 26. FREEMAN, J. D., A. S. CAUSEY, J. W. SHORT, and R. R. HAYNES. 1979b. eae chrearened, and special concern ai of Alabama. Departmental Series urn University, Auburn, Alabam HARPER, R. M. 1942. Natural resources of ne Tennessee Valley Region in Alabama. Geological Survey of Alabama, Special Report 17. University, Alabama. 93 pp. KRAL, R. 1983. A report on some rare, threatened, or endangered forest-related vascular plants of the South. Vol. I. Isoetaceae through Euphorbiaceae. Technical oi R8-TP 2, March 1983. Published by USDA Forest Service, Atlanta, Geor ROLLINS, R. C. 1955. The auriculate-leaved species of Lesquerella (ces are 57:241— 264. ROLLINS, R. C. and E. A. SHAW. 1973. The genus ab os (Cruciferae) in North America. Harvard University Press, Cambridge, Mass. 288 pp. U. S. DEPARTMENT OF INTERIOR. 1980. a and threatened wildlife and plants: Review of plant taxa for listing as endangered or threatened species. Federal Register 45, No. 242, 15 Dec 1980, 82480 — 82569. NOTES A RECOMBINATION IN HYDRANGEA L. (SAXIFRAGACEAE)— HyDRANGEA ARBORESCENS L. f. carnea (Raf.) L. J. Uttal comb. et stat. nov. Hydrangea vulgaris Michx. var. carnea Raf. New Fl. 3:77. 1838. Flowers incarnate. No type having been designated, the description stands as the type. This 1s an apparently very rare color form of the common white-flowered Hydrangea arborescens L. of the eastern United States in which the outer surface of the petals and calyx lobes are the color of raw beef. The cyme is very conspicuous and attractive. Anthocyanic coloration of flowers in Hy- drangea L. is common in Asiatic species in shades of pink, blue, violet, or purple. The deep-flesh color of the present taxon is the only deviation from white flower color known from North American Hydrangea. One specimen: TENNESSEE. Unicoi Co: Unaka Mountain, elev. ca 1300 m, Cherokee National Forest Service Road 132, 1 m shrub with normal white- flowered shrubs among rocks near summit, 8 Jul 1985, Usta/ 13881 (VPI). It and the white-flowered shrubs belong to H. arborescens subsp. arborescens sensu McClintock (Proc. Calif. Acad. Sci. 29:147 — 256. 1957), and lack marginal sterile flowers with dilated sepals. No reports of other specimens or literature references have been found by the author except for synonymy under H. arborescens in Rehder (Bibliography Cultivated Trees and Shrubs. 199. 1949).—L. J. Uttal, Department of Biology, Virginia Polytechnic Insti- tute and State University, Blacksburg, VA 24061, U.S.A. MONERMA CYLINDRICA (POACEAE: MONERMEAE) NEW TO TEXAS.—A recent collection from Northeastern Louisiana University (R. Dale Thomas 88432 & B. E. Dutton 1665) was received by the S. M. Tracy Herbarium (TAES) for identification and identified as Monerma ee (Willd.) Coss. & Dur., an adventive species native to the Previously, this species was reported from San Francisco Bay south to California Norte (Gould & Moran 1981, Gould & Shaw 1983) and recently from southwest Louisiana (Thomas & Dutton 1985). While examining specimens at TAES, an additional collection of M. cylindrica from Texas was located which had previously been misidentified as Parapholis incurva (L.) C. E. Hubb., a closely related taxon of the same tribe. Parapholis is reported growing in saline coastal sites along the Atlan- tic and Pacific Coasts and in Texas along the Gulf of Mexico (Gould & Shaw 1983). Monerma is characterized by having only a single glume per spikelet with the exception of the terminal spikelet where two glumes are present SIDA 11(3):352. 1986. 353 and alternate on the rachilla. Parapholis can be easily distinguished from Monerma by the presence of two glumes for each spikelet, these adjacent to one another (Chase 1951). Our identifications were corroborated by Stephan L. Hatch, curator of the S. M. Tracy Herbarium. After a review of literature and herbarium specimens deposited at TAES and the University of Texas Herbarium (TEX and LL), we concluded M. cylindrica to be a new record for Texas. This species is not included in Correll and Johnston (1970) or Gould (1975). Collection data of M. cylindrica : UNITED STATES. Trxas. Galveston Co.: Occasional, clay loam banks at intake canal of PR. H. Robinson Generating Station near Bacliff, 8 May 1974, Waller and McAden 2661 a TAES, TEX). This is Technical Bulletin TA No. 20687, Texas Agricultural Experi- ment Station.—Chuck R., Coffey and Jesus Valdes R., Department of Range Science, Texas AGM University, College Station, TX 77843, U.S.A REFERENCES CHASE, Ae 1951.2 ALS: ie manual of the grasses the United States, 2nd ed. U.S.D.A. Misc. Publ. CORRELL, D. S. and M. ‘ TF OaNGHON. 1970. Manual of the vascular plants of Texas. Texas Research Foundation, Renner, Texas. GOULD, E W. 1975. The grasses of Texas. Texas A&M University Press, College Station, Texas. GOULD, E W. and R. MORAN. 1981. The grasses of Baja California, Mexico. Society of Natural History, San Diego, California. GOULD, FE W. and R. B. SHAW. 1983. Grass systematics, 2nd ed. Texas A&M University Press, College Station, Texas OMAS, R. D. and B. E. DUTTON. 1985. Monerma cylindrica new to Louisiana. Phyto- logia 57:366. REVIEWS PLANT DISEASES: Infection, Damage and Loss. 1984. Edited by: Wood, R.K.S. and G.J. Jellis. Blackwell Scientific Publications, Ltd., i: London, Edinburgh, nee Palo Alto, Melbourne. 327 pp. + vit. Cloth, $4: The book is based on a symposium held at the University of Surrey in December 1982 by the British Society for Plant Pathology. The objective was to publish the review papers of the symposium, whose theme was to survey the effects of pathogens on plants, for use by senior undergraduates, postgraduates, and scientists in general. The Chapters (review articles) are grouped into four sections: 1) How pathogens cause disease (four Chapters), 2) Physiological responses of SHOA TE 3) 3936-0980; 354 plants to pathogens (five Chapters), 3) Infection and host damage (seven Chapters), and 4) Damage and loss (ten Chapters). Each Chapter presents a review of that topic along with a rather extensive bibliography on the subject. The wide variety of topics concern cultivated crops such as cotton, potatoes, cereal grains, and ornamentals as well as forest crops. An Index to the book occurs at the end. The 26 Chapters were authored by thirty-three plant pathologists from Britain (26), Denmark (1), France (1), Italy (2), Australia (1), and the United States (2). On page 285, a permanent correction has been affixed over the original printed material concerning the authors of the article. Otherwise, the printing and layout exude quality throughout.—WFM THE VASCULAR PLANTS OF SOUTH DAKOTA. 1985. Theodore Van Bruggen. 2nd ed. lowa State University Press, Ames, Iowa, 50010. 476 pp. + xxv. Paperback, $28.95 This manual is an updated version of the 1976 edition. It was printed by The Iowa State University Press from camera-ready pages provided by the author. The Introduction contains information on the geology, physiog- raphy, and climate with supporting maps. Photographs aid in illustrating the vegetation. This section is an excellent treatment of the past and present geological and botanical history of the state. n the Statistical Summary, the number of vascular plant species totals 1608 for South Dakota. This number includes the native and introduced taxa with no distinction between them. There are three principal vegeta- tion regions represented: Eastern Deciduous Flora, Plains and Prairie Flo- ra, and the Rocky Mountain Flora. The following statement addresses the endemic flora: “If one assumes a less than conservative taxonomic interpre- tation of the species present, and their ranges, ic 1s doubtful that any endemics are present in the state.’ The systematic treatment has keys to the major groups and their subcategories down to the generic level that are typical dichotomous keys with one to three characters per unit. However, keys to the species include brief diagnostic characters or descriptions, frequency, distribution, synonymy, flowering or fruiting (months) terminating in the binomial with the author citation. A glossary, general references, and an index to common and scientific names follow the systematic treatment. Even though its odd size may be more common than I realize (15 cm wide by 19.4 cm tall — paperback), it will lie flat when it is open if it is opened near the middle of the text. This is extremely helpful when one is keying out a plant. Strophostyles (p. 275) is not in the Index although its common name ts listed. The specific epithet of Desmanthus illinoensis is mis- SIDA 11(3):354. 19806. 355 spelled and the full author citation should read (Michx.) MacM. ex Robins. & Fern. This manual is well worth possessing, not only for the identification of the flora of South Dakota, but for the past and present history associated with the state from a botanical viewpoint.—W EM ACTA BOTANICA HUNGARICA. Volume 29, numbers | — 4, 1983, pp. |1— 399. Twenty-one papers. Studies in Rondeletieae (Rubies. IV. A new genus: Javorkaea. A. Borhidi, Magda Jarai-Komlodi.—Studies in Rondeletieae (Rubiaceae), V. Los limites del género Swheranthus. A. Borhidi, Mayra Fernandez Zequeira.—Studies in Rondeletieae (Rubiaceae), VI. Estudio taxondmico de la Rondeletia odorata Jacq. Mayra Fernandez Zequeira, P Herrera Oliver.—Studies in Rondeletieae (Rubiaceae), VII. The significance of leaf epidermis for taxonomy tn Neomazaea sensu lato. M. A. ea del polen de las especies cubanas de Gymnospermas. L. Stuchlick, Milagros Moncada.—New names and new species in the flora of Cuba and Antilles, III. A. Borhidi.—Xylotomic examination of some Venezuelan Capparis species, Il]. K. Babos, I. R. Bermudez, I.J.C. Cumana.—Contribucion al estudio anatomico del xilema de la familia Simarubaceae en Cuba, I. A/varadoa Liebm. y Stmaruba Aubl. M. A. Vales, Candida Martinez.—Plant commu- nities of Cuba, I. Fresh and salt water, swamp and coastal vegetation. A. Borhidi, O. Muniz, E. Del Risco. Volume 30, numbers 1—2, 1984, pp. 1—247. Eighteen papers. Revision del género Machaonia H. et B. (Rubiaceae) en Cuba. Mayra Fernandez, A. Borhidi.—Cytological investigation of Sci//a bifolia popula- tions in Hungary I. Z. Kereszty, L. Szilagyi.cComparative anatomy of the androecium of male sterile and fertile sunflowers (Helianthus). M. Szabo, $. Gulyas, J. Frank. —Plant communities of Cuba, I. The riverside scrub vegetation,. A. Borhidi, R. Capote.—Reproductive allocation in the stages of sandy succession, II. Erigeron canadensis L., Polygonum arenarium WW. et K. Erika Melko. Volume 30, numbers 3—4, 1984, pp. 249—480. Fifteen papers. Morfologia de granos de los polen de las Chloranthaceae y Canellaceae cubanas. L. Stuchlick.—Xylotomic examination of some Venezuelan species of the Capparidaceae, I. K. Babos, I. R. Bermudez, [.J.C. Cumana.—Estudio taxonémico del género Ariadne Urb. (Rubiaceae). Mayra Fernandez, A. Borhidi.—A preliminary numerical taxonomic study of the Scilla bifolia agg. (Liliaceae, Scilloideae) 1n Hungary. Z. Kereszty, J. Podant.—BLL SIDA 11(3):355. 1986. S DA CONTRIBUTIONS TO BOTANY VOLUME 11 NUMBER 4 DECEMBER 1986 CONTENTS A new species and taxonomic notes on Gentianella (Genti- anaceae) in South America. James S. Pringle. A new combination in Chimaphila (Ericaceae). Laurence _]. Dorr. Renovation of Dyssodia (Compositae: Tageteae). John L. Strother. Noteworthy plants from north Florida. Ul. Loran C. Anderson. Synopsis of the Florida species of Pectis (Asteraceae). David J. Keil. Bumelia dominicana (Sapotaceae), a new name for an old sapote. R. David Whetstone and T: A. Atkinson. Taxonomic and nomenclatural notes on Vaccinium L. section Cyanococcus (Ericaceae). Leonard J. Uttal. A new combination in Hedyotis L. (Rubiaceae). R. PR Wunderlin. Millerocaulis, a new genus with species formerly in Osmun- dacaulis Miller (Fossils: Osmundaceae). William D. Tidwell. Wolffia papulifera Thompson (Lemnaceae), new to Michigan. William J. Hess. A new variety of Hedeoma hyssopifolium Gray (Lamiaceae). James Henrickson. Sabal etonia (Palmae): systematics, distribution, ecology, and comparisons to other Florida scrub endemics. Scott Zona and Walter S. Judd. Notes about Psoralea sensu auct., Amorpha, Baptisia, Sesbania and Chamaecrista (Leguminosae) in the southeastern United States. Duane Isely. An undescribed Panamanian Vaccinium: Vaccinium bocatorensis (Ericaceae). Robert L. Wilbur. (continued on back cover) a7 370 371 a1? 385 396 397 400 401 407 US ISSN 0036-1488 SIDA, CONTRIBUTIONS TO BOTANY Founded by Lloyd H. Shinners, 1962 Publisher Wm. F. Mahler SMU Herbarium Dallas, Texas, 75275 Editor Associate Editor Barney L. Lipscomb John W. Thieret SMU Herbarium Northern Kentucky University Dallas, Texas, 75275 Highland Heights, Kentucky, 41076 Guidelines for contributors are available upon request. Subscription: $10.00/$15.00 (U.S.) per year, numbers issued twice a year. © Sida, Contributions to Botany, Volume 11, Number 4, pages 357 — 497. Copyright 1986 by Wm. E Mahler A NEW SPECIES AND TAXONOMIC NOTES ON GENTIANELLA (GENTIANACEAE) IN SOUTH AMERICA' JAMES S. PRINGLE Royal Botanical Gardens Box 399, Hamilton, Ontario, CANADA L8N 3H8 A NEW SPECIES FROM NORTHERN PERU GENTIANELLA Chlorantha Pringle, sp. nov. Figs. 1,2. Suffrutex caulibus usque ad 2.5 dm altis, erectis, ramosis. Folia numerosa, adscendentia, elliptica, 6—8 mm longa, 1.5 — 2.5 mm lata, trinervia nervo medio infra anguste carinato, obtusa. Flores solitarii, subsessiles. Calyx 7.5 —9.5 mm longus. Lobi calycis oblongi vel ovato-oblongi, 2.5 — 3.5plo longiores quam tubus, acuti vel subacuti. Corolla 14— 18 mm longa, viridis, lobis obovatis circa 2.25plo longioribus quam tubo et circa 1.6plo longioribus quam latioribus, apicem versus rotundatis et eroso-undulatis. Corolla intus glabra vel sub sinubus trichomatibus paucis minutisque. Filamenta circa 5.5 mm longa. Antherae caesiae. Ovarium stipitatum. Subshrub with several erect or suberect stems 0.7 — 2.5 dm tall; most stems dividing into 2-several strongly ascending branches, with flowering and vegetative branches present simultaneously. -Leaves densely spaced throughout, a older portions of the stems ringed with old leaf bases, the distal. 05 = dm densely leafy at flowering time. Leaves mostly 6 — mm long a 1.5—2.5 mm wide, consisting of an erect, pseudopetiolar portion |— 2 mm long and an ascending, elliptic to ovate-elliptic blade, prominently 3-nerved (sometimes with an additional pair of lesser nerves) with the midrib narrowly carinate below, the apex obtuse, thickened. Flowers solitary (sometimes appearing grouped when terminating 2 or 3 short branches), erect, subsessile or on peduncles less than 2 mm long. Calyx 7.5—9.5 mm long, with erect, oblong to ovate-oblong lobes 2.5—3.5 X as long as the tube, subacute to acute. Corolla 14— 18 mm long, green throughout, with the lobes obovate, ca 2.25 X as long as the tube and ca 1.6 X as long as wide, distally rounded, erose-undulate. Inte- rior corolla surface glabrous or with a very few inconspicuous trichomes below the sinuses. Stamens inserted at ca 0.67 X the length of the corolla tube; filaments ca 5.5 mm long; anthers bluish. Ovary stipitate. ‘Contribution No 58 from the Royal Botanical Gardens, Hamilton, Ontario, Canada. SIDA 11(4):357 — 369. 1986. 358 TYPE COLLECTION: . AMAZONAS. Bagua: Cordillera Colan NE of La Peca, ca 10,400 ft, humid ee ae 3439 (HOLOTYPE: MQ; tsorype: HAM). Known only from the type collection. Previous authors have recognized an excessive number of species of Genti- anella (formerly treated as a subgenus of Gentiana) from much of Pert. Macbride in 1959 reduced many specific names to synonymy, and further reduction has occurred subsequently (Pringle 1981). However, when Gilg’s (1916) and Macbride’s (1959) publications were written, few specimens from the Departamento de Amazonas were available. In view of the relatively restricted ranges of many of the Andean species of Genti- anella, newly discovered species from this part of Pert are hardly surpris- ing. Gentianella chlorantha appears to be relatively closely related to G. radicata (Griseb.) Pringle, which was described from specimens collected in the mountains above Lima. Gentianella chlorantha differs from G. radicata in its fewer and larger flowers; in its green corollas (white or pale '; | fai Wes wi has b Fig. 1. sae a ek a, portion of corolla, interior surface, and stamens; b, portion of calyx, exterior surface; c, pistil; d, 322 ( we sealed ie portion of holotype (MO). Figs. 3 — 6. Small plants of Gentianella spp. oe and 6 pressed with corollas spread out to show lobe shape. Figs. 3 — 4. C. eee pene ; = 30 8). Figs. 5 — 6. G. rupicola, Asplund 17333 (S). Figs. 2— 6 to same scale. 360 violet with deep purple veins in G. radicata, according to Macbride and various label data, although not well preserved in Grisebach’s material {“flavae? . . . striatae”}); and in its broader, elliptic rather than linear leaves. GENTIANELLA CERASTIOIDES and G. RUPICOLA The distinctness of Gentianella rupicola (H.B.K.) Holub from G. cerasti- oides (H.B.K.) Fabris has been a long-persistent question. In 1916, Gilg commented that it was often difficult and sometimes impossible to deter- mine to which of these species a specimen should be assigned. Gilg, how- ever, felt that evidence to support the uniting of these taxa was inadequate at the time, and that the differences between the extreme (and nomenclatu- rally typical) forms of the respective species were much greater than would normally be acceptable within a single species. Fabris (1960) did reduce G. rupicola to synonymy under G. cerastivides, but his wording likewise in- dicated uncertainty. Much of what passed as Gentiana rupicola in Gilg’s time actually was Genttanella cerastioides, as indicated by his descriptions of G. rupicola and his citations of specimens. Very little material of true G. rupicola was avail- able for study by Gilg (1896, 1916) or later by Macbride (1959) and Fabris (1960), and most of the few specimens of G. rupicola that they did see were of poor quality. Consequently, Macbride’s description of Gentiana rupicola, and Fabris’s concept of this species as possibly merely a high-altitude form of G. cerastivides, appear to have been based largely on misidentified materi- al of G. cerastioides. The extensive collecting for the Flora of Ecuador project has increased by severalfold the specimens of this complex available for study, making possible a new approach to this problem. Descriptions of these species will appear in the Flora of Ecuador, but, in view of the statements made by Fabris (1960) and the wide use of his monograph, my acceptance of G. rupicola as a distinct species requires a more thorough discussion than would be appropriate in the Flora. Gilg (1896, 1916), Macbride (1959), and earlier authors differentiated these species primarily on the basis of height, erectness of stems, length of internodes and pedicels, and number of flowers per stem. Their descrip- tions of the flowers indicated virtually no differences. In the present study, however, I found that all specimens in this group could readily be sorted into two taxa on the basis of floral characters. I also noted that those persons who collected both of these taxa in the same area consistently kept them separate, under different numbers. A summary of the more conspicuous differences between these species is presented in Table 1. A more detailed discussion, emphasizing recognition of these species in the herbarium, appears below. 361 Taste |. Summary comparison of the more conspicuous differences between Gentianella cerastioides and G. rupicola. CHARACTER G. CERASTIOIDES G. RUPICOLA Length of flowering stem 0.5 — 17 (-25) cm 0.5—3 (-11) cm Number of flowers per stem Usually | or 2, occasionally Usually 1, occasionally up to 6 2 or 3 Pedicel length 0.4-—5 cm 0.1—2.5 (-4) cm Corolla length (closed) 14-) 17-40 mm (14-) 17-28 mm Position of corolla lobes Spreading, the corolla opening Incurved, the corolla widely scarcely opening Shape of corolla lobes Narrowly cuneate-obovate, not Elliptic, somewhat ventricose ntrico Corolla color Usually sees or tees violet, Usually red (crimson to scarlet), Ss onelly pink or white occasionally orange, rarely yellow Corolla color, when retained, sharply differentiates G. rupicola from G. oe In G. rupicola, it is usually red, often described as “crimson, ’ ‘ ht red,” or “scarlet,” occasionally orange-red, rarely yellow. In con- trast, corollas of G. cerastioides range from pink through various shades of violet to violet-blue or white. Most corollas are medium to pale violet, although a few specimens have fairly deep blue-violet corollas. (Corollas may sometimes appear bluer when dried than when fresh, but the contrast between those of G. rapicola and those of G. cerastioides remains.) In some specimens of G. cerastiordes, the corollas are minutely spotted. Even in an- cient specimens in which corolla color is not well preserved, the corollas of G. rupicola appear darker and more reddish than those of G. cerastioides. Corolla color is correlated with the shape of the intact corolla and with the shape of the corolla lobes (Figs. 3 — 6). Balls (zw sched.) described the corollas of his B.7281 (EK), representing true G. rupicola, as “long- globular in shape. The petals do not appear to open, so that the flower has always the appearance of being in rather full bud.” Pressing tends to push the petals apart, but they remain incurved toward the summit. The corollas of G. cerastioides, in marked contrast, open widely, ranging from funnelform to nearly rotate, with outwardly flaring lobes. This characteris- tic is particularly well shown in photographs taken in the field (e.g., photo attached to Rawh-Hirsch E310 {F}), but is likewise indicated in the aspect of herbarium specimens The corolla lobes of G. rupicola are more or less elliptic, widest near the middle, with the margins convex nearly their full length. They are evidently almost ventricose in the fresh corolla, with the margins often being folded under when pressed. Those of G. cerastioides are cuneate to ” 362 spatulate-obovate, widest well above the middle. Below the widest point, they taper with straight or slightly concave margins to the base. They are not at all ventricose, and generally press flat. Differences in habit are of limited use in distinguishing these species. Gentianella cerastioides does exhibit a greater range in stem length than does G. rupicola, and the longer stems of G. cerastioides are usually erect, whereas even the longest stems of G. rupicola are decumbent. In both species, how- ever, specimens with very short stems (presumably from exposed habitats) are frequent. On the longer stems of G. cerastioides, the upper internodes are often 1—6 cm long, occasionally up to 8 cm, whereas even on the longest stems of G. rupicola the internodes seldom exceed 1 cm (Sparre 15694 {[S], with internodes up to 3 cm long, is an exception). On short- stemmed plants of G. cerastioides, all the leaves are closely spaced. Never- theless, even on acaulescent extremes of G. cerastioides, such as Mulroy 1097 (HAM) and Sparre 15869 (S), the narrow, cuneate-obovate, flaring, pale violet corolla lobes readily identify such specimens as being G. cerastioides; and long-stemmed extremes of G. rupicola, such as Sparre 15694, are like- wise readily identified by the broad, elliptic-obovate, incurved, red corolla lobes. In both species, the majority of the flowering stems bear but one flower; in G. cerastioides, stems bearing up to five flowers are not unusual, but in G. rupicola, even three-flowered stems are rare. Pedicel length is highly variable in both species, but ranges to a greater maximum in G. cerastioides. The altitudinal ranges of the two species overlap considerably, ca 3150-4500 m for G. cerastioides, and ca 3650 — 4600 m for G. rupicola. Gentianella cerastioides has the greater geographic range, from the Depart- amento de Narifio in southern Colombia to the Departamento de Azuay in southern Ecuador. Gentianella rupicola is known only from Ecuador, from Pichincha to Chimborazo. My count (Pringle 1981) of 2n = 18 for G. cerastioides was obtained from a specimen of G. cerastioides s. str., as delimited in the present study. HYBRIDIZATION IN ECUADOREAN GENTIANELLA Twenty-five species of Gentianella are native to Ecuador (Pringle, ms. for Flora of Ecuador). Most grow in paramo habitats ca 2800 to ca 4500 m altitude in the Andes, and several have large, open, blue-violet to rose- violet corollas that presumably attract similar pollinators. To date, no in- terspecific hybrids have been reported. Botanical exploration for the Flora of Ecuador with detailed locality data has for the first time clearly indicated the existence of hybridization in Gentianella in South America. Two 363 hybrids are described below. Descriptions of the parental species have been published by Fabris (1960) and will appear, with further details, in the Flora of Ecuador. GENTIANELLA CERASTIOIDES (H.B.K.) Fabris X G. FoLiosa (H.B.K.) Fabris. Figs. 7—11. Gentianella foliosa, like G. cerastioides (above), is widely distributed at high altitudes in Ecuador and is well represented in herbaria. Both species have corollas similar in size and color. A series of specimens at AAU, all from “Volcan Iliniza, NE slope below the refugio, lee side of loma with bunchgrass and shrubs, alt. 4300 m,” Prov. Pichincha, Ecuador, evidently represents hybridization between these two species. Holm-Nielsen et al. 24956 and 24971 are, respectively typical specimens of G. cerastivides and G. foliosa. Holm-Nielsen et al. 24957 (Fig. 10) is similar in habit to G. cerastioides, but has 5 flowers per stem, on stems ca 6 cm tall, more flowers than is usual on plants of G. cerastioides of that size, and has relatively short, stout pedicels like those of G. folzosa. The flowers are similar in aspect to those of G. fo/zosa, having the relatively abruptly rounded corolla lobes of that species, and the leaves are somewhat wider and more lanceolate than is usual in G. cerastioides. Intermediacy is even more evident in Holm-Nielsen et al, 249064 (Fig. 9), which ts similar in habit to 24957 but has stems 8 — 11 cm tall, bearing up to 6 flowers. The upper leaves are distinctly lanceolate, representative dimensions being 22 mm long, 5 mm wide, i.e., similar in shape to those of G. foliosa but in the size range characteristic of G. cerastioides. Some show a tendency toward arcuate spreading, as in G. foliosa. The corollas are similar to those of 24957. GENTIANELLA FOLIOSA X G. SULPHUREA (Gilg) Fabris. Figs. 11— 13, Table 2. Gentianella sulphurea has a much more restricted distribution than the other two species discussed here. Only 5 collections besides those mentioned below, all from the central part of the Ecuadorean Andes, have been encountered in my studies. Although G. su/phurea differs strikingly from G. foliosa in its leaf shape, corolla color, and lack of corolla trichomes, and usually also in its stature and pedicel length, the close relationship between these species was noted by Gilg (1896) and Fabris (1960). The two species are similar in habit, inflorescence type, and size and shape of the corolla, except that the corolla lobes of G. su/phurea are proportionately wider than those of G. foliosa. A series of 16 specimens at AAU, all from paramo habitats in the Cordil- lera de los Llanganates, from points 3 km SW to 13 km NW of Cerro Hermoso, Prov. Tungurahua, clearly represents a large hybrid swarm. In- cluded are specimens representative of G. folzosa and G. sulphurea, interme- diate specimens closer to each of the parental species, and specimens about — 11. Gentianella from Volcan Iiniza, Ecuador, all to same scale. Fi G. cerastioides, Ss 3 (S). Figs. 9— . ys cerastioides X G. foliosa, Holm-Nielsen et fa ee eee 24957 (AAU), respectively. Fig. 1 1. foliosa, Holm-Nuelsen et al. 24971 (A 365 midway between the parental species in morphology. Several intermediate specimens are compared with the parental species in Table 2. One of the most clearly intermediate plants, as indicated by its description in the ta- ble, is illustrated in Fig. 12. Descriptions of the parental species are based on all specimens examined for the Flora of Ecuador, exclusive of a few anomalous individuals assigned to G. folzosa. Collection numbers are those of Holm-Nielsen & Jaramillo. Detached flowers in packets were examined for internal corolla pubescence in two intermediate specimens, nos. 28168 and 28189, in both of which the corollas were yellow with red suffusions. The former bore tufts of trichomes below the corolla sinuses, more restricted and shorter than those of G. folzosa, but thus differed from G. sulphurea, in which corol- la trichomes are absent or occasionally few and minute. In 28189, the co- rolla trichomes were more numerous and longer, the plant in this respect being more like G. foliosa although otherwise bearing a greater resem- blance to G. salphurea. Hybridization may also account for some of the anomalous specimens obviously allied to G. foliosa that have been grouped as G. stellarioides (Griseb.) Fabris. Fabris (1960) described this “species” as “bastante polimorfa” even though he thus identified only 9 specimens representing perhaps as few as 5 collections. In particular, a specimen collected by Jameson on the “snowy summit of the Andes,” Ecuador (S), differs from typical G. foliosa in having much longer pedicels (2.5 — 9 cm), more deep- ly lobed corollas, and narrower, less abruptly tapering corolla lobes, suggesting that its origin might be G. foliosa X G. rapunculoides (Willd. ex Schultes) Pringle. Jameson’s labels, however, do not provide precise locali- ty data or otherwise indicate which specimens were found in proximity to one another. NOMENCLATURAL TRANSFERS Although most students of the Gentianaceae now accept generic status for Gentianella Moench, they have wisely refrained from making “automat- ic transfers” for all of the South Americn taxa accepted as Gentiana spp. by Gilg (1916) or Macbride (1959). The following species, however, appear from my own studies and from a monograph of the Argentine Gentianaceae by Fabris (1953) to be taxonomically acceptable, and the combinations published here are required for specimens sent to me for identification, for discussion in the present paper, or for use in works by other authors. Combinations herein attributed to “Fabris ex Pringle” were proposed by H. A. Fabris, as seen in sched. on specimens in AAU, E and WIS. HYBRIDS CHARACTER G. FOLIOSA 28676 28115 28120 28168 G. SULPHUREA (closer to G. (closer to G. FOLIOSA) SULPHUREA) length 3rd internode 15-40 mm 40 mm 22 mm 19 mm 9 mm 15 mm 3—10 mm below terminal in- florescence on longest stem length chis internode/ 0.6-1.5 1.1 0.77 0.79 0.34 0.56 0.1-0.7 length subtending leaves shape of lower leaves oblong-oblanceolate oblong elliptic-obovate elliptic elliptic elliptic-obovate —_ elliptic-obovate shape of upper leaves ovate-lanceolate —_ elliptic-oblong elliptic elliptic elliptic elliptic elliptic length largest leaf 20-80 mm 43 mm 30 mm 27 mm 23 mm 23 mm 8-23 mm length/width largest = 12 6.1 2.8 38 oie) 3:3 2-4 leaf number of flowers in 3-21 6-9 5-7 2-6 4-7 26 1—2 (-6) terminal inflorescences pedicel lengths 8—40 mm 14-21 mm 9-15 mm 15-21 mm 7-14 mm 10-17 mm 5—12 mm corolla color (from length/width repre- sentative corolla lobe rose-violet to blue-violet 1.6—2.4 violet light violet pink to red yellow with red veins and margins 1.4 yellow at base pale red above red above 1.4 yellowish-green to yellow, rarely with red suffusion L215 TaBLe 2. Comparison of selected specimens of Gentianella foliosa X G. sulphurea with the parental species. 99% c Q Q zt ca Q wu & o < x o — ka 12. Gentianella foliosa X G. sulphurea, Cordillera de los Llanganates, Ecuador, Holm-Nielsen & Joan 28115 (AAU). Fig. 13. Representative G. sulphurea, Atilio, Ecuador, Harling et al. 6663 HAM) 368 GENTIANELLA Cuspidata (Griseb.) Pringle, comb. nov. Gentiana cuspidata Griseb., Gen. Sp. Gent. 224. 1838 [“1839"], basionym, non Gentiana cuspidata Wallich ex C. B. Clarke, J. Linn. Soc. Bot. 14:45. 1875 According to Macbride (1959), Gentiana cuspidata Grisb. “Seems to be the same as G. multicaulis (G. Don) Gilg, non Griseb., i.e. G. Pavonii Griseb., and then the earlier name” {authors’ names added}. Examination of the respective type collections, however, indicates that G. cuspidata differs from G. pavonii in the greater size of the plants, more diffuse in- florescences, larger flowers, and widely opening corollas, and presents a very different general aspect. Type collection, G. cuspidata: PERU (not more precisely located): Dombey s.n., holotype: P, and isotypes: NY(2)! On G. pavonit, see Pringle (1981). GENTIANELLA dianthoides (H.B.K.) Fabris ex Pringle, comb. nov. Gentiana dianthoides H.B.K., Nov. Gen. Sp. 3:170 (quarto ed.). 1819, basionym. GENTIANELLA dielsiana (Gilg) Pringle, comb. nov. Gentiana dielsiana Gilg, Bot. Jahrb. Syst. 22:316. 1896, basionym. GENTIANELLA dolichopoda (Gilg) Pringle, comb. nov. Gentiana dolichopoda Gilg, Bot. Jahrb. Syst. 54(Beibl. 118):36. 1916, cava GENTIANELLA gilgiana (Reimers) Fabris ex Pringle, comb. nov. Gentiana gilgiana Reimers, Bot. Jahrb. Syst. 62:326. 1928 [" oon basionym GENTIANELLA helianthemoides (Gilg) Pringle, comb. nov. Gentiana helian- themoides Gilg, Bot. Jahrb. Syst. 22:321. 1896, basionym. This species was described from the Provincia de Salta, Argentina. A recent collection from the adjacent Departamento de Tarija, Bolivia, is the first for that country. Voucher specimen: BOLIVIA. Mendez: Tarija 25 kms hacia Camargo, 2765 m, 35° degrees E, Beck 843 (LPB) GENTIANELLA mendocina (Gilg) Pringle, comb. nov. Gentiana mendocina Gilg, Bot. Jahrb. Syst. 54(Beibl. 118):36. 1916, basionym. GENTIANELLA persquarrosa (Reimers) Pringle, comb. nov. Gentiana persqu- arroia Reimers, Bot. Jahrb. Syst. 62:332. 1929, basionym. GENTIANELLA radicata Sia Pringle, comb. nov. Gentiana radicata Griseb., Gen. Sp. Gent. p. 29. 1838 [“1839"], basionym. Geerianeixa riojae Gilg) Fabris ex Pringle, comb. nov. Gentiana rivjae Gilg, Bot. Jahrb. Syst. 22:319. 1896, basionym GENTIANELLA vaginalis (Griseb.) Pringle, comb. nov. Gentiana vaginalis Griseb., Gen. Sp. Gent. p. 215. 1838 {“1839"], basionym. REFERENCES FABRIS, H. A. 1953. Sinopsis preliminar de las Genciandceas argentinas. Bol. Soc. Argent. Bot. 4:232 — 259 369 —_______. 1960. El género Gentianella en Ecuador. Bol. Soc. Argent. Bot. 8:160— 192. GILG, E. rere Beitrage zur Kenntnis der Gentianaceae. I. Bot. Jahrb. Syst. oe Be eee _ Gentanacee andinae. Bot. Jahrb. Syst. 54(Beibl. 118):4— 122. MACBRIDE. 7 E 1959. Gentianaceae. Gentain family. J: Flora of Peru. Field Mus. Nat. : Bote oeh mac 2h BINGE fo Pesta bo: er enna transfers and taxonomic notes on some South Amer- ican Geacnae oe Phytologia 48:281 — A NEW COMBINATION IN CHIMAPHILA (ERICACEAE) LAURENCE J. DORR Missouri Botanical Garden, PO. Box 299 St. Louis, MO 63166-0299, U.S.A. Preparation of floristic treatments of the Pyroloideae (Ericaceae) that oc- cur in the West Indies, Mexico, and Central America necessitates a new combination in Chimaphila from Hispaniola. Collections of Chimaphila from Hispaniola were first reported by Urban (1908) as C. wmbellata (L.) W.P.G. Barton (cited as Chimophila umbellata Nutt.), a wide-ranging and variable species that is found throughout temperate Europe and Asia, Ja- pan, and North America from Canada to Central America. Blake (1914, 1917), emphasizing the smaller leaves, fewer teeth on the leaf margins, glabrous peduncles and pedicels, and glabrous filaments of the material from Hispaniola, described C. domingensis §.R Blake. Comparing this taxon with the species of Chimaphila that occur in North America, C. maculata (L.) Pursh, C. menziesii (R. Br. ex G. Don) Sprengel, and C. umbel- lata, it is evident that C. domingensis is closely related to C. wmbellata. Both taxa have oblanceolate or spatulate to elliptic leaves, teeth confined to the upper margins of the leaves, cuneate leaf bases, linear-subulate bracts, and 2 — 6-flowered, corymbose inflorescences. The relatively minor size and pubescence differences of Chimaphila in Hispaniola could easily have arisen as a consequence of isolation. In order to reflect what is assumed to be the shared evolutionary origin of C. domingensis and C. umbellata the following new combination is proposed. CHIMAPHILA UMBELLATA (L.) W.P.G. Barton subsp. domingensis (S.E Blake) Dorr, comb. et stat. nov. Chimaphila domingensis S.E Blake, J. Bot. 52:169. 1914. Type: DOMINICAN REPUBLIC. Prope Constanza in Valle Nuevo, 2200 m alt., in pineto, Aug 1910, H. von Turckheim 3434 (HOLOTYPE: BM!, GH- photograph!; tsoryvpes: Bt, G-2 sheets!, GH!, K!, MO!, NY!, US-microfiche!). ACKNOWLEDGMENT Research was supported, in part, by grants from Sigma Xi and the Na- tional Science Foundation (BSR 85057 10) REFERENCES BLAKE, S.F 1914. A new Chimapbila from San Domingo. J. Bot. 52:169. ——_——. 1917. The varieties of Chimaphila ve Rhodora 19:237 — 244. URBAN, I. 1908. Symbolae Antillanae 5(3):45° SIDA 11(4):370, 1986. RENOVATION OF DYSSODIA (COMPOSITAE: TAGETEAE) JOHN L. STROTHER Botany, University of California Berkeley, CA 94720, U.S.A. ABSTRACT Reconsideration of circumscription of Dyssodia s.1. has led to resurrection of Adenophyl- lum, Boeberastrum, Comaclinium, Dysodtopsis, and Thymophylla and to change to generic rank for one section: Boeberoides (DC.) Strother with one species, B. Beedifices New combinations a Adenophyllum are: A. anomalum, A. appendiculatum, A. aurantium, A. cooperi, A. glandulosum, A. aun aa A. ee var. cancellatum, A. porophyllum var. radiatum, A. speciosum, A. mosum, and A. wrightii var. pulcherrimum; in Comaclinium: C. montanum; and in nT ymophylle T. acerosa, T. aurea var. polychaeta, T: concinna, T. gentryi, T. gypsophila, T. micropoides, T. mutica, T. pentachaeta var. belenidium, T. pentachaeta var. hartwegii, T. pentachaeta var. puberula, T: setifolia var. radiata, T. tenuiloba var. texana, T. tenuiloba var. treculii, T. tenuiloba var. wrightii, and T. tephroleuca. In a review of Tageteae (Strother 1977), I acknowledged that my earlier (Strother 1969) circumscription of Dyssodia was too broadly drawn and that the subgenera and some sections would be better treated as distinct genera, because some of the subdivisions are more closely allied with other Tageteae than with each other. Those conclusions were drawn from integra- tion of new and re-evaluated, old observations of morphology and chromo- some numbers both within Dyssodia s.1. and across Tageteae. The desirabil- ity of such a reinterpretation of Dyssodia (sensu Strother 1969) has been indicated, at least implicitly, by Robinson (1981) and Downum et al. (1985) The resulting realignments require several nomenclatural changes, which are set forth here. The format is intended to coordinate with my 1969 treatment and to account for the one name added to Dyssodia in the interval. Circumscriptions of species and varieties are unchanged except that of D. decipiens, which is expanded to include D. sanguinea. In the fol- lowing synopses of genera and in the nomenclator for Dyssodia, accepted names are in capitals (for previously published) or boldface (new combina- tions); for all, bastonyms are indicated. Review of revised phyletic postulates for relationships among genera of Tageteae is beyond the scope of this paper. Nevertheless, some of the more SIDA J (4) 37 l=-378. 1986, 372 salient differences that form the bases for recognition of the retained, resurrected, and newly recognized genera may be summarized in the fol- lowing key: a. alenieas free quite to base or nearly s¢ b. Erect annuals and perennial ee leaves not fleshy; calyculus of 3- oe = ep ee c. Leaves pinnatifid with 3 — 17 linear to lance-cuneate or oblanceolate lobes, glands scattered in lamina and/or submarginal; receptacles fimbrillate or finely setose; lobes of disc corollas ane eee to ETH a ere (ol is |e caver eee ae ae PC ae nea ether er ener redo a eral et eee ee Dyssodta c. Leaves ae or crifoliolare, glands mostly in rows between lateral erves; receptacles minutely paleate; lobes of disc corollas lance-linear Se -MUEM He EGR Mars cate ata Ace Bake a senha Comaclinium to subulate. b. Sprawling or decumbent annuals; leaves fleshy; calyculus none ora single PROIEOIGS So aan eee eee eae ee eee a ee a. Phyllaries weakly to strongly conn d. Leaves mostly pinnate or witinatlfe if simple, then lanceolate, linear, or iliform; calyculus of deltate, linear, subulate, or pectinate bracteoles; Boeberastrum some or all phyllaries gland-bearing. . Plants less than 3 dm high; leaves linear-filiform or pinnatifid with linear filiform lobes; phyllaries strongly connate 2/3 + their length, margins of the outer seldom free more than 1/2 their length; recep- acles nigked Or titty SO. 2a sue tibia in ccabeebucee kee ess Thymophyla e. Plants (2—)4— 20+ - high; leaves or lobes linear or broader; phyllaries weakly connate 1/3 — 2/3 their length (somewhat more in Dysodiopsis), often separating in age, margins of the outer usually free to base; receptacles fimbrillate or setose f. Leaves mostly pinnate or singauitid: if simple, then lanceolate, glands marginal and subterminal; lobes of disc corollas lance- linear to subulate; style-branch tips papillate and conic or his- PidclOUS tad sublilate. cu kee seed este eews ce eee Adenophyllum f. Leaves linear, glands scattered along either side of midvein; lobes of disc corollas lance-deltate; style-branch bs papillate, abruptly truncate and with a fine, setaceous appendage. ............. Dysodtopsis d. Leaves simple, oblong to lance-ovate; calyculus of igadly lanceolate bracteotes! pavllarics Slandless. «cogs c ceasing dy bebibede cine Boeberotdes SYNOPSES OF GENERA ADENOPHYLLUM Pers., Synop. Pl. 458. dee =Dyssodia Cav. sect. oe vas /um (Pers.) O. Hoffm. in Engl. & Prantl, Naturl. Pflanz. 4(5):266. 1894.—Ty Adenophyllum coccineum Pers. =Willdenowa pane Cav. =ADENOPHYLLUM GLANDULOSUM (Cay.) Strother. Willdenowa Cav., Icon. Pl. 1:61, t. 89. 1791, non Willdenowia Thunb. 1788, nec Willdenowa eee 1790. a PE: nie a glandulosa Cav. =ADENOPHYLLUM YULOSUM (Cavy.) Strother. Schlechtendalia Willd., Sp. PI. 3) 2125. 1804, nom. rej. vs. Schlechtendalia Less., —Typr: Wilden a glandulosa Cav. = ADENOPHYLLUM GLANDULOSUM (Cav. ) Strother co ee) ne —_) os) Clomenocoma Cass., Dict. Sci. Nat. 9:416. 1817. =Dyssodia Cav. subg. Clomenocoma (Cass.) Stro ‘hee Gniy: Ge Publ. Bor. 48:37. 1969. eee Cav. sect. Clomeno- coma (Cass.) Strother, Univ. Calif. Publ. Bor. 48:49. ee —TypE: Aster aurantins L. =C/lomenocoma aurantia (L.) Cas ADE URANTIUM (L.) Strother. Lebetina Cass., Dict. Sci. Nat. 25: ne 1822. Dyna ae sect. Lebetina (Cass.) O. Hoffm. in Engl, & ae Naturl. Pflanz. 4(5):266. 1894.—Type: Lebetina cancel- mie Cass. =ADENOPHYLLUM POROPHYLLUM (Cav.) Hemsley var. CANCELLATUM Strother. Pada Aen N. Amer. Fl. 34:170. 1915.—Type: Trichaetolepis wrightii (A. Rydb. =ADENOPHYLLUM WRIGHTH A. Gray. Annual or perennial herbs and shrubs, most robust or coarse, (2—) 10 — 20+ dm high; leaves opposite or alternate, pinnate with lanceolate or ovate leaflets or pinnatifid with obovate, linear, or filiform lobes, bases, rachises, and teeth usually setose or bristly, glands marginal, often associ- ated with bases of lobes, and subterminal in lobe tips; calyculus of 12—16(1— 22, sometimes wanting in A. anomalum), subulate, lance- linear, or pectinate, often seta-tipped bracteoles, often equalling or surpassing the phyllaries; involucre 8 — 20(3.5 — 25) mm high; phyllaries 8 — 20(-30), weakly connate 1/3 — 2/3 their length, often separating with age, margins of the outer usually free to base or nearly so, glands round to elliptic or elongate; receptacles fimbrillate or setose; ray corollas golden yellow, orange, or scarlet; lobes of disc corollas lance-linear to subulate; style-branch tips papillate and conic or hispidulous and subulate; achenes stoutly to narrowly obpyramidal; pappus of (8—)15—20 squamellae, variously muticous, aristate, or dissected into 4— 11 bristles; x = 7, BoeBerastRuM (A. Gray) Rydb., N. pas Fl. 34:161. 1916. mee Cav. sect. Boeberastrum A. Gray, Proc. Amer. Acad. Arts 19:39. 1883.— Dyssodia anthemidifolia’ Benth. =BorBERASTRUM ANTHEMIDIFOLIUM Fo Rydb Sprawling annuals with stems to 3 dm long; leaves at first opposite, soon alternate, the blades fleshy, spatulate or pinnatifid with linear lobes, not setaceous at base, on lobes, or at tips, glands submarginal or scattered in lamina; calyculus none or a single bracteole; involucres 5-8 mm high; phyllaries 8, free to base, glands mostly round; receptacles fimbrillate; ray corollas bright yellow; lobes of disc corollas narrowly lance-linear to subulate; style-branch tips papillate, rounded-truncate; achenes stoutly obconical or obscurely obpyramidal; pappus of 16— 20 squamellae, each scale dissected into 5— 10 bristles; x = Boeberoides On ) Strother, stat. nov =Dyssodia Cav. sect. Boeberoides DC., rodr. 53:64( 1836.—Typr: Dycendis grandiflora DC. =BorBEROIDEs GRANDIFLORA a Strother. 374 Coarse herbs (annual?) to 25 dm high; leaves all opposite or subopposite, blades oblong to lance-ovate, often with 2—6 inconspicuous subulate lobules at base, glands scattered in lamina; calyculus of 10 — 20 broadly lanceolate bractlets; involucres ca 20 mm high; phyllaries 18 — 20, strong- ly connate ca 2/3+ their length, glandless; receptacles foveolate; ray corollas yellow-orange; lobes of disc corollas lance-linear; style-branch tips hispidulous, long-subulate; achenes obpyramidal; oe of ca 20 squamellae, each scale dissected into 7 — 10 bristles; x = CoMACLINIUM Scheidw. & Planchon, Fl. Ser. Jard. l'Europe 8:19, t. 756. 1852.—Tver: Comaclinium aurantiacum Scheidw. & Planchon. =COMACLINIUM MONTANUM (Benth.) Strother. Perennial herbs to 10 dm high; leaves opposite, becoming alternate, the blades simple (—trifoliolate), lanceolate, usually with 1—3 pairs of subulate-setose lobules at base, glands mostly in rows between lateral nerves; calyculus of 3 — 12 linear bracteoles; involucres 12 — 15 mm high; phyllaries 10 — 16, free to base or nearly so, streaked with linear to elliptic glands; receptacles minutely paleate; ray corollas orange; lobes of disc corollas lance-linear to subulate; style-branch tips papillate, conic; achenes stoutly obpyramidal; pappus of ca 20 squamellae, the outer shorter, all dissected into 5— 10 bristles; x = a Gray) Rydb., N. Amer. Fl. 34:171. 1915. el sete ct. Dysodtopsis A. Gray, Smithsonian Contr. Knowl. 3(5):116 (Pl. Wright I). 852. —TYPE: a tagetoides Torrey & A. Gray. =Dysopiopsis TAGETOIDES eae & A. Gray) Ryc Annuals or short-lived perennials, 4-— 8+ dm high; leaves opposite at base, alternate above, blades linear, coarsely toothed, obscurely setaceous at base, glands scattered along either side of midvein; calyculus of 5 — conspicuous, subulate or pinnatisect bracteoles nearly as long as eee involucre 9 — 12 mm high; phyllaries 10 — 12, strongly connate but with outer margins free to base, glands round to elliptic; receptacles minutely fimbrillate; ray corollas lemony to greenish yellow; lobes of disc corollas lance-deltate (not short-deltate); style-branch tips papillate, truncate, and with a fine, fragile, setaceous appendage; achenes narrowly eae, pappus 10 — 12 unequal, lanceolate scales ending in 1(—3) aristae; x = Dyssopia Cav., Descr. Pl. 202. 1802.—Tver: Taceres papposa Vent. = Dyssovia PAPPOSA (Vent.) Hitche Boebera Willd., Sp. Pl. 2125. 1804.—Tyee: Boebera chrysanthemoides Willd. =Tagetes happosa Vent. =Dyssopia papposa (Vent.) Hitche Rosilla Less., Synop. a Comp. 245. 1832.—Type: ‘Rel lutea Less. = Dyssovia NNATA (Cav.) Robinson. a1) Syncephalantha Bartling, Ind. Sem. Hort. Goett. 6. 1836 [ex Linnaea 12:80. 1838}. =Dyssodia Cav. sect. Syncephalantha (Bartling) Strother—Tyvre: Syncephalantha decipiens Bartling. =Dyssopia DeEcIPIENS (Bartling) M. Johnston in M. Johnston & B. Turner. Annuals or perennial herbs 1— 3(—9) dm high, leaves opposite, often becoming alternate distally, the blades pinnatisect with linear to linear- cuneate or oblanceolate lobes, little, if at all, setaceous at base, on teeth, or at tips of lobes, glands scattered in lamina and/or submarginal; calyculus of 3 —5(1 —9) deltate to linear bracteoles 1/2 — | times as long as phyllaries; involucres 5 —8(—-10) mm high; phyllaries 4—8(—-16), free to base or nearly so, glands round to elliptic; receptacles fimbrillate to finely setose; ray corollas yellow to yellow-orange; lobes of disc corollas short, deltate to lance-deltate; style-branch tips papillate, rounded-truncate to shortly delrate; achenes stoutly obpyramidal to obconic; pappus of 15 — 20 un- equal to subequal squamellae, each scale dissected into 5 — 10 bristles; x = 1s THyYMopPHYLLA Lagasca, Gen. Sp. Nov. 25. 1816. =Dyssodia Cav. sect ee ee O. Hoffm. in Engl. & Prantl, Nat. Pflanz. 4(5):266. > THYMOPHYLLA SETIFOLIA Lagasca. eee Cass., Bull. Soc. Philom. Paris 1818:183. 1818. =Dyssodia Cav. subg. ee (Cass.) Strother, Univ. Calif. Publ. Bot. 48:57. 1969. =Dyssodta ne ae Cass.) Strother, Univ. Calif. Publ. Bot. 48:78. 1969._"Tv PE: Hymenatherum tenuifolium Cass. =THYMOPHYLLA = TENUIFOLIA (Cass.) Rydb. Dyssodia Cav. sect. Aciphyllaea DC. , ae - 641. 1836. =Aciphyllaea (DC.) A. Gray, em. Amer. Acad. Arts, ser. 2. 4:9. 1849. en: Cass. sect. Aciphyllaea (DC.) A. Gray, Smithsonian nee Knowl. 3(5):115 (Pl. Wright. 1). 1852.—Tyee: Distulec acerosa DC. =THYMOPHYLLA ACEROSA (DC.) Strother. G Ce DC., Prodr. 7:258. 1838. ee Sane: sect. eee (DC.) Be eigonisa Contr. Knowl. 3(5): 116 (PI. _ 1). 18 Dyssodia - sect. Gnaphalopsis (DC.) Strother, ae a Publ. ao 48:57. 1969.—Tyer: Gnaphalopsis micropoides DC. =THyMopHyL_a MicRopoIDES (DC.) —~ y, Strother. ace A. Gray, Mem. Amer. Acad. Arts, ser. 2. 4:89. 1849.—Tyre: Lowellia aurea . Gray. = THYMOPHYLLA AUREA (A. Gray) E. Greene in Britton & A. Brown. rene Cass. sect. Heterochromea A. Gray, Synop. Fl. N. Amer. 1(2):453. 1884.—Type: Hymenatherum concinnum A, Gray. =THYMOPHYLLA CONCINNA (A. Gray) Strother. Dyssodia Cav. sect. Aurantiacae Strother, Univ. Calif. Publ. Bot. 48:64. 1969.—Tyre: Hymenatherum aurantiacum Brandegee. =THYMOPHYLLA AURANTIACA (Brandegee) b Annual or perennial herbs or shrublets less than 3 dm high; leaves oppo- site or alternate, the blades linear-filiform (narrowly spatulate in T. mzcropo- 576 ides) or pinnatisect with linear-filiform lobes, little, if at all, setaceous at base, on teeth, or at tips of lobes, glands scattered along rachis and lobes (submarginal and scattered in lamina of 7. micropoides); calyculus of | — 3(0 — 8), deltate to subulate bracteoles, usually much shorter than the phyllaries; involucres 3—7 mm high; phyllaries 8 — 13(—22), strongly connate 2/3 + their length, seldom with outer margins free more than 1/2 their length, glands mostly round; receptacles naked or nearly so; ray corollas golden to pale yellow (white in T. concinnum); lobes of disc corollas short, delcate to lance-deltate; style-branch tips papillate, delrate or conic; achenes stoutly to narrowly obconic, obpyramidal, or cylindric; pappus of 10(—20) squamellae, variously muticous, aristate, or dissected into 5 — 7 bristles; x = 8 {reports of 7 = ca 13, 13, ca 26 in species of Thymophylla (as Dyssodia, e.g., in Strother, 1969) probably represent misinterpretations of triploid (27 = 24) and hexaploid (27 = 48) individuals]. NOMENCLATOR FOR DYSSODIA S.L. —~ J. acerosa DC., Prodr. 5:641. 1836. = Thymophylla acerosa (DC.) Strother, comb. nov. ). anomala (Guuby & Rose) ee =Hymenatherum anomalum Canby & Rose, Contr. Natl. Herb. 1:105. 1891. =Adenophyllum anomalum (Canby & Rose) St trocher, comb. nov. . anthemidifolia Benth., Bot. Voy. Sulphur 29. 1844. =BorBerasrRUM ANTHEMIDI- FOLIUM (Benth. ) pa N. Amer. Fl. 34:162. 1915 . appendiculata Lagasca, G Gen. Sp. Nov. 28. 1816. =Adenophyllum appendiculatum Lagasca) Strother, ab nov. . auvantia Ass ) Robinson. =Aster aurantius L., Sp. Pl. 877. 1753. =Adenophyllum tium (L.) Strother, comb. nov . auvanttaca bases Robinso ee ee aurantiacum Brandegee, Zoe 5:258. 1908. =THYMOPHYLLA AURANTIACA (Brandegee) Rydb., N. Amer. Fl. 34:175. 1915. D, Bale Gray) Nelson. =Lowellia aurea K. Gray, Mem. Amer. Acad. Arts, ser. 2. 4:91. 49, =THYMOPHYLLA AUREA (A. Gray) E. Greene in Britton & A. Brown, IIlusc. . 3:435. 1898. aurea a Gray) Nelson var. polychaeta (A. Gray) M. Johnston. =Hymenatherum polychae- tum A. Gray, Smithsonian Contr. Knowl. 3(5):116 (Pl. Wright. 1). 1852 =Thymophylla aurea (A. Gray) E. Greene var. polychaeta (A. Gray) Strother, =~ ou ~ VY ~— YS ~— YY u comb. nov. . concinna (A. Gray) Robinson. =Hymenatherum concinnum A. Gray, Synop. Fl. N. Amer. (2):453. 1886. =Thymophylla concinna (A. Gray) Strother, comb. nov. . cooper? A. Gray, Proc. Amer. Acad. Arts 9:201. 1874. =Adenophyllum cooperi (A. Gray ences oo nov D. becipieNs (Bartlin iM, Johnstons in M. Johnston & B. Turner, See 64: 1962. =Sypalane decipiens Bartling, Index Sem. Hort. Goett. 6. ae ee — 38}. So 0 ~ a ‘ Gini ae. fice Southw. Naturalist 3:219. 1959. =Thymophylla gencryi (M. Johnston) Strother, comb. nov. D. glandulosa (Cav.) O. Hoffm. in Engl. & Prantl, 1894 (non Dyssodia glandulosa Cav., 271 1802). =Willdenowa glandulosa Cav., Icon. Pl. 1:61, t. 89. 1791. =Adenophyl- lum glandulosum (Cav.) Strother, eos, nov dD, canine DC. "Prod. 5:640. 1836. =Hechecoides grandiflora (DC.) Strother, Dz wll B. te Madrono 21:421. 1972. =Thymophylla gypsophila (B. Turner) Strother, comb. n D. Iittoralis Benders, Zoe 5:163. 1903. =BorserastRUM LITTORALIS (Brandegee) ydb., N er. Fl.34:162. 1915. D. ae (DC) Loes. =G Rie nl DC., Prodr. 7:258. 1838. mophylla micropoides (DC.) Strother, comb. D. montana Benth ) A. Gray. ene montana ae ms Hartweg. 86. 1841. = aclinitum montanum (Benth.) Strother, comb. D. mutica ne ee Southw. Naturalist 5:225. 1960. aa opr mutica (M. ee comb. nov. _ neomexicana (A. Gray) Robinson. = ‘ymenatherum neomexicanum A. Gray. =ADENOPHY- LLUM WRIGHTH A. Gray, Smithsonian Contr. Knowl. 5(6):92. (PI. Wright. II). 1853, non Dyssodia pneu wrightit (A. Gray) Robinson, 191 So D. neomexicana (A. Gray) Robinson var. Laan Strother, Univ. Calif. Publ. Bot. 48:43. 69. =Adenophyllum wrightit A. Gray var. pulcherrimum ai ae comb. nov D. PAPPOSA (Vent ) A. Hitche., Trans. Acad. Sci. St. Louis 5:503. 1891. =Tagetes papposa Ve scr. Pl. Nouv. Cels. 4th livr. 1801. D. penatbact ae ) Roe ison: =H pe ine acneai DC., Prodr. 5:642. 1836. YMOPHYLLA PENTACHAETA (DC.) Sm Fl. Souther U.S. 1295 ; paca (DC.) Robinson var. saree ce Strother. =Hymenatherum eee DC rodr, 7:292. 1838. oe pentachaeta (DC.) Small var. once (DC.) Strother, comb. Dz pact (DC.) Sa vat. ee (A. Gray) Strother. =H ymenatherum hartwegii . Gray, Smithso Contr. Knowl. 3(5):117 (PI. 852. iy mophyl ee (DC.) Small var. hartwegii (A. Gray) Strother, ~ mb. nov. D. ie (OC) Robinson var. puberula (Rydb.) Strother. =Thymophylla puberula db., N. Amer. Fl. 34:177. 1915. =Thymophylla pentachaeta (DC.) Small var. puberula (Rydb.) Strother, comb. nov. D. PINNATA ee aie Proc. Amer. Acad. Arts 49:501. 1913. =Aster pinnatus Cav., Icon. Pl. 3:6. 1794. D. pINNATA (Cav.) Robinson var. GLABRESCENS Strother, Univ. Calif. Publ. Bot. 48:33. 1969. Dy. Lene (sic) (Cav.) Cav. =Pteronta porophyllum Cay., Team: Plies: 1pts 229.1794: DENOPHYLLUM POROPHYLLUM (Cay.) Hemsley, Biol. Cen. Amer. Bot. 2:218. D. Re (Cav.) Cav. var. cancellata (Cass.) Strother. =Lebetina cancellata Cass., Dict. Sci. Nat. 25:395. 1822. =Ac ee oa porophyllum (Cav.) Hemsley var. cancellatum (Cass.) Strother, comb. D, ee (Cav.) Cav. var. radiata DC. co 5:639. 1836. aaa rophyllum cane ) Hemsley var. eacliatunn (DC.) Strother, comb. D. prpilinds Gray, Mem. Amer. Acad. Arts, ser. 2. 5:322. 1854. Seta pile rophylloides i Gray) Strother, comb. D, ees (Klatt) Strother. =Syncephalanthus sanguineus Klatt, Leopoldina 25:106. 1889. 378 = Dyssopia DECIPIENS (Bartling) M. Johnston in M. Johnston & B. Tur D. sai Gaagasca Robinson. =THYMOPHYLLA SETIFOLIA Lagasca, Gen. Sp. Nov 25. dD. wae a Robinson var. radiata (A.Gray) Strother. =Thymophylla greggii A. os var. radiata A. Gray, Smithsonian Contr. Knowl. 3(5):119 (Pl. Wright. 1). ey setifolia Lagasca var. radiata (A. Gray) Strother, comb. nov. D. speciosa A. Gray, Proc. Amer. Acad. Arts 5:163. 1861. =Adenophyllum speciosum (A. Gra hae comb. D. squamosa A, Proc. re mer. Acad. Arts 19:38. 1883. =Adenophyllum squamosum a ray) Strother, comb. TAGETIFLORA Lagasca, Gen. Sp. Nov. 29. 1816. D. tagetoides Torrey & A. Gray, Fl. N. Amer. 2:361. 1842. =Dysopiopsis TAGETOIDES (Torrey & A. Gray) Rydb., N. Amer. Fl. 34:171. 1915. D. — (Cass.) Loes. =Hymenatherum tenuifolium Cass., Bull. Soc. ere 1818:183. Rydb. =THYMOPHYLLA TENUIFOLIA (Cass.) Rydb., N. Amer. FI. :173. 1915. D. ee (DC.) Robinson. =Hymenatherum temilban DC., Pre ae . 462. 1836. =THYMOPHYLLA TENUILOBA (DC.) Small, Fl. Southea: US. 1 295. 1903. D. tenuiloba (DC.) Robinson var. texana (Cory) Strother. = Dyssodia texana Cory, Rhodora ee 162. ae =Thymophylla tenuiloba (DC. ) Small var. texana (Cory) Strother, mb. D. bane (De. ) Robinson var. trecu/it (A. Gray) Strother. =Hymenatherum treculii A. Gray, Proc. Amer. es Arts 19:42. 1883. =Thymophylla tenuiloba (DC.) Il var. treculit (A ay) Strother, comb. nov D. tenuiloba (DC.) Robie: var. es (A. Gray) eurnee =Hymenatherum wrightii A. Gray, Mem. Amer. Acad. Arts, ser. 2. 4:89. 1849. ee tenuiloba (DC.) Small var. wrightii Se ee Strother, comb. D. ihe S. E Blake, J. Wash. Acad. Sci. 25:320._ 1935. =Thymophylla tephroleuca (S. E Blake) pene comb. ACKNOWLEDGMENTS I thank D. Keil and A. Smith for helpful discussions. REFERENCES DOWNUM, K. R., D. J. KEIL, and E. RODRIGUEZ. 1985. Distribution of acetylenic Seacoast in che Pectidinae. Biochem. Syst. Evol. 13:109— 113. ROBINSON, 1981. A revision of the tribal ae subtribal limits of the Heliantheae TOR Sui heonian Contr. : STROTHER, J. L. 1969. LoS en eae (Compositae: Tageteae). Univ. Calif. Publ. Bot. 48:1— 88. . 1977 {1978}. Tageteae — systematic review. Pp. 769 — 783 in The biology and chemistry of the Compositae, eds. V. H. Heywood et al. London: Academic Press. NOTEWORTHY PLANTS FROM NORTH FLORIDA. I. LORAN C. ANDERSON Department of Biological Science, Florida State University Tallahassee, FL 32306-2043, U.S.A ABSTRACT The following appear to be first reports for the state of Florida: Agrostis elliottiana, Aster hemisphericus, Plantago rugelit, Ratibida columnifera, Salix eriocephala, Stachys tenuifolia, Veronica ae ee Vicia hugeri. Over 30 additions to the flora of the Florida panhandle are documented h This ts the second installment of an anticipated series (Anderson 1984) to update our knowledge of the flora of the Florida panhandle and Clewell’s guide (1985) to the flora. The area of coverage is from Madison, Lafayette, and Dixie counties west to the Alabama state line. New discoveries and significant range extensions are given; a few collections appear to be first reports for the entire state. Exotics that appear to be adventive or natural- ized are also listed. Voucher specimens are at FSU unless noted otherwise. Some adjustments or clarifications in nomenclature are reported to re- flect current understanding. Additional nomenclatural changes that affect our flora are recorded by Wunderlin et al. (1985) AGROSTIS ELLIOTTIANA Schult. Gadsden Co.: Chattahoochee, A. K. Gholson 10845 (FLAS), R. K. Godfrey 81222; native, new to Florida. ALSTROEMERIA PSITTACINA Lehm. Gadsden Co.: large colony natural- ized over 10 years (A. K. Gholson, pers. comm.), Chattahoochee, L. Anderson 8211; new to Florida panhandle. ALYSICARPUS OVALIFOLIUS (Schum. & Thonn.) J. Leonard. This is the common species in the southeastern United States rather than A. vaginalis (L.) DC. The latter ts found in Florida in Dade and Monroe Counties and 1s the common species in the Caribbean region. Our species can be dis- tinguished by its lax inflorescences and larger leaves; A. vaginalis has dense short spikes and smaller leaves (R. Wunderlin, pers. comm.). AMARANTHUs Livibus L. Franklin Co.: Apalachicola, L. C. Anderson 7359; new to Florida panhandle. ARISTIDA RAMOSISSIMA Engelm. ex Gray var. CHASEANA Henr. Bay Co.: Westbay, R. K. Godfrey 73951 (det. K. Allred); new to Florida panhandle. ASCLEPIAS VIRIDULA Chapm. Washington Co.: Rock Hill, 3.5 air mi SE SIDA 11(4):379 — 384. 1986. 380 of Chipley, L. C. Anderson 8464; range extension to the northwest for this Florida endemic. ASTER HEMISPHERICUS Alex. Washington Co.: Rock Hill, 3.5 air mi SE of Chipley, R. K. Godfrey 81896, R. Kral 3521 (det. J. Semple); native, new to Florida. Clewell (1985) said “reportedly from the western panhan- dle.” These collections confirm its presence in the state. AsTeR LONGIFOLIUS Lam. Jefferson Co.: 5.3 mi S of Wacissa, R. K. Godfrey 80175; Wakulla Co.: vic. Newport, R. K. Godfrey 80177 (det. A. G. Jones); new to Florida panhandle. ASTER SIMMONDs!II Small. Franklin Co.: Bloody Bluff, 10 air mi N of Apalachicola, L. C. Anderson 7776, 3 air mi SW of Sumatra, L. C. Anderson 7816, 7817; Wakula Co.: McBride Slough, L. C. Anderson 7765, St. Marks, R. K. Godfrey 67760 (det. A. G. Jones); new to Florida panhandle. CAMPANULA FLORIDANA S. Wats. ex Gray. Franklin Co.: 5 air mi S of Sopchoppy, L. C. Anderson 8250. This is a range extension westward from Taylor County. CERATOPHYLLUM MURICATUM Cham. Franklin Co.: Cape St. George Island, L. C. Anderson 9064. The species was listed from neighboring St. Vincent Island by Chapman (1897); he listed it tentatively (under its synonym) as follows: "Ceratophyllum echinatum Gray?” . Clewell (1985) did not list the species under either name. CRATAEGUS PHAENOPYRUM (L. f.) Medic. Washington Co.: 5.5 air mi SW of Chipley, L. C. Anderson 8200; R. K. Godfrey 81202. Coker and Totten (1934) reported the species (as C. youngii Sarg.) from Wakulla County, but the Wakulla station has not been rediscovered, and the taxon is not listed by Clewell (1985). ELEOCHARIS TORTILIS (Link) Schultes in R. & S. Franklin Co.: Shell Hammock, 5.5 air mi SW of Panacea, L. C. Anderson 8275; Gulf Co.: M & K tract, 8.5 air mi NW of Apalachicola, L. C. Anderson 8833. Clewell (1985) reported this species was known in the panhandle only from one specimen collected by Chapman in Gadsen County. HyprocoryLe sipTHORPIOIDES Lam. Jackson Co.: Neal’s Landing, Lake Seminole, L. C. Anderson 8449, A. K. Gholson 11290; new to Florida panhandle. IMPATIENS BALSAMINA L. Bay Co.: Callaway, W. S. Judd 2307 (FLAS); Leon Co.: Lake Ella, Tallahassee, L. C. Anderson 8082; adventive, new to Florida panhandle. IsOTRIA VERTICILLATA (Muhl. ex Willd.) Raf. Washington Co.: ca. 6 air mi SE of Vernon, L. C. Anderson 8194. Correll (1950) reported this orchid from Gadsden County, but its occurrence there has not been redis- covered. 381 LINDERNIA GRANDIFLORA Nutt. Franklin Co.: 4.7 air mi S of Sopchoppy, L. C. Anderson 6995, 7105; Jefferson Co.: E of Newport, 2 mi W of Aucilla River, R. K. Godfrey 81768; new to Florida panhandle. LupwiGIA CuRTissii Chapm. Franklin Co.: St George Island, R. K. Godfrey 71148 (det. Ching I-Peng); new to Florida panhandle. Lupwicia ERECTA (L.) Hara. Franklin Co.: sandbar at milepost 7.1 on Apalachicola River, L. C. Anderson 8528, 8677; Gulf Co,: Brothers River, 13 air mi N of Apalachicola, L. C. Anderson 8932; new to Florida panhan- dle. Lupwicia LANCEOLATA Ell. Jefferson Co.: Lloyd, R. K. Godfrey 61411; Madison Co.: 8 mi W of Greenville, R. Kra/ 3741; Taylor Co.: 1.2 mi SE of Salem, R. K. Godfrey 64738, Wakulla Co.: NW of Crawfordville, R. K. Godfrey 64876 (det. Ching I-Peng); new to Florida panhandle. This species is listed under L. alata Ell. by Clewell (1985). Both species occur in our area; L. lanceolata has sepals about half as long as the capsule and pollen shed in tetrads, whereas L. a/ata has sepals nearly as long as the capsule and pollen shed singly. Manisuris CYLINDRICA (Michx.) Kuntze. Jackson Co.: Apalachee Game Management Area, N of Sneeds, J. B. Nelson & G. R. Knight 2615; new to Florida panhandle. MARSILEA UNCINATA A. Br. Franklin Co.: N. Market Street, Apalachi- cola, L. C. Anderson 7625, 8009, A. K. Gholson 11212. A second popula- tion of Marsilea from Avenue K in Apalachicola (L. C. Anderson 7356) has not produced sporocarps during the past year, so its identity as M. vestita or M. uncinata has not been resolved. A collection of M. vestita, so annotated by D. S. Correll, from Apalachi- cola (Chapman in 1860) was cited by Ward and Hall (1976). They suggested the label data may be in error because Chapman (1897) did not list the species for Florida. Johnson (1986) has determined M. wncinata to be synonymous with M. vestita ssp. vestita, so all Franklin Country collec- tions are M. vestita. MENTHA PIPERITA L. Franklin Co.: Apalachicola, L. C. Anderson 7357, 7535; Wakulla Co.: St. Marks, R. K. Godfrey 64407. MENTHA ROTUNDIFOLIA (L.) Juds. Franklin Co.: Avenue E Apalachi- cola, L. C. Anderson 7352, Avenue M, Apalachicola, L. C. Anderson 8312; naturalized, new to Florida panhandle. OENOTHERA NuTANS Atkins. Jackson Co.: Caverns State Park, R. S. Mitchell 824; Leon Co.: Tallahassee, W. D. D’Arcy 1299 (FLAS); Liberty Co.: rte 20 near Ochlockonee River, R, R, Smith 1936 (FLAS) (det. W. Dietrich); new to Florida panhandle. 902 PHYSALIS CORDATA Miller. Dixie Co.: Suwannee River, E of Old Town, R. K Godfrey 65885; Gadsden Co.: Quincy, R. K. Godfrey 67548 (det. J. R. Sullivan); new to Florida panhandle. PHysaLis WALTER! Nutt. is the correct name for Florida plants previous- ly referred to as P. viscosa L. (det. J. R. Sullivan). PLANTAGO HETEROPHYLLA Nutt. Calhoun Co.: Chipola River, H. Kurz in 1943 (FLAS); Gadsden Co.: Chattahoochee, R. K. Godfrey 81234, Quincy, E. West in 1933 (FLAS); Jackson Co.: Paramore Landing, Lake Seminole, A. K. Gholson 10309; Leon Co.: Tallahassee, L. C. Anderson 8004; new to Florida panhandle. PLANTAGO RUGELII Decne. Leon Co.: Lakeshore Drive, Tallahassee, L. C. Anderson 5952 (det. J. Bassett), 8442; ruderal native, new to Florida. POLYGALA VERTICILLATA L. Wakulla Co.: Hwy 98, N of Sc. Marks, L. C. Anderson 8290; new to Florida panhandle. POLYGONUM AVICULARE L. Franklin Co.: Apalachicola, R. K. Godfrey 81441; new to Florida panhandle. RATIBIDA COLUMNIFERA (Nutt.) Woot. & Standl. f. PULCHERRIMA (DC.) Fern. Leon Co.: Madiera Circle, Tallahassee, L. C. Anderson 8064, 8078, adventive, new to Florida (possibly escaped from cultivation but not seen in local gardens). SALIX ERIOCEPHALA Michx. Gadsden Co.: E of Little River on I— 10, A. K. Gholson 11533; Jackson Co.: 7 mi W of Malone, R. K. Godfrey 79125, 79360. 79419, 80348, 80757 (det. G. Argus); native, new to Florida as far as the name is concerned. Listed as S. rigida Muhl. in Clewell (1985); the latter does not occur in north Florida. SCOPARIA MONTEVIDENSIS (Spreng.) R. E. Fries. Santa Rosa Co.: Milton, J. R. Burkhalter 9199; Wakulla Co.: NE of Medart near junction hiways 98 and 319, L. C. Anderson 7263. This South American species was first collected in Florida at Port St. Joe, Gulf Co., M. FE Baker in 1939 (FLAS) and at Carrabelle, Franklin Co., M. F Baker in 1940 (FLAS). It has not been re-collected at Port St. Joe, but it is represented by many collec- tions from Franklin County. The species is now spreading. TACHYS TENUIFOLIA Willd. var. PERLONGA Fern. Gadsden Co.: 3 air mi SE of Havana, L. C. Anderson 7378, 8448; native, new to Florida. These plants are atypically bristly with branching inflorescences and are tenta- tively placed as var. perlonga (J. B. Nelson, pers. comm.). Small (1933) listed the species for Florida, but Nelson (198 1) found no extant specimens for the state. Closest known populations of the variety are in northern Georgia and southern Mississippt. STELLARIA PROSTRATA Baldw. Franklin Co.: St. Vincent Island, L. C. Anderson 7024; new to Florida panhandle. 383 THELYPTERIS INTERRUPTA (Willd.) Iwatsuki. Franklin Co.: 1.5 mi S of Bald Point, 7.8 air mi SE of Panacea, L. C. Anderson 7753, 9073. This tropical fern was first reported for north Florida from Dog Island, Franklin Co., by Anderson and Alexander (1985). The Dog Island population was severely damaged by hurricane Elena in September, 1985; the mainland population may have the only extant plants in the area. VERONICA AGRESTIS L. Franklin Co.: Commerce Street, Apalachicola, flowers white, capsule 4-5 mm long with short, stout style, L. C. Anderson 7919; adventive, new to Florida. VERONICA POLITA Fries. Pensacola, J. R. Burkhalter 3482; new to Flori- da panhandle. This species is sometimes included as part of V. agrestis; the two are distinguished by Pennell (1921). VICIA FLORIDANA S. Wats. Franklin Co.: McIntyre on Ochlockonee River, L. C. Anderson 7005; range extension westward from Taylor County. VictA HUGERI Small. Jackson Co.: Three Rivers State Rec. Area, N of Sneads, G. R. Knight 414; native, new to Florida. Viota sororiA Willd. Nearly glabrous plants from Florida with light- colored flowers that have been assigned to this species have been proble- matic. Cooperrider's treatment (1884) is followed here. Viola sororia f. be- ckwithae House. Franklin Co.: cemetery, Apalachicola, R. K. Godfrey 762066; Jefferson Co.: 3.2 mi N of Waukeenah, L. C. Anderson 78406; Leon Co.: Tallahassee, L. C. Anderson 7834, 7848; new to Florida panhandle. Violo sororia f. priceana (Poll.) Cooperrider. Dixie Co.: Suwanee, SE of rte 249, S. W. Leonard 8087; new to Florida panhandle. Yucca GLoriosa L. Franklin Co.: Bald Point, facing Ochlockonee Bay, L. C. Anderson 7747; new to Florida panhandle. The collection was from a sandy, rear-dune setting far from habitation, but the plants are probably adventive rather than relictual. ZEPHYRANTHES CANDIDA (Lindl.) Herbert. Franklin Co.: naturalized along Avenue K (also on Avenue F) in Apalachicola, L. C. Anderson 7537, new to Florida panhandle. ZEPHYRANTHES GRANDIELORA Lindl. Franklin Co.: naturalized at Brickyard Landing, Apalachicola River, L. C. Anderson 73206; new to Flori- da panhandle. ACKNOWLEDGMENTS This work was supported in part by NOAA financial assistance award NA85AA-D-CZ048 for study in the Apalachicola National Estuarine Sanctuary. 384 REFERENCES ANDERSON, L. C. 1984. gee plants from north Florida. Sida 10:295 — 297. ANDERSON, L. C., and L XANDER. 1985. The vegetation of Dog Island, Florida. Florida Sci. 48:2 . CHAPMAN, A. W. 1897. aera southern United States, ed. 3. Am. Book Co., New York, NY. CLEWELL, A. FE 1985. Guide to the vascular plants of the Florida Panhandle. Florida State University Press/Univ. Presses of Florida, Tallahassee. COCKER, W. C., and H. R. TOTTEN. 1934. Trees of the southeastern States. Univ. North Calica Press, Chapel Hill. COOPERRIDER, T. S. 1984. Some species mergers and new combinations in the Ohio . Mich. Bor. 23:165 — 168. CORRELL, D.S. 1950. Native orchids of North America. Stanford Univ. Press, Stanford, CA. JOHNSON, D. M. 1986. eae of the New World species of Marsilea (Marsileaceae). Syst. Bot. Mono 11:1—87. NELSON, J. B. 1981. Stachys (Labiarae) i in southeastern United States. Sida 9:104 — 123. PENNELL, E W. 1921. Veronica in North America. Contrib. New York Bot. Gard. —4]1 SMALL, J. K. 1933. Manual of the southeastern Florida. Facsimile reprint, Hafner Publ. Co., New York, N-Y. WARD, D. B., and D. W. HALL. 1976. Re-introduction of Marsilea vestita into Florida. Amer. Beta: 66:113— 115. WUNDERLIN, R. P, D. E HANSEN, and D. W. HALL. 1985. The vascular flora of central Florida: taxonomic and nomenclatural changes, additional taxa. Sida 11:232— 244. SYNOPSIS OF THE FLORIDA SPECIES OF PECTIS (ASTERACEAE) DAVID J. KEIL Biological Sciences Department California Polytechnic State University San Luis Obispo, CA 93407, U.S.A. ABSTRACT A key, descriptions, revised nomenclature, range statements and maps are presented for the species of Pectzs known to occur in Florida. A new combination, Pectis glaucescens, is published for the plant formerly known as P. /eptocephala. A naturally occurring triploid interspecific hybrid, Pectis X floridana (P ae x P. prostrata), is described and illustrated. Four species of Pectis (Asteraceae) have been reported to occur in Florida: P. humifusa, PB. leptocephala, P. linearifolia, and P. prostrata (Fernald 1897; Rydberg 1916; Small 1933; Keil 1975; Long and Lakela 1976; Cronquist 1980; Wunderlin 1982). Systematic investigations of Pectis necessitate some taxonomic changes for the species occurring in Florida. An examina- tion of type specimens deposited in the herbarium of the Museum National d’Histoire Naturelle (P) revealed that the basionym of P. leptocephala is predated by an earlier available epithet and that a nomenclatural change is required. Recent field studies have revealed the presence of a heretofore unrecognized natural interspecific hybrid that closely resembles Pectzs lineavifolia. The range of several species is greater than indicated in local and regional manuals (e.g., Anderson 1984). PECTIS L., Syst. Nat. Ed. 10. 1221. 1759. Tyee: P sinifolia L. Tap-rooted or fibrous-rooted annual or perennial herbs. Stems prostrate to erect, often several arising together from the base, straw-colored to deep purplish brown, often diffusely branched, glabrous to puberulent. Leaves opposite, linear to oblanceolate, connected at the base by a narrow connate rim, proximally ciliate with slender bristles, dotted on the undersurface with pellucid glands containing scented oils or unscented compounds, glabrous or minutely puberulent on the margin and midvein. Heads radi- ate, solitary or in open to condensed cymose clusters, the peduncles bearing one to several alternate scale-like bractlets. Involucres cylindric to fusiform or campanulate; phyllaries in a single series, of equal length, distinct or cohering in the proximal | mm, linear to obovate, the margins narrowly SIDA 11(4):385 — 395. 1980. 386 hyaline, often overlapping, acute to rounded, the abaxial surface with a prominent, proximally gibbous keel, dotted or streaked with glands, usu- ally ciliolate distally, glabrous or puberulent. Receptacle flat or hemi- spherical, naked, shallowly pitted. Ray florets pistillate and fertile, equal in number to the phyllaries and individually inserted on the phyllary bases rather than the receptacle; corollas yellow, often suffused with red in age, the tube slender, the ligule elliptical, entire or shallowly 2 — 3-lobed. Disk florets perfect; corollas yellow, the tube slender, gradually expanded to the throat, the limb (in Florida species) bilabiate with a 1-lobed anterior lip and a 4-lobed posterior lip; stamens (in Florida species) 5, included, the apical anther appendages very short, rounded or emarginate; style included or exserted, the branches very short, papillose. Achenes cylindrical, black or dark brown, variously puberulent with 2-celled trichomes. Pappus of lanceolate to setiform scales, awns or bristles, sometimes reduced to a low crown. Chromosome base number: x = 12. About 85 species ranging from California, Nebraska and Florida south through much of Latin America to the Galapagos Islands, northern Argen- tina and the West Indies. In Florida, more frequent from mid-peninsular regions southward, uncommon in the panhandle region (Fig. 1). KEY TO THE SPECIES OF PECTIS IN FLORIDA 1. Heads borne on slender peduncles mostly 5 — 2. Foliar glands all or mostly in marginal row; plane falling —— ly, not coherent at base; achenes seed-bearing. ................ . P. glaucescens 2. Foliar glands submarginal and scattered over the undersurface of i blade; phyllaries coherent at base, falling together with the enclosed achenes; achenes with abortive ovules. .........000 00000 even 3. PX floridana , Hexds sessile or borne on peduncles mostly less than 3 mm long. 3. Involucres campanulate; phyllaries obovate, 2— 4 mm wide; disk florets P2OP Ey, cha te heey Chae ee eas Seca es aaa e 5. P humifusa 3: inves cylindrical to fusiform; phyllaries linear to oblong, 1 — 2.5 mm wide; disk florers 4— 10. 4. Leaves mostly more than 2 mm wide; heads fusiform; foliar glands very numerous, scattered on the undersurface of the leaves; herbage not scented. 4 . PB. prostrata 4. Leaves mostly | — 1.5 mm wide; heads cylindrical; herbage scented. 5. Phyllaries prominently keeled, cohering at base and falling cogeth- er with che enclosed achenes; achenes with abortive ps es; NE DApe Spl cysscenteds est candle ds ccre ee ee was PX floridana 5. Phyllaries nor Sromioeuely keeled, falling individually; ne seed-bearing; herbage lemon-scented. .............0.40. 2. P. linearifolta 1. Pectris glaucescens (Cassini) Keil, comb. nov. Chthonia glancescens Cassini, Dict. Sci. Nat. 9:174. 1817. Type: se unknown origin (HOLOTYPE: P-JU!). 4 i i ' ¢ i i H aw \ i ' -! ; = = a one eho PT ST s ‘a H : ne pene : : oe : ° Lan! + ui? v4 i ofc ical Lo. \ ea, ee ee \ : \ A 7 °° ' > s t Nap ent OO ‘ SA Jeans ek 4 , ¢ ? Lent a ped \ \ a é 1 { : ime 5 i \ ee ow” i ne ; Ce Serene ee Serre , ‘ye. Poa A 3 “3 ro = Doe Figure 1. Distribution of Pectis species in Florida. A. P. prostrata Cav. B. P. linearifolia Urban. C. P. glaucescens (Cass.) Keil. D. PR Aamifusa Swartz and PX floridana Keil 388 Chthonia leptocephala Cassini, Dict. Sci. ma 27:206. 1823. Tyrer: of unknown origin (HOLOTYPE: [according to Cassini} in the Desfontaines herbarium at P burt not located during my visit in 1985). Pats ieee (Cassini) Urban, Symb. Antil. 5:280. : Pectis lessingtt Fernald, Proc. Amer. Acad. Arts 33:67. 1897. Type: UNITED STATES: Froripa: Dade Co.: between the Evergli and Biscayne Bay, Curtis 1162 (LEcro- CH! TYPE: GH !; isoLecroryprs: BM!, CM!, F!, K!, M!, MI , NY (2)!, P!, US!). The lectotype is here chosen con seven syntypes. Spicy-scented annuals (sometimes persisting and becoming subligneous at base), simple to much-branched from the base, sometimes mat-forming but not radicant. Stems slender, erect to prostrate, 2—50 cm long, cymosely much-branched, sparsely to densely puberulent, sometimes glabrate. Leaves narrowly linear, 1— 3.5 cm long, 0.2—1.8 mm wide, mucronulate or setose-tipped, often revolute, proximally ciliate with | —5 pairs of bristles 1-2 mm long, submarginally punctate on the abaxial surface with broadly elliptic to circular glands 0.2 —0.3 mm diameter, sometimes with additional scattered glands, minutely scaberulous on the margins, otherwise glabrous. Heads solitary or in diffuse few- to many- headed cymose clusters; peduncles filiform (3-)7 — 35 (-54) mm long, with 2 —%5 slender, scale-like bractlets 0.5 — 1.5 mm long. Involucres cylindric; phyllaries 5, distinct, falling individually from the receptacle at maturity, linear-oblanceolate, 4—5 mm long, obtuse to acute, slender-keeled to near the apex, often bowed-out near the middle, sparsely punctate with ellipti- cal glands 0.1 -0.2 mm long, distally ciliolate, otherwise glabrous. Ray florets 5; corollas 3-5 mm long, the narrow ligule 2 —3.5 mm long, often involute when dry. Disk florets 3— 7; corollas 2—3 mm long; anthers 1 mm long. Ray and disk achenes similar, 2.5-—3 mm long, antrorsely strigillose. Pappus as composed of 0—5 antrorsely scabrid bristles or slender scales 1 — 2 mm long, and 0 — 5 entire to irregu- larly lacerate scales 0.2. —0.7 mm long. Chromosome number: 7 = 24. Common and widespread in southern Florida and the Bahamas; also in Hispaniola and Jamaica. In Florida it occurs from Glades and Martin coun- ties south to Key West (Fig. 1-C). Flowering specimens have been collected throughout the year. Pectis g/aucescens is most Common on lime- stone soils in open grassy sites. Various types of human disturbance, partic- ularly road construction, have created habitats suitable for these plants, and in places this species is an abundant roadside weed. It also occurs as a lawn weed in the Miami area and probably elsewhere. 2. PECTIS LINEARIFOLIA Urban, Symb. Antil. 5:276. 1907. Tyee: UNITED STATES: Froripa: Hillsborough Co.: Tampa, Nash 2479 (tecrorype: US!; ISOLECTOTY PES: E!, GH!, K!, LE!, MO (2)!, MSC!, NY!, P (2)!, PR!, UC!, WU!). The lectotype is here designated from isosyntypes. Urban designated two syntypes, 389 one from Florida and the second from Jamaica. Both specimens were apparently destroyed when the Berlin herbarium burned during World War II. Iam excluding the Jamaican collection JAMAICA: without locality, MacFayden s.n. (GOET! [frag- ment}, K!), which is actually P glawcescens. Lemon-scented annuals, simple or much-branched from the base. Stems slender, erect to decumbent, 4—40 cm long, puberulent, the upper branches mostly short. Leaves linear, 1—5 cm long, |—3 mm wide, mucronate or setose-tipped, often revolute, proximally ciliate with 2— 6 pairs of bristles 1— 2.5 mm long, submarginally punctate on the abaxial surface with round glands 0.2—0.4 mm diameter, scaberulous on the margins, otherwise glabrous. Heads solitary or in congested terminal and axillary leafy-bracted cymose clusters, sessile or on peduncles up to 1 mm long. Involucre cylindric to narrowly campanulate; phyllaries 5, distinct, falling individually from the receptacle at maturity, linear or linear- oblanceolate, 5-6 mm long, | — 1.5 mm wide, acute, indurate-keeled in the proximal half, punctate with scattered elliptical glands 1-2 mm long, apically ciliolate, otherwise glabrous. Ray florets 5; corollas .5—5.5 mm long, the narrow ligules 3—4 mm long, often involute when dry. Disk florets 4— 10; corollas 2.5 — 3 mm long; anthers 0.6— 1 mm long. Ray and disk achenes similar, 2.25 — 3.25 mm long, antrorsely strigillose. Pappus of ray and disk achenes similar, composed of 2—5 antrorsely barbed bristles or slender scales 1.5 — 2.5 mm long and several shorter barb-margined scales. Chromosome number: » = 24. Endemic to mid-peninsular Florida from Pinellas Co. and northern Polk Co. east to Martin Co. and south to northern Collier Co. (Fig. 1-B). The main flowering period is from August to December, but this species sometimes flowers in the spring months as well. The distribution of P linearifolia is mostly to the north of that of P glaucescens. Both species occur in open sandy or gravelly soils with grasses and other low herbs. I have seen the two together at only one site (along a railroad and roadside in Martin Co.) and observed no hybrids. It is likely that the two taxa grow together at other sites as well. Pectis linearifolia is represented by fewer and more widely scattered collections than is P glaucescens and is apparently less common. 3. Pecris X floridana Keil, nothosp. nov. (Fig. 2). E ceteris speciebus Floridae numero triploideo chromosomatum et acheniis sterilibus differt. EP prostrata foliis angustioribus, capitulis gracilioribus et oleis essentialibus grave- olentibus et E P glaucescens glandibus foliorum dispersis, pedunculis brevioribus et phyllariis proximale coherentibus distinguitur. E P /inearifolia phyllariis proximale coherentibus et valde carinatis, pedunculis isapiocins et odoris oleorum essentialium separari potest. 390 Spicy-scented annuals, simple or much-branched from the base, often mat-forming and sometimes radicant. Stems slender, erect to decumbent, 5 — 30 cm long, puberulent. Leaves linear, 1.5 — 3.5 cm long, 1—- 2 mm wide, punctate on the abaxial surface with numerous scattered glands ca 0.2 mm diameter, scaberulous on the margins, otherwise glabrous. Heads solitary or in condensed axillary and terminal cymose clusters; peduncles filiform, 5—25 mm long, bearing 1—4 scale-like bractlets. Involucre cylindric to narrowly fusiform; phyllaries 5, coherent at base and falling as a group together with the enclosed achenes, linear or linear-oblanceolate, 5—7 mm long, 1—2 mm wide, subacute, strongly indurate-keeled to near the apex, punctate with scattered oval glands 0.2 —0.3 mm long, ciliolate apically, otherwise glabrous. Ray florets 5; corollas 3.5 —4.2 mm long, the narrow ligule 2— 2.7 mm long, often involute when dry. Disk florets 4—6; corollas 2.5 mm long, the anthers ca 1 mm long. Ray and disk achenes similar, 3—3.5 mm long, strigillose to short pilose, the pericarp darkening but not swelling, the ovule abortive and shrunken. Pappus of 2 (ray) or 5 (disk) slender, antrorsely scabrid, setose-tipped scales 2 —2.5 mm long, sometimes with one or more additional shorter scales or bristles. Chromosome number: 37 = 36 Tyee: UNITED STATES: Froripa: Collier Co.: 6 mi SE of Royal Palm Hammock along US 41, 18 Nov 1982, Keil 16488 (HOLOTYPE: FTG; Isoryprs: to be distributed). Known at present from Collier and Dade counties (Fig. 1-D). The in- dividuals collected at these sites were in flower in November and had evidently been flowering for at least two months. Additional specimens examined: FLoripa: Dade Co.: along US 41, 18 mi W of Florida Turnpike, Ke‘/ 16476 (OBI); Everglades National Park at Pa Hay Okee Overlook, Gi//is 7121 [mixed with P. glaucescens] (J). Pectis X floridana is apparently a first generation hybrid between Pectis glaucescens and P. prostrata. Second generation hybrids and backcrosses have not been discovered. In south peninsular Florida, the two parental taxa are known to grow together at several locations. At two sites where | encountered mixed populations of these species, I observed morphological- ly intermediate individuals that | at first mistook for P. /inearifolia. Howev- er, chromosome counts of the two species and the intermediates plus the characters of these plants demonstrate that the intermediates are hybrids. Pectis prostrata 1s a diploid and P. glaucescens is a tetraploid. The resemblance to P. linearifolia is apparently coincidental (Keil 1983). The hybrids are triploids and apparently are completely sterile. Meiosis is very irregular. Pollen grains are malformed and variable in size, and have (0) percent staina- bility in cotton blue in lactophenol (400 grains counted). oI SSS ERR — SSS SS SSS SSS See ————S—_- = — : SD SSS See eee ae oe a en em Figure 2. Pectis X floridana Keil. A. Branchlet with heads. B. Principal foliage leaf. C. Phyllary. D. Ray achene. E. Disk achene. E Ray corolla. G. Disk corolla. C-G, same scale. o22 Regeneration of the hybrids from season to season apparently requires new hybridization events. There is apparently no barrier to hybridization between P. prostrata and P. glaucescens. At the type locality the hybrids were almost as common as the parental taxa. At this site the plants formed a dense mixed roadside population, and the branches of individuals of the two parental species and the hybrids frequently were intertangled. Hybrid individuals were apparently as healthy and vigorous as the parents. Pectis * floridana can be expected in other areas where the two parents occur together. The evipioid hybrids are the potential progenitors of a new hexaploid species. Polyploidy is common in the species of Pectis of the Caribbean region. Both parental taxa are facultatively autogamous, and a fertile hexaploid, if formed, would very likely retain this capacity, thus enabling it to establish a population even if it were surrounded by individuals of the parental taxa. 4. Pecris prostRATA Cavanilles, Icon. Descr. Pl. 4:12. 1797. Tyee: grown at the Madrid Botanical Garden from seed collected by Nee in Queretaro, Mexico (HOLOTYPE: MA, photo OBI!). CAthonia prostrata (Cavanilles) Cassini ex Steudel, Nom. Bot. 598. 1821, pro syn. Lorentea prostrata Lagasca, Gen. Sp. Pl. 28. 1816. Type: CUBA: without location, Boldo 7. (HOLOTYPE: MA, photo OBI). Pectis costata Ser. & Merc. ex DC., Prodr. 5:100. 1836. Type: CUBA: without location, . (HOLOTYPE: G-DC!: Isorype: P!). Pectis pat Cavanilles var. urceolata Fernald, Proc. Amer. Acad. arts 33:68. 1897. EXICO: CuinuAHua: Hacienda San Jose, Palmer 53 (HOLOTYPE: GH!; isoTyPEs: BM!, K!, LE!, NY!). Pectss urceolata (Fernald) Rydberg, N. Amer. FI. 34:197 Pectis multisetosa Rydb., N. Amer. Fl. 34:198. 1916. Typr: GUATEMALA: Santa Rosa: Chupadero, Heyde and Lux 4232 (HOLOTYPE: NY!; isorypes: F!, UC! {frag - ment}) Unscented annuals, simple or much-branched from the base, often mat- forming and sometimes radicant. Stems slender, erect to prostrate, 1 — 30 cm long, often much-branched, sparsely to densely puberulent. Leaves lin- ear to oblong or narrowly oblanceolate, | — 3 cm long, 1.5 —5 mm wide, obtuse to subacute, mucronate, proximally ciliate with 4— 12 pairs of bristles 1-3 mm _ long, densely punctate on the undersurface with scattered round glands 0.1—0.2 mm diameter, scaberulous on the margins, proximally villous-ciliolate, otherwise glabrous. Heads solitary, sessile or subsessile in dense terminal or axillary cymose clusters; peduncles up to 2 mm long, bearing 1-several scale-like bractlets. Involucres cylin- drical or fusiform; phyllaries 5, coherent at base and falling as a group together with the enclosed achenes, oblong to narrowly obovate, 5 — 6.5 393 mm long, 1.5—2.5 mm wide, obtuse to subacute, strongly indurate- keeled to near the apex, punctate with scattered oval glands 0.1 — 0.2 mm long, ciliolate apically, otherwise glabrous. Ray florets 5; corollas 2.5—3.5 mm long, the narrow ligule 1.5 — 2 mm long, often involute when dry. Disk florets 3 — 6; corollas 2— 2.5 mm long, the anthers ca | mm long. Ray and disk achenes similar, 2.5 — 3.5 mm long, strigillose to short pilose. Pappus of 2 (ray) or 5 (disk) lanceolate scales 1.5 —2 mm long. Chromosome number: 7 = 12. Widespread from the southwestern United States south throughout much of Mexico and Central America and from Florida to the Bahamas, Cuba, Hispaniola and Puerto Rico. In Florida, known from the panhandle region in Escambia and Gadsden counties and in the peninsula in scattered locations from Pinellas and Highlands counties south into the Keys (Fig. 1-A). Flowering mostly from August to December. This species 1s probab- ly more widespread than present records indicate. It is often overlooked or ignored because it is a rather unattractive roadside weed. In my field studies in southern Florida I found it to be common in disturbed habitats. Pectis prostrata is a variable species represented in Florida by compara- tively small-headed plants. Larger headed plants occur in some areas of Mexico and Central America. It is not certain whether this species is indig- enous in Florida or introduced from some other region. The first collections from the state were made in the 1840’s but by that time Florida had been a part of Spain's commercial shipping network for several hundred years. Pectis prostrata is rather weedy and its occurrence on the islands of Hispa- niola and Puerto Rico are probably a result of human introduction. It 1s facultatively autogamous and readily pioneers roadsides and other dis- turbed open habitats. It is apparently spreading along roadsides and can be expected to become even more common in Florida in the future. This species may spread to Georgia or Alabama if the population sampled in Gadsden County (Anderson 4573) persists. This collection site 1s approxi- mately 10 miles from the Florida-Georgia state line. 5. Pecris HuMirusA Swartz, Prodr. 114. 1788. Type: VIRGIN ISLANDS: Sr. Croix [Santa Cruz}: without location, Swartz 5.2. (LECTOTYPE: BM!; ISOLECTO- type: G-DC!). The lectotype is here chosen from syntypes. Gia humifusa (Swartz) Cassini ex Steudel, Nom. Bot. 598. 1821, pro syn. Lorentea humifusa (Swartz) Lessing, Linnaea 6:719. 1831 Chthonta sae Cassini, Dict. Sci. Nat. 27:204. 1823. Type: eee RICO: without location, Sagra 5.n. (HOLOTYPE: P-JU!; isotypes: P!, P-LA!). Pectis ie Lessing, Linnaea 6:717. 1831. Type: ie WEST INDIES: Marti- re Tee 24 (HOLOTYPE: CW?; isorypEs: HAL!, JE!, K!, L!, M!, MO!, NY!, P ys LW) 394 Pectis serpyllifolia Lessing, Linnaea 6:715. 1831. Type: PUERTO RICO: without loca- tion, Wydler 208 (Lectotype: G-DC!; tsotecrorypers; E!, K!, L!, OXF!, $!, TCD!). The lectotype is here chosen from four syntypes. Non-scented mat-forming annuals or often perennials, the base often more or less woody. Stems several to many from the base, 2— 25 cm long, prostrate, much-branched, densely leafy, often strongly radicant, puberulent. Leaves oblong-oblanceolate to obovate, 3— 17 mm long, 1.5—4 mm wide, obtuse, mucronulate, proximally ciliate with 2—6 pairs of bristles 1—2 mm long, the bases sheathing, on both surfaces punctate with numerous scattered round glands 0.1—0.2 mm diameter, scaberulous on the margins, proximally villous-ciliolate, otherwise glabrous. Heads terminal and axillary, solitary or in few-headed cymes, sessile or on slender peduncles 1 — 12 mm long with 2 — 3 scale-like brac- tlets. Involucres campanulate; phyllaries 5, obovate, 4.5-—6 mm long, 2—4 mm wide, broadly overlapping, broadly obtuse, indurate-keeled in the proximal 1/2 or 2/3, densely punctate with numerous scattered tiny glands, apically ciliolate, otherwise glabrous. Ray florets 5; corollas 3.5 —5 mm long, the tube 1— 2 mm long and the narrow ligules 2.5 — 3 mm long. Disk florets 12 — 21; corollas 2.5 — 3 mm long; anthers ca 1 mm long. Achenes 2.5—4 mm long, puberulent with trichomes 0.2 —0.5 mm long, the ray achenes abaxially glabrous. Ray pappus of 2 — 3 slender, antrorsely scabrid, bristle-tipped scales 1.5—2.5 mm long and 2— 10 shorter lacerate-margined scales or slender bristles. Disk pappus of 4— 15 antrorsely scabrid bristles or slender scales 2—3 mm long and up to 15 shorter bristles or scales. Chromosome number: » = 36. From Puerto Rico eastward and southward throughout the Lesser Antil- les to the coast of Surinam. Known in Florida from a single collection in 1956 from Collier County (Fig. 1-D) where it is probably adventive (Keil 1975). Efforts in 1982 to relocate the collection site were unsuccessful and it is not known whether this species has persisted in Florida. SPECIES LIKELY TO DISPERSE TO FLORIDA Several other species of Pectis are common in the West Indies and may eventually disperse to Florida. Pectis /inifolia var. linifolia is a very common and widespread taxon found on most of the islands of the West Indies in- cluding the Bahamas. It is a tall, slender, unscented plant with a pappus of stout awns. Pectis elongata var. floribunda is an erect, bushy-branching, lemon-scented herb found in both the Greater and Lesser Antilles. Pectis tenuicaulis, ranging from Venezuela and Colombia to the Dominican Re- public, Puerto Rico’and several islands of the Lesser Antilles, is similar to Pectis prostrata but has 3-rayed instead of 5-rayed heads. Pectis ciliaris occurs 32) in northern South America and the Greater Antilles. It resembles P_ pros- trata but it is a tetraploid. In P. cé/iaris the phyllaries are less-prominently keeled than in P. prostrata and fall separately. ACKNOWLEDGMENTS Research was supported by NSF Grant DEB 81-04683. I thank John Popenoe and Roger W. Sanders for making facilities available at the Fairchild Tropical Garden during my field work in Florida and Donald J. Pinkava for providing laboratory space at Arizona State University during my sabbatical leave. REFERENCES ANDERSON, L. C. 1984. Noteworthy plants from north Florida. Sida 10:295 — 297. CRONQUIST, A. 1980. Vascular flora of the southeastern United States. Vol. Asteraceae. Univ. of North Carolina Press, Chapel Hill. xv + 261 pp. FERNALD, M. L. 1897. A systematic study of the United States and Mexican species of Pecs. “Proc. Amer, Acad. Arts 33:57 — 80. KEIL, D. J. 1975. Pectis bumifusa new to the flora of the United States. Rhodora 77:145 — 146. vy —_____. 1983. Chance — and the de of Pectis linearifolia (Asteraceae). Amer. or a FOG; es 2):119. (abs LONG, W., and O. L oe oe 2 iE of tropical Florida. Banyan Books, ae Elona: xvil + ae P A. 1916. ie Tageteae, Anthemidae. N. Amer. FI. 34:181— 288. SMALL, : K. 1933. Manual of the southeastern flora. Univ. of North Carolina Press, Chapel Hill. xxii + 1554 pp. WUNDERLIN, R. P 1982. Guide to vascular plants of central Florida. Univ. of South Florida Press, Gainesville. 472 pp. BUMELIA DOMINICANA (SAPOTACEAE), A NEW NAME FOR AN OLD SAPOTE R. DAVID WHETSTONE Herbarium, Department of Biology Jacksonville State University Jacksonville, AL 36265, U.S.A. T. A. ATKINSON' Department of Botany, Carolina cartel ks Co. Burlington, NC 27215, In his treatment of the Jamaican Bumelia (Sapotaceae), W.T. Stearn (1968) combined the genera Bumelia and Dipholis due to the lack of signifi- cant characters to consistently separate them. Under this interpretation, Dipholis ferruginea Ekman & O.C. Schmidt became Bumelia ferruginea (Ekman & O.C. Schmidt) W.T. Stearn. This is a later homonym of B. ferru- ginea Nuttall, a name applied to a segregate of B. /anuginosa (Michaux) Persoon which occurs in the eastern United States and northern Mexico. Since this new combination is in direct contravention to Article 64 of the Internatonal Code of Botanical Nomenclature (Voss et al. 1983), the name is illegitimate. BuUMELIA pera Whetstone & Atkinson, nom. nov. Sige ida ag Ekman & O.C. Schmidt, Feddes Repert. Spec. Nov. Regni. Veg. 3 1933. ae (Ekman & O.C. Schmidt) W.T. Stearn, J. Arnold cs re 1968. non Nuttall (1849). The senior author gratefully acknowledges financial support in the form of a Faculty Research Grant from Jacksonville State University. REFERENCES NUTTALL, T. 1849. North American sylva. Philadelpt STEARN, W.T. 1968. Jamaican and other species of Bumelia (Sapotaceae). J. Arnold Ar- bor. 49:280 — 289. VOSS, E.G., et al., editors. 1983. International code of botanical nomenclature... Bohn, Scheltema & Holkema, Utrecht. 'Research Associate, NCU Herbarium SIDA 11(4):396. 1986. TAXONOMIC AND NOMENCLATURAL NOTES ON VACCINIUM L. SECTION CYANOCOCCUS (ERICACEAE) LEONARD J. UTTAL Biology Department, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, U.S.A. ABSTRACT Vaccinium simulatum Small is lectotypified, author citation for V. atrococcum Heller is corrected, and the type locality of V. corymbosum is restricted. In studying the genus Vaccinium L. (Ericaceae) in Virginia and West Vir- ginia, the following taxonomic and nomenclatural details were discovered which seem expedient to publish in advance of the regional treatments. 1. VAccINIUM SIMULATUM Small, Fl. Southeastern U.S. 896, 1336. 1903. Syntypes: KENTUCKY. Harlan Co.: Big Black Mountain, Aug 1893, Kearney 5.n. (LECTOTYPE, here designated, NY, photo neg. no. 11894 NY!). VIR- GINIA. Grayson Co.: slopes of White Top Mountain, alt. ft., 20 May 1892, Britton, Britton, and Vail s.n., (NY, photo neg. no. 11892 NY!). The lectotype is a fruiting specimen, the paratype is a ee just past anthesis. The corollas of sect. Cyamnococcus A. Gray are of limited taxonomic usefulness, whereas the berry of this species is distinctive: shining purple- black, spherical, very juicy, tart. For this reason, the fruiting specimen was selected for the lectotype. In describing Vaccinium simulatum from two dried specimens collected by others, Small gave flawed measurements for corolla (3.5 — 4.5 mm long) and berry (5 — 7 mm in diameter). In life, the cylindro-campanulate corol- a is }— 6 mm long, and the berry 6 — 10 mm in diameter. Small also said the berry is "somewhat glaucous.” I have visited the type locality where the species is still abundant and observed the berries are shining purple-black. A topotype specimen (L.J. Uttal 13848) in fruit is deposited in VPI, and it is planned for duplicates to be sent to NY, FLAS, GH, and NCU. On drying, the berries took on a bluish cast, as black berries of other Vaccinium species (1.e., Vaccinium fuscatum Ait.) sometimes do, which presumably caused Small to misinterpret the berry color. In case of a flawed description, the name is based upon the type specimen, bolstered in this case by topotypes. Occasional glaucous-berried specimens attributed to Vaccinium SIDA 11(4):397 — 399. 1986. 398 simulatum may be attributed to introgression with Vaccinium corymbosum L. and Vaccinium pallidum Ait. In hybrid populations, fruit color may vary from purple-black to glaucous. Vaccinium simulatum 1s indigenous to the Cumberland Plateau, Cumberland Mountains, and middle elevations in the southern Appalachian, from southwest Virginia, eastern Kentucky, eastern Tennessee, western North Carolina and northern Georgia, at eleva- tions from 250 to 1200 meters. It is a forest shrub of loamy soils of ridges and slopes, apparently more mesophytic than most highbush crown- forming species of sect. Cyanococcus. It is usually associated with maple forests (Acer rubrum L. or A. saccharum Marsh.), often with yellow birch, Betula allegheniensis Britt., Hydrangia arborescens L., and Kalmia latifolia L. The forb Asclepias exaltata L. is a very frequent edaphic partner of this species. For more detailed discussion of this species see Camp (1945). 2. VaccINIUM aTROcOccUM Heller—correct author citation. Vaccinium atrococcum Heller, Bull. ee Bot. Club 21:24. 1894 Vaccinium corymbosum vat. atrocarpum Gray, Man. ed. 2. 250. 1856. Vaccinium corymbosum var. atrococcum Gray, Man. rev. ed. 250. 1857. Correctly cited by Camp (1945), but ignored, the citation usually being Vaccinium atrococcum (Gray) Heller for almost a century. Even if Heller tn- tended a recombination, he made an inadvertent species designation by not using the basionym, var. afrocarpum. This name is considered synonymous with Vaccinium fuscatum Ait. fide Ward (1974). 3. VACCINIUM CORYMBOSUM L.—type locality. a eae L. ae Pl. 1:350. 1753 (Lectotype: Kalm , North America,” LINN, mic oe — 6!). oe type sheet of een ‘oman bears two el menes (a afy summer branch and (2) a spring branch in t leafing out. Vander ae (1980) ae the name by the ae ae “which he described. Pehr Kalm resided in Raccoon (now Swedesboro), Gloucester County, New Jersey, in the springs of 1749 and 1750 (Benson, 1937). The lectoty- pified specimen is in the condition one would expect in May in New Jersey. The summer specimen was obtained either in Canada in 1749 or in western Pennsylvania or western New York in 1750. It thus seems safe to restrict the type locality of V. corymbosum to the vicinity of Swedesboro, Gloucester County, New Jersey, May 1749 or 1750. This rescriction of type locality 1s considered important because the lectotype is of a morphology common tn the northeastern United States and adjacent Canada, not found in the southern states except in the mountains, a point to be considered by 229 students of the taxonomy of the highbush species of Cyanococcus of the southeastern United States. ACKNOWLEDGMENTS Thanks are due P K. Holmgren, of the New York Botanical Garden, for furnishing me photographs of the type material of Vaccinium simulatum, and to the library staff of the Harvard University herbaria for furnishing copies of relevant literature. REFERENCES BENSON, A. O. 1937. Peter Kalm’s travels in North America. The English Version tn 1770. Revised and edited. 2 vols. 797 pp., Wilson-Erickson, In CAMP. W. H. 1945. The North American blueberries with some notes on other groups of Vacciniaceae Brittonia 5:203 — 275. VANDER KLOET, S. P. 1980. The taxonomy of the highbush blueberry, Vaccensum eee fan. J. Bor. 58:1187 — 1201. WARD, D. B. 1974. Contributions to the flora of Florida —6. Vaccinium (Ericaceae). Castanea 39: 191 — 20 ou ] A NEW COMBINATION IN HEDYOTIS L. (RUBIACEAE) R. RP WUNDERLIN Department of Biology, University of South Florida Tampa, FL 33620 As a result of his continuing studies of Howstonia, Terrell (1985) proposed two new varietal combinations under Howstonia nigricans. These combinations are warranted if Houstonia is accepted as a distinct genus as Terrell contends (cf. Terrell 1959, 1975). Whether Houstonia should be recognized as a distinct genus or combined under Hedyotis is debatable. | feel that Fosberg (1941,194 1a, 1954), Shinners (1949), and Lewis (1961, 1962) argue favorably for reducing Houstonia under Hedyotis. This is the position I have taken for the vascular flora of central Florida (Wunderlin 1982) and, unless convincing evidence is presented to the contrary, intend to maintain for the flora of Florida (in prep.). Thus, the following new combination in Hedyotis is required. HeEDYOTIS NIGRICANS var. floridana (Standley) Wunderlin, comb. nov. Basionym: Hoxstonia floridana page N. Amer. Fl. 32(1):36. 1918. Hedyotis purpurea vat. floridana (Standley) Fos hee Castanea 19:36. 1954. Houstonta nigricans var. floridana (Standley) Terrell, Phytologia 59:79. 1985 REFERENCES FOSBERG, E R. 1941. Observations on Virginia plants, part 1. Virginia J. Sci. 2:106- 111. 1941a. Notes on Mexican plants. Lloydia 4: . 290, 1954. Notes on plants of the eastern United States. Castanea 19:25 — 37. LEWIS, W. H. 1961. Merger . — North American ny and Oldenlandia under Hedyotis. Rhodora 63:216— 1962. Phylogenetic nee of Hedyotis (Rubiaceae) in North America. Amer. J. Bot. 49:855 — 865. SHINNERS, L. H. 1949. eee of Texas species of Houstonia to Hedyotis (Rubiaceae). Field & Lab. 17:166— 169. TERRELL, = E. 1959. A revision of the Houstonia purpurea group (Rubiaceae). Rhodora G15?) = 207. stele, (LOE Dz Sey sane of Hedyotis fruticosa L. to Houstonia L. and Oldenlandia L. hrs 31:418 — 42 nner eee ae “combi in Houstonta and Oldenlandia (Rubiaceae). san A 59: a WUNDERLIN, R. P. oe A guide to the vascular flora of central Florida. University of South Florida Press/University Presses of Florida, Gainesville. SIDA 11(4)-400. 1980. MILLEROCAULIS, A NEW GENUS WITH SPECIES FORMERLY IN OSMUNDACAULIS MILLER (FOSSILS: OSMUNDACEAE) WILLIAM D. TIDWELL Department of Botany and Range Science Brigham Young University, Provo, UT 84602, U.S.A. ABSTRACT Millerocaulis first used by Erasmus and mentioned by Herbst for Miller's previously proposed “Osmundacaulis herbstii group” is discussed and validated. Osmundacaulis, as now restricted, includes only members of Miller's “O. skidegatensis group.” A new informal group “O. hoskingit” is proposed for Osmundacaulis (sensu stricto) containing an el crenate- shaped sclerenchyma mass in their leaf and/or petiole traces. New nomenclatural combina- tions include: Millerocaulis amajolensis (Sharma) Tidwell, Mullerocaulis beardmorensis (Schopf) Tidwell, Millerocaulis dunlopii (Kidston & Gwynne-Vaughn) Tidwell, Millerocaulis estipularis (Sharma, Bohra & Singh) Tidwell, Millerocaulis gibbiana (Kidston & Gwynne-Vaughn) Tidwell, Millerocaulis guptai (Sharma) Tidwell, Millerocaulis hebeiensis (Wang) Tidwell, Millerocaulis herbstii (Archangelsky & de la Sota) Tidwell, Millerocaulis indica (Sharma) Tidwell, Millerocaulis kidstonii (Stopes) Tidwell, Millerocaulis kolbei (Seward) Tidwell, Millerocaulis patagonica (Archangelsky & de la Sota) Tidwell, Millerocaulis rajmahalensis (Gupta) Tidwell, Millerocaulis sahnii (Mittre) Tidwell, and Millerocaulis wadei (Tidwell & Rushforth) Tidwell INTRODUCTION The organ genus Osmundacaulis was established by Miller (1967) as the name of a new taxon, that he thought was a new name for the illegitimate later homonym Osmundites Unger (1854), non Jaeger (1827), but techni- cally was not. In order to have been a new name for Osmundites, Osmunda- caulis would have required the same type as Osmundites. However, Miller (1967, 1971) explicitly excluded Unger’s type 0. schemnicensis (von Pettko) Unger (treated in Osmunda), established a different type for Osmundacaults {O. skidegatensis (Penhallow) Miller}, and gave a validating description. Osmundacaulis was instituted and named for petrified axes exhibiting general anatomical features of living members of the Osmundaceae, but because they differ anatomically from these members, they cannot be assigned to an extant genus of this family. Miller (1967, 1971) proposed three informal groups within Osmundacaulis: the “Osmundacaulis herbsti group,” the “O. braziliensis group” and the “O. skidegatensis group.” These groups were based upon distinctive anatomical features. These features in- SIDA 11(4):401 — 405. 1986. 402 cluded the thickness of the xylem cylinder, the number of clusters of proto- xylem cells in each trace as they departed from the xylem strands and the degree of differentiation of the inner and outer cortices. The “O. braziliensis group,” in which the axis lacks a definite sclerotic outer cortex and stipular wings, was proposed as the new genus Guwairea Herbst. It was removed from the Osmundaceae and placed in the new fami- ly Guaireaceae (Herbst 1981). The" Osmundacaulis herbstii group” was proposed as the new genus Millerocaulis by Erasmus in his unpublished doctoral dissertation (1978). Because this does not constitute effective publication, the generic name was not effectively published (Art. 29). Herbst (1981:37) accepted Erasmus’ M<¢/lerocaulis but failed to validate it with a description (Art. 32.1). Therefore, Mi/lerocaulis is validated here for the first time. TAXONOMY Millerocaulis Erasmus ex Tidwell Foss1] osmundaceous rhizomes, rarely arborescent axes, containing stem or stems surrounded by a mantle of leaf bases and roots. Stele ectophloic- dictyoxylic-siphonostele (Miller 1971) with a xylem cylinder approximate- ly 15 tracheids thick. Leaf trace separates from the xylem cylinder with only one protoxylem cluster and often, but not always, lacks axillary sclerenchyma. Petiole bases stipulate and adventitious roots arise either singly or in pairs. Type: M. dunlopii (Kidston & Gwynne-Vaughn) Tidwell (Osmundites dunlopii Kidston & Gwynne-Vaughn “dyn/op7”) The generic name honors Dr. Charles N. Miller, Jr. of the University of Montana at Missoula for his contribution to our knowledge of the phylogeny of the Osmundaceae. The list of species assigned to Méllerocaulis and their synonyms are as follows: MILLEROCAULIS amajolensis (Sharma) Tidwell, comb. nov. Osmundacaulis amajolensis Sharma, Palaeontographica 140B:156. 1973. Miiterocautts beardmorensis (Schopf) Tidwell, comb. nov. Osmundacanlis beardmorensis Schopf, Can. J. Bot. 56:3034. 1978. Mitterocautts dunlopii pore & Gwynne-Vaughn) Tidwell, comb. OV. Osmundites dunlopti Kid & Gwynne-Vaughn, Trans. Roy. Soc. Edinb. 45(D:759. 1907 ¢ ene ). nee: dunlopit (Kidston & Gwynne-Vaughn) Miller, Contr. Mus. . Univ. Mich. [21:146. 1967 ('dunfop’), nom. invalid. under Art. 33.2 - no page reference to basionym] 23:135. 1971 oe Osmundites aucklandicus Marshall, Trans. and Proc. N.Z. Inst. 56:210. 1924. 403 MiLLerRocAULIs estipularis (Sharma et al.) Tidwell, comb. nov. Osmunda- caulis ne Sharma, Bohra & Singh, Ph ysomerpicleay 8:61. 1979 (“estipu Mitterocautis gibbiana ke & Gwynne-Vaughn) Tidwell, comb. NOV. Osmundites gibbiana Kidston & Gwynne-Vaughn, Trans. Roy. Soc. Edinb. 45(1):763. 1907 Osmundacaulis ae (Kidston & Cane Vaughn) Miller, Contr. Mus. Paleo. Univ. Mich. {21:146. 1967, nom. invalid. under Art. 33.2 - no page reference to Pasonyia 23:136. 1971 MILLEROCAULIS guptai (Sharma) Tidwell comb. nov. Osmundacaulis guptai Sharma, Palaeontographica 140B:154. 197: Mitterocautts hebeiensis (Wang) ate comb. nov. Osmundacaulis hebei- ensis Wang, Rev. Palaeobot. Palyn. 39:93. 1983. Mitterocautis herbstii (Archangelsky & de la Sota) Tidwell, comb. nov. Osmundites herbstti eae & de la Sota, Ameghiniana 3:135. 1963. eee caulis herbstii (Archangelsky & de la Sota) Miller, Contr. Mus. Paleo. Univ {21:146. ma nom. invalid. under Art. 33.2 - no page reference to ee 23:134. 197 MILLEROCAULIS ae (Sharma) Tidwell, comb. nov. Osmundacaulis indica Sharma, Palaeontographica 140B:157. 1973. Mitterocautts kidstonii (Stopes) Tidwell, comb. nov. Osmundites kidstonii ee, Ann. Bot. 35:55. 1921 (“Azdstonz”) oe kidstonii (Stopes) Miller, Mus. Paleo. Univ. Mich. [21:146. 1967 (‘&:dstonr”), nom. invalid. under i t. 3 - no page reference to basionym] 23: 136. 1971 Mitterocautts kolbei (Seward) Tidwell, comb. nov. Osmundites kolbei Seward, Geol. Mag., N.S.V. 4:482. 1907. Ovpandsienlh kolbei (Seward) Miller, Contr. Mus. Paleo. Univ. Mich. [21:146. 1967, nom. invalid. under Art. 33.2 - no page refer- ence to basionym]} 23:136. 1971. MILLEROCAULIS patagonica (Archangelsky & de la Sota) Tidwell, comb. NOV. Osmundites patagonica Archangelsky & de la Sota, Ameghiniana 2(9):153. sek sees patagonica (Archangelsky & de la Sota) Miller, Contr. Mus. Paleo. Uni h. {21:146. 1967, nom invalid. under Art. 33.2 - no page refer- ence to aneter 23:136. 1971. MitterocaAutts rajmahalensis (Gupta) Tidwell, comb. nov. Osmundites rajmahalensis Gupta, Proc. Ind. Sci. Congr. Varanasi, 55:428. 1968. Osmundites rajmahalensis Gupta, Pal acoatosraphite 130B:174. 1970. Osmundacaulis rajma- halensis (Gupta) Sharma, Palaeontographica 140B:152. 1973. MILLEROCAULIS Ssahnit (Mittre) Tidwell, comb. nov. Osmundites sahnii Mittre, reas A135 W955; Oimenaui: sahnit (Mittre) Miller, Contr. Mus. iv. Mich. [21:146. 1967, nom invalid. under Art. 33.2 - no page refer- ence to ee Pola IE L. MILLEROCAULIS wadei (Tidwell & Rushforth) Tidwell, comb. nov. Osmun- dacaulis wader Tidwell & Rushforth, Bull. Torrey Bot. Club 97:137. 1970 OSMUNDACAULIS Miller emend. gen. Emended diagnosis: Fossil osmundaceous axes, usually arborescent or erect, rarely rhizomatous; a xylem cylinder 25 or more tracheids thick dis- 404 sected into relatively high number of xylem strands; leaf trace strongly curved, protoxylem divides before leaving the stele; sclerenchyma usually in adaxial concavity of the trace; petioles stipulate, wings may or may not contain sclerenchyma strands; inner and outer cortical tissue well differentiated. Tyee: O. skidegatensis (Penhallow) Miller (Osmundites skidegatensis Penhallow). The“ Osmundacaulis skidegatensis group” comprises a different taxon among the other members of the Osmundoideae. Thus with the formation of Millerocaulis and Guairea, Osmundacaulis is reserved exclusively for members of the “O. skédegatensis group” (Herbst 1981). Therefore, Osmun- dacaulis, in a strict sense, consists of the species Osmundacaulis skidegatens1s (Penhallow 1902) Miller, O. atherstoner (Schelpe 1956) Miller, O. natalensis (Schelpe 1955) Miller and O. Aoskingii Gould (1973). The inner cortex of Osmundacau/is (sensu stricto) is wider than the outer. Among other members of the Osmundoideae, with the exception of Millerocaulis wadei (Tidwell & Rushforth 1970) Tidwell, the opposite con- dition is the case. The outer cortex of these other members of the Osmun- doideae is very wide and the inner is thinner. In M. wadez, they are about equal in width. According to Miller (1971), cortical cylinders of nearly equal dimensions represent a primitive state, whereas, thicker outer and thinner inner cortices would be more advanced. Further, two groups in Osmundacaulis (sensu stricto) are proposed. One of these groups would contain only O. skdegatensis (Penhallow 1902) Miller and the other, designated the “Osmundacaulis hoskingii group,” is represented by arborescent and rhizomatous taxa having crenate-shaped adaxial sclerenchyma mass in their leaf traces and/or petiole vascular strands. At present, these forms have been reported exclusively from the Southern Hemisphere and include O. atherstonet (Schelpe 1956) Miller, O. natalensis (Schelpe 1955) Miller and O. Aoskingzz Gould. The crenate- shaped adaxial sclerenchyma in the trace is not present in O. skidegatensis. In this species, the sclerenchyma of the trace and petiolar vascular strand is a single mass connecting downward with similar cells in the pith (Miller 1971) ACKNOWLEDGMENT The author ts grateful to Dr. Dan H. Nicolson of the Smithsonian Insti- tution and Professor Stanley Welsh for reviewing this paper and for their aid with the taxonomic nomenclature. 405 REFERENCES ERASMUS, T. 1978. The coral and evolution of Osmundacaulis Miller emend. with notes on the geometry o xylem framework of the osmundaceous stele. Univ. of Pretoria ae PhD: Biter ven, GOULD » LD Proc. ae Soc 5 i A new species of Gaal from the Jurassic of Queensland. —94 HERBST, R. w South Wales oes 1981. as miller’ nov. gen p. y ult ireaceae, nueva Familia de las Osmundales (sensu /ato) del Permico superior de Paraguay. Ameghiniana 18:35 — JAEGER. G. E 1827. Uber die Pflanzenversteinerungen weiche in dem Bausandstein von Stuttgart gee J. B. Metzler, Scuttgart. KIDSTON, R. WY NNE- ees na T. 1907. On the fossil Osmundaceae. Part I. ‘Trans. aoe ae Edinb. 45:759 — MILLER, C. N. 1967. ae of ~ ie genus Osmunda. Contr. Mus. Paleontol. Univ. Mich. es Lo 97 ae OF the fern eo. eral based on anatomical studies. Contr. Mus. Paleontol. Univ. Mich. 28(8):10 sapere D. P1902. Osmandites er n. sp. Trans. Roy. Soc. Canada II ear LPE, a AsCESE. 1955. se hg aie te new fossil fern from the Cretaceous of Zululand. Ann. Mag. Nat 8(12):654 — 656. 1956. Osmundites pan new Cretaceous fern from Cape Province, South Arce Ann. Mag. Nat. Hist. 9:330— 332. TIDWELL, W. D. & S. R. RUSHFORTH. Osmundacaulis wadet, a new osmundaceous species from the Montison Formation (Jurassic) of Utah. Bull. Torrey Bot. Club O16) 197-—. 144; UNGER, E 1854. Ein fossiles Farnkraut aus der Ordnung Osmundaceen. Denkschr. Kaiserl. Akad. Wiss., Math.- Naturwiss. KI. 6(1):143. REVIEW NIXON, E. S. 1985. Trees, shrubs, and woody vines of East Texas. 240 pp., illustrated. Bruce Lyndon Cunningham Productions, 102 Shields Drive, Nacogdoches, Texas 75961. ISBN 0-934115-00-1 (pbk). This paper back manual (9 3/8” X 12 1/2”) was designed to aid woody plant identification in the east Texas forest region which takes in some 40 counties. It is profusively illustrated with line drawings of all 340 species as well as an illustrated glossary. The table of contents includes: Preface, Map of east Texas counties; Introduction; East Texas; Key to groups; Key to genera; Families, genera, and keys to species; Illustrated glossary; Litera- ture cited; Index of common names and scientific names; Metric system of measurement. BLL. BOOKS RECEIVED CLEWELL, A.E 1985. Guide to the vascular plants of the Florida Panhandle. 605 pp., 20 figs. University Presses of Florida, Florida State Univ. Press, Tallahassee. ISBN 0-8130-0779-8. MOHLENBROCK, R.H. 1986. Guide to the vascular flora of Illinois. Ed. 2, 507 pp. Southern Illinois Univ. Press, Carbondale and Edwar- dsville. ISBN 0-8093-1273-5 (pbk) and ISBN 0-8093-1272-7. RZEDOWSKI, J. & G.C. DE RZEDOWSKI. 1985. Flora Fanerogamica del Valle de Mexico. Vol. II, Dicotyledoneae (Euphorbiaceae- Compositae). 674 pp., 130 figs. Instituto de Ecologia, AP 18-845, Deleg. Miguel Hidalgo, CP 11800, Mexico, D.E ISBN 968-72 13-02- re VAN BRUGGEN, T. 1985. The vascular plants of South Dakota. Ed. 2, 476 pp. lowa State Univ. Press, Ames, Iowa. ISBN 0-8138-0650-x (pbk). SIDA 11(4):406. 1986. WOLFFIA PAPULIFERA THOMPSON (LEMNACEAE), NEW TO MICHIGAN WILLIAM J. HESS The Morton Arboretum, Lisle, IL 60532, U.S. A. ABSTRACT Wolffia papulifera Thompson is reported for the first time from Michigan and LaPor County, Indiana. An illustration of W. ee in flower and electron ae eee of its echinate pollen surface are included. Two recent collections to the Morton Arboretum (MOR) herbarium of Wolffia papulifera Thomps. were new records for Michigan and LaPorte County, Indiana. Deam (1940) had reported it from Ohio, Indiana, [l1- nois, Kentucky, Tennessee, Missouri, Arkansas, and Kansas. Daubs (1965) monographed Lemnaceae and cited collections from Maryland, Florida, Louisiana, Illinois, Missouri, Kansas, Mexico, and Argentina. Other reports include it from Texas (Blake 1952) and Oklahoma (Correll & Correll 1975). In the most recent monograph of Lemnaceae (Landolt 1980), W. papulifera and W. punctata Grisebach (in part) are reduced to synonymy under W. brasiliensis Weddell. Now included in its distribution in North America also are Alabama, Massachusetts, North Carolina, Penn- sylvania, and South Carolina. Voss (1972) did not record it from Michigan nor did Swink and Wilhelm (1979) record it from the seven Indiana counties and one Michi- gan county of the Chicago region flora. Deam’s Indiana record of Wolf/fia papulifera was from Posey County in the extreme southwest corner of the state. Lake County, Illinois (Mohlenbrock & Ladd 1978) was the nearest known locality to Michigan and northwestern Indiana until the Morton Arboretum accessioned its first record from Cook County, Illinois (J. Higgens, s.n. 1983). Under W. papulifera or W. brasiliensis, the state and county records are still valid. The collection data are: MICHIGAN: Berrien Co.: Buttonbush swamp in NW quadrant of Bridgman inter- change of I-94, 7 Sep 85, K. Dritz s.n. (MOR) INDIANA: LaPorte Co.: Hudson Lake, ca 2 mi W of New Carlisle, 5 Sep 85, G. Wilhelm & K. Dritz 13082 (MOR). Wolffia papulifera (sensu Daubs) and W. brasiliensis (sensu Landolt) ts dis- tinguished from the other species of Wo//fia on the basis of its central cont- cal papilla on the upper surface of the frond. SIDA 11(4):407 —411. 1986. 408 Both collections from Indiana and Michigan had plants that were in prime flower and here presented are illustrations of the relatively unknown flowers of Wolffia. The smallest known flowering plants are in this genus (Daubs 1965) and its flowers are equally as small. The plants are monoe- cious with a single stamen in the staminate flower and one pistil in the pistillate flower. Both flowers lack a perianth and are housed in a large central cavity. Six stages are shown (Fig. 1). The central papilla is evident in the vegetative and incipient flowering condition (A & B). In C, on top of the frond, is a round opening where the papilla was and through which the style with its slightly impressed, terminal stigma emerges. Often the an- ther sacs can be seen within the central chamber. In all cases examined, the style was exposed prior to the emergence of the stamen and always on the side toward the vegetative reproductive pouch or bud. The stamen emerged (D—F) after the stigma appeared to dry up slightly. Anthesis occurred by the breakdown of a line of cells across the top and between the two anther sacs, afterwhich the sacs spread out laterally upon dehiscence (F). The fruit remains within the flower chamber, however, mature fruit was not evident in our material. Anthers were dissected out of the plants and their anther sacs opened so as to examine pollen grains. They were measured with an ocular microme- ter using a Zeiss Photomicroscope II. They varied in size from 26 — 28 p, somewhat larger than the 18 — 22 w reported by Daubs (1965). The surface of the pollen grains were examined with an ISI SX-40 electron microscope after they had been coated with gold-paladium in a Polaron Sputter Coater. Although the pollen are distorted due to their collapse in the vacuum, they are, as reported by Daubs (1965), echinate (Fig. 2). An effort was made to obtain chromosome counts of Wolffia papulifera, however all the squashes of the anthers resulted in either microspores be- yond the spore tetrad stage or fully developed pollen grains. It is not clear when microsporogenesis occurs but it must be very early in the develop- ment of the staminate flower. Chromosome counts are available for Wolffia (Landolt, 1980) based on somatic cells. For W. brasiliensis, the variation is extreme with 2n= 20, 40, 50, 60, and 80. ACKNOWLEDGMENTS I would like to thank Gerould Wilhelm and Ken Dritz for making their collections of Wolffia available for study. The beautiful drawing of Wolffa papulifera was done by Nancy Bartels. 409 {mm Fig. 1. A—E Wolffa papulife f flowering (Wilhelm & Dritz 13082). A. Vegetative plant. B. Incipient flowering. c Reco cise Biseiare flower. D—E. Maturation of staminate flower. E Anthesis. 410 = Fig. 2. SEM photomicrographs of pollen grain surface of Wolffia papulifera (Dritz s.n.). A. Slighly mane echinate pollen grains (250 X). B. Echinate surface of pollen grain (1000 x) REFERENCES BLAKE, Sok hoe ee papulifera in Texas. Rhodora 54:306 — 307. CORREEL,. B.S B. CORRELL. 1975. Aquatic and wetland plants of southwestern United States, eras Univ. Press, Stanford CA. 1777 pp DAUBS, E. H. 1965. A monograph of Lemnaceae. IJ linois Bialea al Monographs No 34. The University of Illinois Press, Urbana, Illinois. 118 p DEAM, C. C., 1940. Flora of Indiana, Department of Gnd Division of Forestry, lndianapolis Indiana. 1236 pp. LANDOLT, E. 1980. Editor. Se alae unter suchungen in der Familie der ee lingen (Lemnaceae). Veroff. Geobot. ETH Stiftung Rubel Zurich 70:5 — 247. MOHLENBROCK, R. H. & D. M. oo. 1978. Distribution of Illinois vascular rp Southern Ilinois University Press, Cui sondile & Edwardsville, Illinois. SWINK, FE & G. WILHELM. 1979. Plants of the Chicago region. The Morton oe tum, Lisle, Illinois. 922 pp. VOSS, E. G. 1972. Michigan flora. Part 1. Gymnosperms and Monocots. Cranbrook Insti- tute of Science, Bloomfield Hills, Michigan. 488 pp. REVIEW A NEW FERN MANUAL LeLLINGER, Davip B. 1985. A field manual of the ferns and fern-allies of the United States and Canada. 389 pp., 26 figs., 402 color illustrations. Smithsonian Press, Washington, D.C. ISBN 0-87474-603-5 (pbk) and ISBN 0-87474-602-7. This handsome book is a welcome addition to the popular fern-literature of North America, and will be useful in other parts of the North Temperate Zone. It is particularly valuable as a book written for an amateur audience, in popular and understandable language. However, it also seems critically attentive to accuracy ina field noted for its complexity, wide differences of Opinion, and for much superficial popular writing The stated purpose of the work is to aid in identifying all of the species of ferns and fern allies that are native to or naturalized in Canada and the United States, excluding Hawaii. In carrying out this purpose it 1s eminently successful. Descriptions of taxa at the three principal levels, family, genus, and species, are clear and accompanied by discussion and explanation where necessary. Keys are of the “bracket” type, to save space, and they work, so far as I have tried them. Line drawings are provided to illustrate some techincal terms in the excellent glossary. A map and diagrams clarify ideas that are unfamiliar to many. And 402 magnificent color photos mostly by A. Murray Evans are provided for visual recogni- tion, as well as to display the beauty of ferns. The photos are small, to keep the cost of the book down, but beautifully reproduced, nine to a page, in three columns, the numbering, curiously, from right to left on the page. The classification adopted, selected from a number of relatively recent ones available, is not the most conservative in number of families accepted, but neither is it the most extreme. Twenty families of ferns proper are ad- mitted for the United States and Canada, about half of which are readily recognized. The others may require more experience than most amateurs possess. The family key is not easy to follow, but so far as I can see, it will, if carefully used, lead to the correct family. Some of the families are so techni- cal that they are hard to recognize by aspect. An outstanding feature of the book is its recognition and explanation of hybridity in ferns. The prevalence of hybridization in these plants has led to much confusion and description of questionable species. A careful study of the chapter on hybrid complexes will alert the user to this phenomenon. The ample introduction should provide the beginner with an idea of the range and fascination of the study of these plants. The more advanced student will find many of his questions answered, too. The only major suggestion for a second edition would be to include fa- miliar synonyms. The almost cotal lack of synonyms in a work where much new nomenclature is used is frustrating. One cannot locate a plant that one knows by a former commonly used name except by guessing from its posi- tion in the book, unless one already knows the many recent changes in fern classification and nomenclature. In my own case, the only way I could lo- cate what I grew up knowing as Cheslanthes californica was to look in the vicinity of the treatment of Chezlanthes, where I found Aspidotis californica. All in all, one can recommend this book as the best available popular treatment of American ferns, and it is also useful to the more botanically trained non-pteridologists. ER. Fosberg SIDA 11(4):412. 1986. A NEW VARIETY OF HEDEOMA HYSSOPIFOLIUM GRAY (LAMIACEAE) JAMES HENRICKSON Department of Biology, California State University Los Angeles, CA 90032, U.S.A. ABSTRACT Disjunct populations of Hedeoma hyssopifolium from northeastern Chihuahua and western Coahuila differ from Arizona and southwestern New Mexico populations in their more erect, nonrhizomatomous stature, more angular stems with vestiture restricted to decurrent lines, larger inflorescences, shorter, nonciliate lower calyx lobes and are recognized as Hedeoma hyssopifolium vat. chibuahuensis Henrickson. During preparation of the treatment of Lamiaceae for the Chihuahuan Desert Flora it became apparent that the disjunct populations of Hedeoma hyssopifolinm from northeastern Chihuahua and adjacent Coahuila differ in a number of substantial characters from populations of central and southeastern Arizona and adjacent New Mexico, and Sonora. The Chihu- ahuan Desert populations are described below as a distinct variety. HEDEOMA HyYSSOPIFOLIUM Gray var. chihuahuensis Henrickson var. nov. AH. Ayssopifolio var. byssopifolio caudice bene evoluta superficiali (non subterraneo, non polyrhizomifero), internodiis medicaulinis folia longioribus (non brevioribus), caulibus quadrangularibus (non rotundato-quadrangularibus), dichasio axillari (1-)3 — 7(-13)-floro (non 1(-3)-floro), calycis lobis superis acicularibus, rectis vel debile sursum arcuatis (non angustideltoideis et valde recurvatis) lobis inferis 1— 2.2 mm longis debile ciliatis (non (1.4-)2.2 — 3.3 mm valde ciliatis) differt. Erect-ascending suffrutescent herbs 2.5 — 4(-G) dm tall developing from a woody crown; stems several from base, unbranched or rarely branched in inflorescence; lower internodes 5 — 10 mm long, mid-stem internodes (12-) 17 — 26 mm long, 0.7 — 1.3 mm wide, upper stem internodes 3 — 10 mm long; stems glabrous or hirtellous with erect hairs 0.05 — 0.15 mm long in decurrent lines below nodes, quadrangular, shallowly sulcate on surfaces alternate to leaves. Leaves linear-lanceolate, (8-)15 —26 mm long, 1—2 mm wide, acute at tip, cuneate at subsessile base, entire, glabrous, firm, erect-ascending, shorter than nodes, smooth above, with medial and later- al veins slightly raised, straight and with sunken glands beneath; leaf margins not forming an interpetiolar rim across nodes. Flowers borne in upper fourth of plant in paired, axillary,(1-)3 —7)—flowered dichasia, SIDA 11(4):413 — 416. 1986, 414 with lateral branches sometimes proliferating monochasially producing up to 13 flowers per inflorescence; bracts and bracteoles leaf-like in shape and texture; bracteoles 1—2.5 mm long, usually longer than peduncles; peduncles 0.5 — 1(-3); pedicels 1.5 — 2.5(-7) mm long, both short hirtel- lous; calyx tubes cylindrical, slighty tapered at base, slightly ampliate, .5—6.8 mm long, strongly ribbed, glabrous to short hirtellous; upper 3 calyx lobes acicular, 0.5 — 1.2 mm long, mostly straight or very slightly curved upward, lower calyx lobes acicular, 0.9 — 2.2 mm long, weakly ciliate with hairs to 0.1 mm long; calyx annulus at orifice of throat, hairs 0.5—0,7 mm long; corollas lavender to pink, 10-13 mm long, tube straight, slightly ampliate above, upper lobe erect, emarginate, lower 3 lobes spreading, middle lobe longest, emarginate; filaments 3-4 mm long, anthers 1.0— 1.1 mm wide (fig. la—e). Type: MEXICO. CHtnuAnua: road from Castillon to ee via S$. Salvador and Piramide, basalcic cliffs near Virulento, crevices of cliffs, 8—24 inches tall, corollas pinkish, 21 — 22 Sep 1940, 1.M. Johnston & C. H. Muller a 2 (HOLOTYPE: LL; GH) Additional collections; MEXICO. Coanuita: Canon del Indio Felipe, a deep wooded canyon with running water in the igneous Sierra Hechiceros, close to ae Chihuahuan pee crevices in cliffs, abundant, fls. lavender, 27 — 29 Sep 1940, R. M. Stewart 80 (GH, LL); same area; crevices of cliffs, not abundant, flowers lavender, 27 - 29 i 1940, R. M. Stewart 153 (GH, LL); Sierra de Hechiceros: naa n del Indio Felipe, (beyond se del ne from Rancho El Tule; crevices of cliffs in deep canyon, corolla lavender, 18 Sep 1940, 1. M. Johnston G C. H. Muller 1352 (LL). Hedeoma hyssopifolium var. chihuahuensis appears restricted to rocky crevices of cliffs in margins of mesic canyons in igneous mountains (Sierra de Hechiceros, Sierra del Virulento) in the Northern Chihuahuan Desert. These populations lie some 550 km southeast of the nearest populations of H. fh. var. hyssopifolium in the Animas Mountains in southwestern New Mexico (A. Zimmerman pers. comm.). Irving (1980) notes Hedeoma h. var. hyssopifolium is common in mountains of central and southeastern Arizona and adjacent southwestern New Mexico and Sonora from 1800 to 3100 m where it is most frequent in recently burned areas. Hedeoma hyssopifolium var. hyssopifolium differs from the above in a wide series of characteristics involving habit, stems, inflorescence development and calyces. While the new variety is an erect-ascending plant 2.5 — 4(-6) dm tall that grows from crevices in cliffs from a distinct woody crown (fig. la), the type variety is a shorter (15 — 25 cm tall), more delicate, widely spreading plant with stems developing from a system of slender, rhizomes that grow in the forest detritus layer (fig. 1f). Stems are decumbent, often rooting at the nodes, more slender (0.5 — 0.8 mm in diameter), more even- Fig. 1. Hedeoma hyssopifolium. a—e.—H. h. var. chihuahuensts. a. Habit of type specimen showing woody, basal rootstock, erect-ascending leaves shorter than internodes, and dense cymes of flowers in upper stems. b. —Leaf, abaslal cae shoming ¥ venation an punctate glands. c. iugine of well branched infl eral proliferation. d.—S ca 1.0 mm in pi aee showing quaene use outline aes limited to decurrent lines below nodes er lobes. All from 1. M. Jobnston and C.H. peri 1432 (LL). f— j—H. pe var. bspifoliam Eg aaa note slender rhizomatous rootstalks, slender stems with internodes shorter than leaves and paired flowers at upper nodes. Plants are more extensive, with more stems than snow _ Dd. ‘ ee 529, (RS A). g. ee epee view showin venation and puget glad. f infl wered structure. 5 i.— Z mm in diameter showing rounded-q nature and trichomes n all surfaces. Diagram of calyx showing longer, distinctly ciliaee loge labes g—h from M. E. pen 4330 (TEX). Magnifications as indicated. Delination by Bobbi Angell 416 ly hirtellous to puberulent with straight or recurved hairs 0.06 — 0.15 mm long, more rounded-quandrangular (fig. li) rather than abruptly quadran- gular and petiolar margins continue as a distinct shallow rim across the nodes. Mid stem internodes in the type variety are only (3-)}6—11 mm long, shorter that the subtending, (7-)11— 18 mm long leaves. Flowers in the type variety are borne in 1(-3)-flowered, axillary, secund dichasia (fig 1h) while in the new variety dichasia more often contain (1-)3 — 7 flowers, and when lateral dichasial shoots proliferate in a monochasial pattern as many as 13 flowers may form per inflorescence (fig. 1c). Several conspicu- ous differences occur in the calyces. In the type variety the upper 3-calyx lobes are broad at the base and taper to slender tips, the lobes are distinctly recurved, and 0.8 — 1.2(-1.4) mm long (fig. 1j). In the new variety the lobes are more slender at the base, straight or only slightly upcurved, and only 0.5 — 0.9(-1.2) mm long (fig. le). Lower calyx lobes of the type varie- ty are straight, mostly 2.2 — 3.2 mm long, and rather conspicuously ciliate with hairs 0.1 —0.5 mm long (fig. 1j) while in the new variety they are only 1—2.2 mm long and obscurely ciliate (fig. le). Calyx annulus is slightly better developed in the type variety with the hairs measuring 0.6-—0.9 mm long but in both the hairs are exserted. Other minor differences occur in vestiture, with the new variety tending to be more glabrous. Irving (1980) in his excellent monograph of Hedeoma noted that popula- tions of eastern Chihuahua differed from western populations in their glabrous stems, subulate-filiform calyx teeth, and sparse calyx annulus and was perhaps deserving of varietal status. In describing the taxon for a mon- ograph the few disjunct eastern Chihuahuan specimens, while different, can be generally fitted within Hedeoma hyssopifolium with which they are most closely related. But when preparing a treatment for the Chihuahuan Desert flora the consistancy of the many differences become very apparent thus necessitating the nomenclatural recognition of these populations. ACKNOWLEDGEMENTS I thank M. C. Johnston (TEX) for the Latin diagnosis, Bobbi Angell (NY) for the illustration, Robert S. Irving for comments on the manu- script, and the University of Texas Plant Resources Center (TEX) for use of facilities. REFERENCES IRVING, R. S. 1980. The systematics of Hedeoma (Labiatae). Sida 8:218 — 295. SABAL ETONIA (PALMAE): SYSTEMATICS, DISTRIBUTION, ECOLOGY, AND COMPARISONS TO OTHER FLORIDA SCRUB ENDEMICS'! SCOTT ZONA Rancho Santa Ana Botanic Garden Claremont, CA 91711, U.S.A. WALTER S. JUDD Department of Botany, U se ue Florida Gainesville, FL 32611, U.S A taxonomic study of Sabal etonia Swingle ex Nash and related taxa (in- volving field work throughout Florida along with-the study of more than 800 herbarium specimens) has shown that this palm is morphologically and ecologically distinctive and most closely related to S. palmetto (Walt.) Lodd. ex J. A. & J. H. Schultes and S. mzamiensis Zona (see Zona 1983). Sabal etonia differs from the more widespread S. palmetto in its usually subterranean trunk (vs. erect and aerial), crown of usually only 3 — 5 leaves (vs. 14—40), narrower petioles (ca 0.6—2 vs. 2—4 cm), blades with fewer segments [20—46(-56) vs. 44—80(-90)}, shorter median leaf segments [32 —66(-69) vs. (55-)60—110(-120) cm], shorter hastulas {1—2.7 vs. (2.8-)3— 13.3 cm], inflorescence structure (2 orders vs. 3 orders of branching), larger fruits (diameter usually 12—15 vs 9 — 12 mm), and larger seeds (diameter of usually 8— 10 vs. 6-8 mm). Sabal miamiensis 1s intermediate between S. efonia and S. palmetto in most vegeta- tive characters, but has a subterranean trunk like the former and three branch orders in its inflorescences like the latter; its fruits and seeds are typically larger than either species (see Zona 1985, for a detailed discussion of S. mzamiensis). The morphological character most obviously distinguishing Sabal etonia from S. palmetto is acaulescence. Authors of recent treatments of the flora of Florida (Long & Lakela 1976, Wunderlin 1982) have placed great empha- sis on the above-ground trunk of S. pa/metto versus the subterranean trunk of S. etonia. Usually, the trunk of S. pa/metto is emergent and erect, but that of S. etonia is subterranean and sigmoid. An excellent illustration of the 'This paper is Florida Agricultural Experiment Station Journal Series No. 7060. SIDA 11(4):417 — 427. 1986. 418 peculiar trunk of S. etonza may be found in Bailey (1944). While these character states are typical for the species, they are by no means consistent. Occasionally, one encounters individuals of §. eonta with well developed emergent trunks. Bailey (1944) mentioned §. etonia in Marion County with six foot (ca 2 m) trunks; however, such individuals are encountered very infrequently in dry habitats (such as sand pine scrub). Individuals of S. etonia growing in more mesic habitats, such as those of coastal Volusia and Dade counties, also exhibit caulescence. These plants grow in what appears to be “mesic scrub,” a scrub in the process of succession toward a mesic hammock (Kurz 1942). The short-emergent trunks of these individuals may be the result of mesic edaphic conditions. There also exist individuals of S. palmetto with only shortly emergent trunks to | m tall. Such in- dividuals may be seen in dry coastal dunes such as those near Cedar Key (Levy Co.) or Merritt Island (Brevard Co.). There are also acaulescent to short-emergent plants on the Everglades pine keys of Dade County. These “stunted” plants also are likely the result of edaphic conditions. It appears that either excessively dry soil or a confined root system results in acaulescent or short-trunked §. pa/metto. The morphological plasticity of these species has not been fully appreciated by many taxonomists. The trunk of Sabal etonta 1s smooth, even on those plants with emergent trunks. The trunk of S. pa/metto may be mote or less smooth or clad with old leaf bases. The petioles, and consequently, the leaf bases are smaller in S. etonta as Compared with those of S. palmetto. As indicated above, the leaves of Saba/ etonia are smaller on average than those of S. palmetto: the hastula and petiole are narrower and shorter, and there are fewer, shorter and narrower segments. The lamina of S. efonia and of S. miamiensis is characteristically yellow-green, but that of S. pa/metto is slightly blue-green. The inflorescence structure is variable and of taxonomic importance in the genus Saba/. Inflorescences of the group are axillary and enclosed by a series of open-ended tubular bracts which protect the primary axis of the inflorescence. The inflorescence of §. efonia is nearly erect, but as the fruits develop, it frequently becomes prostrate on the ground from its own weight. The branches of the inflorescence of §. pa/metto are arrayed loosely along the main axis. There are three orders of branches (discounting main axis). In S. etonta the branches are crowded tightly on the main axis and are ascending, and there are only two orders of branches. Terminology used here, particularily that of axis enumeration, corresponds to Tomlinson & Zimmermann (1968). Floral morphology varies very little among these species, and the taxonomic usefulness of floral features is not great. The petals of Sabal 419 etonia average 3.1 mm long, while those of S. palmetto average slightly shorter (1.e., 2.9 mm), and those of S. mzamiensis slightly longer (i.e., 3.5 mm). However, fruit and seed size has proved to be of taxonomic value in the species complex. As indicated above, the fruits of S. etonza are interme- diate between those of S. palmetto and S. miamuensis. Sabal etonia is a member of the characteristic and highly endemic flora occurring on the white or yellow sands of the upland areas of the central “backbone” of the Florida peninsula. One of the most common plant com- munities in this region is the sand pine/xerophytic oak scrub (e.g., vegeta- tion dominated by Pinus clausa (Chapm. ex Engelm.) Vasey ex Sarg., Quercus geminata Small, Q. myrtifolia Willd., Q. inopina Ashe, and Q. chapmanii Sarg.). The species occurs in central Florida from Clay Co. south to Highlands Co. along the Trail Ridge, Mount Dora Ridge, and Lake Wales Ridge, and along the Atlantic Coastal Ridge from St. Lucie to Dade County (Figure 1). (See White 1970, for a summary of the geology of these ridges). In contrast, S. palmetto occurs from coastal North Carolina (Brun- swick Co.) south through South Carolina and Georgia to northeastern Flor- ida, throughout peninsular Florida, and in the Bahama Archipelago. This species shows a distinctive affinity for water. It is abundant in mesic to hydric hammocks, tidal flats, river banks, coastal strand and dunes, and pine flatwoods and savannas. The species is the canopy dominate in penin- sular Florida’s tidal flats just above the Juncus roemerianus Scheele zone. Sabal miamiensis ts limited to the pinelands (1.e., Pznus e//iottzt Engelm. var. densa Little & Dorman) occurring on the oolitic limestone of the Miami region. Thus S. etonza is typically ecologically isolated from both S. pa/metto and S. miamuiensis. The authors conclude that Saba/ etonia is clearly distinct from both S. palmetto and S. miamiensis in both morphology and ecology. This study has demonstrated the unreliability of cauline characters in this complex. The morphology of adult leaves, particularly leaf number, lamina color, seg- ment number, and hastula length, has proven taxonomically valuable. In- florescence structure is also an important and useful character, as is fruit and seed size. No one morphological character is reliable enough to sepa- rate the three species due to the amount of overlap in the range of variation, however, S. etonia is clearly delimited from related taxa when a suite of characters is used. A summary of the nomnenclatural synonymy along with a brief description and citation of representative specimens for S. etonta ts given below SABAL ETONIA Swingle ex Nash, Bull. Torrey Bot. Club 23:99— 100. 896. Tver: FLORIDA: vicinity Se Eustis, 16-30 Jun 1894, Nash 999 (HOLOTYPE: NY!; tsoryees: BH!, GH!, MO!, US!). Figure 1. The distribution of Sabal etonta. Sabal adansonii Guerns. vat. megacarpa Chapm. Fl. South. U.S., 2nd ed. 651. 1883. Sabal megacarpa (Chapm.) Small, Fl. SE. U.S. 223. 1903. Type: SOUTH FLORI- Garber 5.n. (LECTOTYPE: MO)!). Plants usually acaulescent, rarely caulescent to 2 m, stems sigmoid, to 10— 15 cm in diameter, or if upright, 15—20 cm in diameter. Leaves 3 —5, usually yellow-green, costapalmate; petiole 23 —55 cm long, 0.6— 2(-2.1) cm wide; hastula triangular, 1— 2.7 cm long, segments 20 — 46(- 56), bifid, filiferous, 32—66(-69) cm long, 1.5—3 cm wide. In- florescence paniculate, densely branched with two orders of branching (discounting main axis), upright at first chen prostrate as fruits develop. Flowers subsessile, perfect, white, sweetly fragrant, protogynous,; calyx three-lobed, 1— 1.4 mm long, cup-shaped; petals three, 3— 3.2 mm 421 long, ovate; stamens 6, the same length as the petals, connate slightly at the base and basally adnate to the petals; gynoecium composed of three fused carpels, 2.5 — 3.5 mm long; ovary superior, stigma papillose. Fruit a one-seeded berry developing from one carpel (rarely more than one carpel develops, the fruit is then a strongly lobed two- or three-seeded berry), globose, shiny black, (1 1-)12 — 15 mm in diameter with a fleshy pericarp; seed oblate, concave on the funicular end, brown, (6-)8 — 10(-11) mm in diameter, embryo sublateral, endosperm bony. (Figure 2, see also Bailey 1944, Small 1925). Distribution and Ecology: Florida, from Clay Co. to Highlands Co. in the sand pine scrub of the Central Florida Ridge, also in the scrub on older coastal dunes of Volusia, St. Lucie, Palm Beach, Broward, and Dade coun- ties, and in isolated patches of scrub in DeSoto, Hernando, Manatee, Okeechobee, and Seminole counties (Figure 1). Associated species are dis- cussed by Harper (1914, 1915, 1927), Kurz (1942), and Mulvania (1931). Flowering occurs from late May through July. Representative Specimens: UNITED STATES: Floripa. Broward Co.: Fr. Lauderdale, pee 473 (BH). Clay Co.: Goldhead Branch State Park, Skean 850 (FLAS), Ward 5490 (FL Dade Co.: Miami, Bailey 6472 (BH); North Miami, Zona 68 (FLAS). De Soto Co.: west of Horse Creek, Shuey 1853 (USF). Highlands Co.: Lake Placid, Brass 33413 (USP): near Sebring, Judd 2498 (FLAS); just west of nie Jackson, oe 28406 (FLAS); without definite locality, Smal/ 11572 (NY); off U.S. Rt. 27, Zona 10 (FLAS); south of reat Creek, Zona 60 (FLAS). Lake Co.: near Eustis, nS 975 an Savage 79 (BH). n Co.: near Salt Springs, Judd 2774 (FLAS); Ocala National Forest, Wunderlin, et al. rey (USF), Zona | (FLAS). Okeechobee Co.: near Okeechobee City, Bailey & Small 6211 sie Osceola Co.: east of ae Lake, Shuey 5.n. 27 Oct 1974 (USF). Palm Beach .: Jupiter, O’Ne7// 5.n. (FLAS, US); Boca Raton, Zona 63 (FLAS); Boynton Beach, Zona i (FLAS). Polk Co.: near Beorook Judd 2840 (FLAS), Zona 3 (FLAS). Putnam Co.: orth of Florahome, W7/son s.n. 16 May 1959 (FLAS). Volusia Co.: near Ormond Beach, a 31 (FLAS), Zona 60 (FLAS). Economic Uses: Sabal etonia is of potential ornamental use in areas where the edaphic conditions would permit its growth. The sand pine/xerophytic oak scrub of the Central Florida Ridge supports many endemic species (Harper 1949, James 1961, Ward 1979). Sabal etonia is almost always found wherever there is scrub, and the taxono- my and distribution of this species may provide some insight into the probable origin of this and other scrub endemics. Noteworthy endemics of this region include: Asclepias curtissii A. Gray, Bonamia grandiflora (A. Gray) Heller, Bumelia lacuum Small, Calamintha ashei (Weatherby) Shinners, Carya floridana Sarg., Chapmannia floridana Torr. & Gray, Chionanthus pygmaeus Small, Clitoria fragrans Small, Conradina brevifolia Shinners, Dicerandra frutescens Shinners, Eriogonum floridanum Smal 422 longifolium Nutt. var. gnaphalifolium Gandoger), Eryngium cuneifolinm Small, Garberia heterophylla (Bart.) Mert. & Harper, Hypericum cumulicola (Small) P Adams, H. edisontanum (Small) PR Adams & Robson, I/ex opaca Ait. var. arenicola (Ashe) Ashe, Lechea cernua Small, Liatris ohlingerae (Blake) Robins., Lupinus aridorum McFarlin ex Beckner, L. cumulicola Small, Nolina brittoniana Nash, Osmanthus megacarpus (Small) Small ex Little, Palafoxia feayi Gray, Paronychia chartacea Fern., Persea humilis Nash, Pityopsis graminifolia (Michx.) Nutt. var. aequilifolia Bowen & Semple, Polygala lewtonii Small, Polygonella bastramia (Small) Nesom & Bates, P. robusta (Small) Nesom & Bates, P. myriophylla (Small) Horton, Prunus geniculata Harper, Quercus tnopina Ashe, Sabal etonia Swingle ex Nash, Sisyrinchium xerophyllum Greene, Stylisma abdita Myint, Warea amplexifolia (Nutt.) Nutt., W. carter? Small, and Ziziphus celtata Judd & Hall. An understanding of the origin of many of the scrub endemics is aided through a study of the distribution and morphological variation of the vart- ous endemics and their close relatives, an appreciation of past geological events in the area of endemism, and a knowledge of the region’s edaphic and climatic conditions [see discussion of endemism in Stebbins (1942), Mason (1946), and Woodson (1947)}. The Central Highlands of the Florida peninsula, as they are called by Cooke (1945), are a series of disconnected ridges which may represent the remnants of a single ridge extending through Florida from Baker to Highlands County (White 1970). This geomorphological feature is now seen as a series of separate smaller ridges including the Trail Ridge to the north in Baker, Bradford, and Clay counties; the Mount Dora Ridge ex- tending through eastern Marion, Lake, and Orange counties; and, by far the largest of the individual ridges, the Lake Wales Ridge of Lake, Orange, Osceola, Polk, and Highlands counties. McNeil (1949) showed that much of the land that is now the Trail Ridge and Lake Wales Ridge was exposed even during the Pleistocene interglacial periods when the sea level was con- siderably higher than its present level. In the Pliocene, portions of the Florida-Georgia border and probably portions of the ridges were above sea level (White 1970). Another ridge which ts relevant to this discussion is the Atlantic Coastal Ridge which extends down the entire Atlantic coast but supports scrub only in its higher areas such as those found in Volusia, St. Lucie, Palm Beach, Broward, and Dade counties. In its southerly parts the Atlantic Coastal Ridge was formed from sand dune deposits overlaid on the Miami Ridge, a calcareous bar formed in Pamlico times (100,000 B.P.) and was submerged until very recently (White 1970). The areas of Florida that were exposed in the Pleistocene presumably were suitable for plant habitation. Watts (1969) noted that sclerophyllous oak pollen is well 423 Figure 2. Habit of Saba/ etonia,; photograph taken in Highlands Co., Fla. represented in Pleistocene lake sediment in Marion County. Discon- tinuities in sedimentation suggest periods of time when the environment was drier than it is now. In addition, Watts (1975) reported that the fossil pollen record from a lake in Highlands County indicated a dominate scrub community in the Pleistocene. The genus Saba/ is well represented in fossil deposits in the southeastern United States (Daghlian 1978, Moore 1973, Read & Hickey 1974). Many of these deposits predate the formation of the land masses in Florida. The fossil record of Sabal suggests that it has long occurred in North America, and a refugium, similar to that suggested by Woodson (1947), may have been instrumental in the speciation of the genus in Florida. Moore and Uhl (1982) stated that dwarfism in palms ts an evolutionarily advanced condi- tion, so it is likely that S. etonia evolved from caulescent ancestors that colonized the ridges of Florida as they formed. Sabal etonia now inhabits some of the geologically oldest formations in Florida. However, the Atlan- tic Coastal Ridge is geologically recent and has been colonized successfully by S. etonta. Thus S. etonia is likely capable of invading new scrub habitats as they form. 424 Mason (1946) stated that the area occupied by any species is determined by environmental conditions and that, of the many factors contributing to the overall environment, only edaphic factors occur in sharply defined, of- ten small areas. The soil of the central ridges of peninsular Florida is clearly extreme; Mulvania (1931) described the sand pine scrub inhabitants as “rooted in a bed of silica, to which the term soil is but remotely applica- ble.” The soil underlying the scrub is a coarse white or yellow silica sand that extends to a depth of ca 3.5 — 4.5 m, and the scrub vegetation, includ- ing Sabal etonia, is found almost exclusively on these xeric sands. This specificity suggests that the distribution of S. efonia is the result of the limited occurrence of xeric white sands, and the species may have developed through selection in response to the extreme xeric environment of these sandy ridges. Stebbins (1942) maintained that species with a great deal of genetic variability within populations are more likely to spread over large areas and many habitats than are species whose populations are genetically homoge- neous. It is of interest that Saba/ etonia is morphologically uniform, espe- cially when compared with S. pa/metto, a species found in many habitats throughout the southeastern United States and the Bahamas (Brown 1976, Zona 1983). Furthermore, Stebbins (1942) defined two different kinds of rare genetically homogeneous species: the “depleted species” and the “insu- lar species.” Depleted species are those which were once more common but their present rarity is due to depletion of the store of genetic variability. An insular species, which need not be found only on islands, is one which was never common but has diverged from a more widespread ancestor and has become established in a small insular or isolated area. Stebbins suggested that if the endemic is closely related to a widespread species, which occurs on an adjacent continental area, it is probably a strictly insular species, but if it is closely related to no other living species, or has its close relatives ina geographically distant region, it is more likely a depleted species. A second criterion given by Stebbins (1942) is that an insular species is typically morphologically and/or ecologically specialized in relation to its continen- tal relatives. According to these guidelines Saba/ etonia is clearly an insular endemic, and likely evolved from mesophytic, more widespread, S. pa/met- to-like ancestors. Many of the other characteristic scrub species also appear to be insular, e.g., Asclepias curtissii (related to A. purpurascens, Woodson 1954), Chionanthus pygmaeus Small (C. virginicus L., Hardin 1974), Ilex opaca var. arenicola (Ashe) Ashe (I. opaca var. opaca, Wunderlin 1982), Osmanthus megacarpa (O. americana (L.) Benth. & Hook. f. ex A. Gray, Hardin 1974), Persea humilis Nash (P. borbonia (L.) Spreng., Wofford 1973), Polygonella basiramia (P. ciliata Meisn., Nesom & Bates 1984), P 425 robusta (P. fimbriata (Ell.) Horton, Nesom & Bates 1984), Prunus geniculata (P. angustifolia Marsh., Harper 1911), Quercus inopina (Q. myrtifolia Willd., Johnson & Abrahamson 1982). Sabal etonia is typical of these endemics. The ancestral species presumably inhabited the southeastern United States, and S. etonia shows the morphological specializations so typical of scrub endemics: dwarf stature, small leaves, large fruits, and large seeds (see Hardin 1974). These morphological features are likely adaptations to the xeric conditions of the sand pine scrub. Isolation, both in past refugia of emergent land masses and present-day “ecological islands,” has preserved the genetically divergent biotypes of the scrub endemics. How- ever, the floristic affinities of the scrub endemics are complex, and some [e.g., Ziziphus celata and Nolina brittoniana (Judd & Hall 1984), Eriogonum longifolium var. gnaphalifolium (Horton 1972), Bonamia grandiflora (Myint & Ward 1968), Palafoxia feayi (Turner & Morris 1976), and Carya floridana (Hardin & Stone 1984)}, appear to have western affinities. Additional sys- tematic studies of the endemic species of the Central Florida Ridge are urgently needed. It is the unique combination of historical and edaphic features which makes the sand pine scrub habitat so rich in endemic species. The Florida peninsula, by virture of its absence of glaciation, recent geological activity characterized by many fluctuations in its coastline, relative isolation from the remainder of the continent, proximity to the species-rich tropics, and unique climatic patterns, is an area which supports many endemics. Penin- sular Florida is also being subjected to rapid urban and agricultural devel- opment, and the natural plant communities of the Central Florida Ridge are very rapidly being destroyed. At this time less than 3% of the total land area of the southern Lake Wales Ridge is currently protected from develop- ment (Peroni & Abrahamson 1985). For these reasons, action to preserve the distinctive flora of the Central Florida Ridge is essential. ACKNOWLEDGMENTS We wish to thank the curators of the herbaria(AUA, BH, E FLAS, FSU, GA, GH, MO, NCSC, US, USCH, and USF) from which specimens were borrowed for this study. Dr. Robert W. Read of the Smithsonian Institu- tion, Washington, D.C., deserves special thanks for his helpfulness. We thank Dr. David W. Hall and Dr. Robert E Thorne for their valuable suggestions concerning the manuscript, and Kent D. Perkins for his assist- ance in the processing of specimen loans. REFERENCES BAILEY, L. H. 1944. Revision of the American palmettoes. Gentes Herb. 6:365 — 459. 426 BROWN, K. E. 1976. Ecological Se . the cabbage palm, Sabal palmetto. Principes 200 10, Le 56, 98-115, COOKE, C 1945. The ae : Ror, Florida Geol. Surv. Bull. 29:1 — 339. DAGHLIAN, = P 1978. Coryphoid palms from the Lower and Middle Eocene of southeastern North America. Lani eet Abt. B, Palaeophytol. 166:44 — 82. HAR me ae bi 1974. Studies in southeastern United States flora. IV. Oleaceae. Sida 5:27 aon i 7 & D. E. STONE. 1984. Atlas of foliar surface features in woody plants, VI. Carya Juglandaceae) of North America. Brittonia 36:140 — 153. HARPER, R. M. 1911. A new plum from the lake region of Florida. Torreya 11:64 — 67. —______. 1914. Geography and vegetation of northern Florida. Florida State Geol. Surv. Ann. Rep, 6: 163 — 437. 1915. aaa types, 7 E. H. Sellards, R. M. Harper, C. N. Mooney, W. J. cee H. Gun E. Gunter. Natural resources of an area in central Blotida, including a part i ‘Marion, Levy, Citrus, and Sumter counties. Florida State Geol. Surv. Ann. Rep. 7:135— 188 1927. Netucal resources soesouchetn Florida. Florida State Geol. Surv. Ann. Rep. isi) 182, ——___—. 1949. A preliminary list of endemics. Florida Acad. Sci. Quart. J. Lio 34, 39 57 HORTON, J. H. 1972. Studies of the southeastern United States flora. IV. Polygonaceae. J. Elisha Mitchell Sci. Soc. 88:92 — 102. JAMES, C. W. 1961. Endemism in Florida. Brittonia 13:225 — 244. JOHNSON, A. E & W. G. ABRAHAMSON. 1982. Quercus is a species to be recog- nized from south-central Florida. Bull. Torrey Bor. Club 109:392 — 395. JUDD, W. J. & D. W. HALL. 1984. A new species of Ziziphus ah amnaceae) from Florida. Rhodora 86:381 — 387 KURZ, H. ae Blerida: dunes and scrub, vegetation and geology. Florida Geol. Surv. Bull. 23:1— 149. LONG, R. W., & O. LAKELA. 1976. A flora of tropical Florida. 962 pp. Banyan Books, Miami, FL. MASON, H. L. 1946. Edaphic factors in endemism. Madrono 8:209 — 226. MCNEIL, E . 1949. Pleistocene shore lines in Florida and Georgia. U.S. Geol. Surv. Prof. MOORE, H. E., JR. 1973. The major groups of palms and their distribution. Gentes Herb. 11:27- 141. MOORE, H.E., JR., & N. W. UHL. 1982. Major trends of evolution in palms. Bot. Rev. 48:1—69. MULVANIA, M. 1931. Ecological survey of a Florida scrub. Ecology 12:528— 540. MYINT, T. & D. B. WARD. 1968. A taxonomic revision of the genus Bonamia (Con- ala ia Phycologia - L21— 239. , ESOM, G. & BATES. 1984. as ng ations of infraspecific taxonomy in Pui Rene Brittonia 36:34 PERONI, P A., & W. G. ABRAH a oa. Veseucion loss on the souchern Lake Wales | The Palmetto 5:6 — READ, R ,&L. J. HICKEY. on A revised classification of fossil palms and palm- like ees. Taxon 21:129— 1. SMALL, J. 1925. The ea ieee etonia. J. New York Bot. Gard. 20: 145.= 7 427 STEBBINS, G. L. 1942. A genetic approach to the problem of endemics. Madrono 6:241—258 TOMLINSON, P B. & M. H. ZIMMERMANN. 1968. Anatomy of the palm Rhapis excelsa, V. ape ere J. Arnold Arbor. 49:29 1 — 306. TURNER, B. L. & M. I. MORRIS. 1976. Systematics of Palafoxia (Asteraceae: Helenieae). Rhodora 78:567 — 628. WARD, D. B. ‘1979. Introduction, pp. x — xix 7 D. B. Ward, biota of Florida, Vol. 5. Plants. 175 pp. Univ. Presses of Florida, Gainesville ATTS, W. A. 1969. A pollen ee from ae Lake, Marion County, north central Florida. Geol) Soc. America Bull. 80:631 pa ee eeeer Oi ye Nelace: aera record of vegetation from Lake Annie, south-central een . Geology 3:344—3 WHITE 1970. The ae eee of the Florida peninsula. Florida Geol. Surv. ed. Rare and endangered FL WOFFORD, B. E. 1973. A biosystematic study of the genus Persea (Lauraceae) in the southeastern United States. Ph.D. dissertation. 160 pp. Univ. of Tennessee, Knoxville, TN. WOODSON, R. E. 1947. Notes on the “Historical Factor” in plant geography. Contr. Gray Herb. 165:12— 25. 1954. The North American species of Ascfepias. Ann. Missouri Bot. Gard. A lieth 21 ls, WUNDERLIN, R. P. aon Guide to the vascular plants of central Florida. 472 pp. Univ. Presses of Florida, Tampa, PL. ZONA, S. 1983. A taxonomic study of the Sabal palmetto complex (Palmae) in Florida. M.S. ane 88 pp. Univ. of Florida, Gainesville, FL. . A new species of Saba/ (Palmae) from Florida. Brittonia 37:366 — 368. NOTES ABOUT PSORALEA SENSU AUCT., AMORPHA, BAPTISIA, SESBANIA AND CHAMAECRISTA (LEGUMINOSAE) IN THE SOUTHEASTERN UNITED STATES' DUANE ISELY Departments : Botany and of Plant Pathology, Seed and Science, lowa State University a TA, 50011, U.S.A ABSTRACT The classic Psoralea is presented as vase Pedtomelum and Psoralidium: Pediomelum digitatum (Nutt. ex T. & G.) comb. ; Orbexilum pedunculatum var. eglandu- losum (Elliott) comb. nov.; ioe pace (Michx.) comb. nov. Amorpha essen- tially follows Wilbur's monograph: Amorpha herbacea Walter var. crenulata (Rydberg) comb. nov. Baptisia lactea (Raf.) Thieret (B. leucantha T. & G.) and B. alba (L.) Vent. of prior literature must change names; Baptisia alba var. macrophylla (Larisey) comb. nov Sesbania concerns the delimitation of Sesbania macrocarpa and S. emerus (Aubl.) Urban, and the distribution of S. virgata (Cav.) Poir. in the United States. An author alteration for Chamaecrista nictitans vat. aspera is noted. This is one of several contributions in which rationale for floristic treat- ment of the Leguminosae for the Vascular Flora of the Southeastern United States (University of North Carolina) is presented. GENERIC DELIMITATION IN THE PSORALEEAE! ORBEXILUM, PEDIOMELUM AND PSORALIDIUM Except for Rydberg (e.g. 1919— 1920, and 1928) and the few other authors who followed him in floristic treatments (notably Small 1933), American authors have maintained a traditional, diverse Psoralea without generic segregation. Rydberg raised the traditional Psoraleeae (spelled Psoraleae by authors prior to correction by Barneby 1977) from subtribal to tribal status and divided North American Psoralea into seven segregate genera. Isely (1962) accepted the first of Rydberg’s premises, but not the second in that he maintained the U.S. species within the confines of a single genus. He felt that Rydberg’s segregates represented, at least in part, natural groups but, because of the reticulate nature of variability, was unsuccessful in delimiting coherent sets on a multiple character basis. ‘Journal Paper No. 12021 of the Iowa Agriculture and Home Economic Experiment Station, Ames, [A 50011 SIDA 11(4):429 — 440. 1986. 430 Since that time, the Psoraleeae has been segregated into two tribes, the Amorpheae and Psoraleeae (Barneby 1977, Stirton 1981). Stirton also divided the Old World representatives of the classic Psoralea into some five genera, Psoralea in this restricted sense being a small (ca 20 species) homo- geneous group limited to South Africa. The circumscription of Psoralea is accepted (I do not pass judgment on the other Eurasian-African segre- gates), and therefore that name is not available for any American species. Were American “Psoralea” yet maintained as a single genus the available generic name would be Orbexi/um. Because of the evident diversity of the North American Psoraleeae, and in connection with impending decisions for the SE Vascular Flora I have reexamined the problem with the hypothesis that fruit-calyx features (rath- er than foliage divergence for example) are the best markers for evolu- tionarily segregate groups and arrived at a classification that works at least for the region concerned. In stripped down key form, it is as follows. . Legume enclosed in the enlarging calyx except for the long, projecting beak; percatp thin, Commonly papery oasis s4 645 been oe ane eadese dd Pediomelum |. Legume exserted above calyx remnants, shortly beaked; pericarp thick and coriaceous. 2. Legume cross tupose, eelandulet cs cen ecw du ba oe yale Orbexilum 2. Legume not cross-rugose, conspicuously glandular. ............... Psoralidium This presentation is the same as Rydberg’s except that his Rhytidomene is included in Orbexi/um and (among southeast species) Psoralidinm digitatum (that has enlarging calyx and long beak) is transferred to Pedsomelum. Because I have conducted no critical phylogenetic study of the American Psoraleeae as a whole, this possibly should be regarded as a working ar- rangement for the immediate purpose of a coherent floristic interpretation. However, a study of New World Psoraleeae, comparable to that of Stirton for Africa-Eurasia (Ca//en to Australia), is presently underway by James Grimes of the University of Texas. The above generic segregation, within its limited context, seems approximately equivalent to his concepts. Three new combinations resulting from these taxonomic decisions are listed under the subject genera. PEDIOMELUM RypserG PeDIOMELUM digitatum (Nutt. ex T. & G.) Isely, comb. nov. Psoralea digitata Nute. ex T. & G., Fl. N. Amer. 1:300. xT. G.) Rydb., N. Amer. FI. 24:16. 1919. _ 838; Psoralidinm digitatum (Nutt. ex Although this species indeed superficially resembles some of Psorali- dium, e.g., P. tenuiflorum (Pursh) Rydb., it has the fruit of Pediomelum. 43] ORBEXILUM RafinesQue ORBEXILUM PEDUNCULATUM (Miller) Rydberg Of the two varieties listed following, var. pedunculatum, the eastern Coastal Plain form, is usually glandular on all plant parts, conspicuously so on bracts, and the undersides of leaflets. Var. eglandulosum, lacking the glands or nearly so, is widely distributed from Texas and Oklahoma east into the Appalachians. Varietal segregation is not entirely “air-tight”; there is some intermediacy in the mountains and a few weakly glandular forms may be encountered farther west. But most material is clearly one or the other and it seems worthwhile to retain this now traditional varietal bifurcation despite, perhaps, its rather trivial nature. OrBEXILUM PEDUNCULATUM (Miller) Rydb. var. pEDUNCULATUM, N. ges Fl. 24:7. 1919. Hedysarum pedunculatum Miller, Gard. Dict. ed. 8. No. 1768; non Psoralea pedunculata (Pursh) Poir. (1816) nec P. pedunculata Ker aac Nee ee Walt., Fl. Carl. 184. 1788; Psoralea psoralioides (Walt.) Cory, 8:406. 19306. As the above citations indicate, the familiar Psoralea psoralioides 1s lost on transfer to Orbexilum where the earlier epithet must be taken up. That the Hedysarum pedunculatum Miller is this species was verified for Freeman by C. . Weatherby who, however, felt that it was a mixture of the two varieties (Freeman 1937). Dr. C. E. Jarvis of the British Museum has kindly sent me a photocopy of the sheet of the Miller material that Weatherby examined. It includes the apices of two separate flowering stems that might or might not have come from the same plant. Both are easily, as Weatherby asserted, the now tradi- tional Psoralea psoraltoides. Dr. Jarvis kindly also examined the specimens and says “There are indeed numerous dark glands on the lower surfaces of the leaflets and bracts” thus contradicting Weatherby’s assertion that the collection(s) includes a mixture of the two conventional varietal types. While these specimens match the protologue, the secondary observa- tions, i.e. “The seventeenth sort . ” starts with the same subject but trails off into the statement that the flowers “are succeeded by jointed pods, straight on one side.” In view of the fact that the specimen and diagnosis are confirmatory, it is reasonable to assume that Miller, perhaps writing his commentary at a later date, became confused about his subject, possibly then thinking of a Desmodium. Dr. Jarvis has also noted that the specimen “bears on the verso the legend ‘America septenrionalis’ and was accordingly filed amongst the North American rather than amongst the cultivated material in our herbarium. 432 Miller clearly had material in cultivation, but I would not like to say whether this sheet was of cultivated or wild origin.” Because of this ambi- guity, I hesitate to designate the sheet as holotypic although I think it would be expedient to regard it in this light. The British Museum also has a specimen in the Dale herbarium, presu- mably collected by Thomas Dale in South Carolina in 1730. It is almost an exact match for the Miller specimen. ORBEXILUM PEDUNCULATUM vat. ee (Elliott) Isely, comb. Psoralea eglandulosa Elliott, Sketch Bot. S. Carolina 2:198. 1822; Psoralea nibnalioede var. eglandulosa (Elliott) Freeman, athe. 39:426. 1937. Freeman (1937) made no reference to Elliott’s original material, and I have not had an opportunity to see it (if extant). Identification, however, seems evident from the protologue. OrBEXILUM lupinellum (Michx.) Isely, comb. nov. Psoralea cg Michx., Fl. Bor. Amer. 2:58. 1803; Rhytidomene lupinellus (Michx.) Ryd BL. 24:12. 1919. Giedin though predominantly of species with pinnately trifoliolate leaves, as O. pedunculatum and 0. simplex (Nutt. ex T. & G.) Rydb., includes O. lupinellum with linear-filiform, palmately foliolate leaves and O. virgatum (Nutt.) Rydb. with simple leaves. They all have the same fruit type. Michaux’s protologue of Psoralea lupinella is explicit as to the identity of his material. AMORPHA L. I follow Wilbur's admirable revision (1964, 1975) of an amorphous ge- nus except for reduction of a couple of taxa, noted following, in the dwarf A. herbacea complex. The major problems, however, are not with the so- called dwarf group. They are instead with the segregation of A. fruticosa and its immediate relatives. These include all of the remaining species save A. schwerinit Schneider, A. paniculata T. & G. and A. californica Nutt. Herein the omnipresent A. /ruticosa not only overlaps with the others in its geographic range, but also in its plethora of variability (in pubescence, number of leaflets, the level and type of exsertion of the leaflet midrib, calyx lobe length, level of glandular development on all plant parts). I follow Wilbur because I believe the species that he delineated from the overlying blanket of A. fruticosa represent real entities, although it is possible that some of the southwestern (e.g. Texas, Arkansas) segregates should be considered peripheral varieties. Also, among those I have recent- ly studied, I find that I cannot always confidently distinguish some 433 specimens, e.g. of A. glabra Poir., A. nitens Boynton, and A. ouachitensis Wilbur from A. fruticosa. | believe that the problem is not hybridization, rather that the evident exomorphic features are less than consistently diag- nostic, and that perhaps we have yet to discern characters that clearly dif- ferentiate the taxa. AMORPHA HERBACEA Walter and A. CRENULATA Rydb. I confirm Wilbur's lucid (1964, 1975) characterization of geographic- morphological variation in this complex that includes the relatively wide- ranging and variable Amorpha herbacea, North Carolina to Lake Co., Florida and A. crenulata, a monotype in Dade Co., Florida. Amorpha herbacea is normally conspicuously pubescent, but the typical form fades to thinly hairy in various parts of the range, and to glabrate in two disjunct areas in Florida, specifically Hillsborough and contiguous counties, and separately in Franklin and Wakulla counties in the Panhan- dle. The glabrate forms are the basis of var. floridana (Rydb.) Wilbur. Be- cause pubescence is a quantitative feature that is regionally variable, | prefer to regard the glabrous forms as local extremes and withdraw var. floridana from nomenclatural listing. The name is, of course, available for those who wish to call attention to glabrate forms. Amorpha crenulata, exclusively of Dade Co., Florida, differs, in Wilbur's analysis, from the glabrate forms of A. erbacea in that it 1s usually white- flowered and that the revolute margins of its leaflets are slightly crenulate. But I have seen white-flowered forms of A. erbacea, and the flowers of A. crenulata (as to herbarium labels) may range to pale lavender. Wilbur (1964) called A. crenulata a “weakly differentiated species,” and I reduce it to a slightly isolated variant of A. herbacea as follows. Plants conspicuously pubescent to almost glabrous; leaflets usually 1.8 —3 times as long as wide, entire or subcrenulate, petioluled 1— 1.5 mm, flowers blue-purple (whit) 2.22 «nsdn Gys £4 be Sa Gaee pee var. herbacea Plants esas leaflets most i — y 2.8—3.5 times as long as wide, usually crenulate along the incurved (revolute) margin, usually Sie 1.5—2 mm; flowers white (-lavender) ........0.......00. var. crenulata AMORPHA HERBACEA Walter var. ieee aa Isely, comb. nov. A. crenulata Rydb., N. Amer. Fl. 24:3 BAPTISIA VENTENAT BAPTISIA LEUCANTHA — LACTEA — ALBA SEQUENCE. The relatively common Baptisia with large, white flowers of the central states has traditionally been known as B. /ewcantha T. & G. The similar eastern plants were referred to the same species by Small (1933), but were 434 mostly relegated to B. pendula Larisey and B. psammophila Larisey in Larisey’s (1940) monograph of the genus. Thieret (1969) identified Rafinesque’s Dolichos lactens with B. leacantha and took up B. /actea (Raf.) Thieret for the subject species. His determination was confirmed by Isely (1981) with reasonable assurance, and B. /eucantha, therefore, was relegat- ed to synonymy. The name Baptisia alba (L.) Vent. has traditionally been applied to the eastern (primarily North Carolina to Georgia) white-flowered Baptisia that, though with entirely different fruits, considerably resembles B. leucantha (lactea) in flower except that the corolla is usually smaller. The identity of the Linnaean basionym (Crotalaria alba) has been assumed rath- er than definitively identified and Isely remarked (1981, p. 219) “Crotalaria alba.... traces to ‘Hort Cliff 499’ and the associated specimen (BM). The specimen in LINN (microfiche!) marked by Linnaeus as ‘a/ba’ lacks fruit and could be Baptisia lactea.” But Isely had no opportunity to see the Hortus Cliffortianus material and indeed was happy to leave the refer- ence of B. a/ba as it has been for over 150 years. That is no longer possible. For Turner (1982), in a critique of Isely’s treatment, wrote that he had had Opportunity with Stearn to examine the Hort. Cliff. specimen at the British Museum and that it was indeed the species that had been called B. leucantha and subsequently B. /actea. The consequences of the Turner- Stearn identification, sadly, result in a further scrambling of names for both of the white-flowered species. B. /ewcantha, recently to B. /actea, now be- comes B. a/ba (L.) Vent. and the shift follows to the varietal names for the eastern and western components of that species because the Linnaean type (Habitat Carolina) is of the eastern rather than the western variety. And the prior B. alba of all U.S. treatments must become B. a/bescens Small. In the following listing only the names mentioned above and essential synonyms are cited; complete synonymy is given in Isely (1981). BapTISIA ALBESCENS Small, Fl. SE. U.S. 600, 1331. 1903. } — Raf., New Fl. N. Amer. 2:47. 1837 sensu Merrill (1949). . alba sensu auct. pl. Merrill (1949) identified the Rafinesque name with Baptisia albescens (as B. alba) but this determination is patently unwarranted because Rafinesque described the pods as obovate. Baptisia albescens has cylindric pods while those of B. alba (=B. leucantha, lactea) are obovate. Further- more the range given, “Carolina to Alabama and Louisiana,” is impossible because B. albescens is exclusively an eastern species. Granting that Rafinesque might have had a mixture of the two, the reference to B. albescens is rejected. 435 BaprtisiaA ALBA (L.) Vent., Dec. Gen. Nov. 9. 1808. Crotalaria alba L., Sp. PI. 716. 1753 Dolichos lactews Raf., Fl. Lud. 102. 1817; B. he (Raf.) Thieret, Sida 3:496. 1969. Baptisia leucantha T. & G., Fl. N. Amer. 1:385. 1840 The identity of Crotalaria alba has been discussed in foregoing text. Eastern and western varieties of B. a/ba may be distinguished as follows: Legume usually 1.5 —2(-3) cm in diam., thin-walled and brittle (-thick- walled); eastern U.S.: North Carolina, south to northern Florida, west to Alaa eae eee ec ae asa, Sg 4 hae at, ado & yaoi an ne. Ne ee acer a ie var. alba gis aly 0.9 — 1.2(-1.5) cm in diam., rigidly coriaceous; central U.S.: Mississippi to eastern Texas, north to Minnesota, Wisconsin and Ohio var. macrophylla BAPTISIA ALBA Var. ALBA B. leucantha 'T. & G., Fl. N. Amer. 1:385. 1840 sensu authors In part. B. pendula Larisey, Ann. Missouri Bot. Gard. 27:170, B. psammophila Larisey, Ann. Missouri Bot. Gard. 27:180. 1940. B. pendula Larisey var. obovata Larisey, Ann. Missouri Bot. Gard. 27:171. 1940; B lactea Rat. var. obovata (Larisey) Isely, anne 30: 471. 1978. BapTisia ALBA var. macrophylla (Larisey) Isely, comb. nov. B. pendula var. macrophylla Larisey, Ann. Missouri Bot. Gard. 27:172. 1940 as to type, not Georgia citations. leucantha TV. & G., Fl. N. Amer. 1:385. 1840. lactea (Raf.) Thieret, Sida 3:446. 1969 as to var. Jactea. mo SESBANIA Scopo.t This genus has been known under two similar names: Sesban Adanson, Fam. 2:327, 604. 1763 and Seshania Scopoli, Introd. 308. 1777. The fact that they were once considered orthographic variants, but are now treated as different names, Seshania being conserved over the earlier Sesban, has produced some interpretational problems in author citation of several species that are listed in both genera. SESBANIA MACROCARPA Muhl., Cat. 65. 1813; also Muhl. ex nee Gen. 2:112. 1818; also Muhl. ex Ellioce, Skecch Bot. S. Carolina 2:221. Darwinia es Raf., Fl. Ludoviciana 106. 1817; Seshan exaltatus (Raf.) Rydb., N. Amer. F 204. 1924, Seshania exaltata (Raf.) Hill, Index Kewensis, Suppl. 7 223, 1929 a authors); Seshania exaltata (Raf.) Cory, Rhodora 38:406. 1936. Sesbania macrocarpa is an abundant weedy species that ranges in the southern United States from Florida to California. In vegetative condition and flower, §. macrocarpa resembles Glottidinm vesicarium (Jacquin) Harper (this species is commonly treated as a Sesbania), which ordinarily has fewer 436 leaflets, considerably smaller flowers and a calyx that is but slightly toothed. It is easily known in fruit because, except for S. emerus (Aubl.) Urban and S. sericea (Willd.) Link of subtropical Florida, the linear pods are essentially unique among our herbaceous legumes. The S. macrocarpa- emerus problem will be discussed under the latter species following. The flowers of S. macrocarpa are ordinarily yellow or mottled, but there is a race in western peninsular Florida (Bay and Santa Rosa counties) that has a conspicuously dark red standard. Merrill and Hu (1949) regarded the Muhlenberg entry for Seshania macrocarpa aS a nom. nud. and attributed first validation of the name to Elliott in 1822. Consequently the synonym S. exa/tata has been taken up y many recent authors. Muhlenberg (loc. cit.) listed two species under Seshania as follows: tubr. 1. platycarpa broad-podded lut. 2. macrocarpa long-podded These statements, contrasting both flower color and pod conformation of the two kinds, meet the minimum, technical requirements of a diagnosis. Sesbania macrocarpa is retained. The synonym Sesbania exaltata (Raf.) Hill enters the literature under the presumption that Hill (then editor of Index Kewensis) made an inadvertant combination. This is not the case, the listing is Sesbania exaltatus Rydberg. Hill only corrected Rydberg’s spelling. SESBANIA EMERUS (Aublet) Urban, Repert. Spec. Nov. Regni Veg. 16: nee 1919. Aeschynomene emerus Aublet, Hist. Pl. Guiane 775, table des noms p. , 1775; Sesban emernus (Aublet) Britt. & Wilson, Sci. Surv. Porto Rico 5:395, 1924 Rydberg, N. Amer. Fl. 24:204. 1924 ?Emerus herbacea Miller, Gard. Dict. ed. 8. Emerus no. 3. 1768. The differences between the herbaceous to suffrutescent Seshania emerus and the ubiquitous, probably derivative United States annual, S. macro- carpa are of uncertain dimensions. On the basis of Antilles and Central American specimens and descriptions (e.g. Standley and Steyermark 1946; White 1980), S. emerus is a branched, presumably perennial or potentially perennial herb or a shrub 1 — 2(-5) m tall with larger flowers (corolla ca 15 — 20 mm long) than those usually possessed by S. macrocarpa. In the United States, interpretations of these species and their distribu- tion have been various. Rydberg (1924) listed both Seshan emerus of Florida and Seshan exaltatus (= Sesbania macrocarpa) “Missouri to Louisiana and Tex- s.” Small (1933) postulated a similar distribution but said of Seshan emerus “one of our very vigorous annuals,” surely primarily with reference to Sesbania macrocarpa. 437 Long and Lakela (1971; subtropical Florida) reported only Seshania macrocarpa without inclusion of Sesbania emerus as a synonym; Ward (1972) listed both species for Florida; and Wunderlin (1982; central Florida) in- cluded only Sesbania emerus “Disturbed sites. Frequent; throughout.” Sesbania macrocarpa (as Sesbania exaltata (Raf.) Cory) was relegated to synonymy. Thus, United States authors have reduced Seshania macrocarpa to Sesbania emerus or ignored the latter (Long and Lakela), or attempted differentia- tion. Distinction, where attempted, as indicated both by descriptions and annotation of specimens seems to be entirely on the basis of flower size. Most Florida material seen has been identified as Sesbania macrocarpa (or by its synonym, Sesbania exaltata) except that a few specimens from the southern half of the peninsula with flowers more than ca 15 mm long are identified as Sesbania emerus. hose United States plants that I tentatively take to be Sesbania emerus are of my own collections (ISC) from Key West where the species 1s not uncommon in ruderal areas. These plants are suffrutescent or woody, to 3 m tall, much branched and spring-flowering, i.e., blooming in April and abundantly fruiting by June. At least some Sesbania emerus from the Antil- les are also spring-flowering. The large-flowered (corolla 15-20 mm long) specimens from peninsular Florida otherwise, which botanists have identified as Sesbania emerus, | presently consider to be the annual, often wand-like and fall-flowering Sesbania macrocarpa. Similar plants are occasi- onal elsewhere in the United States. It is true that duration cannot always be determined from the specimens but, so far as discernible, the plants are not branched as are those of Key West, and all are late summer- or fall- flowering, typical of Sesbania macrocarpa throughout its range. KEY DIFFERENTIATION IN SUMMARY A eee branched, annual herb, flowering middle to late summer; flowers Li m long; a widely distribtued and abundant weed across the entir nee MWS aes soars ace aces Gah ee oe Sesbania macrocarpa A branched shrub, flowering in the spring; flowers 15 — 22 mm long; Fla, Key West. and possibly farther HORN, acs enkKets ee arcea eee Roe emerus The name Aeschynomene emerus of Aublet is derived from a Plumier plate that I have not seen. Since Urban, who made the combination in Sesbania, published an extensive See ee: about the Plumier taxa (Rep. Spec. Nov. Beth. 5:1— 196. 1920), the identity may be reasonably assumed. Britton and Wilson meena the combination Sesban emerus to Urban; i.e., they only corrected the spelling to that of the earlier orthographic variant and did not claim authorship. It was Rydberg (1924) who initially assigned the combination to them. 438 But the epithet “emerus” possibly should be superseded by the earlier Emerus herbacea cited in synonymy above. Miller said, of his Emerus herbacea, “It was found growing in plenty in La Vera Cruz, New Spain by the late Dr. Houston, who sent me the seeds—which succeeded in the Chelsea garden.” and this is verified by the annotations on the type specimen (Photo of type: photographs of Miller collections BH!) that I have examined. The specimen, which consists of pods and a few leaves, ts either Seshania emerus or Seshania macrocarpa, but it seems impossible to tell which, and the name is therefore rejected. SESBANIA VIRGATA (Cav.) Poir., Lam. Encyl. 7:129. 1806. Aeschynomene virgata Cav., Icon. Pl. 3:47, pl. 293. 1797. Sesbania marginata Benth., Mart., Fl. Bras. 15:43. 1859. Sesbanta virgata, introduced from South America, ts similar to the well known S$. drammondii and S. punicea but the flowers are smaller and the fruit, the indehiscent Daubentonia type, lacks wings. United States specimens seen range from coastal Florida panhandle to southern Mississip- pi. Most of them, collected by Demaree in the vicinity of Ocean Springs, Harrison Co., Mississippt match South American S. virgata except that the flowers may be either red or yellow, whereas those of 8. virgata are said to be yellow only. There is evident introgression between S. drummondii and the red-flowered S$. punicea about Ocean Springs, and it is possible that S. virgata 18 «lso genetically affected by the latter species. The preponderance of U.S. collections were initially identified as S. marginata Benth., possibly because Pierce (1942) felt that the original Cavanilles material of Aeschynomene virgata was Glottidinm vesicarinm. The identity of the Cavanilles plant (completely unlike the G/lottidium, both as to description and plate) has been verified by Burkhart (1967). CHAMAECRISTA (L.) MogENcH CHAMAECRISTA NICTITANS var. ASPERA (Muhl. ex Ell.) Irwin & Barneby, Mem. New York Bot. Gard. 35:838. 1982. Cassia aspera Muhl. ex EIL., Sketch Bot. S. Carolina 1:474. 1817; Cassia nictitans var. aspera (Muhl. ex ELL.) T. & G., Fl. N. Amer. 1:396. 1838 Irwin & Barneby inadvertently attributed their combination to Torrey & Gray who published the equivalent trinomial in Cassza. ACKNOWLEDGMENTS This work has been supported by NSF Grant BSR-8306918 and the lowa Agricultural Experiment Station. As always I owe a vote of hearty thanks to the personnel of several herbaria without whose cooperation work 439 like this could not proceed. The facilities and holdings (visits and/or loans) of the following herbaria have contributed to this report: FLAS, FSU, GA, ISC, MO, NCU, NY, UARK and USE Portions of the manuscript have been read by James Grimes, John Thieret, Robert Wilbur, Richard Wunderlin, and Rupert Barneby to whom I extend thanks for suggestions. Jennifer Byrne, Iowa State University student, has painstakingly con- ducted this manuscript through several “Apple” drafts. REFERENCES BARNEBY, R. C. 1977. Dalea Imagines. Mem. New York Bot. Gard. 27:1— 891. BURKART, A. 1967. Leguminosae, pp. 394—647. In: Cabrera, A. 1. Flora de la ee Buenos Aires. Vol. II. Instituto Nacional de Tecnologie Agropecuria, Buenos Aire FREEMAN, E L 1937. The variations of Psoralea psoralivides. Rhodora 39:425 — 428. ISELY, D. 1962. Leguminosae of the north-central states. IV. Psoraleae. Iowa State J. Sct. 37 1037162: . 1981. Leguminosae of the United week Hl. ee Papioncideas tribes Soehoree: Podalyrieae, Loteae. Mem. New York Bot. Gard. 204: LARISEY, M. M. 1940. A monograph of the genus Baptisia. ae Riecoutt Bot. Gard. LOZ ONG, R. W. and O. LAKELA. 1971. Fabaceae, pp. 445 — 502. In: A flora of tropical Florida. University of Miami Press, Coral Gables MERRILL, E. D. 1949. Index Rafinesquianus. 296 pp. Arnold Arboretum of Harvard University, Jamaica Plain. MER a E. D. and SHIU-YING HU. 1949. Work and publications of Henry enberg, with special attention to unrecorded or incorrectly recorded binomials. ae 25266, eee ee J. i 1942. The American species of Dawhentonia (Leguminosae). Trop. Woods re A. 1919— 1920. Fabaceae: Psoraleae. N. Am . 24:1— 136. 1924. Fabaceae: Galegeae (pars.). N. Amer. FI. wa se 250: . 1928. — of North American Fabaceae. HII. and IV. Tribe Psoraleae. Amer. J. Bor. 15: ae , 425-432. SMALL, J. K. 193: eee pp. 669 — 743. In: Manual of the southeastern flora. Au- STANDLEY, P. C. and J. A. STEYERMARK. 1946. Flora of Guatemala, Leguminosae. Fieldiana: Bot. 24:1 — 368. eae Co 1981. Psoraleeae (Benth.) Rydb., pp. 337 — 344. Im: Polhill, R. M. and PH. Raven (eds.). Advances in legume systematics. Part 1. Royal Botanic Gardens, oe England. THIERET, J. W. 1969. Baptisia lactea (Rafinesque) Thieret, comb. nov. (Leguminosae). TURNER, B. L. 1982. Review of "Leguminosae of the United States. Hl. ... by Duane Isely”. oe Bor. 7:350— 352. WARD, D. B. 1972. Checklist of the legumes of Florida. 21 pp. mimeo. Florida Agricul- tural eee ne Station, Gainesville. 440 WHITE, PS. 1980. Sesbania, pp. 760 — 765. In: Dwyer, J. D. and collaborators. Flora of Panama, eae minosae subfamily Papilionoideae. Ann. Missouri Bot. Gard. 671523 WILBUR, R. 7 1964. A revision of the dwarf species of Amorpha (Leguminosae). J. Elisha Mitchell Soc. 80:51—65. ———. 1975. A revision of the North American genus Amorpha (Leguminosae — Psoraleae). Rhodora 77:337 — 40 WUNDERLIN, R. P. 1982. Falsaceae, pp. 203 — 228. In: Guide to the vascular plants of central Florida. University Presses of Florida, Tampa AN UNDESCRIBED PANAMANIAN VACCINIUM: VACCINIUM BOCATORENSIS (ERICACEAE) ROBERT L. WILBUR Department of Botany, Duke University Durham, NC 27706, U.S.A Among the strikingly different species still being discovered along the relatively uncollected Caribbean slope of Panama is a vaccinioid unlike any previously described from Central America. In view of the recent attention played to this group of plants in Panama (Wilbur and Luteyn 1978, 1981), these continuing discoveries are surprising. They emphasize the need for critical collecting even in an area as intensively explored botanically as Panama, a country blessed with the most recent flora of any Central American nation as well as the most intensive collecting of any country in the area—a continuing program that reportedly has resulted in over 300,000 collections. Vaccinium bocatorensis Wilbur, sp. nov. Frutex ut videtur epiphyticus, plus minusve aa pert’ 2—3 mm longi. Laminae foliorum coriaceae, integrae, ovatae vel ovato-ellipticae, 4— 6.5 cm longae et 3.5 —5 cm latae, apice rotundatae, basi rotundatae et cordatae, pinnatinerviae. Inflorescentia racemosa vel subcorymbosa; rhachis 2 — 5 cm longa, pilosula; bracteae ee lanceolatae vel lanceo- ovatae, 3— 6 mm longae; pedicelli puberulenti eglandular1, 10 — 15(20) mm longi. Calyx et hypanthium 7 — 10 mm longus et ca 6 mm diametro, cylindricus, pilosulus; lobi calycis m longi. Corolla carnosa rubra cylindrica, glabra extra, tomentosa intra distalis; lobi corollae 2 mm longi et 2.5 mm lati. Stamina 5 mm longa; filamenta ca 2.5 mm longa, moderate vel dense sericea; antherae ca 3 mm longae, tubulis 2, 1— 1.2 mm longis. Stylus ca 7 mm longus, glaber Reportedly an epiphytic, scandent treelet. Mature stems drying brownish, + terete; bark thin, longitudinally splitting. Leaves coriaceous, entire, ovate to ovate-elliptic, 4—6.5 cm long * 3.5—5 cm wide, basally rounded and slightly cordate, apically + rounded, glabrous above and moderately but inconspicuously erect-pilosulose beneath with widely spaced, slender, hyaline trichomes 0.2 —0.3 mm long, apparently eglanduar; venation pinnate with 3 pair of lateral veins arising within 5 — 10mm of the base, the midvein and secondary veins impressed above and elevated beneath but the tertiary venation not apparent; petiole 2 — 3 mm long, shortly puberulent adaxially, + glabrous abaxially. In- florescence axillary but congregated distally and sometimes appearing SIDA 11(4):441 — 443. 1986. 442 terminal, racemose to subcorymbose, 6— 12-flowered, 3—7 cm long; rachis 2—5 cm long, + terete, moderately and inconspicuously pilosu- lose; floral bracts lanceolate to lance-ovate, acute to acuminate, sparingly puberulent to pilosulose externally, 3 —6 mm long; bracteoles borne on the proximal third of the pedicel, lanceolate to lance-oblong, 4—6 mm long and 0.6— 1.5(1.8) mm wide, adaxially glabrous and abaxially spar- ingly puberulent to pilosulose and marginally ciliate; pedicels terete, densely but inconspicuously puberulent with erect hyaline trichomes, striate, eglandular, 10— 15(20) mm long and about | mm in diameter, articulate with the flower. Calyx (and hypanthium) 7 — 10 mm long and about 6 mm in diameter, inconspicuously and moderately pilosulose with spreading erect trichomes, “purplish red,” the hypanthium broadly cylin- dric, pilosulose and glandular-pustulate throughout, the calyx lobes narrowly triangular to lanceolate, spreading recurved or reflexed, 4—5 mm long and about 2 — 3 mm wide at base; corolla thick and fleshy, 7 — 9 mm long and 6—7 mm in diameter, externally glabrous, internally glabrous or very nearly so for the basal half and increasingly tangled pilose distally and the broadly deltoid lobes exceedingly densely matted tomen- tose and about 2 mm long and 2.5 mm wide at base, “livid red” externally; stamens about 5 mm long, the filaments very weakly connate basally, very much flattened and rather broad, about 2.5 mm long and medially about 0.5 mm wide, marginally and on both surfaces moderately to densely sericeous especially on the connective abaxially, the anthers golden, finely or moderately granular, about 3 mm long, strongly incurved basally, the tubules 2/anther and each 1 — 1.2 mm long, about as wide as the thecae, glabrous, dehiscing by obliquely introrse pores; style included, glabrous, about 7 mm long. Type: PANAMA. Bocas del Toro: headwaters of Rio Colubre, 2400 — 2550 m (Colubre Camp), 3 Mar 1984, L.D. Gomez, 1. Chacon, G. Davidse & G. Herrera 22368 (HOLOTYPE: DUKE; tsoryee: ). Generic limits within the vaccinioid ericads are not as precise as might be implied or expected and the placement of this species is not above ques- tion. It 1s here placed in the genus Vaccinium because of certain resem- blances in shape and pubescence of the corolla and of the androecium to such species as Vaccinium poasanum Donn. Sm., V. floccosum (L.O. Wms.) Wilbur & Luteyn and even V. jefensis Luteyn & Wilbur. At first glance it might equally be placed in the neotropical genus Macleanea but its rela- tively short corolla and very short antherial tubules and finely granular thecae suggest a closer relationship with the very diverse genus Vaccinium. A broadly based reappraisal of generic limits within the vaccinioid ericads seems very much overdue, and until completed, generic placement will in 443 large part be rather speculative. In reference to this it should be noted that this species belongs in the same complex of species that has engendered a continuing debate as to the validity of the segregate genus Symphysia (Vander Kloet 1985). REFERENCES VANDER KLOET, S. P. 1985. On the generic status of Sywphysia. Taxon 34:440 — 447. WILBUR, R.L. & J. L. LUTEYN. 1978. Flora of Panama: Ericaceae. Ann. Missouri Bot. Gard. 65:27 — 144. 1. Additions to the Ericaceae of Panama. Ann. Missouri oe a RY e198 Bot. Gard. 68:154 — 166. SCAEVOLA (GOODENIACEAE) IN SOUTHEASTERN UNITED STATES JOHN W. THIERET Department of Biological Sciences Northern Kentucky University Highland Heights, KY 41076, U.S.A. DAVID M. BRANDENBURG Brooklyn Botanic Garden 1000 Washington Avenue Brooklyn, NY 11225, U.S.A ABSTRACT Two species of Scaevola (Goodeniaceae) occur in southeastern United States: S$. plamiers (L.) Vahl, a native strand plant, and S. sericea Vahl, a shrub used as an ornamental and sparingly escaped a naturalized. Noces on a y, ne keys to and descriptions of taxa, p i) including two varieties of S. sericea, are giv Two species of Scaevola (Goodeniaceae}—a genus of about 80 species—are littoral plants widely distributed in warm regions. The Indo- Atlantic S. plumieri (L.) Vahl (Fig. 1) occurs from Ceylon and India west through the Mascarenes (Guillaumet 1976) and eastern, southern, and western Africa (Somalia to Cape Province to Sao Tome) (Davies 1978) to Florida, Louisiana, Texas, the West Indies, the Galapagos, and continental tropical America. The Indo-Pacific S. sericea Vahl (Fig. 2) is native from eastern Africa (Kenya, Tanzania, and Natal) (Davies 1978), Madagascar, the Mascarenes, and the Seychelles east through India, southeastern Asia, Malaysia, and tropical Australia to Melanesia, Micronesia, and Hawai; it is naturalized elsewhere, e.g., in Florida and in the Bahamas (Correll and Correll 1982). Thus, as Guppy (1917) pointed out, the two taxa divide between them the tropical shores of the world. Scaevola plumieri and S. sericea owe their wide distribution primarily to the ocean-current-dispersed stones of their fruits. The stones of S. sericea are buoyant because of a corky outer layer; in contrast, those of S. plumier? lack such a layer but usually have one empty, watertight locule (Brizicky 1966; Guppy 1906, 1917). Stones of the former species can float in sea water for at least a year (Guppy 1890, 1906, 1917); those of the latter, for only 4 or 5 months (Guppy 1917). Fruits of S. sericea float with or without their fleshy outer layer; this portion, if not worn off by beach sand before the fruits reach the water, is SIDA 11(4):445 — 453. 19860. 446 \\ SS \ —SS | \\ fj | / N of er, adax g branch, X 1; upper left, flow 448 said to disintegrate “during the early days of . . . immersion” (Guppy 1890). Fresh drupes of S. p/wmier7, upon getting into the sea, will, accord- ing to Guppy (1917), “sink in two days, the buoyant stone, on being freed from the decaying fruit, soon floating to the top,” an observation somewhat at variance with ours. Drupes we placed in sea water on 23 July 1985 never did sink but remain, 15 months later, as buoyant as ever. The fleshy portion is still intact (but barely so), doubtless because the fruits have been in still water in a beaker rather than in the moving, often turbulent sea. Lesko and Walker (1969) found that stones of S. sericea showed no “significant loss” in viability after 50 days of floating in sea water. Indeed, such fruits germinated 1—2 weeks sooner than dry ones when placed in non-saline environments. Dispersal of the two taxa may also be accomplished by frugivorous birds. Indeed, the fruits seem as well suited for ornithochory as for hydrochory; Guppy (1917) and Schimper (1891) suggested the agency of birds for local dispersal, that of ocean currents for long distances. A note ona specimen of S. plumiert from Barbuda (J. S. Beard 367, GH) reads: “fruit said to be eaten by gulls” (Altschul 1973). Julia E Morton (pers. comm.) was told in 1982 by D. S. Correll that, on Pine Key in the Caicos Islands, he watched seagulls devouring all the fruits on S. p/wmieri, coming in great flocks to feast on them. Millspaugh (1907), in his paper on the Florida Keys, stated that “the black, pulpy fruits of {8. p/wmier/} form a very attractive food for land birds; 1t thus becomes scattered far throughout the Antillean region.” Louda and Zedler (1985) wrote that, among the four Pacific atoll species whose fruit and seed predation they studied, S. sericea (as S. taccada) had the highest race of fruit disappearance, due to hermit crabs and, possibly, birds. Scaveola sericea 1s apparently a recent introduction to the naturalized flora of Florida. It is mentioned as a plant “confined to the Eastern Hemisphere,” but not as one cultivated or spontaneous in the southeast, by Brizicky (1966) and is absent from Long and Lakela (1971). It is not included in Shelter and Skog (1978) or in the “United States and Canada” section of National List of Scientific Plant Names (Soil Conservation Service, U.S.D.A. 1982) (it ts listed, of course, in the “Hawaii” section). Kartesz and Kartesz (1980) listed the species—in four varieties—but there is no way to tell from their listing which, if any, of the varieties was known to them from conterminous United States as opposed to Hawaii, where all four varieties occur (Hawaiian plants are accounted for in the Kartesz and Kartesz work). Finally, the presence of the species—but only the sericeous variety—in Florida 1s acknowledged in Wunderlin’s (1982) central Florida manual. 449 Scaevola sericea is cultivated as an ornamental in southern Florida, where its use is increasing. We have seen it as a hedge plant on Sanibel Island and on Key Biscayne. With its glossy, bright green leaves, its white to lavender flowers, and its white fruits, it is indeed attractive. The appearance of the fruits is recalled in one common name we heard for the plant, “hailstones” (see also Neal 1965). In Florida, S. sericea spreads from cultivation—presumably by seeds—to nearby thickets, woodland borders, canal banks, and waste places, as on Sanibel Island. Its stones are obviously dispersed by ocean currents, too. For example, on the foredune near Marathon High School, Key Vaca, Monroe County, both varieties of S. sericea grow with S. plumieri—all certainly spontaneous—just 2 or 3 meters from waters of the Straits of Florida. We noted, on Key Biscayne, an individual of S. sericea var. sericea about 40 cm tall rooted in sand between beached sailboats on the upper strand; we suspect that it could have originated from a washed-up stone or from a fruit produced on a hedge of S. serzcea about 100 m away. It certainly was not deliberately planted where it was growing. The pollen-presentation mechanism of Goodeniaceae is an outstanding feature of the flowers (Brizicky 1966; Carolin 1960; Krause 1912; Kugler 1973). The style bears at its apex a pollen-collecting cup (or “cupular in- dusium”) within which ts the stigma. The anthers, more or less coherent, release their pollen introrsely while the flower is still in bud; the pollen collects in the cup as the style elongates. At anthesis, the enlarging stigma pushes at least some of the pollen out of the cup onto the trichomes of the cup’s apex and exterior. Pollen from the cup and the trichomes is then dusted onto visiting insects (bees, wasps, beetles, and butterflies; perhaps only the first two mentioned are effective pollinators). The stigma finally becomes receptive after the pollen is all or mostly gone from the cup. Hairs of the cup then brush pollen off visitors, bringing it to the stigma. The nomenclature of S. plamieri and S. sericea has been reviewed by Jeffrey (1980). It is hoped that his conclusion that S. sericea is indeed the correct name of this species will lay to rest the long-standing controversy and confusion over the matter. Jeppesen (1981) wrote that S. p/amieri and S. sericea (as S. taccada) are “very similar and . . . may prove to be one polymorphous pantropical species,” a suggestion that seems to us untenable. The taxa are, at least in Florida, certainly easily separable. Guppy (1917) noted: “The two plants, as was evident to me when I first met Se. p/wmieri in the West Indes, are quite distinct, and could not be mistaken by any one with both before him.” To the other features that characterize the two should be added the 450 fact that S. sericea is frequently grown as an ornamental and S. p/umieri is not—at least we have not seen it so, although the species is offered by at least one southern Florida nursery specializing in native plants. Scaevola serivea 1S propagated with ease by cuttings; S. p/wmier’, only with difficulty by this means (A. Sprunt, Jr., pers. comm.). TAXONOMIC ACCOUNT The account below follows, in general, the format established for Vascular Flora of the Southeastern United States. SCAEVOLA L. Evergreen, monoclinous shrubs. Leaves simple, alternate, often crowded at branch tips, blade tapering to a sessile or subpetioled base. Inflorescence cymose, axillary, bracteate. Flowers perfect, zygomorphic, epigynous; calyx of 5 sepals; corolla of 5 united petals, split to base adaxial- ly, tube woolly inside, limb 1-lipped, spreading, resembling an outstretched hand, lobes more or less equal, narrowly ovate, apex acute, margins with thin, induplicate wings; stamens 5, free, the pollen subpro- late to prolate, 3-colporate (Dunbar 1975; Moreira and Belem 1978); carpels 4 but appearing as 2 (Carolin 1959, 1966); stigma within a ciliate, pollen-collecting cup; style elongate; ovary inferior, 2-loculed, ovule | per locule. Fruit a fleshy drupe, 1 — 2 seeded Cyme with central flowers sessile, laterals pedicelled; calyx more or less obsolete to clearly 5-lobed, the lobes broadly ovate to broadly triangular, 0.2—1 mm long; drupe black Saar at lane tne dag ae asta inte te Sepa Reeenete acne a Gace Gann Seta Soa le d's Baya, ane _ SS. plamieri Cyme with all flowers pesicelled calyx 5-lobed, the lobes linear to ners obovate or narrowly elliptic, 3.5 —5 mm long; drupe white to yellowish- Pee et ct ee ee S. sericea 1. S. prumieri (L.) Vahl. Beachberry, inkberry. Shrub 15 — 150 cm tall, often colonial; stems glabrous or essentially so, often glandular, the leaf axils white-pilose with hairs 0.5 — 2 mm long. Blades narrowly to broadly elliptic or obovate, 2.5 — 10 cm long, 0.5 — 5 cm wide, glabrous or nearly so, sometimes minutely glandular, apex rounded to obtuse, sometimes apiculate, margin entire. Cymes simple to 1-compound, or 1-flowered by reduction; central flowers sessile, laterals pedicelled. Calyx more or less obsolete to clearly 5-lobed, the lobes broadly ovate to broadly triangular, 0.2— 1 mm long, glabrous, rounded to obtuse; corolla white to pinkish- white adaxially, pale greenish to pale yellowish abaxially, 1.8— 2.4 cm long. Drupe subglobose to ellipsoid, |— 1.8 cm long, 1— 1.6 cm wide, black. (n= 8; Kausik 1939). Spring-fall. Coastal dunes and beaches; cp. 4S] Fla (all coastal counties north to and including Hillsborough and Pinellas on the west coast and Brevard on the east); La Jefferson Parish [Timbalier Island}, where collected in 1937: E. B. West 123, LSU; Brown 1945); and Tex (Kleberg County, where collected in 1954: FB. Jones L082, SMU). The Louisiana and Texas (?) populations may be “one-time” introductions and possibly not persistent. 2. S. sericea Vahl. Beach naupaka, hailstones. Shrub 0.5 — 2 m tall (7 m maximum recorded in Flora Malesiana; Leenhouts 1957); stems glabrous or essentially so or lightly to heavily sericeous or glandular-sericeous, the leaf axils conspicuously white-pilose with hairs 1.5 —7 mm long. Blades elliptic to spatulate or obovate, 4— 21 cm long, 1.8—9 cm wide, often minutely glandular, apex truncate or slightly emarginate to obtuse, sometimes apiculate, margin entire to shallowly crenate. Cymes usually 2—4 compound; all flowers pedicelled. Calyx 5-lobed, the lobes linear to narrowly obovate or narrowly elliptic, 3.5 —5 mm long, acute to obtuse; corolla white to pale lavender adaxially, pale greenish-yellow abaxually, 1.8—2 cm long. Drupe subglobose to ellipsoid, |— 1.7 cm long, I — 1.5 cm wide, white to yellowish-white. (n= 8; Carr 1978; Skottsberg 1953). Spring-summer. Coastal dunes, beaches, woodland borders, thickets, canal banks, and waste places; cp. Fla {S. fratescens Krause; S. Roenzgzz Vahl} The two taxa below are cultivated in Fla; both are sparingly escaped and extending their ranges as naturalized plants. Leaves lightly to heavily sericeous (Fig. 3a) 2.2... ee 2a. var. sericea Leaves-clabrous or essentially so (Fig. 3D)ino sca. iwi eine veaey eres 2b. var. taccada 2a. var. seRICEA. Stem lightly to heavily sericeous; inflorescence sericeous. Fla: Dade, Lee, and Monroe counties. 2b. var. raccaba (Gaertn.) Thieret & Lipscomb. Stem glabrous or essentially so; inflorescence glabrous to sericeous. [S. ftaccada (Gaertn.) Roxb.} Fla: Monroe County. The status of the infraspecific taxa of S. sericea, a most variable species, 1s subject to verification following study of the plant throughout its natural range. ACKNOWLEDGMENTS Thanks are due to the curators of FLAS, FSU, LSU, NCU, and SMU for loan of specimens; and to Barney L. Lipscomb, Julia EF Morton, Tony Powell, A. Sprunt, Jr., and Nancy Lee Thieret for aid during the prepara- tion of this paper. 452 Figure 3. Scaevola sericea, leaves, ies ts vesticure: left, var. sericea, leaf, X 1/2, inset, X 20; right, var. taccada, leaf, X 1/2, inset, 20 REFERENCES ALTSCHUL, S.v.R. 1973. Drugs and foods from little-known plants. Harvard University Press, Cambridge BRIZICKY, G. K. 366, The Goodentaceae in the southeastern United States. J. Arnold Arbor. 47:293 — 30( BROWN, C. A. 1945. ee on additions to the flora of Louisiana. Proc. Louisiana Acad. Sci. 9:4-1 CAROLIN, R. ¢: 1959. Floral structure and rime in the family Goodentaceae Dumort. Proc. Linn. Soc. New South Wales 84:242 — 255. . 1960. The structures involved in the presentation of pollen co visiting insects in the Gndes Campanulales. Proc. Linn. Soc. New South Wales 85:197 — 207. 1966. Seeds and fruit of the Goodeniaceae. Proc. Linn. Soc. New South Wales MEY: pt = op CARR, a D. 1978. Chromosome numbers of Hawaiian flowering plants and the signifi- cance of eyeoleey in selected taxa. Amer. J. Bot. 65:236 — 242 CORRELL, D. S. and H. B. CORRELL. 1982. Flora of the Bahama Archipelago (includ- ing the atkes a Caicos Islands). J. Cramer, Vaduz. ee EG. 1978. Goodeniaceae. In: R. M. Polhill, editor, Flora of Tropical East Africa. wn Agents, London. oo A. 1975. On pollen of Campanulaceae and related families with special reference to the surface ultrastructure Il. Campanulaceae subfam. Cyphioidae and sub- fam. Lobelioidae; Goodeniaceae; Sphenoc st Bot. Not. 128:1 02— 118. aun AUMET, J.-L. 1976. Goodéniacées. In: e des Mascareignes 110:1—4. GUPPY, H. B. 1890. The dispersal of plants as ceeonee by the flora of the Keeling or Cocos Islands. Trans. Victoria Inst. 24:267 —301; 304—306 (“Authors further rep joe 1906. Observations of a naturalist in the Pacific between 1896 and 1899. II. Plant Gisieredl Macmillan and Company, London. 453 . 1917. Plants, seeds, and currents in the West Indies and Azores. Williams and Norgate, Lon ee C. 1980. for the nomenclature of the strand Scaevola species (Goodeniaceae). Bull. 34:537 — 545. ce S. 1981. Goodeniaceae. In: ie of Ecuador 14:177 — 178. KARTESZ, J. T. and R. KARTESZ. 198( onymized checklist of the vascular flora of the United States, Canada, and oa val 2. The biota of North America. Sey of North Crolina Press, Chapel H KAUSIK, S. 1939. A cytological study of Sal Lobelia Linn. [Scaevola plumiuert (L.) Vahl}. oe Indian Acad. Sci. B. 9:39 — KRAUSE, K. 1912. Goodeniaceae. oan IV. 227(Heft 54): 1 — 207. KUGLER, H. 1973. Zur Bestaubung von Scaevola plumieri (L.) Vahl ae so pes-caprae Sweet, zwei tropischen Strandpflanzen. Flora (Jena) 162(4):381— LEENHOUTS, P W. 1957. Goodeniaceae. J: Flora Malesiana I. 5: cc 344, LESKO, G. L. and R. B. WALKER. 1969. Effect of sea water on seed germination in two Pacific atoll beach species. Ecology 50:730 — 734 LONG, R. W. and O. LAKELA. 1971. A flora of opie Florida. University of Miami Press, Coral Gables. LOUDA, S. M. and P H. ZEDLER. 1985. Predation in insular plant dynamics: an ex- perimental assessment of postdispersal ~s and seed survival, Enewetak Atoll, Marshall Islands. Amer. J. Bot. 72:438 — MILLSPAUGH, C. FE 1907. Flora of the sand is of Florida. Publ. Field Columbian Mus. 118, Bot. Ser. 2(5): ee MOREIRA, A. X. and C FE BELEM. 1978. Confronto entre os polens de Scaevola Plumieri ee e Scaevola es (Mill.) Krause [S. sericea Vahl]. Revista Brasil. Biol. 38:837 NEAL, M. e 1965. In gardens of Hawaii. Special Publ. Bernice Pauahi Bishop Mus. 50. SCHIMPER, A. FE W. 1918. Die Indo-Malayische Scrandflora. Gustav Fischer, Jena. SHETLER, S. G. and L. E. SKOG, eds. 1978. A provisional checklist of species for Flora North America (revised). Missouri Bot. Gard. Monog. Syst. Bot. 1. SKOTTSBERG, C. Rie ae romosome numbers in flowering plants. Sniclemmae? report. Ark. Bot. 3(4):63 — 70. SOIL Connon Gnuiee: U.S.D.A. 1982. National list of scientific plant names. Vol. 1. List of alt THIERET, J. W. and B. L. LIPSCOMB. 1985. Scaveola es Vahl var. taccada (Gaertn.) Thieret & Lipscomb, comb. nov. (Goodeniaceae). Sida 11:103. WUNDERLIN, R. P. 1982. Guide to the vascular plants i central Florida. University Presses of Florida, Tampa. REVIEWERS FOR VOLUME 11 OF SIDA The following individuals kindly provided time and energy 1n review- ing manuscripts published in SIDA Charles M. Allen Loran C. Anderson oe K. Arp K . Brown he E Clewell Lincoln Constance Helen B. Correll Allison W. Cuscik George see H.R. DeSelm Paul A. Fryxell Robert R. Haynes eS S. Irving Surrey Jacobs Jeffrey Marshall Johnston Almut G Walter S. Judd Robert Kral Walter H. Lewis Wm. E Mahler, Publisher robe L. Lipscomb, Editor, Jol in W. Thieret, Associate Baier SIDA 11(4):454. 1986. Elizabeth McClintock J. McNeill Guy L. Nesom Elray S. Nixon Kittie E Parker A. Michael Powell se K. Rabeler eed C. — in Rollins B. Smith John L. Strother Eric Sundell Connie Taylor Edward E. Terrell Ralph Thompson Robert E Thorne Alfred Traverse S.P. Vander Kloet Warren L. Wagner Dan B. Ward Dieter C. Wasshausen Natalie W. Uhl Richard P Wunderlin ADDITIONS TO THE FLORA OF NEW MEXICO RICHARD SPELLENBERG Department of Biology, New Mexico State University Las Cruces, NM 88003, U.S.A. RICHARD WORTHINGTON Department of Biology, University of Texas at El Paso E/ Paso, TX 79968, U.S.A. PAUL KNIGHT Resource Surveys, New Mexico Dept. of Natural Resources Villagra Building, Santa Fe, NM 87503, U.S.A. REGGIE FLETCHER U.S. Forest Service, 517 Gold Ave., SW Albuquerque, NM 87102, U.S.A. The publication of A Flora of New Mexico (Martin and Hutchins 1980, 1981; for brevity cited as M&H from hereon) was a long awaited and wel- come event. To have access to geographic ranges and nomenclature updated over the 65 years since the publication of the Flora of New Mexico (Wooton & Standley 1915) has been a stimulus to floristic research in the state and has allowed the easy determination of plant species here-to-fore unknown in New Mexico or in portions of the state. Also, since 1975 in- creased knowledge of the flora of New Mexico has resulted from vegetation studies concerned with the distribution of endangered or threatened plant species, from environmental impact studies, or both. In addition to the 79 records that we present here, at least another 138 are now scattered in the literature and are compiled in the Appendix to aid individuals who wish to identify plants from New Mexico or are interested in floristics of the South- west. The taxa presented here add 8 genera, 43 species, and 9 subspecific taxa to the NM flora. Those given in the appendix bring the total taxa reported here and not recorded as present in NM by M&H to one family, 25 genera, 144 species, and 28 subspecies or varieties. The following records are arranged in alphabetic order by family and genus. Those that represent taxa not recorded for the state in M&H are marked by an asterisk before the name of the taxon in this list or in the appendix; most of these are first records as indicated by specimen citations. Others are significant range extensions within the state. Abbreviations for SIDA Li(4)455 —470. 1936, 456 collectors are: F = Fletcher, K = Knight, S = Spellenberg, W = Worthington. Deposition and date of the first known collection is in- dicated; later collections by EF K, S, or W can be found at ALBU, NMC, UTEP, or UNM. ARISTOLOCHIACEAE ARISTOLOCHIA *wriGHTu Seem. Luna Co.: § 1/2 Florida Mts, 9 Apr 1978, W 2581, 5316, 7046 (COLO, UTEP); Melntosh, s.n. (NMC). A species of w TX and n Mex. ASTERACEAE *ANTHEROPEAS LANOSUM (Gray) Rydb. (Eriophyllum lanosum Gray). Hidalgo Co.: Peloncillo Mts, Granite Gap, 9 Apr 1982, S 6455 (NMC, NY, UNM, ASU); near Steins, W 10221, 11721.5, near Lordsburg in Pyramid Mts, W 99/9. Common spring annual formerly known from se AZ to nw Mex and s CA, in our region most frequent after moist winters. AsTeR LAEVIS L. var. *GUADALUPENSIS A. Jones. Eddy Co.: Guadalupe Mts in riparian of Big, Black, and Gunsight canyons, 15 Sep 1982, K 2370, 2374 (UNM). Variety previously known only from nw TX. MACHAERANTHERA *RIPARIUS (Kunth) A. G. Jones. Known from sever- al collections in Hidalgo Co., NM (NMC, UNM). Jones (1983) follows other authors such as Kearney and Peebles (1960) (who use the name Aster riparius Kunth) and has considered this and A. sonorae Gray as conspecific. None of these names appear in M&H, but A. b/epharophyllus does, which is interpreted as a rare form of M. riparia by Almut Jones (pers. comm. to S., 26 Feb 1986). CHRYSOTHAMNUS PARRYI (Gray) Greene subsp. *parryi. Rio Arriba Co.: Toltec Mesa 8 km airline ne of Lagunitas, 13 Oct 1984, F 7912 (ALBU, UNM). This subspecies was previously known to range from WY to NV, UT, CO. Cirstum *scopULORUM (Greene) Cockll. Taos Co.: Sangre de Cristo Mts, above E. Fork Santa Barbara Cr., 6 Aug 1985, F 8414 (ALBU, NMC). Extension from the alpine of s CO. Conyza *RAMOsiIssSIMA Crong. Dofia Ana Co.: Las Cruces, 4 Oct 1985, McIntosh 1824 (NMC). Common weed in e US; established in Las Cruces in lawn and cracks of sidewalk. ERIGERON *BiGELOviI Gray. Hidalgo Co.: Sierra Rica, 13 May 1980, § & Ward 5524 (NMC, NY). Otero Co.: Sacramento Mts., Soreng 2808 (NMC). M&H map as potential for se NM; previous United States records apparently from TX near El Paso. Nesom (pers. comm.) says species 1s frequent in CHIH. 457 ERIGERON *sCOPULINUS Nesom & Roth. Sierra Co.: Black Range, ca 3 km W of James Brothers Cabin, T11S, RIOW, S28, 24 May 1981, Todsen 810524-01 (NMC). Catron Co.: Mogollon Mts, K 1759. Socorro Co.: San Mateo Mts, K & F 1781, etc.). Previously known only from se AZ. ERIGERON *uRsiNus D.C. Eaton. Rio Arriba— Mora cos line: Sangre de Cristo Mts, vicinity Santa Barbara Peak, 6 Aug 1985, F 8384 (ALBU, NMC). Extension from the alpine of S CO; only ca 6 plants observed. *FILAGO CALIFORNICA Nutt. Hidalgo Co.: U.S. Hwy. 80, Peloncillo Mts, 20 Apr 1973, S 3039 (NMC, NY); Pyramid Mts, W 9908. Grant Co.: Little Hatchet Mts, W 8/95B. Luna Co.: Tres Hermanas Mts, W 10371. Spring annual formerly known from se AZ to s CA and nw Mex. The species may occur in collections under the names Sty/ocline micropoides Gray or Evax multicaulis DC., both reported for the region, all three of similar appearance, and for the casual collector, at least, difficult to distin- guish. HyMENOPAPPUS FILIFOLIUS Hook. var. *PAUCcIFLORUs (I. M. Johnston) B. L. Turner. San Juan Co.: ca 10 km ssw of Waterflow in sand, 31 May 1985, S 8199 (NMC, TEX). Confirms presence for NM; said by M&H to be expected in nw corner. MALACOTHRIX *SONORAE Davis & Raven. Grant Co.: Little Hatchet Mts, 24 Apr 1982, W 8157 (UTEP). Luna Co.: Tres Hermanas Mts, W 9545. Catron Co.: ca 9 km se of Glenwood, Sheridan Gulch Trail, Soreng & Ward 21306 (NMC). Known previously from s AZ and n SON. *PRENANTHELLA EXIGUA (Gray) Rydb. Hidalgo Co.: Pyramid Mts, 7 Apr 1984, W 11739 (UTEP). Mostly a species from the Sonoran Desert, but type collection is from near El Paso, TX. Postulated distribution by M&H for se NM counties unlikely. Prionopsis CILIATA Nutt. San Juan Co.: 21 km w of Farmington, US Hwy 550, 16 Sep 1984, S 7875 (NMC, NY). Dona Ana Co.: Las Cruces (S in 1985, visual w/o specimen). Eddy Co.: Carlsbad, § 8259. A plains species recorded by M&H from Otero Co. SENECIO *cCANUS Hook. Harding Co.: 26 km ne of Roy, 2 Jul 1981, $ et al. 6053, § 7203 (NMC, KSC), a record ca 150 km se of nearest known populations. Diminutive, monocephalic form on caliche, with Parthenium alpinum. Other collections from n CO and central UT resemble this phase but none are as dwarfed (specimens examined at BYU, CSU, ID, UNM, UT, UTC, WS). SENECIO *DIMORPHOPHYLLUS Greene var. DIMORPHOPHYLLUS. Taos Co.: Wheeler Peak Cirque, 5 Aug 1980, F 4690 (ALBU, UNM). Previ- ously known from high mts of CO and WY. SENECIO *sOLDANELLA Gray. Taos Co.: Red Dome e of Wheeler Peak, 458 5 Aug 1980, F 4659 (ALBU). Previously known from high mts of CO. *SILYBUM MARIANUM (L.) Gaertn. Dofia Ana Co.: NM Hwy 404 11 km e of Interstate 25, roadside, 8 May 1985, S 8155 (NMC, UTEP). A widespread weed introduced here in straw used to stabilize roadside. Con- firms presence for New Mexico; said to be expected in northern tier of coun- ties by M&H. THELESPERMA SIMPLICIFOLIUM Gray. San Miguel Co.: Rowe Mesa s of Pecos, 22 Aug 1985, F 8443 (ALBU). A N range extension of ca 200 km from Chaves Co. BORAGINACEAE CRYPTANTHA *BARBIGERA (Gray) Greene. Luna Co.: Florida Mts, 4 Apr 1980, W 5709 (UTEP). Common to the w, reported as far eas w TX (Correll & Johnston 1970), and said to be expected in s NM by M&H. Numerous collections by W (UTEP, etc.) from Dona Ana, Grant, Hidalgo, Luna cos. confirm its presence. PECTOCARYA *HETEROCARPA (I. M. Johnst.) I. M. Johnst. Luna Co.: Victorio Mts, 23 Apr 1983, W 9995 (NMC, UCR, UNM, UTEP). In- dicated by M&H to be expected in three sw counties of NM; easternmost record in w TX (Correll & Johnston 1970). Recent records of this, P. platy- carpa (Munz & I.M. Johnst.) Munz & I.M. Johnst., and P recurvata 1. M. Johnston well e of previously known ranges may be due to spread of bristly fruits by livestock. However, a number of spring annuals, common to the w, have been found in this region during the past two decades. Increased botanical activity may be one explanation, their comparatively introduc- tion from the w another, and their spread from small sites of suitable micro- habieat after degradation of arid grassland and increase of desert scrub (Van Devender 1980, York & Dick-Peddie 1969) yet a third. The region was inadequately botanized earlier to give strong support to any one explana- tion, BRASSICACEAE CHORISPORA TENELLA (Pall.) DC. Dofia Ana Co.: Las Cruces, 4 Apr L981, Leyendecker s.n. (ASU, NMC, NMCR, NY, TEX, UNM). Reported by Marley & Wagner (1981) and mapped by M&H 300 km to n; also in w TX near NM border (Lipscomb 1984). Apparently rapidly spreading and to be expected throughout NM. DipLoraxis *TrENUIFOLIA (L.) DC. Otero Co.: US Hwy 82, 8 km e of US Hwy 70, 15 Aug 1972, Jackson 295 (NMC). Collections by W and students (UTEP) from the same general vicinity, and sightings of the 459 species 1n Dofia Ana Co. and in El Paso, TX, indicate this Old World weed to be well established and spreading. LepipiumM *CAMPESTRE (L.) R. Br. Sierra Co.: n end of Black Range e of Beaverhead, 2 Jun 1982, F & K 6124 (ALBU, UNM). Confirms presence for NM; said by M&H to be expected in n. LepipiuM *LaTIFOLIUM L. Donia Ana Co.: near TX border, 8 Jun 1976, Alpers 19 (NMC, GH). Guadalupe Co.: 5 km n of Pastura, S ef a/. 5984. European weed now widespread in N. Amer., said M&H to be ex- pected in central NM. Leprpium *RAMOsiIssiMuM A. Nels. Santa Fe Co.: nw portion of co. in Guaje Canyon, | Jun 1984, S 7766 (NMC). Confirms presence for NM; said to be expected by M&H. STREPTANTHUS *spARSIFLORUS Rollins. Eddy Co.: Guadalupe Mts, Black River Canyon, 17 May 1979, F 3827 (UNM); others in general vi- cinity F 3807, K 1966. An endemic of the Guadalupe Mts, previously known only in TX. THYSANOCARPUS AMPLECTENS Greene. M&H write that this occurs in sw NM, but map it “to be expected.” Rollins (pers. comm. to W) says that the type locality is in that portion of the state, and that the name is a synonym of Thysanocarpus curvipes Hook. var. elegans (E & M.) Robins. Records at NMC, UNM, and UTEP, and probably elsewhere, indicate it to be fairly common there. CAMPANULACEAE TRIODANIS *BIFLORA (Ruiz & Pavon) Greene. Grant Co.: 11 km nnw of Buckhorn, base of Mogollon Mts, 22 May 1983, Soreng G Ward 2143a (NMC). A species of the s U.S. from CA to VA, s to MEX and S.A. Inter- mixed with T. perfoliata (L.) Nieuwl. at this site. CARYOPHYLLACEAE CERASTIUM *AXILLARE Correll. Otero Co.: ca 3 km nw of Orogrande at s end of Jarillo Mts, 12 Apr 1985, McIntosh 1605 (NMC). Previously known from w TX and n CHIH. STELLARIA *WEBERI Boiv. Taos Co.: Latir Peaks, 1 Aug 1979, F 4028 (ALBU, UNM). S extension of taxon from CO. Considered a dwarf alpine form of S umbellata Turcz. (Weber 1976), the latter known from several localities in NM. CHENOPODIACEAE ATRIPLEX *PLEIANTHA W. A. Weber. San Juan Co.: Navajo Mine, Fruitland, 5 June 1979, Pace et al. 2388 (RM); collected in same area with- 460 out knowledge of previous record in 1983, 1984, § 7087, 7091, 7795. Sporadic on Fruitland Clay knolls; common at 709/ in 1983, where co- occurring with two other state records, Phacelia demissa, Mentzelia thompsoni (Soreng 1984a), but absent at this site in the drier spring of 1984. Previ- ously known only from one site in extreme sw CO (Ecology Consultants 1978), but now also known in se UT. CUCURBITACEAE CirRULLUS VULGARIS Schrad. var. *crrroipes Bailey. Eddy Co.: ca 42 km airline ese of Carlsbad on large, active dunes at Los Medanos, 21 Jul 1978, K 394 (UNM); K 786 (UNM). About 30-40 plants were found over an area of ca 20 km near an abandoned dwelling. Collections from two consecutive years indicates that population is at least weakly persistent. IBERVILLEA *rRIpARTITA (Naud.) Greene. Eddy Co.: ca 45 km airline ese of Carlsbad in Los Medanos dunes, 29 Apr 1983, K 2467 (UNM). A species common in the Trans-Pecos region of TX. CYPERACEAE CAREX PITYOPHILA Mackenz. Cibola Co.: ca 40 km ssw of Grants at $ end of Grants Malpais, 19 May 1979, S G Repass 5131 (NMC, NY); Zuni Mts, F 6210 (ALBU). M&H map this for Rio Arriba Co., ca 100 km to ne. Hermann (1970) indicates this species to be rare and local in s CO and n EUPHORBIACEAE EUPHORBIA *ERIANTHA Benth. Eddy Co.: Guadalupe Mts, Big Can- yon, infrequent in riparian, 16 Sep 1982, K 2400 (UNM). A weedy species common in much of s U.S EUPHORBIA GEYERI Engelm. var. *wHEELERIANA Warnock & M. C. Johnston. Dona Ana Co.: near boundary monument 4 on Mexican border, 25 Jul 1984, Melntosh 1588 (NMC). Formerly known only from dunes of extreme w TX, where it is said to be rare and local (Correll & Johnston 1970). It has also been found very near the NM border at Anthony, TX (UCR, UTEP). FABACEAE ACACIA ANGUSTISSIMA (P. Miller) Kuntze var. *surFRUTESCENS (Rose) Isely. Hidalgo Co.: Peloncillo Mts, Guadalupe Pass, T34S, R21W, s edge S16, 15 Aug 1979, S & Repass 5301 (NMC). Isely (1973) shows this variety to be widespread in s AZ. Variety texensis (T. & G.) Isely is common in sw NM. 4G] ACACIA *MILLEFOLIA S. Wats. Hidalgo Co.: very near MEX and AZ borders in Guadalupe Canyon and tributaries, 17 Aug 1979, S & Repass 5371 (NMC, NY). Isely (1973) indicates this species to occur in extreme se AZ, virtually on the NM border. Observed several times within | km of AZ. border by S. ASTRAGALUS *DESPERATUS Jones var. DESPERATUS. San Juan Co.: near AZ border, 3.5 km airline s of Hwy 504 near Beclabito, 27 Apr 1983, K 2446 (UNM); 14 km airline n of “The Thumb,” K 2448. Members of the section Desperati are common in se UT and ne AZ, but uncommon in NM. ASTRAGALUs FEENSIS M. E. Jones. Hidalgo Co.: low pass s of Pyramid Mts e of Animas, | Jun 1985, Barneby 17985 (NMC). A 370 km range extension to sw for a species previously known only as an endemic of central ASTRAGALUS FUCATUS Barneby. Socorro Co.: 22 km n of Magdalena, 8 May 1977, S 4653 (NMC). A single plant in heavily grazed grassland, sandy soil, ca 200 km to se of localities in San Juan Co. in M&H and Barneby (1964). DaLEA PURPUREA Vent. vat. PURPUREA (= Petalostemum purpureum {Vent.} Rydb). Grant Co.: 3.2 km s of Silver City on Hwy 90, 26 Apr 1983, W 10777 (NMC, UTEP). A plains species mapped as far sw in NM as the central counties by M&H; this a range extension of 250 km to the sw, where probably introduced. LatHyrus *Latiro.ius L. Otero Co.: Cloudcroft, 19 Oct 1969, Smart 84 (UTEP); Haynes Canyon, Gonzalez 44. Lincoln Co.: 3 km nw Ruidoso, Berry 19. Sierra Co.: Kingston, W 13302. This is a common escape from cultivation and is used in roadside stabilization in many parts of the w U.S. PROSOPIS GLANDULOSA Torr. var. GLANDULOSA. Dona Ana Co.: 5 km e of Akela on I-10, 16 Apr 1984, S & Leva 7797 (NMC). Along freeway right-of-way where it receives somewhat more water than the sympatric var. torreyana (L. Benson) M. C. Johnston. M&H map the variety as far w as Lincoln and Eddy cos. in NM, and Johnston (1962) does not indicate it to occur in NM at all. It is a common ornamental in Las Cruces, a possible source of seed for plants along the highway. *SCHRANKIA UNCINATA Willd. Hidalgo Co.: along I-10 at Lordsburg, 5 Apr 1983, W 10688 (UTEP); Peloncillo Mts, Granite Gap, W 10703. Grant Co.: 3.2 kms Silver City, W 10779. Known from much of central U.S., probably introduced in sw NM through hay to control roadside ero- sion. Vicia *vittosa Roth. Doha Ana Co.: College Farm (in Mesilla Park), both 16 Apr 1895, Sweet s.n. and Peacock s.n. (NMC). M&H indicate ina footnote that species “probably occurs near cultivated areas” in the state. 462 Apparently has not been seen in NM since these two simultaneous early records, probably collected by students of Wooton. HALORAGACEAE MyriopHyYLLUM *pPINNATUM (Walt.) B.S.P Harding Co.: 1 km ne Abbott Lake, 1 Jul 1981, F GS 5339(UNM). A species widespread to the east. LAMIACEAE *CALAMINTHA ARKANSANA (Nutt.) Shinners. Otero Co.: Sacramento Mts, T15S, RIZE, $22, | Aug 1971, Todsen 5.n. (NMC); e of Cloudcroft, T15S, RIZE, $33, 7 Oct 1978, F 5.x. (ALBU). Nearest known localities are apparently in central TX. HepEOMA *DENTATUM Torr. Hidalgo Co.: Peloncillo Mts, Skeleton Canyon, 6 Sep 1981, S$ 6287-B (NMC, NY); 15 air km ssw of Animas, W 13395. Irving (1980) shows this species to approach NM only a few km to the w in se AZ. MONARDA CITRIODORA Ceryv. var. cIirRIODORA. Hidalgo Co.: Peloncillo Mts, Granite Gap, Hwy 180, 6 Jun 1983, W 10705 (NMC, UTEP). A species of s-c U.S. reaching se NM (Scora 1967), possibly introduced in sw NM through hay used in control of roadside erosion. LILIACEAE ALLIUM *ACUMINATUM Hook. San Juan Co.: above e side of Navajo Lake adjacent to sandstone rimrock, T30N, R8W, S25, 2 May 1984, K 3031 (UNM). Hidalgo Co.: 10 km ne of Virden, S 8408. A species wide- spread to the n and w. ALLIUM GOODDINGU M. Ownbey. Lincoln Co.: Sierra Blanca, 7 Jul 1977, K 148 (UNM) with subsequent independent collections by each of us since then. Known for a number of years in c AZ, and for about a decade in w-c NM. Extension ca 240 km se of closest populations, where it is sporadic in moist areas in spruce/fir forest on ne slope of the mountain. This population has passed as A. brevistylum Wats. in M&H. * ASPHODELUS fistuLosus L. Luna Co.: 34 km w of Deming on I-10, 19 Apr 1981, § 5975 (NMC, NY, TEX); W 10783, 11848. A Mediterra- nean plant common as a weed along roadsides in nc Mex. Also introduced ins CA (Munz 1968). LINACEAE Linum Lewisn Pursh. This perennial ts widespread in the west. In the sw corner of NM there are annual plants with styles about 3 mm long or 463 less, shorter than or equal to the anthers. These plants do not have the erose or ciliate inner sepals characteristic of L. angustifolium Huds. or L. usitatis- simum L. They will "key” to L. pratense (Nort.) Small in Correll & Johnston (1970) and in M&H, and match the description of this species rather well. Linum pratense is recorded in M&H only from extreme e-c NM. Observa- tions in Hidalgo Co., near Cloverdale, by Spellenberg indicate that populations are entirely annual and short-styled. To the north, near Virden, perennials and plants blooming their first year grow together, less than 10% are short-styled. These populations of short-styled annuals may be arid grassland ecotypes derived from adjacent perennial L. /ewzs71, and probably do not represent L. pratense of the Great Plains. Records are: Hershey 141, S 3070, Hess & Sticky 3446, NMC, all from Hidalgo Co., and Metcalfe 1580, NMC from Sierra Co. The phase also barely enters AZ in se Cochise Co. w of Guadalupe Cyn. (F 3187, ALBU). MALVACEAE ABUTILON *sONORAE Gray. Hidalgo Co.: Peloncillo Mts, Skeleton Cyn., 6 Sep 1981, S 6294 (NMC). Luna Co.: s end Florida Mts., McKin- tosh s.n. (NMC). Formerly known from w TX, n MEX, se AZ. *AVLTHAEA ROSEA (L.) Cav. Dofia Ana Co.: Las Cruces, Del Rio Drainage Canal, 20 Sep 1984, Green s.n. (NMC). Lincoln Co.: Ruidoso, Berry s.n. (UTEP). Sparingly established at both sites; young plants appar- ent. NYCTAGINACEAE BOoERHAVIA spicATA Choisy. Many of the collections of spicate Boerhavia from sandy areas in NM will key to B. spicata in Kearney & Peebles (1960) or in Standley (1918). These are comparatively robust plants with densely-flowered racemes, ovate floral bracts that are as long as or longer than the ovary at anthesis, and that have stems that are viscid- pubescent below the inflorescence. They will key to B. torreyana (Wats.) Standl. in M&H, acommon form with narrower, shorter bracts, and usual- ly with less densely pubescent stems. The extremes of this form are on the rockier sites. If the two forms are kept as separate species, then B. spicata occurs in NM; if not, then all forms are included in B. spicata, the older name, a taxonomy followed by C. E Reed in Correll & Johnston (1970). ONAGRACEAE CALYLOPHUS *BERLANDIERI Spach. subsp. pinirotius (Engelm. ex Gray) Towner. Socorro Co.: 13 kin s of San Antonio along 1-25, 24 May 1983, § 7017 (NMC, MO). San Juan Co.: Little Water, road shoulder, S 464 8492. About 160 km to w of known range of species in NM (subsp. berlandier?) and a state record for the subspecies, a native of central Oklahoma and Texas (Towner 1977). It may have been introduced in hay spread on road banks and shoulders. Species not in M&H. CALYLOPHUS SERRULATUS (Nutt.) Raven. Hidalgo Co.: Peloncillo Mts, Granite Gap, Hwy 80, 5 Apr 1983, W 10708 (NMC, UTEP). Mapped mostly e of mountains in NM by M&H, and shown to "skip” to e AZ in Towner (1977). Possibly introduced in hay to control roadside ero- sion, OROBANCHACEAE OROBANCHE LUDOVICIANA Nutt. var. *ARENOSA (Suksd.) Crong. San Juan Co.: ca 6 km n of La Plata, 11 Aug 1981, S 6119 (NMC). Confirms presence for state; said to be expected by M&H (as 0. mu/tiflora). OXALIDACEAE Oxatis *pitosa Nutt. Hidalgo Co.: Peloncillo Mts, Guadalupe Cyn, T348, R21W, S6, 7 Apr 1979, § 5072. Infrequent; a species of AZ, CA, and SON. POLEMONIACEAE [PpOMopsiIs CONGESTA (Hook.) V. Grant. Sandoval Co.: 35 km nw of San Ysidro, 4 Apr 1984, S 7789 (NMC, ID, NY, CSU, ASU, UNM). Dominant on clay road bank, about a 100 km se range extension from San Juan Co. PHLOX *VARIABILIS Brand. Taos Co.: Little Costilla Peak, 29 Jul 1982, F 6396 (ALBU, NMC). Previously known from the high mts of CO and UT. POLYGON ACEAE ERIOGONUM *FrLexum M. E. Jones. San Juan Co.: 10.5 km s of Water- flow, Navajo Mine, 5 Jun 1985, § 8222 (NMC, NY, RSA, UNM). Rare on steep, n-facing clay slope. Reveal and Ertter (1976) map this species (in Stenogonum) very near to the border of NM in sw CO. Oxyria DIGYNA (L.) Hill. Otero Co.: Sierra Blanca, cliff face at 3350 m elev, 16 Aug 1980, W 6355 (COLO, UTEP). M&H indicate this occurs in NM only tn high mountains of northern counties; this a range extension of ca 400 km to south. RANUNCULACEAE DeLeHinium *ayacis L. Eddy Co.: ca 30 km e of Carlsbad, roadside 465 below Maroon Cliffs, 14 May 1979, K 878 (UNM). A garden escape, na- tive of Europe, found occasionally in several states in U.S. RANUNCULUS *ABORTIVUS L. vat. ABORTIVUS. Taos Co.: Rio Santa Barbara 5 km airline se of Penasco, 4 Jul 1982, F 6252 (ALBU, UNM). A transcontinental species. ROSACEAE POTENTILLA CONCINNA Rich. in Frankl. var. *RuBRipEs (Rydb.) C. L. Hitchc. Taos Co.: Latir Peaks, 1 Aug 1979, F 4062 (ALBU, UNM). An alpine variety widely distributed in w N. Amer. UBUS *pROCERUS P. J. Muell. Sierra Co.: Kingston, along N. Percha Creek, 23 Aug 1980, W 6392 (UTEP). Catron Co.: Glenwood, weed along road and in creek, S 7742. Doma Ana Co.: Organ Mts, Haynor Resort (S visual, w/o record). A blackberry adventive in AZ (Kearney and Peebles 1960) and a well established and unwelcome weed in the Pacific States. SCROPHULARIACEAE CoRDYLANTHUS *LAXIFLORUsS Gray. Grant Co.: 5 km n of Redrock, single plant in heavily grazed area, 7 Sept 1985, $ 8286 (NMC, NY, UC). Species previously known only from AZ. Linaria vutcaris Mill. Lincoln Co.: s of Nogal along Hwy 37, 25 Aug 1984, Soreng 2001 (NMC). An aggressive Eurasian weed 200 km s of the two NM counties where mapped by M&H. Found throughout much of the U.S.; to be expected in much of NM. PENSTEMON BRIDGESI! Gray. San Juan Co.: 6.4 km n of La Plata, 11 Aug 1981, S 6126 (NMC, NY). Kearney and Peebles (1960) indicate this species to occur from sw CO across much of n AZ. It was known from NM only from one old record in extreme w Catron Co. nearly 300 km to the south, SOLANACEAE SOLANUM AMERICANUM Mill. Hidalgo Co.: Peloncillo Mts, Guadalupe Cyn, 14 Aug 1979, S & Repass 5276 (NMC, NY). Dona Ana Co.: Organ Mts, W 6594. These records extend the known range ca 300 km to s from central NM and central Arizona. SOLANUM *CAROLINENSE L. Mora Co.: 29 km w of Roy, 3 July 1981, Ward et al. 81-209 (NMC, NY). M&H indicate the species is to be ex- pected in the extreme se counties. TY PHACEAE TyeHa *DOMINGENSIS Pers. San Juan Co.: Navajo Coal Mine, ca 5 km 466 s of Fruitland, 8 Sep 1983, S & Ward 7585 (NMC), where it grew with T. latifolia L. Correll & Correll (1972) indicate the species to occur across the s U.S., but M&H do not include the name. ULMACEAE Ceitis *occIDENTALIS L. Harding Co.: along the Canadian River 11 km airline w of Mills, 3 Jul 1981, F & K 5353 (UNM). A western exten- sion from TX & OK, possibly escaped from cultivation. VIOLACEAE Viota *viarum Pollard. Eddy Co.: 46-59 km sw of Whites City in Guadalupe Mts, shaded, moist crevices of limestone in Big, Black, and Gunsight canyons, 27 Mar 1982, K 1822, 1823 (UNM). A w range exten- sion of ca 800 km from OK. The plants from this region have been previ- ously misidentified as V. missouriensis; R. Barneby kindly identified Knight’s material. The Guadalupe Mts presently harbor several relict species of the southeastern hardwood forest, and this violet may have been part of a more inclusive flora of such affinity. Plants are consistently differ- ent from those of more eastern populations, and study may show them to be taxonomically distinct. APPENDIX Additional records and literature citation for distribution records of plant taxa reported for New Mexico. This list serves to up-date Martin & Hutchins (1980, 1981) but does not repeat range extensions or new species mapped in Fletcher, et al., 1984. For taxa new to the state or otherwise not included in Martin & Hutchins (1980, 1981) an asterisk appears next to the name of the taxon added to the flora. Carlowrightia *texana; *Justicia u ies * Tetramerium NEYVOSUM; tr “1984, APIACEAE: Lomatinm *nevadense var. parishii—Mathias & Con- stance 1945, ees sae subsp. parishii—Chuang & Cie 1969. Preryxia “onstance 1945. APOCYNACEAE: Amsonta “in Melauh ra ASCLEPIADACEAE: *Cynanchinm arizonicum—Yodsen 984. ASTERACEAE: Chaetopappa *elegans—Soreng & Spellenberg 1984; *C. ersheyi—Shinners 1946. C asain nauseosus subsp. *arenarius, C. mn. subsp. *nitidus—Anderson 1978; C.n. ssp. *fexensis—Anderson 1980; sige haan vat. *consimilis—M & Wagner 1981; E. *eximius—Nesom 15 982; Evax sprotfore—Spellenberg 1984; Helianthus *para eile t “al 1981; ees vta *arizonittca—Tomb 1980; Parthenium *alpinum vat. pe len 1984; erityle *lemmoni (Gray) MacBride—Todsen 1973; P teh eae var. Ui camera 81 _ Se fase es Marley & Wagner 19 EAE: Cryptantha * baker *oracilis; C. ee i Sites Sivin see in press. Eee CACEAE: hse minus vat. micranthtum—Hartm t al. 1980; * Diplotaxis muralis—Marley & Wagner 1981; Lepidinm oblongum— Wagne 198 35 Magis Re ace aaes & Wagner 1981; Thelypodiopsis *purpustimWard & Spellenbere 1981. *BROMELIACEAE: Tillandsia 467 recurvata—Wagner 1979. CACTACEAE: ges enneacanthus var. *enneacanthus, Opuntia *ficus-indica—Benson 1982. CA ULACEAE: *Nemacladus glenda var. orientalis—Ward & Spellenberg ne CAPPARACEAE: Polanisia sala ae : 1979. CARYOPHYLLACEAE: Arenaria *stricta subsp. texana— FlS35 hae *nitens—Soreng & Spellenberg 1981. ia ee DIACEAE: iia andes 1985. Suaeda suffrutescens vat. aes a—Hopkins & Blackwell 1977. COMMELI- ACEAE: Tradescantia *wrightit spent op. CONVOLVULAC one Ipomoea he —Soreng & Spellenberg 198 ROSSOSOMATACEAE: *Apacheria chirtcahu- nsis—Knight et al. 1984; i ae ‘lantan-—oks 1984. CUPRESSACEAE: ee *erythrocarpa—Adams & Zanoni ch CYPERACEAE: Cyperus *ovularis var. cylindricus—Kessler 1984. | EUPHORBIACEAE: ie thus *abnormis vat. abnormis—Urbatsch et al. 1975. ee are cobrensis var. perme 1984; A. *monumentalis—Marley & Wagner 1981; A. puniceus cf. var. *punieus; A. *sericoleucus—Spellenberg 1984; Dalea *cylind Pan) 77; * Tephrosia tenella—Todsen 1984. HYDROPHYLLACEAE: Nama hispidum var. *mentzelit, *retrorsum, N. *stevensii—Bacon 1984; Phacelia *bakeri—Atwoo 1975, P *demissa—Soreng 1984a; P. integrifolia var. *texana—Atwood 1975, P. *splendens—Knight 1984. LAMIACEAE: Hedeoma *reverchonit var. serpyllifolium; H. Naeem: 1980; Monarda punctata vas. *occidentalis (replacing var. /asiodonta, mis-mapped for NM in M&H)—Scora 1967; Salvia *microphylla var. wislizenti—Marley & Wagner ‘1981: ; *summa—Worthington 1982. LOASACEAE: Mentzelia eee ne LYTHRACEAE: *Nesaea Jongipep—Graham 1977. MAL HIACEAE: ee *hirtella—Todsen 1982. MALVACEAE: Sphaeralcea ‘ebrone Duke 1985 MARTYNIACEAE: Proboscidea *sabulosa—Bretting cfs HIOGLOSSACEAE: Boas chium *matricartifoliam—Spellen ae erg 1978. POAC ve rrostis gles) A lopecurus * pratensis; Apera *interrupta; Aristida St nit; seeays *bladhii, B. *ischaemum vat. songarica; *Brachiaria ciliatissima, B. *texana; Bromus ae a mollis; Chloris *submutica; Dichanthelium *linearifolium; Echinochloa *crus-pavonis, E. murtcata vat. i *microstachys. _ om. var. *muricata; Elymus *saundersti; Elytrigia *pontica; Eragrostis */utescens, E. *palmeri, E. *superba; Pie *tritweum; Festuca *arundinacea, F. ovina vat. *rydbergii; Hilaria *rigida; Leymus bvee _ L. *salinus; Muhlenbergta *villosa; Panicum *antidotale, P. *coloratum, P. *hillmanii, P. *straminenm; Paspalum *setaceum vat. setaceum; Pennisetum *ciliare; Poa *arachnifera, “ * bulbota: P. fendleriana var. *albescens, P. *trivialis; *Psathyrostachys juncea; *Rhynchelytrum repens; *Sclerochloa dura; Setaria *leucoprla, S. *ramiseta; Spor ne ean *Urochloa panicoides—Allred et al. 1986. POLEMONI- AE: ee longiflora subsp. *australis—Fletcher & Wagner 1984; 1984b POLYGALACEA innata—Sore : OLYG pA: Polygala rimulicola var. oe & Todsen 1982. POLYGONACEAE: Eriogonum *gliquantum—Reveal 1976, E. *atrorubens var. atrorubens—Soreng ; *cordonii—Reveal 1976; E. *hookeri—Soreng 1984a; E. — *mucrothecum, — E. *palmerianum—Reveal 1976, E. *scabrellum—Soreng 1984a; E. *schockleyi; E. P ; *umbellatum—Re ‘veal 1976. POLYPODIACEAE: oo oe 1979). AE: Ranunculus Sila teae arley & Wagner 1981. ROSACEAE: Cercocarpus *intricatus—Marley & Wagner 1981. RUBIACEAE: Galinm *emeryense subsp. emeryense—Wagner 1983. SAXIFRAGACEAE: eae *¢lomerulata—Todsen 1982. SCROPHULARIACEAE: *Mecardonia —_ vandellioides—Yodsen 1984 Penstemon eae Aaa 1966. § NACEAE: Chamaesaracha *pallida—Averett 1973. VERBENA EAE: Glandularia bipinnatifida var. *brevispicata, G. *chirtcahensis—Umber 1979. rer Viola *pratincola—Russell 1965 468 REFERENCES ADAMS, R. P. and T. A. ZANONI. 1979. The distribution, synonomy, and taxonomy of three junipers of southwestern United States and northern Mexico. Southw. Naturalist 24:323 — 329. ere K., S. HATCH, and R. SORENG. 1986. A verified checklist of the grasses of w Mexico. New Mexico Agric. Exp. Sta. Res. Rep. 579. oe . C. 1978. New taxa in Chrysothamnus, section Nauseosi (Asteraceae). hye rr 210 dae FA . 1980. eo and biogeography of sy asec NAUSCOSUS SSP. texensts (Asteraceae): a new Guadalupe Mountains endemi w. Naturalist 25:197 — 206. ATWOOD, N. D. 1975. A revision of the eee ree group (Hydrophyllaceae) for North America. Great Basin Naturalist 35:12 saree - E. 1973. Biosystematic study 7 Chamaesaracha (Solanaceae). Rhodora 75:32 BACON, a 1984. Chromosome numbers and taxonomic notes in the genus Nama (Hydrophyllaceae). II. Sida 10:267 — 275 a R. C. 1964. Atlas of North American Astragalus. Mem. New York Bot. Gard. 13:1-— 1188. es on Daleae Imagines. Mem. New York Bot. Gard. 27:1 — 892. BENSON, L. 1982. The cacti of the United States and Canada. Stanford Univ. Press, Stanford. ix + 1044 pp. BRETTING, PK. 1982. Range extension for Proboscidia sabulosa Correll (Martyniaceae), a sand dune endemic. Southw. Naturalist 27:228. BR S, R. E. 1984. An radios to the New Mexico flora, G/ossopetalon planitierum (Crossosomé ae Sida 10:321— 322. CHUANG, T. - CONSTANCE. 1969. A systematic study of Perideridia (Umbel- Iherae — rene - Calfi. cor oe Per aaa: CORRELL, D. S. & H. B. CORRELL. 1972. Aquatic and wetland yon of - south- western United States. - S. Govt. rinsing Office, Washington, D.C. xv + 1777 pp. CORRELL, D. S. & M. C. JOHNSTON. 1970. Manual of the vascular ine ] Texas. Texas Research Bowidation, Renner. xiii + 1881 pp. CROSSWHITE, E S. 1966. Revision of the Penstemon section Chamaeleon (Scrophulari- aceae), Sida 2:339 — 346. DANIEL, T. E 1984. The Acanthaceae of the southwestern United States. Desert Pl. 5:162— 17 ECOLOGY CONSULTANTS, INC. 1978. An illustrated guide to the proposed threatened and endangered plant species in Colorado. U.S. Dept. Interior, Fish & Wildlfie Service, Denver. 114 pp FLED ey R. B. ISAACS, P KNIGHT, W. MARTIN, D. SABO, R. SPELLENBERG, T. TODSEN. 1984. A handbook of rare and endemic plants of New Mexico. Univ. New Mexico Press, Albuquerque. xvii + 291 pp. FLETCHER, R. and W. L. WAGNER. 1984. A new are of Ra a longiflora enone from Arizona, New Mexico, and norther Madrono 31:20 — 23. GRAHAM, S. A. 1977. The American species of Nesaea (Lythraceae) and their relation- ships to Hermia and Denon Syst. Bot 2:61 —7 HARTMAN, R.L., B. E. ieee oe K.H. DUEHOLM. 1980. Noteworthy records (Wyoming). Madrono 7: 180 — HERMANN, E J. 1970. Manual a - carices of the Rocky Mountains and Colorado Basin. U.S. Dept. Agric., U.S. Forest Serv., Agric. Handbook 374. 1— 394 pp 469 HOPKINS, C. O. and W. H. ame JR. 1977. Synopsis of Svaeda (Chenopodi- aceae) in North America. Sida 7:147 IRVING, R. S. 1980. The systematics of Hedeoma (Labiatae). Sida 8:218— 295. ISELY, D. 1973. Leguminosae of the United States: I. Subfamily Mimosoideae. Mem. New York Bot. Gard. 25:1— 152 —____.. 1984. Astragalus L. ia Papilionoideae) II: Species summary A— E. lowa Sere J. Res. 59:99 —2 JOHNSTON, M. C. 1962. The North American mesquites, Prosopis sect. Algarobia (Leguminosae). Brittonia 14:72 —90. JONE hie 1983. | transfer from Aster to Machaeranthera (Asteraceae). . 8:85. KEAR NEY, T H. and R. H. PEEBLES. 1960. Arizona flora (2nd ed. with supplement). Univ. Calf. Press, Berkeley. viii + 1085 KESSLER, J. W. 1984. Cyperus ovularis (Michx. i" Torr. var. cylindricus (Ell.) Torr. new to KNIGHT, P J. 1984. oeaoehy collections (New Mexico). Madrofio 31:63. KNIGHT, P. J., C. R. WAHL and R. FLETCHER. 1984. Noteworthy collections (New Mexico). Madrono 31:63. LA DUKE, J. C. 1985. A new species of Sphaeralcea from New Mexico. Southw. Naturalist 30:433 — 436. LEVIN, G. A. 1985. Noteworthy collections (New Mexico). Madrono 32:192. LIPSCOMB, ie L. 1984. New additions or otherwise noteworthy plants of Texas. Sida 10:226 — MARLEY, a A. and W. L. WAGNER. 1981. Noteworthy collections (New Mexico). Madrono 28: 41 — 43 MARTIN, W. C C mle R. HUTCHINS. 1980, oe 1 & 2, respectively). A flora of New Mexico. J. Cramer, Germany. xi + 2591 MATHIAS, M. E. and L. CONSTANCE. 1945. Umbelliferae (pars). North American Flora 28B. McLAUGHLIN, S. P. 1985. A new species of Amsonia (Apocynaceae) from central New Mexico. Southw. Naturalist 30:563 MUNZ, P A. 1968. A supplement to a flora oF California. U. C. Press, Berkeley. 224 p NESOM, G. L. 1978. Chromosome numbers in Erigeron and Conyza (Compositae). ae Foo [= 38 . 1982. Systematics of the Erigeron rusbyi group (Asteraceae) and delimitation of sect. Peregrinus. Syst. Bot. 7:457 — 470. REVEAL, J. L. 197 34:409 — REVEAL, J. L. and B. J. ERTTER. 1976. Reestablishment of Stenogonum Nutt. (Polygo- naceae). Great Basin Naturalist 36:220— 280. RUSSELL, N. H. 1965. sles (Viola) of central and eastern United States: an intro- ductory survey. Sida 2:1— 113 76. Eriogonum (Polygonaceae) of Arizona and New Mexico. Phycologia SCORA, R. 1967. interspecific relationships in the genus Monarda (Labiatae). Univ. Calif. Publ. Bor. 41:1 SEILER, G. J., L. CUK and C. E. ROGERS. 1981. New and interesting distribucion records for Helzanthus aan Heiser (Asteraceae). Southw. Naturalist 26:43 1 — 432. 1946 SHINNERS, L. H. . Revision of the genus Chaetopappa DC. Wrightia 1:63 —81. care ne R. C. - ae New Cryptantha (Boraginaceae) records for New Mexico. . Natura Se R. ae ‘orewordy collections (New Mexico). Madrono 31:126. 470 oe: ee [pomopsis pinnata (Polemoniaceae) in the United States. Madrofio 31:189 — SORENG, % o R. amiga 1981. Noteworthy collections (New Mexico). Madrono 28:87 te 4. An unusual new Chaetopappa (Asteraceae — Astereae) from New Mexico. Syst. Be ae SPELLE ae R. 1978. New plant distribution records from the southwestern United States and northern Mexico. Madrono 25:169— 170 1979. Chromosome numbers from some derail proposed threatened or eddanuered southwestern angiosperms and other miscellaneous taxa. Southw. Natura- list 24:187 — 186 «1984 £ Morewerthy collections (New Mexico). Madrofio 31:193, STANDLEY, P G 1918. Alliontaceae. North American Flora 21, pt. 3, 171-254. TODSEN, T. K. 1973. Comments on New ae flora. Sida 35. 1982. en collections (New Mexico). Madrono 29: 60. 1983. A new variety of Perityle staurophylla (Asteraceae) from New Mexico. Madrono s l : = Ll. 4. Noteworthy collections (New Mexico). Madrofio 31:125. TOMB, A. . 1980. eoeee ot Lygodesmia (Asteraceae). Syst. Bot. Monogr. 1:1—5 TOWNER, H. E 1977. The biosystematics of Calylophus (one raceae). Ann. Missoni Bot. Gard. 64:48 — ns UMBER, R. 1979. The genus Glandularia (Verbenaceae 4:72— 10 vee it E., J. D. BACON, R. L. HARTMAN, M. C. JOHNSTON, T. J. WATSON, JR., and G. L. WEBSTER. 1975. Chromosome numbers for North Ameri- can Euphorbiaceae. Amer. J. Bot. 62:494 —500 VAN DEVENDER, T. R. 1980. Helpeae plant remains from Rocky Arroyo and Last Chance Canyon, Eddy County, New Mexico. Southw. Naturalist 25:361— 372. WAGNER, W. L. 1979. New records to the Animas Mountain fom: New Mexico. Southw. Naturalist 24:291 — 296. ————. 1983. Noteworthy collections (New Mexico). Madrofo 30: 126. WARD, D. and R. SPELLENBERG. 1981. Noteworthy collections (New Mexico). Mad- rono 28:185 — 186. WEBER, W. A. 1976. Rocky Mountain Flora. Colo. Assoc. Univ. Press, Boulder. xii + 479 pp. WENDT, T. and T. K. TODSEN. 1982. A new ius of Polygala rimulicola (Poly galaceae) from Dona Ana Co., NM. Madrono 29:19 — 2 ON, J.,R. SPELLENBERG, and H. OLE 1979. Noteworthy collections (New Mexico). Madrono 26:4 WOOTON, E. i P. . STANDLEY. 1915. Flora of New Mexico. Contr. U.S. Natl. Herb. 19:1— WORT ine on - D. 1982. Noteworthy collections (New Mexico — Texas). Madrono 29:217 YORK, i C. and W. A. DICK-PEDDIE. 1969. Vegetation changes in southern New Mexico during the past hundred years. Pp. 155 — 166. In: W. G. McGinness and B J. Goldman, eds. Arid lands in perspective. Univ. Ariz. Press, Tucson. wee in North America. Syst. Bot. TAXONOMIC AND NOMENCLATURAL NOTES ON HOUSTONIA NIGRICANS (RUBIACEAE)! EDWARD E. TERRELL Department of Botany, University of Maryland College Park, MD 20742, U.S.A. ABSTRACT Herbarium and field studies on Florida and Texas representatives of Houstonia nigricans sens. lat. resulted in che recognition of three varieties in Florida: var. nigricans, vat. dana, and var. pulvinata. Taxonomic and nomenclatural sau one the name H. filifolia and its parent combinations are discussed, and these names and the Texas taxa H. salina and H. tenuis are relegated to synonymy under vat. nigricans. A key to Florida varieties and a nomenclatural summary are provided. INTRODUCTION Houstonia nigricans (Lamarck) Fernald is a perennial, tap-rooted, polymorphic species occurring from southern Michigan to southern Flori- da and the Bahamas, west to eastern Colorado and Arizona, and south through northern Mexico. Over this wide range it grows 1n a great variety of habitats, including prairies and plains in the central United States, shale outcrops in southern Ohio, cedar glades in Tennessee, and sandy sea coasts in Florida. These notes deal with varieties and so-called species from Flort- da and Texas related to H. nigricans and conclude that var. mgricans, vat. floridana, and var. pulvinata should be recognized in Florida. Houstonia filifolia and its parent combinations are relegated to synonymy under var. nigricans, as are also H. salina and H. tenuts. Terrell (1975) discussed generic delimitations and pointed out the dis- tinctness of the type species of Hedyotis, Houstonia, and Oldenlandia and their heterogeneity if all included under Hedyorzs. Furthermore, recent re- search (Terrell et al. 1986) shows that Howstonia itself in North America includes several distinct groups of species based on seed and pollen mor- phology and chromosome number. The present paper deals with Houstonia, which is restricted to North America, while Hedyoris has an Asian type and is centered in Asia. ‘Scientific Article No. A-4380, Contribution No. 7369 of the Maryland Argicultural Experiment Static SIDA 11(4):471—481. 1986. 472 MATERIALS AND METHODS Herbarium and field studies were carried out on Howstonia nigricans sens. lat. using standard taxonomic methods and emphasizing morphological, chromosomal, ecological, and geographical data. Taxa described from Florida and Texas were the main objects of study. Specimens examined to compile the data appearing in the tables were lent from herbaria at FSU and US. Other specimens came from the herbaria cited in the nomenclatural summary. Additional Florida collections were recorded during visits to FLAS and USE Field observations and collections in Florida were accom- plished in 1965 and 1980. Samples of the handwriting of A. W. Chapman were accessed at NA. RESULTS AND DISCUSSION HOUSTONIA FLORIDANA In 1918 Standley described H. floridana from the vicinity of Miami, Florida (see nomenclatural summary below). The protologue and type specimens clearly apply to a group of populations that differ from typical H. nigricans primarily in having globose or subglobose instead of turbinate or oblong capsules. These populations are restricted almost entirely to oolitic limestone outcrops and soil over limestone in open places and rocky pinelands in Dade County from the Miami area south to Everglades Na- tional Park, in Monroe County on Big Pine Key and Key West in Florida (Fig. 1), and in the Bahama Islands (Abaco and Grand Bahama Islands). The name here adopted for these populations is var. floridana. In contrast, Houstonta nigricans vat. nigricans occurs in Florida in coastal sands along the Gulf coast from Collier County north to Pinellas and Levy counties, in these and other habitats in several northwest counties, and in Palm Beach and Osceola counties (Fig. 1). Variety nigricans and var. floridana are allopatric. Comparison of the more important morphological characters in ten samples of var. floridana and seven of var. nigricans from the southern one- half of Florida shows some overlap except in the capsule character (Table 1). Capsules in var. floridana are always globose or subglobose (superficially suggesting a relationship to the H. purpurea group), whereas var. nigricans has turbinate or oblong capsules. Variety floridana always has filiform leaves, but var. nigricans leaves vary from filiform to somewhat wider. There appear to be clines in var. nigricans: plants from northern Florida tend to have wider leaves, taller stems, and longer capsules; thus, southern Florida populations of var. nigricans are more like var. floridana than are northern Florida populations. SIDA Contributions to Botany volume 11 (in 4 numbers) Copyright 1985, 1986 by Wm. E Mahler SMU Herbarium Dallas, Texas 75275 DATES OF PUBLICATION No. 1, pp. 1— 106: 19 Jun 1985 Wo, 2, pp. 107 =.209: 12 Dee: 1985 No. 3, pp. 255-355: 20 May 1986 No. 4, pp. 357-497: 16 Dec 1986 Index, pp. 491 — 497 For contents, see the unnumbered pages forming front cover of the separate issues. SIDA sseuxro™ Volume 1] 1985 — 1986 473 var. nigricans ® var. floridana A var. pulvinata @ Figure 1. Distribution in Florida of three varieties of Houstonta nigricans, Bah f var. floridana not shown. Subglobose capsules and other parts of plants of var. floridana from the Bahamas were illustrated by Correll and Correll (1982) under the name Hedyotis nigricans var. filifolia. Seed characters are very important to differentiate among species in Houstonia. Seeds of the two varieties of H. nigricans differ only in size. Seeds of three collections of var. floridana were 0.45—0.75 mm long and 0.3 —0.45 mm wide compared to 0.45— 1.15 mm long and 0.3—0.6 mm wide for var. nigricans for collections from Mexico and United States. Chromosome counts for var. nigricans are n=9 and 10 (Lewis 1959, 1962). The chromosome number for var. floridana is n=9 for two Dade County collections (Lewis 1962, reported as Hedyotis nigricans var. filifolia). To these may be added another recently published count of 7 = 9 (Terrell et al. 1986). 474 HOUSTONIA FILIFOLIA The epithet, /i//folia, has been used in both varietal and specific combi- nations to refer to south Florida specimens of H. nigricans with filiform leaves (see synonymy for var. nigricans). The epithet began with Chapman (1860) who described Oldenlandia angustifolia vas. fiulifolia. The protologue includes the phrase, “Flowers and capsules very small.” This could refer to var. floridana which tends to have small flowers and capsules, but so also do some plants of var. mzgricans from south Florida. Chapman's protologue also mentions obcordate capsules rather longer than the calyx teeth. This and other phrases do not differentiate var. fi/ifolia from var. nigricans, especially as Chapman did not mention the hallmark character of the Dade County- Monroe County populations (var. floridana)—the globose or subglobose capsules. The second and third editions of Chapman’s flora repeat the protologue of the first (1860) edition. In the third edition, however, Chapman (1897) changed the name to Houstonia angustifolia var. filifolia, following Gray (1884). Chapman did not designate a type specimen. To determine whether there are any extant collections by Chapman which might typify var. filifolia, 1 contacted or visited a number of herbaria, including most of those mentioned by Stafleu and Cowan (1976) as having Chapman collec- tions: AUA, BM, E, GH, K, MO, NA, NY, OXE PH, US. Results were negative from all but two of these herbaria. The GH has one collection which is discussed below. The US has three collections bearing Chapman's handwriting (verified by comparison with specimens of Chapman's hand- writing on file at the National Arboretum) and variously labelled O/den- landia angustifolia, Oldenlandia angustifolia var. filifolia, and Houstonia angustifolia vat. filifolia. These three specimens are included in Table 1 for the purposes of comparison with other collections of both varieties. All three of the collecions fit var. wigricans rather than var. floridana, as they have longer, more turbinate capsules with length-width ratios of 1.3 to 1.9. One of the three collections, US 83375, labelled as O/denlandia angus- tifolia var. filifolia from “S. Florida,” is here chosen as the lectotype of var. filifolia (see synonymy below). Consideration of Chapman's (1860) protologue and his available collec- tions suggests, therefore, chat var. filifolia applies to filiform-leaved southern Florida i ae of var. nigricans and perhaps also to var. flori- dana, indiscriminately; 1.e., Chapman did not distinguish the Dade Co.- Monroe Co. populations (var. /loridana) as being distinct from other southern Florida populations. Subsequently, two other authors made new combinations using the epi- Taste 1. Comparison of some Florida collections of var. nigricans, var. floridana, and Chapman collections (var. mgricans). Var. NIGRICANS var. FLORIDANA CHAPMAN COLLECTIONS: var. NiGricans n=7 n= 10 US 83375 US1390549 US 956984 Height (cm) 17-33 12-35 30 = 15-35 Leaf width (mm) 0.5 — 2(-5) 0.5—1.2 0.5-—1.2 51 0.4-—1.3 Corolla length (mm) 37 Ns 4— 4.8(one) 4-7 estiture pubescent to glabrate to densely lensel densely densely hirsutulous densely pubescent pubescent pubescent pubescent Anther length (mm) 1.( 0.8—1.2 -- - =12 Capsule length (mm) 1 eo a L.2= 2.5 2.0—2.8 ae ko 1.9 3.0 (mm) L320 L222 led 1.8 Pg 2.2 LO 1.7 Ll. : ea L.3—1275 13S 17 1 Wes Yarra a) shape turbinate or oblong subglobose or turbinate turbinate turbinate 2lobos globose GLY 476 thet /ilifolia. Gray (1884) in his Synoptic Flora of North America cited Chapman’s name and transferred it as Howstonia angustifolia var. filifolia. His description does not mention the subglobose capsules of var. /loridana, thus could refer to either var. /loridana or var. nigricans, although he mentions “Rocky pine barrens near the coast, Florida,” suggesting the habitat of var. floridana. The GH has on file a designated “type specimen” which matches var. floridana. The collection is labelled as Houstonia angustifolia var. filifolia and “Syn. FI. N. Amer.,” “Blodgett,” and “Key West.” None of the label- ling is in Chapman’s handwriting. Key West probably supported var. flori- dana at one time, although I have not seen any other collections from there. “Blodgett” must refer to John Loomis Blodgett (1809 — 1853), born in Massachusetts. He went to Key West in 1838 and while there sent specimens to Torrey and Nuttall. This particular specimen may indicate Gray's idea of var. filifolia, but has no connection with Chapman nor any validity as a type specimen of Chapman's name. Later, Small (1903) raised var. filifolia to the rank of species, but in- correctly cited Gray instead of Chapman. Small’s description fits var. /lor7- dana by recognizing the subglobose or globose capsules. The combinations by Gray and Small are discussed above to present a complete record of the nomenclature, but do not alter the facts that the var. fi/ifolia originated with Chapman (1860) and the application of the name depends on Chapman's protologue and type. Furthermore, adherence to the present nomenclature code does not allow the citation of “Chapman ex Gray,” as Chapman provided a full description. Standley (1918) in describing H. floridana dealt with the vars. filifolia of Chapman and Gray by relegating them to synonymy under Houstonia angustifolia (it was not until 1940 that Fernald showed that the name H. nigricans must supplant H. angustifolia). Standley listed Small’s name, H. fulifolia, as a synonym of H. floridana as to the description only. Thus, Standley reached essentially the same conclusions expressed here. HOUSTONIA PULVINATA Small (1899) described Houstonia pulvinata from St. Augustine, Florida, based on collections by Mary C. Reynolds and A. P. Garber (see nomencla- tural summary). He believed that the short leaves, congested cymes, smaller corollas, and shorter, more obovoid capsules of H. pulvinata were “sufficient to warrant its treatment as a species.” Fosberg (1954), however, merely listed H. pulvinata as a variety of Hedyotis nigricans. In addition to the type specimens six collections of Howstonia pulvinata 477 have been examined from several herbaria (cited below). Houstonia pulvi- nata is now known only from St. Johns and Flagler counties in northeast Florida along the Atlantic coast (Fig. 1). The usual habitats are beaches, tn sand, and more recent collections prior to 1984 mention sea shells and coquina rock. Because of the rapid expansion of building along the ocean front, there is some question whether H. pu/vinata should be considered threatened or endangered; however, at present there are not definitive data concerning its abundance. A 1984 collection of H. pulvinata by K. J. Wurdack has provided the first chromosome count and new information on its habitat. The locality for the collection was in southernmost St. Johns County in a vacant lot between two houses in a developing beach front community, where plants grew in sand among a dense mass of coquina shells about 100 feet from the ocean. Wurdack searched unsuccessfully for H. pxlvinata at the Bunnell locality (cited below) in Flagler County, but he did not search for it else- where in St. Johns County. Without a complete survey it is uncertain whether H. pulvinata always occurs with coquina shells, thus the role of a possibly physiologically stressful habitat is not known. To judge H. pulvinata more objectively, ten geographically scattered Florida collections (from herbarium FSU) of H. nigricans were compared with the isotype, lectoparatype, and the four recent cited collections of H. pulvinata. The results (Table 2) show that H. pwlvinata differs mainly in having a consistently subprostrate habit (described as cushion-like masses), shortened internodes and pedicels, and congested inflorescences. Other differences, including corolla size and capsule shape and size, overlap greatly with var. nigricans. Walter H. Lewis obtained a chromosome count of = LO for var. pulvi- nata from buds collected as part of Wurdack 100 (Terrell et al. 1986). Gen- erally, H. nigricans var. nigricans is known (Lewis 1959, 1962) on the basis of several counts to have both »=9 and 10; however, the »= 10 counts came from plants in Brewster Co., Texas, and Nuevo Leon, Mexico. The n=9 plants came from New Mexico, Texas, and Mexico. In addition, H. nigricans var. floridana has n=9. There are no counts of H. migricans vat. nigricans from Florida plants; consequently we do not know whether there are x= 10 plants elsewhere in Flori Considering the polymorphic nature of H. nigricans sens. lat., it seems best to recognize H. pulvinata on the varietal level. Its morphological differences are consistent but mainly in vegetative characters. It 1s allopatric (Fig. 1). The importance of the 7 = 10 chromosome number can- not be evaluated without chromosomal data on H. nigricans var. nigricans in Florida. 478 Tani 2. Comparison of 10 Florida collections of var. nigricans and 6 of var. pulvinata (see text). Var. NIGRICANS Var. PULVINATA Height or dia. (cm) abit erect or decumbent cushion-like masses Internode length (mm) —46 2-20 Leaf length (mm) = 32 a= 15 width (mm) 0.3-—2.4 0.4-—2.1 Inflor. congested somewhat to very very Pedicel length (mm) 0-3 O— 1.5 Corolla length (mm) a 3.5-6 color purplish or pink to white pinkish to white Capsule length (mm) 2.0— 3.0 2.0— 3.0 width (mm) 1.32.2 OH 2.2 L/W ratio 1251.9 ‘ ; shape oblong to turbinate, oblong to turbinate sometimes broadly or obovoid ellipsoid HOUSTONIA SALINA and H. TENUIS Houstonta salina A. A. Heller appears to be possibly an ecological equiva- lent of var. pulvinata, as it occurs on beaches along the Gulf coast of Texas. The type collection has linear-oblong leaves and somewhat congested inter- nodes, but otherwise falls within the limits of var. xigricans. There are a number of other collections from the Gulf coast of Texas including at least one from or near the type locality. Some of these collections are smaller and more congested, but others are very similar to var. nigricans from Texas and elsewhere. Generally, H. salina is variable and does not seem sufficiently distinct from var. nigricans to warrant recognition. There is a chromosome count of 2 =9 reported for Hedyotis nigricans f. salina (Lewis, 1962). Later, Lewis (1970) merely listed Houstonia salina as a synonym of Hedyotis nigri- cans. Houstonta tenuis Small was described from east central Texas. The leaves are more filiform than most Texas populations of H. nigricans. Although H. tenuis somewhat resembles H. nigricans var. floridana, the capsules in the protologue and type specimen are distinctly longer than wide, which places it in var. nzgricans. Standley (1918) listed H. tenuis, H. pulvinata, and H. salina all as synonyms of H. angustifolia (H. nigricans). A key to the three varieties and a nomenclatural summary of var. nigri- cans are provided below. Following these is an appendix of specimens exam- ined for Tables 1 and 2 479 ABBREVIATED KEY TO FLORIDA VARIETIES OF H. NIGRICANS A. Stem internodes short, 2— 20 mm long; inflorescence congested; plants in A hoa de geee) later feck otek wath eats eon aod be aaa var. pulvinata re internodes usually 6—46 mm long (or sometimes longer); 1 guia: open or at least not ane plants not matted. ature capsules subglobose, 1.0 — 1.3 times longer than wide... . var. floridana aa Mature capsules turbinate, oblong, or broadly ellipsoid, 13=2.2 times longer chan wide (Florida plants only) Dae sm NRA el a Pe ee Ae Nt Pe eet tata tie Spade tev ty. ai tgs ho ahece pei a ao nega ace oe eee var. nigricans NOMENCLATURAL SUMMARY HoustoNiA NIGRICANS (Lamarck) Fernald var. NiGRICANS, Rhodora 42:299. tee Gentiana nigricans Lamarck, Encycl. 2:645. 1788. Type: Herb. Jussieu (HOLC P; pHoTO: FERNALD 1940). Hedyotis nigricans (Lamarck) Fosberg, Lloydia re 287. — . Partial synonymy follows: Houstonia angustifolia Michau Bor. Amer. 1:85. 1803. Type: “submaritimis oridae” (HOLOTYPE: P?). aa angustifolia (Michaux) A. Gray, Pl. Wright. 2:68. ae Chamisme angustifolia (Michaux) Nieuwl., Amer. Midl. Naturalist 4:92. Olden enti var. Nee eds FI. S. U.S. 181. 1860. Type: “S. Flori- da,” Chapma yPE: US-83375). here designated. Houstonia Peaiaie var. ifolia man) A Gray, Syn. A 1(2):27. 188 Satie ae ail seeks 8s oe 1338. 1903, as “(A. Gira) Sm Hedyotis nigricans vat. Bs fe hapman) Shinners, Field and Lab. 17:168. 19 i Houstonia salina A. A. Heller, Contr. Herb. Franklin and Marshall Coll. 1:96, pl. 9. 1895. Type: TEXAS. Corpus Christi, shell deposit along beach, 31 May 1894, A A. Heller 1812 (LecroryeE: GH!; isorypes: BM!, ILL!, K!, MO!, NY!, PH!, US- 3!). Lectotype here designated. Hedyotis salina (A.A. Heller) Shinners, Field and Lab 17:169. 1949. Hedyotis nigricans f. salina (A.A. Heller) W.H. Lewis, Rhodora d 961. Houstonia tenuis Small, Fl. S.E. U.S. 1109, 1338. 1903. Type: TEXAS. San Saba Co.: San Saba, October eae ), Thurber 67 (HoLoTyPE: NY). . HoustTonia NIGRICANS var. FLORIDANA (Standley) Terrell, Phytologia 59:79. 1985. Houstonia floridana Standley, N. Amer. Fl. 32(1):36. 1918. Type: FLORIDA. Dade Co.: Cocoanut Grove, Biscayne Bay, Jul 1895, A. H. Curtiss 5484 (HOLOTYPE: US!; isotypes: FLAS!, NY!). Hedyotis purpurea var. floridana (Standley) Fosberg, Castanea 19:36. 1954. i) HousTONIA NIGRICANS val. PULVINATA (Small) Terrell, Phytologia 59:79. 1985. Houstonia pulvinata Small, Bull. New York Bot. Gard. 1:289 — 290. 1899. Type: FLORIDA. Sct. Johns Co.: St. Augustine, sandy soil, Jul 1876, Mary C. Reynolds s.n. (LECTOTYPE: ee IsoTYPE: NA!). Lectoparatypes, same locality and date, A. P. Garber s.n. (NY!, US-2!). Lectotype here designated. Hedyotis nigricans vat. pulvinata (Small) area yee 19:37. 1954. ies) 480 ACKNOWLEDGMENTS I wish to thank Kenneth J. Wurdack for his collection and cytological material of var. pulvinata. Type and other collections were lent or seen dur- ing visits, thanks to the curators of BM, FLAS, FSU, GH, ILL, K, MO, NA H, US, and USE D. H. Nicolson helpfully contributed advice on nomenclature, but is not responsible for any errors. Appreciation is also expressed for help from E G. Meyer, Julia Morton, J. L. Reveal, and R. P. Wunderlin. > APPENDIX Specimens included in Table | (all in herbarium US): Variety NIGRICANS: FLORIDA. Lee Co.: Sanibel Island, Tracy 7493; Lower Captiva Island, Brana 5709. Manatee Co.: Longboat Key, Wi/bur and Webster 2518. Pinellas 20.: St. Petersburg, Beckwith 80; Gleawater Beach, Beckwith 803. Sarasota Co.: Keys, Sarasota, Jun 1876, Garber s.n. County unknown: Tampa Bay, Ruge/ 323. Variety FLORIDANA: FLORIDA. Dade Co.: Miami, Tracy 9247; Miami, Pollard and Collins 222, south of Miami, Fosberg 56758; Homestead, Killip 44238; 6 mi east of Royal Palm Park, O'’Ne:// 7584; Sykes Hammock, Smal/ et al. 6755; between Everglades and Biscayne Bay, Curtiss 1137 (type). Monroe Co.: Big Pine Key, Ki/lip 31434, 43372, Killip and Swallen 40443 Specimens included in Table 2 (all FSU): Variety NIGRICANS: FLORIDA. Collier Co.: NW _ of Naples, ms et al, 28121. Franklin Co.: Camp Weed, Godfrey 58816; Alligator Point, Godfrey 79017. Holmes Co. 3.7 mi west of Ponce de Leon, McDaniel 4845. Jackson Co.: 6 mi south of Sneads, ce 75206. Okaloosa Co.: west of Laurel Hill, Godfrey 51075. Pinellas Co.: town of St. AS Beach, D. B. and S. S. Ward 2345. Sarasota Co.: Longboat ey Godfrey . Wakulla Co.: Shell Point, Godfrey 58899; Live Oak Point, Ward 27 —_ Variety PULVINATA: all collections seen are listed as ie but the Williamson and Crawford collections were not included in data in Table FLORIDA. St. Johns Co.: dry sand, Anastasia Island, Aug 1894, C. S. Williamson s.n. (PH); Ne asia Island, 26 Jul 1923, ). Crawford s.n. (PH); Marineland, 2 Jul 1939, W. A. Murrill 741 (GA, US); abundant on sand-shell flats behind the foredunes along the coast between Matanzas and Marineland, 5 Aug 1971, R. K. Godfrey 70664 (FSU); caespitose mats, fls. pink to white, tap root angling oan the ocean, inner dunes of coquina sands, west side of old Fla. ALA, 0.8 mi north of Flagler Co. line, $39, T9S, R31E, 10 Aug 1984, K. J. Wurdack 106 (NY, US). Flagler Co.: near coquina — quarry just east of Bunnell, flowers pink, 23 May 1979, J. Popenoe et al. 1783 (NC REFERENCES Cc pda, A. W. 1860. Flora of the southern United States. Ivison, Phinney, and Co., New York. ees a Flora of the southern United States. 3rd edition. American Book Co., New York. 48 | CORRELL, D. S. and H. B. CORRELL. 1982. Flora of the Bahama Archipelago. J. Cramer. FOSBERG, E R. 1954. Notes on plants of the eastern United States. Castanea 19:25 — 37. GRAY, A. 1884. Synoptical flora of North America. Vol. 1. Ivison, Blakeman, Taylor, and Co., New York. LEWIS, W. H. 1959. Chromosomes of east Texas Hedyotis (Rubiaceae). Southw. Naturalist 3:204 — 207. —_______.. 1962. Phylogenetic study of Hedyotis (Rubiaceae) in North America. Amer. J. Bot. “ ee 865. 2 sear pp. 1487 — 1490. In: Correll, D. S. and M. C. Johnston. Manual of he Bs plants of Texas. Texas Research Foundation, Renner, Texa SMALL, J. K. ee Udescib species from the southern United States. Bull. New York Bot. Gard. 1: — 290. e035. ce. a - southeastern United States. Publ. by author, New Yo Se ane F A. and R. S.C . 1976. — literature. Second edition. Vol. —G. Bohn, Scheltema, and Holkema, Utrec STANDLEY P 1918. Rubiaceae, Ol sae In: North American Flora pil ON Aico TERRELL, . E. 1975. Se of Hedyotis fruticosa L. to Houstonia L. and Oldenlandia L. Phytologia 31:418 — 424 TERRELL, E. E., W. IS, H. ROBINSON, and J. W. NOWICKE. 1986. Phylogenetic implications of diverse seed types, chromosome numbers, and pollen mor- phology in Houstonia (Rubiaceae). Amer. J. Bot. 73:103 a ae . ee ee DOCUMENTED PLANT CHROMOSOME NUMBERS 1986:1. MISCELLANEOUS COUNTS IN ASTER (COMPOSITAE) JERRY G. CHMIELEWSKI Department of Biology, University Le Ge Calgary, Alberta, CANADA. N4 Semple et al. (1983) stated the need for chromosome counts of asters and other genera for the purpose of determining cytogeographic patterns. These patterns in turn could be used to determine the limits of distribution of infraspecific variants; in reconstructing biogeographic histories of species; in testing hypotheses on the evolution of infraspecific polyploidy, and in biosystematic studies (Strother 1972, Semple et al. 1983). With this purpose in mind, this paper reports chromosome number determina- tions to supplement the available counts in the genus Aster. MATERIALS AND METHODS Chromosome counts were made from freshly prepared material follow- ing the procedures of Chmielewski and Semple (1983). Root-tips were tak- en from transplanted rootstocks of plants collected in the wild or from seedlings grown in the greenhouse. Mitotic metaphase cells were examined to determine the chromosome number of each individual. Voucher specimens were deposited in the Herbarium of the Department of Biology, University of Waterloo (WAT). In citation, Chmielewski is abbreviated to C RESULTS AsTeR (VIRGULUS) X AMETHYSTINUS Nutt. 2n = 10. CANADA. Onrario. Brant Co.: Tuscarora Township, 2.3 km SW of Sixty-nine Corners, C. 1768. Perth Co.: Ellice Township, Ellice Con. 4—5, 2.5 km W of Perth 12, C. 1864. oe Co.: Asphodel Township, W of Asphodel Line 6 and Hwy 7, ASTER BOREALIS (T. & G.) Daag: on = te. CANADA. SAsKATCHEWAN. 14.5 km SW of Armit, W of North Armit River, Hooper & Baker 84091315, 84091310. —, 2n = 32. CANADA. Ontario. Bruce Co.: St. Edmunds Township, Hwy 6, S of entrance to Cyprus Lake Provincial Park, C. & Ringuis 2040. Oxford Co.: East Oxford Township, intersection of Hwy 2 SIDA 11(4):483 — 485, 1986. 484 and Hwy 53, SE of Eastwood, C. 1600. Wellington Co.: Puslinch Town- ship, Puslinch 7, 5 km N 5 Gor Rd. C. 1355. SASKATCHEWAN. Nitenai River, 7 km W of Hwy 163 and Hwy 9, Hooper & Baker 84090800. ASTER CILIOLATUS Lindl. 2n=48. CANADA. SaskarcHewan. 8 km N of Usherville, coulee W of Hwy 9, Hooper & Baker 84090703. Aster LANCEOLATUS Willd. 2n=32. CANADA. Onrario. Wentworth Reg. Mun.: West was Township, Con. 6, 1.3 km W of Hamilton-Wentworth 4, C. 136( —. 2n = 48. CANADA. ‘Onna. Brant Co.: Onondaga Township, Creek Rd., 2.1 km E of McBay Rd, C. 1323. Essex Co.: Sand- wich West Township, Windsor Salt Factory, C. 1515. Halton Co.: Burlington City Limits; North Service Rd., S of Aldershot, C. & C. 1541. Kent Co.: Walpole Island, SE of town of Walpole Island, C, 1526. Nipiss- ing District: Hwy 11, 1.4 km N of Strathcona Township line, S of Temagami. C. & C. 1306. Nipissing District: Gladman Township, Hwy Ll, Marten River, C. & C. 1310. Parry Sound District: South Himsworth Township, Hwy 11, Trout Creek rest area, C. G C. 1315. Timiskaming District, Dymond Township, New Liskeard, intersection of Hwy 11 and 11B, C. GC, 1281. Timiskaming District: Casey Township, 3.4 km S of Belle Vallee, C. & C. 1291. Timiskaming District: Chamberlain Town- ship, Hwy 11, picnic area S of intersection with Hwy 112, C. & C. 1302. Timiskaming District: Hwy 11, 5.9 km S of Latchford, C. & C. 1304. Waterloo Reg. Mun.: Wellesley Township, C. 1337. Wellington Co.: Guelph Township, Wellington 72, 0.5 km SE of Speedvale Rd. C. 1350. QurBec. Hwy 101, 9.6 km N of Ville Marie, Ile de College, C. & C. 1285. 0.5 km W of Hwy 101, at curnoff to [le de College Rd., N of Ville Marie, C. GC, 1289. SaskarcHewan. 8 km SE of Somme, Hooper & Baker 84090304. Pepaw Lake, Hooper G Baker 84090306. McBride Lake, McBride Lake Rd. 1, Hooper & Baker 84090307. 3 km E and 3 km N of Somme, Hooper 84091301. Bertwell, 15 km NE Reserve, Etommami River, Hooper & Baker 84091302. 20 km S of Hudson Bay, intersection of Hwy 9 and Pepaw River, Hooper G Baker 84091310. 40 km N of Arran, Mink Creek and Woody Lake Rd, Hooper 84091505. 15 km NE of Weekes, S shore of Neely Lake, Hooper 84091601. ASTER aff. LONGIFOLIUS Lam. 2n = 64. CANADA. Onrario. Oxford Co.: East Zorra Township, NE of Cassel, C. 1607. ASTER NEMORALIS Ait. 2n = 18. CANADA. Onrario. Nipissing Dis- trict: E of Huntsville, C. G C. 1579, 1580. Aster pitosus Willd. 2n = 32. CANADA. Onrario. Gray Co.: Bentinck Township, 11 km E of Elmwood, C. 2269 485 ACKNOWLEDGEMENTS The author thanks Susan Chmielewski and Dr. Gordon Ringius for their assistance and D.E Hooper and L. Baker for collections from Saskatchewan. This work was funded by University of Waterloo and Ontario Graduate Scholarships to JGC and Natural Sciences and Engineer- ing Research Council of Canada operating grants to Dr. J. C. Semple, Unti- versity of Waterloo. REFERENCES CHMIELEWSKI, J. G. and J. C. SEMPLE. 1983. | Se of Aster lanceolatus. II]. Cytoecology in southern Ontario. Cand. J. . 61:1879 — 1886. SEMPLE, J. C., J. G. CHMIELEWSKI, and C. C. ee ee Chromosome number determinations in Aster L. (Compositae) with comments on cytogeography, phylogeny and chromosome morphology. Amer. J. Bot. 70:1432 — 1443 seagrass J. L. 1972. Chromosome studies in western North American Compositae. mer. J. Bot. 59:242 — 247. 486 NOTES PANAX QUINQUEFOLIUS L. (ARALIACEAE) REDISCOVERED IN LOUISIANA—The American species Panax quinquefolius L. commonly known as ginseng and other Asiatic members of the genus have long been esteemed for their medicinal properties (Dixon 1976). Panax quinquefolius is fairly widespread in rich cool woods of the eastern United States. Accord- ing to Fernald (1950) the species ranges from Quebec to Manitoba south to Florida and Oklahoma. Previously, ginseng was documented for Louisiana by a single specimen collected by C.A. Brown in 1938. The locality data on the specimen label is rather vague: “West Feliciana Parish in the vicinity of Plettenberg” [Brown 7293 (LSU)]. The species was recently found in the same general area. Twelve plants growing in two clumps of two and ten plants each were located on rich soil of a west-facing slope in West Feliciana Parish; aerial parts of the plant were collected to document its occurrence [private property along local road ca 5.0 mi NW of the jct. of LA 968 and LA 66; R4W, TIS, sect 47, 17 Jul 1986, Urbatsch, Meier, Cox, Lievens, and Harris 4106 (LSU)}. The ginseng plants were growing under Arundinaria gigantea (Walt.) Chapm.; Schisandra coccinea Michx. was observed nearby. Canopy trees included Liriodendron tulipifera L., Carya glabra (Mill.) Sweet, and Fagus grandifolia Ehrh. The ginseng plants were about 15 to 20 cm tall and a few had red fruit on them while others had immature or under- veloped green fruit. By a later visit to the site the plants (5 Sep 1986) had lost their fruit and their riddled leaves were senescent as evidenced by their yellow to brown color.—Lowell E. Urbatsch, Department of Botany and Albert Meter, School of Forestry, Wildlife, and Fisheries, Louisiana State University, Baton Rouge, LA 70803, U.S.A. REFERENCES DIXON, P. 1976. Ginseng. Gerald Duckworth and Co., Ltd. London. FERNALD, M.L. 1950. Gray’s manual of botany, eight edition. American Book Co. New York. MUHLENBERGIA_ FRONDOSA (POACEAE) NEW TO > LOUI- SIANA— According to Chase (1951) and Gould (1975) the range of Mubhlenbergia frondosa (Poir.) Fernald extends from New Brunswick and Maine to North Dakota, and south to Georgia and Texas. A distribution map published by Pohl (1969) shows this species to be absent from all southern states including South Carolina, Georgia, Florida, Alabama, Mississippi, Louisiana and Arkansas and it is not reported from Louisiana SIDA 11(4):486. 1980. 487 by Allen (1980). In Texas, the grass is known only from specimens taken in Dallas and Grayson counties (Correll and Johnston 1970, Gould 1975). On May 16, 1986, we collected ae specimens, McKenzie 284, LSU, (duplicates sent to FLAS, LAE TAES, US) of this species growing in shady woods in moist sandy-loam along the eens River ca 0.1 mile south of the School of Veterinary Medicine on the LSU campus in Baton Rouge, East Baton Rouge Parish. Two small populations, each containing approxi- mately 30 plants were located ca 300 meters apart. The grass was growing on a slightly elevated natural levee between the river and a constructed levee in association with Arthraxon hispidus (Thunb.) Makino, Digitaria ciliaris (Retz.) Koel., Leersia lenticularis Michx., Leptochloa attenuata (Nutt.) Steud., Panicum capillare L., P. dichotomiflorum Michx., Paspalum fluitans (EM) Kunth, Brannichia ovata (Walt.) Shinners, Polygonum a Alternanthera philoxeroides (Mart.) Griseb., Celtis laevigata Willd., Acer negundo L., Forestiera acuminata (Michx.) Poiret, Physalis angulata L., Phyla lanceolata (Michx.) Greene, Teacrium canadense L., Cephalanthus occidentalis L., Artemisia annua L., Bidens cernua L., B. frondosa L., Eclipta alba (L.) Hassk., Eupatorium coelestinum L., E. serotinum Michx., Pluchea camphorata (L.) DC., and Xanthium strumarium L. The dominant overstory plants were of Populus deltoides Marsh. and Salix nigra Marsh. The discovery site ts irregularly flooded during periods of high water and silt is deposited be- tween the river and the constructed levee. Because this species 1s common in some north central states bordering the Mississippi River (Pohl 1969), it may have been introduced to Louisiana by floodwaters carrying seeds or rhizomes. Although areas north and south of the discovery sites were searched for additional specimens, no other populations were located. These collections constitute the first record for Louisiana and extend the range of this species ca 600 km.—Paul M. McKenzie, Loutstana Cooperative Fish and Wildlife Research Unit, and Lowell E. Urbatsch, Department of Bota- ny, Louisiana State University, Baton Rouge, LA 70803, U.S.A REFERENCES ALLEN, C. M. 1980. Grasses of Louisiana. The University of Southwestern Louisiana, Lafaye CHASE, ie 1951. Rev. of Hitchock’s Manual of the grasses of the United States, 2nd ed. U.S.D.A. Misc. Publ. 200. CORRELL, D. S., and M. C. JOHNSTON. 1970. Manual of the vascular plants of Texas. Texas eran Foundation, Renne GOULD, E W. 1975. The grasses oft ais Texas A & M University Press, College Station. SIDA 11(4):487. 1980. 488 POHL, R. W. 1969. Muahlenbergia, subgenus Mublenbergia (Gramineae) in North America. Amer. Midl. Naturalist 82:5 12 — THASPIUM TRIFOLIATUM (APIACEAE) AND RANUNCULUS MARGINATUS (RANUNCULACEAE) NEW TO TEXAS—A collection of Thaspium trifoliatum (L.) Gray var. flavum Blake (13 Apr 1985, McFar- lane 30 [SBSC, SMU, TEX}) from the Damuth Nature Sanctuary near Cleveland, Liberty County is the first reported for Texas. Thaspinm tri- foliatum ranges throughout the eastern United States southwest to Louisi- ana, Arkansas and Oklahoma. Variety flavum is yellow-flowered and more common westward whereas variety frifoliatum is purple-flowered but more common eastward. (Cooperrider 1985, Gleason & Cronquist 1963, Smith 1978) Ranunculus marginatus d'Urv. was taken 28 April 1984 from a weedy area in the Armand Bayou Nature Center near Clear Lake City in Harris County (Brown 7249 [PAC, SBSC, SMU}). Because of the somewhat similar achene faces these plants keyed to R. muricatus L. (Correll & Johnston 1979), but they differ from R. muricatus in their longer peduncles, more pubescent habit, and shorter achene beaks (Keener 1976). Ranunculus marginatus 1s native to the Mediterranean region and Shinners (1962) reported it (as R. trachycarpus Fisch. & Meyer) new to North America from Avoyelles Parish, Louisiana. I wish to thank Carl S. Keener (PAC) for providing the identifi- cation of my collection.—Larry E. Brown, Spring Branch Science Center and Houston Community College, 726 Horncastle Street, Channelview, TX 77530, A REFERENCES COOPERRIDER, T. S. 1985. Thaspinm and Zizia (Umbelliferae) in Ohio. Castanea 50: Fo 119: CORRELL, D. S. and M. C. JOHNSTON. 1979. Manual of the vascular plants of Texas. Univ. of Texas, Dallas. GLEASON, H. A. and A. CRONQUIST. 1963. Manual of vascular plants of northeastern United States and adjacent Canada. D. Van Nostrand Co., New KEENER, C. S. 1976. seas in ig Ranunculaceae of the Southeastern United States. V. Ranunculus L. Sida 6:266 — SHINNERS, L. H. 1962. ees feral (Ranunculaceae) in south-central Louisi- ana: new to North America. Sida 1:104— 105. SMITH, E. B. 1978. An atlas and annotated list of the vascular plants of Arkansas. Book- store, i of Arkansas, Fayetteville SIDA 11(4):488. 1986. 489 SCUTELLARIA MINOR (LAMIACEAE) NEW TO NORTH AMERICA—Scutellaria minor Hudson, a small skullcap with three leaf shapes and a pinkish corolla with purple spots, (fig. 1) was taken in April 1984 from a roadside ditch along East Knox Street near the junction of Loop 494 & North Park Drive near Kingwood, Montgomery Co., Texas (Ward 1378; SMU). In July 1985 a larger population was found, again ina roadside ditch, along highway U.S. 59, 0.5 miles north of junction with highway 1314 near the village of Porter, Montgomery Co. (Ward 1406; SBSC). These two collections are presumably the first in North America for this small mint which ts native in Europe from the Azore and British Isles to East Germany and northern Italy (Clapham et al. 1962, Tutin et al. Lo 2) The authors wish to thank Elizabeth Messenger for the illustration and Raymond McBen of the Houston Community College for some technical assistance.—Larry E. Brown, Spring Branch Science Center, 8856 Westview Drive, Houston, TX 77055, U.S.A., and John R. Ward 16010 Buccaneer St., Houston, TX 77062, U.S.A. REFERENCES CLAPHAM, A. R., T. G. TUTIN, and E. EF WARBURG. 1962. Flora of the British Isles. Cambridge University Press, Cambridge, England. TUTIN, T. G., V. H. HEYWOOD, N. A. BURGES, D. H. VALENTINE, S. M. WALTERS, and D. A. WEBB. eds. 1972. Flora Europaea, Vol. 3. Cambridge Univer- sity Press, Cambridge, England. SIDA 11(4):489. 1986. 490 A 1 cm \ Figure |. Sevtellaria minor: A. habit, B. ovate lower leaves; C. hastate middle leaves; D. linear upper leaves; E. flower (all from Ward 1378). SIDA 11(4):490. 1980. INDEX TO VOLUME 11 Names of contributing authors are in CAPITAL letters. New names and combinations are in boldface. combination in Chimaphila Ericacen) by Laurence J. ae / combination in L ie eas by R.P. Ses “400 new species and taxonomic notes on Gentianella (Gentianaceae) in South America by James S. _ Pringle 357 — 369 A new species of Wallenia (Myrsinaceae) iti by Walter S. Judd A new variety of Hedeoma hyssopifolinm Gray (Lamiaceae) by James Henrick- son 413 — A Secret inH ste es L. (Saxifra- gaceae) by L.J. U 352 A revision of se i ee leaved Potamogeton (Potamogetonaceae) by obert R. Haynes 173 — 188 A white—flowered form of Spigelia marilandica L. (Loganiaceae) new to Tennessee by Gene S. Van Horn and 249 y Max E. Medley et al. 92 —95 Additional notes on Coreopsis by Edwin B. Smith 245 — 246 Additions and noteworthy vascular plants me oo notes by Steve L. Orz et al sal 22 Additions to the flora of New Mexico by Richard Spellenberg et al. 455 — 470 Adenophyllum 372; anomalum 376; wrightit 373, 377; var. pulcher- rimun 377 Agrostis duces 319 Aletris farinosa 2 8 Aletris farinosa, Lm as and y ‘yperus pilosus new for Florida b James R. Burkhalter 247 — = Alstroemeria psittacina 379 Alysicarpus ovalifolius 379 Amaranthus lividus 379 29, 432; herbacea 433; var. mus scabrivalvis 2 A mphibromus scabrivalvis (Gramineae) in . Lynn Calaway and ohn W. Thieret 207 — 214 An undescribed ae eae Vaccinium: Vaccinium bocatorensis (Ericaceae) by bert L. Wi 1— 443 ANDERSON, LORAN C., Forestiera godfreyi (Oleaceae), a new tia from Florida and South Carolina 1—5. ANDERSON, LORAN C., Noteworthy plants from north Florida. I]. S723 ANDERSON, LORAN C., see Danin and Anderson 318 — 324 Anisacanthus quadrifidus 286, 292; var. brev sa sees 298; var. quadrifidus 294, var. us 296; var. wrightii 296 ee ie Gee sensu lato (Acanthaceae) by James Henrickson 286-296 Aristida ramosissima 379 Asclepias viridula 379 Aster X amethystinus 483; borealis ee ciliolatus 484; hemisphaericus 380 sea ce 484; nT 380, 484: nemoralis 484; pilosus 484; simmon- dsii 3 ieeNcoR T.A., see Whetstone and Atkinson 396 BACON, JOHN D., LOUIS H. BRAGG, and GARY L. HANNAN, Systematics of Nama (Hydrophyl- — comparison of seed morphol- ogy of sects. Arachnoidea and Cinerascentia with five species of Eriodictyon and = Turricula parryt 271281 BAKER, M.M. and B.L. NER, Taxonomy of ee (Asteraceae- Eupatorieae) 300 — 3 Baptisia 429, 433; as 435; var. macrophylla 435; albescens 434 492 pee anthemidifolium 376; litcoralis 377 bo 373 F sapeuese BRAGG, LOUIS H., see a et al. 271-2 BRANDENBURG, DAVID and Ne HN W. ream see ae a Brandenburg 445 — 45° BRIDGES, EDWIN te see Orzell et al. 226231 BROWN, LARRY E., Campanula rapun- aoe (Campanulaceae) new to Texas oe ARRY Siege ol ee tri- foliatum pice and R marginatus ae ie new to Texas ee BROWN, RRY E. and JOHN R. WARD, Scutellaria minor (Lamiaceae) new to North America 489 — 490 BRYAN, HAL, see Medley et al. 92 —95 aa floridana 226 Bume ana 396 Rae sete —. a new see e by R. David "Atkinson 396 BURKHALTER, UME i$ ., Aletris farinosa, Cyperus difformis, aaa ee pilosus new for Florida 247 — 5 2 = S Cactaceae 200 CALAWAY, M. LYNN and JOHN W. THIERET, Amphibromus ang (Gramineae) tn Louisiana 207 Campan floridana 380; eae C sna rapunculoides (Campanulaceae) w to Texas by Larry E. Brown 102 Coun muricatum 380 —— 429, > nictitans var. Chimaphila oe omingensis 370 CHMIELEWSKI, JERRY G. Oc mented plant chromosome aumbers 198 Miscellaneous counts from Ontario and Quebec 251 — 253 CHMIELEWSKI, JERRY G., Docu- mented plant chromosome numbers 1986: 1. Miscellaneous counts in Aster (Compositae) 483 — 485 Cirsium muticum 226 CLEWELL, ANDRE E, Observations on umbellata subsp. the ee te Mosquitia in Honduras 258 Clintonia borealis vt COFFEY, CHUCK R. and JESUS VALDES R., = Monerma cylindrica (Poaceae: Monermeae) new to Texas Comaclinium 374; montanum 374 ee 245; bigelovii ie calliopsidea douglasi1 246; sect. Leptosyne v0 stillmanii 24 CRANFILL, RAY et al., Taxonomy, dis- tribution and rarity status of Leaven- worthia and Lesquerella (Brassicaceae) in Kentucky 189 — 19¢ Crataegus ee 380 USICK, ISON W., see Spooner ect al. 2 vs — a Cyperus difformis 247, 248; pilosus 247, 248 Dactyloctenium aegyptium 245 Dactyloctenium aegyptium (Gramineae) new to Belize by James S$. Pringle 2 ANDERSON, Distribution of Porty- laca oleracea (Por Orage subspecies in Florida 318— 3 DIAL, STEVE, see Stalter Dial 25 — 328 Dioecy in North American Cactaceae: a review by Bruce D. Parfitt 200 — 206 Distribution of Portulaca oleracea L. (Portulacaceae) subspeci¢ es In — y Avinoam Danin and Lora Anderson 318 — 324 Documented plant chromosome numbers 1985: Miscellaneous counts from Ontario and Quebec by Jerry G 3 Documented plant eiesaaeuie numbers 986: 1. Miscellaneous counts in Aster (Compositae) Dy Jerry , A new combina- tion in Chima (Bricaceae) 370 Drymaria MSCOSA Sed bee correct re distion and range ex- tension to the United States i. Bruce Parfitt and Wendy Hodgson 96-98 Drymaria viscosa 96 Dysodiopsis 374; tagetoides 374, 378 eee 371, 374, 376; decipiens 375, 376, 378; Pape. 374, 377; pinnata : glabrescens ati: tagetiflora 378 Eleocharis cortilis 380 Eremogeton 107, 167; grandiflorus 167 Erigeron 249; i dca iad 249; scaber- rimus Eriodictyon on angustifolium ers Erythronium americanum 253 FLETCHER, Svea see Spellenberg etal. 490-; Flyriella 300, ne harrimanii 308 leonensis 314; parryi 310; stantordi: 308 Forestiera godfre Forestiera godfreyi (Osc a new species ; orida and South Carolina by nC. Anderson 1 5. Bee “L. DAVID?T, see Henrickson & Flyr 107-172 FREEMAN, JOHN R., see Van Horn & Freeman 248-249 FRYXELL, PAUL A., Sidus sidarum — V. The North and Central American species of Sida Gentianella 357 Gentianella cerastioides X foliosa 363; dolichopoda 368; _ foliosa sulphurea 363; gilgiana 368; helian- themoides 368: mendocina 368; persquarrosa 368; radicata 368; riojae 368; vaginalis 368 HALL, D.W., see Wunderlin et al. 232- aes GEORGE E, see Spooner et al. 215-2 ine a he L., see Bacon et al. TL HANSEN, "4 E, see Wunderlin et al. 244 ace STEPHAN L, see Morden and Hatch 282 — 285 HAYNES, ROBERT R.., A revision of the clasping — leaved Potamogeton (Pota- 493 mogetonaceae) 173 — 188 Hedeoma hyssopifolium 413; var. chihu- ahuensis4 13 Hedyotis 400; nigricans var. floridana 40 HENRICKSON, JAMES and L. DAVID YRt, Systematics of Leacophyllum and = Eremogeton — (Scrophulartaceae) 107 — 172 HENRICKSON, JAMES, A new variety f Hedeoma si ifolium Gray (Lamiaceae) 413 HENRICKSON, nn Anisacanthus quadrifidus sensu lato (Acanthaceae) 286 — 299 HESS, biege ee a Thompson (Lem Michigan 407 — END see Parfitt and Hodgson ‘92 =D Houstonia filifolia 474; floridana 472; nigricans 471, 479; var. floridana 479; var. pulvinata 479; pulvinata 476; salina 477; tenuis 4 oe 352; arborescens f. carnea ae eee e sibthorpioides 380 Impatiens balsamina 380; capensis 251 ISELY, DUANE, Notes about Psoralea sensu auct Amorp Baptista, Sesbanita aad Chamaecrista (Legumi- nosae) in the southeastern United States 429 — 440 Isotria verticillata 380 JUDD, WALTER S., A new species of eae (Myrsinace ae) from Haiti Bo 33 JUDD. WALTER S., see Zona and Judd 417 —427 Juncus capitatus 102 KEENEY, TONEY M. and BARNEY L. LIPSCOMB, Notes on two Texas plants 102 — 103 KEIL, DAVID J., Synopsis of the Florida (Asteraceae) 3 395 KNIGHT, ae see Spellenberg et al. 455 KRAL, seca see Webb and Kral 447-351 494 LANDRY, GARRIE P, see Reese and Landry 99 — 102 Leavenworthia 189, 190; uniflora 190; eotlose 192; exigua var. laciniata Lectorypifcaton of oe subcarnosus a cE NSIS eas ae) by L. er n ia L cst 189, scuril 195; lyrata 34 Leucophylivm 107, 131; ambiguum 162; drews 255 — o globosa ee 4 ucoph llum 133; minus si pringlei 164; pruinosum 159; revolutum = 15 ee 160. zygophyllum 4 ns PATRICIA A., see Matthews & Levins 45-61 Lindernia grandiflora 381 Liparis loeselii 22 LIPSCOMB, BARNEY L., see Keeney & Lipscomb 102-103; see Thieret & ipscomb 103 Lobelia spicata 251 Ludwigia curtissii 381; erecta 381; lance- olata 381 Lupinus subcarnosus 255; texensis 255 MacGREGOR, JOHN, see Medley et al. Pea | McKENZIE, PAUL M. and LOWELL E. BATSCH, ecaseeles ron (Poaceae) new to Louisia Mammillaria dioica 200; delat ul ci Manisuris cylindrica 381 Marsilea uncinata 381 ear ae JAMES E and PATRICIA VINS., Portulaca As L., sates ILM. Johns. paren Gray in the southwest 45 MEDLEY, . et al., Ayan japonica’ (Miq.) Nakai (Am thacea in entucky and est Virginia: n to North America 656 MEDLEY, MAX E., see Cranfill et al. 189 — 1 Mentha piperita 381; rotundifolia 381 Millerocaulis amayjolensis 402; beardmorensis 402; dunlopii ean i) — 402; estipulans 402; gibbiana 402 guptai 402; hebeiensis 403; herbstii kidstonii 403; ; patagonica 403; rajma- halensis 403: sahnii 403; wadei 403 Millerocaulis, a new genus with species ormerly in Osmundacaulis Miller (Fossils: Osmundaceae) by William sikae and Mohlenbrock oe. — 346 ma cylindrica 352 Mie ‘ond (Poaceae: Monermeae) xas by Chuck is = offey and Jesus ie R. 352-— MORDEN, cao . 4 STEPHAN L. HATCH, Vegetative epOlntts in Muhlenbergia repens (Poaceae: Eragrostideae) 282 — 285 Muhlenbergia frondosa 486; repens 282 MARE ie Jrondosa ee new to Louisiana by Pau Mc oo and Lowell E. ae 486 — Nama 271; ee Zo. pas 27 2 , rothrockit 2 eee “GUY L. and SCOTT SUNDBERG, New combinations in Erigeron Cstercen 249 — 250 Nestronia umbellul Nestronia umbellula Raf (Santalaceae) new to Mississippi by Timothy E. Smith 46 — 247 € Soucuaur (Asteraceae) by Scott Sundberg ve — Psoralea sensu auct., Erigeron a a and Chamaecrista (Leguminosae) in the ree cae as States by Duane Isely 429 — Notes on two oie plants by Toney M. Keeney and Barney L. Lipscomb 102103 ie ee plants from north Florida. I] an C. Anderson 379 — 384 Ol discributi nd ecol- ogy of Sida hapa (L. ) Rusby (Malvacea €) by David M. Spooner et al. 215— Observations on the vegetation of the Mosquitia in ea by Andre E Clewell ae a Oenothera nuta aay Opuntia nee pbb: grandis 200, stenopetala 200 Orbexilum 431; pedunculatum 431; var. eglandulosum 432; lupinellum 432 Osmundacaulis 401, 403; skidegatensis 404 ax quinquefolius 486 Dae GUE L. (Araliaceae) redis- cc in faoeta by Lowel Urbatsch PARFITT, ie D. and WENDY maria VISCOSA (Caryophyllaceae): correct author cita- tion and range extension to the United ates 96 — PARFITT, BRUCE ee ee in North Sai Cacta a review Sone S. LANCE, see Orzell et al. 226-231 ae 385; glaucescens 386; linearifolia xX floridana 389; humifusa 393; ptoscrata 3892 Pedicularis lanceolata 227 Pediomelum 430; digitatum 430 ios aac ay, 12; Sa Ze pees 20; saxifraga 15; velut Paonia ee ince), o North by Richard K. Rabeler Physalis cordata 382; walter: 382 Pinus glabra 325 Plancago heterophylla 382; rugelit 38 Podophyllum peltatum 251 major 253; 5 Portulaca mundula 45; oleracea 318; subsp. granulatostellulata 319; subsp. nicaraguensis 321; subsp. nitida 321; subsp. Seem clare 321; subsp. stellata 321 Oe _ piles L., PB mundula 1.M. Johnst. and P parvula Gray in the Potamogeton 173; floridanus 340; subsect. Praelongi 174; praelongus 495 174; subsect. Perfoliati ia perfoli- atus 178; ee 18 PRINGLE, JAM ope Dasyactnim aegyptinm (eae ee new to Be PRINGLE, JAMES S., A new species and taxonomic notes on Gentianella (Gentianaceae) in South America 7 — 369 RABELER, RICHARD K., Petrorhagia a ae of Nort h America a marginatus 488; repens 253 Ratibida columnifera f. pulcherrima 382 Recent collections and status of Lesquerella des Rollins (Cruciferae) by David and Robert Kral 347 — 351 REESE, WILLIAM D., as Laisa valvis (L.) une (Poaceae) in Louisiana 99 — 102 Renovation of Dyssodia (Compositae: Tageteae) by John L. Strother 371—378 Rhynchospora capillacea 228 Sabal elonia (alinge): Systematics, dis- a and Walter S. Judd 417 — 427 Salix eriocephala 382 Scaevola 445, 450; plumieri oe sericea 451; var. taccada 103, 4 Scaevola (Goodeniaceae) in eee Un me eee by John W. Thieret and M. Brandenburg 445 — 453 ie 228 Scoparia montevidensis 382 Scutellaria minor 489 Scutellarta minor oe new to North merica by Larry E. Brown and John R. Ward 189 190 Selenicereus innesii 200 Sesbania 429, 435; macrocarpa 435; emerus 436; virgata 438 Sida’ 62; acuta 87; aggregata 76; sect. Ellipticifoliae 82; glabra 70: haenkeana 88; hermaphtodita 66, 215; sect. Hookerianae 66; hyalina 79; inflexa 84; imeri 84; lindhe- longipes 84; jussieana 7 1; linifolia 65; Pseudo — napaeae c Su 65; tragiifolia 80; troyana seas ae viarum 76; xanti 80; as SM ITH, “EDWIN B., Additional notes on Coreopsis 245 — 24 46 SMITH HY E., Nestronta ales lula Rap (Ganedacens) new to Mis sippt 246 — 247 SONGBED 334; altissima 252; bicolor 252; var. strictula 229; ptarmicoides 252; riddellii 229, 252; rugosa 252; speciosa var. pallida 338; spathulata var. neomexicana 246; squarrosa 253, uliginosa 2 Solidago (Asteraceae) of limited distribu- tion in the central United States by aera E.S. Taylor and R. John elo Seg Se) Solidago. spathulata DC. var. neomexicana (G ronq. (Asteraceae) new to aa xas by Richard D. Worthington 246 Some observations on bes glabra Walter (Pinaceae) by chard Stalter and Steve Dial 325 —- 328 Spigelia marilandica 248; marilandica f. rnea 249 SPOONER, DAVID M. et al., Observa- tions on the distribution and ecology of Sida hermaphrodita (L.) Rusby (Malvaceae) 215 — 225 Stachys tenuifolia var. perlonga 382 STALTER, HA and STEVE AL, Some observations on Pinus glabra (Pinaceae) 325 — 328 Stellaria prostrata 382 STROTHER, JOHN L., Renovation of Dyssodta (Compositae: Tageteae) 11378 Synopsis of the Florida species of Pectis (Asteraceae) by David J. Keil tsb Systematics of Leacophyllum and E remogeton (Scrophulariaceae) by Ja kson and L. David Flyr 107 — 172 Systematics of Nama (Hydrophyllaceae): with five species of Eriodictyon and Turricula parryi by John D. Bacon et al. 271-281 Tanacetum vulgare 253 Taxonomic and nomenclatural notes on Houstonia nigricans (Rubiaceae) by Edward E. Terrell 471 —481 Taxonomic ~ nomenclatural notes on Vaccinin L. section Cyanococcus ee by Leonard J. Uttal 7 — 395 Taxonomy, distribution and rarity status of Leavenworthia and — Lesquerella (Brassicaceae) in Kentucky by Ray Cranfill et al. 199 Taxonomy of — Flyriella Eupatorieae) ie . Turner 300 TAYLOR, CONSTANCE E.S. TAYLOR, oltdago (Asteraceae) of limited distribution in United States 334 — 339 OHN, see Taylor and 9 (Asteraceae- Baker and B. and R. E., Taxonomic nclatural notes on ilies (Apiaceae) and marginatus (Ranuncu- laceae) new to Texas by Larry E. Brown 488 The rediscovery of Sila pce Small (Potamogetonaceae) by Gerould S. Wilhelm and Robert H. Mohlen- brock 340 — ° The vascular flora of Central Florida: taxonomic nomenclatural an changes, sauces taxa by R Wunderlin et al. =A Thelypteris eae ve Themeda quadrivalvis 99 Themeda quadrivalvis L.) Kuntze (Poaceae) in Louisiana by William D. Reese and Garrie P- Landry 99 — 102 THIERET, JOHN W., see Medley et al. 92-95; see Calaway and Thieret 207 — 214 THIERET, John W. and BARNEY L. LIPSCOMB, Scaevola sericea Vahl var. taccada (Gaertn.) Thieret & sa comb. nov. (Goodeniaceae) | THIERET, JOHN W. and D ane M. BRANDENBURG, Scaevola (Goo- den ae in southeastern United — 453 Thymophlla os acersoa 375; 376; : rea smears 376; mutica ogee Waa ee oa var. belenidium 377; var. hartwegii 377; var. puberula 377; setifolia 37), 378: var. wrightii 378. ee WILLIAM D., Millerocaulis, us with species formerly in Osmundacaais Miller (Fossils: Osmun- daceae) 401 5 TURNER, B.L. cone JEAN ANDREWS, Lectotypification of mee abana and L, texensis (Fabaceae) 255 — 257 TURNER, eae see Baker Turner 0-317 URBATSCH, LOWELL E., Panax quinquefolius L. Gualiaceas) redis- covered in Louisiana 486 URBATSCH, LOWELL see McKenzie and ee ee ~ 488 UTTAL, LEONAR Cyanococcus ( Brieacese) a2 UTTAL, LEONARD J., A recombin tion in Hydrangea L. eee 352 Vaccinium 397 atrococcum 398; bocatorensis 441; corymbosum 398; simulatum VALDES R aos see Coffey and Valdes VAN ORR. ean S. and JOHN R. FREEMAN, A white — flowered form - Spigelia marilandica L. eal ew to Tennessee 248 — 249 497 Vegetative apomixis in Muhlenbergia repens (Poaceae: Eragrostideae) by Clifford . Morden and Stephan L. Hatch 282 — 285 Veronica agrestis 383; americana 103; polita 38° Vicia floridana 383; hugert 383 Viola sororia 383 Wallenia 329; formonensis 329 ns JOHN R., see Brown and Ward 489 WEBB, DAV H. and ROBERT K ae Aas collections and status ta Rollins n Bumelia dominicana (Sapotaceae), a new name for an old Sapote 396 WILBUR, ROBERT L., An undescribed Panamanian Vaccinium: Vaccinium bocatorensis (Ericaceae) 441 — 443 WILHELM, GEROULD SS. nd ROBERT H. MOHLENBROCK, The rediscovery of Potamogeton flori- dans Small (Potamogetonaceae) we 346 a papulifera 407 Wali hase ees pas. n by William J. Hes Zon WORTHINGTON, RICHARD D., Solidago spathulata DC. var. neomexi- cana (Gray) Cronq. (Asteraceae) new to exas 246 WORTHINGTON, RICHARD D., s RLIN, R.P et al., The vascular flora of Central Florida: taxonomic and nomenclatural changes, additional taxa 232 — 244 WUNDERLIN, R.P, A new bina- tion in Hedyotis i bas "400 Yucca gloriosa 383 Zephyranthes candida 383; grandiflora ZONA , SCOTT and WALTER S. JUDD, Sabal etonia (Palmae): ese oe ecolo and compar er Florida "scrub aes (Contents continued from front cover) Scaevola (Goodeniaceae) in southeastern United States. John W. Thieret and David M. Brandenburg. 445 Additions to the flora of New Mexico. Richard Spellenberg, Richard Worthington, Paul Knight and Reggie Fletcher. 455 Taxonomic and nomenclatural notes on Houstonia nigricans (Rubiaceae). Edward E. Terrell. 471 Documented plant chromosome numbers 1986: 1. Miscel- laneous counts in Aster (Compositae). Jerry G. Chmielewski. 483 OTES. Panax eee L. (Araliaceae) rediscovered in Louisiana. Pet nc ti frondosa (Poaceae) new to Louisiana. 486—Thaspium trifoliatum (Apiaceae) and R culus mergnat (Ranunculaceae) new to Texas. 488—Scutellaria minor je new rth America. 489. REVIEWS 406, 411 REVIEWERS 454 INDEX TO VOLUME II 491