Selbyana 23(2): 137-180. 2002.

NEW SPECIES AND NOTES ON MYRTACEAE FROM NORTHERN SOUTH AMERICA

BRUCE K. HOLST Marie Selby Botanical Gardens, 811 South Palm Ave., Sarasota, FL 34236-7726 USA. E-mail: [email protected]

ABSTRACT. In preparation for the treatment of Myrtaceae for the Flora of the Venezuelan Guayana, two new combinations (Eugenia yatuae, Myrcia intonsa) and 12 new species are presented (Calyptranthes conduplicata, Calyptranthes nigrescens, Eugenia conduplicata, Eugenia magna, Marlierea convexivenia, Marlierea foveolata, Marlierea mcvaughii, Marlierea subcordata, Marlierea ventuarensis, Myrcia kylisto­ phylla, Myrcia liesneri, Myrciaria puberulenta), as well as notes on species distribution and circumscription.

Key words: MYliaceae, Guayana Shield, Venezuela, Guianas, Brazil, Colombia, Peru, Bolivia, Ecuador

INTRODUCTION The neotropical genera of M yrtaceae most in need of study are Myrcia and Eugenia. Because In preparation for the treatment of the Myr­ of their large sizes, no one has attempted to treat taceae for the Flora of the Venezuelan Guayana, them as a whole since the major publications of 12 new species were found as well as numerous Berg (1855-1856, 1857-1859, 1861). A serious new records for the Venezuelan Guayana and for problem is the lack of naturally delimited sub­ the Amazonian regions of Colombia, Peru, Bra­ groups within the genera that would provide zil, and Bolivia. Additionally a number of col­ workable units. The previous attempts have lections in the past two decades have provided proven either artificial, i.e., Berg's division of critical material needed to augment previously Eugenia based on shape of the inflorescence, or incomplete descriptions. unworkable on a large scale, i.e., the sections by Notes presented here are not limited to spe­ McVaugh (1969) in Myrcia, which even he not­ cies in the Venezuelan Guayana since, it was ed were somewhat artificial. Perhaps the greatest necessary to study Myrtaceae from the Guianas, contribution then that anyone could make to­ eastern Colombia, northern Brazil, northern Ve­ ward increasing the attractiveness of working on nezuela, and the Antilles. Sixteen genera are such huge genera is laying out the groundwork known to be native in the Venezuelan Guayana. for an effective subgeneric classification. That is Pimenta, Eucalyptus, and Syzygium have been beyond the scope of this work, but it should be found cultivated in the area being treated in the one of the principal goals for future research. Flora of the Venezuelan Guayana (the states of Amazonas, Bolivar, and Delta Amacuro), with Additionally, several morphological characters only Syzygium becoming naturalized. Certain in need of more study in the fleshy-fruited, species groups will be treated in a broad sense mainly neotropical Myrtoideae are anthers (mor­ in the Flora, especially in Eugenia and the sec­ phology, dehiscence), leaf venation, trichomes tion Myrcia of Myrcia. Select citations of type (types, branching), acicular appendages, and specimens and additional material examined are bark characters. presented here, as the format of the Flora of the Venezuelan Guayana and other checklists do not allow for extensive specimen citations. Any attempt to treat the species in any part of NEW SPECIES AND COMBINATIONS northern South America would be very difficult jf not for the previous works of McVaugh (l958b, 1969) in Bassett Maguire's Botany of Calyptranthes conduplicata B. Holst, sp. nov. the Guayana Highlands, that of Amshoff (1942, TYPE. Venezuela. Amazonas: Carretera San 1948, 1950) which culminated in her Myrtaceae Carlos-Solano, 2-6 km SO de Solano, 100 treatment in the Flora of Suriname (1951), and m, 25 Apr. 1974 (ft), G. Morillo, B. de Mor­ the Flora Neotropica treatments of Landrum illo & C. Wood 3940 (HoJotype: VEN; Iso­ (1981, 1986). types: MICH, MY). FIGURE 1.

137 138 SELBYANA Volume 23(2) 2002

Calyptranthes nigrescens B. Holst affinis, sed dis­ two collections from evergreen lowland forests tincta: petiolo crasso, 1.5-2 cm Ion go (nec 1-2 mm at 100-160 m elevation in southwestern Ama­ longo); foliis basi obtusis ad rotundatis (nec subcor­ zonas state, Venezuela. datis). Calyptranthes nigrescens B. Holst, sp. nov. Tree 6-10 m tall; inflorescence and vegetative TYPE: Rio Pasimoni, entre la boca y Piedra buds lightly pubescent with appressed whitish Arapacoa, 175 m, 01050'N, 66°45'W, 20-30 trichomes; branchlets laterally compressed in Oct 1986 (ft); B. Stergios & O. Martinez cross section. Leaves stoutly petiolate, the peti­ 9529 (Holotype: PORT; Isotypes: MO, NY, oles 1.5-2 cm long, 2 mm diameter, drying dark, shallowly channeled and mostly terete in cross SEL, US). FIGURE 2. section; blades drying dark reddish brown, ellip­ Calyptranthes multiflora O. Berg affinis, sed dis­ tic to oblong, 11-24 X 5-8.5 cm, coriaceous tineta: foliis 9-15 cm longis, subeordatis, eoriaceis when dry ("quebradiza" when fresh, fide Mor­ (nec 5-10 cm longis, aClltis ad cuneatis, membrana­ illo et al. 3940), glabrous, glands numerous, im­ ceis); baga 1.5-2.5 cm diam. (nec 0.5-0.8 cm diam.), pressed-punctate to plane on upper surface, parieti crassa (nee tenui); foliis, infiorescentiis et fruc­ slightly elevated on lower surface; apex abruptly tibus sieeatis uigris (nee atroviridibus nee olivaceis). acuminate-caudate, conduplicate-recurved in apical lA, base obtuse to rounded, slightly de­ Shrub or tree 4-12 m tall; glabrous with the exception of vegetative buds, which are ap­ current on petiole; midvein broad, plane or with two convex ridges on upper surface, secondary pressed-pubescent; the whole plant typically veins inconspicuous, ca. 16-20 per side, margin­ dries blackish; branchlets terete in cross section, al vein 2-3 mm from the margin and parallel to the nodes slightly thickened. Leaves subsessile, it. Inflorescences paniculate, 7-13 cm long in­ the petioles 1-2 mm long, stout; blades drying cluding peduncle, to 4.5 cm broad, axillary or blackish, particularly on upper surface, ovate to subterminal, 25-35-flowered, main axis laterally elliptic, 9-15 X 2-6 cm, coriaceous, glands in­ compressed in cross section; peduncles 1-2.4 conspicuous; apex acuminate, usually abruptly cm long; bracts not known, deciduous before an­ so and conduplicate, sometimes rounded and thesis; pedicels 0-2 mm long; bracteoles not with a short (ca. 1 cm long) acumen, base sub­ known, deciduous before anthesis. Flower buds cordate, appearing at times nearly peltate due to obovoid, ca. 3 mm long, glabrous without, apic­ the reflexed angle of the blade and overlapping ulate; hypanthium prolonged ca. 2 mm above blade tissue at the base; midvein convex to bi­ summit of ovary, strongly reflexed after anthe­ convex, broad, secondary veins numerous, in­ sis; calyx closed in bud, circumscissile, very thin conspicuous, marginal vein 0.5-3 mm from the and mostly deciduous after anthesis; petals 2 and margin and mostly parallel to it. Inflorescence perhaps a few vestigial ones as well, broadly of paired panicles in the leafaxils, 8-11 cm ovate to suborbicular, ca. 1.5 mm long. Stamens long, with relatively slender branches, axillary ca. 100; anthers ca. 0.3 mm long; filaments ca. to subterminal, 9-ca. 30-flowered; peduncles 1- 2.5 mm long. Ovary bilocular, ovules 2 per Joc­ 3(-4) cm long; bracts deciduous well before an­ ule; style stout, 4-5 mm long; stigma punctate. thesis; pedicels lacking in terminal flowers of 3- Fruits not known. Seeds not known. flowered branches or 4-7 mm long in lateral ones; bracteoles deciduous well before anthesis, Paratype: VENEZUELA. Amazonas: Dept. Ca­ acicular glands often present. Flower buds ob­ siquiare, Calio Miguel, 160 m, 02°50'N, ovoid, ca. 3 mm long, apiculate, calyptrate, oc­ 66°50'W, 20 Apr. 1991 (st), Aymard 9037 casionally tearing irregularly, glabrous; hypan­ (PORT, SEL). thium prolonged ca. 1 mm beyond the summit of the ovary, explanate after anthesis; calyptra Calyptranthes conduplicata is related to C. ni­ ca. 3 mm wide, glabrous within; petals 5, grescens B. Holst, C. multiflora O~ Berg, and C. clawed, 2~3 mm long. Stamens numerous; an­ forsteri o. Berg. It differs from C. nigrescens by thers oblong to nearly globose, ca. 0.3 mm long; having stoutly petiolate leaves (petioles 1.5-2 filaments ca. 5 mm long. Ovary 2-locular, the cm long) that are obtuse to rounded at base (vs. walls thick, ovules 2 per locule; style 4-5 mm 1-2 mm long and subcordate at base). From C. long; stigma punctate. Fruits purple-black to multiflora and C. forsteri, it is distinguished by black, 1.5-2.5 cm diameter, wall 2-3 mm thick. having the leaves abruptly acuminate-caudate Seeds 1, kidney-shaped, ca. 1.5 cm long, hypo­ and conduplicate (vs. short-acuminate and acute cotyl elongated, cotyledons contortuplicate. to obtuse at apex and plane) as well as by the stout and long pediceJs (vs. pedicels < 1 cm Paratypes: BRAZIL. Amazonas: Munic. S. Ga­ long). briel da Cachoeira, CUCU!, Igarape Bonte, Mata Calyptranthes conduplicata is known from de Igapo, OJ °21 'N, 66°50'W, 27 Oct. 1987 (past HOLST: MYRTACEAE 139

FIGURE 1. Calyptranthes conduplicata (Morillo et al. 3940, Isotype, MICH). 140 SELBYANA Volume 23(2) 2002 ft); Farney et al. 1807 (MO); Foz do rio Marie, 7514 (PORT, SEL); Dept. Rio Negro, Rio Yatua, aftuente do rio Negro, 00025'N, 66°24'W (fr), 4 entre la piedra Catipan y Ipaca, 2 Dec. 1984 (fr), Dec. 1987, Kawasaki 356 (MO, NY). COLOMBIA. Stergios & Aymard 7545 (MO, PORT, SEL); Guainfa: just south of jct. of Rio Negro, Rio Dept. Rio Negro, Rio Emoni, hasta aprox 8 km Guainia and Casiquiare, 65 m, 01°59'N, arriba de la boca, 17-18 Apr. 1985 (fr), Stergios 67°06'W, 25 Jun. 1984 (fr), Davidse & Miller & Aymard 8239 (PORT, SEL); Dept. Rio Negro, 26576 (MO, SEL, VEN); Rio Guainia, near Pta. Rio Pamoni, 02°48'N, 65°57'W, 24 Sep.-5 Oct. Colombia on Colombia side, opposite Maroa, 1987, Stergios & Stergios 11404 (NY); Dept. 130 m, 12 Oct. 1957 (bud, ft), Maguire, Wur­ Rio Negro, Rio Varia, 90 m, 01°25'N, 66°24'W, dack & Keith 41835 (F, MO, NY, US, VEN); Apr. 1991 (fl), Velazco 1919 (MO, SEL); Rio Vaupes, Rio Guainfa basin, Rio Naquieni, vicin­ Guainfa, Maroa, 11 Feb. 1942, Williams 14290 ity of Cerro Monachi, June 1948 (fr), Shultes & (F, VEN, US); Cafio San Miguel, Rio Guainia, Lopez 10049 (US). VENEZUELA. Amazonas: Rio 127 m, 25 Mar. 1942 (fr), Williams 14877 (F, Cuao, desde raudel de Danta, entre Calla Raya NY, VEN, US). y Picure, 04°54'-05°03'N, 6r34-46'W, 23 Sep. Calyptranthes nigrescens is apparently com­ 1996 (fr), Castillo 4188 (SEL); Dept. Rio Negro, Cafio Cuweje, 30 km S of San Carlos, 2 May mon along periodically flooded margins of 1980 (fr), Clark 7548 (MO, NY, VEN); Cafio blackwater rivers, at 80-120 m elevation in cen­ tral to southern Amazonas State, Venezuela, as Cuweje, 32 km S of San Carlos de Rio Negro, well as adjacent Colombia (Guainia) and Brazil 20 Mar. 1980 (fr), Clark 7459 (MO, NY, VEN); (Amazonas: upper Rio Negro). Cafio Pasi, 19.2 km S of San Carlos de Rio Ne­ Calyptranthes nigrescens is a species that is gro, 8 May 1981, Clark & Maquirino 8017 (NY); Lower part of Rio Baria, 80 m, 01°10- somewhat intermediate with Marlierea because of the calyptra that occasionally irregularly tears, 27'N, 66°25-32'W, 22-23 Jul. 1984 (ft, fr), Dav­ instead of the typical cleanly circumscissile ca­ idse 27636 (MO, VEN); Lower Rio Baria, 80 m, 01°1O-27'N, 66°25-32'W, 22-23 Jul. 1984 (fr), lyx in Calyptranthes. It is distinguished from all other Calyptranthes species in the Guayana Davidse 27663 (MO, SEL, VEN); Rio Pasimoni, Shield by its feature of the inflorescence, leaves, between its mouth and its jct. with Rio Baria and and fruits drying black in combination with the Rio Yatua, 80 m, 01°27-53'N, 66°32-35'W, 23- 25 Jul. 1984 (fr), Davidse 27722 (MO, SEL, subcordate leaf blade bases. Close relatives ap­ VEN); Dept. Atabapo, SE bank of middle Cafio pear to be Calyptranthes mUltiflora O. Berg, C. conduplicata B. Holst, and Calyptranthes for­ Yagua at Cucurital de Yagua, 120 m, 03°36'N, steri O. Berg. The fruits of C. nigrescens are 66°34'W, 8 May 1979 (fr), Davidse, Huber & Tillett 17336 (MO, NY, US, VEN); Laguna Car­ described as edible but astringent. ipo, Rio Orinoco, J an-Feb. 1969 (fr), F arifias, Eugenia condupiicata B. Holst, sp. nov. TYPE: Velasquez & Medina 511 (NY, VEN); Dept. Ca­ Brazil. Amazonas: Rio Urubu, between siquiare, Maroa, 17 Nov. 1978 (bud, ft), A. Fer­ Cachoeiras Lindoia and Iracema, nearer to nandez 3422 (MY, NY); Alto Rio Atacavi, 7 Iracema; flooded river bank; 3 June 1968 Sep. 1960 (ft), Foldats 3757 (MICH, VEN); (ft), G.T. Prance, D. Philcox, W.A. Ro­ Dept. Atures, Rio Autana, Raudal Ceguera, 100 drigues, J.F. Ramos & L.G. Farias 4958 m, 04°48'N, 67°28'W, 11 Nov 1984 (f1, fr), (Holotype: US; Isotypes: F, JNPA, MICH, Guanchez & Melgueiro 3486 (SEL, TFAV); NY). FIGURE 3. Dept. Atabapo, Calla Negro, en las cabeceras del Cafio Yagua, 130 m, 03°49'N, 66°33'W, 25 May Species propria foliis conduplicatis, coriaceis, nitidis 1981 (fr), Guanchez 1246 (MO, TFAV); Dept. et racemis brevis, fioribus numerosis, breviter pedice­ latis. Casiquiare, Rio Atacavi, 03°06'N, 67°11 'W, 16 Jul. 1981 (fr), Guanchez 1385 (MO, TFAV); Shrub or treelet to 4 m tall; glabrous with the bajo Rio Yatua, 25 Jul. 1982 (fr), Stergios & exception of vegetative buds and inflorescence Aymard 4131 (US); Alto Rio Pasimoni, 25 Jul. axes, which are appressed-pubescent to puberu­ 1982 (fr), Stergios & Aymard 4123 (PORT); lous with reddish trichomes; branchlets slightly Alto Rio Pasimoni, 25 Jul. 1982 (fr), Stergios & compressed in cross section. Leaves petiolate, Aymard 4127 (PORT); Dept. Rio Negro, Rio the petioles 9-11 mm long, channeled; blades Pasimoni, 01 °40'N, 66°35'W, 30 Nov. 1984 (fl, drying yellowish brown to olive-brown, lustrous fr), Stergios & Aymard 7417 (MO, PORT, SEL); on upper surface, ovate to broadly ovate, 8.5- Dept. Rio Negro, Laguna Yui, boca del Rio Ya­ 12.5 x 5-8 cm, thick-coriaceous, glands incon­ tua, 01°25'N, 66°30'W, 30 Nov. 1984 (past fl), spicuous; apex abruptly acuminate, recurved­ Stergios & Aymard 7450 (MO, PORT); Dept. conduplicate, base obtuse to broadly rounded or Rio Negro, Rio Yatua, entre la piedra Catipan y nearly truncate; midvein broad, convex to prom­ Ipaca, 2 Dec. 1984 (bud, fl), Stergios & Aymard inently convex, secondary veins 10-15 per side, HOLST: MYRTACEAE 141

FIGURE 2. Calyptranthes nigrescens (Stergios & Martinez 9529, Isotype, NY). 142 SELBYANA Volume 23(2) 2002 inconspicuous and scarcely elevated on upper cence velutinous or villous; trichomes simple, and lower surfaces, marginal vein equal in white or grayish with reddish orange bases; prominence to the secondaries and slightly arch­ branchlets com,pressed in cross section near the ing between them at 1.5-2.5 mID from the mar­ nodes. Leaves petiolate, the petioles stout, terete, gin. Inflorescences much-abbreviated racemes, 10-20 mID long, 3-5 mm wide; blades drying axes 1-4 mID long, solitary or clustered in the olive-green to dull reddish-brown, coriaceous, leafaxils or at leafless nodes, 4-8 flowered; elliptic to narrowly elliptic or oblong, 20-32 X bracts persistent, drying blackish, broadly ovate, 7-10 cm, 2.6-3.4 times longer than broad, ca. 0.8 mID long, apex acute to obtuse;pedicels glands scarcely visible above or not, evident be­ slender, laterally compressed, 1-3 cm long; low as small numerous convex black dots; apex bracteoles ca. 0.8 mID long, ovate, ciliate, per­ acuminate, base cuneate to broadly obtuse, mar­ sistent, connate basally. Flowers 4-merous, gla­ gin subrevolute to nearly plane, glabrescent brous; hypanthium flaring above ovary due to above except along the puberulent midvein, ve­ broad staminal ring; calyx lobes broadly ovate, lutinous below, densely so along the secondary 1-2 mm long, rounded at apex; petals broadly veins; midvein plane or concave, convex at ex­ elliptic to orbicular, 3-6 mID long. Stamens nu­ treme base, prominent below, secondary veins merous; anthers oblong, ca. 0.8 mm long; fila­ 11-14 per side, ascending, convex above, more ments ca. 2 mm long. Ovary 2-locular, ca. 10 prominent below, marginal vein 5-12(-20) mID ovules per locule; style ca. 4 mID .long; stigma from the margin and arching slightly between punctate. Fruits globose to broadly ellipsoid, 8- the secondaries, a second less prominent mar­ 10 mID diameter, reddish, crowned by the per­ ginal vein 1-4 mID from the margin and parallel sistent calyx lobes. Seeds 1, cotyledons fused. to it. .Inflorescence short-racemose, appearing umbellate, axillary at leafy or leafless nodes, Paratypes: BRAZIL. Amazonas: Rio Cuieras, densely grayish velutinous, 4-8 flowers per axis; Igarape Lobisomen, 02°31'S, 60015'W, 22 Jul. bracts ovate, 2-3 mID long, glabrous except for 1991 (fl, fr), Mori & Gracie 21814 (NY). VE­ ciliate margins; pedicels 4-6 mID long, velutin­ NEZUELA. Amazonas: Dept. Atabapo, southeast­ ous, fruiting pedicels stout, 1-2.2 cm long; brac­ ern bank of the middle part of Cafio Yagua, 120 teoles linear to narrowly clavate, 5-8 mID long, m, 03°36'N, 66°34'W, 8 May 1979 (fr), Davidse, usually deciduous before anthesis. Flower buds Huber & Tillett 17327 (MICH, MO); Planicie obpyriform,8-13 mID long, open flowers large, inundable riverifia del Cafio Yagua, 180 m, to 3 cm across, densely tomentose without; disk 03°37'N, 66°35'W, Nov. 1989 (fr), Marin 461 ca. 4 mID across, villous; calyx lobes spreading (NY, PORT, SEL); en la ribera derecha de del at anthesis, oblong to ovate, 4.5-8 mID long, 4- alto Atacavi, aprox. 100 m, entre 03°05' y 6 mID wide, cupulate, villous on both surfaces, 03°50' Lat. N y entre los 66°50' y 67°30' Long. broadly rounded at the apex, the interior pair W, 10-24 Nov. 1980 (past fl), Pifiate & Mon­ slightly longer; petals ligulate to elliptic, 1.2-1.4 dolfi 1051 (MICH). cm long, 6-10 mm wide, glabrous. Stamens aprox. 180; anthers oblong, 1-1.2 mm long; fil­ Eugenia conduplicata is known from black­ aments 10-12 mID long. Ovary 2-locular, aprox. water riverine and floodplain forests, ca. 100- 16 ovules per locule; style 14-15 mID long, vil­ 120 m elevation in Amazonas state, Venezuela, lous in the lower half or not. Immature fruits and the Urubu and Cuieras rivers in Amazonas, obpyriform, 1.5-2 X 1-1.5 cm, disk 6-7 mm Brazil. It is quite distinctive with broad, thick­ wide, .sepals persistent and reflexed, persistently coriaceous, lustrous, leaves that are recurved­ velutinous. Seeds unknown. conduplicate in the apical 1,4, and slender-pedi­ cellate flowers on a short-racemose .inflores­ Paratypes: BRAZIL. Amazonas: Base of Serra cence. Araca, 0-3 krn South of Central Massif, 3 krn Eugenia magna B. Holst, sp. nov. TYPE: Ve­ E of Rio Jauari, 00049'N, 63°19'W, 8 Feb. 1984 nezuela. Bolivar: 4 krn west of EI Pauji, 2 (fr), Prance et at. 28897 (MO); arredores de ser­ to 5 krn north of road, Rio ChaberIi, 750- ra Araca, 26 Feb. 1977 (fr), Rosa & Cordeiro 900 m, 04°30'N, 6l036'W, 12 Nov. 1985 1676 (SEL).GUYANA. U. Takutu-U. Essequibo, (fl), R. Liesner 19973 (Holotype: MO; Iso­ Sipu River, 0-5krn E of camp, 240 m, 01"25'N, types: SEL, VEN). FIGURE 4. 58°57'W, 11 Sep. 1998 (past fl), Clarke et at. 7757 (SEL); Gunn's, Essequibo River, 249-260 Ab Eugenia gomesiana O. Berg differt: ealieibus m, 01°39'N, 58°38'W, 2 Sep. 1989 (past fl), Jan­ lobis longioribus, villosis (nee sparsim tomentulosis, sen-Jacobs et al. 1411 (SEL); Kamoa River, glabratis); fruetibus tomentosis (nee glabratis) et foliis longioribus, ultra 20 em (nee < 20 em longis). Toucan Mountain, 260-360 m, 01°33'N, 58°50'W, 22 Sep. 1989 (fl), Jansen-Jacobs et al. Tree 3-15 m tall; twigs, leaves, and inflores- 1750 (SEL). VENEZUELA. Bolivar: Orillasde Rio HOLST: MYRTACEAE 143

FIGURE 3. Eugenia conduplicata (Prance et al. 4958, Ho!otype, US; detail of Mori & Gracie 21814, NY). 144 SELBYANA Volume 23(2) 2002

Caroni, cerca de los raudales de Mureyema, 740 ers per raceme; calyx splitting irregularly to m, 9 Oct. 1946 (fl), Cardona 1805 (NY, US, form 2-3 lobes, persistent into fruit; anthers ob­ VEN); same locality, 800 m, 10 Oct. 1946 (fl), long, gray; style ca. 1 cm long at anthesis; fruits Cardona 1814 (US, VEN); 1-2 kms NE of E1 oblong, 3-3.5 X 1-1.5 cm, apparently with Pauji, Rio Cabass below SaIto and adjacent for­ much flesh, maturing orange-red, spicy, and ed­ ested slope, 800-900 m, 04°30'N, 61°35'W, 7 ible (fide Maguire); embryo massive, cotyledons Nov. 1985, Liesner 19639 (MO, SEL, U, VEN); homogeneous, hypocotyl evident. Distr. Sifontes, Munic. Aut6nomo Gran Sabana, Specimens examined: BRAZIL. Amazonas: Cuenca alta del Caron!, a 1 km al Oeste de Carro Tucano, Rio Cauaburi, 125 m, 17 Nov. Apoip6 en el bajo Kukenan, riberas del Suruktin, 1965 (ft) , Maguire, Steyermark & Maguire 830 m, 04°45'N, 61°30'W, 30 Apr. 1991, Rosales 60188 (MICH, MO, NY). VENEZUELA. Amazon­ & Dezzeo 815 (MO); 10-40 kms al ENE de Ica­ as: Dept. Rio Negro, Carro Baria, between head­ barti, 700-800 m, 04°25'N, 61°32-40'W, 15 waters of Rio Mawarinuma and Rio Baria, 130 Dec. 1978 (imm. fr.), Steyermark et al. 117657 m, 00053'N, 66°15'W, 29 Mar. 1984 (fl), Liesner (MICH, MO, VEN). 16997 (MO, NY, VEN). Marlierea convexivenia B. Holst, sp. nov. Eugenia magna is related to E. gomesiana O. TYPE: Venezuela. Amazonas: Dept. Cas i­ Berg, a lowland riverine species widespread in quiare, Carro San Miguel, sector Las Tinajas Amazonia. Eugenia magna tends to occur at y el Carro Iqueven, 160 m, 02°39'N, higher elevations (250-900 m elevation) in terra 66°45'W, 24 Apr. 1991 (fr) , G. Aymard firme forest. It is distinguished from E. gome­ 9151 (Holotype: PORT; Isotypes: NY, siana by having longer calyx lobes (4.5-8 mm SEL). FIGURE 5. long vs. 1.5-3 mm long) that are permanently villous (vs. sparsely tomentulous and glabres­ Marlierea caesariata MeVaugh affiuis, sed distinc­ cent), the fruits permanently tomentose (vs. ta: foliis angustatis, 3.5-4.5 em latis (nee 5-10 em fruits glabrescent), and the leaf blades being rel­ latis), oblongis ad ellipticis, apieibus nee reftexis (nee atively long (> 20 cm long vs. < 20 cm long). elliptieis ad obovatis, apicibus reftexis); petiolis 7-9 mm longis (nee 10-20 mm longis), et inftoreseentiis Eugenia yatuae (McVaugh) B. Holst, comb. plerunque raeemosis (nee paniculatis). nov.-Calycorectes yatuae McVaugh, Tree to 5 m tall; young stems and leaves, and Mem. New York Bot. Gard. 18(2): 226. all parts of inflorescence coppery sericeous; 1969. TYPE: Venezuela. Amazonas: Rio Ya­ branchlets laterally compressed in cross section. tua, 100-140 m, 22-26 Oct. 1957(fl), Ma­ Leaves petiolate, the petioles 7-9 mm long, guire et al. 41947 (Holotype: MICHl; 1so­ shallowly channeled, older ones becoming types Fl, NY!). corky-rimose; leaf blades mostly oblong to el­ Calycorectes differs from Eugenia by the pro­ liptic, rarely obovate, 7.5-12.5 X 3.5-4.5 cm, longation of the hypanthium above the summit thinly coriaceous, older leaves sparsely coppery of the ovary and the irregular tearing of the ca­ sericeous to glabrescent, upper surface glabrous, lyx lobes, scarcely a valid method of delimiting impressed-punctate, glands drying dark on lower two genera given the demonstrated variability in surface, slightly raised; apex abruptly and sharp­ fusion of the calyx lobes in several other genera: ly acuminate, base acute; midvein prominently Psidium, Myrcia, Campomanesia. Without tak­ convex on upper surface, secondary veins ing the major step of reducing Calycorectes to scarcely visible, 12-14 per side, marginal veins the synonomy of Eugenia before a revision of ca. 2-3 mm from the margin and shallowly arch­ the eugenioid genera can be done, it is obvious ing between the secondaries. Inflorescence ra­ that E. yatuae is very closely related to other cemose, but apparently so by reduction of the species in the "feijoi" complex of Eugenia, no­ lateral branches, 2.5-8 cm long including pe­ tably E. dipiocampta Diels. The midvein is el­ duncle, solitary or occasionally paired in axils evated into a narrow line, the calyx lobes are of leafy or leafless nodes; peduncles 1.5-3.5 cm united into a developed hypanthium, fruits ellip­ long; bracts triangular, mostly deciduous by post soid-oblong, trichomes appressed, dibranchiate anthesis, 3-4 mm long, acute at apex; pedicels or laterally attached near the base, coppery, an­ lacking; bracteoles subulate, curved, ca. 3 mm thers narrow, gray. long, mostly deciduous at anthesis. Flowers ses­ Previously only known from the type collec­ sile, 4-merous, densely coppery sericeous with­ tion in flower, two additional collections have out; hypanthium prolonged ca. 1.5 mm above been made in fruit. The following notes are in­ summit of ovary, reflexed after anthesis, irreg­ tended to supplement the original description. ularly tearing through the staminal ring into Tree 3-15 m high; leaves elliptic, lateral veins mostly 4 lobes, but with some irregular pieces to 25 per side; inflorescences with up to 6 flow- visible; calyx probably closed in bud or lobes HOLST: MYRTACEAE 145 partially connate, reflexed after anthesis with turning gray with age; branchlets terete or slight­ some persistent; petals incompletely known, ly compressed in cross section, current year's based on shape and position of 1 petal seen, like­ growth tomentose. Leaves nearly sessile, the ly 4 present, obovate, ca. 2.5 mm long, ciliate. stout petioles ca. 1 mm long; blades broadly Stamens numerous; anthers ca. 0.3 mm long; fil­ ovate to broadly elliptic or nearly orbicular, 2.0- aments ca. 4 mm long. Ovary bilocular, ovules 3.3 X 1.6-2.6 cm, stiff-coriaceous, young blades unknown; style 5 mm long; stigma punctate. Im­ densely ferruginous-tomentose, glabrescent on mature fruits globose, ca. 8 mm diameter, upper surface and patchily so on lower surface, sparsely sericeous, crowned by remnants of hy­ impressed-punctate on upper surface, slightly panthium and calyx lobes. Seeds not known. raised on lower surface; apex bluntly obtuse to rounded, base subcordate, margins slightly rev­ Paratypes: VENEZUELA. Amazonas: Cafio olute; midvein impressed on upper surface, Duapo, que desemboca en el Rio Siapa, abajo prominent below, secondary veins 12-14 per del campamento de los Yanomami-Siapateri, 18 side, slightly elevated on lower surface, plane or Feb.-4 Mar. 1986 (fr), Stergios & Aymard 9221 delicately engraved ou upper surface, tertiary (PORT, SEL); Lugar "Sabana Grande," del Rio veins not visible, marginal vein ca. 1 mm from Pasimoni, a altura de Pueblo Viejo, Rio Casi­ and parallel to the margin. Inflorescences 3- quiare basin, 10-22 Feb. 1989 (fr), Stergios et flowered, 9-10 mm long, paired in the leafaxils; al. 13341 (NY, PORT); Dept. Atabapo, Rio Ata­ peduncles 6-7 mm long; bracts deciduous at or bapo, a 30 kms de la confluencia con el Rio before anthesis, oblong, canaliculate, 2-5 mm Caname, 110 m, 03°48'N, 67°38'W, Nov. 1989 long, apex obtuse; pedicels lacking; bracteoles (fr), Yanez 223 (MO, PORT). 2-3 mm long, similar to bracts. Flower buds ca. 3 mm long, 5-merous, tomentose without; hy­ Marlierea convexivenia is related to Marli­ panthium prolonged above summit of ovary ca. erea caesariata McVaugh, but differs by having 1 mm, tearing longitidinally between the calyx narrower leaf blades (3.5-4.5 cm wide vs. 5-10 lobes at anthesis; calyx lobes irregular but most­ cm), oblong to elliptic leaf blades that are not ly ovate, ca. 2 mm long, obtuse at apex, decid­ reflexed apically (vs. elliptic to obovate leaf uous after anthesis; petals (4?)5, orbicular to ob­ blades that are apically reflexed), petioles 7-9 long, ca. 3 mm long, villous without. Stamens mm long (vs. 10-20 mm), and mostly racemose numerous; anthers oblong, ca. 0.4 mm long; fil­ (vs. paniculate) inflorescences. Racemose inflo­ aments ca. 3 mm long. Ovary 2-locular, 2 ovules rescences are unusual in Marlierea (or any Myr­ per locule; style 4 mm long; stigma punctate. ciinae), but appear to be by reduction of the lat­ Immature fruits globose, ca. 8 mm diameter eral branches as evidenced by the presence of when dried (to 2 cm diameter when fresh, fide two flowers at the same node and numerous bas­ Huber 13090). Seeds unkown. al bracts. The only known specimens of M. con­ vexivenia are past flower or are in fruit. Paratypes: VENEZUELA. Amazonas: Dept. A species of white-sand forests and scrub for­ Atures, Cerro Yavi, cabeceras del Rio Parucito, ests along black-water rivers, Marlierea convex­ afluente oriental del Rio Manapiare, 2100 m, ivenia is known from 110-160 m elevation in 05°43'N, 65°52'W, 29 Oct. 1986 (fr), Huber southern Amazonas state, Venezuela. Marlierea 11903 (MO, MYF); Sierra de Maigualida, sector caesariata is found in lower montane Clusia for­ nor-oriental, cabeceras nor-orientales del Rio ests and is known only from the Sierra de la Iguana, afluente del Rio Ventuari, 2150 m, Neblina at 650-700 m elevation. 05°40'N, 65°08'W, 24 Nov. 1989 (fr), Huber 13090 (MYF, SEL). Bolivar: Distr. Cedefio, Si­ Marlierea foveolata B. Holst, sp. nov. TYPE: erra de Maigualida, sector nor-oriental, cabec­ Venezuela. Amazonas: Dept. Atures, Sierra eras del Rio Chajura, afluente occidental del Rio Maigualida, NW sector, small valley along Erebato, aprox. 100 km en linear recta al SW an upper tributary of Cafio Iguana, 2000 m, del Campamento Entrerios, 2100 m, 05°33'N, 05°30'N, 65°15'W, 28 Feb.-3 Mar. 1991 65°13'W, 18 Nov. 1988 (fl, fr), Huber & 1zquier­ (fl), P.E. Berry, O. Huber & J. Rosales do 12829 (MYF, SEL). 4874 (Holotype: VEN; Isotype: SEL). FIGURE 6. A distinctive species due to the abundant fer­ Marlierea buxifolia Amshoff affinis, sed distineta: rugineous indument, small, stiff-coriaceous foliis basi subeordatis (nee obtusis), plerunque planis leaves, and few-flowered inflorescences. In the (nee manifeste revolutis), et infioreseentiis 3-fioris (nee Venezuelan Guayana, Marlierea foveolata is 5-fioris vel ultra). most similar to Marlierea buxifolia Amshoff, Shrub or small tree to 2 m tall; all parts but differs by having subcordate (vs. obtuse) leaf densely ferruginous-tomentose, the trichomes bases that are mostly flat (vs. margins strongly 146 SELBYANA Volume 23(2) 2002

1 l :J.. 6 1 a Marie Selby Botanical Gardens ~r:ium (SEL) .. I Scans of Type CoIIection"ScaIe 10 em

FIGURE 4. Eugenia magna (Liesner 19973, Isotype, SEL). HOLST: MYRTACEAE 147

revolute), and 3-flowered inflorescences (vs. 5- with an apiculum 0.1-05 mm long, tearing or more-flowered) inflorescences. lengthwise into 2 or3 irregular lobes or circum­ Marlierea foveolata is known from open scissile with one large lobe, but then with some rocky areas at 2000-2150 m elevation along the calyx tissue remaining attached to the hypanthi­ Bolivar-Amazonas border on Sierra de Maigual­ urn, glands conspicuous, convex; petals 4, orbic­ ida. The specific epithet refers to the deeply im­ ular and truncate at the base, ca. 2 X 2 mm, pressed-punctate leaf blades. white, sericeous without, glabrous within, inte­ Marlierea mcvaughii B. Holst, sp. nov. TYPE: rior petal slightly smaller with a prolonged apex Venezuela. Amazonas: Dept. RIO Negro, that in bud is tucked into the ring formed by the Cano Darigua, 11 km S of San Carlos de stamens. Stamens ca. 90-110; anthers basifixed, RIO Negro, 119 m, 01°56'N, 67°03'W, 1 0.1 mm long, parallel or slightly oblique, de­ Dec. 1981 (fl), H. Clark & P. Maquirino hiscing longitudinally; filaments 2.5-3 mm long, 8302 (Holotype: MO; Isotypes: INPA, NY, with few long trichomes towards the base. Ova­ SEL) FIGURES 7, 8. ry bilocular, ovules 2 per locule; style slender, 4-5 mm long; stigma slightly broader than the Species propria fructibus muricatis, trichomatibus style. Fruits slightly broader than long, 0.6-0.9 sericeis, fiavidis vel albidis, basi rubellis; foliis apice et basi rotundatis ad obtusis, et calicibus longitudinal­ X 0.7-1.0 cm, shallowly muricate, greenish yel­ iter vel circumscissis fissis. low, not fleshy, crowned by the hypanthium and persistent reflexed calyx lobe(s). Seeds 1; em­ Shrub or tree 3-12 m tall; branchlets, buds, bryo contortuplicate, hypocotyl elongate. petioles, lower leaf surface, inflorescence and flowers with whitish or dull yellowish antrorse Paratypes: BRAZIL. Amazonas: Municip. Sao sericeous indument; trichomes simple, unicellu­ Gabriel de Cachoeira, Rio Negro, Rio Demiti, lar, elongate, red-orange at base; bark of two 00050'N, 66°53'W; 1 Nov. 1987, Daly et al. year old wood gray-white, smooth to shallowly 5536 (MO); Rio Uneiuxi, 5 km above mouth, 8 longitudinally fissured; acicular appendages oc­ Nov. 1971, Prance et al 16184 (F, MICH, NY, curring around bases of petioles and bracts. US). COLOMBIA. Meta: ca 7 km S of San Fer­ Leaves petiolate, the petioles 3-6 mm long; 1 nando de Atabapo, 95 m, 3°55'N, 67°43'W, 28 mm thick, channeled; blades reddish-brown Apr. 1979, Davidse 16842 (MICH, MO, NY, above and drying dull, gray or gray-brown be­ VEN). VENEZUELA. Territorio Federal Amazon­ neath, elliptic-oblong to slightly obovate, 504- as: Dept. Atabapo, Laja Cucuta, RIo Atacavi, 9.0 x 2.4-3.7, coriaceous, upper surface gla­ 100 m, 03°12'N, 67°24'W, Nov. 1989 (fl), Ve­ brous except for at extreme base, persistently se­ lazco 1080 (SEL). Dept. Casiquiare, RIo Ata­ riceous below, glands drying dark; apex gener­ bapo, from San Fernando to 12 km upstream, ally rounded and then apically shortly obtuse to 110 m, 14 Jan. 1988, Aymard, Stergios & Cuello acute, margins revolute or just barely so,. base 6487 (NY, PORT, SEL); Dept. RIO Negro, Cano cuneate or acute and shortly decurrent on the Darigua, 11 km S of San Carlos de RIo Negro, inner angles of the petiole; midvein channeled 119 m, 1°56'N, 67°03'W, 27 Nov. 1981, Clark above for % the length of the blade, then plane & Maquirino 8294 (MO, NY); Dept. Rio Negro, towards the apex, prominent below, secondary Cano Adobo, ca. 25 km S of San Carlos de RIo veins 15-20 pairs, slightly elevated on both sur­ Negro, 120 m, 01°38'N, 66°58'W; 27 Jan 1980, faces of the leaf or nearly completely flat above, Liesner 8644 (MO, VEN); same locality and marginal vein about as strong as the secondaries, date, Liesner 8761 (MO, MICH, VEN); Playa 05-1.0 mm from the margin. a Inflorescence Tigre; San Fernando, RIO Atabapo, Feb. 1971, panicle, 3~8 cm long including the peduncle, 3- Medina 546 (VEN). 6 cm broad, or sometimes racemose and then only 2 cm broad, axillary or subterminal~ branches slightly compressed laterally, flowers From the specimens on hand, it appears that in groups of three at the end of the branches; flowering occurs during the early part of the dry bracts deciduous before anthesis; pedicels 0-4 season from November through January, and mm long;. bracteoles ovate-elliptic, acute, 1-1.5 fruiting from January to April. Marlierea mm long, deciduous at or before anthesis. Flow­ mcvaughii is found in low altitude, 100-120 m er buds broadly obpyriform, flowers 3-4 mm elevation, periodically inundated white sand long, 2.5-3.5 mm broad; hypanthium prolonged, shrubby savannas. It is confined to the RIo Ne­ ca. 1 mm beyond the sericeous summit of the gro basin of the Amazon River in Venezuela, ovary,. densely sericeous with ascending tri~ Colombia and Brazil. chomes within and without, explanate after an­ Marlierea mcvaughii represents yet another thesis; calyx glabrous within except for some intermediate between Marlierea and Calyptran­ reddish trichomes at very base, closed in bud thes. The calyx, completely closed in bud, tears 148 SELBYANA Volume 23(2) 2002

---M

FIGURE 5. Marlierea convexivenia (Stergios et ai. 13341. Paratype, NY). HOLST: MYRTACEAE 149

1 234 5 6 7 8 9 Marie Selby Botanical Gardens Herbarium (SEL) 1.... __ ..;;.;:;:;.;;;;~S;.;;ca;;;;.ns of Collection - Scale 10 em

FIGURE 6. Marliereafoveolata (Berry et al. 4874, Isotype, SEL). 150 SELBYANA Volume 23(2) 2002

either irregularly around, or through the apicu­ stout acicular appendages present at nodes of lum, to form 2 or 3(4) lobes (usually 2 larger stem and inflorescence; branchlets slightly com­ ones along with a smaller one, or, circumscis­ pressed in cross section. Leaves subcordate, pet­ sally to form a single large calyptra. \Vhile the ioles ca. 1 mm long, ca. 3 mm wide; blades dry­ direction of tearing of the hypanthium is usually ing dark on upper surface, elliptic to nearly ob­ a good specific character, in this case, both type long, a few leaves slightly ovate or obovate, of calyx opening can be found on the same 6.5-10 X 3-4.5 cm, coriaceous, upper surface plant! In either case, the hypanthium splits ir­ slightly impressed-punctate, lustrous, glands regularly to the summit, or even into the sta­ otherwise inconspicuous; apex obtuse to round­ minal zone to produce the ragged appearance of ed, base subcordate; midvein shallowly im­ Marlierea. pressed on upper surface, prominent below, nu­ These perplexing intermediates may eventu­ merous (ca. 16-20 per side), slightly elevated on ally end up in Calyptranthes, but until the ge­ upper and lower surfaces, marginal vein ca. 2 neric limits of the Myrciinae can be better de­ mm from and parallel to margin. Inflorescences fined, I believe it best to follow the more tradi­ few-branched panicles, 5-10 em long, axillary tional concept. Also, a very closely related spe­ or subterminal, solitary or paired in leafaxils; cies has been described as a Marlierea, M. peduncles 2-4.5 cm long; bracts ovate, ca. 3 mm spruceana O. Berg. It is fitting to dedicate this long, acute at apex, sericeous without, glabrous species to Rogers McVaugh, who first pointed within, deciduous at or before anthesis; pedicels out the "gray area" between Calyptranthes and 0-2 mm long; bracteoles ovate, 2-3 mm long, Marlierea (McVaugh, 1958b). Marlierea mcvaughii is distinguished by its sericeous without, glabrous within, deciduous at warty fruits, yellowish or whitish sericeous tri­ or before anthesis. Flower buds obpyriform, 3- chomes with redish bases, and flower buds that 4 mm long, obtuse apically but not apiculate, open by the two methods described above. Its densely sericeous without, sessile or on short, closest relatives are M. spruceana and M. sub­ stout pedicels; hypanthium prolonged ca. 2 mm cordata B. Holst. Marlierea spruceana has the above summit of ovary, explanate at anthesis; same general appearance, including the warty calyx closed in bud, tearing irregularly and fruits, but they are larger, 1.5-1.7 X 1.7-1.9 cm, mostly longitudinally into 2-4 irregular lobes, as are the leaf blades 9-15 X 4-6.5 cm, which persistent past anthesis; petals 4, orbicular to are abruptly acuminate at the apex. Marlierea broadly ovate, 2~3 mm long, sericeous without, subcordata has subsessile, subcordate, oblong glabrous within. Stamens numerous; anthers 0.3 leaves that are lustrous above. All three species X 0.3 mm; filaments 5-6 mm long. Ovary bi­ are sympatric. locular, ovules 2 per locule; style slender, ca. 6 Many species of Neotropical Myrtaceae bear mm long; stigma punctate. Fruits unknown. acicular appendages around leaf bases, bases. of Seeds unknown. inflorescence branches, bracts and bracteoles. It is not clear what the significance of these are, Paratype: VENEZUELA. Dept. Rio Negro, Rio either biologically or taxonomically, but, it is in­ Pasimoni, between its mouth and its junction teresting to note their occurence since in some with the Rio Baria and the Rio Yatua, 80 m, species they are noticeable, and most descrip­ Olo27-53'N, 66°32~35'W, 23-25 Jul. 1984 (past tions omit them. fl), Davidse 27751 (MO, SEL, VEN). Marlierea subcordata B. Holst, sp. nov. TYPE. Venezuela. Amazonas: Rio Bacia, entre la Marlierea subcordata belongs to a groups of boca de la Laguna Yuruvi y la Laguna Laja species that includes Marlierea mcvaughii B. Bajiiio, aproximadamente a 25 vueltas de la Holst and Marlierea spruceana O. Berg. It is boca, 7 Nov. 1994 (fl), B. Stergios, M. Nino, distinguished by having subsessile, subcordate /. Sikoura, N. Piven, P. Martinez & N. leaves that are lustrous on the upper surface and Friederich 16348 (Holotype: PORT; Iso­ with the midvein nearly plane (vs. leaves round­ type: MO). FIGURE 9. ed to cuneate at base with distinct petioles, Marlierea mcvaughii B. Holst et M. spruceana O. leaves drying dull on. the upper upper surface Berg affinis, sed distineta: foliis subsessilibus, subeor­ and the midveins noticeably sulcate near the datis, supra nitidis, nervo medio fere plano (nee foliis base). rotundatis ad euneatis, supra obseuris, nervo medio Marlierea subcordata is known from only manifeste su1cato; petiolo distincto). two collections in lowland evergreen forests Tree 8-10 m tall; branchlets, lower surfaces along periodically flooded river banks at about of leaves, and inflorescences densely sericeous 100 m elevation, along the Rio Barfa and Rio to nearly strigose with straw-colored trichomes; Pasimoni in Amazonas state, Venezuela. HOLST: MYRTACEAE 15]

FIGURE 7. Marlierea mcvaughii (Clark & Maquirino 8302. Holotype, MO). 152 SELBYANA Volume 23(2) 2002

Marlierea ventuarensis B. Holst, sp. nov. TYPE: Marlierea ventuarensis is related to Marlierea Venezuela. Amazonas: Dept. Atabapo, Cer­ schomburgkiana O. Berg, but differs by having ro Moriche, 04°40'N, 66°17'W, Oct. 1989 the leaves with shallowly sulcate midvein (vs. (past fl), L. Delgado 893 (Holotype: PORT; prominently convex), mostly elliptic to rarely Isotypes: NY, SEL). FIGURE 10. ovate blades (vs. obovate), and the upper surface Marlierea schomburgkiana O. Berg affinis, sed dis­ impressed-punctate (vs. not impressed punctate). tineta: nervo medio parum sulcato (nee manifeste Known from only two collections, Marlierea prominente); foliis plerunque elliptieis, rare ovatis (nee ventuarensis occurs in shrubby savannas or obovatis), supra impresso-punetatis. scrub forests at 200-350 m in the Rio Ventuari Shrub or slender tree to 8 m tall; branchlets basin of Amazonas state, Venezuela. and inflorescences sparsely sericeous with pale yellowish trichomes, leaves and flowers gla­ Myrcia intonsa (McVaugh) B. Holst, comb. brous; branchlets terete or slightly compressed nov.-Marlierea intonsa McVaugh, Mem. below the nodes. Leaves petiolate, the petioles New York Bot. Gard. 10(1): 85. 1958. TYPE. 2-3 mm long, shallowly channeled; blades dry­ ing dark brown to reddish brown with the upper Bolivar: Chimanta Massif, ChimanHi-tepui, surface darker, elliptic to less often ovate, 4.5- base of bluffs, 1600-1700 m, 5 Jun. 1953 8.5 x 2.0-3.2 em, coriaceous, upper surface im­ (fr), Steyermark 75670 (Holotype: MICHl; pressed-punctate, glands on lower surface dry­ Isotypes: Fl, MO!, NYl, VEN!). ing dark; apex abruptly acuminate, the tip 1.5- 2 em long, base acute to obtuse, margins plane; After having described this as a Marlierea, midvein shallowly sulcate on upper surface, sec­ McVaugh (1969), placed it into the synonomy ondary veins scarcely visible, ca. 8--10 pairs, of Myrcia defiexa (Poir.) DC. with some ques­ marginal vein ca. 1 mm from the margin and tion. With several more recent collections avail­ parallel to it, higher order venation not visible. Inflorescences panicles, 3-5 cm long including able, it is clear that M. intonsa is distinct from the peduncle, 2-3 cm wide, solitary or paired in M. defiexa, though belonging to the same spe­ the leafaxils, axillary or subterminal, 9-15- cies complex. It is distinguished within the "ca­ flowered, axes compressed in cross section, the lycampe" complex (see under M. calycampa) by lateral branches nearly perpendicular to the main the long petioles 0.5-2 cm long), and the round, axes; peduncle 1.5-2.5 cm long; bracts decidu­ permanently densely pubescent fruits. Myrcia ous at or before anthesis, not known; pedicels defiexa has smaller flowers, ellipsoid and sparse­ lacking in terminal flowers of 3-flowered ly pubescent or glabrescent fruits, and petioles branches or 4-5 mm long in lateral ones; brac­ less than I cm long. teoles not known. Flower buds obovoid, im­ Collections from northern Brazil extend the mature buds ca. 3.5 mm long, closed in bud, range. They differ slightly from the Venezuelan apiculate at apex, the apiculum pubescent and collections in the less-pronounced pebbled leaf with a tiny hole similar to an ostiole, flowers undersurface and that the pubescence of the otherwise glabrous except for apex of petals; hy­ fruit, branches of the inflorescence, and under­ panthium prolonged ca. 2 mm above summit of surface of the leaves is tightly appressed and lus­ ovary, strongly reflexed after anthesis; calyx trous as compared with the undersurface of the nearly closed in bud, tearing regularly or irreg­ ularly and mostly longitudinally into 3 or 4 leaves being strongly pebbled, and the trichomes lobes, 2-3 mm long, persistent past anthesis, dull and slightly more spreading. apices persistent; petals 4, coalesced into a cap, Specimens examined: BRAZIL. Amazonas: broadly ovate, margins irregularly lobed on Plato da Serra Araca, parte S.E. da Serra Norte, some or dentate, apically ciliate. Stamens nu­ 1150-1250 m, 00°51 'N, 63°22'W, 14 Feb. 1984 merous; anthers 0.2 x 0.2 rum; filaments ca. 3 (fl), Amaral 1572 (NY); Distr. Agropecuario, mm long. Ovary bilocular, ovules 2 per locule; Reserva 1501, km 41 of the WWF/INPA MCS style slender, ca. 5 rum long; stigma punctate. Project, 50-125 m, 02°24'26"-2°25'31", Fruits globose, 1-1.3 cm diameter, calyx lobes 59°43'40"-59°45'50"W, 26 Jun. 1989 (fr) , Mori absent. Seeds 1; embryo contortuplicate, hypo­ et al. 20538 (MO); Serra Araca, western slope cotyl elongate. of southern massif, 800 m, 00040'N, 63°18'W, Paratype: VENEZUELA. Amazonas: Cuenca 19 Mar. 1984 (fr), Pipoly & Samuels 6845 (MO, del Rio Manapiare, cerros ubicados entre el Cer­ NY). VENEZUELA. Bolivar: Gran Sabana, 7 km ro Morrocoy al sur y la Serrania Colmena al al N. de la Misi6n de Santa Teresita de Kavan­ norte, 200-350 m, 05°20'N, 66°1O'W, 29 Jan. ayen, 1160 m, 20 Feb. 1978 (fr), Steyermark et 1977 (fr) , Huber 446 (MICH, VEN). al. 115519 (MICH, MO, VEN). HOLST: MYRTACEAE 153

E E C\I c

3cm

E E ('I)

FIGURE 8. Marlierea mcvaughii. A. Habit. B. Detail of inflorescence branch. C. Fruit. D. Petal. E. Calyx lobe. F. Acicular appendages. G, H. Detail of calyx opening. Drawn from Clark & Maquirino 8302. Holo­ type, MO. IS4 SELBYANA Volume 23(2) 2002

Myrcia kylistophylla B. Holst, sp. nov. TYPE: 64°26'W, 22-27 Mar. 1967 Cfr), Steyermark Venezuela. Bolivar: Distr. Cedeno, Sierra de 97855 (NY, VEN). Maigualida, sector nor-oriental, altiplanicie tepuyana disectada sobre granito en las ca­ Myrcia kylistophylla is related to Myrcia gen­ beceras del Rio Chajura, afluente occidental tryi B. Holst from Cerro Huachamacari, Ama­ del Rio Erebato, aprox. 100 kIn (en linea zonas state, Venezuela, but differs principally by recta) al SW del Campamento Entrerfos, having strongly revolute leaf blades (vs. plane) 2100 m, Oso33'N, 6So13'W, 18 Nov. 1988 that are 3-S.S X 1.S-4 cm (vs. 1.4-1.8 X 0.7- (fl), O. Huber & L. Izquierdo 12794 (Ho­ 0.9 em). Myrcia kylistophylla grows on rocky lotype: MYP; Isotype: SEL). FIGURE 11. outcrops or along streams on tepui summits at Myrcia gentryi B. Holst affinis, sed distincta: foliis ca. 1900-2100 m elevation in central-western valde revolutis (nec planis), 3-5.5 X 1.5-4 cm (nec Bolivar state of Venezuela. Species in this com­ 1.4-1.8 X 0.7-0.9 cm). plex, including M. gentryi, M. bolivarensis CSteyerm.) McVaugh, and M. minutifiora Sagot Shrub 1-3 m tall; vestitute softly yellowish have greatly reduced, often unbranched inflores­ pilose to villous (fading to whitish) throughout cences and 4- or S-merous flowers and im­ on young vegetative growth and inflorescence; pressed-punctate glands on the upper leaf sur­ branchlets whitish tomentose, laterally com­ face. The specific epithet refers to the im­ pressed. Leaves petiolate, the petioles 2-3 mm pressed-punctate leaves. long, channeled; blades drying whitish green on lower surface, ovate to broadly elliptic, though Myrcia liesneri B. Holst, sp. nov. TYPE: Vene­ the strongly revolute leaves obscures the shape, zuela. Amazonas: Dept. Rio Negro, on hills 3-S.S X 1.S-4 cm, stiff-chartaceous; young 1.S kIn South of Cerro de la Neblina Base leaves softly long-pilose on both surfaces, gla­ Camp which is on Rio Mawarinuma, 140- brescent on upper surface, persistently pilose on 340 m, OooSO'N, 66°lO'W; 4 Dec. 1984 lower surface, drying whitish green; glands (past fl), R. Liesner & D. Bell 17494 (Ho­ scarcely apparent on lower surface, deeply im­ lotype: ASU; Isotypes: SEL, VEN). pressed-punctate on upper surface; apex obtuse FIGURE 12. to rounded, base obtuse, appearing acute due to Species propria foliis breviter petiolatis, petiolis the strongly revolute margins; midvein deeply suberoso-rimosis, nervo medio supra prominente, fio­ sulcate, secondary veins 6-8 per side, slightly ribus minute puberulis, 5-meris, hypanthio valde prod­ engraved on upper surface, slightly elevated on ucto. lower surface, tertiary venation inconspicous, Tree to 4 m tall; mostly glabrous except for marginal vein slightly arching between the sec­ light brown villous indument on midveins of ondaries, ca. 2-3 mm from the margin. Inflo­ young leaves and branchlets; stems eventually rescences 3-flowered (possibly occasionally 1- glabrous, the bark peeling in strips. Leaves flowered), O.S-1 cm long, solitary in the leaf ax­ short-petiolate, the petioles 3-S mm long, stout, ils;peduncle O.S-nearly 1 cm long, densely yel­ corky-rimose, diverging at an angle from the lowish villous; bracts unknown; pedicels plane of the blade; blades drying pale green on lacking; bracteoles spatulate, 2-3 mm long and lower surface, lustrous on upper surface, ovate equalling the flower bud, deciduous before an­ or lanceolate to narrowly elliptic, 12-19 X 4- thesis. Flowers sessile, 4-merous, villous with­ 6.S cm, chartaceous to coriaceous, glabrous ex­ out; hypanthium prolonged above summit of cept on midvein when young, glands inconspic­ ovary ca. 1 mm; calyx lobes broadly ovate, ca. uous; apex evenly to abruptly acuminate, base 2 mm long, obtuse to rounded at apex; petals ca. rounded to subcordate; midvein prominently 2 mm long, glabrous, rounded at apex. stamens convex to binconvex on upper surface, second­ relatively few, ca. 20-30; anthers oblong; fila­ ary veins numerous, parallel, lightly impressed ments ca. 3 mm long. Ovary 2-locular, 2 ovules on upper surface, lightly to prominently raised per locule; style ca. 2 mm long; stigma punctate. on lower surface, marginal vein 2-4 mm from Fruits globose, ca. 7 mm diameter, I-seeded, margin and fairly parallel with margin, an ad­ glabrous, calyx lobes mostly persistent. Seeds 1, ditionalminor marginal vein evident on larger embryo contortuplicate, hypocotyl elongate. leaves and < I mm from margin. Inflorescence Paratypes: VENEZUELA. Amazonas: Sierra paniculate, sparsely branched, 4-7 cm long, sub­ Maigualida, sector noroccidental, en pequeno terminal, 3-S-fiowered(but fragmentary on valle afluente izquierdo del Cano Iguano medio, available specimens, and perhaps more), axes 2000 m, OSo30'N, 6So1S'W, 2 Mar. J 991 (fr) , puberulent; peduncles 2.S-4 cm long; bracts ca­ Huber, Berry & Rosales 13115 (SEL); Bolivar: ducous before anthesis; pedicels lacking or to 3 Cerro Jaua, cumbre de la porci6n Central-Occi­ mm long; bracteoles deltoid, ca. 1 mm long, ca­ dental de la meseta, 1922-2100 m,04°4S'N, ducous at anthesis. Flowers S-merous, minutely HOLST: MYRTACEAE 155

FIGURE 9. Marlierea subcordata (Stergios et al. 16348, Isotype, MO). 156 SELBYANA Volume 23(2) 2002

puberulent without; hypanthium constricted marginal veins 2, the inner, more prominent one above ovary, broadly flaring above, prolonged ca. 2 mm from the margin and mostly parallel 1.5-2 mm long; calyx lobes wider than long, to it. Inflorescence an abbreviated, bracteate ra­ slightly tearing basally post-anthesis; petals un­ ceme, solitary or clustered in leafy or leafless known. Stamens ca. 3 mm long, number un­ axils, 5-7 mm long, the axis itself 4-5 mm long, known; anthers oblong, ca. 0.2 mm long. Ovary a few flower pairs with only one flowering de­ 2-10cular; ovules 2 per locule; style and stigma veloping to anthesis, giving the inflorescence a unknown. Fruits oblate-globose, ca. 1 cm di­ seemingly alternate arrangement (it is not clear ameter, glabrous, calyx persistent. Seed 1, testa if the non-developing flowers are natural or chartaceous; cotyledons fleshy, scarcely folded, caused by external agents); bracts ovate, 1-1.5 hypocotyl elongate. mm long, acute to rounded at apex, puberulent without, glabrous within; pedicels ca. 1 mm Paratype: VENEZUELA. Amazonas: Dept. Ata­ long, puberulent; bracteoles 3-4 mm long, bapo, Calio Iguapo, alto Rio Orinoco, 180 m, prominently exceeding the summit of the ovary 03°08'N, 65°27'W, 20 Feb. 1990 (fr), Aymard & by 2-3 mm, ovate, keeled, acute at apex, puber­ Delgado 8182 (SEL). ulent within, glabrous without. Flowers 4-mer­ ous, puberulent without, open in bud; hypanthi­ Myrcia liesneri is known from two collections urn prolonged ca. 1 mm above the summit of the from southern Amazonas State in Venezuela, in ovary, spreading at anthesis, irregularly tearing lowland evergreen moist forests between 140 through the staminal ring into 4 lobes; calyx and 400 m elevation. It is characterized by the lobes somewhat irregular, generally orbicular, short-petiolate leaves, flaky petioles, convex ca. 1.5 mm diameter, ciliate, glabrous within, midvein, minutely puberulent, 5-merous flowers. puberulent without; petals similar to the calyx Perhaps most closely related to Myrcia revolu­ lobes in size and indument. Stamens numerous; tifolia McVaugh, M. liesneri is distinguished by anthers ellipsoid, ca. 0.5 mm long; filaments 2- the relatively slender leaves (2.5-3.5 times as 3 mm long. Ovary bilocular, densely puberulent, long as wide vs. < l.5 times), acuminate leaves ovules 2 per locule; style ca. 1 cm long; stigma (vs. rounded to emarginate), and the minutely punctate. Fruits 2.5-3.5 cm diameter, dark pur­ puberulent flowers (vs. mostly glabrous). ple at maturity. Seeds 1, embryo kidney-shaped, Myrciaria puberulenta B. Holst, sp. nov. TYPE: cotyledons fused. Venezuela. Amazonas: Dept. Atures, vicin­ Paratypes: VENEZUELA. Amazonas: Margen ity of and upstream from damsite, along derecha del Rio Cuao, desde raudal de Danta, north side of Rio Cataniapo, 45 kIn south­ entre Calio Raya y Picure, 04°54'-05°03'N, east of Puerto Ayacucho, 05°35'N, 67°34-46'W, 23 Sep. 1996 (fr), Castillo 4191 67°15'W, 13 May 1980 (fl), J. Steyermark, (SEL); Headwater of Calio Kaenaeruoto, right G. Davidse & F. Gudnchez 122403 (Holo­ bank tributary of upper Cuao River, 750-800 m, type: VEN; Isotypes: MO, SEL). 05°40'N, 66°47'30"W, 28 Sep. 1985 (st), Zent FIGURE 13. 985-14 (MYF). Ad omnibus speciebus generis regiis differt: intlo­ rescentiis axibus elongatis, 4-5 mm longis (nee 0-2 Myrciaria puberulenta is known from tall ev­ mm longis) et bracteolis 3-4 mm longis, supra ger­ ergreen moist forests at 100-800 m elevation in mine valde productis (nec < 1 mm longis, vix prod­ uctis). the Cataniapo and Cuao river basins of north­ western Amazonas state, Venezuela. Myrciaria Tree to 15 m tall; branchlets, petioles, midvein puberulenta is apparently rare since only three on lower surface of leaves, and inflorescences collections have been made from the heavily puberulent with whitish trichomes; branchlets collected area just south and southeast of Pto. slightly laterally compressed in cross section. Ayacucho. It is distinguished from all other Leaves petiolate, the petioles 8-10 mm long, Myrciaria species in the region by the pro­ slightly channeled apically, mostly terete in longed, 4-5 mm long inflorescence axis (vs. 0- cross section; blades mostly oblong-elliptic, 2 mm) and the long bracteoles, 3-4 mm long rarely lanceolate, 10-18 X (2.5-)3-5 cm, char­ and exceeding the summit of the ovary (vs. < 1 taceous, thin, abundantly glandular with minute, mm long and not exceeding the ovary summit). closely spaced pellucid glands; apex acuminate, The prominent bracteoles give the inflorescence the apicule sharply acute; margins plane to the aspect of a Plinia, though the cleanly cir­ scarcely revolute; base obtuse to rounded; mid­ cumscissile hypanthium and fused cotyledons vein plane to convex on upper surface, but clearly place this species in Myrciaria. somewhat channeled between the halves of the The fruits of Myrciaria puberulenta are de­ blade, lateral veins numerous, inconspicuous, scribed on the label of Zent 985-14 as "fruit HOLST: MYRTACEAE 157

3 5 6 8 Marie Selby Botanical Gardens Herbarium (SEL) Scans of Collection - Scale 10 cm

FIGURE 10. Marlierea ventuarensis (Delgado 893, Isotype, NY). 158 SELBYANA Volume 23(2) 2002

! f i I Ii 1 3 5 6 7 8 Marie Selby Botanical Gardens Herbarium (SEL) Digital Scans of Type Collection - Scale 10 cm

FIGURE 11. Myrcia kylistophylla (Huber & Izquierdo 12794, Isotype, SEL). HOLST: MYRTACEAE 159

3 7 8 Marie Selby Botanical GardensHerbarium (SEL) Digital Scans of Type Collection - Scale 10cm

FIGURE 12. Myrcia liesneri (Liesner & Bell 17494, Holotype, ASU). 160 SELBYANA Volume 23(2) 2002

FIGURE 13. Myrciaria puberulenta (Steyermark et al. 122403, Isotype, MO). HOLST: MYRTACEAE 161

dark purple skin (with reddish tint), clear-cream frutos verdes; 130 m, 03°20'S, 72°55'W; 21 Feb. meat inner (fruit collected by informant separate 1989 (fr), Vasquez & Jaramillo 11771 (ASU, trip; looks and tastes like a large grape)." The MO). plant is in cultivation by the Piaroa Amerindians Calycolpus roraimensis Steyerm., Fieldiana, of the region, who call it 'kuyaeri', and plants Bot. 28: 1009. 1957. were observed in an "old garden/secondary for­ est" area. Another collection (Castillo 4191) de­ This somewhat rare species in the Gran Sa­ scribes the fruit as being green with tinges of bana of southeastern Bolivar State in Venezuela wine red when mature. and adjacent Guyana was collected in a cerrado type patch of vegetation in southern Peru along the border with Bolivia. It represents a consid­ COUNTRY RECORDS AND NOTES erable range extension for this species. Following, in alphabetical order by genus and PERU. Madre de Di6s: Tambopata Province, species, are country records and notes of select­ Pampas del Heath, 200 m, 12°50'S, 68°50'W, 26 ed species. Specimen citation is alphabetical by Feb. 1990 (fr) , Gentry & Nunez 69605 (MO); country, major political division, collector, and same region, 200 m, 12°57'S, 6S053'W, 21 Jun. then numerically by collection. Some of the new 1996 (st), Beltran et al. 2276 (SEL). country records have appeared in published checklists elsewhere, but usually without full 10- Calycorectes O. Berg cality information. Types are cited for entries other than new distribution records, and those The sole basis that has been used for distin­ seen for this study are indicated with an excla­ guishing Calycorectes from Eugenia is the pro­ mation mark. longation of the hypanthium above the summit of the ovary and the irregular tearing of the ca­ lyx lobes (McVaugh 1968). As Landrum (1986) Blepharocalyx O. Berg and McVaugh (1958b, 1968, 1969) have pointed Blepharocalyx eggersii (Kiaersk.) Landrum, Fl. out, the degree of fusion of the calyx lobes into Neotr. 45: 121. 1986. a hypanthium for delimiting genera has not been found to be reliable and appears to have evolved New genus and species for Guyana and separately in several genera. Therefore, it seems Peru; previously known from Guadeloupe, St. unlikely that this method of distinguishing Ca­ Vincent, Venezuela, and Amazonian Brazil. lycorectes from Eugenia will hold. Likewise, to GUYANA. Potaro-Siparuni Region, summit of maintain Eugenia as a single, large, and likely Mt. Kopinang, 1700-1800 m, 05°00'N, polyphyletic group in the Eugeniinae, based pri­ 59°55'W, 8 Apr. 1988 (fr), Hahn et al. 4359 marily on the four distinct calyx lobes is unten­ (MO), 4389 (MO). PERU. Loreto: Maynas Prov­ able. ince, Iquitos, Allpahuayo, Estaci6n Experimen­ No other characters to maintain Calycorectes tal del Instituto de Investigaciones de la Ama­ apart from Eugenia can be offered here, but it zonia Peru ana CHAP), 29 May 1990 (fr) , Vas­ is premature to sink it. Nevertheless, one of the quez et at. 13774 (MO). two species in the Venezuelan Guayana recog­ nized by McVaugh, Calycorectes yatuae, is Calycolpus O. Berg transferred to Eugenia (see under E. yatuae be­ low) because of its close relationship to several Two species previously thought restricted to species in that genus (see under E. diplocampta the Guayana Shield have been found in Ama­ for additional notes). The other species of Ca­ zonian Peru. lycorectes that has been described form the Ve­ nezuelan Guayana, Calycorectes enormis Mc­ Calycolpus calophyllus (H.B.K.) O. Berg, Lin­ Vaugh, appears to be closely related to the type naea 27: 381. 1856. species of the genus, Calycorectes grandifolius A common species of southwestern Venezue­ O. Berg, and will be maintained in that genus la, upper Rio Negro of Brazil, and southeastern until a revision of the Eugeniinae genera can be Colombia. The Peruvian collection represents a made. new record for the country and an unusually tall plant for the genus. Calyptranthes Sw. PERU. Loreto: Maynas Province, Quebrada Sucusari, aftuente izquierdo del Rio Napo; Calyptranthes is generally a well distin­ bosque primario, arbol, 30 m; fuste fuertemente guished genus by virtue of the calyptrate calyx hendido, con ritidoma en tiras largas, dejando and myrcioid embryo. See McVaugh (1958b, una superficie lisa de color marr6n amarillenta, 1969) for a discussion of its relationship with 162 SELBYANA Volume 23(2) 2002

Marlierea and, in comments included with Mar­ C. multiflora. McVaugh noted that the buds of lierea macvaughii. C. mult!flora generally darken in drying, where­ as in C. fiorifera they dry pale yellow brown. McVaugh, Mem. Calyptranthes amshoffae That however may be a result of the drying New York Bot. Gard. 10(1): 70. 1958. method used, or, it may be characteristic of the Previously only known from Suriname, a col­ plants from Rio Pargueni, the type locality of C. lection from French Guiana. extends the range jlorifera, but does not seem to be of significance. eastward. The main confusion in Calyptranthes jlorifera FRENCH GUIANA: Montagne de la Trinite, som­ though, was it being based on two species. The met NE in creek, in loco forest below rocky first, C. multiflora as noted above, includes the slopes of inselberg; 25 Jan. 1984 (bud); de type and one of the paratypes, Wurdack & Mon­ Granville et al. 6283 (MO) .. achino 39781. The second species includes the other paratype cited, Schultes & Lopez 8918, Calyptranthes macrophylla O. Berg in Mart., and two fruiting specimens cited (McVaugh Fl. Bras. 14(1): 45. 1857. 1958b) as being referable to C. multiflora, but This species was included by McVaugh in his later placed by McVaugh (1969) in C. jlorifera. treatment of Myrtaceae for the Flora of Peru as The two fruiting specimens were used as the ba­ possibly occuring in Peru because of the prox­ sis of contrasting what was thought to be the imity of the type collection in Acre, Brazil, but "large fruited" C. jlorifera with the "smaller until now, no collections have been reported fruited" C. multiflora. Also referred to C. .fiori­ from within the borders. The Colombian collec­ fera by McVaugh (1969) is a flowering collec­ tion cited below is also a new record for the tion of Wurdack & Adderley 43644 which be­ species in that country. longs with the second species. These, together COLOMBIA. Vichada: rio Guaviare, regi6n with a large number of other collections from Amanabel, 250 m, 23 Nov. 1948 (fl), Molina & widespread localities in northern South America Barkley 18V (US). PERU. Loreto: Mariscal Cas­ are here recognized as C. forsteri O. Berg, with tilla, Caballo Cocha, 106 m, 03°55'5, 70030'W, the type collections from the Lesser Antilles. 12 Jul. 1987 Cfr) , Vasquez & Jaramillo 9251 The South American material differs from the (MO). Antillean in that the leaves are distinctly petio­ late and acute to rounded at the base as com­ Calyptranthes multiflora O. Berg in Mart., FI. pared with sub sessile and broadly rounded to Bras. 14(1): 42. 1857. TYPE: Peru. Poeppig subcordate .. Some of the collections from the 2684 (Isotype: F!). Lesser Antilles have buds that open by longitu­ Calyptranthes jlorifera McVaugh, Mem. New dinal, tears as in Marlierea, and, according to York Bot. Gard. 18(2): 72. ]969. TYPE: Vene­ the original description by Berg and a more re­ zuela. Amazonas: Rio Pargueni, 90 m, 10 Dec. cent one by Urban (1895), a calyptrate type 1955, Wurdack & Monachino 39782 (Holotype: opening, though I have not seen any specimens NY!; Isotypes: MICH!, VEN!). myself. The South American material is strictly Calyptranthes multiflora is an often collected, calyptrate. Also, the flowers are generally larger widespread shrub or small tree in the Upper Am­ in the Antillean specimens. Both Urban (1895) azon and Orinoco basins. It is characterized by and McVaugh (1969) point out that C. forsteri small elliptic leaves with a convex midvein, could just as well be placed in Marlierea as in paired panicles in the leafaxils with some de­ Calyptranthes. The method of opening of the ca­ gree of reddish pubescence on their branches, lyx. is not a viable generic character in the Myr­ and relatively small fruits. In the original pub­ ciinae. A closely related species, C. nigrescens lication of C. jlor!fera, McVaugh suggested that B. Holst, is described and contrasted below. it was so much like C. multijlora that the two Following are representative specimens of C. might conceivably be conspecific. In my opinion multiflora. The collection for Bolivia is a new they are. record of the species for that country. See also A paratype of Calyptranthes jlorifera, Wur­ McVaugh (1958b, 1969) for additional collec­ dack & Monachino 39781 (F), matches an iso­ tion citations. type C. multiflora, (F!), nearly exactly except BOLIVIA. Santa Cruz: Provo Velasco, Parque that the inflorescence branches are only sparsely Nacional Noel Kempff Mercado, Estaci6n Flor pubescent and the buds tend to dry a lighter col­ de Oro, ca .. 200 m, 13°33'S, 61000'W, 23 Sep. or. More recent collections of C. mult(fiora how­ 1995 (fl), Vargas et al. 3847 (MO): BRAZIL. ever show that amount of pubescence to be var­ Amazonas: Entre os Municipios de Maraa e Ja­ iable. Even the type of C. jlorifera, has a fair puni, Rio Japura, 0l045'S, 60000'W; 8 Nov. 1982 amount of the reddish trichomes on the branches (past fl), Cid & Lima 3530 (MO, NY); Rio Ne­ of the inflorescence that are commonly found in gro, rio Waupes, Matapi, 8 Dec. 1978 (fr), Dam- HOLST: MYRTACEAE 163

zao 3023 (MO); Tapuruquara, Rio Negro, 14 4025 (MICH, MO, VEN); Dept. Atures, Rio Ca­ Oct. 1978 (bud), Nascimento 635 (MO). Rorai­ taniapo, 200-300 m, 05°35'N, 67°15'W, II May rna: Boa Vista, regiao do Rio Cauame, afluente 1980 (fr), Steyermark, Davidse & Guanchez do Rio Branco, 12 Feb. 1977 (fr), Rosa & Cor­ 122270 (MICH, MO, VEN); 100 m, 13 May deiro 1479 (MO). VENEZUELA. Amazonas: Sta. 1980 (fr), Steyermark, Davidse & Guanchez Barbara del Orinoco, 90 m, 03°59'N, 67°03'W, 122421 (INPA, MICH, MO, U, VEN). 12 May 1978 (fr), Steyermark et al. 117122 The following collection is also referable to (MO, VEN). Anzoategui: Rio Mapire, afluente Calyptranthes ruiziana, but has abundant red­ norte del Rio Orinoco media, 07°30'-08°00'N, dish loose dibrachiate trichomes on the inflores­ 64°30'-65°00'W, 11 Oct. 1986 (bud), Rosales & cence and young leaves. BRAZIL. Amazonas: Rio Valles 066 (MO). Apure: Dist. Pedro Camejo: Negro, Above and below junction with Rio Rio Cinaruco, 35 m, 06°32'N, .67°22'W, 30 Apr. Branco, east of Carvoeiro, OJ °25'N, 61°05- 1977 (fr), Davidse & Gonzalez 12424 (MO, 20'W, 26 Jun. 1979 (imm fr), Poole 1637 (MO, VEN); Dist. Munoz, Cano Caicara, 80 m, NY). 07°34'N, 69°15'W, 3 Mar. 1979 (fr), Davidse & The group of species to which Calyptranthes Gonzalez 14738 (MO, VEN). Bolivar: Cano ruiziana belongs is very easy to recognize, but Pablo, tributario del Rio Caura, 240 m, 06°14'N, as far can be discerned, has no name. Specific 64°24'W, 7 May 1982 (fr), Morillo & Liesner limits within the group, however, are not quite 8961 (MO, VEN); Dist. Cedeno: Cano Chaviri­ so easily defined. In South America the species pa, carretera Caicara-EI Burro, 16 Apr. 1984 all have carinate branchlets with low longitudi­ (fr), Stergios et al. 8606 (MO, PORT); Rio Par­ nal wings or keels on both sides of the young gueni, 1-10 km above mouth, middle Rio Ori­ stems that terminate at the nodes midway be­ noco, 90m, 10 Dec. 1955 (fl), Wurdack & Mon­ tween the bases of the opposite leaf pairs, a achino 39781 (MICH, US); Wurdack & Mona­ deeply sulcate midvein,dibrachiate trichomes, chino 39782 (MICH, NY, type of C. jlorifera). numerous parallel ascending lateral veins, and frequently have paired panicles in the leafaxils. Calyptranthes rniziana O. Berg, Linnaea 27: The characters that have been used for separat­ 22. 1855. ing the species have been: leaf shape, relative The specimens from the Rio Negro basin cit­ amount of pubescence and relative number and ed below do not match perfectly with a "prob­ size of flowers. It is interesting to note that Mar­ able" isotype of C. ruiziana [Peru. Ruiz 5105 lierea sect. Myrciopsis is very similar in the (US!)], and are referred here with some doubt, same general appearance, but the wings of the the flowers are slightly smaller, and the inflores­ stem terminate directly below the petioles and cence tends to be more spicate with the lateral are sometimes continuous with them. Other branches very short or even absent, noticeably South American species in this complex are: Ca­ so in Gentry et at. J0910. Calyptranthes affinis typtranthes pulchella DC., C. Lucida DC., C. af­ O. Berg from Rio de Janeiro, Brazil is closely finis o. Berg, C. ruiziana O. Berg, C. bipennis related. O. Berg, C. tridymantha Diels, and C. pullei A single collection from the Sierra de la Neb­ Burr. ex Amsh. This complex is much in need lina at 1100 m elevation, Maguire et al. 42185 of a revision, as well as the entire genus. (MICH, VEN) is completely glabrous except for There are also some Central American and the vegetative buds, but otherwise matches col­ West Indian Calyptranthes that seem to belong lections from lower altitudes. to this assemblage, but at least the Central Specimens examined: BRAZIL. Amazonas: American ones are not as consistant with the ori­ Mouth of Rio Castanho and Raudales Uayaniri, entation of the wings. Some have 4-angled Rio Padauiri, 1-15 May 1946 (bud), Cardona branches and others have the wing orientation of 1256 (MICH, US, VEN); Rio Negro, near Ilha Marlierea sect. Myrciopsis. da Silva, 16 Jan. 1978 (imm fr), Steward et al. A Field Museum photograph of the type of C. 370 (MO, NY); Rio Negro, near Ilha Proved­ ruiziana O. Berg, #29480, is of a Eugenia in the encia, 22 Jan. 1978 (bud), Steward et al. 509 "florida" group and not of a Calyptranthes. (MO, NY). VENEZUELA. Amazonas: Cano Pavon, Rio Orinoco, 25 Mar 1974 (fl), Gentry et al. Eugenia L. 10910 (MICH, MO, VEN); Dept. Rio Negro, Rio Negro, Piedra de Coeui, 22-23 Dec. 1947 The largest ·of the New World genera, Euge­ (fr), Schultes & Lopez 9429 (F); Dept. Atures, nia is conspicuously absent from the summits of rio Cataniapo, 70-90 m,05°34'N, 67°31'W, 6 the tepuis or sandstone mesas, as well as above Aug. 1980 (fr), Guanchez 1482 (TFAV); Dept. approximately 2500 melevation in the Andes, Rio Negro, San Carlos de Rio Negro, 120 m, where it is largely replaced by the closely related Ol °55'N, 67°05'W, 29 Nov. 1977 (fr), Liesner genus Myrcianthes. 164 SELBYANA Volume 23(2) 2002

Eugenia armeniaca Sagot, Ann. Sci. Nat. ser 6, lleweiynii display several leaf forms. A type 20: 190. 1885. photograph of E. chrysophyllum shows clearly the leaf shape found in many Orinoco area col­ The following two collections are the first that lections such as Gentry & Stein 46356, Davidse have been made of this species in ca. ]40 years, & Gonzalez 12420, 13885, 15643, Plowman as well as a new record for Guyana. The spec­ 13479. The color of the pubescence also varies imens match perfectly with a Field Museum apparently with age of the specimen and drying photograph (No. 23512) and the large, densely conditions. Young, well preserved leaves have yellow-brown pubescent fruits in combination lustrous golden sericeous trichomes on the un­ with the looping marginal veins are quite dis­ dersurface, but lose color and turn grayish with tinct. age. A peculiar feature that shows up well in the GUYANA. Kanuku Mountains, 150 m, type photograph of E. chrysophyllum as well as 03°06'N, 59°25'W, 16 Feb. 1985 (fr), Jansen­ in many specimens from Venezuela, is the Jacobs et ai. 252 (US); Kanuku Mountains, sharply angled mosaic appearance of the leaves 280-400 m, 03°08'N, 59°23'W, 21 Feb. 1985 when dry. It is caused by a discoloration of cer­ (fr), Jansen-Jacobs et ai. 358 (MO, NY, US). tain leaf areas that are distinctly bordered by the Eugenia cachoeirensis O. Berg in Mart., FI. secondary veins. It does not occur on all collec­ Bras. 14(1): 265. 1857. TYPE: Brazil. Ama­ tions however and is not obvious in fresh leaves. zonas: prope San Gabriel da Cachoeira, ad The following description slightly amplifies Rio Negro, Jan.-Aug. 1852 (bud), Spruce Amshoff's in the Flora of Surinam to include the 2274 (Isotype: NY!). Venezuelan, Colombian, and Peruvian speci­ mens cited below. Eugenia lingua O. Berg, syn. nov. TYPE: Bra­ Shrub or small slender tree 3-6 m, branchlets, zil. Amazonas: Provo Rio Negro, in vicinibus inflorescence, petioles, densely covered with Barra, Apr. 1851 (fl), Spruce 1161 (Isotype: crisped reddish-brown trichomes, sericeous on NYl). lower leaf surface. Leaf blades drying olive­ A species restricted to the upper Rio Negro and its affluents and previously known from Ve­ green to yellow-green above, gland dots not ap­ parent or present as small depressions above and nezuela and Brazil. Not surprisingly, it has been only slightly so below, oblong to elliptic, 8-13 found on the Colombian side of the river. Eu­ (-16) cm long, 3-6 cm wide, 2-3.3 times longer genia lingua is merely a narrow-leaved form. than broad, sub-coriaceous, midvein plane to COLOMBIA. Guianfa: Between the tri-junction slightly concave, occasionally convex or wrin­ of the Rio Negro, Brazo Casiquiare, Rio Guainia kled towards the base, prominent below, primary and San Carlos de Rio Negro, 75 m, 01 °55'N, 67°04-0TW, 26 Jul. 1984, Davidse 27995 (MO, lateral veins convex above and below, ascend­ VEN). ing, parallel, aprox. 15-25, marginal vein from less than 1 to 2.5 mm from the margin and about Eugenia chrysophyllum Poir. in Lam., Encyc. as prominent as the secondaries, glabrous above supp!. 3: 129. 1813. TYPE: French Guiana. except occasionally along midvein, persistently Poiret s.n. (FI). sericeous below with lustrous coppery to golden Eugenia llewelynii Steyerm., Fieldiana, Bot. trichomes, apex acuminate to abruptly acumi­ 28: 1011. 1957. TYPE: Venezuela. Amazonas: nate, the acumen obtuse, base rounded to acute; Sanariapo, Puerto Ayacucho, 124 m, 15 Apr. petiole 5-10 mm long, 1-1.5 mm wide. Inflo­ 1942, Ll. Williams 14962 (Holotype: F!; Iso­ rescence of 1-3 racemes clustered together at types: MICHl, NY!, US!, VENl). leafy or leafless nodes, raceme axis 1-5 (-9) mm McVaugh (1969) distinguished E. chrysophyl­ long, densely ferruginous-tomentose with 1-4 lum from E. llewelynii as follows: pairs of flowers, bracts and bracteoles persistent, glabrous within, tomentose without, bracteoles 1. Leaves long acumiuate, oblong to lanceolate-ob­ ovate, ca. 1 mm long; pedicels, 2-7 mm long, long, 2.5-3 times as long as wide; pubescence tomentose. Flowers 4-merous, ferruginous to­ golden to red; French Guiana; Suriname .... mentulose without; calyx of 2 pairs of unequal ...... E. chrysophyllum I. Leaves obtuse or short-acuminate, broadly elliptic cupulate lobes, exterior lobes 1.5-2.2 mm long, to obovate, about twice as long as wide; pubes­ interior lobes 3-3.5 mm long, glabrous within; cence brownish-red; Orinoco lowlands ..... petals white, oblong, cupulate, 4-5 mm long, ...... E. Llewelyn;; glabrous; stamens aprox. 70-100, filaments white, 5-6 mm long, anthers oblong with a Since then, numerous collections have been gland at the apex; style 7-8 mm long, with made in the middle Orinoco River area that sparse long trichomes, stigma minute; ovary 2- show a considerable amount of variation of leaf locular, aprox. 13-15 ovules per locule. Fruits shape and size. Even the various isotypes of E. grayish to olive when dry, widely elliptic to ob- HOLST: MYRTACEAE 165

long, 1-1.5 cm long, 0.7-1.0 cm wide, I-seeded; moist or wet areas along gallery forests, and in embryo homogeneous. periodically inundated forest at low altitudes, Specimens examined: COLOMBIA. Vichada: 100-300 m. Cano Arepa, Rio Torno, 100 m, 05°26'N, A single collection from Merida State in Ve­ 68°39'W, 14 Mar. 1985, Zarucchi & Barbosa nezuela has longer racemes and slightly larger 3710 (MO); Meta: Pto. Gaitan, Rio Yucao, 250 leaves than normal, but otherwise matches well. m, 12 Mar. 1971, Pinto E. & Sastre 1178 It is mentioned here separately because of the (MICH); Vaupes: Rio Vaupes, between Mitu and abnormally high altitude where it was collected. Javarete, 14-24 May 1953, Schultes & Cabrera VENEZUELA. Merida: Pueblos del Sur, 1140 m, 19280-A (US). PERU. Loreto: Maynas Province, Jun. 1955, Bernardi 2237 (MER). Santa Maria de Nanay, Mishana, 90 m, 03°55'S, The "chrysophyllum species complex" con­ 73°35'W, 2 Oct. 1990 (fr), Pipoiy et al. 12734 tains the following: Eugenia baileyi Britton, E. (MO); Iquitos, 122 m, 03°48'S, 73°25'W, 19 Sep peregrina McVaugh, E. chrysophyllum DC., E. 1988 (past f1), Vasquez & Jaramillo 11050. VE­ cricomolensis StandI., E. kaiteurensis Amshoff, NEZUELA. Bolivar: Dist. Piar, Rio Acanan, 480 and E. heterochroma Diels. Eugenia myroba­ m, 05°55'N, 52°16'W, 2 May 1986, Holst, Lies­ lana DC. may also belong to this group. The ner & Steyermark 2775 (ASU, MO, PORT, chrysophyllum complex is perhaps related to an­ VEN); 470 m, 05°56'N, 62°17'W, 18 May 1986, other species complex that includes: E. ferrei­ Steyermark, Liesner & Holst 131908 (MO, raeana O. Berg, E. omissa McVaugh, and E. VEN); Dist. Cedeno, Rio Parguaza, 50 m, ramiflora Desv. 06°J 5'N, 67°07'W, 10 Sep. 1985, Steyermark, All of the species of the chrysophyllum com­ Holst & Manara 131693 (MO, VEN). Apure: plex are found in South America north of the Reserva Forestal San Camilo, Cerro Nulita, Rio Amazon, with one or possibly two species ex­ Nulita, 280 m, 3 Apr. 1968, Steyermark et al. tending into southern Central America. They are 101790 (NY, US, VEN); Cerro La Cristalina, shrubs or small trees that occur in moist mon­ Rio Nulita, 300 m, 5 Apr. 1968, Steyermark et tane or wet lowland forests. Trichomes colored, at. 101880 (NY, VEN); Dist. Pedro Camejo, S coppery to rusty-red; leaves permanently seri­ bank of Rio Cinaruco, 35 m, 06°32'N,~67°22'W, ceous below, glabrous or glabrescent above, 30 Apr. 1977, Davidse & Gonzalez 12420 (MO, midvein plane or prominently convex above. VEN); Rio Meta, 70 m, 06°13'N, 68°49'W, 16- secondary venation convex above and below; in­ 18 Feb. 1978, Davidse & Gonzalez 14326 (MO, florescence densely tomentose to sericeous, VEN); Galeras de Cinaruco, 50 m, 06°35'N, composed of I-several short-axis racemes per 67° 15'W, 23 Feb. 1979, Davidse & Gonzalez node, frequently appearing fasciculate; calyx of 15643 (MO, VEN); Laguna La Guacharaca, 70 2 unequal imbricate sepal pairs, completely en­ m, 06°42'N, 67°27'W, 24 Feb. 1979, Davidse & closing the glabrous petals in bud; ovary fre­ Gonzalez 15694 (MO, VEN); Dist. San Fernan­ quently ridged when young; hypanthium slightly do, Rio Meta, 60 m, 06°19'N, 67°50'W. 9-11 prolonged above the summit of the ovary; fruit Feb. 1978, Davidse & Gonzalez 13885 (MO. elliptic-oblong, grayish pubescent. VEN). Amazonas: Rio Cataniapo, mouth on Rio Eugenia coffeifolia DC., Prodr. 3: 272. 1828. Orinoco, 37 m, 06°25'N, 67°25'W, 23 Jul. 1981, Castillo 1284 (MO); 24 Jul. 1981, Castillo 1299 In Venezuela, previously known only from the (MO); 14 Feb. 1983, Castillo 1597 (MO); 15 eastern part in Bolivar and Delta Amacuro Aug. 1986, Castillo 1601 (MO); Castillo 2416 states, a significant range extension for the coun­ (MO). 2441 (MO); Rio Orinoco, Pto. Ayacucho, try is a collection from Carabobo state in north­ 100 m, 05°40'N, 67°40'W, 3 Apr. 1984, Gentry central Venezuela. & Stein 46263 (MO, VEN); 5 Apr. 1984, Gentry VENEZUELA. Carabobo: 18-20 km S of Pto. & Stein 46356 (ASU, MO, VEN); San Carlos Cabello, 350-700 m, 10023'N, 68°04'W, 15 Apr. de Rio Negro, 120 m, Olo55'N, 67°05'W, 4 Dec. 1982, Liesner 13722 (MO, VEN). 1977, Liesner 4165 (MO, VEN); Pto. Ayacucho, Eugenia cuaoensis McVaugh, Mem. New York 4 Apr. 1984, Plowman 13479 (F, NY); Rio Sia­ Bot. Gard. 18(2): 174. 1969. pa, Rio Casiquiare, 130 m, 4 Apr. 1953, Maguire & Wurdack 34832 (NY, US, VEN); Dept. Atu­ Previously only known from the type collec­ res, Sanariapo, 120 m, 15 Apr. 1942, Williams tion, Maguire & Politi 27336 (Holotype: 15962 (MICH, NY, US, VEN, type of E. llew­ MICH!; Isotypes: NY!, VEN!) which was in elynii); Cano Sanariapo, 120 m, 2 Jul. 1942, Wil­ flower, immature fruits have been found. They liams 15958 (NY, US, VEN). are strongly and vertically 14-16 ridged, 1-1.5 Eugenia chrysophyllum is also known from cm long, 1-1.3 cm wide, persistently densely Suriname. The collections from Venezuela, Co­ puberulent. The calyx-lobes are reflexed and lombia, and Peru were found to be occurring in persistent, but rather thin and fragile. The apex 166 SELBYANA Volume 23(2) 2002 of the fruit is narrow and then abruptly expands nearly their entire length, with only the tips of into the 4-4.5 mm broad staminal ring. A col­ the lobes free. lection from Miranda state in northern Venezue­ Specimens examined: BRAZIL. Amazonas: Km la is surely conspecific, differing only in that it 26 of the Manaus-Itacoatiara Highway, Reserva is persitently minutely pilose on the lower leaf Forestal Ducke, 14 Dec. 1966 (fl), Prance et al. surface and on the fruit. The relationship be­ 3631 (F, NY, US); Km 202 of the Manaus-Ita­ tween Eugenia cuaoensis and E. marowynensis coatiara Highway, near Rio Urubu, 19 Dec. of Suriname and French Guiana needs to be 1966 (bud, 6), Prance et al. 3711 (F, NY, US, more closely examined. VEN); Km 204 of the Manaus-Itacoatiara High­ VENEZUELA. Amazonas: Dept. Atures, Rio Ca­ way, 21 Dec. 1966 (bud, 6), Prance et al. 3772 taniapo, entre San Pedro de Cataniapo y comu­ (F, NY, US); Basin of Rio Negro, Rio Uneiuxi, nidad EI Milagro, 90-100 m, 06°25'N, 67°25'W, 200-300 km above mouth, 22 Oct. 1971 (imm. 12 Aug. 1986 (imm. fr), Castillo 2189 (MO, fr), Prance et al. 15529 (F, MICH, NY, US). VEN). Miranda: Cerrosdel Bachiller, between VENEZUELA. Amazonas: Dept. Rio Negro, Neb­ base and summit, above Quebrada Corozal, 20- lina Massif, Canon Grande, along the Rio Ma­ 700 m, 10009'N, 65°48'W, 20 Mar. 1978 (imm. warinuma between the mouth of the canyon and fr), Steyermark & Davidse 116509 (VEN). the first major fork of the river, ca. 7 km ENE of Puerto Chimo, 350 m, 00050-51'N, 66°02- Amshoff, Rec. Trav. Bot. Eugenia cupulata 06'W, 9-14 Jui. 1984 (fl), Davidse & Miller Need. 39: 160. 1942. 27176 (MO, SEL, VEN); 0-0.5 km northeast of New to French Guiana; previously known San Carlos de Rio Negro, 120 m, 01°55'N, from Suriname. 67°5'W, 17 Nov. 1977 (fr), Liesner 3577 (MO, FRENCH GUIANA. Proximite de Zidockville, VEN). Haut Oyapock, 7 Aug. 1980 (st), Prevost & Eugenia diplocampta is a member of the "fei­ Orenand 930 (MO). joi" species complex, a well-characterized group where most of the members have the ca­ Eugenia denigrata McVaugh, Mem. New York lyx lobes connate for part or nearly all of their Bot. Gard. 18(2): 176. 1969. length. This complex challenges the validity of Previously known from two collections from the genus Calycorectes, which is essentially dis­ the Brazilian side of the Rio Oiapoque in Ama­ tinguished from Eugenia by the hypanthium that pa, the following collections represent the first is prolonged above the ovary and closed in bud records for French Guiana. or nearly so. Eugenia typically has no prolon­ Specimens examined. FRENCH GUIANA. Crique gation and the calyx-lobes are free. Gabaet, Bassin de I' Oyapock, entre embouchure The species complex can be characterized as et Crique Merignan, 005 m, 03°55'N, 51°48'W, follows: Trees or shrubs occuring in the Amazon 12 Apr. 1988 (fl), Cremers 9883 (U, NY, US); Basin and the Guayana Shield. Trichomes fre­ St. Georges de l'Oyapock, Crique Gabaret, 19 quently dibrachiate. Leaves with the midvein Dec. 1978 (fr), Lescure 800 (MICH). raised in a characteristic narrow vertical line; lower secondary veins strongly ascending, the Eugenia diplocampta Diels, Verh. Bot. Vereins first and sometimes the second arising to form Provo Brandenbburg 48: 191. 1907. the marginal vein, not anastamosing with the The leaves on the holotype of E. diplocampta ones above. Inflorescence of short racemes, the (ute 6151, G) are apparently from a vigorously flowers sometimes appearing solitary, fascicu­ growing shoot, and the flowers are .bomeon a late or umbellate. Flowers glabrous or pubes­ thick, leafless branch. The calyx-lobes are free cent, calyx-lobes connate for at least part of their for nearly their entire length. The calyx-lobes length, sometimes nearly wholly connate; an­ and ovary are densely pubescent without with thers linear to lanceolate, frequently drying coppery appressed trichomes. The ovary shows grayish. Fruits known from few collections, el­ a slight amount of longitudinal ridging. liptic or oblong, occasionally ridged. More recent collections of E. diplocampta The "feijoi" complex includes at least the fol­ have the calyx-lobes glabrescent. The ovary is lowing species: E. chartacea McVaugh, E. cos­ variable in the degree of pronunciation of the tata O. Berg, E. cuspidifolia DC., E. diplocamp­ longitudinal ridges, collections from Venezuela ta Diels, E. fasciculifiora O. Berg, E. feijoi O. more prominently so, and those from around Berg, E. marlierioides Rusby, E. media Sagot, Manaus, Brazil less so, but still discernible. The E. pisonis O. Berg, E. pleurosiphonea Diels, E. degree to which the calyx-lobes are connate also percrenata McVaugh, E. schunkii McVaugh, E. seems to be variable, the single collection avail­ sigillata Amshoff, E. solimoensis O. Berg, E. able from Venezuela with flowers (Davidse & subterminalis DC., and E. yatuae (McVaugh) B. Miller 27176) has the calyx lobes connate for Holst. Based on herbarium material, there .exist HOLST: MYRTACEAE 167 another 3 or 4 other species that do not match riname. The single collection that McVaugh any of the above. The center of diversity is in (1969) used as the basis for the citation of E. the upper Amazon basin. ramiflora var. montana from Venezuela, Stey­ ermark 90325, is here cited below under E. fer­ Eugenia egensis DC., Prodr. 3: 281. 1828. reiraeana. New for Panama; previously with a broad dis­ The following collections are referred to Eu­ tribution in South America. genia ferreiraeana. The Guyana collection cited PANAMA. Veraguas: Isla Coiba, Distr. de Mon­ below is a new record for that country. The leaf tijo, Estaci6n Biol6gica, Norte de la Isla, ca. 0 shape varies from ovate to narrowly elliptic, but m, 07°37'N, 81°43'W, 17 Feb. 1995 (fr), Gal­ they share the same characters of having the dames et al. 2012 (SEL). leaves densely ftoccose-tomentose when very young, becoming glabrescent, the marginal vein Engenia eurycheila O. Berg; Linnaea 27: 213. nearly parallel and close to the margin, lateral 1856. veins numerous and scarcely evident. They are Previously known only from the interior of separated below for convenience into 2 forms Guyana, this has been. collected from northern based on the length versus width ratio of the leaf Brazil. See Berg (1856) for a description, and blades. Amshoff (1948) for citation of Guyanan collec­ 1. Typical broad-leaved form as per the type, tions. leaf blades 2-2.5 times longer than broad: VE­ Specimens examined: BRAZIL. Roraima: On NEZUELA. Amazonas: Dept. Atabapo, Salto Yu­ an azimuth of 49.5' from Boa Vista at a distance reba, Cafio Yureba, bajo Rio Ventuari, 120-150 of 50 kms (BR-401) Fazenda Quixabeira, 120 m, 04°03'N, 66°01'W, 24 Oct -4 Nov. 1981, m, 13 Oct. 1977, Coradin & Cordeiro 648 Delascio & Gudnchez 10665 (MO); Dept. Rio (NY); on an azimuth of 49°05' from Boa Vista Negro, Neblina. Massif, Rio Mawarinuma, 350 at a distance of 50 kms on BR 401, Fazenda m, 00050-51'N, 66°02-06'W; 9-14 Jul. 1984, Quixabeira, 120 m, 15 Oct. 1977 (ft), Coradin Davidse & Miller 27193 (MO, VEN); Rio Siapa, & Cordeiro 686 (NY); Cauame region, 8 km near base of Cerro Aracamuni, 250 m, 01°39'N, NW of Boa Vista, by Cauame river, 100 m, 16 65°40'W; 4 Nov. 1987, Liesner & Carnevali Oct. 1977, Coradin & Cordeiro 714 (NY); Boa 22754 (MO, VEN). GUYANA. U. Takutu-U. Es­ Vista, regHio de Rio Cauame, aftuente do Rio sequibo, Acarai Mts., Kashinar Mt., summit and Branco, 12 Feb. 1977 (imm. fr.), Rosa & Cor­ surrounding slopes, 825-975 m, 01°17'N, deiro 1471 (MO, NY); margens do Rio Cauame 58°39'W, 2 Mar. 1994 (ft), Henkel et al. 4911 nas proximidades das praias do Casari, 18 Feb. (SEL). 1977 (imm. fr), Rosa & Cordeiro 1554 (NY). 2. Slender-leaved riverine form, leaf blades 3.5-6(-7) times longer than broad: BRAZIL. Para: Eugenia ferreiraeana O. Berg in Mart., Fl. Basin of Rio Trombetas, jct. of Rio Tombetas Bras. 14(1): 285. 1857. TYPE: Brazil. In vi­ and Mapueira, 30 May 1974, Campbell et at. cinibus Barra, Provo Rio Negro, Dec.-Mar. P22319 (NY). Amazonas: Rio Urubu, between 1850-1851 (ft), Spruce 1160 (F!). Cachoeira Iracema and Manaus-Caracarai Road, ?Eugenia ramiflora var. montana Amshoff, 6 Jun. 1968, Prance et at. 5015 (F,. NY); Rio Bull. Torr. Bot. Club 75: 535. 1948. 'TYPE: Su­ Curuquete, vic. of Cachoeira Santo Antonio, 16 riname. Tafelberg (Table Mountain), rim of Ar­ JuI. 1971 (fr), Prance et at. 14290 (MO); Rio rowhead Basin, 725 m, 24 Aug 1944, B. Ma­ Maturaca, between the Missao Salesiana and guire 24481 (Isotype: NY!). Rio Cauaburi, 8 Jan. 1966, Silva & Braziio Eugenia ramiflora var. montana was distin­ 60788 (MO). VENEZUELA. Bolivar: Sierra Ichtin, guished from the typical variety on the basis of Rio Ichtin, 500 m, 04°46'N, 63°18'W, 28 Dec. having glabrous and narrower leaves. As Mc­ 1961, Steyermark 90325 (MICH, VEN). Ama­ Vaugh pointed out (1969), this variety is closer zonas: Dept. Rio Negro, Between Neblina Base to E. ferreiraeana since the venation and pubes­ Camp and Pto. Chimo, Rio Mawarinuma, 130- cence, or lack of it, are like that described for 200 m, 00050'N, 66°06-lO'W, 6-7 JuI. 1984, E. ferreiraeana below. The fruits· of var. mon­ Davidse & Miller 27065 (MO; VEN); Cerro tana are also permanently pubescent, a possible Neblina, Rio Mawarinuma, vicinity of Pto. Chi­ character of E. ferreiraeana that needs to be mo, 190 m,. 00050'N, 66°05'W,. 24 Apr. 1984, confirmed. Many more recent collections from Gentry & Stein 46908 (MO, VEN); Pto. Chimo the base of Cerro Neblina in southern Amazonas Camp, Rio Mawarinuma, 150 m, 00050'N, state in Venezuela and northern Brazil cannot be 66°07'W, 13 Feb 1984, Liesner 15878 (MO, distinguished from var. montana, some having VEN); Foothills of Cerro Neblina, above Pto. the leaves even more narrowed and the apices Chimo Camp, Rio Mawarinuma, 150-500 m, longer acuminate than the collections from Su- 00050'N, 66°07'W, 14 Feb. 1984, Liesner 15933- 168 SELBYANA Volume 23(2) 2002

A (MO, VEN); Along Rio Mawarinuma, at Pto. celos, 14 Jan. 1978, Steward et al. 328 (F, MO, Chimo Camp, 170 m, 00050'N, 66°07'W, 23 Apr. US); Rio Negro, Manaus, May 1910, Ule 8930 1984, Thomas 3218 (MO, NY, VEN). (US). FRENCH GUIANA. Riviere Petite Ouaqui, See McVaugh (1969) for additonal specimen Saut Baille-Nom. Rive droite, 16 Jul. 1973 (fl), citations. See also under E. ramifiora for addi­ Granville B4980 (NY). PERU. Loreto: Provincia tional comments on this and related species. Maynas, Pto. Almendras, Rio Nanay, 130 m, 29 Apr. 1978, D(az & Jaramillo 279 (MO); Caserio Eugenia gomesiana 0. Berg in Mart., Fl. Bras. Sungaro Cocha, 16 Feb. 1976, Revilla 173 14(1): 254. 1857. TYPE: Brazil. Amazonas: (MO); Distrito Iquitos, Rio Nanay, 22 Mar. in vicinibus Barra, Provo Rio Negro, Apr. 1976, Revilla 343 (MO); Rio Nanay above lqui­ 1851 (past fl), Spruce 1163 (lsotype: NY!). tos, 120 m, 03°40'N, 73°20'W, 15 May 1980, Eugenia prosoneura 0. Berg, Linnaea 31: Vasquez & Jaramillo 197 (MO); Maynas, Pto. 255. 1862. TYPE: Brazil: Amazonas: prope San Almendras, 03°48'N, 73°25'W, 122 m, 28 Jul. Carlos ad flumen Rio Negro Brasiliae borealis, 1983 (fr), Vasquez & Jaramillo 4260 (MO). VE­ Spruce 3112 (Isotype: NY!). NEZUELA. Amazonas: Dept. Rio Negro, Cafio Cu­ In the Flora of Peru, McVaugh (1958a) distin­ weje, 32 km S of San Carlos (119 m, 01°56'N, guished Eugenia prosoneura by the whitish tri­ 67°03'W), 14 May 1980, Clark 7570 (MO, NY, chomes of the inflorescence as compared with VEN); Brazo Casiquiare between Culimacare reddish or purple in E. gomesiana. Numerous and its junction with the Rio Negro, 75 m, collections from the along the Rio Negro in the 01°58'-02 OO'N, 66°50'-67 07'W, 26 Jul. 1984, Manaus region, the type locality of E. gomesi­ Davidse 27984 (MO, VEN); Prope San Carlos, ana, show the trichomes to be whitish or grayish Rio Negro, Spruce 3112 (type of E. prosoneura; en masse, but they all have distinct reddish or isotype at NY!); San Carlos de Rio Negro, red-orange bases. A collection from Rio Curu­ Olo50'N, 66°55'W, 25 Nov.-9 Dec. 1984, Ster­ quete, in southern Amazonas, Brazil (Prance et gios & Aymard 7562 (MO, PORT); Rio Pasiba, al. 14478) has the trichomes of the inflorescence 02°25'N, 66°23'W, Stergios & Aymard 7671 darkened as McVaugh describes for E. gomesi­ (MO, PORT); Cafio Chimoni, 200 m, 02°02'N, ana. Microscopic observation, however, shows 66°24'W, 17-20 Sep. 1986, Stergios et al. 9418 the trichomes to be covered with hypha! growth (MO, PORT); Rio Casiquiare, along river near giving the overall appearance of dark purple tri­ Solano, 100-130 m, 9 Jul. 1959, Wurdack & chomes. The red-orange bases typical of the Adderley 43364 (F, MICH, MO, NY, VEN). "white" -haired collections can still be seen A population from the base of Serra Araca in though the hyphal covering. Amazonas, Brazil, has larger leaves and fruits Eugenia gomesiana is uniform in leaf size, than the above cited collections, but is otherwise pedicel length, flower size and amount of pu­ similar in features. bescence throughout its range in the Amazon ba­ BRAZIL. Amazonas: 0.3 km ao S da parte cen­ sin. All of the collections studied indicate that it tral de Serra Araca, 00049'N, 63° 19'W, solo ar­ is a riverine shrub or tree. See McVaugh (l958a) enoso, mata de igap6, 29 Feb. 1984 (fr), Miralha for a description under both E. gomesiana and 53 (NY); Serra Araca, southern massif, west fac­ E. prosoneura. A collection from the Bolivian ing talus slope, black water seasonally inundated Amazon represents the first collection for that forest on sandy soil, 0048'N, 63°18'W, 27 Feb. country and extends the range southward, and a 1984 (fr), Pipoly et al. 6714 (NY); base of Serra collection from French Guiana extends the range Araca, 0-3 km south of Central Massif, 0 49'N, eastward. 63 19'W, tall forest on white sand, 8 Feb. 1984, Specimens examined: BOLIVIA. La Paz: Provo Prance et al. 28897 (NY). Iturralde, Siete Cielos, Rio Manupare, 180 m, 12°27'S, 67°37'W, 3 Jun. 1987, Solomon 16883 Eugenia griseiflora McVaugh, Mem. New York (LPB, MO, SEL). BRAZIL. Amazonas: Rio Uau­ Bot. Gard. 18(2): 187. 1969. pex, acima de Santa Rosa, 14 May 1973, Silva et al. 38623 (MO); Rio Negro, NW of Sao Ga­ Most previous collections of this species are briel, near mouth of Rio Uaupes, 00005-08'S, from Suriname; following are two that match 67°lO'W, 21 JuI. 1979, Poole 2090 (US); Rio quite well from Peru and Brazil. Negro, opposite Manaus, 8 Apr. 1971, Prance et PERU. Cuzco: Provo de La Convencion, Rio al. 11754 (NY, US); Rio Curuquete, Sao Paulo, Apurimac, 10 km NW of Pto. San Francisco on 30 Ian above mouth of Rio Coti, 20 Jul. 1971, road to Pichari, 600 m, 24 Sep. 1976 (ft), Was­ Prance et al. 14478 (NY, US); Rio Negro, shausen & Encarnaci6n 706 (NY). BRAZIL: mouth of Rio Caures and Barceios, 12 Oct. Mun. Oriximana, Para, BR 163 a 7 km ao N de 1971, Prance et al. 15113 (NY, US); Manaus, Cacho Porteira, vicinal ES-7. 1002'N, 57°02'W, 1882, Schwecke 25439 (US); Rio Negro, Bar- 20 Aug. 1986, Cid Ferreira et (II. 7928 (NY). HOLST: MYRTACEAE 169

Eugenia latifolia Aublet, Hist. PI. Guiane 502. vegetative growth and ovary, calyx lobes that pI. 199. 1775. are glabrous and erect in bud, the linear brac­ teoles, and the nearly featureless leaves. The first New species to Guyana and Venezuela; pre­ collections known from French Guiana also ex­ viously known from Suriname and French Gui­ tend the distribution further to the North-East, ana. and a new collection can be reported from Ec­ GUYANA. Potaro-Siparuni region, Pakaraima uador also. Most specimens examined have Mountains, Mt. Wokomung, Wusupubaru Creek, smaller calyx lobes than the type, but they are 2 km from juncture with Suruwabaru Creek, variable in size throughout the range. See 975-1125 m, 05°03'N, 59°53'W, 13 Feb. 1993, McVaugh (1969) for synonomy and additional Henkel et al. 1342 (SEL). VENEZUELA. Sucre: collections. Peninsula de Paria, Cumbre de las Estrellas, Specimens examined. FRENCH GUIANA. Crique west of Manacal, 800-830 m, 10048'N, Solitaire, Montagne de Kaw, 100 m, 14 Dec. 62°40'W, 30 Nov. 1979 (fl), Steyermark & Lies­ 1986, de Granville 9089 (MO, U); Saiil, La Fu­ ner 120803 (MO, VEN); Peninsula de Paria, vic. mee Mountain Trail, 250 m, 03°37'N, 53°12'W, of Manacal, above Rio Grande Arriba, 800 m, 9 June 1988 (fr), Mori & Gracie 18839 (MO); 25 Feb. 1980 (fr), Steyermark et al. 121785 Fleuve Maroni, Montagnes Fran<;aises (Gaa (MO, VEN). Caba), 04°33'N, 54°27'W, 27 Aug. 1986 (imm. Eugenia Iigustrina (Sw.) Willd., Sp. PI. 2: 962. fr), Sastre et al. 8092 (U); Maroni River, above 1800. mouth of Marouini River, 140 m, 03°19'N, 54°06'W; 20 Aug. 1987, Weitzman 238 (US). New species to Bolivia and Guyana; scattered ECUADOR. Napo: Augrico Canton, Reserva elsewhere in the West Indies and northern South Faunistica Cuyabeno, Laguna Zancudo Cocha, America. 230 m, 75°32'W; 00033'S, 28 Sep. 1991 (fl), Pa­ GUYANA. U. Takutu-U. Essequibo Region: lacios et al. 7756 (MO). VENEZUELA. Aragua: Rupununi area, Surama village, 65-90 m, Parque Nacional, 1000 m, June 1938 (bud); Val­ 04°08'N, 59°04'W, terra firma forest, 19 Feb. le de Ocumare, 800 m, 2 Apr. 1937, Pittier 1990 (fr), Acevedo et al. 3301 (MO). BOLIVIA. 13956 (F, NY, US); Km 29, Parque Naciona1 La Paz: Inquisivi Province, "Lakachaka," trail Henry Pittier, 900 m, 11 Jan. 1939, Williams between the mouth of Rio Aguilani and the for­ 11075 (F, US). taleza of Choquecamiri, 1650-1800 m, 16°40'S, 67°20'W, 23 Sep. 1991 (fl), Lewis 40424 (MO); Eugenia marlierioides Rusby, Bull. New York Inquisivi Province, Loma de Playa Verde, 30 km Bot. Gard. 8: 108. 1912. TYPE: Bolivia. San north of Choquetanga, 1500-1600 m, 16°35'S, Juan, 3200 fr, 22 Mar. 1902 (bud), R.S. Wil­ 67°19'W, 14 Nov. 1991 (fl), Lewis 40530 (MO); liams 218 (Holotype: NY!; Isotype: US!). Eugenia luciae Amshoff, Rec. Trav. Bot. Neerl. Originally described from Bolivia, this was 42: 14. 1950. collected twice in Amazonian Venezuela by Ronald Liesner. The Venezuelan material has the New species to Guyana; previously known leaves much larger, 12-19 cm long, vs. 5-6.5 from Suriname. than in the type collection though the second GUYANA. U. Takutu-D. Essequibo Region: known collection from Bolivia has leaves that Rupununi area, new road from Lethem to 25 km are equivalent in size to the Venezuelan collec­ past Surama village entrance, 90-110 m, tions. The two populations share several fairly 04°15'N, 58°56'W, 28 Feb. 1990 (fl), Acevedo unique characters that unite them well: abbre­ 3435 (MO). viated, few-flowered raceme; pedicels relatively Eugenia macrocalyx (Rusby) McVaugh, Fiel­ long and slender; calyx-lobes connate, thin, diana, Bot. 29: 212. 1956. TYPE: Bolivia. forming a globe much larger than the ovary; Vic. Pampas near Lake Rogagua, 3000 fr., thickened leaf blade margin; double marginal 12 Sep. 1921 (past fl), Rushy 666 (Isotype: vein with the interior strongly looping; midvein NY!). convex. Fruits are not known from the Venezue­ lan populations, but Daly et al. 6595 from Bo­ Previously known from Bolivia, Peru, Brazil livia describes the immature fruits as 5 cm di­ (Amazonas, Para), Guyana, and Suriname, Eu­ ameter and depressed-globose. genia macrocalyx was collected 50 years ago in Specimens examined. BOLIVIA. Beni: Provo Venezuela by Pittier and Williams but never ful­ Ballivian, Rio Beni, above confluence with Rio ly identified. Though the two Venezuelan spec­ Quiquibey, 3.5 hr upstream from Rurrenabaque, imens cited below are not in full flower, they 14°44'S, 67°25'W, 320 m, 23 May 1990 (bud, can be determined positively by the light cop­ fr), Daly et al. 6595 (SEL). VENEZUELA. Terri­ pery to white sericeous indument of the young torio Federal Amazonas: Dept. Rio Negro, 0- 170 SELBYANA Volume 23(2) 2002

0.5 km NW of San Carlos de RIo Negro, forest 1900 m, 03°22'N, 61 0 20'W, Hopkins ef al. 659 and open areas 120 m, Olo55'N, 67°05'W, 28 (MO). FRENCH GUIANA. Massif des Emerillons, Nov. 1977, Liesner 3979 (MO); 0-0.3 km NW zone nord "savane roche" au sommet d'une col­ of San Carlos de Rio Negro, 120 m, 01°56'N, line sur la gauche de la Haute Approuague, 350 67°03'W, 8 Apr. 1979, Liesner 6380 (MO, m, 20 Sep. 1980 (fr) , de Granville 3925 VEN). (MICH); 23 Sep. 1980 (fr) , 3950 (MICH); (ft) 3951 (MICH); Monts Bakra, versant sud Boret Eugenia mimus McVaugh, Mem. New York basse et broussailles sur affteurements grani­ Bot. Gard. 18(2): 193. 1969. tiques a 4 km I'Ouest du PicCoudreqau, 400 m, New species for French Guiana; previously 30 Sep. 1980, de Granville 4042 (MICH); Mont known from Guyana and Brazil. Atachi Bacca, Region de l'Inini, 420 m, FRENCH GUIANA. Eaux Claires, Sentier Bota­ 03°33'N, 53°55'W, 10 Feb. 1989 (fr), de Gran­ nique, 3-5 km from entrance, 320 m, 24 May ville et al. 10555 (MO, NY). PERU. Loreto: Prov­ 1992 (fr), Acevedo-Rodriguez et al. 5011 (MO); incia Maynas, Dist. Iquitos, Rio Nanay above Lely Mts., SW plateaus covered by ferrobauxite, Iquitos, 27 Oct. 1976, Revilla 1692 (MO). VE­ along road S of camp 4 on plateau, 550-710 m, NEZUELA. Piedra de Culimacare, conftuence of 24 Sep. 1975 (ft), Lindeman et al. 115 (NY); 312 Brazo Casiquiare and Rio Pasimoni, 125 m, Apr. (MO); SaUl, Monts La Fumee, 200-400 m, 1981, Delascio & Christenson 9590 (MO, 03°37'N, 53°12'W, 9 Oct. 1982 Cimm fr), Mori VEN); Dept. Atabapo, Cucurital de Caname, et al. 15074 (MO). middle Cafio Caname, 100 m, 03°40'N, 67°22'W, 30 Apr.-l May 1979, Davidse et al. Eugenia multirimosa McVaugh, Fieldiana, Bot. 16987 (MO, VEN). 29: 213. 1956. Eugenia punicifolia (H.B.K.) DC., DC. Pro dr. New species for Brazil and Pacific Colombia. 3: 267. I 828.-Myrtus punicifolia H.B.K., The Colombian specimen lacks the multi-rimose Nov. Gen. Sp. (quarto ed.) 6: 149. 1823. petioles of the Amazonian collections. BRAZIL. Acre: Mun. Cruzeiro do SuI, Reserva The most commonly collected species of Estrativista do Alto Rio Jurmi, Serringal Sao Myrtaceae in Venezuela and known also from Joao, 09°12'S, 72°41'W, 14 Mar. 1992 Cfr), Daly the West Indies, the three Guianas, Brazil, Peru, et al. 7429 (MO). COLOMBIA. Choc6: Parque Na­ Bolivia, and Africa can now be reported from cional de Vtria, Isla Playa Blanca, nor-oeste de Colombia. It will also likely be found in Arauca, la Ensefiada de Utria, 0-100 m, 06°20'N, Vichada and Guainia since it is very common in 77°20'W, 29 May 1990 (ft), Garda C. & Agu­ the areas opposite the border in Venezuela. alimpia 294 (MO). COLOMBIA: Magdalena Valley, Rincon Hondo, 6 Aug. 1924 (ft), Allen 307 (MO). Eugenia omissa McVaugh, Mem. New York Bot. Gard. 18(2): 197. 1969. TYPE: Surina­ Eugenia quadriovulata Amshoff, Rec. Trav. me. Lanjouw & Lindeman 2134 (Isotype: Bot. Need. 39: 160. 1942. NY!). New for French Guiana; previously known Originally described from collections from from Guyana and Suriname. Suriname and Brazil, the following collections FRENCH GUIANA. Riviere Tampoc, 18-24 Sep. from Venezuela, Brazil, French Guiana, Perti, 1961 (ft), BAFOG 7898 (U). and Bolivia amplify the range considerably. ft is Eugenia ramiftora Desv. in Hamilton, Pro dr. FI. closely related to E. ramijlora and perhaps not Ind. Occid. 43. 1825. distinct. It differs only slightly in that the mid­ vein is finely wrinkled or elevated into a narrow This species belongs to a complex that in­ median line, and the orangish tomentum of the cludes at least two other species: Eugenia om­ lower leaf surface is permanent and completely issa McVaugh, and E. ferreiraeana O. Berg. obscures the actual surface of the leaf below. Both Amshoff (1948) and McVaugh (1969) See under E. ramijlora and E. ferreiraeana for point out that there are strong similarities be­ additional comments. tween them. McVaugh (1969), in describing E. BOLIVIA. La Paz: Privince of Larecaja, Tuiri, omissa gives little basis for separating it from E. Rfo Mapiri, 490-750 m, 12-30 Sep. 1939, Kru­ ramijlora, using mostly the character of the mid­ kofl10922 (MO). BRAZIL. Amazonas: Between vein to differentiate it from the other closely re­ Rio Solimoes and Rio ls;a, 19 Aug. 1973, Lleras lated species. Following is the key that he used: et al. P17416 (MO, NY). Acre: Municipio Sen­ 1. Midvein fiat or concave above, often finely wrin­ ador Guiomard, estrada para Placido de Castro, kled or shrunken into a narrow elevated median 17 Oct. 1980, Cid & Nelson 2931 (MO, NY). line; lower leaf surface and hypanthium fiocco­ Roraima: Ilha de Maraca, Mun. Alto Alegre, se-tomentose, the bracteoles and calyx-lobes gla- HOLST: MYRTACEAE 171

brous or pubescent at base only; Suriname to It is premature to reduce Eugenia omissa to Para ...... E. omissa the synonomy of E. ramijlora, but it is clear that 1. Midvein in dried leaves definitely convex above, the two are closely related. Until these species smooth or somewhat wrinkled; bracts and brac­ can be studied in the field, there will remain teoles, pedicels and calyx lobes at least at base, cottony-tomentose ...... 2 doubts about the proper relationships of the spe­ 2(1). Leaves relatively broad, 2-2.5 times as long as cies in this complex. See under E. ferreiraeana wide; veins inconspicuous beneath, scarcely el­ and E. omissa for additional comments. evated, the marginal vein 1-2(-3) mm from the Specimens examined: SURINAME. Brokopondo margin; interior of Para to the upper Orinoco District, north of village Brokopondo, 14 Feb. lowlands ...... E. ferreiraeana 1966 (fr), van Donselaar 3112 (U). FRENCH Gm­ 2. Leaves relatively long, (2-)2.5-3 times as long as ANA. Riviere, grand Inini en aval et en amont de wide; veins conspicuous and elevated beneath, the Degrad Fourmi, 13 Sep. 1985 (fr), de Granville marginal vein 2.5-5 mm from the margin; Guian- et al. 8205 (MO); Camp N° 1 Ouman fou Langa as to eastern Para ...... E. ramiflora Soula-Bassin du Haut-Marouini 2 km en aval, rive gauche, 02°53'N, 54°00'W, 150 m, 14 Aug. Upon seeing specimens determined by Mc­ 1987 (fl), de Granville et al. 9267 (NY, U, US); Vaugh as Eugenia omissa and E. ramiflora, I Haut-Marouini basin, 150 m, 02°53'N, 54°00'W, find it difficult to separate the two based on the 23 Aug. 1987 (fl), de Granville et al. 9603 (US); above characters. Eugenia omissa tends to have Montagne Alikene, 29 Aug. 1973 (st), Oldeman the lower leaf surface obscured by the perma­ T.914 (US); Bas Oyapock, 3 Jun. 1970 (fr), Old­ nent dense orangish tomentum, whereas E. ram­ eman B-3345 (U). BRAZIL. Amazonas: Dist. iflora has only the youngest leaves arachnoid­ Agropecuiirio, Reserva 1501 ("km 41 ") of tomentose, later becoming sparsely appressed­ the WWFIINPA MCS project, 50-125 m, arachnoid allowing the lower surface of the leaf 02°24'26"-02°25'31"S, 59°43'40"-59°45'50"W, to be visible. The type of E. juivipes Sagot, 2 Dec. 1988 (fl), Boom et al. 8701 (MO). placed in the synonomy of E. ramiflora by both Eugenia subterminalis DC., Prodr. 3: 263. Amshoff and McVaugh has nearly glabrous ca­ 1828. lyx-lobes and bracteoles and can be distin­ guished from E. omissa only by the rnidvein be­ New for Colombia; previously known from ing convex vs. elevated into a narrow median Amazonian Peru and Brazil. line and the sparse indument on the lower leaf COLOMBIA: Amazonas: Municipio de Leticia; surface. Parque Nacional Natural Amacayacu; Quebrada The flowers in the various species in this com­ Bacaba, 100 m, 03°44'S, 700 15'W, 1 Mar. 1991 plex vary in size and the amount of pubescence, (ft), Rudas et al. 1441 (SEL). and length and thickness of the pedicels, but are Eugenia tafelbergica Amshoff, Bull. Torrey Bot. not at all consistent within a "species." Vena­ Club 75: 535. 1948. tion is perhaps the best character for separating Eugenia omissa-ramiflora from E. ferreiraeana. New for Guyana; previously known from Su­ Eugenia omissa and E. ramiflora have 9-13 riname. conspicuous primary lateral veins, the lower­ GUYANA. Potaro-Siparuni Region, 4-5 km most of which is strongly ascending and arching north of Surama village along trail to confluence between the other lateral veins to form the dis­ of Burro-Burro & Surama Rivers, 75 m, tinct marginal vein well within the leaf margin, 04°IO'N, 59°03'W, 1 May 1992 (fl), Hoffman et appearing almost triplinerved. Eugenia ferrei­ al. 1525 (MO). raeana has numerous, close-together and incon­ Eugenia tapacumensisO. Berg, Linnaea 27: spicuous lateral veins, and the marginal vein 222. 1856. runs close to the margins and barely arches be­ This was mentioned by McVaugh (1969) as tween the laterals. Eugenia omissa and E. ram­ being known from Venezuela, but he cited no ijlora have leaves generally larger, 10-20 cm, as specimens. well as more abundant indument on the lower Some good representative collections are: VE­ leaf surface, while those in E. ferreiraeana rare­ NEZUELA. Bolivar: 20 km east of La Paragua, 200 ly reach 10 cm and are glabrescent. m, 06°56'N, 63°15'W, 23 Jul. 1978, Liesner & Another character that needs to be looked into Gonzalez 5406 (MO, MICH, VEN). Apure: Dist. more closely is the fruit. From the few collec­ Pedro Camejo, 11 km directly east of Paso de tions at hand that are in fruit, it seems that those San Pablo, Rio Capanaparo, 45 m, 07°02'N, of E. omissa-ramiflora haveglabrescent fruits 67°39'W, 8-9 May 1977, Davidse & Gonzalez that turn red and then black at maturity while 12891 (VEN). those of E. ferreiraeana have the fruitsperma­ nently gray or brown floccose. Eugenia trinervia Vahl, Eclog 2: 36. 1798. 172 SELBYANA Volume 23(2) 2002

A distinct species well known from the Lesser Material examined: BRAZIL. Amazonas: Rio Antilles and Trinidad has now been collected Maturaca, between Missao Salesiana and the Rio several times in forests in Venezuela and Brazil. Cauaburi, 8 Jan. 1966, Silva & Brazao 60796 The material from Venezuela differs from the (MO, NY). VENEZUELA. Territorio Federal Ama­ Antillean collections in that the leaves are acute zonas: Dept. Rio Negro, 4.5 km NE of San Carlos at the base compared to rounded or slightly sub­ de Rio Negro, 120 m, 01056'N, 67°03'W, 5 Apr. cordate, and slightly smaller and thinner. 1979, Liesner 6182 (MO, VEN); Neblina Massif, These collections from the mainland of South Canon Grande, along the Rio Mawarinuma, 200 America revive the possibility that the type, as m, 00050-51'N, 66°02-06'W, 9-14 Jul. 1984, stated by Vahl, actually does come from French Davidse & Miller 27155 (MO, INPA, NY, VEN). Guiana. McVaugh postulated (1969) that since no collections of E. trinervia were known from Marlierea Cambess. the mainland, it was likely that the type came from the Antilles. See Mc Vaugh (1969) also for Marlierea is a poorly defined genus that is a list of synonyms. Eugenia perplexans R.O. becoming increasingly difficult to maintain. It is Williams, with type from Trinidad & Tobago distinguished entirely by a single character; the (R.O. Williams 12200; Isotype: NY!), may be a method of opening of the calyx (closed in bud synonym of E. trinervia. or with small lobes and tearing longitudinally Material examined: BRAZIL. Amazonas: Pico into the hypanthium). As more collections are Rondon, 0-3 km north of km 211 of Perimetral studied, however, that is proving to be a variable N Highway, 01°32'N, 62°48'W, 25 Mar. 1984, character, at times, even on the same individual. Pipoly, Samuels & de Oliveira 6913 (NY). At one extreme it is like Calyptranthes. In GUYANA. U. Takutu-U. Essequibo. Kassikai­ these cases, the inflorescence is of paired pani­ tyu R., 0-1 km north of landing at terminus of cles and the calyx is closed in bud and roughly trail from Kuyuwini R., 240 m, 01 °50'N, calyptrate (the edges of the hypanthium with one 59°05'W, 18 May 1997 (fl), Clarke 4663 (SEL). to three smaller lobes, but the major lobe con­ VENEZUELA. Miranda: Dist. Paez, Fila La Tigra, taining the apicula intact). Some examples of Quebrada San Juan, 18 km SW of Cupira, 600 this are Marlierea schomburgkiana O. Berg, M. m, lOo04-05'N, 65°45-47'W, 2-7 Sep. 1977, cana McVaugh, M. ligustrina McVaugh, and M. Gonzalez & F. Ortega 1326 (MO, VEN); Gua­ suborbicularis McVaugh. At the other extreme, topo, 600 m, Jun. 1958, Aristeguieta 3181 (MY, the calyx is open, if only slightly so, and tears NY, US, VEN); Bolivar: 0-6 km SE of El Pauji, longitudinally usually into 4 lobes [Marlierea 800-900 m, 04°30'N, 61°35'W, 9 Nov. 1985, umbraticola (H.B.K.) O. Berg]. Liesner 19753 (MO, VEN, U); Chimanta Mas­ See McVaugh (l958b, 1968, 1969) for more sif, Agparaman-tepui, near Rio Tirica, 1365 m, discussion on the delimitation of the Myrciinae 5 Mar. 1955, Steyermark & Wurdack 1252 (NY, genera. See also in this paper under Marlierea VEN). mcvaughii B. Holst for additional comments. Eugenia tumulescens McVaugh, Fieldiana, Bot. Marlierea caudata McVaugh, Fieldiana, Bot. 29: 22l. 1956. TYPE: Brazil. Amazonas: 29: 176. 1956. Porto Curucuhy, Rio Negro, 6 Oct. 1945 Originally described from Peru, more recent (fl), Froes 21106 (Isotype: NY!). collections show that this species has a fairly A species previously known from only two broad distribution in the Amazon Basin. See collections on the lower Rio Negro in Brazil has McVaugh (1958a) for description. now been collected just over the border from This species is closely related to Marlierea Brazil in Venezuela. It is still only known from schomburgkiana O. Berg with whose distribu­ the Rio Negro basin. tion it meets in the upper Rio Negro basin. They The original collections describe the habit as are easily told apart in flower by the long cop­ a mound forming shrub to 70 cm, but in the pery or reddish-yellow trichomes on the branch­ upper Rio Negro it grows to a slender tree 4-8 es of the inflorescense and densely so on the m high and to 5 cm diameter. The fruit at ma­ flowers, some of which are persistent into fruit. turity is yellow to orange, elliptic-oblong, l.5-2 In M. schomburgkiana such trichomes are pre­ cm long, 1-1.5 cm wide (when dry), glabrous sent only on bracts and bracteoles, the flowers or with some of the light brown tomentum per­ themselves are glabrous except for a few tri­ sisting, conspicuously 8-ridged or winged and chomes on the apiculum. Marlierea schomburg­ crowned by the persistent, erect, 5-10 mm long kiana also has larger fruits and the leaves obI an­ calyx lobes, borne on stout pedicels topped by ceolate-obovate as compared with elliptic in M. the persistent bracteoles. See McVaugh (1958a) caudata. Both species have a prolonged caudate for a description of the rest of the plant. leaf apex. HOLST: MYRTACEAE 173

Specimens examined: BRAZIL. Acre: Serra da unidentified collection made by Llewelyn Wil­ Moa, 29 Apr. 1971, Maas et al. P12652, (F, NY, liams. US). Amazonas: Vicinity of Rio Uatuma, Estra­ Specimens examined: COLOMBIA. Amazonas­ da Manaus-Caracarai (BR-174) estrada Balbina Vaupes: Rio Apaporis, Soratama, 250 m, 21 Jun km 69, no date, Cid et al. 062 (F, US); Upper 1951, Schultes & Cabrera 12766 (Holotype: Rio Solimoes, Mun. Sao Paulo de Olivens;a, Es­ US; Isotype: MICH); 24 Aug 1951 (fr); Schultes trada Bomfin, 03°30'S. 68°57'W, 23 Nov. 1986 & Cabrera 13722 (US). PERU. Loreto: Maynas (imm fr), Daly, Lima & Pinho 4397 (MO); Ma­ Province, Dist. Fernando Lores, caserio Serafin, naus-Porto Velho Highway, crossing with BR- 125-130 m, 04°08'S, n055'W, 9 May 1991 (fr), 319, 15 Mar. 1974, Prance et al. 20553 (MO, Grandez et al. 2569 (MO). VENEZUELA. Ama­ NY, VEN, US). Mato Grosso: Mun. Sinop, 7 km zonas: Dept. Casiquiare, Maroa, Rio Guainia, 12 east of BR-163, north of Rio Celeste, 51 km S Feb 1942 (imm fr), Williams 14302 (F, US). of Sinop, 12°18'N, 5S032'W, 18 Sep. 1985, Marlierea schombnrgkiana 0. Berg, Linnaea Thomas, Guedes & Lima 3864 (MO, NY). Para: Cuiaba-Santarem Highway, BR-163, km 1234, 29: 209. 1858 18 Nov. 1977, Prance et al. 25536 (MO, NY). New for Ecuador; previously known from Su­ PERU. Loreto: Provo Maynas, Iquitos, Rio Nanay, riname, Guyana and Venezuela. 122 m, 03°40'N, 73°25'W, 7 Mar. 1987 (fr), Vas­ ECUADOR. Santiago-Zamora: Taisha, 1500 m, quez et al. 8952 (MO). VENEZUELA. Amazonas: 26 Jan. 1962 (fr) , Cazalet & Pennington 7525 Dept. Atabapo, Cucurital de Caname, middle (NY). Cano Caname, 100 m, 03°40'N, 67°22'W, 30 Collections from Peru approximate the spe­ Apr.-l May 1979, Davidse, Huber & Tillett cies, but have shorter petioles, more marked ve­ 17009 (MO, VEN); Dept. Guainia, Maroa, Rio nation, and the leaf shape tends to be elliptic Guainia, 127 m, 18 Feb. 1942, Williams 14410 instead of obovate in the more northern popu­ (cf., sterile, VEN). lations. PERU. Loreto: Alto Amazonas, Rio Pas­ taza, 200 m, 04°20'S, 76°35'W, 25 Jan. 1979 Marlierea cuprea Amshoff, Bull. Torrey Bot. (fr), Diaz & Ruiz 934 (MO, NY); Maynas Prov­ Club 75: 530. 1948. ince, Iquitos, Pto. Almendras (Rio Nanay), 122 This is very much like Myrcia decorticans m, 03°48'N, 73°21'W, 24 May 1990 (fr), Vas­ DC. in the leaf texture and rimose petioles, but quez et at. 13749 (MO). differs in that the bracts are deciduous, the flow­ Marlierea scytophylla Die1s, Verh. Bot. Vereins ers are coppery-sericeous, and the fruits are Provo Brandenburg 48: 187. 1907. TYPE: smaller. See under Myrcia albidotomentosa for Brazil. Amazonas: Sao Joaquim, Rio Negro, a Jist of what I consider to be related species. Ute 6044 (G). The collections cited below from Venezuela represent the first for that country. The collec­ This was reported by McVaugh (1958a) as oc­ tion from Rio Emoni is very much out of alti­ curing in Venezuela, but he did not cite any col­ tudinal and geographical range but I cannot sep­ lections. The fmiting specimen cited below cer­ arate it from material from the type locality. tainly belongs to this species; it matches the de­ Specimens examined. GUYANA. Kamarang, scription and a type photo in almost all details. 505-545 m, 05°50'N, 600 40'W, 25 Jun. 1987 It differs though, in that the branches of the in­ (st), Boom et al. 8253 (MO); Karowtipu Moun­ florescence are strigose, the fruits have some tain, 920-1080 m, 05°45'N, 60035'W, 24 Apr. persistent sericeous trichomes, and the inflores­ 1987 (bud), Boom & Gopaul 7688 (NY); Mak­ cences are longer (to 15 cm). The type (Ule reba Falls, Kurupung River, 24 Feb. 1939 (fr) , 6044, collected nearby at Sao Joaquim on the Pinkus 261 (MO, NY). VENEZUELA. Amazonas: Rio Negro), is completely glabrous except for Dept. Casiquiare, Cano San Miguel, 160 m, 20 the vegetative buds and ciliate margins of the Apr. 1991 (bud), Aymard 9050 (SEL); Dept. Rio calyx-lobes, and the inflorescence is only 5-6 Negro, Rio Emoni, 17-18 Apr. 1985 (bud), Ster­ cm long. Its closest relative appears to be Myr­ gios et at. 8269 (MO, SEL). Bolivar: Km 109 cia grandis McVaugh. carretera Sta. Elena, 450 m, 24 Feb." 1959 (past Specimen examined. VENEZUELA. Amazonas: fl), Bernardi 7234 (MER). Callo Cuweje, 32 km S of San Carlos de Rio Negro, 119 m, 01°56'N, 67°03'W, 4 Apr. 1980 Marlierea insignis McVaugh, Fieldiana, Bot. (fr), Clark 7485 (MO, NY, VEN). 29: 176. 1956. Marlierea subulata McVaugh, Fieldiana, Bot. Previously only known from the Rio Apaporis 29: 177. 1958. TYPE: Pem. Loreto: Mishuy­ region of Colombia, this species can now be re­ aca, near Iquitos, 100 m, May-Jun. 1930 ported from Peru and Venezuela from a long- (fl), Klug 1341 (Holotype: F! ; NY! , US! ). 174 SELBYANA Volume 23(2) 2002

See McVaugh (1958a) for description. Previ­ (fr), Liesner 6171 (VEN); 1-2 kID southeast and ously known only from the type collection from east of San Carlos de Rio Negro, 120 m, Peru, this has more recently been found to be a Olo56'N, 67°03'W, 22 Apr. 1979 (fr), Liesner species widespread in the Amazon Basin; col­ 6880 (MO, VEN); Neblina Massif, mouth of lected now in southern Venezuela, Brazil, and Calion Grande, 500-700 m, 00050'N, 66°06'W, Bolivia. 15 Jul. 1984 (fr), Davidse & Miller 27389 (MO, The leaves are variable in size. In Revilla VEN). 2065 they measure up to 30 cm in length and summa McVaugh, Mem. New York nearly 13 cm wide, much larger than the type, Marlierea Bot. Gard. 10(1): 89. 1958. but exactly alike in detail. Venezuelan material differs slightly from that of other countries in A species restricted to higher elevations of the that the leaves are more glandular-punctate, and sandstone massifs of the Guayana Shield. Pre­ it has reddish appressed trichomes on the viously only known from Venezuela, it is here branches of the inflorescence, instead of yellow­ reported as a new species for the Guyanan and ish erect trichomes in the Peruvian material, or Brazilian floras. the denser, ascending trichomes of the Brazilian BRAZIL. Amazonas: Southern extremity of and Bolivian material. Northern Plateau of Serra Araca, 1200 m, Additional specimens examined: BOLIVIA. 00051-57'N, 63°21-22'W, 13 Feb. 1984, Prance Pando: Nicolas Suarez, en la zona del Narueda, et al. 29040 (MO, NY); Arredores de R. da serra Campoana, junto ala barraca San Jose, 290 m, Araca, 28 Jan. 1978, Rosa & Silva 2264 (MO). 13 Jan. 1983 (fr), Casas & Susanna 8213 (MO); GUYANA. Upper Potaro River region, summit of west bank of Rio Madeira, 2 kID above Riverao, c Mt. Wokomung, 1650 m, 05°05'N, 59°50'W, 7 26 Jul. 1968 (fr), Prance et al. 6480 (F, US); La Jul. 1989 (fl), Boom & Samuels 9125 (MO); Mt. Paz: Provo Iturralde, Siete Cielos, Rio Manupare, Wokomung, 1530 m, 05°04'N, 59°52'W, 20 Feb. 180 m, 12°27'N, 67°37'N, 4 Jun. 1987 (fl), Sol­ 1993 (bud), Henkel et al. 1519 (SEL). omon 16908 (MO). BRAZIL. Amazonas: Rio Cu­ ruquete, halfway between Cachoeiras Sao Paulo Marlierea uniftora McVaugh, Mem. New York and Republica, 22 Jul. 1971 (US), Prance et al. Bot. Gard. 18(2): 69. 1969. 14509 (US); Rondonia: Basin of Rio Madeira, Described only from a single specimen in north bank of Rio Abuna between Cachoeira flower, several more collections have been gath­ Tres Sand Fortaleza, 4-16 kID above mouth, 18 ered since, including one in fruit, which is de­ Jul. 1968 (bud), Prance et al. 6168 (F, US); Vi­ scribed briefly. A Brazilian collection cited be­ cinity of Jam, Porto Velho to Cuiaba highway, low represents the first collection for that coun­ 16 Aug. 1968 (fr), Forero & Wrigley 7119 (F, try. US); Basin of Rio Madeira, 8 kID NE of Porto Fruit round, 6 X 6 mm, black at maturity, 2- Velho, 7 Nov. 1968 (fr), Prance et al. 8245 (F, 4 irregular calyx-lobes persisting. US). PERU. Loreto: Prov. Alto Amazonas, Rio BRAZIL. Amazonas: Mun. de Sao Gabriel da Pastaza, Andoas, Capahuari Sur, 11 Nov. 1979 Cachoeira, Rio Cubate, afluente do I<;ana, 4 Nov. (fr), Ayala 2218 (MO); Rio Nanay, University 1987 (fl), Farney et al. 1886 (MO). VENEZUELA. Arboretum at Puerto Almendra, 27 Jul. 1972 Amazonas: Dept. Rio Negro, Rio Pasimoni, 80 (fr), Croat 18530 (MO); Provo Maynas, Rio Na­ m, Olo53'-01°27'N, 66°32'-66°35'W, 23-25 Jul. nay, Caserio Lupuna, 17 Dec. 1976 (fr), Revilla 1984 (fl), Davidse 27836 (MO, SEL);Calio Mo­ 2065 (MO); Provo Requena, Jenaro Herrera, Rio moni, medio Casiquiare, 18 Feb. 1986 (fr), Ster­ Ucayali, Aucayacu, 140 m, 04°55'N, 73°50'W, gios & Aymard 9161 (NY); Dept. Atabapo, Rio 16 Jul. 1980, Vasquez & Jaramillo 279 (MO); Atabapo, 15 kID SE de San Fernando de Ata­ Quebrada Santa Cruz, 130 m, 03°50'S, 73°35'W, bapo, 110 m, 03°55'N, 67°40'W, 10-16 Jan. 21 Mar. 1982 (fr), Vasquez & Ruiz 2959 (ASU, 1988 (imm fr), Stergios et al. 11668 (SEL); MO); Alpahuayo, Estacion IIAp, 19 Oct. 1984 Calio Monomi-Casiquiare, 2 Nov. 1962 (fl), Va­ (bud), Vasquez & Criollo 5779 (MO, CAS); reschi 7803 (VEN); Rio Atabapo, cerca Isla Prov. Maynas, Pto. Almendras, 122 m, 03°48'S, zapo, 90 m, 18 Aug. 1964 (fl), Vareschi & Jaffe 73°25'W, 29 Sep. 1986 (bud), Vasquez & Jar­ 8011 (VEN). amillo 8053 (MO); Prov. Maynas, Iquitos, Pto. Almendras, 122 m, 03°48'S, 73°25'W, 3 Nov. 1987 (fl), Vasquez & Jaramillo 9971 (MO). VE­ Myrcia DC. NEZUELA. Amazonas: Dept. Rio Negro, 4 kID east Myrcia albidotomentosa (Aubl.) Amshoff, of San Carlos de Rio Negro, 120 m, 01056'N, Mem. New York Bot. Gard. 18(2): 79. 67°04'W, 10 Nov. 1977 (bud), Liesner 3283 1969. (VEN); 4.5 kID northeast of San Carlos de Rio Negro, 120 m,01°56'N, 67°03'W, 5 Apr. 1979 A distinctive species well known from middle HOLST: MYRTACEAE 175 elevation forests of the Guayana Shield in north­ or Guayanan species. As currently circum­ ern Amazonas state and the Gran Sabana region scribed, it is centered in southern Brazil, with of Bolivar state, Venezuela, as well as Guyana only one species known from as far north as the and Suriname. It can now be reported from Bra­ Greater Antilles. I believe that this complex is zil from a collection on the slopes of Serra Araca more widespread, occurring west to the Andes of the Brazilian Guayana. and as far north as Panama, but the definite as­ BRAZIL. Amazonas: Encosta da Serra Araca, 4 signment of names in that area will be difficult Feb. 1978, Rosa & Lira 2371 (NY). without an examination of several types in Eu­ The leaves of this species are remarkeably ropean herbaria. It is still not clear whether like those of several species of the "decorti­ Gomidesia is sufficiently distinct from Myrcia, cans" alliance, most notably Myrcia decorti­ so I am retaining the species in Myrcia for the cans. They are coriaceous, dry reddish brown, present. and the short petiole is corky-rimose. The slen­ Myrcia calycampa Amshoff, Recueil Trav. Bot. der, neady spike-like inflorescences and 4-mer­ Need. 39: 153. 1942. ous pubescent flowers reduced in number how­ ever resemble a group of species that includes Calycampe latifolia O. Berg, Linnaea 27: 130. Myrcia minutiflora Sagot, M. gentryi B. Holst, 1856, non Myrcia latifolia O. Berg. TYPE: Guy­ M. ptariensis (Steyerm.) McVaugh, M. induta ana. Schomburgk 927 (Syntype, NY!). McVaugh, M. cuprea Amshoff, and M. bolivar­ Calycampe angustifolia O. Berg, Linnaea 27: ensis (Steyerm.) McVaugh, all of which are also 131. 1856, non Myrcia angustifolia (0. Berg) montane species. Nied. Myrtus latifolia (0. Berg) Badillo, nom. nud. Myrcia aliena McVaugh in Macbride, Field Once considered as the monotypic genus, Ca­ Mus. Nat. Hist., Bot. Ser. 13(4.2): 627. lycampe O. Berg based on the rounded sinuses 1958. -Aulomyrcia chilensis O. Berg, Lin­ between the calyx-lobes, this species, previously naea 27: 38. 1855, non Myrcia chilensis O. known from Guyana and Brazil, can be reported Berg, 1855. TYPE: Peru. Huanaco: Montes now from Venezuela from the following collec­ sylvat. Ad Cassapi, Poeppig 133 (G; Field tions. Museum negative 23315!). VENEZUELA. Bolfvar: Rio Cuyuni, entre Ana­ A species new for the Venezuelan flora, this coco y la Isla de Cano Negro, 16 Jan. 1983 (bud, has been previously reported from Bolivia, Peru, fl), Stergios et at. 5092 (MO, PORT); Reserva Ecuador, and Amazonian Brazil (McVaugh Forestal Imataca, Rio Cuyuni, alrededores de la 1958a). Isla Anacoco, 19-20 Jul. 1983 (fr), Stergios et The following specimens are tentatively al. 6269 (MO, PORT). placed in Myrcia aliena, the oldest name that I Myrcia c1usiifolia (H.B.K.) DC., Prodr. 3: 255. can find for this species complex in northern and 1828. eastern South America. The other related species are: M. neesiana McVaugh (Amazonia) and M. New for Colombia; previously known from clausa McVaugh (Venezuela), Myrcia impeifec­ Brazil (Amazonas, Roraima) and Venezuela ta McVaugh (Guyana), and possibly M. directa (Amazonas). McVaugh (Peru). Limited to white sand savannas Myrcia clu­ Specimens examined. BRAZIL. Amazonas: siifolia is well distinguished for a member of Manaus, 24 Dec. 1936 (bud), Ducke 363 (MO). Myrcia section Myrcia by the ridged hypanthi­ VENEZUELA. Amazonas: Dept. Atabapo, Salto urn and oblong or elliptic coriaceous leaves that Yureba, Cano Yureba; Bajo Rio Ventuari, 120- tend to tear along the midrib at the tip when 150m,. 04°03'N,66"01'W, 24 Oct.-4 Nov. 1981 pressed flat. (bud,. fl), Delascio & Guanchez nOlO (MO, COLOMBIA. Guainia: Rio Negro, San Felipe, VEN); Rio Cunucunuma, base of Huachamacari, below confluence of Rios Guainia and Casiqui­ 220 m, 03"39'N,. 65°41'W; T Mar. 1985 (fr), uare, 12 Dec. 1947, Schultes & LOpez 9322 Liesner 18409 (MO,. VEN). (US). Myrcia aliena, along with the several other Myrcia cuprea (0. Berg) Kiaersk., Enum. Myrt. related species mentioned above from eastern Bras. 95. 1893. and northern South America, and Panama seem to be related to what is called Gomidesia in Previously only known from Brazil, the first southern South America, distinguished mainly collection from any of the Guianas is listed be­ from Myrcia by the completely or imcompletely low. four-locular anthers, though it is not always FRENCH GUIANA. Haut Oyapock, sur la savane readily visible. Gomidesia was reviewed by Le­ roche au Nord du Mont Saint Marcel, 31 Jul. grand (1958), but he did not treat any Andean 1975 (bud), de Granville B. 5302 (MICH). 176 SELBYANA Volume 23(2) 2002

Myrcia fenestrata DC., Proclr. 3: 251. 1828. COLOMBIA. Guainia: 0.5 km north of Boca de Casiquiare, 120 ill, 01°57'N, 67°08'W, 5 Feb. New record for Bolivia. 1980, Liesner & Clark 9103 (MO, VEN); Near BOLIVIA. Santa Cruz: Prov. Velasco, Parque Coitara, ca. 7 km S of San Fernando de Atabapo, Nacional Noel Kempff Mercado, campamento 95 m, Q3°55'N, 67°43'W, 28 Sep. 1979, Davidse Las Gamas, 850 m, 14°48'27"S, 60°24' 11''w, 15 16841 (MO). Nov. 1993 (ft), Arroyo et al. 351 (SEL); Parque Nacional Noel Kempff Mercado, Campamento Myrcia graciliflora Sagot, Ann. Sci. Nat. ser 6, Las Gamas, 850 m, 14°48'23"S, 60°23' 12''w, 27 20: 185. 1885. Oct. 1995 (ft), Foster et at. 493 (SEL). New to Guyana; previously known from Su­ Myrcia fenzliana O. Berg in Mart., Fl. Bras. riname. 14(1): 196. 1857. GUYANA. U. Takutu-U.Essequibo. Kassikaityu Gomidesia lindeniana O. Berg, Linnaea 29: R., between camp and unnamed outlier of Kamoa 208. 1858. Mts., 5 km S of Camp, 240 m, 01°49'N, 58°44'W, Previously believed to have a disjunct range 23 May 1997 (fr), Clarke 4871 (SEL); Tumatu­ between the West Indies and Minas Gerais, Bra­ mari, 18 Jun.-8 Jul. 1921, Gleason 169 (NY). zil, this has been collected several times in the Myrcia magnoliifolia DC., Prodr. 3: 248. 1828. Gran Sabana of Bolivar state, Venezuela, Guy­ ana and Colombia, Dept. Santander. New species for Bolivia; previously known Specimens examined. COLOMBIA. Dept. San­ from northern Amazonia and the Guayana tander: Mesa de los Santos, 1500 m, 11-15 Dec. Shield. 1926 (fr), Killip & A.C. Smith 15060 (US). GUY­ BOLIVIA. Santa Cruz: Parque Nacional Noel ANA. Potaro-Siparuni region, northern Pakarai­ Kempff Mercado, Estaci6n Flor de Oro, 200 m, mas, Ciong Valley, 9 km north of Kato Village, 13°33'S, 61°00'W, 20-21 Sep. 1995 (ft), Vargas Manawarrai Mountain, 2000 m, 04°42'N, et at. 3836 (MO). 59°50'W, 1 Jun 1995, Mutchnick 1448 (SEL). Myrcia minutiflora Sagot, Ann. Sci. Nat. ser 6, VENEZUELA. Bolivar: 2-10 km from El Dorado­ 20: 185. 1885. Sta.Elena Rd. on road to Kavanayen, 1200-1250 m, 16 Mar. 1974 (fr), Gentry et at. 10514 (MO); Previously known from the Guyana, Surina­ Distr. Sifontes, Mun. Urdaneta, 5 km E de San me, French Guiana, and Brazil (Amapa, Mar­ Ignacio, 1120 ill, 05°00'N, 61°00'30''w, 23 Apr. anhao, Para), this distinctive species has been 1985 (fr), Dezzeo & Hernandez 146 (MO); Ay­ collected twice at a considerable distance to the avapani, 15 km al oeste de Wadakapiapue, 1110 west in Dept. Loreto, Peru. It is closely related m, 05°16'30''N, 61°00'W, 6 Mar. 1987 (fr), L. to Myrcia bolivarensis (Steyerm.) McVaugh, but Hernandez 392 (MO); Murric. Gran Sabana, differs principally by the venation. Myrcia bo­ aprox. 10 km S.W. del Wadakapiapue-tepui, livarensis has numerous secondary veins and the 1100 m, 05°15'N, 600 58'W, 18 Feb. 1987 (fr), marginal vein is near and parallel to the margin, Huber & Hernandez 11946 (SEL); Dtto. Sifon­ whereas in M. minutiflora, there are few second­ tes, Parupa via Toron meru, 1300 m, 05°45'N, ary veins and the marginal vein arches notice­ 61°43'W, 25 Apr. 1986 (fr), Picon 1095 (MO). ably between them and is more removed from the leaf margin. See McVaugh (1969) for addi­ Myrcia gigas McVaugh, Mem. New York Bot. tional specimen citation and a summary of dis­ Gard. 18(2): 88. 1969. tinctive characters. Previously only known from northeastern PERU. Loreto: Provo Maynas, Quebrada Su­ Brazil, two collections are now known from cusari, Rio Napo below Mazan, 140 m, 6 Nov. French Guiana. 1979 (fr) , Gentry et at. 27563 (F, MO); 7 Nov. FRENCH GUIANA. Saii1: Monts la Fumee, 200- 1979 (sterile at MO), Gentry et at. 27639 (MO); 400 m, Q3°37'N, 53°12'W, 1982 (imm. fr.), Pasco: Shiringamazu, ca 20 km S of Iscozacin, Boom & Mori 2431 (MO); 9 Sep. 1982 (bud), Rio Pa1cazu Valley, 300 m, 100 48'S, 75°1O'W, 8 Mori et at. 14893 (MO); 22 Sep. 1982 (ft), Mori Jul. 1988 (sterile), Gentry et at. 63400 (MO); & Boom 14966 (MO); 7 Oct. 1982, Mori et at. 63404 (MO). 15065 (MO). Myrcia multiflora (Lam.) DC., Prodr. 3: 244. Myrcia grandis McVaugh, Mem. New York 1828. Bot. Gard. 18(2): 114. 1969. Aulomyrcia vinacea Steyerm., Fieldiana, Bot. A well collected species in the upper Rio Ne­ 28: 1008. 1957. gro and Rio Orinoco basins of Venezuela and This common and widespread species in Brazil. The following collections represent the South America can be reported from Venezuela first records for Colombia. from the following collections. HOLST: MYRTACEAE 177

VENEZUELA. Amazonas: Dept. Atures, Pto. 2 pairs of unequal calyx lobes, convex midvein, Ayacucho, 30 Jan. 1978, Huber & Cerda 1520 nearly glabrous flowers and reddish dibrachiate (MICH, NY, YEN). Bolivar: Represa Guri, 0.5 trichomes on the inflorescence. kIn SSW of dam, 250-350 m, 07°46'N, VENEZUELA. Bolivar: De Botanamo a Corumo, 63°00'W; 31 Mar. 1981, Liesner & Gonzalez 17 Dec. 1959, Bernardi 8045 (NY, VEN). 11040 (MO, VEN); Represa Guri, 250-300 m, Myrcia revolutifolia McVaugh, Mem. New 07°35'N, 63°07'W, 4-5 Apr. 1981, Liesner & Gonzalez 11218 (MO, NY, VEN); Dist. Roscio, York Bot. Gard. 18(2): 121. 1969. TYPE: Venezuela. Amazonas: Cerro Coro Coro, 2 Calio Orocopiche, carretera Cd. Bolivar-Cd. Mar. 1953, Maguire 35462 (Holotype: Piar, carretera Cd. Piar-Pto. Ordaz, 16 May MICH!; Isotype: F!, VEN!). 1982, Stergios et al. 3503 (MO, NY, PORT); Dist. Heres, Cerro Arimagua, Embalse de Guri, Myrcia planifolia McVaugh, Mem. New York 480 m, Jan. 1984, A. Fernandez 794 (MO, Bot. Gard. 18(2): 118. 1969. TYPE: Venezuela. PORT). Sucre: Valley of Cocollar, 820 m, 28 Amazonas: Cerro Sipapo, Sabana Grande, vicin­ Apr. 1945, Steyermark 62375 (VEN, type of Au­ ity of Campo Grande, 21 Jan. 1949, Maguire & lomyrcia vinacea Steyerm). Politi 28685 (Holotype: MICH!; Isotypes: F!, MO!, YEN!). Myrcia pistrinalis McVaugh, Mem. New York At the time of publication, Myrcia planifolia Bot. Gard. 18(2): 117. 1969. was distinguished from M. revolutifolia by hav­ New for Venezuela; previously known from ing a glabrous inflorescence and flat leaves (vs. Suriname. pubescent inflorescence and the leaves markedly VENEZUELA. Bolivar: Dist. Roscio, Base de revolute). With abundant additional collections Peray-tepui y Guy-tepui, 7 kIn NW del Caserio made from the type locality and numerous other El Pi16n, 58 kIn al. Oeste de Sta. Elena, 850- collections in the Guayana Region, it is clear 1100 m, 04°40'N, 61 0 33'W, 23 Oct. 1986, Ay­ that neither the degree of leaf margin revolution mard 4831 (MO, PORT). nor the amount of pubescence can be used suc­ cessfully to separate the two species. Pubes­ Myrcia platyclada DC., Prodr. 3: 244. 1828. cence varies from nearly completely glabrous in New for Brazil; previously known from Ve­ populations on Cerro Sipapo, Cerro Vinilla and nezuela and the Guianas. in Colombia to puberulent on Cerro Yutaje, Cer­ BRAZIL. Para: KIn 100 Belem-Brasilia, 14 ro Marahuaca, and Cerro Pam. The calyx lobes Aug. 1964 (fr), Prance & N.T. Silva 58730 and staminal ring also vary as to the amount of (MO); vic. Cachoeira, BR 22, KIn 98, 24 Aug. pubescence within, ranging from glabrous to 1964 (fr), Prance & N.T. Silva 58855 (MO). thinly pubescent. Leaf margins are typically slightly revolute (induding the holotype at Myrcia nubicola McVaugh, Mem. New York MICH and an isotype at NY of Myrcia revolu­ Bot. Gard. 18(2): 116. 1969. tifolia), and are less commonly plane, so the M. New for Guyana; previously only known from revolutifolia is selected for use. Venezuela in Amazonas and Bolivar states. Myrcia revolutifolia can be confused with M. GUYANA. Cuyuni-Mazaruni Region: Mt. Ay­ rotundata McVaugh var. rotundata because of the anganna, easternmost peak, 1350-1380 m, similar leaf shape and texture. Myrcia rotundata 05°25'N, 59°57'W; 11 Mar. 1987 (fr), Pipoly et can be distinguished, howeve~ by the four calyx­ al. 11115 (SEL). lobes, the single panicle per leaf axil, and the nu­ merous secondary veins that are closely parallel. Myrcia pyrifolia (Desv. ex Ham.) Nied. in Engl. Myrcia revolutifolia can also be confused with M. & Prantl, Nat. Pflanzenfam. 3(7): 76. 1893. guianensis which also occurs frecuently in the New species for Peru; previously known from same area. Myrcia guianensis can be distinguished northern Amazonian Brazil, Colombia, Vene­ by the trilocular ovary, calyx lobes densely pubes­ zuela, and the Guianas. cent within, and a single panicle per leaf axil. Myr­ PERU. Loreto: Rio Mamon near Rio Nanay, 1 cia revolutifolia has the ovary bilocular, the 5 ca­ Sep. 1972, Croat 19889 (SEL). lyx lobes are only sparsely pubescent or glabrous within, and the inflorescence is of paired panicles Myrcia quitarensis (Benth.) Sagot, Ann. Sci. with an aborted bud inbetween. Nat. ser 6, 20: 184. 1885. Myrcia revolutifolia is by no means limited to Although I have not seen the type, the pre­ high altitudes of the sandstone table mountains, viously only known collection, there is little and has been found in white-sand savannas doubt that the specimen cited below belongs down to 100 m above sea level. Additional col­ with this species, and represents a new record lections have also extended the range to include for Venezuela. It has the unique combination of northern Brazil and eastern Colombia. 178 SELBYANA Volume 23(2) 2002

Selected specimens examined: BRAZIL. Ama­ ermark et al. 124524 (MO, NY, VEN); same lo­ zonas: encosta da serra Araca, 600 m, 31 Jan. cality, same date (sterile), Steyermark et al. ]978 (fr), Rosa & Lira 2324 (NY). COLOMBIA. 124543 (NY, VEN); Dept. Rio Negro, N. del Vaupes: Mitu and vicinity, lower RIO Kubiyu, 5 Cerro Vinilla, 30 km S.S.W. de Ocamo, 440- Jul. 1975 (fl), Zarucchi 1385 (MICH); same lo­ 600 m, 02°31 'N, 6S023'W, 1-2 Mar. 1984 cality, 30 Jun. 1976 (imm. fr), Zarucchi & Bal­ (bud); Steyermark et al. 130321 (MO, VEN); ick 1789 (MICH). VENEZUELA. Amazonas: Dept. Dept. Atabapo, Cerro Marahuaca, slopes, 1140 Rio Negro, N. de la Serranfa A vispa, 300-3S0 m, 03°43'N, 65°31'W, 21-22 & 24 Feb. 1985 m, 02°46'N, 6S029'W, 28 Feb. 1981 (bud), (bud), Steyermark & Holst 130527 (MICH, Guanchez 878 (MO); Dept. Rio Negro, Sierra MO, VEN). Bolivar: Meseta del Jaua, Cerro de Untunin, vertiente, 550-6S0 m, 01 °32'N, Sarisarifiama, summit, 1400 m, 04°41'40"N, 6S011'W, 26 Feb. 1981 (imm. fr), Guanchez 839 64°13'20"W, 16-18 Feb. 1974, Steyermark et al. (MO, TFAV); Dept. Atures, CaBo Cabeza de 109176 (MICH, VEN). Manteco, 90-100 m, 04°51'N, 67°32'W, 11 Dec. A single collection from Cerro Marahuaca in 1984 (sterile), Guanchez & Melgueiro 3564 Amazonas state, Venezuela may represent a va­ (MO, TFAV); Dept. Atures, Valley of Rio Coro­ riety of Myrcia revolutifolia or perhaps a distinct Coro, W. of Serrania Yutaje, 6S0 m, OS 41'30"N, species. The leaves are broadly elliptic-oblong 66 07'30"W, 21 Feb. 1987 (imm. fr), Holst & 4-7 cm long, 3.S-S cm wide, apex rounded to Liesner 3106 (MO, VEN); same locality, 1100 truncate-emarginate, rounded at the base, and m, OS041'N, 66°09'W, 3 Mar. 1987 (imm. fr), the individual panicles are many flowered. Holst & Liesner 3367 (MO, VEN); Dept. Atu­ VENEZUELA. Amazonas: Dept. Atabapo, Be­ res, 6-8 km S.E. del medio Rio Guayapo, 110 low Salto Los Monos, Cerro Marahuaca, lSOO- m, 04°1O'N, 67°2S'W, 29 Jun. 1979 (bud), Hub­ 16S0 m, 03°3S'N, 6S023'W, 12 Mar. 1985 (bud), er 3912 (NY, US, VEN); Dept. Atures, 10 km Liesner 18577 (MO, VEN). S. del Rio Autana, 15 km S.W. del Cerro Au­ Myrcia sipapensis McVaugh, Mem. New York tana, 100 m, 04°44'N, 67°33'W, 2 JuL 1979 Bot. Gard. 18(2): 140. 1969. TYPE: Vene­ (bud), Huber 4059 (MICH, MO, NY, US, VEN; zuela. Amazonas: Cerro Sipapo, east es­ Dept. Atures, 20 km al N.W. del Cerro Autana, carpment, 2000 m, 14 Jan. 1949, Maguire 100 m, 04°54'N, 67°36'W, 14 Jul. 1980 (fl, fr), & Politi 28342 (Holotype: MICH!; Iso­ Huber & Tillett 5304 (MICH, NY, US, VEN); types: NY!, VEN!). Dept. Casiquiare, 20 km N.W. de Yavita, 120 m, 03°01'N, 6r33'W, 11 Feb. 1981 (bud), Huber Previously known only from the type collec­ & Medina 5968 (MICH, VEN); Dept. Rio Ne­ tion in bud and possibly another in fruit, the fol­ gro, N. del Cerro Vinilla, 400 m, 02°31'N, lowing collection provides the first known flow­ 65°23'W, 14 Feb. 1981 eft), Huber 6031 (MICH, er (described below) and extends the range of VEN); Dept. Atabapo, Cerro Duida, plateau, the species to extreme southern Venezuela. 1250 m, 03°36'N, 66°42'W, 13 Mar. 1985 (bud), Inflorescence axillary or terminal; hypanthi­ Liesner & Morillo 18625 (MO, VEN); Dept. um prolonged 1.S mm above the ovary, tearing Atures, W. side of valley of Rio Coro-Coro, longitudinally between the calyx lobes partially SOO-1000 m, OS041'N, 66°08'30"W, Liesner & or completely to the disk, velutinous within and Holst 21431 (MO, VEN); Dept. Rio Negro, Cer­ without; calyx lobes shallowly triangular, broad­ ro Aracamuni, slope, 600 m, 01024'N, 65°38'W, ly obtuse to rounded at the apex, 1.5 mm long, 21 Oct. 1987 (fl), Liesner & Carnevali 22274 2 mm wide, sparsely sericeous on both sides, (MO, VEN); Dept. Rio Negro, Cerro Aracamu­ ciliate; petals oblong or obovate 4.5-S mm long, ni, summit, 1400 m, 01032'N, 6so49'W, 1 Nov. 3-3.S mm wide, glabrous within, sericeous with­ 1987 (bud), Liesner & Carnevali 22702 (MO, out, margins thinly membranous and somewhat VEN); Cerro Sipapo, 1600 m, 20 Dec. 1948 (fl), irregular; filaments 4-S mm long, anthers ca. 0.5 Maguire & Politi 27812 (F, MICH, MO, NY, mm long; style ca. S mm long, broad at the base, VEN); Cerro Sipapo, 1400 m, 1 Jan. 1949 (bud), narrowing gradually to the flattened, minute Maguire & Politi 28113 (MICH, NY); Cerro Si­ stigma, sparsely sericeous towards the base. papo, 2000 m, 3 Jan. 1949 (bud), Maguire & VENEZUELA. Territorio Federal Amazonas: Politi 28137 (NY); Cerro Sipapo, 1500 m, 20 Dept. Rio Negro, Cerro de la Neblina, Camp Jan. 1949 (fl), Maguire & Politi 28474 (MICH, VII, 1730-18S0 m, 000S2'N, 6soS8'W, 7 Feb. NY, VEN); Cerro Sipapo, 1S00 m, 21 Jan. 1949 1985, Renner 2089 (US). (fl), Maguire & Politi 28685 (MICH, NY, MO, Myrcia sipapensis, as well as other closely re­ VEN, type of M. planifolia); Cerro Paraque (Si­ lated species, Myrcia bonnetiasy/vestris (Stey­ papo), 1600 m, Feb. 1946 (fl), Phelps 33 (VEN); erm.) Steyerm., Marlierea rugosior McVaugh, Dept. Atures, Serranfa Sipapo, summit, IS00 m, point out yet another problem in the traditional aprox. OsoN, 67°30'W, 17 Feb. 1981 (bud), Stey- delimitation of the Myrciinae genera. The hy- HOLST: MYRTACEAE 179

panthyium clearly shows elongation above the aima, Rio Carrao, 400 m, 06°15'N, 62°47'W, 18 summit of the ovary and the flowers show some Jul. 1972, Steyennark 106322 (VEN). tearing of the hypanthium between the calyx A sterile specimen that matches very well lobes. Ordinarily these should all be placed in with the collections cited above was found at a Marlierea as McVaugh did with M. rugosior. much higher altitude. VENEZUELA. Amazonas: However, they are like Myrcia in the well de­ Dept. Atures, Cerro Ualipano, Rio Parucito, veloped 5 calyx lobes in bud, and the terminal 1700 m, 06°N, 65°43'W, 1 Feb. 1962 (st), Car­ or axillary, and mostly solitary panicles. There dona 2829 (MICH, MO, NY, US, VEN). may also be some connection with this complex, and what is occasionally recognized as Gomi­ Myrciaria vismeifolia (Benth.) O. Berg, Lin- desia. They share the same characters of the ca­ naea 27: 336. 1856. lyx persistent into fruit and forming a truncate With a previous distribution of French Guiana, crown, the brownish indument, and the leaves Suriname and Guyana (McVaugh 1969), it is not usually darkening upon drying. surprising that this species has now been collect­ Myrcia subobUqua (Benth.) Nied. in Engler & ed from eastern Venezuela and northern Brazil. Prantl., Nat. Pflanzenf. 3(7): 76. 1893. BRAZIL. Amazonas: Munic. de Itapiranga, Rio Uatuma, 26 Aug. 1979, Cid et al. 810 (US, NY); New to French Guiana; previously known Rio Uatuma, entre rios Pitanga e Uatuma, 1-2°S, from Guyana and Suriname. 59-60°W, 18 Mar. 1986 Cfr), Cid Ferreira et al. FRENCH GUIANA. Trois Saut (Haut Oyapock), 6786 (NY). Para: Mun. Oriximana, Rio Cachorro, Aval de Zidock ville, 20 Dec. 1974 (fl), Gren­ proximo a Cachoeria Sao Pedro, 00053'S, and 584 (SEL). 57°13'W, 25 Aug. 1986 Cfr), Cid Ferreira et al. 8032 (NY). VENEZUELA. Bolfvar: Reserva Forest­ Myrciaria O. Berg al Imataca, Rio Cuyuni, Isla Anacoco, 15 Jan. 1983, Stergios et al. 5000 (MO, PORT); Reserva Myrciaria dubia (H.B.K.) McVaugh, Fieldiana, Forestal Imataca, Rio Cuyunf, entre Isla Anacoco Bot. 29: 501. 1963. y boca del Rio Botanamo, 17 & 19 Jan, 1983, Stergios et al. 5165 (MO, PORT); Rio Venamo, New for Guyana and Ecuador; previously entre Rio Cuyuni y Cano Apanao, 18 Jul. 1983, known from most other countries of the upper Stergios et ai. 6134 (MO, PORT); same locality Amazon basin. and date, Stergios et al. 6145 (MO, PORT). GUYANA. Cuyuni-Mazaruni Region: Essequi­ bo River, between Omai and Dennison Mine Camp (near Kimaka River), 10-15 m, 05°21'N, Plinia L. 58°45'W, 27 May 1989, Gillespie & Persaud Plinia costata Amshoff, Recueil Trav. Bot. 1435 (MO). ECUADOR. Rio Cuyabeno, Laguna Neerl. 42: 11. 1950. de Canangueno, 21 Aug. 1986 (fl), Jaramillo 9038 (NY); Sucumbios: Lago Agrio Canton, Re­ New for Guyana; previously known from Su­ serva Cuyabeno,. 230 m, 00°00', 76°11'W, 16 riname. Nov. 1991 (fl), Palacios et al. 9002 (MO). GUYANA. Kanuku Mountains, Jardin Falls, 330 m, 03°21'N, 59°27'W, 24 Nov. 1987 (fl, Myrciaria tenella (DC.) O. Berg in Mart., Fl. imm fr), Jansen-Jacobs et al. 1214 (MO). Bras. 14(1): 368. 1857. Plinia rivularis (Camb.) Rotman, Bol. Soc. Ar- A well known species of southern South gentina Bot. 24(1-2): 196. 1985. -Eugenia America, can now be reported from the north. rivularis Camb. in A. St.-Hil. Fl. Bras. Mer­ FRENCH GUIANA. Haute Camopi, Mont Bel­ id. 2(2): 337. 1832. TYPE: Brazil. Rio de vedere, 30 Nov. 1984 (imm. fr.), Granville 7080 Janeiro: St. Hilaire S.n. (P, fragment at F; ("cf." U). VENEZUELA. Bolivar: Dist. Piar, Sa­ photo of type, MO i). vanna of Amaruay-tepui, 500 m, 05°55'N, 62°16'W, 29 Apr 1986, Holst & Liesner 2750 Myrcia granulata R.O. Williams, Fl. Trinidad (MO, PORT, VEN); Rio Aparaman, affluent of 1(6): 340. 1932. Rio Acanan, SW base of Amaruay-tepui, 500 m, This variable species in leaf and inflorescence 05°55'N, 62°15'W, 6 May 1986, Holst & Liesner size and shape, was previously thought to be 2796 (MO, VEN); Rio Acanan, SW comer of limited to southern South America (Argentina, Amaruay-tepui, 500 m, 05°55'N, 62°16'W, 22 Paraguay, Uruguay and southern Brazil). It ac­ Mar. 1987, Holst 3474 (ICN, MO, VEN); Rio tually seems to be quite common north of the Aparaman, affluent of Rio Acanan, SW base of Amazon Basin, even as far as Trinidad. Amaruay-tepui, 500 m, 05°55'N, 62°15'W, 22 There has been much confusion over the cor­ Mar. 1987, Holst 3476 (ICN, MO, VEN); Can- rect placement of this species, and the circum- 180 SELBYANA Volume 23(2) 2002 scription of it, but more recent workers (Rotman, NY, US, VEN). Barinas: Dist. Obispos, 20 km 1982, 1985; Sobral, pers. com.) suggest its place­ al NO de Barrancas, 300-600 m, 29 Mar. 1972 ment in Plinia because of the bilocular ovary (ft), Marcano Berti & Torres Lezama 3009 with 2 ovules per locule, distinct cotyledons, and (VEN). Dist. Fed.: Cerros alrededor de Caruao, the non-circumscissile hypanthium. There is con­ hacia Los Caracas, 1 May 1969 (fl), Aristeguieta siderable variation in the size and texture of the 7109 (NY, VEN). Yaracuy: Dtto. San Felipe, leaves, and in the length of the inflorescence. In Munic. Veroes, al sur de Bella Vista, lOo23'N, flower, it is difficult to tell apart from several 4- 68°24'W, 11 Jul. 1972 (fr), Agostini 1757 (MO). merous Myrcia species with convex midvein. The only known collections of Myrcia gran­ ACKNOWLEDGMENTS ulata are in flower or young fruit, so the embryo was not seen, however a paratype from Trinidad I thank Maria Lucia Kawasaki, Leslie Lan­ (cited below) is nearly identical with many Bra­ drum, Fred Barrie, Francesca Grifo, and Marcos zilian and Argentinian collections, and I have no Sobral, for their many discussions of Myrtaceae doubt that it belongs there. McVaugh (1969) in­ taxonomy over the years, for their identifications dicated that the Williams' Venezuelan collec­ of specimens, and for their collaboration on pro­ tions cited below would key out to Myrcia gran­ jects. I also thank the herbaria cited for loans of dis McVaugh because of the convex midvein specimens, the individual collectors listed, and and 4-merous calyx, but probably would repre­ especially Paul Berry and Kay Yatskievych for sent a distinct species. These have wider leaves their untiring devotion to the completion of the than many specimens of Plinia rivularis, but Flora of the Venezuelan Guayana. otherwise there is no difference. See Rotman (1985) for a complete list of syn­ LITERATURE CITED onomy, and Rotman (1982) for additional material Amshoff, J.H. 1942. Notes on the Myrtaceae of Suri­ cited of this species from Argentina and Uruguay. name. Rec. Trav. Bot. Need. 39: 147-165. Representative material examined. BRAZIL. ---. 1948. Myrtaceae. In Maguire et a1: Guiana Acre: Estrada Rio Branco/Quixada, km II, Plants. Bull. Torrey Bot. Club 75: 528-538. aprox. lOoS, 67°50'W, 18 Oct. 1980 (bud), Nel­ ---. 1950. Notes on Guiana Myrtaceae III. Rec. son 731 (F, NY, US); (imm. fr), Nelson 733 Trav. Bot. NeerL 42: 1-27. (MO, NY, US). Amapa: Perminentral Norte, ---. 1951. Myrtaceae. In A. Pulle, ed. Flora of Surinam 3(2): 56-158. Munguba, 23 Jan. 1978 (fl), Santos 213 (NY). Berg, O. 1855-1856. Revisio Myrtacearum Americae. Amazonas: 0-3 km do Km 211 da estrada Per­ Linnaea 27: 1-472. imetral Norte, encosta do Pico Rondon, 01032'N, ---. 1857-1859. Myrtaceae. In c.F. P. von Martius, 62°48'W, 4 Feb. 1984 (ft), Amaral 1479 (NY); FL Bras. 14(1): 1-655. Pico Rondon, 0-3 km north of Km 211 of Per­ ---. 1861. Mantissa II. Ad Revisionem Myrta­ imetral N Hwy. 01°32'N, 62°48'W, 25 Mar. 1984 . cearum Americae. Linnaea 30: 647-713. (fr), Pipoly et al. 6905 (NY); Lower slopes of Laridrum, L. 1981. Campomanesia, Pimenta, Blephar­ Pico Rondon, 3 km from Km 211, Qlo32'N, . ocaiyx, Legrandia, Acca, Myrrhinium, and Luma (Myrtaceae). FL Neotrop. 45: 1-178. 62°48'W, 2 Feb. 1984 (ft), Prance et al. 28755 ---. 1986. A monograph of the genus Myrceugen­ (NY). Maranhao: Rio Alto Turias;u, Nova Es­ ia (Myrtaceae). FL Neotrop. 29: 1-137. perans;a, 0-100 m, 02°55'S, 45°45'W, 7 Dec. Legrand, C.D. 1958. Las Especies Tropica1es del Ge­ 1978 (bud), Jangoux & Bahia 292 (NY); Alzi­ nero Gomidesia. Com. Bot. Mus. Hist. Nat. Mon­ landia, Rio Pindare, 0-100 m, 03°45'S, 46°05'W, tevideo 3(37): 1-30. 13 Dec. 1978 (fr), Jangoux & Bahia 477 (NY). Mc Vaugh, R. 1958a. Myrtaceae. In Flora of Peru. Field Parana: Mun. Cianorte, Fda. Lagoa, 24 Aug. Mus. Nat. Hist., Bot. Ser. 13(4.2): 561-818. 1967 (ft), Hatschbach 16955 (MO); Munic. --- 1958b. Myrtaceae (Calyptranthes and Marli­ erea). Mem. New York Bot. Gard. 10(1): 61-91. Guarapuava, Passo do Jacu, 18 Oct. 1984 (ft), ---. 1968. The genera of American Myrtaceae­ Hatschbach 32717 (MO). PARAGUAY. Dept. Par­ an interim report. Taxon 17: 354-418. aguari, Parque Nacional Ybycu'i, 15 Aug. 1984 ---. 1969. The Botany of the Guayana High1and­ (ft), Perez & Dun! 272 (MO). TRINIDAD. Marac part VIII. Myrtaceae. Mem. New York Bot. Gard. Road, S. Watershed Reserve, 26 Jun. 1929 (imm 18(2): 55-286. fr), R.O. Williams 12167 (NY). VENEZUELA. Rotman, A. 1982. Los generos Calycorectes, Hexachla­ Amazonas: Dept. Atures, 25-30 km SE of Pto. mys, Myrciaria, Paramyrciaria, Plinia y Siphoneu­ Ayacucho, 100 m, 12 May 1989 (ft), Steyermark gena en 1a flora argentina. Darwiniana 24: 157-185. ---. 1985. Nota Sobre un Epiteto Especifico en e1 et al. 122300 (ICN, MO, VEN); Esmeralda, Up­ Genero Plinia (Myrtaceae). Bol. Soc. Argentina per Orinoco, 143 m, l3 May 1942 (imm fr), Lt. Bot. 24(1-2): 196. Williams 15306 (F, ICN, SEL, US, VEN); 15 Urban, I. 1895. Additamenta ad cognitionem florae Indiae May 1942 (imm fr), Ll. Williams 15356 (F, MO, occidentalis. Particula II. Bot. Jahrb. 19: 562-681.