SystematicBotany (1984), 9(2): pp. 229-235 ? Copyright1984 by the American Society of Taxonomists ,a New of fromSouth America

GRADY L. WEBSTER Botany,University of California,Davis, California 95616

ABSTRACT.A new monotypicgenus, Jablonskia, based on the South American Securinegacon- gesta (Euphorbiaceae, Phyllanthoideae), resembles genera of the in pollen and wood characters,but the floralstructure is more similar to that in genera of Aporuseae such as Ashtonia and .From ,where it was originally placed, and the taxa of Antidesmeae and Aporuseae, Jablonskiamay be separated by a combination of characteristics:monoecious; inflores- cences axillary; flowersconspicuously bracteate,the staminatesessile; pollen grains prolate, exine tectate-perforate,germ pore lalongate; seeds with fleshy exotesta, paired in each locule of the irregularlydehiscing capsule; embryowith radicle equalling the cotyledons.

Among the taxa of Euphorbiaceae that were Muell. Arg. was originallydescribed as a never revised by Pax and Hoffmann in their of Phyllanthusby Mueller (1863) and later long series of monographictreatments is Phyl- transferredto Securinegaby Mueller (1873). lanthinae. As circumscribedby Pax and Hoff- Mueller referredthis species to his sect. Secu- mann (1931), this subtribe included five gen- rinegastrum[=sect. Securinega],which had pre- era: Zimmermannia,Securinega, Pleiostemon, viously included only the species from ,and Reverchonia.Relationships in Mauritius, S. durissimaJ. F. Gmelin. Although this complex have remained rather poorly Bentham (1878, 1880) concurredwith its place- understood because of the lack of comprehen- ment in Securinega,the South American species sive study since the last, now badly outdated, is clearly anomalous because of its monoecy, treatments of Mueller (1866) and Bentham conspicuouslybracteate flowers, and seeds with (1880). In an attemptto improve the classifica- fleshyexotesta. tion of this group, I have considered problems The discordant position of S. congestais also of generic delimitation in studies on Phyllan- apparent in its pollen characters.In his survey thus(1956-58), Reverchonia(1963 with Miller), of Euphorbiaceous pollen, Punt (1962) indicat- (1965), and (1979). ed that the pollen of S. congestadiffers from The mostchallenging problem of generic de- that of other species of Securinegain its prolate limitationinvolves SecurinegaJuss. Although shape and elongated colpus transversalis.Al- Pax and Hoffmann(1931) treated Securinegain though Kbhler (1965) did not reporton the pol- a broad sense, including FlueggeaWilld., Ben- len of S. congestain his pollen survey of 'bio- tham (1878) characterizedit as a "purely arti- vulate' Euphorbiaceae, it appears to fitinto his ficial genus" and maintained as dis- 'Antidesmatype' because of the prolate grains tincton the basis of its fruitand seed structure. with long narrow colpi and tectateexine with Aftera study of this problem over the last 15 small lumina. This may be seen in the figures years,I have concluded that Bentham was cor- included here, which show that the pollen of rect in separating Flueggeaand that the artifi- S. congesta(fig. 1) is much more similar to that ciality of delimitation of Securinegashould be of Hyeronimaalchorneoides (fig. 2) and Celianella resolved by segregatingthose elements whose montana(fig. 3) in Antidesmeae than it is to the diagnostic featuresare incompatible with the pollen of Richeriagrandis (fig. 4) and Ashtonia type species, S. durissimaJ. F. Gmelin. Part of praeterita(fig. 5) in Aporuseae or Securinegadur- this task was accomplished with the reinstate- issimain (fig. 6). The tectate-per- ment of Meineckia(Webster 1965), but a num- forate exine of S. congestapollen is strikingly ber of discordantgroups remain,including the similarto thatin pollen of Euphorbioideae and South American that are the subject of is unique in Phyllantheae; however, it appears this paper. to be derived fromthe finelyreticulate exine The South American taxon currentlyknown of genera such as Celianella(fig. 3). as Securinegacongesta (Benth. ex Muell. Arg.) The indications of affinitysuggested by pol- 229

This content downloaded from 169.237.8.35 on Mon, 16 Sep 2013 18:29:31 PM All use subject to JSTOR Terms and Conditions 230 SYSTEMATICBOTANY [Volume 9

5~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~9 6 - ';,

FIGS.1-6. Pollengrains of somegenera of Euphorbiaceae-Phyllanthoideae;scales =10 sm. 1. Jablonskia congesta(Rimachi 4838, DAy). 2. Hyeronimaalchorneoides (Alves Lages RB80704,DAy). 3. Ceizanellamontana (Maguireand Maguire35071, NY). 4. Richeriagrandis (Webster 13477, DAy). 5. Ashtoniapraeterita (Soepadmo 783, K). 6. Securinegadurissima (For. Dept. 11399,DAy).

This content downloaded from 169.237.8.35 on Mon, 16 Sep 2013 18:29:31 PM All use subject to JSTOR Terms and Conditions 1984] WEBSTER: JABLONSKIA 231 len morphology are consistentwith the avail- able data for wood anatomy. In the brief,and only available, previously published descrip- tion of the wood of S. congesta,Record (1938) mentioned that it differsfrom the wood of Se- curinegaacidoton by having vessel elementswith scalariformperforations, thin-walled wood fi- bers, and reticulatexylem parenchyma. In the classification of Phyllanthoidean woods by Mennega (1984; ined.), S. congestaappears to fit best into group A2, which includes Aporusaand other genera of Aporuseae (as delimited by Webster 1975). Mennega (pers. comm.) notes, however, that genera of Aporuseae such as Aporusa,Ashtonia, and Richeriadiffer from S. congesta in having diffuse parenchyma and longer non-septatefibers. In gross morphologicalcharacters, S. congesta has much in common with the South American

Richeria,both in general habit (fig. 7) and in lcm floral details (figs. 8-10). In the treatmentof Phyllanthoideae by Pax and Hoffmann(1922), it keys to Richeriaif treated as a member of Antidesminae.However, S. congestadiffers from Richeriain its monoecy,flowers in axillaryclus- ters (rather than spikes), thin-walled capsules with paired seeds in each locule, and embryo with elongated radicle (about equalling the cotyledons).In some respects,such as the thin- walled capsules and fleshy seeds, S. congesta FIG. 7. Branchof Jablonskiacongesta with imma- more closely resemblesthe Old World Aporusa. ture fruits(Revilla 22, DAV). However, Aporusais dioecious; the staminate flowers usually have only two or three sta- mens, a reduced pistillode, and no disk; the pistillate flowerslack a disk and usually have S. congestaand those of ,which is also two carpels with fimbriatestyles; and the fruits monoecious and has pentamerous flowers. have at most one seed per locule with cotyle- However, Amanoa differsin a number of sig- dons much longer than the radicle. Further- nificantcharacters, particularly the flowerswith more, S. congestadiffers from both Richeriaand well-developed petals, sessile stigmas, drupa- Aporusain pollen and wood characters. ceous fruits,and seeds with endosperm scanty More similar to S. congestain some respects or absent. Furthermore,the pollen of Amanoa is the recentlydescribed Malesian Ashtonia(Airy is very different,with a coarsely reticulate or Shaw 1968, 1972, 1974),which has pentandrous bizarrely ornamented exine, as illustrated by staminateflowers and capsular fruitwith fleshy Punt (1962) and Kohler (1965). seeds. However, the flowersof Ashtonialack a Mention should also be made of Celianella, well-developed floraldisk; the staminateinflo- recentlydescribed fromthe Guyana highlands rescences are racemose; the sepals are three or by Jablonski (1965). The pollen of Celianella four,not five; the pollen grain exine (fig. 5) is montanaJabl. (fig. 3), although finelyreticulate reticulate; and the styles are undivided and ratherthan tectate-perforate,is somewhat sim- stigmatiform.None of the other genera in An- ilar to the pollen of S. congesta(fig. 1) and of tidesmeae or Aporuseae (sensu Webster 1975) Hyeronima(fig. 2) and other genera of Antides- appears to be more similar.There is some sim- meae. Furthermore,in the accompanying pa- ilaritybetween the bracteateflower clusters of per Mennega (1984) notes a resemblance be-

This content downloaded from 169.237.8.35 on Mon, 16 Sep 2013 18:29:31 PM All use subject to JSTOR Terms and Conditions 232 SYSTEMATIC BOTANY [Volume 9

8 ~~~~~~~~~~~~~~~~~~10, IAp"

FIGS. 8-10. Flowers of Jablonskia.8-9. Staminate flower (one sepal removed in fig. 9); Rimachi4838 (DAV). 10. Pistillateflower (2 sepals cut away; gynoecium removed to show disk); Revilla22 (DAV).

tween Celianellaand Securinegacongesta in wood curinegas.str. differt foliis glanduligeris,plantis anatomical characteristics.However, Celianella monoicis,floribus bracteatis, calyce glabro,pol- differsfrom S. congestain having racemose ped- linis granulis prolatis,perforato-tectatis, semi- icellate staminate flowerswith pendulous an- nibus carnosulis. thersthat are rathertypical for Antidesmeae. It Monoecious glabrous or . Leaves also differsin the stronglyaccrescent pistillate alternate,distichous, simple, entire, pinnately calyx and in the seeds, which are apparently veined, short-petiolate, with small laminar solitaryin each locule of the capsule and have glands near base; stipules elongate, caducous. an embryowith the radicle much shorterthan Flowers apetalous, sessile or subsessile in axil- the cotyledons (in this respect fittingAporu- larybracteate glomerules. Staminate flower: se- seae better than Antidesmeae). These differ- pals 5, imbricate;disk of 5 segments; stamens ences appear to rule out any close relationship 5, free; filamentselongated (equalling or ex- between Celianellaand S. congesta. ceeding the calyx); anthers versatile, introrse, In view of the floral,anatomical, and paly- anther sacs not pendulous; pollen grains pro- nological distinctivenessof the South Ameri- late, 3-colporate, germ pore lalongate, exine can plant treatedas Securinegacongesta, it is best tectate-perforate;pistillode shorterthan calyx. treatedby assignmentto a separate new genus. Pistillateflower: sepals 5, imbricate,persistent; It is appropriate to dedicate this genus to Eu- disk patelliform;carpels 3; styles free, bifid, gene Jablonski (1892-1975), monographer of erect; ovules paired in each cell, anatropous. Euphorbiaceae and studentin his finalyears of Fruit a thin-walled baccate capsule, dehiscing the neotropical Euphorbiaceous florato which irregularly; columella subpersistent; seeds Jablonskiabelongs (see Barneby 1975, for bio- paired in each locule; outer seed-coat thin and graphical details). fleshy,inner seed-coat crustaceous;endosperm copious; embryo without chlorophyll, cotyle- jablonskia Webster, gen. nov.-TYPE: Jablon- dons broader than and about equal in length skia congesta(Benth. ex Muell. Arg.) Web- to the radicle. ster Securinegacongesta (Benth. ex Muell. A monotypic genus including only the fol- Arg.) Muell. Arg. Phyllanthuscongestus lowing species: Benth. ex Muell. Arg. Subfam. Phyllanthoidearum,Richeria simile, Jablonskia congesta (Benth. ex Muell. Arg.) differtfoliis glanduligeris, plantis monoicis, Webster,comb. nov.-Phyllanthuscongestus floribusbracteatis, calyce glabro, pollinis gran- Benth. ex Muell. Arg., Linnaea 32:25. ulis prolatis, seminibus carnosulis in quoque 1863.-Securinegacongesta (Benth. ex Muell. loculo binis, radicula embryoniselongata; a Se- Arg.) Muell. Arg., Fl. Brasil. 11(2):76, t. 11,

This content downloaded from 169.237.8.35 on Mon, 16 Sep 2013 18:29:31 PM All use subject to JSTOR Terms and Conditions 1984] WEBSTER: JABLONSKIA 233

fig.I. 1873.-LECTOTYPE(here chosen): Bra- sistent,2.5-3.5 mm long, 0.5-0.6 mm thick at zil, Amazonas, along the Rio Negro be- base, slightly thickened apically; seeds plano- tween Barra[Manaos] and Barcellos,Spruce convex, with outer coat thinly fleshy (thicker 1900 (microficheof specimen at G!; isotype: near micropylarend), yellowish, smooth, 2.6- K!). 3 mm long, 1.8-2.2 mm broad. Glabrous or 1-8 m high, typically Distribution.Lowland rainforests,below 200 sparsely ramifiedwith long slender branches; m, in riverine thicketsor varzea forest,Vene- twigsbrown, terete below, distallyangled with zuela and Guyana to Amazonian Brazil and Peru narrow ribs decurrentfrom the nodes, sparsely (fig. 11). Flowering Jan-May,apparently fruit- lenticellate; foliage evergreen. Leaf blades ing throughout the year. However, the sam- chartaceous, elliptic-lanceolate, mostly 7-12 pling over the large area of distributionis in- (-14) cm long, 2-4.5 cm broad, somewhat de- adequate to give an accurate picture of any currenton the petiole; blade surfaceabove oli- seasonal variation in flowering. vaceous, beneath paler, with minute pigment- Specimens examined.Guyana. Schomburgk372 ed streakson midrib and irregularblotches on (K, U, US); Berbice, 1837, Schomburgk(K). Sur- lamina; foliar glands imbedded, visible be- iname. Marowyne River, Versteeg3026 (K, U). neath,paired (one on eitherside of midribnear Venezuela. Amazonas: Brazo Casiquiare, Holt base), 0.3-0.7 mm long; veins mostly 12-15 on and Blake656 (US); Alto Casiquiare, Capihuara, a side, slightly raised above and more promi- Williams15687 (VEN). Brazil. Amazonas: Rio nulous beneath, rather tenuous, veinlet retic- Vaupes, Panure, Spruce 2637 (K); Esperanta, ulum fine and open; margins plane, unthick- Igarape Umarizal, boca do Javari,Ducke 1631 ened, bordered by lines of pigmented cells; (K, NY, RB, UC, US); Esperansa, Benjamin Con- petiole subterete, (1-)2-4 mm long; stipules stant,Murqa Piresand Black 1031 (U, VEN), Rio narrowly lanceolate, attenuate-acuminate,ca- Javary,Traill 723 (K); Rio Negro, Tapuruquara, ducous, (3-)7-13 mm long. Monoecious; inflo- Murqa Pires304 (VEN); Rio Negro, Isla Maraca, rescences axillary, glomerulate, bracteate; cerca de la confluencia del Rio Padauiri, coll. glomerules unisexual or bisexual, flowerssub- anon. (VEN); Rio Negro, opposite Rio Branco, tended by persistententire lanceolate bracts1- Spruce1900 (K); Rio Negro, Barcellos, Black48- 2 mm long; staminateflowers 5-10 per cluster, 2432 (NY, U, VEN), Froes 22071 (U). Roraima: pistillate 1-5. Staminate flowerssessile; sepals Rio Branco, Kuhlmann3304 (RB); Rio Uananau erect,elliptic-oblong, ca. 2 mm long and 1 mm [=Anaua], Spruce1990 (K). Para: Bella Vista, R. broad, obtuse, entire, cells bullate and darkly Tapajoz, Ducke 10559 (RB, U); Boa Esperansa, pigmented on abaxial surface, venation ob- abaixo do Cachoeira Capoeira, Murqa Pires3650 scure; disk-segmentselliptic, flattened,angu- (NY). Peru. Loreto: Maynas, Florida, Rio Putu- lar, 0.2-0.3 mm across; filamentsslender, 1.5-2 may6, at mouth of R. Zubineta [=Yubineto], mm long, hirsutulous;anthers 0.4-0.5 mm long, Klug 227,4(K, NY, US); Rio Ampiyacu, Revilla connective not enlarged; anthersacs dehiscing 1000 (DAV); Rio Momon, Rimachi4838 (DAV); longitudinally, not apically confluent; pollen Rio Napo, Mazan, Mexia 6452 (K, MICH, NY, grains ca. 30 Amlong, 22.5 Ambroad; pistillode U, UC, US); Padre Isla, frentea Iquitos, Ferreyra cylindrical, apically lobed, striate with pig- 3397 (US), Revilla22 (DAV); Moena Cano, Gen- mented cells, 0.8-1 mm high. Pistillate flowers tryet al. 15644 (DAV); Rio Itaya, Gentryet al. subsessile, pedicel becoming 1-2 mm long in 81402 (Dav, MO), Williams82 (US); Iquitos, As- fruit;sepals erect with recurving tips, obtuse plund 14357 (US), Killip and Smith27394 (US). or minutely apiculate, entire, coriaceous, per- Loreto: Rio Marafion,near mouth of Rio Tigre, sistent in fruitand becoming 3.2-4 mm long, Killipand Smith27524 (NY, US). 1.2-2.2 mm broad; disk tenuous, entire,ca. 1.2 Until a more thorough morphological and mm across; ovarysmooth, carpels 3 (very rarely anatomical surveyof Phyllanthoideae has been 4); styles free,erect, moderately thick, ca. 0.8- done, the tribal position of Jablonskiaappears 1.1 mm high, bifidat tips; style-branchesblunt, uncertain, because it does not clearly fit into ca. 0.4-0.5 mm long. any of the tribes previously recognized (Web- Capsules oblate, shallowly 3-lobed, 5-6.5 mm ster 1975). In its glandular leaves, fleshyseeds, in diameter,thin-walled (walls ca. 0.1-0.15 mm and overall floral morphology,it shows simi- thick); columella deciduous or sometimes per- larities to Aporuseae, but it differsin its wood

This content downloaded from 169.237.8.35 on Mon, 16 Sep 2013 18:29:31 PM All use subject to JSTOR Terms and Conditions 234 SYSTEMATIC BOTANY [Volume 9

0

/ ~~~~~~~~~0

10N

I ~0,-~~~~~~~~~ 200 400 600 800 1000km

v --

21L 1A~~~~~~~~~~~-

lr etoic eis2 ubihdb et FI. 1 Ditiuino(Jbosi ogst.Bs a ro

Syt o.,StaeUi.. Urct

structure (septate fibers, absence of paren- as a peripheral member of Phyllantheae, Se- chyma),perprolate pollen grains,monoecy, and curineginae. embryowith radicle as long as the cotyledons. ACKNOWLEDGMENTS. This studywas supported in in pollen form, Jablonskia Anatomically and partby NSF GrantDEB-7724263. Floral drawingswere agrees much more closely with taxa of Anti- provided by Ms. Dana Hawley. Scanning micrograph desmeae, but it does not show any close re- photographs of pollen were taken by Dr. Steven P. semblance to any particulargenus of thattribe, Lynch, Mr. Michael Klungness, and Ms. Lynn Gil- and it differsfrom the South American Celi- lespie. Assistance in preparing the manuscriptwas anella by its reduced bracteate inflorescences given by Dr. Daryl Koutnik, Ms. Mary Barbercheck, borne monoeciously.There are also similarities Ms. Lynn Gillespie,and Mr. Norman Deesing. I thank between Jablonskiaand genera of Phyllantheae, Dr. JulianSteyermark (Instituto Botanico, Caracas) and where it was originally assigned. In its sessile Dr. A. M. W. Mennega (Instituutvoor Systematische for informationand staminateflowers with introrsestamens, it re- Plantkunde, Utrecht) sugges- tions and the curatorsof the herbaria cited forloans sembles Securinegas.str.; however, both the of specimens. pollen and wood structureof the latterseem to preclude any close affinity.As pointed out by Dr. Mennega (1984), Chascothecais more simi- LITERATURECITED lar anatomically,but it also has differentpollen grainsand divergentreproductive features (e.g., AIRYSHAW, H. K. 1968. Notes on Malesian and oth- dioecy, annular staminate disk, solitary seeds er Asiatic Euphorbiaceae LXXXV. A new genus with dry exotesta). fromBorneo. Kew Bull. 21:357-360. Overall,-it appears that Jablonskiais a rela- 1972. Notes on Malesian and other Asiatic tively isolated genus of uncertain systematic Euphorbiaceae CL. A new Ashtoniafrom Malaya. Kew Bull. 27:4-5. positionwith varyingdegrees of affinityto An- . 1974. Ashtoniapraeterita. Hooker's Icon. P1. tidesmeae, Aporuseae, and Phyllantheae. Per- 38:t. 3702. haps the least objectionableassignment that can BARNEBY,R. 1975. [note on] Eugene Jablonski.Tax- be made at presentis to regard it provisionally on 24:580.

This content downloaded from 169.237.8.35 on Mon, 16 Sep 2013 18:29:31 PM All use subject to JSTOR Terms and Conditions 1984] WEBSTER: JABLONSKIA 235

BENTHAM,G. 1878. Notes on Euphorbiaceae. J.Linn. Phyllantheae.Pp. 1-349 in Das Pflanzenreich,Heft Soc. Bot. 17:185-267. 81, ed. A. Engler. Berlin: W. Engelmann. . 1880. Euphorbiaceae. Pp. 239-349 in Genera and . 1931. Euphorbiaceae. Pp. 11-233 plantarum,vol. 3, by G. Bentham and J.D. Hook- in Die NathrlichenPflanzenfamilien, 2nd ed., vol. er. London: L. Reeve and Co. 19c, ed. H. Harms. Leipzig: W. Engelmann. JABLONSKI,E. 1965. Euphorbiaceae. [The botany of PUNT, W. 1962. Pollen morphology of the Euphor- the Guyana highland VI, by B. Maguire et al.] biaceae with special reference to . Mem. New York Bot. Gard. 12(3):150-178. Wentia 7:1-116. KOHLER, E. 1965. Die Pollenmorphologie der bio- RECORD, S. J. 1938. The Americanwoods of the fam- vulaten Euphorbiaceae und ihre Bedeutung fur ily Euphorbiaceae. Trop. Woods 54:7-40. die Taxonomie. Grana Palynol. 6:26-120. WEBSTER, G. L. 1956-58. A monographic study of MENNEGA, A. M. W. 1984. Wood structureof Jablon- the West Indian species of Phyllanthus.J. Arnold skiacongesta (Euphorbiaceae). Syst. Bot. 9:236-239. Arbor. 37:91-122, 217-268, 340-359; 38:51-80, MUELLER, J. 1863. Euphorbiaceae. VorlauifigeMit- 170-198, 295-373; 39:49-100, 111-212. teilungen aus dem furde Candolle's Prodromus 1965. A revision of the genus Meineckia(Eu- bestimmtenManuscript. Linnaea 32:1-126. phorbiaceae). Acta Bot. Neerl. 14:323-365. . 1866. Euphorbiaceae. Pp. 189-1261 in Pro- . 1975. Conspectus of a new classificationof dromussystematis naturalis . .., vol. 15(2), ed. A. the Euphorbiaceae. Taxon 24:593-601. de Candolle. Paris: Treutteland Wiirtz. . 1979. A revision of Margaritaria(Euphorbi- . 1873. Euphorbiaceae [Pars I]. Pp. 1-292 in aceae). J.Arnold Arbor. 60:403-444. Flora Brasiliensis,vol. 11(2), ed. A. W. Eichler. and K. I. MILLER. 1963. The genus Rever- Leipzig: F. Fleischer. chonia(Euphorbiaceae). Rhodora 65:193-207. PAX, F. and K. HOFFMANN. 1922. Euphorbiaceae-

This content downloaded from 169.237.8.35 on Mon, 16 Sep 2013 18:29:31 PM All use subject to JSTOR Terms and Conditions