Aristolochia gorgona (Aristolochiaceae), a new species with giant ¯owers from Costa Rica and Panama

MARIO A. BLANCO1

Blanco, M. A. (Instituto Centroamericano de InvestigacioÂn BioloÂgica y Con- servacioÂn, P.O. Box 2398-250, San Pedro de Montes de Oca, San JoseÂ, Costa Rica; and JardõÂn BotaÂnico Lankester, Universidad de Costa Rica, 1000 San JoseÂ, Costa Rica). Aristolochia gorgona (Aristolochiaceae), a new species with giant ¯owers from Costa Rica and Panama. Brittonia 54: 30±39. 2002.ÐAristolochia gorgona is described from the Atlantic watershed of Costa Rica and central and eastern Panama. It is similar to A. grandi¯ora Sw., with which it has been con- fused in the past. The extensive reported synonymy of A. grandi¯ora is reviewed and compared with the new entity. Some new terms are proposed for better de- scribing ¯oral structure in this species complex, and suggestions are made for its collection and study. Aristolochia gorgona is one of the largest-¯owered plant species (in terms of perianth area) in the Neotropics. Key words: Aristolochia, Aristolochiaceae, collecting techniques, Costa Rica, ¯ower structure, Panama.

Blanco, M. A. (Instituto Centroamericano de InvestigacioÂn BioloÂgica y Con- servacioÂn, P.O. Box 2398-250, San Pedro de Montes de Oca, San JoseÂ, Costa Rica; y JardõÂn BotaÂnico Lankester, Universidad de Costa Rica, 1000 San JoseÂ, Costa Rica). Aristolochia gorgona (Aristolochiaceae), a new species with giant ¯owers from Costa Rica and Panama. Brittonia 54: 30±39. 2002.ÐAristolochia gorgona se describe de la vertiente AtlaÂntica de Costa Rica y del centro y este de PanamaÂ. Es similar a A. grandi¯ora Sw., con la cual se ha confundido en el pasado. La extensa sinonimia reportada para A. grandi¯ora es revisada, y com- parada con la nueva entidad. Se proponen algunos teÂrminos nuevos para describir mejor la estructura ¯oral en eÂste complejo de especies, y se ofrecen sugerencias para su recoleccioÂn y estudio. Aristolochia gorgona es una de las especies ve- getales con ¯ores maÂs grandes (en cuanto al aÂrea del perianto) en el NeotroÂpico.

Aristolochia grandi¯ora was described by caped from cultivation. It is considered the Swartz (1788), based on a specimen from largest-¯owered plant in the Neotropics Jamaica. As presently conceived by most (Standley & Steyermark, 1946; Barringer, taxonomists, the species is common but var- 1983a). Similar large-¯owered aristolochias iable in ¯ower size and shape, ranging from have been described (see synoymy in Gon- southern Mexico to Ecuador and the West zaÂlez, 1990), but all are considered conspe- Indies. It has become naturalized in Java ci®c with A. grandi¯ora by recent monog- (Cammerloher, 1923), Sri Lanka (Petch, raphers (e.g., GonzaÂlez, 1990, 1994). A 1924), and Florida (Wunderlin, 1998), es- striking new species with giant ¯owers, closely related to A. grandi¯ora, is here de- 1 Current mailing address: University of Florida, De- scribed as new. Its ¯owers are distinctive in partment of Botany, 220 Bartram Hall, P.O. Box living specimens, but herbarium specimens 118526, Gainesville, Florida 32611-8526 U.S.A. have been confused with the latter species.

Brittonia, 54(1), 2002, pp. 30-39. ISSUED: 29 May 2002 ᭧ 2002, by The New York Botanical Garden Press, Bronx, NY 10458-5126 U.S.A. 2002] BLANCO: ARISTOLOCHIACEAE 31

In the following description and discus- popodium and prophyll not atrophied in sion, the terminology for ¯oral parts fol- branches (pseudostipules absent). Leaves lows Pfeifer (1966) and GonzaÂlez (1990, alternate, distichous, simple; petiole 2.5±10 1994). Due to the higher structural com- cm long; blade triangular-cordate, 5±27 cm plexity of the calyx in this and related spe- long (from tip of basal lobes to apex), 5± cies (compared to other aristolochias), how- 21 cm wide, cordate basally with a wide- ever, the following terms are newly coined open sinus 2±6 cm deep, and then attenu- or modi®ed from Cammerloher (1923) to ated into petiole, entire along margin, acute simplify description. These terms might to acuminate apically, glabrous adaxially, also prove useful in future pollination stud- glabrous to hispidulous abaxially; venation ies. The ``vestibule'' refers to the chamber palactinodromous with 3 main veins di- formed by the distalmost part of the tube. verging from the base, the main lateral Cammerloher's Hintergrund, or ``back- veins curving into the basal lobes, with 4± ground,'' refers to the internal wall of the 5 main branches each; tertiary veins more vestibule. The ``gullet'' (Cammerloher's or less perpendicular to the primaries and ``Reuse'') refers to the curved, proximal secondaries; veins impressed adaxially, part of the tube, in between the vestibule prominent abaxially. In¯orescences axil- and the syrinx. These two parts of the tube lary from young stems, single-¯owered, are separated by a short, narrow passage pendent; peduncle 6±10 cm long, bracteo- (the ``bottleneck''), and differ in internal late, green; bracteole suborbicular, perfoli- texture, indument, and coloration, even ate, with an acute apex, to 2 cm wide, when they are similar in appearance exter- green. Pedicel plus ovary to 8 cm long, nally. The ``nectary'' refers to a wide, glan- straight, ribbed, reddish. Flower with a dular zone located proximally on the dorsal, strong putrid smell. Calyx complexly three- internal surface of the utricle (described for dimensional in structure; utricle pyriform, A. grandi¯ora by Cammerloher, 1923); the gibbous, 14±19 ϫ 5.5±10 cm, with 6 prom- utricle wall is thicker in the zone of the nec- inent external ribs, cream-colored with pur- tary. In A. gorgona, the limb folds in such plish ribs and reticulated venation external- a way that it presents two faces: a ``frontal ly, white with cottony hairs and a some face'' and a ``lower face'' (see Fig. 1D and bare, purple spots near the base internally. E, respectively). Nectary oval, 3.2 ϫ 1.4 cm, orange-brown with a purple band distally. Syrinx cylin- Aristolochia gorgona M. A. Blanco, sp. drical, to 3.5 cm long, 1.4±1.7 cm diam. at nov. (Fig. 1) the opening, directed obliquely into the utri- cle, cream with a purple rim. Tube sigmoid, TYPE: COSTA RICA. Heredia: Puerto divided into a ``gullet'' and a ``vestibule,'' Viejo de SarapiquõÂ, EstacioÂn BioloÂgica La 12.5 cm long, same color as utricle exter- Selva, Quebrada SuraÂ, near the Laboratory, nally. Gullet U-shaped, 7.5 cm long, 2.8 cm 10Њ26ЈN, 83Њ59ЈW, 50 m, 4 Jan 2001 (¯, wide at the bottom; white internally. Bot- bud), M. Blanco 1752 (HOLOTYPE: USJ, in- tleneck 2 mm wide (in longitudinal section) cluding material in spirit; ISOTYPES: CR, in one measured ®rst-day ¯ower. Vestibule MO, NY). a wide, gibbous, more or less transversely bilobed chamber, 5 cm long, 7.5 cm wide, Aristolochiae grandi¯orae Sw. af®nis, sed ¯oribus 2.8 cm deep at the middle; internal wall majoribus, limbus calyce plicatus frontalis atque infra cream-colored, heavily blotched with trans- tubus calyce, projecturae parvulae vermiformis interius obtecta, cauda minore. verse dark maroon bands that become solid in the distal third, covered by downward- Herbaceous vine, becoming semi- pointing cream hairs. Annulus thin, sharp- woody toward base with age, but never tru- edged, to 1.5 cm high, dark orange-brown, ly woody; peridermis not corky; stem and opening (fauces) 6.4 cm high (longitudinal- leaves aromatic. Stem twining, ribbed, 3±4 ly), facing obliquely downward. Limb mm diam. when young, reaching 1.5 cm abruptly expanding from the annulus, mas- diam. at the base of the plant, glabrous; hy- sive, the distal half folding under the tube, 32BRITTONIA [VOL. 54

FIG.1. Aristolochia gorgona (Blanco 1752, USJ). A. Stem and leaves. B±F. Flower (drawn from photographs of fresh ¯owers). B. Side view. C. Oblique view from back. D. Front view. E. View from below; front of ¯ower toward the top. F. Longitudinal section. G. Detail of limb margin and tentacles. H, I. Gynostemium. H. Lateral view. I. Top view. 2002] BLANCO: ARISTOLOCHIACEAE 33 thus presenting two faces, cordiform with form a corky bark, contrary to the asser- two wide upper lobes, 25±31 cm long, 25± tions of Barringer (1983a) and Cook 30 cm wide when held ¯at, marginally (2001). Some branches produced low in the fringed, the fringes ¯attened, cirrhose, 3±10 main stem grow toward the ground and be- mm long; apically obtuse with a terminal come runners that produce adventitious cauda (appendix) 5 cm long, 2 mm wide; roots at each node. These resume the climb- cream suffused with purple externally, the ing mode many meters away from the base internal surface cream with clear orange- of the plant, establishing new clumps by brown blotches (almost solid in some ¯ow- vegetative reproduction. ers), darker brown around the opening, vel- Flowering occurs throughout the year. vety in texture, the whole surface densely Flowers are normally produced 3±15 m covered by small vermiform projections from the ground. Numerous muscid and (tentacles), these cylindrical, minutely pa- phorid ¯ies (as well as staphylinid beetles) pillose, clear orange-brown with a dark have been found trapped inside the utricles brown apex, 5±20 mm long, 1±2 mm thick of ®rst-day ¯owers. Similar ¯ies were (0.5±1 mm thick and totally black when found by both Cammerloher (1923) and dry). Frontal limb face directed obliquely Hilje (1984) in Aristolochia grandi¯ora, upward with re¯exed sides, 15±22 cm long, suggesting that there is a potential for hy- 15±18 cm wide (when sides re¯exed); low- bridization if both species occurred togeth- er limb face almost horizontal, with sides er. However, available herbarium specimens and apex re¯exed, 24 cm long, 13 cm wide do not indicate sympatry (in Costa Rica, A. (when sides and apex re¯exed); appendix grandi¯ora is mostly found on the Paci®c directed upwards and variously twisted. lowlands). Gynostemium coroniform, sessile, radially Fruits have been collected in March, symmetrical, 13 mm high, 7 mm wide at April, and July, but undoubtedly occur the apex, cream-colored with purple at the year-round as well. As in A. grandi¯ora, the base; anthers 6, yellow; stigmatic surface capsules are completely dehiscent (the car- white. Fruit a 6-carpelate, many-seeded pels do not remain united at the apex in the capsule, cylindric-hexagonal with a blunt characteristic ``inverted parachute'' fashion apex, 6±7 cm long (excluding pedicel), 3.5 of most wind-dispersed aristolochias). The cm wide, septicidally dehiscent, the carpels seeds are thick and lack wings; their ability separating completely at maturity; green to ¯oat for long periods and the riparian with purplish longitudinal ribs when devel- distribution of the plants suggest water dis- oping, dry at maturity. Seeds obovoid-del- persal. toid, 13±18 mm long, 11±17 mm wide, 2± Etymology. Named after the Gorgons, 2.5 mm thick, seed coat suberized, brown. the three monstrous sisters of Greek my- Distribution and ecology. Currently thology that had snakes for hair and poi- known from throughout the Atlantic low- sonous breath, in allusion to the grotesque lands of Costa Rica and two sites in central appearance of the ¯owers, the tentacle-cov- and eastern Panama (a third Panamanian ered limb, and the putrid smell that ema- collection lacks locality information), from nates from it. sea level to 100 m (to 600 m in the Cor- dillera de TilaraÂn). It is expected to occur Additional specimens examined: COSTA RICA. Al- both in southern Nicaragua and in northern ajuela: San Carlos, Hacienda Platanar, 17 km E of La department of Choco in Colombia, since it Fortuna, 75 m, 28 Jun 1985 (bud), Haber et al. 1823 (CR, MO); San RamoÂn, Bosque Eterno de los NinÄos, has been collected close to the Nicaraguan Cordillera de TilaraÂn, 5 km W of Chachagua, Quebra- and Colombian borders. da Chachaguita drainage, 10Њ24Ј00ЉN, 84Њ39Ј00ЉW, Occurring along or near streams, in sec- 800±900 m [actually 600 m, W. Haber, pers. comm.], ondary growth thickets. At La Selva, it 17 Jul 1993 (bud), Haber 11545 (INB, MO). Heredia: climbs to the crowns of small to medium- SarapiquõÂ, Puerto Viejo, La Selva Biological Station, 10Њ26ЈN, 83Њ59ЈW, 50 m, 29 Mar 2001 (¯), Blanco sized trees, but not tall canopy trees with 1852 (USJ), 30 Mar 2001 (¯), Blanco 1853 (USJ), 4 straight boles. As in Aristolochia grandi¯o- Jul 2001 (fr), Blanco et al. 1946 (USJ), 11 Jun 1981 ra, the stems are herbaceous and never (¯), Hammel & Trainer 10851 (DUKE), 14 Sep 1981 34BRITTONIA [VOL. 54

(¯), D. Smith 211 (DUKE, MO). LimoÂn: Reserva to ¯at, smooth limb, and has a long caudal BioloÂgica Hitoy Cerere, Valle de la Estrella, appendix. GonzaÂlez (1990) provided a de- 9Њ40Ј30ЉN, 83Њ01Ј20ЉW, 100 m, 21 Sep 1993 (bud), Carballo 246 (INB, MO); Matina, Cordillera de Tal- tailed illustration of A. grandi¯ora, that in- amanca, along RõÂo Barbilla, to ca. 0.5 km upstream cludes material from the type. The utricle from jct. with Quebrada CanÄabral, 10Њ01ЈN, and tube are similar in shape in both spe- 83Њ24.5ЈW, 100 m, 8 Sep 1988 (¯.), Grayum et al. cies, which probably accounts for their con- 8891 (CR, MO); Refugio Nacional Barra del Colorado, forests and pastures between RõÂo Chirripocito and RõÂo fusion in herbaria, but both structures are Sardina (``Sardinal'' on Chirripo AtlaÂntico quadran- considerably larger in A. gorgona (Fig. 2). gle), 10Њ38ЈN, 83Њ45ЈW, 12 m, 20 Apr 1990 (fr), Gra- Images of living ¯owers of both species are yum 9781 (CR, MO); Sipurio [Talamanca, Amubri, available online at the W3 TROPICOS 9Њ32ЈN, 82Њ57ЈW, 50 m], Apr 1894 (¯), Tonduz s.n. website (http://mobot.mobot.org/W3T/ (CR 8744); Talamanca, Sixaola, rd. between Punta Uva & San Rafael, Cerros de Manzanillo, 9Њ37Ј28ЉN, Search/vast.html). 82Њ42Ј01ЉW, 0 m, 22 Mar 2001 (bud, fr), Valverde Only a handful of collections of Aristo- 1369 (USJ). lochia grandi¯ora have ¯owers comparable in size with those of A. gorgona. Among PANAMA. Without further locality data: 1962 them are Gentry & Lott 32509 (MO) from (bud, fr.), Duke 6142 (MO). DarieÂn:RõÂo Pirre between Pirre & El Real, 30 Dec 1972 (bud), Gentry & Clewell Veracruz, Mexico; R. & E. West s.n. 7093 (MO). PanamaÂ: SerranõÂa de MajeÂ, trail along RõÂo (FLAS) from a cultivated plant in Florida; IpetõÂ, near con¯uence with RõÂo Ambroino, 8Њ57ЈN, and Allen 867 (MO) from DarieÂn, Panama. 79Њ32ЈW, 29 Jan 1984 (bud, ¯), Churchill & de Nevers The latter might represent a different, as yet 4476 (MO). undescribed species (see below). These col- Aristolochia gorgona has been collected lections have gigantic ¯owers with limbs in several times in the past but has been con- excess of 25 cm in length (excluding the fused with A. grandi¯ora in herbaria. It was appendix). ®rst collected by Tonduz in 1894. All the The calyx limb of Aristolochia gorgona collections from La Selva Biological Sta- has a wide surface area, and once open it tion attributed to A. grandi¯ora appear to folds in a complex way that is dif®cult to be really A. gorgona. There are ®ve Aris- describe (Fig. 1B±E). First, there is a me- tolochia species known from La Selva dial-longitudinal crease from top to bottom, (Wilbur, 1994), three of which have been rendering the limb conduplicate. The mar- described as new from the site (Barringer, gins are somewhat re¯exed, exposing the 1983b; Pfeifer, 1976; and the present one). inner surface to the sides of the ¯ower. Fi- Aristolochia grandi¯ora and A. gorgona nally, there is a transverse crease at the mid- are closely related. Shared characters in- dle, so that the limb folds under the genic- clude herbaceous (non-woody) stems, sin- ulate tube. Because of this second fold, the gle-¯owered in¯orescences with a perfoli- limb actually presents two faces, one to- ate bracteole at the base of the pedicel; a ward the front, and another toward the bot- tube divided into gullet and vestibule that tom. From a distance, each individual ¯ow- curves away from the median calyx lobe er looks like two open ¯owers too close to- (GonzaÂlez & Stevenson, 2000); a well-de- gether, one pushing the other to a down- ®ned annulus, a massive limb that extends ward-facing position (Fig. 1B). This ¯ower more or less perpendicularly from the tube shape is constant among different plants opening, carpels separating completely at seen at La Selva. dehiscence, and thick, ¯oating seeds (Gon- An easy way to distinguish Aristolochia zaÂlez, 1990, reports ¯at seeds for of A. gorgona is to look for the ``tentacles'' cov- grandi¯ora, in reality an occasional condi- ering the internal limb surface. In dried tion caused by seed abortion; pers. obs.). specimens, these tentacles turn black and However, A. gorgona can be distinguished shrivel, looking almost like hairs. While by its generally larger ¯owers, its ®mbriate drying, many of them stick to the newspa- calyx limb that folds under the tube, the per and break off easily, but many others presence of ``tentacles'' on the limb sur- remain on the limb. These structures prob- face, and the relatively short caudal appen- ably act as osmophores, judging from their dix. Aristolochia grandi¯ora has a concave papillate surface (a common feature of 2002] BLANCO: ARISTOLOCHIACEAE 35

FIG. 2. Shapes of pre-anthesis ¯ower buds (about same size of open ¯owers) of Aristolochia grandi¯ora and A. gorgona in pro®le. A. A. grandi¯ora (JimeÂnez & Castro 1015, CR). B. A. gorgona (Blanco 1752, USJ). scent glands in ¯owers; Vogel, 1990). In species consist mainly of unopened buds many Aristolochia species with an appen- (see below). daged calyx limb, the appendix is the major Even though ¯ower color may be vari- source of volatile compounds (Vogel, able in both species (apparent variation 1990), and this is the case in A. grandi¯ora probably owes much to the lack of unifor- (pers. obs.). It is conceivable that in A. gor- mity in color description among collectors), gona the tentacles replace the appendix as it seems that the calyx limb of Aristolochia the scent-producing organ, providing a gorgona has large pink to clear brown ir- much larger surface-area for diffusion. In A. regular blotches (sometimes becoming al- gorgona, the appendix is relatively short most solid) on a cream surface, while that (4±5 cm), whereas in A. grandi¯ora it is of A. grandi¯ora has smaller, dark maroon normally 20±50 cm, and there are a few spots in a checkered, radiating pattern. In reports of lengths of1mormore (Standley, both species the blotches become darker 1938). and solid around the annulus. Even in bud the two species can be read- Flower size also appears to be highly var- ily distinguished (Fig. 2). In buds of A. gor- iable in Aristolochia gorgona, both within gona, the limb is sharply bent under the individuals and among populations. For in- tube and has a very short caudal appendix, stance, the holotype has a larger ¯ower whereas in A. grandi¯ora the limb is fun- (both in limb area and utricle length) than nel-shaped and the appendix is relatively the isotypes. Other specimens have even long, even in small buds. If cut open, im- smaller ¯owers (with the exception of mature buds of A. gorgona show the ``ten- Grayum et al. 8891 at CR, which is even tacles'' developing on the internal limb sur- larger), but some of these clearly represent face. These details are important for iden- immature buds. Most ¯ower measurements ti®cation because most collections of both in the description were taken from the type 36BRITTONIA [VOL. 54 collection only, because it is dif®cult to dis- zaÂlez (1990) as a Linnaean name under the tinguish between true variations in size and synonymy of A. grandi¯ora. Nevertheless, artifacts from pressing. the former was originally described by Jac- Vegetatively, the two species are very quin (1760) based on a Jamaican collection, similar. The leaves of Aristolochia gorgona, and explicitly cited as such by Linnaeus however, tend to be relatively wider (mean (1762). Pfeifer's (1966) illustration and de- length-to-width ratio of 1.2 vs. 1.6 in A. scription of A. caudata agree well with Jac- grandi¯ora), with a rounder outline, and quin's (1760), and is quite different from A. lighter green when exposed to similar light grandi¯ora. Thus, Aristolochia caudata levels. The leaf sinus of A. grandi¯ora Jacq. should no longer be considered a syn- tends to be U-shaped, while in A. gorgona onym of A. grandi¯ora. it is more omega-shaped, with the basal A similar case applies to the name Aris- lobes curved inward, sometimes even tolochia gigantea Mart. & Zucc. Both Pfei- touching. However, there is some overlap fer (1966) and GonzaÂlez (1990) included in leaf shape between both taxa. Fruits and this name under the synonymy of A. gran- seeds are similar in both species. di¯ora, as having been published by Hook- Aristolochia gorgona might correspond er. In his account of A. gigantea, Hooker to what Pfeifer originally called the ``true'' (1846) explicitly cited the original descrip- A. grandi¯ora (Pfeifer, 1960). Flowers were tion by Martius, and mentioned the pres- described as ``pilose'' and extremely large, ence of ``perfoliate stipules'' (pseudostipu- and the color description also matches that les), characteristic of this Brazilian species of A. gorgona (cited specimen, Allen 647, (GonzaÂlez, 1990: 136). The associated il- not seen). However, Pfeifer described it as lustration, however, shows a plant without having a long caudal appendix, and the giv- pseudostipules, very similar to A. grandi¯o- en measurements are probably much exag- ra, but in which the calyx limb points up- gerated (¯owers 3 m long, utricle 45 cm ward and has a short appendix. Pfeifer's long). Pfeifer himself (1966) later refrained (1966) ®gure 37 includes a scheme of the from recognizing this taxon as distinct. A ¯ower in Hooker's illustration. It clearly be- specimen at MO (Allen 867 from DarieÂn, longs to the A. grandi¯ora complex, and Panama) has the same general shape as the could possibly represent a malformed ¯ow- illustration in Pfeifer (1960: ®g. 97), with a er. Duchartre (1864: 473) became aware of very large ¯ower that has a very short tube Hooker's confusion, and christened the il- with a single curve (instead of sigmoid as lustrated plant as A. grandi¯ora ssp. hook- in A. grandi¯ora and A. gorgona), no ob- eri. Therefore, A. gigantea should be ex- vious vestibule, and a long appendix; how- cluded from the synonymy of A. grandi¯o- ever, it lacks any visible pilosity. This spec- ra. imen might represent a different, unnamed Pfeifer (1966) included Aristolochia cor- taxon, but it is necessary to study living di¯ora Mutis ex H.B.K. in the synonymy material. of A. grandi¯ora. GonzaÂlez (1990) dem- The type of Aristolochia grandi¯ora onstrated that A. cordi¯ora really corre- (Swartz s.n., S 2174, annotated as such by sponds to what has until recently been F. GonzaÂlez in 1987) shows a smooth calyx called A. gigantea Mart. & Zucc. in south- limb with a long appendix, and the utricle ern Central America (Pfeifer, 1966; Croat, shape and proportions match those of pop- 1978; Barringer, 1983a), and provides de- ulations on the Paci®c lowlands of Costa tailed differences among the three species. Rica. Pfeifer (1960: ®g. 96) originally clas- The real A. gigantea only occurs in Brazil. si®ed these plants as A. arborescens L. Similarly, the name A. cordi¯ora Mutis ex Even though Linnaeus's (1753) name pre- H.B.K. should be excluded from the syn- dates A. grandi¯ora Sw., the poor original onymy of A. grandi¯ora. description and apparent absence of a type Other names attributed to the synonymy specimen hinder its application (GonzaÂlez, of Aristolochia grandi¯ora that need to be 1990: 98). compared with A. gorgona are A. scandens Aristolochia caudata was cited by Gon- P. Br., A. foetens Lindl., A. gigas Lindl., A. 2002] BLANCO: ARISTOLOCHIACEAE 37 gigas var. sturtevantii W. Watson, and A. elongate tube gullet, and the ¯owers are pichinchensis Pfeifer (Pfeifer, 1966; Gon- smaller than those of regular A. grandi¯ora. zaÂlez, 1990, 1994). Browne (1756: 329) de- The associated illustration shows an incom- scribed three different species from Jamaica pletely dehiscent fruit (carpels remaining as A. scandens. The description of the sec- united at the apex), but this is not evident ond one (``A. 2 scandens'' in Pfeifer, 1966, in the paratype (Dodson & Gentry 6537, and GonzaÂlez, 1990) agrees well with typ- MO). The taxonomic status of A. pichin- ical A. grandi¯ora Sw. It points out the chensis needs to be reassessed with careful large size of the ¯owers and the long ap- study of living material. pendix of the calyx limb. The descriptions More study is needed to clarify the tax- of the other two species correspond respec- onomy of the Aristolochia grandi¯ora com- tively with A. odoratissima L. and A. tri- plex over its wide range in the Neotropics. lobata L. (Pfeifer, 1966). It is possible that several other species are The descriptions and associated illustra- still lumped under this name (Pfeifer, 1966). tions of both Aristolochia foetens, based on For example, the ¯ower illustrated by Patz- a plant from ``the West Indies'' (Lindley, elt (1985: 123, as A. gigas Lindl.) also be- 1836), and A. gigas, based on a plant from longs to this group, but the shape of the Guatemala (Lindley, 1842), clearly show utricle and tube is different from Central the characteristic features of A. grandi¯ora, American populations of A. grandi¯ora, including a concave calyx limb with a long and its appendix is also very short. Differ- appendix. Lindley differentiated his two ences that are obvious in fresh ¯owers are species by ¯ower size and color pattern, obscured in herbarium specimens (Pfeifer, strength of smell, and plant indument. As 1977). presently understood, these features are var- When collecting these plants, the ¯owers iable within A. grandi¯ora. are easily damaged by pole pruners and by Watson (1891) proposed the name Aris- falling to the ground. In Aristolochia gran- tolochia gigas var. sturtevantii for a larger- di¯ora, individual ¯owers take up to 40 ¯owered form of A. gigas Lindl. Again, this days to develop and are open for two days can easily be considered a variation of A. only (Hilje, 1984), so many collections or grandi¯ora. Watson didn't provide a formal duplicates consist only of immature buds description or a Latin diagnosis, but since (Pfeifer, 1960; pers. obs.). Because buds he made reference to the illustrations pre- burst open easily when pressed, care should viously published by Sargent (1890: 597± be taken to note whether unopened buds are 599), it complies with article 44.1 of the being included in the collection. As the ap- Code (Greuter et al., 2000). However, since pendix is easily damaged or lost during col- the name was explicitly proposed (ostensi- lecting, the length of this structure should bly in anticipation of its future acceptance), be included in the description. And since it is to be considered invalidly published the ¯owers deliquesce rapidly after the sec- under article 34.1(b) of the Code. The com- ond day of anthesis, they should be pressed bination ``Aristolochia grandi¯ora var. immediately (preferably between two layers sturtevantii'', occasionally used in the hor- of waxed paper, since they often stick to ticultural trade and in some herbarium de- newspaper). Being three-dimensional, they terminations, has never been validly pub- are also hard to press in such a way that lished. permits appreciation of the original shape. Aristolochia pichinchensis was described The complex ¯ower shape of A. gorgona is from Ecuadorian material by Pfeifer (1977). impossible to infer from a pressed speci- GonzaÂlez (1990) considered its diagnostic men, and the enormous ¯owers scarcely ®t features as those of incompletely open on a standard herbarium sheet when ¯at- ¯owers of A. grandi¯ora, and sunk it under tened. Photographs of fresh, open ¯owers the latter's synonymy. It also has a smooth from different views showing a scale are calyx limb with a relatively long appendix, thus useful. Careful, standardized measure- though the limb folds transversely (but in a ments of different ¯oral parts deserve par- different way than in A. gorgona), has an ticular attention (cf. GonzaÂlez, 1994: ®g. 2). 38BRITTONIA [VOL. 54

The ideal approach for the study of this ÐÐÐ. 1983b. Notes on Central American Aristolo- complex would be to develop a common chiaceae. Brittonia 35: 171±174. Browne, P. 1756. The civil and natural history of Ja- garden, especially since even a single plant maica. T. Osborne & J. Shipton, London. can produce ¯owers of different sizes, as Cammerloher, H. 1923. Zur Biologie der BluÈte von presumably determined by a number of en- Aristolochia grandi¯ora Swartz. OÈ sterr. Bot. vironmental factors (Hilje, 1984: 23). Zeitschr. 72: 180±198. With ¯owers considerably bigger than Cook, K. 2001. Aristolochiaceae Juss. Pages 229±233. In: W. D. Stevens, C. Ulloa Ulloa, A. Pool & O. those of most Aristolochia grandi¯ora M. Montiel, editors. Flora de Nicaragua, Tomo I. specimens (in terms of perianth area), A. Missouri Botanical Garden Press, Saint Louis. gorgona is one of the largest-¯owered spe- Croat, T. B. 1978. Flora of Barro Colorado Island. cies in the genus, and one of the largest- Stanford University Press, Stanford. ¯owered plant species native to the Amer- Duchartre, P. 1864. Aristolochiaceae. In: A. P. De Candolle & A. L. P. De Candolle, editors. Prodr. icas. It is surprising that a plant with such Syst. Universalis Regni Veg. 15: 421±498. large ¯owers and relatively common in a GonzaÂlez, F. A. 1990. Aristolochiaceae. Pages 3±184. well-botanized country as Costa Rica had In: J. O. Rangel, A. Cadena, G. Correa & R. Ber- been overlooked until now. It would be nal, editors. Flora de Colombia. MonografõÂa No. worthwhile to introduce this striking new 12. Instituto de Ciencias Naturales, Universidad species to cultivation. Stem cuttings from Nacional de Colombia, BogotaÂ. ÐÐÐ. 1994. Aristolochiaceae. Pages 1±42. In: G. the type locality have been established at Harling & L. Andersson, editors. Flora of Ecuador. Lankester Botanical Garden, and are grow- No. 51. University of GoÈteborg, Stockholm. ing vigorously at the time of this writing. ÐÐÐ & D. W. Stevenson. 2000. Perianth develop- ment and systematics of Aristolochia. Flora (Ger- Acknowledgments many) 195: 370±391. Greuter, W., J. McNeill, F. R. Barrie, H. M. Burdet, I sincerely thank Orlando Vargas (La Sel- V. Demoulin, T. S. Filgueiras, D. H. Nicolson, P. va Biological Station) for authorizing the C. Silva, J. E. Skog, P. Trehane, N. J. Turland collection of the type specimen of Aristo- & D. L. Hawksworth. 2000. International Code of lochia gorgona and for forwarding images Botanical Nomenclature (Saint Louis Code). Koeltz Scienti®c Books, KoÈnigstein. of ¯owers and herbarium specimens. Jorge Hilje, L. 1984. FenologõÂa y ecologõÂa ¯oral de Aristo- GoÂmez-Laurito (USJ) translated the diag- lochia grandi¯ora Swartz (Aristolochiaceae) en nosis into Latin. Mia Ehn Salter (S) and Costa Rica. Brenesia 22: 1±44. Michael Grayum (MO) kindly provided im- Hooker, J. D. 1846. Aristolochia gigantea. Curtis's ages of many collections, including the type Bot. Mag. 72: 4221. of A. grandi¯ora. I deeply thank Steve Caf- Jacquin, N. J. 1760. Enumeratio Systematica Plantar- um. Theodorum Haak Lugduni Batavorum, Leiden. ferty (BM, Linnaean Plant Name Typi®ca- Lindley, J. 1836. Aristolochia foetens. Bot. Reg. (Ed- tion Project) for sending copies of original ward's) 21: 1824. descriptions and associated illustrations of ÐÐÐ. 1842. Aristolochia gigas. Bot. Reg. (Ed- several names attributed to the synonymy ward's) 28: 60. of A. grandi¯ora, and for providing critical Linnaeus, C. 1753. Species Plantarum. Holmiae Im- pensis Laurentii Salvii, Uppsala. information on the publication of A. cau- ÐÐÐ. 1762. Species Plantarum. Ed. 2. Holmiae Im- data. Judith Magee (Botany Library, Nat- pensis Laurentii Salvii, Uppsala. ural History Museum, London) provided Patzelt, E. 1985. Flora del Ecuador. Banco Central del some references. I also thank Michael Gra- Ecuador, Quito. yum (MO) and two anonymous reviewers Petch, T. 1924. Notes on Aristolochia. Ann. Royal for many helpful suggestions to an earlier Bot. Gard. Perideniya 8: 1±108. Pfeifer, H. W. 1960. Aristolochiaceae. In: R. E. version of the manuscript. Fieldwork at La Woodson Jr. & R. W. Schery, editors. Flora of Pan- Selva Biological Station was funded by a ama. Ann. Missouri Bot. Gard. 47: 309±323. research grant from the Andrew W. Mellon ÐÐÐ. 1966. Revision of the North and Central Foundation and the Organization for Trop- American hexandrous species of Aristolochia (Ar- ical Studies. istolochiaceae). Ann. Missouri Bot. Gard. 53: 1± 114. ÐÐÐ. 1976. Aristolochia translucida, a new species Literature Cited from Costa Rica. Brittonia 28: 348±351. Barringer, K. 1983a. Aristolochiaceae. In: W. Burger, ÐÐÐ. 1977. A new Ecuadorian dutchman's-pipe, editor. Flora Costaricensis. Fieldiana, Bot. n.s. 13: Aristolochia pichinchensis. Selbyana 2: 29±30. 79±87. Sargent, C. S. 1890. New or little known plants. The 2002] BLANCO: ARISTOLOCHIACEAE 39

Pelican-Flower (Aristolochia grandi¯ora). Gard. & Smithsonian Institution Libraries, and The National Forest 3: 596. Science Foundation, Washington, DC. Standley, P. C. 1937. Flora of Costa Rica. Part II. Watson, W. 1891. Foreign correspondence. London Publ. Field Mus. Nat. Hist., Bot. Ser. 18: 401±780. letter. Gard. & Forest 4(195): 544±546. ÐÐÐ & J. A. Steyermark. 1946. Aristolochiaceae. Wilbur, R. L. 1994. Vascular plants: an interim check- In: P. C. Standley & J. A. Steyermark, editors. Flora list. Pages 350±378. In: L. A. McDade, K. S. of Guatemala. Fieldiana, Bot. 24(4): 93±101. Bawa, H. A. Hespenheide & G. S. Hartshorn, edi- Swartz, O. 1788. Nova genera et species plantarum. tors. La Selva: ecology and natural history of a Holmiae, Uppsalae & Aboae In Bibliopoliis Acad. neotropical rain forest. University of Chicago M. Swederi, Stockholm. Press, Chicago. Vogel, S. 1990. The role of scent glands in pollination: Wunderlin, R. P. 1998. Guide to the vascular plants on the structure and function of osmophores. of Florida. University Press of Florida, Gainesville.