Fairchild Tropical Garden, 11935 Old Cutler Road, Coral Gables (Miami), Florida 33156, USA E-mail:

Summary: Agave decipiens Baker is shown to be morphologically and geographically dis­tinct from A vivipara L. of Mexico.The species differ in ovary length and anther length. Agave de­cipiens is endemic to southern Florida, USA, where it occurs in small, scattered populations in coastal habitats, some of which are currently protected. Its conservation status is classified as Conservation Dependent.


The genus Agave L. (Agavaceae) achieves its greatest exuberance in the desertic regions of western North America (Gentry, 1982), but sev­eral additional species are found in northern South America, Central America, the Caribbean, and Florida (Trelease, 1913; Berger, 1915). Flori­da claims two endemic species. One is the aptly-named and poorly-known Agave neglecta Small, which is a member of Gentry's group Sisalanae. The other is A. decipiens Baker (Fig.l), a species described from coastal areas of southern Florida. In this paper, the taxonomic identity of A. decipiens is critically reexam­ined and its conservation status is assessed.

Agave decipiens (false sisal) was first de­scribed by Baker (1892) from an unnumbered collection made in southern Florida in February 1892 by C. R. Dodge, who was investigating fiber crops for the U.S. Department of Agriculture. Neither Baker's (1892) description nor Dodge's specimen label indicates whether the material was taken from cultivated specimens or from the wild. Berger (1915) noted that the type ma­terial is at Kew (K), and this specimen should be considered the holotype. An isotype is deposited at the National Herbarium (US). The existence of type specimens nullifies Gentry's (1982) selec­tion of an illustration from Curtis' Botanical Mag­azine (122: t. 7477 [1896]) as the "lectotype" (Gentry designated a neotype).

Some disagreement as to the nativity of A. decipiens was noted by previous workers. Mulford (1896), Berger (1915), Small (1903, 1913), and Wunderlin (1998) believed it to be native to Florida. In contrast, Long and Lakela (1971) claimed that the species was "introduced from Mexico". Johnson and Barbour (1990) regarded it as an introduced exotic. Small (1933) noted that "it has been suggested that they [popula­tions of A. decipiens] represent prehistoric in­troductions from Mexico". Small concluded, "If this be so, their Mexican ancestors have been lost, or the Florida plants have greatly changed."

One publication, Sousa & Cabrera (1983), cited a collection (Lundell & Lundell 7817 at MICH) of A. decipiens from the state of Quintana Roo, Mexico, but my examination of this collection revealed it to be another taxon, A vi­vipara L. var. vivipara (see below). Likewise, a specimen collected in Mexico and deposited in the Field Museum (Gaunter 23164, photo at CICY) was annotated by Trelease in 1915 as 'Agave decipiens?" It too is A. vivipara. Espejo Serna and Lopez-Ferrari (1992) made no men­tion of A. decipiens in their checklist of Mexi­can Agave.

Similarity to Agave vivipara

The idea that A. decipiens came from Mexi­co arose from the obvious similarity between A. decipiens and some populations of A vivipa­ra (Fig. 2), a species more familiarly known as A. angustifolia Haw. (Forster, 1992). Indeed, the two species are nearly identical in all vegeta­tive characters. Berger (1915) and Gentry (1982) placed both species together in group Rigidae, based on the following characteristics: long narrow leaves, tepals reflexed upon dry­ing and capsules broadly ovate (although Gen­try described the fruits as "ellipsoid to oblong" in his description of the species). Gentry (1982) implied an even closer taxonomic rela­tionship between the two species in his Figure 20.12, captioned "The extensive distribution of Agave angustifolia [A. vivipara] and varieties." In this distribution map, A. decipiens is mapped along with six varieties of A. vivipara. No other group Rigidae agaves appear in the map. Gentry did not explain his reason for in­cluding A. decipiens among the varieties of A. angustifolia in this map, but the fact that he singled out A. decipiens from among all the other group Rigidae agaves suggests to me that he believed A. decipiens and A. vivipara to be more closely related to each other than to other agaves in the group.

Nevertheless, Gentry kept the two species separate. The major difference noted by Gentry (1982) between A. decipiens and A. vivipara is the presence of an obvious stem 1.5 m or more in length in the former species and a shorter stem in the latter, a difference I am unable to affirm. The shape, size, and texture of the leaves are similar in both species, and characters of the marginal teeth and terminal spine are simi­lar and similarly variable.

There are similarities in floral structures as well, but the ovary is consistently longer in A. decipiens than in A. vivipara. The mean length of the ovary is 30.3 mm (±6.0) in A. decipiens and 22.0 mm (±38) in A. vivipara (Table 1). The latter figure closely agrees with the mean of 22.2 mm for A. vivipara reported by Colunga-García Marín et al. (1996). This consistent difference is the easiest characteristic by which the two species can be reliably distin­guished. Another difference is the length of the anther. A small sample of anthers averaged 16.8 (±4.2) mm long in A. decipiens and 21.4 (±5.4) mm long in A. vivipara. Colunga-García Marín et al. (1996) reported a mean of 22.8 mm for A. vivipara. The difference is significant (Table 1), although larger sample sizes for both species are needed. The filament, which lengthens during anthesis, is difficult to compare in herbarium specimens gathered at various stages of floral maturity.

Figure 1. Agave decipiens at the edge of a tropical harrdwood hammock, Big Pine Key, Monroe Co., Florida.

Figure 2. Agave vivipara (formerly known as A. angustifolid) in Mpio. Dzindzantum, Yucatan, Mexico.

Description of Agave decipiens

The overall appearance of A. decipiens is quite unlike any other species of Agave native or naturalized in Florida (A. neglecta, A. sisalana Perrine, and A. desmetiana Jacobi [Wunderlin, 1998]), although it bears a striking resemblance to some Mexican plants of A. vi­vipara (Fig. 2). Mature plants possess an erect stem 5.5 to 10.5 cm in diameter and up to 3 m in length (Fig. 3). The tall stems, either bare or clothed with marcescent leaves and topped with rosettes of stiff leaves, give these plants a superficial resemblance to some species of Yucca L. Some plants of A. decipiens produce suckers.

Plants bear ca. 36 to 72 leaves per rosette. The individual leaves are stiff and hard, 24 to 200 cm long and 1.3 to 8.0 cm wide at their widest point, narrowly lanceolate in outline, with a thickened, succulent, clasping leaf base. The leaves are bright green or yellowish green but not the bluish green of some clones of A. vivipara. In full sun, the margins of the leaves are turned up, giving the leaf a U-shape in cross-section; in shade, the leaves are nearly flat. As in A. vivipara, the leaves are armed with prominent marginal teeth 1.1 to 3.6 mm in length, which may point toward the leaf base or toward the tip. The teeth are brown or brownish black and may be borne on low, in­conspicuous teats, but the leaf lacks a continu­ous corneous margin between the teeth. The apex of the leaf is armed with a stout, straight terminal spine, brown or brownish black in color, 10.3 to 28.9 mm long, with or without decurrent margins.The length and shape of the spine is so highly variable that spine characters cannot be used to differentiate A. decipiens from A. vivipara.

The inflorescence of A. decipiens is a termi­nal panicle up to 4.5 m in length. The 8 to 12 lateral branches are up to 45 cm long, and flowers are borne in clusters of ca. 23 flowers. The flowers are greenish yellow, and short­lived. The pedicel is 5.3 to 11.0 mm long, and the ovary is 22.2 to 43.5 mm long. The hypanthium (floral tube) is 5.8 to 10.1 mm long and funnelform. The six tepals are more or less equal. The outer tepals are 17.1 to 25.8 mm long, greenish yellow, linear, and quickly with­ering. Tepal apices are hooded and bear tufts of short white hairs. The filaments are terete, at­tached more or less at the middle of the tube, 26.8 to 47.5 mm long, and the anthers are 13.1 to 23.5 mm long. The capsules were not seen, although Gentry (1982) described them as "el­lipsoid to oblong, 3.5 to 5 cm long, stipitate."

Figure 3. A small population of A. decipiens, Big Pine Key, Monroe Co., Florida.

Figure 4. A heavily shaded population of A. decipi­ens in tropical hardwood hammock, Elliott Key, Miami-Dade Co., Florida. The population persists, but the plants do not reproduce. All photos are by the author.


The habitat of A. decipiens is exposed areas near the coast, especially on shell mounds or middens. Shell mounds are anthropogenic habi­tats characterized by fertile, well-drained soil, ideal for crops, and consequently much of the original (pre-Colombian) vegetation of shell mounds was destroyed soon after permanent agricultural settlements were established in southern Florida. A. decipiens also occurs on the margins of tropical hardwood hammocks, on coastal dunes inland from the Uniola paniculata L. zone (Johnson & Barbour, 1990), and be­hind the mangrove zone, that is, inland from the zone of the buttonwood mangrove, Conocarpus ectus L. In shaded areas of coastal tropical hardwood hammocks, A. decipiens will persist, but it will not thrive (Fig. 4). Plants in these habitats likely await light gaps in the hammock canopy in order to mature and reproduce.

The distribution of A. decipiens is illustrated in Fig. 5. This species was reported to be natu­ralized in South Africa, near Pretoria, by Smith and Steyn (1999), but their description more likely refers to A. vivipara. Their voucher speci­men at MO was not available for study.

Figure 5. The distribution of A. decipiens in southern Florida.

Threats to survival

Urbanization and suburban sprawl have devastated many of the areas in which Agave decipiens once grew. It is found in coastal scrub, shell mounds, pinelands, and the margins of tropical hardwood hammocks (Small, 1933; Wunderlin, 1998). These areas have been especially hard-hit by development. Although A. decipiens occurs on some lands protected at the federal, state, or local level (e.g., Biscayne National Park, Windley Key State Geologic Site), the species is difficult to find outside these areas. A. decipiens is experi­encing loss of habitat, decline in habitat quality, and fragmentation of existing populations. It is classified as Conservation Dependent (following IUCN criteria) in that the species would move into threatened categories if it ceased to receive protection of federal, state, or local parks.

Specimens Examined

Agave decipiens Baker

USA. Florida: locality unspecified, Dodge s.n. (isotype BUS): Collier Co., ca. 1 mi S of Marco Is­land on Horrs Island. Hansen et al. 11836 (FLAS, NY. USF), Marco Island S of Collier City, Lakela 29446 (USF), vicinity of Marco, Standley 12812 (US); Lee Co., NE tip of Fine Island E of Bokeelia, Beckner 1756 (FLAS), eastern Sanibel Island, Brumbach 6086 (FLAS), Sanibel Island, J. N. "Ding" Darling National Wildlife Refuge, NW end of refuge, Hansen & Hansen 5698 (USF), Mound Key, Estero Bay, Todd 54 (USF); Miami-Dade Co., top of cliff 2 mi S of Miami, Britton 66 (NY), rocks on shore of Biscayne Bay, Curtiss 2836* (US), Bull Key, opposite Lemon City, Small or Carter s.n. (NY), between Cutler and Black Point, Small & Carter 985 (NY), between Miami and Coconut Grove, Small & Small 4625 (NY), Biscayne National Park, Elliott Key, Scorpi­on Bite, Zona 830 (FTG), Biscayne National Park, Elliott Key, coastal strand near docks, Zona 831 (FTG), Meigs Key, Small 7367 (FLAS, NY, US); Monroe Co., Big Pine Key, Killip 31427 (US), same locality, Killip 31696 (FLAS), same lo­cality, Killip 41053 (US), Content Keys, Long et al. 2715 (USF), Lower Matecumbe Key, Moldenke 5833 (NY); Sarasota Co., S tip of Casey Key, Beckner 1717 (FLAS), Mansota Key, Coons s.n. (FLAS), Longboat Key, Long & Lakela 27562 (USF), same locality, Long & Lakela 28146 (FLAS. USF). CULTIVATED. California: Los Angeles Co., San Marino, Huntington Botanical Gardens, Gentry 19749 (US). Florida, Miami-Dade Co., Coral Gables, Fairchild Tropical Garden, plot 135, accession 71-378A, Zona 816 (FTG), Parthasarthay 12 (FTG), Fantz 3973 (FTG).

Agave viviptira L.

MEXICO. Morelos. 20 km NE of Huautla, Camp 30 (MICH). Oaxaca. Between Totolapan and San Carlos, Nelson 2554 (US). Puebla: 15 miles from Puebla on Puebla-Tlaxcala road, Ogden 5134 (MICH, US). Quintana Roo: Coba, E of ruins, Lundell & Lundell 7817 (MICH). Yu­catan: near Telchac, on low dunes, Lundell & Lundell 8110 (MICH); Mpio. Chicxulub, be­tween Chicxulub Puerto and San Benito, E of Progreso, Zona & Orellana 857 (CICY, FTG, MEXU); Mpio. Dzemul, Rancho San Antonio, ca. 8 km S of Xtampu, Zona & Orellana 859 (CICY, FTG, MEXU); Mpio. Dzindzantum, between Chabihau and Sta. Clara, ca. 2 km W of Porvenir, Zona & Orellana 858 (CICY, FTG, MEXU); Mpio. Progreso, between Progreso and Chuburna, Zona & Orellana 856 (CICY, FTG, MEXU); Mpio. Rio Lagartos, Parque Nacional Ria Lagertos, ca. 4 km W of Las Coloradas, Zona & Orellana 861 (CICY FTG, MEXU), ca. 3 km E of San Felipe, Zona & Orellana 860 (CICY, FTG, MEXU); Mpio. Tizimin, Parque Nacional Ria Lagertos, W of El Cuyo, Zona & Orellana 862 (CICY, FTG, MEXU).


This work was supported by a grant from the Research Committee of the Cactus and Suc­culent Society of America. I thank David Bogler (Florida International University), Dena Garvue (FTG), Carl E. Lewis (FTG); Steve Woodmansee (Institute for Regional Conservation), Angelica Narvaez (Science Affairs Specialist, US Embassy, Mexico), and the staff of Centro de Investi­gation Cientifica de Yucatan (CICY), including Patricia Colunga-Garcia. I am especially grateful to Dr. Roger Orellana, my most excellent host in the Yucatan, who was a convivial field com­panion and assisted in so many ways. I am grateful to Richard Curry for granting me per­mission to collect within the boundaries of Biscayne National Park. I also thank the curators of CICY, FLAS, FTG, MICH, NY, US, and USF for allowing me access to their specimens.

Table 1. Statistical comparison (two-tailed t-test) of selected morphological features of the leaves and flowers of Agave decipiens and A. vivipara. The asterisk (*) denotes samples that

are statistically significant, i.e., p<0.5000


Baker, J.G. (1892). False sisal of Florida. Kew Bull. Misc. Inform. 1892: 183-184.

Berger, A. (1915). Die Agaven: Beiträge zu einer Mono­graphie. Gustav Fischer, Jena. (1988 reprint with supplement).

Colunga-Garcia Marin, P., E. Estrada-Loera & F. May-Pat. (1996). Patterns of morphological variation, diver­sity, and domestication of wild and cultivated pop­ulations of Agave in Yucatan, Mexico. Amer. J. Bot. 83: 1069-1082.

Espejo Serna, A., & A.R. Lopez-Ferrari. (1992). Las monocotiledones Mexicanas. Una sinopsis floristica. 1. Lista de la referenda. Parte I. Agavaceae, Alismataceae, Alliaceae, Alstroemeriaceae y Amaryllidaceae. Consejo Nac. Flora Mex., Univ. Autonoma Metropolitana Iztapalapa, Mexico D.F.

Forster, P.I. (1992). New varietal combinations in Agave vivipara (Agavaceae). Brittonia 44: 74-75.

Gentry, H.S. (1982). Agaves ofcontinental North America. Univ. of Arizona Press. Tucson.

Johnson, A.E, & M.G. Barbour. (1990). Dunes and mar­itime forests, pp. 429-480. In Myers, R.L. & J.J. Ewel (eds.). Ecosystems of Florida. Univ. of Cen­tral Florida Press, Orlando. Long, R.W., & O. Lakela. (1971). A flora of tropical Florida. University of Miami Press, Miami.

Mulford, A.I.(1896) A study of the agaves of the United States. Missouri Bot. Gard. Ann. Rep. 7: 47-100, plates 26-63.

Small, J.K. (1903). Flora of the southeastern United States. Published by the author, New York.

Small, J.K. (1913). Flora of the Florida Keys. Published by the author, New York.

Small, J.K. (1933). Manual of the southeastern flora. Univ. of North Carolina Press, Chapel Hill.

Smith, G.F., & E.M.A. Steyn. (1999). A first record of Agave decipiens naturalised in southern Africa. South African f. Bot. 65: 249-252.

Sousa S., M., & E.E Cabrera C. (1983). Listados floristicos de Mexico. II. Flora de Quintana Roo. Institu­te de Biologia. UNAM, Mexico D.F.

Trelease, W. (1913). Agave in the West Indies. Mem. Natl. Acad. Sci. 11: 1-55, plates A-E, 1-116.

Wunderlin, R.P. (1998). Guide to the vascular plants of Florida. Univ. Press of Florida, Gainesville.

© Haseltonia, 2001