The confusion of Agave vivipara L. and A. angustifolia Haw., two distinct taxa

ABISAI, GARCIA-MENDOZA AND FERNANDO CHIANG

García-Mendoza, Abisaí (Jardín Botánico, Instituto de Biología, Universidad Nacional Autónoma de México, Apdo. Postal 70-614, Del. Coyoacán, 04510 M'exico, D. F., México; email: abisai@ibiologia.unam.mx) & Fernando Chiang (Departamento de Botánica, Instituto de Biología, Universidad Nacional Autónoma de México, Apdo. Postal 70-233, 04510 México, D. F., México).

The confusion of Agave vivipara L. and A. angustifolia Haw., two distinct taxa. Brittonia xx: 000-000. 2003.- Agave vivipara L. and A. angustifolia Haw., considered conspecific by Wijnands, are shown to be two distinct species. A previous neo-typification (as a lectotypification) of A. angustifolia by Gentry is superseded because it is in conflict with the protologue; a new neotype is designated herein. The economic importance of A. angustifolia is summarized.

Key words: Agavaceae, Agave angustifolia, Agave vivipara, Agave fourcroydes, Agave tequilana, neotypification.

Among the many Agave species with biological, cultural, and economic importance in Mexico and Central America, A. angustifolia Haw. has played a particularly significant role in the historical man-plant relationship. The broad geographical distribution and many forms of this species have resulted in the exploitation of all parts of the plant. The widespread use has resulted in a complex pattern of cultivation that has led to taxonomic confusion.

The uses of this species are many and varied: food for people (flower buds and flowers, stems, leaf bases, young flower peduncles, and fruits); construction (inflorescence stem as posts, rafters, and fences; leaves for thatching); forage (leaves); fuel (dried plants); fibers for rope making; beverages (the stems are cooked and their juice extracted, fermented, and distilled into alcoholic beverages); traditional medicines (juice of cooked leaves and stems and a root infusion are taken internally or used as poultices for both internal and external swelling, as well as for bruises, liver and kidney diseases, arthritis, and dysentery); making of diverse utensils (spines used as nails or needles); ornamental purposes (whole plant). The most important of these applications are its use as a source of hard fibers, for food, and for making distilled alcoholic beverages (García-Mendoza et al., 1993).

Among the domesticated taxa cultivated for their fibers, Gentry (1982) recognized Agave angustifolia var. deweyana (Trel.) Gentry in Tamaulipas and Veracruz, A. angustifolia var. letonae (Taylor) Gentry in El Salvador and Guatemala, and A. angustifolia var. nivea (Trel.) Gentry in Guatemala.

Other studies (Colunga-García Marín et al., 1996; Colunga-García Marín & May-Pat, 1997; Colunga-García Marín, 1998), based on morphological and isoenzymatic variation, concluded that cultivated henequén (Agave fourcroydes Lem.) originated from wild populations of A. angustifolia from the Yucatán Peninsula.

Apparently, several cultivars have been selected from the same gene pool of Agave angustifolia, first as sources of food and lat-er as sources of distilled alcoholic beverages (Colunga-García Marín, pers. comm.). This is also the case for the beverages known generically as "mezcal" that are produced in the Mexican states of Chihuahua, Durango, Guerrero, Mexico, Nayarit, Oaxaca, Sinaloa, and Sonora, as well as "tequila", obtained from A. tequilana Weber, whose separation from A. angustifolia seems to be only nominal (Gentry, 1982).

Agave angustifolia occurs from Sonora and Tamaulipas, Mexico, south to Panama, and grows in several vegetation types, from coastal dunes at sea level to oak-pine forests at 2200 m. In these environments, several forms have appeared. These are difficult to distinguish from one another, which has made the infraspecific taxonomy of A. angustifolia extremely complicated. The morphological variability of this species has led to the publication of more than 20 taxa, whose descriptions were based on cultivated plants that were only incompletely known. Gentry (1982) considered most of the names of these taxa to be synonyms of A. angustifolia. Presently, this species complex is being analyzed in both its morphological and molecular characteristics in order to clarify its classification and settle the nomenclatural disputes.

The present paper was motivated by a publication of Wijnands (1983), in which his treatment of Agave angustifolia is completely different from our concept of the species. The economic importance of A. angustifolia makes it essential that its taxonomy and nomenclature be set straight.

Wijnands (1983) lectotypified Agave vivipara L. and relegated Agave angustifolia Haw. to synonymy. Even though he cited Trelease (1913), Berger (1915), and Hummelinck (1936), he disagreed with these authors' treatments of A. vivipara and A. angustifolia as distinct species. He also chose not to follow Gentry's (1982) account of the Agaves of continental North America, in which A. angustifolia was treated as a distinct species.

We agree with Wijnands' choice of the lectotype of Agave vivipara, namely Commelin's illustration (t. 15, 1703), wrongly cited as t. 65 by Linnaeus (1753) in the protologue; however, we disagree with his treatment of A. angustifolia Haw. as being conspecific with A. vivipara.

The crux of the problem resides in Wijnands' statement when dealing with Agave vivipara. After lectotypifying this name, he stated: "A. angustifolia Haw. is a homotypic synonym [of A. vivipara L.] as the Commelin plate is cited in the protologue, also taxonomically it is considered to be identical." We find this statement completely unwarranted; in Haworth's original description of A. angustifolia, no mention was made of Commelin's plate. In addition to the description of A. angustifolia, four other species are treated in the same page. All are numbered, appearing as follows: 3. Agave flaccida; 4. Agave veracrucis; 6. Agave karatto; and 7. Agave vivipara. We fail to see how Wijnands could have overlooked Haworth's separate treatment of A. angustifolia and A. vivipara and why he totally disregarded the treatments by Berger, Hummelinck, Trelease, and Gentry.

Commelin's illustrations (t. 15) was partially reproduced in Wijnands' work (1983: 35), lacking the foliage. The complete illustration was shown by Trelease (1913: pl. 1, fig. 1); in it, various details that serve to define the species can be observed: leaves that are succulent, broadly lanceolate, somewhat narrowed toward the base, with large marginal teeth and a long terminal spine, recurved or erect (except for the ex-ternal leaves which are prostrate); and an inflorescence that is oblong, with massive short, lateral branches, apparently loaded with bulbils. An important detail is the length of the peduncle, which is shorter than the inflorescence; the bracts of the peduncle are triangular and close together. Commelin's illustration closely agrees with Hummelinck's (1938) tab. II.a, which is a photograph of Agave vivipara from Isla Blanquilla, a Caribbean island off the coast of Venezuela. Hummelinck's (1936: 232) concept of A. vivipara was the same as Trelease's, as demonstrated by his words: "I entirely agree to the interpretation and conception of these species [A. vivipara], as given by Trelease in his monograph of 1913." Hummelinck's concept of this species has not changed, as shown in his later publications (Hummelinck, 1993a, 1993b, 1993c). Alvarez de Zayas (1995), dealing with the Agaves from the Antilles, recognized A. vivipara (native) and A. angustifolia (introduced) as distinct species.

Gentry (1982) treated Agave angustifolia Haw. as a valid species, choosing plate 6 of Targioni Tozzetti (1810) as the lectotype. Steyn and Smith (2000) point out that this plate is actually a neotype. Wijnands' (1983) treatment of A. angustifolia as conspecific with A. vivipara is untenable, because the broadly lanceolate leaves shown in the lectotype of A. vivipara (Commelin's pl. 15) are in conflict with Haworth's (1812) protologue of A. angustifolia: "folia angustissima omnium." Unfortunately, Gentry's (1982) neotypification (intended originally as a lectotypification) must be superseded, for it is also in conflict with Haworth's (1812) protologue. In discussing the leaves of A. angustifolia, Gentry (1982) states: "Generally they are narrow outstanding ensiform leaves, but an atypical broadened form appears out of nowhere, especially in Oaxaca, quite similar to that of Figure 20.6." In citing a figure that is a reproduction of Targioni Tozzetti's figure, Gentry implies that this figure shows a broad leaf form that would thus not match Haworth's (1812) protologue. Thus, the designation of this plate by Gentry (1982) as the lectotype (neotype) might be disputed due to lack of concordance with the protologue.

The geographical ranges of the two species do not overlap. Agave vivipara occurs in the Leeward Islands (Curacao, Aruba, Bonaire) and other islands off the Venezuelan coast (Hummelinck, 1936, 1938)-cited simply as Antilles by Wijnands (1983), whereas A. angustifolia is restricted to continental America (Gentry, 1982; García-Mendoza et al., 1993) and was introduced into the Antilles (Trelease, 1913; Howard, 1979; Alvarez de Zayas, 1995). Also, these two species have been placed in different groups (or sections) of the genus: A. vivipara in group Viviparae (Trelease, 1913), in sect. Viviparae (Hummelinck, 1938, 1993a; Alvarez de Zayas, 1995), or series [Reihe] Viviparae (Berger, 1915); A. angustifolia in ser. [Reihe] Rigidae (Berger, 1915), group Rigidae, (Gentry, 1982) group Sisalanae (Trelease, 1913), or subser. [Unterreihe] Sisalanae (Berger, 1915).

Trelease (1913: 9) showed the contrasting characters of the groups in a key, taking the representative species as typical of each group. Agave vivipara (taken as the representative species of group Viviparae) and A. angustifolia (as representative of group Sisalanae) are thus contrasted:

Leaves broad or curved, ?eshy. Plants moderately large, suckering, bulbiferous. Leaves green, transiently glaucous; spine acicular, involutely grooved; prickles rather close together. Flowers and seeds medium-sized. Native to the West Indies. (Viviparae) -------A. vivipara

Leaves narrow, spreading, firm. Plants mod-erately large, suckering, bulbiferous. Leaves numerous, grayish; spine conical; prickles rather distant. Flowers and seeds large. Introduced (into the West Indies). (Sisalanae) --------A. angustifolia

The following paragraph, taken from Gentry (1982: 560), indicates that Haworth (1812), Trelease (1913), Berger (1915), and Gentry (1982) shared the same concept of Agave angustifolia: "Trelease gave us a clarifying account of the complex and incomplete history of this name (Agave angustifolia Haw.). The plants figured by (Targioni) Tozzetti were growing on the island of St. Helena. 'It was at about this time that the plants were received from St. Helena at Chelsea (England), which Haworth recognized as being his own A. angustifolia' (Trelease 1908: 283)."

Wijnands' (1983) work served as the basis for Forster's (1992) new combinations under Agave vivipara and these, in turn, were the basis for the "correction" of the name A. angustifolia to A. vivipara (Smith & Steyn, 1999).

Researchers working with introduced agaves in Australia (Forster, 1992) and southern Africa (Smith & Steyn, 1999; Steyn & Smith, 2000), with native agaves of the Sonoran Desert, Mexico (Hodgson, 2001), and with agaves around the world (Thiede, 2001) have adopted Wijnands' (1983) point of view that A. angustifolia and A. vivipara are conspecific.

We have presented conclusive evidence showing that Agave vivipara L. and A. angustifolia Haw. are two separate species whose distribution ranges do not overlap. When Forster (1992) made his combinations, he stated that those new combinations were necessary to continue the use of the names in the sense understood by Gentry (1982). It is outside the scope of this paper to deal with the varietal categories of A. angustifolia, but it is our contention that Forster's proposed new combinations of A. vivipara were made employing a mechanical method, i.e., automatically transferring names published under A. angustifolia to combinations under A. vivipara, following Wijnands' (1993) misconception of that taxon. Because of the economic importance of A. angustifolia, whose varied uses are mentioned above, it is of paramount importance that the nomenclature be clear and precise.

FIG. 1. The neotype of Agave angustifolia Haw. (From García-Mendoza & Palma 5654, MEXU.)

The nomenclature should, therefore, stand as follows:

Agave vivipara L. Sp. Pl. 1: 323. 1753. LECTOTYPE (designated by Wijnands, 1983: 35): Illustration in Commelin, Prael. Bot. 65, t. 15. 1703.

Agave angustifolia Haw., Syn. Pl. Succ. 1: 72. 1812. NEOTYPE, here designated: MEXICO. Oaxaca: Distrito Huajuapan de León, Mun. San Andrés Dinicuite, 3 km N of Tutla, 2220 m, 29 Jun 1992, García-Mendoza & Palma 5654 (MEXU No. 1022865) (Fig. 1). This designation supersedes Gentry's (1982, p. 560) neotypification (originally called a lectotypification), Targioni Tozzetti, 1810, pl. 6.

Agave angustifolia var. deweyana (Trel.) Gentry, Agaves Continental N. Amer. 564. 1982.

Agave deweyana Trel., Trans. Acad. Sci. St. Louis 18: 35. 1909. HOLOTYPE: MEXICO. Tamaulipas: plantations about Victoria, 2 Feb 1907, Dewey 649 (US).

Agave vivipara var. deweyana (Trel.) P. I. Forster, Brittonia 44: 74. 1992.

Agave angustifolia var. letonae (F. W. Taylor ex Trel.) Gentry, Agaves Continental N. Amer. 564. 1982.

Agave letonae F. W. Taylor ex Trel., J. Wash. Acad. Sci. 15: 393, fig. 1. 1925. HOLOTYPE: EL SALVADOR. San Miguel: cult. at Sucesión Letona, Oct. & Nov. 1923, Milner s.n. (ILL, fide Trelease).

Agave vivipara var. letonae (F. W. Taylor ex Trel.) P. I. Forster, Brittonia 44: 74. 1992.

Agave angustifolia var marginata hort. ex Gentry, Agaves Continental N. Amer. 564. 1982. TYPE: None designated.

Agave vivipara cv. Marginata (Hort. ex Gentry) P. I. Forster, Brittonia 44: 74. 1992.

Agave angustifolia var. nivea (Trel.) Gentry, Agaves Continental N. Amer. 566. 1982.

Agave nivea Trel., Trans. St. Louis Acad. Sci. 23: 143, pl. 22. 1915. HOLOTYPE: GUATEMALA. Dept. Progreso: about El Rancho, Apr 1915, Trelease 11 (ILL, fide Trelease).

Agave vivipara var. nivea (Trel.) P. I. Forster, Brit-tonia 44: 74. 1992.

Agave angustifolia var. rubescens (Salm-Dyck) Gentry, Agaves Continental N. Amer. 566. 1982.

Agave rubescens Salm-Dyck, Hort. Dyck. 8: 306. 1834. NEOTYPE: Annual Rep. Missouri Bot. Gard. 18: pls. 32-34. 1907 (designated by Gentry, 1982: 566).

Agave vivipara var. rubescens (Salm-Dyck) P. I. Forster, Brittonia 44: 74. 1992.

Agave angustifolia var. sargentii Trel., Annual Rep. Missouri Bot. Gard. 22: 99. 1911. LECTOTYPE: Annual Rep. Missouri Bot. Gard. 22: pls. 100-101. 1911 (designated by Gentry, 1982: 567).

Agave vivipara var. sargentii (Trel.) P. I. Forster, Brittonia 44: 75. 1992.

Agave angustifolia var. variegata Trel., Annual Rep. Missouri Bot. Gard. 19: 287. 1908. TYPE: None designated.

Agave vivipara cv. Variegata (Trel.) P. I. Forster, Brittonia 44: 75. 1992.

Acknowledgments

We thank Gerrit Davidse (MO), Tom Wendt (TEX), Lawrence Kelly (MEXU), and Mark Olson (MEXU) for their many useful comments and suggestions. Lourdes Rico (K) kindly provided literature and important information concerning that literature.

Literature Cited

Alvarez de Zayas, A. 1995. Los agaves de las Antillas. Bol. Soc. Bot. México 57: 37-48.

Berger, A. 1915. Die Agaven: Beiträge zu einer Monographie. Fischer, Jena. [Reprint 1988. Fischer, Stuttgart.]

Colunga-García Marín, P. 1998. Origen, variación y tendencias evolutivas del henequén (Agave fourcroydes Lem.). Bol. Soc. Bot. México 62: 109-128.

---, E. Estrada-Loera & F. May-Pat. 1996. Patterns of morphological variation, diversity, and domestication of wild and cultivated populations of Agave in Yucatán, Mexico. Amer. J. Bot. 83: 1069- 1088.

--- & F. May-Pat. 1997. Morphological variation of henequén (Agave fourcroydes, Agavaceae) germplasm and its wild ancestor (A. angustifolia) under uniform growth conditions: diversity and domestication. Amer. J. Bot. 84: 1449-1465.

Commelin, C. 1703. Praeludia botanica. F. Haringh, Lugduni Batavorum.

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

García-Mendoza, A., P. Colunga-García Marín & R. Bye, Jr. 1993. Los usos de Agave angustifolia Haw., ancestro silvestre del henequén en su a´rea de distribución geográfica. Pages 92-112. In: P. Peniche R. & F. Santamaría B., editors. Memorias de la Conferencia Nacional sobre el Henequén y la Zona Henequenera de Yucatán. Gobierno del Estado de Yucatán, CONACYT, Universidad Autónoma de Yucatán, Sostenibilidad Maya, Instituto Nacional de Investigaciones Forestales y Agropecuarias. 25- 28 Oct. 1992. Mérida, Yucatán, Mexico.

Gentry, H. S. 1982. Agaves of continental North America. University of Arizona Press, Tucson.

Haworth, A. H. 1812. Synopsis plantarum succulentarum, p. 72.

Hodgson, W. C. 2001. Food plants of the Sonoran Desert. University of Arizona Press, Tucson.

Howard, R. A. 1979. Flora of the Lesser Antilles, Leeward and Windward Islands. Vol. 3. Monocotyledonae. Arnold Arboretum of Harvard University, Jamaica Plains, Massachusetts.

Hummelinck, P. W. 1936. Notes on agave in Aruba, Curacao, Bonaire and some parts of the South American continent. Recueil Trav. Bot. Ne´erl. 33: 223-249, 8 tab.

---. 1938. Notes on agave in the Netherlands West Indies and North Venezuela. Recueil Trav. Bot. Ne´erl. 35: 14-28, 4 tab.

---. 1993a. Agaven op Curaccao, Aruba en Bonaire

---. 1993b. Agaven op Curaccao, Aruba en Bonaire

---. 1993c. Agaven op Curaccao, Aruba en Bonaire

Linnaeus, C. 1753. Species plantarum. L. Salvi, Holm.

1. Succulenta (Netherlands) 72: 5-11.

2. Succulenta (Netherlands) 72: 104-109.

3. Succulenta (Netherlands) 72: 214-223.

Smith, G. F. & E. M. A. Steyn. 1999. Agave vivipara: the correct name for Agave angustifolia. Bothalia 29: 100.

Steyn, E. M. A. & G. F. Smith. 2000. Agave vivipara: a naturalised alien in Southern Africa. Bothalia 30: 43-55.

Targioni Tozzetti, O. 1810. Observationum botanicum. Ann. Mus. Imp. Fis. Firenze 2: 1-103.

Thiede, J. 2001. Agavaceae. Pages 5-102. In: U. Eggli, editor. Illustrated Handbook of Succulent Plants: Monocotyledons. Springer-Verlag, Berlin & Heidelberg.

Trelease, W. 1908. Agave rigida-Furcraea rigida-Agave angustifolia. Annual Rep. Missouri Bot. Gard. 19: 273-287 + 7 pl.

---.1913. Agave in the West Indies. Mem. Natl. Acad. Sci. 11: 3-299.

Wijnands, D. O. 1983. The botany of the Commelins. Balkema, Rotterdam.

© Brittonia, 2003