The Domestication of Henequen (Agave fourcroydes Lem.)

Transcription

1 Chapter 24 The Domestication of Henequen (Agave fourcroydes Lem.) Patricia Colunga-GarcíaMarín INTRODUCTION Henequen (Agave fourcroydes Lem.) is a cultivar of the Agavaceae family that has been cultivated for its fiber in the Maya area since pre- Hispanic times. Henequen is a different species from sisal (A. sisalana Perrine), even though the fiber of both plants is named sisal hemp in English. Sisal is also a Mexican cultivar, probably originating in the state of Chiapas (Gentry 1982). Sisal is presently cultivated in Africa and Brazil, while henequen originated in the Yucatán Peninsula and, at present, is cultivated only in Mexico and Cuba. The main morphological difference between both species is that sisal lacks lateral spines in its leaves (teeth), while henequen has a coarser fiber than sisal. The evolutionary history and diversity of this crop before the arrival of the Spanish is not known, but several illustrations in the Dresden and Tro- Cortesian codices (presumably made in the thirteenth and fifteenth centuries, respectively) indicate the social importance that fiber production had for the lowland ancient Maya. Irigoyen (1950) reproduced some of these illustrations from the codices, which show several human figures hunting, fishing, or carrying bags for trade, while using cordage or nets that the author asserts were made from Agave fibers. In addition, Irigoyen (1950) included an illustration of the God F where the God is holding what is probably a bundle of Agave fibers. What we do know for certain is that at the moment of the Maya-Spanish contact, henequen was already a domesticated plant. In his work Relación de las Cosas de Yucatán, Fray Diego de Landa said that henequen was cultivated in house-gardens, and it was of much better quality than the wild (de Landa [1566] 1978:128). 439

2 440 THE LOWLAND MAYA AREA But what is the wild ancestor of henequen? According to Gentry s (1982) classification of the Agaves of North America, all the wild Agave populations in the Yucatán Peninsula belong to A. angustifolia Haworth. This is the species most widely distributed of the genus. It is found from Costa Rica, on both Atlantic and Pacific coasts, to Tamaulipas and northwestern Sonora in Mexico. The major vegetation formations in which it occurs are tropical savannah, thorn forest, and deciduous tropical forest, at heights from sea level to 1,500 meters (or sometimes higher). Agave angustifolia is an extensive variable species. Gentry (1982) could not separate subspecies populations consistently with any combination of characters; in this species he included 21 synonyms. Gentry considered that variation of this species illustrate that it is a freely-seeding outbreeding complex, which has been widely assorted by circumstances of habitat, changing climates over a long period of time, and human s interventions (Gentry 1982). According to Gentry (1982), four cultivars selected for fiber seem to have appeared from this gene reservoir: (1) A. angustifolia var. deweyana (Trel.) Gentry, named zapupe verde, and grown in Tamaulipas and Veracruz states, Mexico; (2) A. angustifolia var. letonae (Taylor) Gentry, named agave letón, and grown in El Salvador and Guatemala; (3) A. angustifolia var. nivea (Trel.) Gentry, grown in Guatemala; and (4) A. sisalana Perrine, sisal, a possible hybrid between A. angustifolia and A. kewensis, which may have originated in Chiapas but is grown extensively in Africa and Brazil. To these four cultivars we should add (5) A. fourcroydes Lem., named Henequen, grown in the Yucatán Peninsula, Tamaulipas, and Cuba (Colunga-GarcíaMarín et al. 1999), and (6) the hybrid H [(A. amaniensis Trelease & Nowell x A. angustifolia) x A. amaniensis] (Lock 1962) extensively cultivated in Africa. The distribution of all these fiber cultivars in Mesoamerica coincides with the Maya area. From this same gene pool, and from outside the Maya area, several cultivars have been selected that were important as a food source in pre- Hispanic times. Today, they are important sources of alcoholic beverages such as tequila (A. tequilana Weber) and the mezcales, the most famous of which come from the Mexican state of Oaxaca. Gentry (1982) considers that the separation of A. tequilana from A. angustifolia is only nominal. This multiplicity of cultivars derived from the same gene pool makes A. angustifolia a singular case of multiple domestication of a species throughout its geographical distribution. This multiple domestication is probably associated with its use by different human cultures, in contrasting environments, and for different purposes. This domestication pattern is very different from that of other crops of Mexican origin, such as corn and common bean, for which a unique domestication area has been defined.

3 The Domestication of Henequen (Agave fourcroydes Lem.) 441 The results summarized in this chapter are derived from an investigation whose main objective has been to understand the origin, variation, and evolutionary trends of henequen under human selection. MATERIALS AND METHODS Three types of evidence were compiled about wild and cultivated populations: 1. Ethnobotanical evidence This evidence consists of the study of past and present man-agave interactions, which were analyzed by ethnohistorical sources and based on ethnobotanical exploration. 2. Morphological evidence This evidence was compiled through the numeric, statistic, and phylogenetic analysis of 66 morphologic characters evaluated in populations grown under both natural and homogeneous conditions. Characters included stem, leaf, inflorescence, flower, fruit, and seed characteristics. 3. Genetic evidence This evidence was obtained through the analysis of three isozymatic systems: Malate Dehydrogenase (MDH), Acid Phosphatase (ACP), and Cathodic Peroxidase (PRX). (Details of the materials and methods are explained in Colunga-GarcíaMarín & May-Pat 1993; Colunga-GarcíaMarín et al. 1993; Colunga-GarcíaMarín, Estrada-Loera, & May-Pat 1996; Colunga-GarcíaMarín & May-Pat 1997; and Colunga-GarcíaMarín et al ) RESULTS AND DISCUSSION Ethnobotanical evidence Ethnohistorical analysis indicated the lack of evidence about the diversity of henequen during the pre-hispanic Maya times, but it is assume to be at least equal to or larger than that recorded in the agronomy manuals published at the beginning of the twentieth century (de Echánove 1814; Regil and Peón 1853; Espinosa 1860; Barba ; Bolio 1914). These manuals included seven varieties of henequen as well as the experimental cultivation of wild plants. Analysis of these manuals suggests that, under traditional agriculture, Maya peasants selected varieties for their different type of fiber as well as their adaptability to soil and precipitation conditions. Our ethnobotanical exploration of the seven varieties revealed that only three of them could now be identified: (1) Sac Ki, or white henequen, the most

4 442 THE LOWLAND MAYA AREA abundant variety and the one preferred by the cordage industry; (2) Yaax Ki, or green henequen, with fiber very similar to that of Sac Ki, but of lower yield; and (3) Kitam Ki, or wild boar henequen, with a very soft fiber and low yields. This variety is very scarse. Sac Ki is the variety that corresponds with Gentry s (1982) diagnosis of A. fourcroydes; the other two varieties have not been formally described. Subdeciduous forest wild populations, called Chelem, are collected for their fiber and used for handicrafts. Artisans who use wild plant fibers distinguish three variants within these populations according to fiber quality: Chelem white is the one considered as more similar to cultivated variants, while both Chelem green and Chelem yellow are seen as of lower quality (in that order). Kitam Ki and wild population varieties have fibers that are preferred for textile use when there is a direct contact with the skin, such as in hammock and sandal manufacture (Colunga-GarcíaMarín & May-Pat 1993). The loss of henequen varieties during the end of the nineteenth century and the beginning of the twentieth century resulted from the establishment of extensive plantations of Sac Ki (white henequen), the variety preferred for the cordage industry, and the elimination of the other existing varieties. The cordage industry was strongly developed in the Yucatán Peninsula as a response to the great demand for binder twine by the wheat harvest in the United States. This great demand was a consequence of the adaptation of binder twine to a harvesting machine for cutting and threshing grain in the field in The agronomy manuals at the beginning of the twentieth century explicitly encouraged producers to cultivate Sac Ki and to eliminate all other varieties. The loss of henequen varieties was accelerated by the invention in 1895 of the fiber-extracting machine for rasping and cleaning henequen, which imposed very narrow criteria for the length of leaves that could be rasped. This length was specifically that of Sac Ki. Henequen genetic erosion was hastened by the exclusive practice of vegetative propagation, which ensured that only Sac Ki germ plasm could be propagated (Colunga- GarcíaMarín and May-Pat 1993). In this way, the evolutionary trends that were not oriented towards cordage use and adaptation to the rocky soils of the north lowland Maya area were eliminated. Ethnobotanical research uncovered 41 traditional uses for henequen. These uses comprise all morphological structures and all the anthropocentricuse categories, including food use, which was so important as a domestication incentive in other areas of Mesoamerica. The specialized used of only one product the fiber is a result of exclusive exploitation by the cordage industry (Colunga-GarcíaMarín and May-Pat 1993; Colunga-GarcíaMarín et al. 1993).

5 The Domestication of Henequen (Agave fourcroydes Lem.) 443 Morphological evidence Morphological variation analysis under uniform grown conditions indicated that cordage-cultivated varieties (Sac Ki and Yaax Ki) differ from wild populations in similar direction and magnitude, a trend that may be described by four domestication syndromes (i.e., a combination of characteristics with anthropocentric interest, or related to the process of artificial selection). These syndromes include greater fibrosity, less thorniness, less reproductive capacity, and gigantism. Greater fibrosity may be illustrated by the linear relationship that exists between the fresh leaf mass and the dry leaf mass. For a given fresh leaf mass, cordage-cultivated plants have 1.9 times more fiber than wild plants. This difference suggests that domestication has changed the physiological resource allocation strategy of wild plants, resulting in a more fiber-rich domesticated plant that can be harvested more efficiently. Something analogous can be observed with the other domestication syndromes. Less thorniness may be evaluated through the relationship between leaf length and the number of teeth. For a given leaf length, cordage-cultivated plants have 1.6 times less teeth than wild plants. In the same way, the diminished reproductive capacity may be illustrated by the linear relationship between the number of ovules and the number of normal seeds. For a given number of ovules, cordage-cultivated plants have 5.3 times less normal seeds than wild plants. Gigantism can be observed specially in the length and width of the leaves. Leaves from the domesticated varieties are 1.34 times larger and 1.85 times wider than those of wild ones (Colunga-GarcíaMarín, Estrada-Loera, and May-Pat 1996; Colunga-GarcíaMarín and May-Pat 1997). These syndromes have an obvious correlation with the cordage industry interests that have guided the process of henequen evolution during the last century. The textile-cultivated variant Kitam Ki is the one variant that most resembles wild plants. In a cluster analysis using UPGMA (Unweighted Pair- Group Method using Arithmetic averages) as a grouping method, Kitam Ki grouped with wild populations. This variant differs from wild populations in two out of the four domestication syndromes larger fibrosity (although in a much lesser degree than is observed in cordage variants) and a diminished reproductive capacity. These results suggest a process of artificial selection with an orientation and intensity mainly focused on textile use (Colunga- GarcíaMarín, Estrada-Loera, and May-Pat 1996; Colunga-GarcíaMarín and May-Pat 1997). With respect to A. angustifolia populations, morphological evidence suggests the existence of two ecotypes one including dunes and deciduous forest populations, and the other corresponding to subdeciduous forest populations. Subdeciduous forest populations have the longest fiber and the

6 444 THE LOWLAND MAYA AREA greatest quantity of fiber. Populations from the subdeciduous forest used by artisans for their fiber were grouped in a separate cluster using UPGMA as a grouping method. The artisans classification of Chelem varieties according to their fiber quality agrees with morphological evidence as long as the variety Chelem white has the longest fibers and the greater quantity of fiber, which exactly matches the artisans description (Colunga-GarcíaMarín, Estrada- Loera, and May-Pat 1996; Colunga-GarcíaMarín and May-Pat 1997). Genetic evidence Phylogenetic analysis using parsimony, based on morphologic and isozymatic data combined, resulted in a single most parsimonious tree (Figure 24.1). Two main lineages were revealed one for wild populations, and the other for extant henequen varieties used for their fiber. Within this last lineage, two other lineages were evidenced: (1) the lineage of cordagecultivated plants whose sister group is Chelem white, which is the variety recognized by artisans as the most similar to them; and (2) the lineage of the textile-cultivated variety. The position of the other Chelem varieties in the phylogram suggested their hybrid origin between wild and cultivated varieties. For all plants studied, the isozyme Malate Dehydrogenase (MDH) had the same electrophenotype, with the exception of the textile-cultivated variant Kitam Ki, which had an electrophenotype also found in four plants from the Mexican states of Veracruz and Oaxaca. On the other hand, the cordagecultivated variants Sac Ki and Yaax Ki showed identical electrophenotypes to those of several wild plants for the three isozymatic systems. These results support the hypothesis of a Yucatecan origin of Sac Ki and Yaax Ki and the introduction of Kitam Ki from outside of the peninsula. This hypothesis is also supported by ethnobotanical evidence because of the scarce knowledge that people have about Kitam Ki, as compared with Sac Ki or Yaax Ki. Isozymatic evidence, as a genetic diversity estimator, showed that while the wild variants had relatively high genetic variation, no variation was observed within the three henequen varieties. In A. angustifolia, partition of the variation levels indicated that 40% of the variation could be found between populations, and 60% within them. In contrast, all the variation in henequen varieties is found between populations, and none can be found within them. These results are to be expected in a cultivar that is always propagated vegetatively. The extremely small fraction of total isozymatic variation observed within cultivated varieties is, so far as we know, the most dramatic reported for a cultivar.

7 The Domestication of Henequen (Agave fourcroydes Lem.) 445 FIGURE Phylogenetic relations of extant henequen variants (SK, YK, and KK) and wild A. angustifolia populations (D, DF, SF, CHY, CHW, and CHG) from the Yucatan Peninsula, based on 66 morphological characters and 27 isozymatic bands. The single most parsimonious tree derived from an exhaustive search is shown. The tree was rooted using the D, DF, and SF cluster as an outgroup. Consistency index (Cl) = (Cl = when noninformative characters were excluded); homoplasy index (HI) = (HI = when noninformative characters were excluded); retention index (RI) = Tree length = 164 steps; length is shown above each branch. Bootstrap values (%) over 50, based on 1,000 replicates, are in parentheses below branches. Results obtained in this study of the cordage-cultivated varieties Sac Ki and Yaax Ki agree with the predictions of Doebley (1989) in cases of wildcultivated derivatives when isozymes are analyzed: (1) the cultivar falls within the variation range of the putative wild progenitor, (2) the cultivar has a subset of the allelic diversity found in the wild progenitor, and (3) the genetic variation is distributed in a different way. Cultivated plants have more variation between varieties than within them, while the opposite is true for wild plants.

8 446 THE LOWLAND MAYA AREA LITERATURE CITED Barba, R El henequén en Yucatán. Boletín de la Sociedad Agrícola Mexicana Bolio, A. J. A Manual práctico del henequén, su cultivo y explotación. Empresa Editorial Católica, S.A. Mérida, Yucatán. Colunga-GarciaMarín, P., and F. May-Pat Agave studies in Yucatan, Mexico I. Past and present germplasm diversity and uses. Economic Botany 47: Colunga-GarcíaMarín, P., and F. May-Pat Morphological variation of henequén germplasm and its wild ancestor under uniform growth conditions: diversity and domestication. American Journal of Botany 84: Colunga-GarcíaMarín, P., E. Estrada-Loera, and F. May-Pat Patterns of morphological variation, diversity and domestication of wild and cultivated populations of Agave in Yucatan, Mexico. American Journal of Botany 83: Colunga-GarcíaMarín, P., J. Coello-Coello, L. Espejo-Peniche, and L. Fuente-Moreno Agave studies in Yucatan, Mexico II. Nutritional value of the inflorescence peduncle and incipient domestication. Economic Botany 47: Colunga-GarcíaMarín, P., J. Coello-Coello, L. Eguiarte, and D. Piñero Isozymatic variation and phylogenetic relations between henequén Agave fourcroydes Lem. and its wild ancestor A. angustifolia Haw. American Journal of Botany 86: de Echánove, P.A Cuadro estadístico de Yucatán en Boletín de la Sociedad Mexicana de Geografía y Estadística, de Landa, D. [1566] Relación de las cosas de Yucatán. Editorial Porrúa, S.A. Mexico. Doebley, J Isozymic evidence and the evolution of crop plants. Pages in D. E. Soltis and P. S. Soltis, editors. Isozymes in plant biology. Advances in Plant Sciences Series 4. Dioscorides Press, Portland, Ore. Espinosa, J. D Manual de mayordomos de las fincas rústicas de Yucatán. Imprenta del autor, Mérida, Yucatán. Gentry, H. S Agaves of Continental North America. University of Arizona Press, Tucson, Ariz. Irigoyen, R Los Mayas y el henequén. Ediciones Zamná, Mérida, Yucatán. Lock, G. W Sisal. Twenty-five years Sisal research. Longmans, Green & Co. Ltd., London. Regil, J. M., and A. M. Peón Estadística de Yucatán. Boletín de la Sociedad Mexicana de Geografía y Estadística 3: