Economic Botany, 38(4) 417-432 (Oct-Dec 1984)
New Evidence for Ancient Cultivation of Canna edulis in Peru
Donald Ugent, Shelia Pozorski and Thomas Pozorski

Preserved rhizomes of Canna edulis from 5 archaeological sites in the Casma Valley of Peru are illustrated and described. These were identified by their extant surface features and by their starch grains, which are different from those of any other known flowering plant species. Based on radiocarbon assays, these specimens range in date from 2250-295 B.C. Materials used for comparative purposes in this study included a collection of edible canna from the Peruvian archaeological site of Pachacamac and a modern-day specimen. A theory on the place of origin and time of domestication of this species is given in the conclusions of this paper.

The achira plant (Canna edulis Ker-Gawler), a domesticate of the New World, produces an edible, scale-covered rhizome that is divided into several tuber-like segments. These may be trimmed and cut apart with a knife and prepared for table use by baking or boiling. Nutritionally, the achira rhizome is valuable largely as a source of starch. The area of indigenous cultivation of achira extends from Mexico and the West Indies, southward to Argentina and Brazil (Standley and Steyermark, 1952; Kriinzlin, 1912). However, the species has also been introduced to Hawaii, Africa, Asia and Australia. In these places it is utilized primarily in the manufacture of commercial starch, or as a stock feed (Purseglove, 1972; Schery, 1972).

Throughout its range in the New World, achira is grown primarily for home use in dooryard gardens or small outdoor plots. In Peru, however, where this food plant is most appreciated, it is grown frequently in sizable plantations. Gade (1966), for example, reports extensive cultivation of achira in the Upper Apurimac Valley of Peru at elevations between 2,400 and 2,600 m. He also reports that it is common near Ica, where it is grown near sea level in oases along the desert coast. Sauer (1950) reports that it is commonly sold in marketplaces along the coast.

* O. F. Cook (1925) had previously recognized Peru as
a center of domestication for this and other crop plants.

Achira has had a long history of cultivation in South America. Vavilov (1935),* who pinpointed the centers of present-day genetic diversity for this and many other crops, proposed that achira arose somewhere in the central Andes. But while some later authors have agreed with him on this point, others have not. Thus, Herrera (1942 a,b) suggests the mountains of southern Peru as the most probable place for the domestication of this species, whereas Cohen (1978) favors the northern desert coast. Gade (1966) and Sauer (1952), on the other hand, opt for a more northerly, Colombian rainforest or montania origin. Of the various areas suggested, the only one that has so far yielded any archaeological evidence is the very arid Peruvian coast.

In the present account, a series of previously undetermined archaeological samples of achira rhizome, excavated from sites in the Casma Valley of Peru, are described and illustrated. These materials, of Late Preceramic (2250-1750 B.C., uncorrected radiocarbon scale) and later origin, represent some of the oldest samples of this species currently known to science. Their discovery fills a gap in our knowledge of the distribution of achira in ancient times, and adds a measure of support to the idea that achira was originally domesticated in the area that now forms the desert coast of Peru.