2 



AGRICULTURE HANDBOOK 495, U.S. DEPT. OF AGRICULTURE 



of two closely related wild species, 

 D. hamiltonii Hook and D. persi- 

 milis Prain et Burk., suggested 

 that D. alata had been selected 

 from these or from their hybrids. 

 Both are characterized by long, 

 deeply buried tubers that super- 

 ficially resemble some cultivated 

 but inferior varieties of D. alata. 

 Burkill hypothesized that deep 

 tubers evolved as a protection from 

 wild pigs and that human selection 

 resulted in the short-tubered, com- 

 pact varieties. For varieties with 

 upward-curving tubers, which 

 eventually push their way out of 

 the soil, he hypothesized another 

 human selection, since this growth 

 habit makes harvesting easy. 



There is little evidence elsewhere 

 to support BurkilFs hypothesis. 

 Wild species related to D. alata are 

 also found in Papua New Guinea. 

 Moreover, the variation in D. alata 

 is enormous, and some varieties 

 are unique to the region. Indo- 

 nesia also has extremely diverse 

 varieties that must have existed 

 for centuries. 



Relying on what is known about 

 historical movements of people in 

 Southeast Asia and on the scanty 

 information about distribution 

 and variation in yams, Alexander 

 and Coursey (1) placed the origin 

 of D. alata in an area lying be- 

 tween the distribution of D. hamil- 

 tonii (East India and West 

 Burma) and that of D. persimilis 

 (Indochina). The difficulty of 

 their theory is that cultivated 

 races in this area (roughly 

 Burma) do not include many ex- 



amples of the short-tubered, com- 

 pact types found in Indonesia 

 (especially Celebes) and Papua 

 New Guinea. 



It is highly likely that simple 

 or primitive forms of an initial 

 species, perhaps well distributed, 

 were selected locally to give rise 

 to the principal varietal types. This 

 could have happened independent- 

 ly on widely separated islands. 

 Interchange and hybridizations in- 

 creased variation. Inspection of 

 varieties from worldwide sources 

 suggests to the present author that 

 at least two distinct species were 

 involved in the evolution of D. 

 alata, in one of which the stem 

 was not winged; study of materials 

 on hand is needed to clarify this 

 problem. No systematic breeding 

 now appears possible, and existing 

 varieties appear to be ancient 

 dead-end relics of the evolutionary 

 process. A remarkable story of 

 ennoblement of the species thus 

 remains undisclosed. 



Geographic Distribution 



Dioscorea alata is a plant of the 

 hot humid Tropics; it seldom oc- 

 curs where cool temperatures or 

 dry periods prevail during the 

 growing season. There is no single 

 area, however, where D. alata is 

 the chief starchy food; it is almost 

 always utilized as one of several 

 farinaceous crops (other , yams, 

 cassava, sweetpotatoes, and 

 aroids) that are used to some ex- 

 tent interchangeably. 



In Southeast Asia D. alata is 

 grown chiefly as a dooryard crop. 



