194 

 many residents delayed planting, replanted, or abandoned their gardens. Some garden crops dried up, 

 while weeds grew faster than did crops in other gardens. Although harvests were not measured, 

 gardeners reported that only about 10-20% of the gardens produced an average crop in 1989, while the 

 other gardens produced below-average yields. 



Due to the irregular nature of the rains in 1989, data for the timing of horticultural activities 

 that year were limited to the date of first planting. The median first date of planting was 5/12-5/13 

 (n = 150 gardens). The earliest reported planting was on 4/1, while the latest first planting was on 

 8/7. At least seven gardens were completely replanted once and two were completely replanted twice. 

 Other gardens were partially replanted as large numbers of seedlings died due to the irregular and 

 inadequate rains. 



The 1990 annual horticultural cycle, according to residents, was about normal and most 

 gardeners achieved an average harvest. The median date to begin initial site clearing was 2/15 (range 

 = 9/15/89-6/1/90, n = 40 gardens; month/date/year). The median date to bum the garden site was 74 

 days later, on 4/30 (range = 3/1-6/30, n = 40 gardens). The median date to plant was 32 days later, 

 on 6/1-6/2 (range = 4/30-7/15, n = 40 gardens). Weeding and doubling took place during 8/90- 

 10/90. The com harvest began about 10/30 and was complete by about 12/8 (n = 7 gardens). 



Garden type and years of previous use . Most gardens were planted on sites that had not been 

 used as gardens the previous year. In 1989, the average number of years that a site had been used 

 previously as a garden was 0.1 (SE = 0.05, n = 150 gardens, range = 0-5 y). About 91% of the 150 

 gardens were on sites categorized as new, while 9% were on sites categorized as old. In 1990. the 

 average number of years that a site had been used previously as a garden was 0.0 (SE = 0.02, n = 40 

 gardens, range = 0-1 y). About 98% of the 40 gardens were on sites categorized as new, while 2% 

 were on sites categorized as old. The number of years of previous use for gardens was not 

 significantly different between years (x' approximation = 1.7769, d.f. = 1, P = 0.1825). 



One gardener indicated that he had cleared a field in Late Secondary Forest in 1989. All other 

 gardeners reported that they had cleared fields in Early Secondary Forest. The importance of this is 



