root and tuber crops consumed by the farm family. The species grown was Cush-cush (D. trifida), which was planted in March- April 1983 and harvested in December 1983-February 1984. Operations during the crop were two hand-weedings, one fertiliz- ing (57g NPK/plant) and one staking, using only family labour. The role of the yam in that farm system was to diversify pro- duction, to provide food, to avoid total dependence on the bananas and to utilise farm labour at a time when the bananas could not employ the available labour. The latter was during the drier parts of the year when the banana harvests are less frequent and weed growth slower. Table 3, based on the data from an annual longitudinal survey of the farm, shows a partial economic analysis of yam production. Attention was directed to the cash flow into and out of the system. TABLE 3. A summary cash analysis of commercial yam production on 0.1 ha of a small farm in Dominica. CASH INPUTS VALUE EC$ 0 Fertilizers 40.00 transportation 40.00 labour no cash paid (family labour) 740 m/h Total 80.00 OUTPUT Total Unit Value Total Value yield 2273 sold 2136 1,76 3,760.00 home use 137 TOTAL CASH INCOME 3,760.00 TOTAL CASH SPENT 80.00 NET CASH INCOME $3,680.00 Productivity of labour: Total yield = $5.41 per man/hr Total labour 740 m/hr On this farm a total of 2,600 man-hours were spent on the bananas which brought in $12,000 during that year. The produc- tivity of labour on the bananas was calculated at $4.60/man hour, which was less than that of the yams. On Station Tests Consistent with the farming system research methodology, (Fig. 1), the initial planting material introduced into Dominica was evaluated largely at La Plaine Agricultural Station, a southeast coastal area. This material included anthracnose resis- tant cultivars from INRA, Guadeloupe, and virus tested material from CARDI, Barbados (Table 4.) At La Plaine, total annual precipitation, mean monthly maximum temperature, mean monthly minimum temperature and monthly wind run ranged FIGURE 1. Process for introduction of yam into Dominican farm systems. 1 On F F.T. / On Farm Validation 'arm Tests Applicability Testing \On Farm P roductr.ion Systren5 Analysis MOA EXTENSION training training CARD FUTURE WORK TABLE 4. Description of yam planting material introduced to Dominica 1981-84. White Lisbon A Belep SEA 189 V 1712 White Lisbon B White Lisbon C Oriental B Coconut Lisbon B White cush-cush Crop Species D. alata D. alata D. alata D. cayenenss D. alata D. alata D. alata D. alata D. trifuda Lisbon D. alata CARDI Barbados INRA Guadeloupe INRA Guadeloupe INRA Guadeloupe CARDI Barbados CARDI Barbados CARDI Barbados CARDI Barbados INRA Guadeloupe CARDI Barbados VT AT AT AT VT VT VTAT VT VT 1981 1982 1982 1982 1982 1982 1984 1984 1984 1984 Barbados New Caledonia Phillipines N.A. Barbados Barbados Barbados Barbados N.A. N.A. Quantity Year Origin Character Source Cultivar 12 kgs 25 kgs 75 kgs 25 kgs 46 kgs 46 kgs 5.0 kgs 46 kgs VT = virus tested; AT = anthracnose tolerant; N.A. = not applicable; A, B, and C = 1st, 2nd and 3rd crop harvested by certified farmer after initial green house source. PROCEEDINGS of the CARIBBEAN FOOD CROPS SOCIETY-VOL. XX .CL I 256