This is the old United Nations University website. Visit the new site at

Contents - Previous - Next

Appendix: upland rice yields in Northern Thailand

Subsistence crop yields are perhaps the one area of concern where agricultural specialists remain the most sceptical about the possibility for the improvement of swidden agriculture. That cash crops can be profitably grown in small plots, permanent or rotated, in the mountain areas of Northern Thailand seems hardly in question now after the excellent work of the UNPDAC Project (UNPDAC Report Series 1973-1979). Crops grown for subsistence (rice in Northern Thailand), however, cannot justify very expensive inputs, and it is perhaps the difficulty of finding innovative, low-expense inputs that has led some specialists to envision instead an altitude-interdependent Northern Thailand, with rice being sent up from the valleys and cash crops sent down from the hills. Eventually this may come to pass if people's real needs bear out this vision, but in the meantime the overwhelmingly felt need in the hills is to increase rice yields. In a recent "problem census'' of one area, rice problems seemed paramount, and in this category low yields accounted for 86 per cent of the problems cited by the villagers for which they desired assistance (USDA 1979).

Since swidden farmers face a variety of rice-growing problems unaided by modern technology, yields are highly variable. Insect damage may ruin a crop, or lower temperatures at higher altitudes may reduce yields due to spikelet sterility (TAHAP 1979, pp. 6-7). Kunstadler (1978, p. 109) found swidden rice yields among Lua and Karen swiddeners in an area of Mae Sariang averaging about 1,000 kg/ha (weighed several months after harvest but not adjusted to standard moisture content), comparing favourably with rainfed lowland paddy (but only about half the yield of lowland irrigated rice). Hinton (1978, p.194) found slightly higher yields among the Karen he studied-about 1,450 kg/ha (threshed, unhusked, moisture content unknown). However, both these figures were measured in the late 1960s when swidden cycles were probably longer. Yields of less than 800 kg/ha are certainly not unknown today, whereas swidden rice yields in excess of 1,500 kg/ha would now be quite uncommon in most hill areas of Northern Thailand.28

The question is, then, is it reasonable to expect, for rice in Northern Thailand, that some careful, low-cost inputs may indeed be able at least to double swidden yields as suggested by Greenland (1975)? The difficulty in answering this question is due to the lack of data on minimum, low-cost inputs under native swidden cultivator conditions (although there is absolutely no reason why such trials could not be cheaply and easily conducted). That rice yields can be raised on permanent upland plots, even on relatively poor soils in Northern Thailand, with inputs that non-swidden (Northern Thai) farmers can afford has been demonstrated by the TALD Project. In several years upland rice yields rose spectacularly to a mean yield of nearly 1,600 kg/ha, with a few farmers obtaining nearly 3,000 kg/ha (Chapman 1973, pp. 200202).

To examine whether or not yields can be raised in a rotational system, however, we need to look at "one year only" data. Trials of this nature have been conducted in native swidden areas (although not wholly under native cultivator conditions) by the TAHAP Project (1979) and by the Samoeng Royal Agricultural Research Station in conjunction with the Sanpatong Rice Experimental Station (Mr. Vitoon Khunthicula, personal communication). Using specially crossed, locally adapted new rice varieties, the latter experiments were able to produce over 5,300 kg/ha (standard moisture content) in a one-year trial. Inputs were 187.5 kg/ha triple phosphate plus 62.5 kg/ha urea applied two times (in field preparation and at planting), plus a basic dressing of manure and ash. The TAHAP Project (1979, pp. 5-8), in initial trials, used far fewer inputs, obtaining a range of responses, but with a similar maximum yield of 5,300 kg/ha.

The inputs used in these trials are more than most shifting cultivators could afford, but they do indicate at least the possibility of obtaining somewhat less spectacular but meaningful results if Greenland's recommendations were to be seriously explored. As the IRRI upland rice report (1975, p. 11) puts it: "The yields are generally low, but can be increased by the development of improved varieties and cultural practices to suit the soil, climate, and social conditions. These improvements are entirely possible through research and extension, and through changes in national policies."

Contents - Previous - Next