Spatial identification of the crop-fallow rotation cycle and potential capacity for regeneration of fallow in northern Laos by satellite imagery

　　In northern Laos, shifting cultivation, namely slash-and-burn agriculture, has been conventionally practiced for upland rice (Fig. 1). However, the crop-fallow rotation cycle is tending to shorten due to forest conservation policies and population pressure, causing deterioration of productivity and in turn affecting farmers’ livelihoods in the region. To investigate the land condition in these areas, we developed a robust method of identifying the crop-fallow rotation cycle spatially using periodical observed satellite imagery, i.e., Landsat/Thematic Mapper (TM) and Enhanced Thematic Mapper + (ETM+). The study site was an area of 442,000 ha within N19˚30’; E101˚45’; – N20˚00’; E102˚30’ in Luang Prabang Province. 

　　According to the typical calendar for slash-and-burn agriculture in northern Laos, land is cleared from February to early April, and burned in March and April. Sowing is practiced from late April to June, and harvesting is from September to November. The most drastic change in the land surface is slashing to clear the fallow plants in the late dry season: the area used for cropping in the imminent rainy season changes to non-vegetation as a result of slashing for land preparation, but fallow areas show uninterrupted vegetative growth. The classification of vegetation/non-vegetation is one of the most reliable applications using remote sensing; therefore, land use in each year was identified by the presence of vegetation in the late dry season. In this study, 8-scene imagery of TM and ETM+ acquired annually from 1995 to 2003 were applied.

　　The results of classification of 8 scenes were represented by an 8-digit code to track the land use change year by year. For 1995-2003, approximately 77,000 ha (17.3%) had never been cropped but 41,000 ha (9.2%) had been used for cropping every year. The former was regarded as forest and the later as sedentary farms growing rice. The rotated area between cropping and fallow was classified by crop intensity. The areas cropped 1-2 times, 3-4 times, and 5-6 times occupied 129,000 ha (29.1%), 83,000 ha (18.7%), and 54,000 ha (12.2%), respectively (Fig. 2).

　　Vegetation in fallow shows a succession from shrub to bush, and ultimately to forest. Normalized Difference Vegetation Index (NDVI), derived from TM and ETM+, is known to be related to biomass. Hence, NDVI was applied to assess the process of plant regeneration in fallow. Fallow was classified by fallow length (age) from the 1st to 7th years and the mean NDVI was calculated for each fallow age. The mean NDVI increased constantly with fallow age, and it was estimated that 11 years was needed to reach to the same NDVI as that of forest. However, the cropping intensity mentioned above showed that long-term fallow exceeding 7 years accounted for only 17.3% of the area, implying that a large area might be re-used before vegetation is sufficiently regenerated. In addition, the potentiality for vegetation regeneration in fallow was assessed by comparison to the mean NDVI corresponding to the same fallow age. The results indicated that low-potential areas showed a lower NDVI than the mean for 66,000 ha, confirming that low-potential area increased as cropping intensity increased.

　　Due to restrictions on slash-and-burn agriculture, farmers in northern Laos have been urged to switch to more productive and conservational farming systems. The method and spatial information provided in this study will be useful for regional scaled land resource management.

(Y. Yamamoto)

Fig. 1. View of hills in northern Laos where slash-and-burn agriculture is implemented. Fig. 2. Cropping intensity for 1995-2002.

For further details log on website :
https://www.jircas.affrc.go.jp/english/publication/highlights/2005/2005_04.html