Application of system of rice intensification practices in the arid environment of the Timbuktu region in Mali

摘要

Cereal production is chronically deficit in the Timbuktu region of Mali, sufficient for only 4.5 months of annual household consumption. Small-scale, village-based irrigation schemes, usually 30–35 ha in size, irrigated by a diesel motor pump, have become important to improve food security in this arid region. The NGO Africare has worked during the past 12 years with farmers in Goundam and Dire circles to establish irrigation schemes and provide them with technical assistance. In 2007, Africare undertook a first test of the System of Rice Intensification (SRI) in Goundam circle. After farmers observed a yield of 9 t ha−1 of paddy compared to 6.7 t ha−1 in the control plot there was interest in larger scale testing of the SRI system. In 2008, Africare, in collaboration with the local Government Agriculture Service and with support from the Better U Foundation, implemented a community-based evaluation of SRI with 60 farmers in 12 villages. Farmers in each village selected five volunteers, who each installed both SRI and control plots, side by side, starting the nurseries on the same day and using the same seed. For SRI plots, seedlings were transplanted one plant hill−1 at the two-leaf stage (on average, 11.6 days old), with spacing of 25 cm × 25 cm between hills and aligned in both directions. This allowed farmers to cross-weed with a cono-weeder, on average 2.4 times during the season. In the control plots, farmers planted 3 plants hill−1 with seedlings 29.4 days old and spaced on average 23.7 cm, not planted in lines. Weeding was done by hand. 13 t ha−1 of organic matter was applied under SRI management, and 3 t ha−1 in the control plots. Fertilizer use was reduced by 30% with SRI compared to the control. Although alternate wetting and drying irrigation is recommended for SRI, this was not optimally implemented due to constraints on irrigation management within the scheme; thus water savings were only 10% compared to the control. Average SRI yield for all farmers reached 9.1 t ha−1, with the lowest being 5.4 t ha−1 and highest being 12.4 t ha−1. SRI yields were on average 66% higher than the control plots at 5.5 t ha−1, and 87% higher than the yields in surrounding rice fields at 4.9 t ha−1. Number of tillers and panicles hill−1, number of tillers and panicles m−2, and panicle length and number of grains panicle−1 were clearly superior with SRI compared to control plants. Farmers tested five varieties, all of which produced better under SRI. The SRI system allowed for a seed reduction of 85–90%: from 40–60 kg ha−1 for the control plots to 6.1 kg ha−1 under SRI. Although production costs per hectare were 15% higher for SRI, revenue was 2.1 times higher than under the control. Farmers were very satisfied with these results. In 2009/2010, Africare and the Government’s agriculture service worked with over 270 farmers in 28 villages to scale up SRI practices and to test innovations, including composting techniques, optimization of irrigation, and techniques to reduce labor requirements and production costs. The good crop performance along with other advantages was confirmed in this third year with SRI yields of 7.7 t ha−1 (n = 130 farmers) compared to 4.5 t ha−1 in farmers’ fields.