From Standards to Practices: The Intensive and Improved Rice Systems (SRI and SRA) in the Madagascar Highlands

Abstract

In the Betsileo Highlands of Madagascar, the supposedly great potential of SRI (System of Rice Intensification) and SRA (Improved Rice System or systeme de riziculture amelioree) has not been concretized by massive adoption by farmers. The lowest adoption rate occurred in and around forest protected areas where SRI extension was implemented early on to help farmers face conservation constraints. The hypothesis was put forward that the SRI model has either encountered specific ecological and socioeconomic constraints or has not matched the promises made by those who promoted it. An in situ agronomic survey at field level (crop management sequences, yield component analysis) was implemented on random samples of farmers’ fields, with about fifteen fields per system. Three systems were compared: neither SRI nor SRA (control S1), application of SRI (S2), application of SRA (S3) at two sites (near the forest, near the markets) for 3 years. The control S1 was more intensive and higher-yielding than SRI/SRA promoters indicated: 4.2 t/ha against 2 t/ha in official discourses. The management of S2 and S3 matched the prescribed models SRI and SRA for key features overall. However, S2 generally received more organic manure and more fertilizers than S1 and S3 and benefited from the best preceding crops and the best paddies. The average yield of practiced SRA (S3) was not different from S1, because S1 had sometimes already adopted some features of SRA. Near the forest, with peat soils and mainly mineral fertilizing provided by conservation projects, there was no difference between S1 and S2 yields in 2006 (drought in vegetative phase), therefore the 2007 yield of S2 exceeded S1 by 14 % (only due to bigger ears). Yet near the markets in 2008, with mainly mineral soils, counter-season crops, earlier rains, and more manure, the S2 yielded +40 % more than S1 with more panicles. The average gain of +24 % included effects of fertilizer and manure additions, and so the real SRI effect, ceteris paribus was much lower. This low attractiveness, the associated risks, the cost of specialized hired labor, and the lack of manure and fertile mineral soils probably reduced SRI adoption by the poor farmers of the forest and forest edge. Near roads and markets, more people had enough manure and money for hiring workers to invest and run the risks of SRI. But the real SRI effects have yet to be checked, in ceteris paribus yields, environmental and social benefits.