How farmers permanently adapt to climate evolution by testing new options and caring for food security: case of long-cycle sanyo millet comeback in Serer area in Senegal. [P-3330-58]

During the last decade, Serer farmers of the Sine region in the central and western part of Senegal have started to grow again the sanyo millet (Pennisetum glaucum), a long cycle (110-140 days) and photoperiodic traditional variety that had disappeared for 30 years from this region due to the rainfall decrease which has affected the Sahelian and Sudano-Sahelian zones starting from 1970, leaving only the short-cycle (90 days) souna millet in the fields. We made the assumption that the reintroduction of the sanyo millet could be an agronomic “marker” of the increase in rainfall observed in Senegal since the mid-1990s (Salack and al., 2011) attesting to the capacity of farmers to adapt to the evolution of their environment. We wanted to check, however, whether this necessary climatic opportunity was sufficient to explain farmers choices. We investigated how important was the sanyo comeback in local farming systems, its geographical diffusion, and its biophysical, economic, social and cultural drivers. We carried out (a) simulations of souna and sanyo annual development during the 1950-2013 period using the SarraH©Cirad model; (b) several Focus Group Discussions with farmers; (c) a large survey by questionnaire on farming systems among 1,061 farms in the 30 villages of the IRD human and health observatory zone of Niakhar (monitored since 50 years); and (d) a rapid survey in 240 villages of the region located between Bambey and Diourbel in the North and Fatick in the South (about 1000 km2) on whether the sanyo millet was grown and its date of reintroduction. The SarraH©Cirad crop model was parametered according to previous works. Survey data were carefully analysed using relevant statistics to assess the factors underlying sanyo reintroduction. Crop model simulations confirmed that sanyo reappearance is due to the recent rainfall improvement which now allows getting again grains with this variety whereas since 1970 grains production was only possible with souna. But simulations also show that sanyo yields remain very risky due to the rainfall interannual variability whereas souna yields are higher and surer. They also show that sanyo provides important biomass (stalks and straw). Farmers comments and surveys data analysis complement and confirm the results of the simulations. First, peasants report that they prefer the taste of sanyo and above all the quality of its stems rather than those of souna. In addition, adopting sanyo does not require any specific know-how or investment as it is grown like souna. Its qualities explain why it has quickly spread throughout the area. Sanyo was present in 61% of the sites surveyed in 2013 compared with 23% in 2000. But farmers stress that it is a risky cultivation and that they will not endanger their food security by substituting souna with sanyo. Souna and groundnuts still remain their main staple and cash crops, respectively. They also deplore that sanyo (similarly to souna) cannot provide cash, unlike groundnuts or watermelon that has also expanded in recent years. Statistical analyses highlight some social and family influences since farmers belonging to the “warrior caste” and those whose fathers formerly cultivated sanyo are more likely to cultivate it. But “land resources” appear to be the key factor. Yet, farmers explain they would plant fields dedicated to sanyo only if they were sure to produce enough souna. This is why sanyo is mainly cropped in association with souna, in an average proportion of 1 line of sanyo for 4-5 lines of souna. Moreover farmers point out that they would easily stop sanyo if they had other opportunities, particularly if they provide cash. Thus, despite its spectacular diffusion, sanyo cover very little surfaces: in the zone of Niakhar it was present in 2013 on 2.8% of the cropped areas and on 7.3% of millet areas. Sanyo reappearance and its important and rapid diffusion is clearly an agronomic “marker” of the recent climate evolution (rainfall increase) observed in Senegal. It attests to the adaptive capacity of farmers to quickly and autonomously adapt to the evolution of their environment by permanently looking and trying new options, but underlines how cautious they are to not endanger their food security, and confirms that a climatic opportunity is not sufficient to account for farmers' choices. Comparisons can be made with recent watermelon emergence in the same area and developments of maize and rainfed rice areas in the southern areas of Senegal, which have been enabled by the rainfall evolution but benefit of solid economical drivers. (Texte integral)