Analysis of socio-ecological determinants for adoption of legume-cereals intercropping and improved seeds uses in the face of agricultural livelihood diversity in Satiri sub-district, Houet province, Burkina Faso

Abstract

Following the identification of the different agricultural livelihoods systems (ALS) type in the Satiri sub-district, western Burkina Faso, the presented study aims at assessing behaviour of these different ALS types in adoption of cereal-legume association and improved seeds use, and adoption determinants. Based on the sustainable livelihoods framework capital, literature revue and expert knowledge were used to identify candidate variables affecting this adoption. The binary logit model was used for the analyses. A total number of 390 farm-households growing millet, sorghum and cowpea (i.e. main crops targeted by CRP GLDC). Bi-logistic regression models were run for whole population and individual ASL types. The results showed that factors determining farmers’ adoption are related to labour availability, farmer remoteness from paved road and market, equipment level, income, policy intervention and farmer perception of the rainfall variability risk. The study identified determinants for all ALS types (common determinants) and specific determinants shaping adoption behaviour of each individual ALS type (ALS type-specific determinants). The existence of both common and type-specific determinants demonstrated that the presented ALS typology-based adoption analysis approach add new values compared to business-as-usual adoption analysis approaches that treat only sample whole population. The common determinants confirmed factors having wide effects on adoptions. The ALS type-specific determinants reflect new causal effects which the consideration of whole sample population alone cannot find. With the ALS typology-based approach, the number of identified causal relationships explaining adoption can increase 50% (improved seed uses) or 30% (cereal-legume intercropping practices) compared with the use of business-as-usual adoption analysis method. The study therefore recommends accounting for agricultural livelihoods systems heterogeneity in designing policy interventions and for conducting research studies.