Factors affecting the adoption of soil organic carbon enhancement technologies and their spatial distribution among small-scale farmers in Kenya and Ethiopia.

Abstract

Land degradation has poised great threats on food production and the sustainability of conservation areas. This has resulted from the depletion of soil organic carbon which forms the basis of soil fertility rendering farmlands unproductive. Efforts to resuscitate the productivity of farmlands have been made by the stakeholders promoting the adoption of soil organic carbon enhancing technologies (SOCETs). This study therefore, sought to investigate the extent of adoption of SOCETs and the factors affecting the adoption of these SOCETs among smallholder farmers and across the geographical space in the study areas in Kenya and Ethiopia. The dataset consists of 381 households in Ethiopia and 334 households in western Kenya. Probit model was used to predict the factors affecting adoption of SOCETs including fertilizer use, manure use, grass strips, crop rotation, intercropping, agroforestry and residue management. This was followed by the spatial modelling of the factors affecting adoption using Random forest to predict adoption spatially, and geographically weighted regression to show the relationship of adoption to each factor across space. The survey results indicated that fertilizer use was the most adopted technology in both Kenya and Ethiopia at 99% followed by intercropping at 80%, manure at 50 %, use of crop residues at 50%, crop rotation at 40% and grass strips at 30%. Factors constraining adoption were identified as those related to access of information, access to inputs and credit, household characteristics and biophysical characteristics such as rainfall, plot erosion and slope. Farmers who had access to information through extension (p<0.01) or involvement in farmer associations (p<0.05) and those who had access to education (p<0.05) had higher adoption of SOCETs. Those who lived closer to the markets (p<0.05) had higher adoption to fertilizer and agroforestry as compared to manure and grass strips. Farmers who perceived their soils to be fertile (p<0.1) had low adoption of SOCETs while those that perceived their plots to be susceptible to erosion (p<0.05) had low adoption to manure and fertilizers. Adoption prediction using the random forest model and further analysis using geographically weighted regression model showed that, factors affecting the adoption of SOCETs affect the farmers’ decision differently across space. Access to information in an area leads to increased adoption to fertilizer which is discouraged by long distance to markets and difficult access to credit reducing adoption. Therefore, farmers with access to information and who live closer to markets had a higher adoption of SOCETs compared to those who live further away. The study therefore, recommends that the government should support the farmers’ initiative by improving transport and market infrastructure. Also, the governments and the non-governmental organizations involved should invest in farmer education and dissemination of information so as to improve the knowhow of the farmers. Finally, the government should leverage micro-credit services to the farmers such as promoting affordable and appropriate credit facilities. The spatial aspect in the adoption of SOCETs should never be ignored by future research as it proves that failure to consider it would lead to wrong impressions and results. Keywords: Soil organic carbon, Kenya, Ethiopia, sustainable land management technologies, land degradation, adoption, small-scale farmers