N2Africa is a large-scale project with the aim of putting nitrogen fixation to work for smallholder farmers growing legume crops in Africa. It is a science-based “research-in-development” project, with constant learning loops to find the best technology for every farmer. The main goal of N2africa’s agronomic research is to understand the major constrains on legume productivity, with a special focus on the causes of yield variability and how to reduce it. One of the four activities in this cluster is the adaptation trials, that evaluate the performance and adaptation of the proposed technologies under farmer’s management. Focal farmers are chosen to collect detailed information about changes and management of the proposed technologies, as well as agronomic and household characteristics that might affect yield. The surveys are done using electronic tablets with the ODK software. Since a fundamental step in the N2africa learning loops is having enough and reliable data, the first objective of this internship was to revise the quality of the data collected in the Focal adaptation trials from 2015 and 2016. The second objective was to evaluate if the data collected gives information about a) the changes in the technologies, b) their performance and c) information that will help targeting the technology. Several household and farm characteristics were evaluated for their relationship with yield with a linear mixed model. Some of the variables were chosen for principal component analysis (PCA) and subsequent hierarchical clustering per each country to separate farms into separate groups. Overall, the data was complete but some inconsistences were found, ambiguous or incomplete questions were further analysed and changes were suggested. Frequency of hired labour, education level, farm size and inputs (among others) were significantly related with yield, and an interaction with treatment shows they might be relevant for the performance of the N2Africa technology. The clustering for each country was rather arbitrary due to the big amount of NAs values, however, farm size, the ownership of livestock, the labour dynamics and the market orientation of the farm showed differences among clusters which suggest these variables could be the base of a farm characterization. Labour requirements and marketability play fundamental rolls for adoption, so understanding the several types of farms, will improve the targeting of the technologies. Based on the results and feedback from N2Africa stuff, the survey for 2017 was adapted. This included changing the general structure of the survey, questions that recorded the changes of the farmers in the technologies were systematized for easier analysis and others were removed or changed. The guidelines for the focal adaptation trials were also updated. Finally, recommendations were made on how to adapt the survey to have more information about the potential of farmers to adopt certain technology, which will help in targeting the best-fit legumes for specific situations.
MSc and Bcs thesis, internship reports