Qualitative data on land use change and ecosystem services from participatory surveys in northeastern, Kenya (August-October, 2013)

The data comprises of two datasets. The first consists of text files of anonymised transcripts from focus group discussions (FGDs) on livelihood activities, ecosystem services and the prevalent human and animal health problems in irrigated and non-irrigated areas in northeastern Kenya. The second comprises of scores from proportional piling exercises which showed the distribution of wealth categories and livestock species kept. The study was conducted between August and October, 2013 and the data were collected as open-ended meeting notes and audio clips captured using digital recorders. Written/thumb print consent was always obtained from each individual in the group. All the discussions were also recorded, with the participant's permission. Thirteen FGDs were held in the irrigated areas in Bura and Hola, Tana River County involving farmers who grew a variety of crops for subsistence and commercial purposes. The others were held in Ijara and Sangailu, Garissa County inhabited by transhumance pastoralists. Each group comprised of 10 to 12 people and the discussions were guided by a check list. The transcribed documents were formatted in Microsoft Word (2013) and saved as text files in preparation for analysis. The aim of the study was to collate perceptions of land use change and their effects on ecosystem services. The data were collected by enumerators trained by experienced researchers from the University of Nairobi and the International Livestock Research Institute (Kenya). This dataset is part of a wider research project, the Dynamic Drivers of Disease in Africa Consortium (DDDAC). The research was funded by NERC project NE-J001570-1 with support from the Ecosystem Services for Poverty Alleviation Programme (ESPA). Additional funding was provided by the CGIAR Research Program Agriculture for Nutrition and Health. Records from the digital recorders were transcribed and cross-checked with the notes made in the course of the meeting for quality control. Written/thumb print consent was always obtained from each individual in the group. All the discussions were also recorded, with the participant's permission. The transcribed documents were formatted in Microsoft Word (2013) and saved as text files in preparation for analysis. Transcripts were analysed according to the framework developed by Hennink et al. (2011). The data were split into two: (i) responses to open ended questions stored as text files, and (ii) scores from proportional piling exercises which showed the distribution of wealth categories (rich, moderate and poor categories) and livestock species kept (cattle, sheep, goats, camels, chicken and donkeys). These scores were entered into Microsoft Excel and saved as a csv file for use in statistical analysis that employed non-parametric tests. Variables that are common between these two datasets are: date of the field survey (date), study site, village and group (men only, women only or mixed) involved in the discussions. Data was exported from the Excel files as .csv files for ingestion into the EIDC. The Microsoft Word files were exported as .txt files for ingestion into the EIDC.