These data are from a household survey conducted in Nyangatom District, South Omo Zone, Ethiopia, in 2018/19. The primary research question we sought to address was how the end of the annual flood of the Omo River, which was integral to the flood recession agricultural systems of downstream communities, affected food security. Data relate to agricultural production before and after the end of flood retreat agriculture which followed the construction of the Gibe III dam upstream. [2005 and 2010 in the Ethiopian Calendar]. We also sought to account for changes in non-farm wealth including livestock holdings over the same period.Rapid changes in the natural, social, and economic environment are occurring in Ethiopia's Lower Omo Valley, as part of a state-led development vision of repositioning the region as a major sugar exporter. At the same time, these changes raise risks of environmental degradation, and the emergence of new kinds of inequality and conflict. The Lower Omo is home to a large number of pastoralist groups, and is a major centre of ethnic, linguistic and cultural diversity - reflected in its designation as a UNESCO World Heritage Site (Schlee, 2009; Turton, 1991). While new sugar estates promise to create >150,000 new jobs, they are also drawing labour migrants from other regions, setting up new hierarchies of wealth and opportunity, and raising social tensions (Tewolde and Fana, 2014). As part of its ambitious Growth and Transformation Plans (FDRE 2010, 2015), Ethiopia's government has allocated 175,000 hectares to the sugarcane project (more than the entire area under irrigation in Kenya), and aims to increase national sugar production from 17 million to 42 million tons. Pastoralists who, according to the government, do not use the land optimally, have had their "underutilized" lands repurposed for sugarcane plantations and industrial sites that will produce sugar worth 661.7 million USD and produce alternative energy sources (304,000m3 of ethanol per year and 607 MW electricity). This scenario raises urgent questions about the social justice dimensions of current development models, and their implications for socio-ecological resilience. The following interlinked research questions will structure our exploration of the on-going changes and their consequences: 1. How have recent developments affected the spatial and temporal availability of and access to natural resources in the region? (Environmental sustainability / degradation) 2. How are changing resources affecting conflict dynamics in the region? (Conflict and resilience) 3. How are these changes influencing relations of material in/equality? (Wealth and poverty / resource security) These questions will be addressed by three Working Groups with expertise in the fields of environmental sustainability, conflict studies, and poverty research respectively. The group will first conduct Research involving knowledge generation across disciplines and stakeholder groups, and second work towards Application in the form of knowledge exchange and synthesis. The first step in the research design will accommodate a cycle of stakeholder identification, scoping and testing of research assumptions; followed by an intensive period of data collection using both conventional (survey, focus group) and participatory research methods. Crucially, our tools will include participatory video methods that will be used to stimulate conversations about current models of development that are rarely possible due to barriers of language, distance, and power. The project will thereby facilitate knowledge sharing, processing, and utilisation, and explore how knowledge regarding the environmental changes and their implications for poverty, peace, and security is best integrated in decision-making for diverse stakeholders.

The data were collected through a household survey. Sampling combined a purposive selection of communities, chosen to reflect a variety of subsistence types ranging from greater dependence (pre-dam) on flood retreat agriculture to those with less dependence on agriculture and more dependence on herding. The site with greater dependence on recession farming was Ayepa, that where herding was more important was Kopriye, and Napasmuria was a middle case; however all three combined livestock and recession farming to some degree. A two stage cluster randomised sampling of households within traditional settlements (circular enclosures containing multiple independent households) was used to choose participants. The respondents were knowledgeable adults usually either the male or female household head. A total of 66 households responded to the survey. More information on the project from which these data derive, plus results published to date, are available at https://www.canr.msu.edu/oturn/briefing_notes