Sand dams have become popular in many parts of the arid world as a relatively cheap and effective water harvesting technology. Kenya is one of the countries with the highest number of such dams, with semi-arid Kitui County having become a major hub in recent decades. These sand dams are used for water storage in the beds of Kitui's seasonal rivers. The water is used for households and small-scale economic activities. Generally, sand dams are evaluated as very successful, but this paper shows that such success is not guaranteed. Field research conducted in Kitui County in October 2016 suggests that from 116 sand dams surveyed, about half did not have any water during the time of the assessment. This study assesses how various environmental factors affect sand dams' ability to supply water for community use during dry periods in Tiva River catchment in Kitui County. Most of the assessed environmental factors did not show consistent patterns to draw inferences on how they affect sand dams' ability to supply water, with the exception of rainfall amount, water indicating vegetation percentage of clay in a soil and stream orders. More overarching factors like agro-ecological zones and stream order do show a pattern of influence on dams' performance. These results have global significance due to the widespread use or plans to use of sand dams worldwide. There is a clear need to build a better understanding of sand dams performance to define more reliable sand dams' site identification criteria.

Approaches that bridge the technology and knowledge gaps between policy makers and local communities towards natural resource management are required. This study employed participatory Geographical Information System (PGIS) to assess and identify drivers of land use and land cover changes in Nguruman Sub-catchment. Data were collected following a focus group discussion (FGDs) during which a resource mapping exercise was also carried out. FDGs consisting of 12 members were used to establish changes that had taken place between 1994 and 2004, and between 2004 and 2014. The ten year interval was purposively chosen in order to cater for temporal sensitivity in resource changes by local communities. The participants listed the land use in their respective villages for the years 1994, 2004 and 2014. The land uses listed included forestland, irrigated and rainfed cropland, woodlands and water bodies. This was later presented graphically on manila papers for the respective years to showcase land use and land cover changes as perceived by the local communities. Using a digital camera, photographs of these maps were then taken. Features that acted as boundaries and were also found within the areas drawn were mapped using Global Positioning System (GPS). These features were used for geo-referencing of the mental maps in order to analyze natural resource changes as perceived by the local communities. Results from the PGIS were further transferred and analyzed by the GIS in order to determine the extent and magnitude of changes, based on the local knowledge. Results indicate that local communities have knowledge about the causes and consequences of land use and land cover changes occurring in their areas. Significant (p < 0.05) changes were observed in irrigated cropland areas. The study provides an effective basis to describe and explain the patterns of land use and land cover change including their root causes and consequences based on the community perspective. PGIS is a suitable tool to involve local communities in planning, evaluating, monitoring and managing their own natural resources.