3.3 Final set of suitable sites
The final set of suitable sites include 16 locations spread across the 10 communities, with distribution depending on the considered water-related problems (Figure 5).
The distribution of the final set of chosen suitable sites is uneven and was determined by the severity of the water-related problems in each community, the presence and direction of transhumance fluxes and the general trade-off between the need of this technology and potential other water solutions. First, not all the communities present biophysically suitable sites for sand dams construction strictly within their territory. This is for example the case of Tchitemo and Nascente. However, there are other water solutions which could be more fit for their conditions and resources. For example Nascente has a water spring that could be better managed and, to solve their high human-animal and violent conflict problems, some other animal-specific water points could be used – for example chimpacas, which are the local water harvesting micro-basins for animal use. On the other hand, a larger number of suitable points were selected in the Virei communities of Tchicueya and Sayona (four in each community), because there is no other better suited water harvesting technology in such an arid environment – apart from flood wells built along the sandy rivers, which serve to use the water harvested in the sand. Other suitable points are located in strategic areas to serve both the local communities and the transhumant herders passing by during the seasonal caminho. This is, for example, the case of SD4, which is located in the northern part of Tchicueya and it serves both some local kimbos and the other herders coming from the south and directed to Bibala area.
The abductive approach has demonstrated to pragmatically deliver the identification of a fair number of suitable points integrating both scientific and indigenous knowledge. In some occasions, allowing the community to freely suggest suitable suites was a successful strategy. For example in the SD12 site in Munhino, we let the key informants (a group of elder herders) guide us to a suitable location based on their experience and knowledge of the place. They first led us to a not-functioning sand dams, which was a sign that they understand the type of water harvesting technology that they were called to support us to identify. They also asked for the rehabilitation of the not-functioning dam, because it used to be a key water point for the cattle. After then, they took us to a point near a biophysically suitable one, which represents the ideal combination of bottom-up and top-down approaches to site identification. However, not always communities’ indications provided useful insights. In other occasions, the points suggested by the community were in areas of larger mulola were there is high water accumulation in the sand, but the dimensions of the mulola were often too large for the type of small sand dams assessed here, or there were no rocky banks. Also most of these larger mulola are subject to floods during intense rainfall (which was noted during the focus group discussions and confirmed during field inspections), therefore possibly not suited for sand dams. This indicated that the bottom-up approach relying on indigenous knowledge is much more insightful when communities are familiar with (or well trained and informed on) the technology and they had seen or used it in the past. This is particularly important considering that, regardless of the methodology for sand dams siting that is followed, there is still a significant geomorphological and geophysical assessment work to be performed – increasing the costs to refine the top-down selection.