After the International Conference on Water and Environment in Dublin in 1992, the four socalled Dublin Principles created new managerial approaches for the water sector. This case study, which was conducted in the Limpopo Basin in Mozambique, examined the performance in the implementation of the principle related to “water development and management based on a participatory approach involving users, planners and policy makers at all levels”. The study was motivated by the fact that experience with the implementation of this principle in the Limpopo Basin of Mozambique has not delivered the anticipated outcomes. To this effect, this study analysed the history and performance of the decentralization process in the Limpopo Basin, and the factors that might have contributed to the outcomes we observe The methodology employed by the study was based on the framework for institutional analysis of decentralization reforms in natural resource management proposed by Dinar et al. (2005), Kemper et al. (2006), and Blomquist et al. (2008). This framework recommends that for the decentralization process to be successful, the following pre-requisites must be in place: (1) financial assistance from the state to enable basin level stakeholders to establish some of the organisations; (2) actors' participation and equitable representation of different segments of society, and acceptance of it from the communities; (3) the presence of basinlevel institutions, availability of forums for information sharing, communication and for conflict resolution; and (4) legitimacy, relevant human capacities and adequate financial resources among the River Basin Organizations (RBOs). The results from this study indicate that in as much as the Water Law, the Regulations of Water Services Provision and the Water Policy in Mozambique created the basis for the decentralization of water resources management, the operationalisation of the process has not been successful, considering that the prerequisites for an effective and sustainable decentralization process as postulated by Blomquist, Dinar and Kemper are still lacking. The study established that incentives for the decentralization process were not linked to the scarcity of water. The study further established that most prerequisites postulated by Blomquist, Dinar and Kemper were not satisfied, in particular: (1) the financial assistance from the state to enable basin level stakeholders to establish some of the organisations is inadequate; (2) the actors' participation and equitable representation of different segments of society with interest in water resources management is not satisfactory; and (3) the legitimacy, relevant human capacities and adequate financial resources for effective functioning of the Water Users Associations (WUAs) are still lacking. Copyright / Dissertation (MSc(Agric))--University of Pretoria, 2012. / Agricultural Economics, Extension and Rural Development / unrestricted

Severe flooding events in subtropical southern Africa are not uncommon and can cause damage to infrastructure, lead to huge economic losses, and fatalities. Although extreme rainfall events can have far-reaching negative consequences, they can also provide large amounts of freshwater within a short time span, which supports the rain-fed farming upon which much of the population depends. However, the mechanisms through which extreme rainfall is produced in southern Africa are still not well understood. In particular, relatively little is known about where the moisture, a key ingredient in the rainfall, is sourced and how it is transported into the region. This thesis aims to address some of the gaps in this understanding by examining the moisture sources and subsequent moisture transport moisture into one of the key river basins in southern Africa, the Limpopo River Basin (LRB). The LRB, located in eastern southern Africa and spanning four countries, has experienced a number of extreme flooding events over the last three decades. Using CHIRPS satellite merged rainfall data for 1981-2016, the thesis identified the top 200 heavy extreme rainfall events in the LRB and the associated weather systems. It was found that tropicalextratropical cloud bands account for almost half of the events and tropical lows are responsible for just over a quarter. The remaining quarter of the events are associated with mesoscale convective systems and cut-off lows, the latter more important during transition seasons. Most of the events occur in the late summer when tropical lows and cloud bands are more common. Some relationships between the frequency of heavy rainfall events over the LRB and interannual climate modes of variability such as ENSO, SIOD, and SAM were found. Having examined the annual cycle of the top 200 heavy rainfall events, the analysis then applied the Lagrangian trajectory model HYSPLIT, with NCEP II reanalysis data as input, to backtrack air parcels from the LRB to their moisture source on seasonal scales and in terms of the types of weather systems involved. The resulting trajectories show that the seasonal transport of moisture over the LRB originates from seven moisture source regions; namely, local continental, tropical southeast Atlantic Ocean, midlatitude South Atlantic Ocean, tropical Northwest Indian Ocean, tropical southwest Indian Ocean, subtropical southwest Indian Ocean, and the Agulhas Current. Important differences in moisture source regions and pathways exist between early (OctoberDecember) and late (January-April) summers, with the tropical northwestern Indian Ocean and the northern Agulhas Current sources more prominent during JFMA than OND. Generally, moisture v source regions and transport pathways for LRB tend to be influenced by both the regional summer season circulation and the synoptic systems involved. Thus, it was found that cloud band and tropical low events within the top 200 tend to have the Congo Basin as an important moisture source whereas this source is less evident for cut-off low events. To help assess the robustness of the composite synoptic approach, the final part of the thesis applied the Lagrangian analysis to the most severe case in the top 200 events over the LRB (11- 21 January 2013). It was found that this case was largely linked to three main moisture sources: (1) tropical northwest Indian Ocean, (2) the Agulhas Current / Mozambique Channel, subtropical Southwest Indian Ocean, and (3) continental sources over the Congo Basin and northern Tanzania. Generally, the moisture source regions and pathways for the January 2013 event agreed with the climatological moisture source regions over the LRB, apart from the obvious absence of the tropical southeast Atlantic source in this case. In general, the thesis has provided a better understanding of the characteristics of heavy rainfall events over the LRB in terms of their associated weather systems, seasonality, interannual variability, and moisture source regions and trajectories.

Disaster risk reduction policies and practitioners alike emphasise the importance of vulnerability reduction. However, the concept of vulnerability is highly dynamic, and research still strives to understand and capture its complexity. The purpose of this study was to improve the understanding of flood vulnerability in rural disaster-prone communities in Mozambique. To explore previous experiences of floods, I conducted semi-structured interviews with local risk committee members and community members in the lower Limpopo river basin. The findings were analysed with an analytical framework consisting of the Disaster Pressure and Release (PAR) model, drawing on political ecology and the Access model. Disaster was studied as a process revealing important factors, capabilities and strains affecting peoples’ vulnerability. This paper illustrated that rural communities in the lower Limpopo river basin are vulnerable to floods in a variety of ways. The findings presented unsafe conditions such as the fragile local economy, unsafe natural resources, strained physical resources and limited access to human and social capital. Several factors deriving from political, social and economic structures were found to influence specific forms of vulnerability expressed in relation to floods. Therefore, this paper contributes to new insights of how flood vulnerability can be described and explained in Mozambique.