Understanding wetlands reclamation and soil-transmitted helminthes and schistosomiasis incidence patterns in Rwanda (2001-2012)
Wetlands comprise the transitional zone between land and water with land covered by shallow water or with a water table at or near the surface, and an ecosystem associated with long-term soil inundation. Worldwide, wetlands sustain large communities of people, who depend on their natural resources that maintain them. Also in Rwanda, wetlands are ecosystems of critical importance to almost 12 million people, generally engaged in self-sufficient smallholder agriculture. Since the 1970’s agricultural development programs have been implemented in inland valley swamps, larger inland deltas and lacustrine wetlands in Western and Eastern Africa. In Rwanda, the conversion of wetlands for intensified agricultural production is a more recent phenomenon. In the framework of Vision 2020, the long-term strategy for national development, and the Economic Development and Poverty Reduction Strategy (EDPRS), rural development and agricultural transformation are spearheads for rapid and sustainable development. Wetland conversion is one of the major mechanisms to increase agricultural production, ensure food security for the growing population, and achieve MDG1 (eradicating extreme poverty and hunger). Apart from benefits, the “created irrigation scheme landscape” can also have negative impacts. One of these is the increased risk of transmission of schistosomiasis and soil-transmitted helminths, which can negatively affect the health status of communities that depend on their produce. This dissertation contributes to improved understanding of the linkages between wetlands and wetland use on the one hand, and transmission risk of schistosomiasis and Soil Transmitted Helminths (STH) on the other.
The first stage of the study modelled and mapped wetlands and their characteristics both at a national and regional scale. Findings show that the interactions between geomorphologic setting (elevation, slope), hydrological conditions (contributing area) and climatic factors (temperature, rainfall) are fundamental for characterizing Rwandan wetlands. Logistic regression analysis was conducted to generate wetland probability maps with very high accuracy both at national and subnational level. Further analysis was then done to estimate in how far Rwandan wetlands are sensitive to climate change. Results indicate that, countrywide, a 1% increase in temperature would cause a net probable wetland area decline of 12.4%. A 1% increase in rainfall would cause a net probable wetland increase of 1.6%. The analysis also showed that wetlands in the central part of Rwanda are the most sensitive to climate change. Geographic variability of climate change effects implies that future wetlands use planning should be location-specific.
The second stage of the research analysed the linkages between environmental (including wetlands and wetland use) and socio-economic risk factors on the one hand, and transmission risk of schistosomiasis and STHs on the other. The novelty of our analysis is that it is based upon routinely collected data of confirmed cases recorded at primary health facility level. Confirmed case data were aggregated to health facility service areas (HFSAs) with a high spatial resolution, and merged with population data to generate incidence rates. Compared to traditional District-level prevalence surveys, our fine-grained spatial resolution enables a geographically much more accurate detection of transmission hotspots and associated risk factors. Outcomes clearly show that the spatial pattern of schistosomiasis incidence has a localised distribution and is therefore better monitored at HFSA then at the much larger District level. Regarding risk factors, a significant positive relation was identified between schistosomiasis incidence and presence of and proximity to rice cropped areas, rainfall and specific soil properties. For STHs incidence, population density and the proportion of rural residents emerged as significantly associated risk factors.
The third research stage consisted of a spatiotemporal cluster analysis using SaTScan statistics. Findings demonstrate that areas of high schistosomiasis transmission vary considerably across space and over time. The results show a clear impact of the national intervention program, but not leading to full eradication. The analysis revealed existence of three types of clusters: areas with persistently high rates of transmission (close to lakes), areas where transmission declines over time (impact of NTD program), and areas of high transmission that emerge over time. Newly emerging clusters are associated with increased agricultural use of wetlands, irrigated rice cultivation in particular. Our findings illustrate that the spatial and temporal distribution of schistosomiasis incidence is very dynamic which makes elimination difficult to achieve. Identified spatio-temporal dynamics should, therefore, be considered in future disease control activities.
The fourth stage of this study focused on forecasting the future spatial pattern of schistosomiasis transmission risk using a Bayesian model that accounts for false zero cases. The zero-inflated Poisson model incorporated anticipated climate–induced changes in rainfall and temperature as well as the planned expansion of rice cultivation in wetlands. Other factors were assumed to be constant. Results show that, changes between relative risk of 2009 and forecasted risk in 2050 identify persisting and emerging areas with high relative risk of schistosomiasis transmission. Model outcomes indicate that the risk of schistosomiasis transmission can be expected to be 69% higher in areas with rice cultivation, 29% higher in areas in close proximity to rice farming, and 50% higher in areas close to water bodies. The prediction and forecasting maps provide a valuable tool for monitoring schistosomiasis risk in Rwanda and planning future disease control initiatives.
Overall, this research has confirmed that the anticipated impacts of climate change on wetlands will not be uniform across Rwanda but will vary from place to place. Wetlands in the central part of Rwanda are the most sensitive to climate change. In addition, the research confirmed that the anticipated future spatiotemporal expansion of rice cropping areas is positively correlated with increased risk of transmission of schistosomiasis. Finally, routinely collected health data are proven to be a valuable information source for monitoring the spatial, and spatiotemporal distribution of schistosomiasis. That distribution is very dynamic and routine health data can be efficient to monitor monthly or seasonal variability. Conducting the analysis at the level of the health facility service area, as the basic spatial unit of analysis, proved to be a suitable spatial scale for modelling, monitoring and mapping of schistosomiasis transmission across space and over time.
Biography
Elias Nyandwi was born on 5 July 1971 in Ruhango, Rwanda. In 2003, he graduated from the National University of Rwanda where he obtained a bachelor degree in Human and Physical Geography. He pursued his post-graduate studies at the International Institute of Geo-Information Science and Earth Observation (ITC), The Netherlands and graduated in March 2008 with a degree of Master of Science in Geo-Information Science and Earth Observation for Environmental System Analysis and Management. Since July 2012 he was awarded the NUFFIC Scholarship, under the NICHE/RWA/071 project, to pursue his doctoral research at the Faculty of Geo-Information Science and Earth Observation of the University of Twente (ITC-UT), the Netherlands. His research outputs were presented and published in high profile (regional and international) conferences and journals and resulted in this thesis.
Elias was respectively employed at National Service of Census (SNR) until March 2005 and at the Centre for Geographic Information System and Remote Sensing of the University of Rwanda (CGIS – UR) as an Assistant Researcher, since June 2005 up to now. He was involved in several research projects, teaching and tailor-made training, consultancy and community service related to GIS and RS and applications. He also worked as Head of Environmental and Natural Resources Management research unit and National Coordinator of three years (2009-2012) Pan-African Research Project on Participatory GIS for Forest Resources Management funded by International Development Research Centre (IDRC).
