Publication Date



Open access

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Ecosystem Science and Policy (Graduate)

Date of Defense


First Committee Member

Douglas O. Fuller

Second Committee Member

John C. Beier

Third Committee Member

Kenneth Broad

Fourth Committee Member

Justin B. Stoler

Fifth Committee Member

Kristopher L. Arheart


With malaria control in Latin America firmly established in most countries and a growing number of these countries in the pre-elimination phase, malaria elimination appears feasible. A review of the literature indicates that malaria elimination in this region will be difficult without locally tailored strategies for vector control, which depend on more research on vector ecology, genetics and behavioral responses to environmental changes, such as those caused by land cover alterations, and human population movements. An essential way to bridge the knowledge gap and improve vector control is through risk mapping. Malaria risk maps based on statistical and knowledge-based modelling can elucidate the links between environmental factors and malaria vectors, explain interactions between environmental changes and vector dynamics, and provide a heuristic to demonstrate how the environment shapes malaria transmission. Changes in land use and land cover (LULC) as well as climate are likely to affect the geographic distribution of malaria vectors and parasites in the coming decades. At present, malaria transmission is concentrated mainly in the Amazon basin where extensive agriculture, mining, and logging activities have resulted in changes to local and regional hydrology, massive loss of forest cover, and increased contact between malaria vectors and hosts. Thus, employing presence-only records, bioclimatic, topographic, hydrologic, LULC and human population data, I modeled the distribution of malaria and two of its dominant vectors, Anopheles darlingi, and Anopheles nuneztovari s.l. in northern South America using the species distribution modeling platform Maxent. This was done to address the gap in knowledge about the spatial and temporal distribution of malaria and its vectors in this region. Results from the land change modeling indicate that about 70,000 km2 of forest land would be lost by 2050 and 78,000km2 by 2070 compared to 2010. The Maxent model predicted zones of relatively high habitat suitability for malaria and the vectors mainly within the Amazon and along coastlines. While areas with malaria are expected to decrease in line with current downward trends, both vectors are predicted to experience range expansions in the future. Elevation, annual precipitation and temperature were influential in all models both current and future. Human population mostly affected An. darlingi distribution while LULC changes influenced An. nuneztovari s.l. distribution. Secondly, I set out to assess the risk of malaria transmission and vector exposure in northern South America using multi-criteria decision analysis, as well as examine experts’ perceptions of strategies needed for malaria elimination. The risk of malaria transmission and vector exposure in northern South America was assessed using multi-criteria decision analysis, in which expert opinions were taken on the key environmental and population risk factors. Results from the risk maps indicated areas of moderate-to-high risk along rivers in the Amazon basin, along the coasts of the Guianas, the Pacific coast of Colombia and northern Colombia, in parts of Peru and Bolivia and within the Brazilian Amazon. When validated with occurrence records for malaria, An. darlingi, An. albimanus and An. nuneztovari s.l., t-test results indicated that risk scores at occurrence locations were significantly higher (p<0.0001) than a control group of geographically random points. Public education, better environmental management and effective anti-malaria drug administration were the experts’ most highly ranked strategies for malaria control. Finally, armed conflicts are considered important obstacles to achieving malaria control and elimination; however, such association is seldom documented. Here, I test the hypothesis that armed conflicts have reduced effectiveness of efforts for elimination of malaria and other infectious diseases in Colombia. I utilized diverse spatio-temporal data aggregated to the municipal level in the Pacific Coastal region of Colombia to analyze how socio-political and environmental variables may explain trends in malaria cases over a 15-year period from 2000 to 2014. The spatial trends revealed subtle differences and patterns in the distribution of malaria cases at the municipality level, which were not evident when aggregated at the state level. The results show that when environmental and conflict-related variables are combined in a single parsimonious linear regression, temperature, conflict-related homicides, precipitation, and elevation produced a highly robust model (adjusted R2 = 75.7%). A model that included only socio-political variables such as internally displaced persons, coca cultivation, and population density explained much less of the variance (adjusted R2 = 43.0%) relative to a model that included only environmental variables (adjusted R2 = 69.1%). The results establish a novel quantitative link between conflict-related variables, particularly conflict-related homicides per municipality through time, and malaria trend in a war-torn country indicating that political stability may be essential to achieve malaria elimination. Hence, as the region tackles the challenge of malaria elimination, prioritizing areas for interventions by using spatially accurate, high-resolution (1 km or less) risk maps may guide targeted control and help reduce the disease burden in the region. The findings also provide information to the public health decision maker/ policy makers to give additional attention to the spatial planning of effective vector control measures. Finally, investigations such as this could be useful for planning and management purposes and aid in predicting and addressing potential impediments to elimination.


Malaria; risk mapping; Anopheles; South America; Colombia