Publication Date




Embargo Period


Degree Name

Master of Arts (MA)


Geography (Arts and Sciences)

Date of Defense


First Committee Member

Douglas O. Fuller

Second Committee Member

Shouraseni Sen Roy

Third Committee Member

John Beier


The advent of developing outdoor malaria vector control methods creates a demand for distribution models of Anopheles mosquitoes at regional (~30 meters) and fine spatial scales (~2 meters). The distributions Anopheline mosquitoes in West Africa have been modeled in the past, yet always at relatively coarse resolutions. In this study, I worked to develop methods to ascertain the distribution of Anopheline mosquitoes at these little studied spatial scales. The species distribution modeler Maxent was used to create a species distribution model at a regional scale for Anopheles gambiae and Anopheles arabiensis which relied on Landsat derived environmental indices. Models for both species preformed reasonably well with a training Area Under the Curve value (AUC) of 0.767 & a test AUC of 0.783 for Anopheles gambiae, and a training AUC of 0.822 & a test AUC of 0.680 for Anopheles arabiensis. The result of the created models agrees with the known bionomics of these species and demonstrated the reliance on the area around and in urbanized areas as being important to both species. The second aim of this research was to observe the distribution of mosquitoes at a fine spatial scale by mapping possible areas of resting habitats that these malaria mosquitoes use to rest during daylight hours. This was performed by using two different models, Maxent and Dempster-Shafer modeling, along with the high resolution satellite images from the WorldView 2 satellite. The results of the two modeling methods appear to agree with the results of the other fairly well with a linear regression R-squared value of 0.428 (p< .001) and both appear to be capable of mapping out the presence of areas likely used as resting sites by mosquitoes. Yet to accurately determine which resting sites are more important than others may require additional data that is difficult to determine by using remote sensing alone.


Mosquitoes; Spatial Modeling; Malaria; Mali; Species Distribution