Off-campus South Dakota State University users: To download campus access theses, please use the following link to log into our proxy server with your South Dakota State University ID and password.

Non-South Dakota State University users: Please talk to your librarian about requesting this thesis through interlibrary loan.

Document Type

Dissertation - University Access Only

Award Date


Degree Name

Doctor of Philosophy (PhD)

Department / School

Geospatial Science and Engineering

First Advisor

Matthew C. Hansen


The present dissertation reports wetlands characterization results in the Central African Congo River Basin. Multi-source and multi-resolution remote sensing data are used along with a state-of-the art decision tree classification method to map the wetlands extent in the Democratic Republic of the Congo (DRC). This map is then combined with vegetation and forest cover change maps published in Potapov et al. (2012) to derive a map including forested wetlands as a thematic class. Using population density data, we investigated whether high terra firma population density and low percent remaining terra firma forest are related at the lowest administrative level (Sector); specifically, we tested these two variables as predictors of wetland forest cover loss. Additionally, a network of 548 circular permanent carbon inventory plots of 15 meters radius were established in forested wetlands and dense semi-deciduous forest in order to estimate Above Ground Biomass (AGB) for each forest type and to compare values between forested wetlands and non-wetlands forests. Results showed that wetlands are a significant part of the landscape in the Cuvette Centrale and in the DRC. They cover approximately 32% of the Cuvette Centrale located in the western part of the Democratic Republic of the Congo and in the eastern part of the Republic of the Congo. They also cover an estimated 440,000 square kilometers, or 19.2% of the total area of the Democratic Republic of the Congo. They predominate in the Lake Tumba-Lake Tele area where they cover more than 55% of the land area. By quantifying available upland forest resources and overlaying with population density, it was possible to identify stressed areas inside of the forest domain (traditionally known for having generically high levels of forest resources). A polynomial regression relating population and primary terra firma forest to wetland forest cover loss yielded an r2 of 0.76, illustrating a nascent and significant land cover change dynamic in response to terra firma forest resources exhaustion. Areas most at risk for future wetland forest loss lie in the western Cuvette and coincide with the Lake Tumba and Lake Tele landscape that include (north to south) the Sud-Ubangi, Mongala, Equateur and Mai-Ndombe Districts. The scarcity of terra firma forests in these districts has forced local communities to develop cropping methods that allow for cultivation in periodically flooded areas, specifically in marginally elevated areas within the wetlands. Biomass estimates in the Lake Mai-Ndombe area have revealed that forested wetlands hold important AGB stocks. These stocks estimates varied significantly with the forest type and the allometric equation used to derive AGB from DBH measurements. A local allometric model developed herein has estimated AGB values from 308 to 554 Mg biomass per hectare in flooded forests. The lowest values were estimated for the open swamp forest (308 Mg/ha) and the AGB estimates were significantly higher in the dense swamp forest (554 Mg/ha). Compared to the two mature undisturbed terra firma forest classes (Forescom logged and unlogged), this equation showed that the dense flooded forest has significantly higher AGB stock (Forescom logged: 532.346 Mg/ha and unlogged: 526.718 Mg/ha). The Chave’s Pan-tropical allometric equation yielded substantially larger estimates of AGB values for the undisturbed mature forests. The Forescom logged area has the highest AGB estimate (634.5 Mg/ha), followed by unlogged forest stratum (549.9 Mg/ha) and the dense swamp forest (538.5 Mg/ha). However, there was no significant difference in AGB stocks between the dense flooded forest and the dense semi-deciduous unlogged forest. These results show the need for addressing issues related to wetland forest management and protection in the Congo Basin, especially where increasing populations are exhausting primary terra firma forest resources. Thus, regular forest monitoring and verification using remotely sensed data coupled with on the ground data collection such as forest inventories will be an asset for wetlands preservation and management.

Library of Congress Subject Headings

Wetlands -- Congo River Valley -- Remote Sensing
Wetlands -- Congo (Democratic Republic) -- Remote sensing


Includes bibliographical references.



Number of Pages



South Dakota State University


In Copyright - Educational Use Permitted