Extreme erosion characterizes Madagascar's central highlands. The country has a "world-record erosion rate" according to the World Bank, and the widespread hillslope gullies called lavaka are featured in every guidebook as examples of the catastrophic denudation of the country.s landscape. Lavakas, which have been described as .cat claw-marks. on the landscape, initiate suddenly and grow rapidly, shedding sediment volumes on the order of 8000 m3 in a few months. They cause substantial infrastructural problems because they undermine roads and bridges, and because the debris flows that issue from them swamp agricultural land. They are widely cited as being largely or entirely anthropogenic in origin, and the major source of high sediment loads in Madagascar's rivers. Neither of these interpretations is based on hard data, however, and in spite of the importance of this issue, both from an economic and an environmental point of view, of erosion are largely unstudied. Strong statements abound, but there are no data to constrain the rates at which lavakas form, or their relative contribution to regional sediment budgets; and the geological controls on their formation are largely unknown. The proposed project will address the issue of modern erosion in Madagascar by investigating rates of lavaka formation and geological controls on their initiation. Analysis of changes in lavaka distribution and size will be achieved by combining air photo mapping of lavaka distributions 60 years ago and field mapping of present-day lavaka distributions in a GIS database. This will permit evaluation of decadal-scale sediment generation volumes, as well as lavaka evolution rates. Cosmogenic isotope analysis will produce sediment generation rates on the millennial scale, which can be compared with the mapping-derived rates of recent erosion to evaluate whether erosion rates have in fact increased in recent time as has been frequently asserted but never demonstrated. Quantification of human effects on lavaka formation and growth will be achieved by GIS analysis of mapped distributions of lavakas and their relationship to different land-use categories. Understanding of geologic controls on the location of lavakas will come from detailed mapping of the structural geology of underlying saprolitized bedrock and slope characteristics. The results of this project will be of significance to studies of human interactions with the landscape in general. They will have direct relevance for environmental planning and erosion mitigation strategies for Madagascar. Establishment of geologic criteria predisposing some slopes to lavaka formation could save enormous amounts of money by providing preventive planning strategies for road and dam building organizations. Above all, understanding the role of lavakas in sediment generation; the role of humans in lavaka formation; and whether Madagascar is naturally characterized by high erosion rates or whether they are of recent and anthropogenic origin, will solidify thinking about ecological strategizing in Madagascar and contribute to a better understanding of the islands exogenic system as a whole. This project will be a wonderful way to train students in the process of approaching a scientific problem that has direct societal relevance. Linking education and international collaboration is also a key component of this project. Undergraduate students, both American and Malagasy, will get intense one-on-one mentoring and training in the field and in the lab, as well as hands-on experience with a variety of analytical tools and GIS/remote sensing techniques. The project will provide opportunities for third-world Malagasy students to travel to the U.S. for education and training, and will also create career training and research opportunities for Malagasy collaborators.

This award is co-funded by the Division of Earth Science and the Office of International Science and Education.

National Science Foundation (NSF)
Division of Earth Sciences (EAR)
Standard Grant (Standard)
Application #
Program Officer
Enriqueta Barrera
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Williams College
United States
Zip Code