African mountain areas rarely receive the attention offered regions such as the Himalaya or Alps, although 45% of the continent has slopes greater than 8% and some 3 million km² lie above 2000 m. The mountain zones are often the most productive, having a better combination of soils and rainfall than many lowland regions. About 100 million people are thought to live in these areas. However, of greater economic and political importance today is the fact that the highlands are vital sources of water, timber, and minerals for the economic development of the lowlands. The lowlands have become the focus of most development initiatives, leading to tensions between highlands and lowlands in some areas. In Africa, as elsewhere, many highlands are also border zones (eg, in Burundi) and remain areas of political unrest.

This book is the proceedings of a workshop organized by the African Mountains Association (AMA) and the African Highlands Initiative (a program within the International Centre for Research on Agroforestry, ICRAF) in Madagascar in 1997. The workshop was the fourth of a series that began in 1986 in Ethiopia and met subsequently in Morocco (1990) and Nairobi (1993). During this period, the international framework for mountain research and policy changed dramatically. First was the adoption of Chapter 13 on mountains as part of Agenda 21 at the Rio conference in 1992. This was followed by FAO's designation as Task Manager for Chapter 13 in 1994 and the inauguration of the Mountain Forum in 1995.

The papers in this book represent some of the work carried out in Africa during this exciting period. Of the 18 main papers, 12 are in English and 6 are in French; all have appropriate French or English summaries. The editors point out that they represent just part of the research currently under way in Africa, concentrating on Eastern Africa and especially Madagascar. The range of topics includes soil and water management, biodiversity, and farming systems.

Research on soil and water conservation provides a focus for 7 chapters, ranging from studies of soil analysis through GIS in Kenya to a discussion of the World Overview of Conservation Approaches and Technology (WOCAT) strategy for decision support and management for soils and water. The GIS analysis by Mati et al, for example, was used to explore relationships between key soil variables and water availability and their susceptibility to erosion in the Upper Ewaso Ng'iro. This is part of the largest basin in Kenya, and the analysis clearly highlights how the best soils were found on the mountain slopes that were also susceptible to erosion. Another study in this basin assessed the utility of streamflow modeling, with particular reference to developing an understanding of the flow patterns between highlands and lowlands in a situation where much of the new investment and water demand is in lowland areas. A particularly interesting point regarding the improvement in accuracy was that, despite instrumentation, it was the local knowledge of staff that really counted in making sensible judgments about the validity of flow data.

From a more traditional perspective, Léa et al explore the link between the local geology, erosion processes, and climate in an intensive rice production zone in Madagascar, leading to a strong call for policies to prevent erosion, especially through detailed land-use planning to avoid zones of particular sensitivity. Finally, in this group, work on alpine wetlands in Lesotho by Grab and Morris highlights problems arising from the commercial exploitation of water in the Highlands Water Project. Road construction plays an important role in removing the turf cover, and local climate change may alter the nature of alpine degradation. These factors, in addition to the more commonly documented grazing problems, need to be taken into account when discussing the dynamics of highland wetlands in Lesotho.

Another group of papers focuses on biodiversity conservation, particularly problems associated with deforestation. Most of the papers reflect the concern in Madagascar resulting from the clearance of highland tree cover for agriculture. A study using satellite images in the Manongarivo Massif (Gautier et al) indicates that clearance has taken place despite the area being part of a protected zone. Once again, the principal contributor is the pressure for rice production, reflecting increased population. Similarly, Ralaiarivony argues that conservation policies are needed to preserve biodiversity but that simple ‘top down’ policies will not work. The economic pressures to maintain livelihoods are far too strong. He reports on a project that tries a more participatory approach and seeks to find a common language between the different perspectives of local communities and the state. As Ralaiarivony states, such work requires enormous patience and compromise and is unlikely to be resolved through relatively short-term projects.

A final group of papers represents the input of social scientists who examine the impact of farming systems, gender, and broader management practices on highland landscapes. Several of the papers noted earlier point the finger at slash-and-burn agriculture. However, only the work of Messerli and Pfund begins to explore the complex net of relationships that drives farmers to continue to practice ‘Tavy,’ or pluvial rice cultivation by slash-and-burn, which is so common in the highlands of Madagascar. They employ a form of sensitivity analysis that requires the detailed specification of the key actors and all the relationships between them as a starting point. It is then possible to trace the impact of particular policies on both households and the landscape through the way farming systems are chosen. It is a very time-consuming approach but highlights the fact that the concentration on Tavy, far from being simply the result of poor household agricultural practice, reflects the impact of the wider policy arena where self-sufficiency in rice is an established priority. They argue that a sustainable approach would require more intense farming on the valley floors, which would meet the food demand targets but, at the same time, help limit the removal of upland tree cover.

In Ethiopia, Wood describes the problems facing highland wetlands development, which until relatively recently was the focus of much activity. As with ‘Tavy,’ Wood reports that the real problems facing sustainable development of this biotype come from the way that state policies are interacting with local management techniques. While many farmers possess detailed and appropriate knowledge of the dynamics of the wetlands, policies on land ownership and production are beginning to lead to their overexploitation in order to meet state and household demands. In this example, the problems lie less with technology than with the development of suitable institutions that can shield local ecologies from pressures generated from outside.

This collection provides a useful addition to the growing number of works dedicated to the study of African mountains. The case studies are not just valuable as statements of local interest but provide an indication of the state of research into African mountain systems. It is noticeable that there is a gradually increasing awareness that such research must move beyond an exclusive concentration on physical denudation processes to encompass an understanding of the social and political framework of livelihoods. Nevertheless, there remains a dearth of studies examining the policy arena and incorporating a greater awareness of the directions of general development policy. In this respect, work carried out in other highland areas in Latin America and Asia could provide some useful methodological guidelines.