Deforestation and Biodiversity Preservation in Madagascar: The View from Above and Below by Conrad P. Kottak, Lisa Gezon, Glen Green For further information contact: Professor Conrad P. Kottak Department of Anthropology 1054 LS&A Building The University of Michigan Ann Arbor, Michigan 48109 313-663-7102 Copyright 1994 Conrad P.Kottak, Lisa Gezon, Glen Green. All Rights Reserved Deforestation and Biodiversity Preservation in Madagascar: The View from Above and Below by Conrad P. Kottak, Lisa Gezon, Glen Green Deforestation is a key issue in global environmental change. At the local and regional levels degradation of forests jeopardizes food products (honey, game, and wild fruits), medicinal plants, and wood used to build houses, fences, and stock houses. Forests are vital watersheds. As such they supply electricity and water for towns and cities. In societies that rely on irrigated agriculture, forests also serve the economically valuable functions of conserving water and impeding erosion and siltation of canals. Other effects of deforestation have a global dimension. Forest degradation makes a major contribution to greenhouse gas production (CO2), implicated in global warming. The destruction of tropical forests is also a major factor in the loss of biodiversity. This is because of the many species, often of limited distribution, in these forests. Tropical forests may contain more than half of Earth's species while covering just six percent of the planet's land surface. It is particularly with respect to preserving biodiversity that Madagascar, the world's fourth largest island, is important. Madagascar (home of the Malagasy people) is often called an "island continent," because of its many cultures, ecosystems, and topographic regions. The island separated from Africa around 100 million years ago and became biologically isolated sixty million years ago. Madagascar's isolation, size, and varied environments have produced a rich array of species. There are at least 10,000 species of flowering plants. In some forests over 80% of the species are endemic. In biodiversity, Madagascar is one of the richest areas on earth, and its plants and animals are among the most endangered. Madagascar has 2% of the world's plant species, while covering .07% of the earth's land area. Many biologists regard Madagascar as the top conservation priority in the world. In 1991 a team directed by Dr. Conrad P. Kottak1 of the University of Michigan Department of Anthropology used satellite images to identify and study areas of deforestation in northern Madagascar. This research was supported as a "pilot project" by CIESIN, Consortium for an International Earth Science Information Network. CIESIN has been funded by NASA to gather, store, and dispense information on the human dimension of global environmental change. In 1991 a dozen CIESIN pilot projects fostered integrated research (remote sensing specialists, human, and natural scientists) on the human dimension of environmental change in several world areas. For northern Madagascar, "from above" (using satellite images from 1972, 1984, and 1990), we located areas of evident deforestation. We then investigated those deforestation hot spots "from below," through on-the-ground research by a multidisciplinary team. Comparison of the satellite images showed us that the deforestation rate has increased. Change images (generated by plotting the remotely-sensed data--the satellite image information--across time) revealed more deforestation in the six years from 1984 to 1990 than in the twelve years before 1984. The focus of our multidisciplinary research (cultural anthropology, remote sensing, geography, biology, natural resources, public health) has been the human dimension of deforestation. We are studying the extent, nature, and causes of forest degradation in Madagascar. We are also developing strategies for remedial action to guide Malagasy and international policy makers. In our attempt to integrate research in the human and natural sciences, we combined remotely-sensed data with an "on-the-ground" investigation of forest distribution and socioeconomic (i.e., social, cultural, and economic) features. "From below," by mapping vegetation types and the extent of forest change, we began to discover which species are endangered and which are spreading. We were able to compare places that have experienced heavy degradation with areas where little forest has been cut. On-the-ground study of socioeconomic factors allowed us to observe the human uses of forested and deforested land and forest products. Our views from above and below provide up-to-date information about a process of degradation that has affected Madagascar's forests since humans arrived some 2000 years ago. The island's forests were last mapped in 1965, using aerial photographs taken in 1949. Using those maps and Landsat satellite imagery, Glen Green (co-leader, with Lisa Gezon, of our 1991 field team) and Robert Sussman studied the reduction of the rain forests of eastern Madagascar (Figure 1). They monitored the rate of deforestation there over 35 years (Green and Sussman, 1990). In 1985, 3.8 million hectares of eastern forests survived--about half the 7.6 million hectares that had existed in 1950, and one-third the estimated original extent (11.2 hectares). For the eastern rain forests, Green and Sussman found that degradation correlated with topographic relief and human population density (Figure 22). High relief zones have more inherent protection than other areas do. Forest loss was greatest on land of low and moderate slope (Figure 2B). Rising population density in the context of slash-and-burn subsistence cultivation was a major force behind eastern deforestation. More than half the forests in the sparsely- populated regions survived through 1985, compared with just 19% in the high density areas. The Green-Sussman study exposed an important fact for policy makers: Although low relief areas have the greatest need of protection, nearly all Madagascar's forest reserves are in regions of steep slope. Particularly in eastern Madagascar, forest degradation is demographically driven--one manifestation of a national population explosion. The Malagasy population is doubling each generation. It increased from six million in 1967 to twelve million in 1992. The national population is growing at a rate around 3% annually, but with regional differences. The increase is greatest in the central highlands, the homeland of two of the island's most populous ethnic groups, the Merina and the Betsileo (Kottak 1980). For both those groups the subsistence economy is based on irrigated rice cultivation in paddy fields. Population pressure is fueling agricultural intensification and emigration, including rural- urban migration. There is need to inform and educate the people of Madagascar about links between forest preservation and the survival of subsistence agriculture. As watersheds disappear, the productivity of rice fields will decline. (Rice is the basis of the national diet.) The devastating effects of deforestation on rice production are already obvious in the central highlands. Increasing runoff causes siltation in canals and erosion of low-lying rice fields near swollen rivers. What can be done? To curb deforestation in "developing" nations like Madagascar, policy makers are beginning to see the need for new conservation strategies. The traditional approach was to restrict access to isolated forests in areas designated as parks, then employ park guards and punish violators. Modern strategies are more likely to consider the needs, wishes, and abilities of the people (often impoverished) living in and near the forests. Since effective conservation usually depends on the cooperation of the local people, their concerns must be addressed when planning a viable conservation strategy. In summer 1990 Conrad Kottak was asked to consult for USAID, which was designing a conservation project for Madagascar. The United States Agency for International Development now requires its development (including conservation) projects to be evaluated for their social soundness. Social anthropologists are often hired as consultants on the project "design team," to do the social soundness analysis. USAID3 hired Kottak to be the social analyst for their SAVEM project (Sustainable and Viable Environmental Management). The project's goal was conservation--specifically the preservation of biodiversity in Madagascar, with an initial focus on a number (five or six) of protected areas. The project reflected American and international concern with preserving biodiversity and conserving finite global resources--particularly tropical forests. The SAVEM project was intended to help the people of Madagascar. In the long run millions of Malagasy might benefit. This figure includes the urbanites who rely on the protected areas for water and electricity and the rural people whose rice cultivation depends on forest watersheds. More immediately, some 50,000 to 100,000 people live in the villages near the five or six protected areas on which the project focused. The SAVEM project, if successful, will be good for Madagascar in the long run. Still, it must be implemented in the short run and in local communities, where there are obvious obstacles. People who have relied on a broad spectrum of resources in the protected areas are now being asked to give up many of their traditional economic and cultural activities without immediately obvious alternatives, substitutes, or incentives. They are being asked or forced to make these sacrifices in order to satisfy long-term, national and global, rather than immediate and local, goals and interests. Kottak's task was to design culturally appropriate and socially sensitive ways to ensure or enhance conservation in the protected areas. He visited parks and reserves in many parts of the large island. One was Ranomafana, where, largely through the efforts of a primatologist, Dr. Patricia Wright, a national park has recently been established in the eastern rain forest. Kottak also visited the southeast--the geographically divided Andohahela reserve, which covers several ecological zones and ethnic groups. He surveyed the area around Andringitra reserve, a granite mountainous region in the extreme southern highlands. Finally Kottak visited, in northern Madagascar, the city of Antsiranana (formerly Diego-Suarez), where he renewed contact with Dr. John Hough, a natural resources specialist who was working there as a technical advisor for WWF. (In the United States and Canada, WWF stands for the World Wildlife Fund; elsewhere it designates the Worldwide Fund for Nature.) In his 1990 travels in Madagascar Kottak found that the areas chosen for "protection" by the SAVEM project have considerable economic and cultural utility for the communities in and near them. These areas have supplied firewood, wood for house and granary construction, fences, and technology (ox carts, mortars and pestles). Some protected areas have been used for food and cash crop production, including tavy (slash-and-burn shifting cultivation). Some have irrigated rice fields. Others have fields with tree crops (bananas, fruits, coffee). Hunting and gathering also proceed within the protected areas. Women in the eastern rain forest gather pandanus to make mats and catch crayfish to eat and sell. Men hunt and cut down century-old trees for tree fern pots, which women sell along the main road for a few dollars. Madagascar's forests also contain vital cultural products (including medicinal plants and unguents considered essential for the proper growth of children). In one ethnic group rice from a forest field is part of the ceremony needed for a successful and fertile marriage. Another ethnic group has its most sacred tombs in the forests. Such areas have been tabooed for burning and wood cutting traditionally. They are part of the native conservation system. What happens when resources are placed off-limits by an external agency? Conservation schemes are like economic development projects in that they often ask people to change the way they have been doing things for generations so as to satisfy planners' (not local) goals. In places as diverse as Madagascar, Brazil, and the Pacific Northwest of the United States people are asked, told, or compelled to modify or abandon activities essential to the local economy because to do so is good for "nature" or "the globe." When people are asked to give up traditional activities on which they depend for their livelihood, they often resist. Well-meaning conservation schemes can be as culturally insensitive as development schemes that promote radical changes without involving the local people in planning and implementing the policies and programs that affect them.4 Yet the need for environmental preservation is real. The task is to devise culturally appropriate and effective strategies to reduce air and water pollution, curb forest degradation, retard global warming and the greenhouse effect, and preserve biodiversity. The ethnocentric rhetoric that Americans usually hear simplifies the problem of designing viable conservation strategy to a battle between environmental preservation and employment. The complexity of the issue is obscured by pitting the higher, future-oriented, and global morality of environmentalism against the pragmatic and immediate goal of preserving jobs. Development agencies, NGOs (nongovernmental organizations), and other groups will not succeed if they try to impose their conservation goals on communities (or nations) without paying attention to the practices, customs, rules, laws, beliefs, values, and organizations of the people to be affected. This caveat applies even when intentions are noble and the project is unquestionably "good for the globe." In contemporary Third World nations like Brazil (where Kottak has also done research on issues of human ecology, economic development, and social change since 1962) politicians routinely inveigh against Northerners (North Americans and Europeans) as interfering hypocrites. Brazilians berate Northerners for preaching about global consciousness and Amazonian preservation after having destroyed Northern forests in promoting First World economic growth. Northerners are also blamed for having sent predatory enterprises to countries like Brazil, where multinational corporations contribute to economic development and ecological devastation. Brazilians easily perceive the irony when American conservationists tell them to preserve the Amazon while the American government drags its feet on emissions standards and energy-efficiency measures needed to reduce global warming, to which our culture of consumption contributes far more than Brazil does. Moralism aside, effective environmentalism requires culturally informed negotiation with political and economic interests at the local, regional, national, and international levels. Environmentalists must also learn to deal with divergent sets of environmental values--different "ethnoecologies." (An ethnoecology is any society's traditional set of environmental perceptions, its cultural model of the environment and its relation to people and society.) The end of ecocide (or genocide, or ethnocide) requires much more than raising ecological awareness, getting nations to sign agreements, or merely declaring vast areas to be natural or indigenous reserves. American culture makes fetishes of "the law," "trust," and "honesty" (as values and ideology, if not as practice). We use the law to evaluate and settle issues all the time. But in many nations earnest people--including government officials--routinely make and sign laws they never expect to see enforced. Anthropologists have good reason to be suspicious about the effectiveness of laws, decrees, and treaties. In remote areas of the world (like rural Brazil and Madagascar) we have been observing "the human dimension of global change" for years. In such locales we too rarely encounter regulatory officials or mechanisms capable of doing much to halt the predatory expansion of the modern world system and its culture of consumption. In the face of that predation, if conservation measures are to have even limited success, they must be socially sensitive and culturally appropriate. Effective conservation strategy requires extensive and ongoing knowledge of affected areas and the socioeconomic and cultural practices of the local people. Conservation can no longer be equated with laws and fences. What are the alternatives? Researchers have an important role to play in the conservation of Earth's resources. Two associates of our 1991 project, Dr. Robert Sussman (a primatologist at Washington University, St. Louis) and Dr. Glen Green (a geologist and remote sensing specialist at Macalester College, Minnesota), had previously done research of the sort described in this paper. They did field research in the Beza Mahafaly protected area of southwestern Madagascar. This research included the study of nonhuman primates (lemurs), human use of the forest for medicinal plants (conducted by Dr. Linda Sussman), and the rate of deforestation (studied with the aid of remotely-sensed images). Preliminary research in Madagascar in summer 1990 by Conrad Kottak and Lisa Gezon5 and consultation with Green and Sussman led us to believe that northern Madagascar would be an ideal area to continue deforestation research, for several reasons. First, Dr. John Hough, the technical advisor for the WWF conservation and development project in northern Madagascar, had expressed an interest in working further with us to find ways of uniting researchers and policy makers to enhance conservation. Second, the northern area to be protected by the WWF, in collaboration with the government of Madagascar, included varied forest types, dynamics of deforestation, and social constellations. Third, the deforestation in the region was threatening the water supply of the rural areas and the nearby city of Antsiranana. Fourth, the biodiversity of the forests had not been extensively studied, nor had the forest types been mapped. The WWF project in northern Madagascar encompasses two large protected areas--Amber Mountain and Ankarana. These form the dominant features in two SPOT6 satellite images (Figures 4B and C) of our study area. To the north, the Amber Mountain National Park includes rain forests on Amber Mountain (Ambohitra). At lower elevations, the rain forest gives way to grasslands and gallery forests (river valley forests). In summer 1991 we surveyed the village clusters of Ankorefo and Bemanevika, located at the forest-grassland boundary on the western slopes of Amber Mountain. To the south, the Ankarana Special Reserve includes a mostly inaccessible massif dominated by pinnacle karst. Ankarana is surrounded by a ring of forests, and several valleys in the massif also contain forest. The largest river in the area has deposited alluvial soils. Forested until recently, this alluvium now supports cash crop agriculture. We studied the village of Antsambalahy, just south of the Ankarana massif, in an area of denser population than Amber Mountain. Our "from above" site selection procedure depended on the fact that Malagasy villages are identifiable in satellite images. Because the villagers sweep away the litter from between their dwellings, to drive off insects and scorpions, the villages show bare ground not usually present in the natural environment. This makes them spectrally distinguishable in SPOT multispectral satellite images (Figure 16B). People also plant trees around their village for food and materials. This ring of trees provides a sharp and high-contrast edge between the village and the surrounding terrain. Thus villages can be located and village size estimated using satellite images. Armed with information "from above," our 1991 field team arrived in northern Madagascar in June. We renewed our contacts with John Hough and other staff members of the WWF Integrated Project for the Amber Mountain Forest Reserves. We studied our satellite images and aerial photos in consultation with WWF staff and local people. We made final choices of village sites where integrated research seemed promising. We developed questionnaires to be used at our field sites. Out "in the field," Gezon and Green worked together at each village cluster. In the villages Gezon gathered basic social and economic information through structured interviews. She also compiled land use histories and elicited local perceptions of forest degradation in relation to resources. Green (accompanied by an assistant, a Malagasy translator and guide) did "ground truthing" in the surrounding countryside and forest. (Ground truthing is a term that geographers use to describe the process of correlating features, including vegetation types, on-the- ground with images from above.) Green often met people working in their fields or the forest. He interviewed many of them about local vegetational changes. He asked about plant names, irrigation practices, water availability, weed species, and dates of exploitation of wood by commercial timber companies. Gezon's socioeconomic research was aided by a three- person team hired by the WWF. In 1990 Kottak had recommended to USAID-Madagascar that the SAVEM project train educated Malagasy to be "para-anthropologists"/social workers at each protected area. This would be part of a flexible strategy that would encourage an ongoing dialogue with people in the project area and respond to their concerns. The para- anthropologists would collect baseline data, monitor project effects, and respond to problems in and near the protected areas. WWF, a major recipient of USAID funds for its conservation work in northern Madagascar, followed Kottak's recommendation. They hired three educated local people for us to train. These para-anthropologists would collect socioeconomic data for and monitor the WWF project in the affected villages. They became important members of our 1991 research team, and their work continues. The environmental, economic, and demographic diversity in our study area allowed us to study forest degradation in different contexts. We worked in rain forests and dry forests. We observed different economic processes (cash cropping, lumbering, slash and burn, and sustainable indigenous use). We found the Amber Mountain area to be less deforested than Ankarana. (Both were less degraded than the eastern rain forests previously studied "from above" by Green and Sussman [1990].) Our field sites illustrated different relationships between northern Malagasy populations and their environments. Ankorefo and Bemanevika, the villages west of Amber Mountain, have different socioeconomic characteristics and resource demands than do the villages south of Ankarana (e.g., Antsambalahy). The natural environment is also significantly different. Amber Mountain is covered mostly with rain forest. Ankarana is dry lowland forest. On the west side of Amber Mountain, the population isn't large enough to put a strain on the forest. Villagers' extraction of trees for house building is not so great as to have a negative impact. The threat to the forest here is commercial logging. The village of Ankorefo has had selective logging since early colonial times (circa 1900). About 50 hectares of forest have been lost since 1950; this area is now secondary grass land. At Bemanevika village the first commercial loggers arrived just a week before we did. The 20 kilometers of road we drove on from Ankorefo to Bemanevika had been hastily constructed days earlier. Before then, the route had been impassible even with four-wheel drive. While we were in Bemanevika, we heard chain saws daily and saw enormous felled trees carried out on trucks (see Figures 13C and 13D). These trees were going to a sawmill in Antsiranana, to supply the urban market there. We could observe first-hand the local reactions to this disruptive event. The villagers were concerned that the cutting might endanger their water supply, used for subsistence rice production. The company's legal right to log this land was questionable; their logging permit may have been falsified. Still, local opinion about the logging was divided. Some people wanted to try to force the company out; others wanted to let the company cut certain forests in exchange for compensation. Commercial logging degrades forests in several ways. Even when it is selective, logging can increase vulnerability to weather extremes and fires. Obviously, logging deforests because it removes trees. Less obvious are the destructive effects of road building, tree dragging, and other activities associated with commercial logging. The temporary logging road we followed into Bemanevika would soon degrade into a swath for erosion. Villagers complained that loggers kill a dozen trees for each log they drag out. We chose our field sites in northern Madagascar from space, but we needed on-the-ground research to determine the varied causes of deforestation. The reasons for forest degradation on Amber Mountain contrast with other parts of Madagascar, where the main causes are subsistence cultivation (the eastern rain forest), fuel wood collection (the southwest), or commercial agriculture (Ankarana--to be considered next). Cash cropping (sugar, cotton, and tobacco) is common just south of the Ankarana reserve. Some local people work for wages on plantations belonging to a sugar company. Others are sharecroppers for local land owners, or individual farmers if they have access to land. Commercial agriculture has attracted many immigrants and has created pressure on the land. The conversion from subsistence to cash crops has been the major cause of forest depletion here (e.g., in the village of Antsambalahy, our field site south of Ankarana). We can tell from aerial photos where cash-cropped fields are located, by their geometrical shapes. Their borders usually have straight lines. This reflects the use of tractors, which cultivate in straight lines. These contrast with the curved lines of subsistence rice fields. Subsistence farmers don't use any form of mechanization, and most don't own a plow. As a result, there is nothing to keep their boundaries from following natural contours. Our views from above and below let us test several hypotheses about forest degradation, which has multiple causes in Madagascar, as it does globally. Madagascar provides an ideal laboratory for studying the human dimension of global deforestation, since most of the global scenarios of deforestation are present. These include: demographic pressure (from births or immigration) on subsistence economies (including shifting, slash-and-burn cultivation), commercial logging, road building, cash cropping, fuel needs associated with urban expansion, and clearing and burning associated with livestock and grazing. The fact that forest loss has several causes has a policy implication: different deforestation scenarios require different conservation strategies. Even in Madagascar, conservation plans for the SAVEM project had to be site-specific, designed to fit varying circumstances. In the north, commercial rather than subsistence pressures were behind forest degradation. In the east population pressure on the subsistence economy ("slash- and-burn" cultivation) is the main cause of deforestation. We suspect this is also true in the northeast, especially on the large Masoala peninsula--the next ecological hot spot we plan to study. As another possible reason for deforestation, we also investigated, but found inconclusive support for, a regional model of population-environment dynamics. According to this model, immigrants, not natives, drive forest degradation. Madagascar's population is growing fastest in the central highlands, an area of intensive agriculture where people lack knowledge of and respect for the forest because their homeland is denuded. Intensive cultivators often see trees as a kind of weed, and we suspected that highlands emigrants would be strong deforesters. Moving to other parts of the island, including the north, they would cut back the forests to replicate the irrigated fields and bare landscapes of their homelands. We found some slight evidence for this in northern Madagascar, but commercial factors rather than population increase, either direct or indirect, was the main reason for degradation in our study area. Although the Malagasy highlands are denuded, scholars still debate the nature of the original highlands vegetation and whether its degradation is due to natural or human causes. One often-mentioned human cause is the annual burning of grassland just before the rainy season. The Malagasy do this to promote pasture regrowth for zebu cattle, which are ubiquitous in the highlands and play vital roles (manure, traction, transport) in highlands agriculture. In the north, we also investigated the possibility that the burning of grasslands leads to deforestation. We found a few examples of grass fires' having entered and degraded forests, but with damage to just a few hectares, and after logging had made the ecosystem vulnerable. Grass fires don't appear to cause deforestation. However, government attempts to curb forest fires have led to an ecologically unsound criminalization of the local population. Since fires are a natural phenomenon in savannas, making burning illegal creates tensions between villagers, the government, and conservation groups. (Fearing an accusation, villagers run away from natural fires, instead of trying to control them.) Our integrated study of northern Madagascar shows that local-level scarcity and poverty aren't the sole causes of deforestation. Although population increase poses a long- term threat to Madagascar's ecological balance, demographic pressure is not yet a direct threat to the northern forests. Even on this "island at the end of the world," national and international commercial interests play a strong role in ecological devastation. The culture of consumption stands alongside population increase as a driving force behind global environmental change. References Cited Green, Glen M. and Robert W. Sussman 1990 "Deforestation History of the Eastern Rain Forests of Madagascar from Satellite Images." Science 248(4/13/92): 212-215. Kottak, Conrad P. 1980 The Past in the Present: History, Ecology and Cultural Variation in Highland Madagascar. Ann Arbor: University of Michigan Press. 1990 "Culture and Economic Development." American Anthropologist 92(3): 723-731. 1991 "When People Don't Come First: Some Lessons from Completed Projects." In Putting People First: Sociological Variables in Rural Development, 2nd. edition, ed. Michael Cernea. pp. 429-464. Oxford University Press. 1Kottak has been studying Madagascar since 1966-67, when he did an extended ethnographic and socioeconomic study of the Betsileo of the southcentral highlands. His research focused on the social, political, economic, and cultural implications of the expansion of wet rice cultivation. 2Figure numbers refer to another presentation; they will be changed as used in the current article. 3Through Tropical Research and Development of Gainesville, Florida. 4In a 1983-84 study for the World Bank Kottak compared 68 economic development projects from all over the world, in order to assess the social and cultural factors that promoted project success or failure. He was also asked to assess the extent to which the social soundness and cultural appropriateness of innovations had been considered when the development projects were being designed and implemented. Kottak also examined the effect of social soundness and cultural appropriateness on the financial success of the projects. He found that the socially sound projects were twice as successful in their financial rates of return as the less sound projects were (See Kottak 1990, 1991). 5Kottak and Gezon are cultural anthropologists at the University of Michigan. 6SPOT stands for Systeme Probatoire d'Observation de la Terre.