Natural resource systems and human exploitation systems : physiobiotic and ecological linkages

THE WORLD BANK POLICY PLANNING AND RESEARCH STAFF Environment Department Natural Resource Systems and Human Exploitation Systems: Physiobiotic and Ecological Linkages Norman Myers November 1988 Environment Department Working Paper No. 12 This paper has been prepared for internal use. The views and interpretations herein are those of the author(s) and should not be attributed to the World Bank, to its affiliated organizations or to any individual acting on their behalf. This paper has been prepared by Norman Myers, a consultant to the Environment Department of the World Bank. The author is indebted for valuable comments and suggestions to a large number of people inside and outside the Bank; in particular he would like to thank John English, Ernst Lutz, Bill Magrath, David Pearce, Robert Repetto, Alfredo Sfeir-Younis, John Spears, and Jeremy Warford. Departmental Working Papers are not formal publications of the World Bank. They present preliminary and unpolished results of country analysis or research that are circulated to encourage discussion and comment; citation and the use of such a paper should take account of its provisional character. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Because of the informality and to present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. - ft - ABSTRACT This paper analyzes the causes, course, and consequences of human activities (also of natural processes where pertinent) that lead to resource depletion in developing countries and to delineate the lessons to be learned. Emphasis is placed on three stocks of natural resources -- soil, water, and forest cover -- and their exploitation is illustrated principally through deforestation (among other forms of land clearance), degradation of watersheds, soil erosion, siltation and sedimentation of water bodies, overgrazing of rangelands, and desertification. Thus the discussion focuses on natural resources that are in principle renewable, but that in practice tend to become nonrenewable. Each form of resource depletion is examined in turn to assess the development sector outputs affected (such as forestry products, agricultural activities, energy production); the differences in regional patterns and practices; and the future outlook for natural resource systems under various scenarios. Special attention is given to controversial issues and spillover effects, principally in the form of lateral and temporal externalities. Some distributive consequences are also discussed. Given the complexities of the issues, they are generally discussed in broad terms; a large bibliography permits the interested reader to go more deeply into spccific subject areas. The paper serves to set the scene for policy responses, such as sectoral initiatives and programs, project analyses at country level, and measures to engage in natural resource accounting. But while presenting a scientific and technical framework for policy responses, the paper emphasizes that many gaps remain in our knowledge and understanding of natural-resource factors and environmental processes. So the paper concludes with a chapter that raises some further key questions to be addressed, setting out an agenda for research priorities. "It is clear that the costs to economic progress, resulting from adverse environmental change, can be tremendous... It is equally clear that, in many cases, a small investment in prevention could be worth many times over what would have to be expended to repair the damage." Robert S. MacNamara, Speech to the U.N. Economic and Social Council, 1970 "Degradation and destruction of environmental systems and natural resources are now assuming massive proportions in some developing countries, threatening continued sustainable development... It is now generally recognized that economic development itself can be an important contributing factor to growing environmental problems in the absence of appropriate safeguards... A greatly improved understanding of the natural resource base and environment systems that support national economies is needed if patterns of development that are sustainable can be determined and recommended to governments." World Bank Annual Report, 1985 -iv - NATURAL RESOURCE SYSTEMS AND HUMAN EPLOITATION SYSTEMS: PHYSIOBIOTIC AND ECOLOGICAL LINKAGES Table of Contents Page No. I. CONCEPTUAL BACKGROUND ............................................1 The Case of Deforestation.................................... 1 Information Base: The Role of Uncertainty...................2 Environment as an Over-Arching Sector of Development.........3 II. ENVIRONMENTAL SECTOR: SOIL......................................5 Soil Erosion: Mechanisms....................................7 Extent of Erosion............................................8 Impact on Crop Productivity.................................12 Sedimentation of Hydropower Facilities......................13 Summary and Conclusions ..................................... 16 III. ENVIRONNENTAL SECTOR: WATER....................................18 Water and Disease........................................... 18 Water for Irrigation........................................ 19 Watershed Dynamics .......................................... 20 Summary and Conclusions.....................................22 IV. ENVIRONMENTAL SECTOR: FORESTS..................................23 Nonwood Products............................................23 Genetic Resources: Extinction of Species...................25 Environmental Services:.....................................27 1. Protection of Soil..................................27 2. Hydrological Services...............................28 3. Fisheries Connection................................30 Additional Services: ........................................30 1. Fuelvood-Fertilizer Linkages........................30 2. Tree Browse as Livestock Fodder.....................32 Summary and Conclusions .....................................32 -v-- V. GAPS IN KNOWLEDGE AND RESEARCH PRIORITIES ............... . 32 Assessment of Resource Stocks ................. ........... 33 Vegetation/Atmosphere Interactions..............................36 Ecological Discontinuities......................................39 Asymetry of Evaluation.........................................40 REFERENCES...........................................................43 NATURAL RESOURCE SYSTEMS AND HUMAN BEULOITATION SYSTEMS: PHYSIOBIOTIC AND ECOWOGICAL LINKAGES I. CONCEPTUAL ACKGROUN Certain components of the natural resource base in developing countries are being depleted. About this, there is little dispute. However, the precise character and extent of the depletion are poorly understood. If we are to formulate policy responses precisely tailored to the problem, we must gain a better idea than we have had to date of just what is going on. The critical questions to answer are these: How much resource depletion is actually taking place? What is its scale and scope? What causes it -- that is, to what extent is it due to man's activities as opposed to natural processes? What thresholds of irreversible injury are being approached? How does the depletion affect development of sectors such as agriculture, forestry, energy, and public health? How much depletion is too much? What types and levels of depletion are significant, serious, critical, intolerable? And over what time horizons? Our answers to these questions will help us to deal with the biggest question of all: What can and should we do about it? The Case of Deforestation Consider the case of deforestation in tropical areas, which should clearly illustrate some of the dynamic linkages between "natural resource systems" and "human exploitation systems." Forests not only produce wood -- both commercial timber and subsistence fuelwood, plus a range of so- called minor forest products -- but they also protect soils, retain moisture, and offer all manner of other environmental services. When they are unduly degraded or destroyed -- whether through extensive logging or fuelwood gathering, or through clearing for agriculture -- the loss amounts to far more than the elimination of a source of wood. The productive capacity of the exposed soil is rapidly depleted through nutrient leaching, laterization, dessication, and outright erosion. (Of course much depends u-on the type of land-management regime that follows the removal of forest cover; all too often, it consists of minimal management.) Soil fertility in formerly forested lands can be markedly impaired or even irreversibly reduced. Of course, we have to be careful not to be too absolute with our use of the term "irreversibility." Over a sufficiently lengthy period of time, generally measured in centuries rather than decades, even IMpersta grasslands and other degraded types of post-deforestation vagetation can be recolonized by woody species and eventually revert to forest cover -- or advance to a fresh form of forest cover. Serious repercussions can also arise in areas far removed from the site of deforestation. Water flows to downstream river systems may become excessive during the rainy season and unusually low during the dry season, as illustrated by the flood-drought cycles that now characterize many river systems below deforested watersheds. These cycles have an adverse effect - 2 - not only on flood-prone lands, but also on irrigation-dependent croplands. Furthermore, soil runoff and other erosion debris can cause downstream river beds to build up and thus may increase the risk of flooding. Runoff is also responsible for the sedimentation and siltation of irrigation channels, reservoirs, natural lakes, harbors, and offshore waters. In turn, these processes can have an adverse effect on irrigation agriculture, hydropower installations, domestic water supplies, port facilities, and inland and coastal fisheries. Unwanted consequences also derive from less direct ecological- economic linkages. Once forests are "mined" for fuelwood and their potentially renewable resource is harvested almost to extinction, rural households start to divert animal manure and crop residues from farm fields to house hearths. Thus, fertilizer benefits give way to fuel needs, even though cropland productivity may well be further reduced depending on season of diversion and soil type (Bornoud and Christoffersen, 1985) (though some recent analysis [McIntyre et al., 19853 suggest that in Africa at least the linkage is not nearly so pervasive as has been supposed). In addition, deforestation results in the extinction of large numbers of wild species, along with their genetic reservoirs and their potential contribution to agriculture, medicine, and industry. Clearly, the consequences of the misuse and overuse of the forest resource spill over not only onto other natural resources, but also on to other development sectors. These side effects may be felt by generations to come since the improper use of natural resource stocks today reduces the future capacity of the resource base to recover its productivity, or to withstand adverse circumstances such as severe shifts in rainfall patterns. Whereas a healthy resource base may be able to tolerate sizable fluctuations in rainfall, for example, a degraded resource base may not. The problem becomes compounded for small-scale farmers whose socio-economic status does not allow them to withstand major damage to their resources or other assets. They may be forced to join other "shifted cultivators" in practicing slash-and-burn agricultule in the upstream forestlands, and thus may bring about further deforestation. The process feeds back on itsalf, amplifying and adding to the harmful consequences of each round of the vicious circle. Information Base: The Role of Unetainty To understand the true sequence of the causes and effects of deforestation -- and eventually to identify the "pressure points" for policy interventions -- we must undertake a proper analysis of the problem. Such an analysis should proceed as follows: (1) we should compile a sufficiently accurate data base on deforestation; (2) we should delineate the direct linkages between the removal of tree cover and the environmental ills that are supposed to ensue; and (3) we should document and demonstrate the indirect ecological-economic linkages, such as the diversion of animal manure from fields to hearths. Up to now, however, we have encountered varying degrees of certainty, whether through lack of data or lack of understanding. Admittedly, many if not most environmental services, such as soil retention -3- and water-systems regulation, are difficult to quantify. Thus our uncertainty in this area is not surprising. Nor is the fact that much of the conventional wisdom on deforestation emanates from divergent statistical information and divergent interpretations (though certain of these divergences, notably between the YAO/UNEP [1982] study and the Myers study [1980) have proven to be more apparent than real). All the same, an effort must be made to be as precise as possible -- that is, to strive for an appropriate degree of "precise imprecision" -* when dealing with these various environmental-economic linkages. Environment as an Overarching Sector of Development Deforestation illustrates the many interdependency relationships between the natural wor'd and man's world. Nan is plainly dependent on nature for its many goods and services. Likewise nature is often dependent on man for its full productivity in support of human welfare. Without human intervention via management, technical inputs, and the like, nature cannot contribute much (whether food, energy, or water) in a form that is immediately suitable for consumption. Therefore, those concerned with the effective management of natural resource would recommend that policymakers introduce two overriding goals to their agendas: (1) they must strive to avoid depletion of the resource base, both quantitatively and qualitatively; and (2) they must find ways to engage in all such forms of legitimate exploitation as are sustainable (this applies especially to the developing world). This second goal requires clarification. A rational approach to the management of natural resources in developing countries, even in the face of profound degradation, must guard against an attitude of "hands off" or extreme preservationist spirit (except in a few salient instances, as explained below). This would be economically undesirable and politically unacceptable in developing countries. Rather, we should engage in exploitation to the full -- in the sense of generating many more goods and services than has been the case to date -- r&ovided, of course, that the exploitation is sustainable and the resource stocks are made more self- renewing than ever. This view is in accord with that of a leading U.S. environmental group, the Global Tomorrow Coalition, which defines sustainable development as "a process of change to meet the needs of people, as defined by them, without diminishing the potential for meeting their future needs, the needs of other societies, or those of future generations." It also links up with the definition of conservation in the World Conservation Strategy: "The management of human use of the biosphere so that it may yield the greatest sustainable benefit to present generations while maintaining its potential to meet the needs and aspirations of future generations" (IUCN, 1980). This approach to natural resources has evolved over the past 30 years (see, for example, the broad-ranging evaluations presented in Carpenter, 1984, and Dixon and Hufschmidt, 1986). This is well illustrated by the forestry sector. During the 1960s, the great concern was how to take a greater harvest of tropical forest prodlicts, especially hardwood timber. During the 1970s, the emphasis was on how to better develop - 4 - tropical forests. And in the 1980s the question has been how to manage forestry so that it can contribute to development overall. We should adopt the same approach to soil, water, and all the other critical resources. In other words, the environmental cause should not be perceived as a constraint on development. Rather, it should be regarded as a valid aspect of development itself, or as an overarching aspect of development that not only promotes the rational and sustainable use of natural resource stocks, but seeks to integrate environmental concerns into all related sectors (notably, agriculture, water, forestry, fisheries, energy, and public health). The problem should no longer be conceived as too much exploitation of natural resources, but as the wrong types of exploitation (for further analysis along these lines, see Myers, 1987a and 1988a). Thus the key question is no longer "How can we best safeguard the environment?" but "How can we best make more productive and expanded use of our natural resources in order to further human welfare now and forever?" Because this question has not been systematically and substantively answered in the past, some development planners (both in developing country government and international organizations) remain less than convinced that environmental values are intrinsic to development. Rather, they believe that such concerns divert funds and efforts away from the goal of economic growth. Until the economic benefits of sound management of the natural resource base, together with the economic costs of environmental deterioration, are better defined and quantified, where possible, the environmental cause will continue to go by default. This issue is all the more important for developing countries since they tend to depend more on their natural resource base than do developed nations (Collard et al., 1987; Peskin, 1987, Upadhyay, 1986). For one thing, mnst developing countries have extensive subsistence populations. A survey of 48 countries (Repetto, 1986a) reveals that in 19 of them subsistence activities account for more than 20 percent of GDP -- and this figure, large as it may seem, must surely be an underestimate as it does not account for many nonmarket production activities such as water carrying, food preparation, and various transportation activities within the household. Still more important, many developing country economies are primarily based on agriculture. In 1980 the agricultural sector in Asia as a whole accounted for 70 percent of employment. The country figures were 57 percent for Pakistan, 58 percent for Indonesia, 69 percent for the People's Republic of China, 71 percent for India, and 74 percent for Bangladesh (in contrast, the figure was only 12 percent for Japan). In 1981, exports based on natural resources amounted to 40 percent of all exports in Asia. The breakdown by country was 44 percent for China, 64 percent for India, 81 percent for Pakistan, and 88 percent for Bangladesh (only 14 percent for Japan) (World Bank, 1984). Central American economies are rather more advanced than those of most Asian countries, yet the sectors based on natural resources -- such as agriculture, forestry, fisheries, and hydropower-derived energy (including secondary industries) -- account for more than half of all employment and for the bulk of export earnings. This is not surprising in a region where almost 60 percent of the populations are rural (Leonard, 1986). - 5 - Furthermore, we can realistically assume that if a lower income country finds its natural resource base is decliting, it will have difficulty switching to other economic activities, because of severe shortages of capital and trained manpower. In contrast, industrialized, higher-income nations, by virtue of their sophisticated technologies and capital stocks, can buffer themselves against, and even compensate for, any rundown of their natural resources. When oil prices rise, for instance, high-income countries can employ all types of strategies to engage in more efficient consumption of oil, or to devise alternative sources of energy. Low-income countries have far less "maneuvering room" when one of their main sources of energy -- such as stocks of fuelwood -- declines. Soil erosion in developed nations can be offset in major measure through increased and sophisticated application of fertilizers or appropriate conservation measures, but few farmers in developing countries can afford this option. In the process of development, agriculture in most industrialized, high-income countries has become relatively concentrated in areas of good soil, perhaps with major irrigation and/or drainage investments. Marginal lands have either been abandoned or allowed to revert to forests, e.g., in north-central or eastern U.S., or is used for extensive, low-value grazing, e.g., in U.K. or France. In lower income countries this process is not far advanced. Agriculture is still expanding the area in use, and therefore, some at least of the population are attempting to survive in unsuitable conditions, and with minimal capital available to offset them. As a result, LDC agriculture tends to be extensive rather than intensive, and to occupy much larger sectors of available lands than is the case with the more concentrated farmlands of developed nations. This situation leaves LDCs all the more vulnerable to environmental deterioration such as erosion of topsoil and climatic vagaries. II. ENVIROMENTAL SECTOR: SOIL In the next three sections we take a closer look at several categories of environmental resources, starting with a basic one, soil. Generally speaking, soil is a meter-thick layer of debris that serves to support all terrestrial ecosystems and food chains. Topsoil, the sublayer with rich organic content and hence of particular interest, is often no more than one-quarter of a meter deep. There can hardly be a more critical factor in agriculture than this thin covering of soil. Even so, soil is being eroded and otherwise degraded at such high rates in certain parts of the developing world (the developed world, too) that the process threatens to limit some forms of agricultural productivity for the foreseeable future. Some authorities estimate that erosion may well cause the productivity of rainfed crops in developing countries to decline by 19-29 percent during the next 25 years (Higgins at al., 1982; Sfeir-Younis, 1986). (For some recent references on soil and soil erosion, see American Society of Agronomy, 1982; Blaikie and Brookfield, 1987: Brown and Wolf, 1984; Crosson, 1983; El-Swaify et al., 1985; Food and Agriculture Organization, 1984; Greenland and Lal, 1977; Kirkby and Morgan, 1980; Lee, 1984; Morgan, 1986; Pimentel et al., 1985; Quansah, 1981; Sfeir-Younis, 1986; Southgate at al., 1984; Stocking, 1980; and Wolman, 1985.) - 6 - Although soil is a mixture of renewable and nonrenewable elements, it tends to be treated overall as a nonrenewable resource with respect to the time horizons relevant to present human communities. Under normal conditions, soil forms at rates of 0.01-0.5 mm per year (occasionally as high as 0.83 mm per year in carefully cultivated lands). This means that tans if not hundreds of years are required to renew 25 mm of soil, or the equivalent of 400 tons per ha (Hall et al., 1982; Hudson, 1981; Pimentel et al., 1985). Yet an erosion rate of 10 tons per ha per year is all too common in developed countries, and 50 tons in developing countries. Although the rate of 10 tons is widely recognized as the top boundary of "acceptable" erosion, it is clearly far too high under most circumstances for natural soil co form in compensation. In other words, soil that is created at rates of centimeters per millenium is being eroded away at rates of centimeters per decade. The loss can be made good only by increasing the amounts of chemical fertilizer -- a form of "heroic use" of technology that soon leads to declining returns. Nitrogenous fertilizer on an eroded soil is often only one-third as effective as on uneroded soil (Lal, 1983). In any case, the option of ever-increasing amounts of fertilizer is a remedy not available to many developing country farmers. Since the development of agriculture some 12,000 years ago, soil erosion is said by some to have ruined about 4.3 million km2 of agricultural lands, or an area equivalent to rather more than one-third of today's croplands (Kovda, 1983). The amount of agricultural land now being lost through soil erosion, in conjunction with other forms of degradation, can already be put at a minimum of 200,000 km2 per year (Dudal, 1981; Food and Agriculture Organization, 1984). Soil erosion is therefore a principal and pervasive mode of degradation of an essentially nonrenewable resource. Some sizable costs may be involved: direct costs in the form of declining agricultural productivity, and indirect costs (or external costs) in the form of sedimentation and siltation of water bodies, such as irrigation canals, reservoirs, and navigable waterways. Another serious problem is that soil erosion is often difficult to discern. A loss of 40 tons per ha per year is enough to cause the productivity of most soils to plunge. Yet this rate generally amounts (depending on the type of soil and its constituents, plus their weight and volume) to less than 3 mm off the top of the soil profile per year -- even though it is enough to reduce the depth of a shallow (i.e., 40 cm thick) soil by 15 percent in 20 years. Conversely, a deep (i.e., 100 cm thick) soil with a moderate erosion rate may suffer only a 1 percent reduction in 20 years. Without doubt, soil erosion can be a deceptively insidious process, making it all the more difficult for policymakers to confront. Moreover, the removed 41 may contain more nutrients than the remaining soil (which is likely to feature a greater proportion of sand and clay); and erosion is often non-uniform, in that it can result in a surface seal, or crusting, that makes cultivation more difficult. Because soil erosion has a critical effect on agriculture, not to mention several other development sectors, we need to understand as much as we can about the mechanisms and extent of this phenomenon. Otherwise, we will be unable to appreciate the true overall costs of erosion, or the true overall benefits of countermeasures. -7- Soil Erosion: Mechanisms Soil erosion occurs through the action of water or wind, or both, but erosion through water tends to be more predominant. In the Great Plains of the United States, for example, wind erosion is only about half as great as water erosion. Note, however, that wind-blown soil crossing the Atlantic from North Africa can alone amount to 100-400 million tons a year (Prospero et al., 1981), and spring ploughing in China becomes detectable to scientists in Hawaii by virtue of the quantities of dust blown across the Pacific (Parrington et al., 1983). Erosion depends on various factors, including soil type, slope, vegetative cover, and rainfall intensity, plus the interactions of these factors with those of splash detachment and particle transport. None of these factors is independent of the others: the simple effects of each differ, and the magnitude of any one effect varies, in accordance with the level of the other interacting factors. After particles of soil have broken loose, the main "pathways" of erosion include suspension, surface creep, siltation, gulleying, and saltation (which refers to the action of wind when it carries particles in short jumps). There are two main sources of "erosive energy": rainfall and the slope of the land. Rainfall detaches particles from the surface of soil and transports them downslope. As rainfall on the soil surface, its volume and intensity produce a splashing effect that removes particles of soil. "Splash erosion" has been defined as a function of mean raindrop size, storm kinetic energy, and rainfall intensity (Hudson, 1981; Quansah, 1981; Stocking, 1978). In the tropics, rainfall is often torrential: a downpour can deposit 25 mm of rain in 15 minutes (or 10 times as much water as an average shower in the northeastern United States), and can remove 15 tons of soil from 1 ha. Not surprisingly, rainfall erosivity in the tropics can be 16 times greater than in temperate zones. The erosive energy of the slope of the land derives from its length. The faster or longer the slope, the faster the runoff, hence the greater the power of the rainwater to abrade the soil. In other words, the slope factor can affect erosion exponentially (McPherson, 1984; Posner and McPherson, 1982; Wischmeier and Smith, 1978). In Nigeria, the rate of erosion on cassava croplands is 3 tons per ha per year on a 1 percent slope, 87 tons on a 5 percent slope, and 221 tons on a 15 percent slope (Aina et al., 1977). Strongly related to erosive energy is the erodibility of the soil. When the texture, structure, and other properties of the soil enhance infiltration, for example, they reduce erosion. The rate at which water infiltrates the soil is determined in part by the permeability of the soil. A strongly compacted soil has few pores to accommodate water, whereas a soft and well-tilled soil allows water to soak in rapidly. Similarly, the distribution of soil particles is important, insofar as the stability of soil aggregates prevents particles from splashing away when rainfall strikes them; and silty soils are more easily eroded than sands or clays. - 8 - A fourth factor influencing erosion is the amount of organic matter, whether particulate or dissolved, in the soil. Soil that is deficient in organic matter is much more readily eroded (Pimentel, et al., 1985). Organic matter improves soil structure and aggregation, promotes water infiltration, enhances water-retention capacity, increases the cation exchange (particularly in acid soils), fosters aeration and tilth, retains otherwise fixable or leechable nutrients in a form available to plants, and supplies most of the nitrogen and phosphorus required by plants (Sanchez, 1976). Moreover, in tropical and subtropical regions, because of higher temperatures organic matter is more readily decomposed or washed away than in temperate zones. A final factor reflects the type of vegetation remaining on erodible land. The key concern is not the amount of vegetation but its density, together with the amount of litter underneath. Removal of forest cover, for instance, does not necessarily mean that the ground surface remains permanently denuded. On many deforested mountain slopes in the Himalayas there can be very dense ground cover of one sort or another and it can often serve a worthwhile soil-protection function, even though its makeup is very far from what was there before. Thus, a variety of factors must be considered before one can even begin to determine the probable erodibility of a particular soil or landscape. Extent of Erosion Now for the key question: How much erosion is actually occurring? Since erosion is a natural process, a certain amount will take place regardless of how humans use the land. Soils covered with natural vegetation lose 0.02-1 mm of soil per year, or one-quarter of a ton to 13 tons of soil per year (Schumm and Harvey, 1982). But when humans remove the natural vegetation for agricultural or other purposes, erosion rates can accelerate to as high as 45 mm per year, or almost 600 tons per year -- more in the case of certain "weighty" soils (Troeh et al., 1980). When grassland is converted to row crops, for example, erosion can increase by 20-100 times; and when forestland is converted, erosion may increase 100- to 1000-fold (El Swaify et al., 1982). As for the volume of soil lost, field findings range from just a few tons per hectare per year in flatish or carefully cultivated lands, to 100 tons per year on steep slopes in the Andes, in the middle reaches of the Yellow River basin in China, and in the black soil sectors of the Indian Deccan; and to 200-500 tons in some gullied localities of Nepal -- even to 1000 tons or more in some exceptionally severe cases of steep slopes with friable soils. An average rate postulated by a World Bank survey of developing countries as a whole (Sfeir-Younis, 1986) is 53 tons per ha per year, which accords well with a mean for nine separate studies of developing countries. The average figure is only 6 tons for Africa, 12 tons for Latin America, and 138 tons for Asia. Of course, a low average for all of Latin America is of scant interest to a farmer in the Acelhuate River basin of El Salvador, where the rate is likely to approach 200 tons per ha per year (Wiggins, 1981). *9- Despite the immense significance of soil erosion, detailed data are generally scarce in developing countries. According to several experts (El Swaify et al., 1982), "There is little or no documentation of the extent, impact, or causes of erosion in tropical zones." Fortunately, this situation is improving, albeit slowly. As Sfeir-Younis (1986) points out, "While the data are far from complete, statistics at the national and local levels are comprehensive enough to support the assertion that soil erosion should be an issue of public concern at all levels of decision making." Most developed nations are fortunate to have enough data to make use of the so-called Universal Soil Loss Equation, which is based upon a number of biophysical factors -- namely, soil erodibility, rainfall erosivity, angle and length of slope, crop management (e.g., rotational cropping), and conservation or erosion-control practices (e.g., minimum tillage and terracing) (Peterson and Swan, 1979; Wischmeier and Smith, 1978). Thus the equation supplies an empirically derived formula to predict average annual soil losses from croplands. Developing countries, in contrast, make little use of the measure because the necessary data about soil characteristics and topography are not widely available. They rely mainly on estimates obtained by applying sediment-delivery ratios to estimates of sediment delivered at watershed outlets. Although the technology is fairly simple and the costs reasonably low, these estimates relate to all land within watershed basins, not just agricultural land. However, some data are available for individual areas and particular croplands, such as maize fields, which thus afford a measure of comparison even though soil types and rainfall regimes are far from constant. In Nigeria, alfisols on 5 percent slopes have been losing an average of 7.1 tons per ha per year when cultivated continuously with maize; and 28 tons on 15 percent slopes (Lal, 1976). In Zimbabwe, gently sloping maize fields have been losing 12.3 tons per ha per year (Hudson, 1971). And in southern Brazil, 8-12 percent slopes have been losing 8-30 tons per ha per year (Quintiliano et al., 1961). In the main, then, maize- field soil losses in certain developing countries appear to amount to 10-20 tons per ha per year, within a range extending from 1 to 30 tons. In comparison, the United States is believed to lose an average of 18 tons of soil per ha per year from its croplands -- and some of the best agricultural areas, such as Iowa and Missouri, are losing twice as much (Pimentel et al., 1985). To cite further examples -- although of a more general nature and with additional variables mixed in -- Turkey is estimated to experience some degree of erosion in 85 percent of its lands, with 20 percent being moderately eroded, 37 percent severely eroded, and 17 percent very severely eroded (Food and Agriculture Organization, 1981). In Peninsular Malaysia, measurements indicate that three catchments of the Cameron Highlands lose 0.44 tons of soil per ha per year in forested areas (0.9 tons in hilly catchments), 8.8 tons in tea plantations, and 13.1 tons in vegetable croplands (Challeau, 1956; Joseph, 1985). In Madagascar, certain hilly deforested areas are reputed to be losing 25-250 tons per ha per year (Randrianarijaona, 1983). - 10 - Data from Mexico for the early 1950s (no better or more recent quantified findings are available) reveal that pockets of severe erosion, notably in the highlands of the north and in the mountainous tropical lands of the south, amount to 4 percent of the country's land surface; slightly to moderately eroded lands in central and southeastern sectors amount to 30 percent; and moderately to severely eroded areas in the northern and western sectors, to 33 percent (Olson, 1973; Sancholuz, 1984). According to a World Bank agriculturalist (Milford, 1986) with 15 years of intermittent field experience in Mexico, soil erosion is generally considered to be a "serious problem," especially in rainfed areas totaling some 30,000 km2 that are on 6 percent or greater slopes (comprising one- fifth of all rainfed cropped areas); watershed erosion is widespread in virtually all the catchments supplying the country's major irrigation areas in the northern and northwestern sectors. In other words, soil erosion affects some 1.4 million km2 of land: 28 percent mildly, 60 percent moderately, and 12 percent severely. Still more pronounced erosion, and on a broader scale, is found in India. Land degradation of all forms, principally soil erosion, affects almost 1.75 million km2 of the country's 3.3 million km2 of territory, with 60 percent of 1.43 million km2 of croplands suffering significant erosion. Overall, the country's croplands are reckoned to be losing at least 6 billion tons of soil per year (some officials believe twice as much), much of it going eventually into the sea (Indian Council for Agricultural Research, 1984; Nayarana and Sastry, 1985). (Of course if, as is usually the case, soil is merely transported from one part of a watershed to another without being lost to the region, that is not so harmful; see below.) During the past few decades India has achieved the remarkable feat of tripling its wheat output. But this has occurred mainly in the irrigated flatlands of Punjab and Haryana. In the rainfed lands where 80 percent of India's croplands are located, erosion is widespread, and according to conservative estimates, 30-50 million tons of produce are lost each year as a result (Indian Council for Agricultural Research, 1984). In China, total soil loss is roughly 4.3 billion tons per year from some 100 million hectares of croplands, or an average of 43 tons per ha (Brown and Wolf, 1984). One-third of Java's 4.4 million ha of upland cultivated lands are seriously eroding and 1 million ha of cropland have already been taken out of production, with another 200,000 ha being taken out each year; soil erosion there is reputed to account for at least 770,000 tons per year (Daryadi, 1981; Repetto, 1986a). In Nepal soil loss is some 240 million tons a year. In comparison, the total loss from U.S. croplands can be put at 2.8 billion tons -- with society-wide costs, both direct and indirect, estimated at US$35.4 billion annually, or US$220 per ha (Pimentel et al., 1985). Becau,;e the data are meager and sometimes poorly interpreted, soil erosion rates need to be treated with extreme care. Consider the case of Ethiopia, where it is regularly stated that one billion tons of soil are eroded away each year - - a figure that derives from an assessment made more than 25 years ago (Fournier, 1962). Ethiopia is reputed to have some 9 million ha of agricultural lands in its highland zone, which cover almost 500,000 km2 (or 43 percent of the country), receive 90 percent of total 3 ��� �� �� � �� � ��� � � � � � � . �� ' � � � ��� �� �� �� �� � �� � � � � �� �� �� �� �� �� �� � �$ . �� �� � �� � �� � ��� � ���� � � � � о Х Ч . р х х Х g х х + $ х ,ц � � g х �'�;��� � �� % х х х х х � � р � � х � � О х �� � � . ����'�� �' ��' ��33� � � й $�� 8i 8 S х х � а�� '� х а t% ������ �7.. ! � g �gg g � и � � S $ � � .цг'ti� R х � � х д �� �$� �� � f' х ��� . � � � ��� � � � � � ��� � � � �,,.. � �� � � � � Н й �� � ���� � � ���� ��� �� ��� � #� ��'.�� �� �� � Н . � �� � � .� �� �� � � � � � � � �� � �� � 4 � g а g " tS � � " х ^3 S.' $ й х � н � ��lt�=� и � х х �� х х � х х �� Q х� �� # 388� � S� � х 3 �Е й ""! � � tR � � 8 TR п����� � � У � х � � дt � � х х � х � �ц� . 1 rainfall, and support some 88 percent of the populace (Constable, 1985). It is further asserted that only about one-fifth of the soil erosion occurs in the and lowlands, which support only 12 percent of the populace and 35 percent of the livestock population; and that the rest of the erosion occurs in the highlands, where it is split so that roughly two-thirds occurs in cultivated lands (18 percent of the zone) and one-third in noncultivated lands. This means that soil erosion in the cultivated highlands must average around'60 tons per ha per year -- a high but far from unlikely figure for a mountainous zone being heavily farmed with rudimentary methods. At the turn of the century, the country was still 40 percent forested, almost all in the highlands, whereas now only 3 percent is forested. This drop has contributed highly to soil Orosion. In fact, what seems to be a more substantive figure has just been calculated for soil erosion in EthiopPW through a diversified series of test-plot measurements right across the highlands: 3.5 billion tons per year (Hurni, 1985). Using the crude calculations set out above, this suggests an average of just over 210 tons per ha of agricultural land per year, which is surely too high. Data on rates of erosion in selected countries has been sunn-grized in Table 1 reproduced from World Resources Institute (1987). Table 1 SOIL EROBION IN SELECTED CQM=9A, 12ZQ-19§6 * 12 - Impact on Cron Productivity Let us now move on to some of the impacts of soil erosion. It is not so much the rates of soil loss that ultimately count but the repercussions. The main direct consequence of soil erosion is that crop productivity declines (Food and Agriculture Organization and University of East Anglia, 1985; Follett and Stewart, 1985; Lal, 1983; Rijsberman and Wolman, 1984; Schmidt et al., 1982). As noted earlier, different soils tolerate widely differing rates of erosion before productivity is significantly reduced. The deep loess soils of western Iowa, for example, have been losing well above 10 tons per ha per year .for decades without suffering a serious decline in their inherent produdtivity. In contrast, a loss of less than 10 tons per ha per year appears enough to threaten the productivity of shallow soils in the southeastern United States (Crosson, 1983). Erosion reduces water availability, removes nutrients (such as phosphorus, Pitrogen, potassium, and calcium), reduces organic matter, and restricts ro. Lng depth as the soil thins. The loss of water is the major limiting factor for crop production. Erosion reduces the water-holding capacity of the soil by selectively removing organic matter and finer soil particles. When this happens the closely correlated property of water infiltration may decline by as much as 93 percent (Pimentel et al., 1985). In tropical areas, organic matter in the topsoil is more rapidly lost through erosion than it is in temperate areas. Next in order of importance is the depletion of nutrients. One ton of rich topsoil may contain as much as 4 kg of phosphorus, 10 kg Of nitrogen, 66 kg of potassium, and 72 kg of calcium (Allison, 1973). When certain U.S. croplands lose topsoil at an average rate of about 18 tons per ha per year, the loss of nitrogen fertilizer is equivalent to almost half as much as total chemical fertilizer applied to those croplands (Larson et al., 1983; Pimentel et al., 1985). Soil erosion in India costs farmers some 8.4 million tons of nitrogen, phosphorus, and potash, among other critical soil nutrients each year, which are worth, in terms of chemical- fertilizer replacements, US$6 billion (Sfeir-Younis, 1986). In Zimbabwe, losses of nitrogen and phosphorus (not only through soil erosion, but also through water runoff) are estimated to be worth US$20 to US$50 of fertilizer equivalent per year for each hectare of arable land. The country-wide total is US$150 million compared with fertilizer applications worth only US$50 million in 1984 (Elwell, 1985; Stocking, 1986). How much are crop yields actually depressed through soil erosion? Again, the data are meager at best. In general, crop yield declines exponentially as erosion increases, the drop being most marked during the early stages of erosion (Stocking, 1984). As a rule of thumb in the United States, the best maize yields appear to be obtained from not less than - 25-30 cm of topsoil. If this amount is reduced to 15 or only 10 cm, the yield falls away by 18 percent; and if to 10 or 5 cm, by 27 percent (Pimentel et al., 1982). That is, each centimeter of soil loss imposes a reduction in maize yield of almost 400 kg per ha (Lyles, 1975; Pimentel et al., 1985). To compensate for that decline requires at least 250 liters of oil equivalent in the form of fertilizer (Pimentel et al., 1981). - 13 - The yield decline is likely to be more rapid in tropical countries, largely because tropical soils tend to be intrinsically less productive in terms of their organic matter, nutrient content, structural properties, and water-holding capacity. Thus, a unit loss of soil in the tropics can be as much as 50 times more serious in reducing crop yields than the same soil loss in the U.S. grain belt (Stocking, 1985). For instance, erosion of 10 mm of soil in the U.S. maize fields leads to a 2-4 percent decline in yield, while the same level of loss from the same soil type in Nigeria results in a 75 percent decline. Put another way, erosion of 25 cm of some Iowa soils will reduce the maize yield by about 50 percent, whereas a loss of only 5 cm of thin soil in Nigeria can have the same effect (Sanders, 1981). According to extensive field research in the tropics, maize yields may decline by 10-30 percent with the erosion of "good* soils, and 30-70 percent, occasionally even more, with the erosion of "poor" soils (i.e., shallow and/or impoverished soils) (Langdale and Shrader, 1982). In parts of Mexico, maize yields have dropped from 3.8 tons to only 0.6 tons per ha as a result of severe erosion (Sancholuz, 1984), while in parts of Nigeria they have declined from 6.5 tons to only 1 ton (Lal, 1983). We should bear in mind, of course, that the eroded soil, with its nutrients, is not necessarily swept away to sea, and thereby lost to human use. As far as we can measure, only 20 percent or so is thus removed beyond human reach. Rather, the soil is simply shifted from an upstream part of a watershed to a downstream part (which raises questions of social equity as well as economic efficiency). This process gave rise to some of the great civilizations of the past, based as they were on rich alluvial deposits in valley bottomlands. Note, too, that today the process is inhibited in the Nile Valley, where the Aswan High Dam holds back most of the silt that formerly fertilized the croplands of lower Egypt, and thus served as a natural defense against desert encroachment. In some 7200 km2 of croplands, or one-third of Egypt's total cultivated area, the decline in annual silt renewal has led to an average reduction in crop production of about 25 percent, which is equivalent to some 8 percent of GDP in agriculture (1Kishk, 1985). Sedimentation of Hydropower Facilities The second main impact of soil erosion, an indirect impact this time, lies in the sedimentation of hydropower facilities. Again, this problem tends to be worse in tropical countries (primarily because of upstream deforestation on organically poor tropical soils) than in temperate zone countries. In central Europe, typical rates of storage- capacity depletion through sedimentation are 0.5 percent (Cyberski, 1983), and in the United States as low as 0.1 percent (Dendy, 1983; Sundborg and Rapp, 1986). In the tropics, however, rates are commonly about 2 percent, sometimes even higher. According to a review (World Bank, 1984) of development programs of 100 developing countries for the period 1982-1995, hydropower is expected to account for 43 percent of their electricity production by 1995, at an investment cost of about US$10 billion. But this 1995 level will * 14 - nonetheless represent only about 16 percent of technical hydropower potential (as of early 1982, only 6 percent on average had been developed, with another 5 percent under construction). Just six countries account for 60 percent of the potential: China, almost 25 percent, Brazil 12 percent, Zaire almost 7 percent, Colombia almost 7 percent, and India and Peru, 5 percent each. In certain extensive areas of each of these countries, except Zaire, erosion caused by deforestation is already a serious problem. The country with the greatest hydropower potential compared to land area is Costa Rica, yet the water catchment above virtually every large hydropower facility already constructed in the country has been deforested; at the Cashi Dam in particular, which has been operating for only two decades, revenues lost in the wake of sedimentation are estitiated at US$133 million to US$274 million (Leonard, 1986). The country with the eighth largest hydropower potential in relation to land area is El Salvador, yet it ranks as one of the most eroded countries on Earth, with its forests a matter of history. According to the World Bank study cited earlier (Sfeir-Younis, 1986), which simulates the effects of sedimentation in 200 major dams built since 1940, a 2 percent sedimentation rate -- which can be used for a constant average -- means that the live storage capacity of these dams will be reduced by one-third by the year 2000. If the sedimentation rate were to increase to as much as 4 percent per year (not unlikely in certain instances), the loss would amount to two-thirds of capacity by the end of the century. Even a 1 percent reduction in capacity would mean a loss of some 148,000 gigawatt hours of electricity (hydropower in the 100 countries contributed some 445,000 gigawatt hours in 1980, or 41 percent of total electricity demand of 1,090,000 gigawatt hours). To produce an equivalent amount of electricity by thermal means would require some 37 million tons of oil; and at US$80 per ton, or US$12 per barrel, the sedimentation-caused loss would constitute a direct cost of US$3 billion in electricity output foregone in the year 2000 alone. Consider that US$3 billion is equal to about one-tenth of all OECD and OPEC development assistance in 1984; (see Pearce, 1987). Nor would it be worthwhile to dredge the reservoirs: in Latin America dredging generally costs US$2.50 to US$3 for each cubic meter of sediment removed, in contrast to the present value of additional energy produced (were the active storage capacity available) throughout a dam's lifetime, namely way below US$2 per m3 removed (Southgate, 1986). Consider, for example, how sedimentation is affecting the Paute hydroelectric project in southeastern Ecuador, which cost US$600 million. The dam's watershed features a steep terrain and is heavily populated. Sedimentation appears to be increasing by 4-6 percent a year, and even the lower figure means that the minimum active storage capacity will persist for only 32 years instead of the projected 50 years. If an effort was made to control erosion in the watershed, at a cost of US$20 million during an initial 5-year period, the in-reservoir benefits (in present values) would be almost US$77 million (Southgate, 1986). In Pakistan, siltation has reduced the life of the US$600 million Mangla Dam reservoir from a planned 100 years or more to 57 years or less (West Pakistan Water and Power Development Authority, 1977). Pakistan also has the world's largest reservoir, the Tarbela, which is losing its storage * 15 - capacity of 12 billion m3 at a rate that will leave the dam useless within just 40 years (Indus Basin Research Assessment Group, 1978). At the Nizamsagar Dam in India, the siltation rate is 16 times greater than anticipated, and the reservoir has lost 60 percent of its storage capacity already (when 80 percent is lost, the economic value of storage capacity can be assumed to have become zero) (Agarwal and Narain, 1985). When finished, the Tehri Dam in India will be the sixth highest dam in the world; but owing to deforestation-derived sedimentation, the Indian Geologic Survey has already reduced its expected life span from 100 years to 40 or even 30 years. Sixth, the Ambuklao Dam in the Philippines, opened in 1956 with a planned 56 years of operational lifetime, but it is now reduced to only 32 years (Abernethy, 1984). Seventh, the Peligre Dam in Haiti, completed in 1956 as the central feature of a large-scale hydropower/irrigation plan, was originally expected to last 50 years, but will barely make 30 because of sedimentation (Williams, 1983). Probably the most notable example of sedimentation of a water body is the Panama Canal (Alvarado, 1985; Robinson, 1985). The Canal's 3230 kmz watershed is vital to the continued operation of the waterway, which produced gross direct revenues expected in 1985 of US$350 million (or around 8 percent of Panama's expected GNP.) Shipping is expected to grow by 1.2-2 percent per year from 1985 to 2010. The Canal is also a source of potable water for the conurbation of Panama City/Balboa. By 1983 some 70 percent of the watershed's forest cover had been cleared, as the land has been steadily appropriated for croplands and cattle ranches since the early 1900s. The Lake Gatun basin has been 90 percent deforested, but, because of the gently-sloping terrain, it has experienced little erosion and sedimentation. In contrast, the Lake Madden basin, which harbors most of the remaining forest in the Canal area, covers only 20 percent of the Canal's entire watershed, yet it supplies 40 percent of the Canal's water. Its slopes average 40-45 degrees, and the soil, being red volcanic soil, is unusually friable: within 3 to 5 years after deforestation, one-third of a meter of soil, being all the topsoil available, can readily be eliminated. Not surprisingly, erosion in the Canal's deforested areas is severe. During the 49-year period up to 1983, storage loss in Lake Madden has declined by 4.7 percent -- mainly as a result of the 20 percent deforestation of the watershed since 1960. If the present rate of deforestation in the watershed, about 800 ha per year, is held constant, the forest cover will decline from 77,000 ha to 56,000 ha by the year 2010; but if it doubles or triples, then the forest cover will shrink to 34,000 or 12,000 ha (Table 2). Even at a rate of 800 ha per year, the lake's storage loss will amount to 13.7 percent by 2010, and 18.4 percent by 2020. This will be serious enough in itself (while a good deal less than the 40 percent postulated for the year 2000 by Wadsworth, 1978). But the more likely rates, which are double or even triple the 800 level, will seriously restrict the Canal's operations and economic returns. *16 - Table2 PANAMA CANAL: LOSS O STORAGE CAPACITY AT LAKE MADDEN Forest Remaining in Watershed. at Various Deforestation Rates Double Triple Deforestation Deforestation Deforestation Rate - Rate - Rate - 800 halv. _1600 hab 2400 ha/vr 83 77,000 77,000 77,000 100 64,000 50,000 36,000 110 56,000 34,000 12,000 Loss of Storane Canacity at Lake Madden Interval Storage Period in Yeara I=s 1934-1983 *49 4.7% 1983-1990 7 6.5% 1990-2000 (vith present deforestation rate) 10 9.75 2000-2010 10 13.7% 2010-2020 10 18.4% * 25 years (since roughly 1958) of significant deforestation. Source: Alvarado, 1985. See also Robinson, 1985. Storage loss in the upstream reservoirs is not the only problem. During drought years when the level of water flows declines, sedimentation occurs in the Canal itself. In certain recent years, a number of outsize cargo ships have had to be re-routed around Cape Horn as a result. Summary and Conclusions Soil, one of our most valuable resources, is being progressively depleted at a substantial cost not only to agriculture but also to a good number of other leading development sectors. In the agricultural sector alone, if the widely accepted figure for erosion in Ethiopia -- one billion tons a year -- is correct, and if two-thirds of the erosion occurs in cultivated lands, grain cror output foregone is close to one million tons, - 17 - or two-thirds of all the relief food shipped to Ethiopia in 1985. Of course many other considerations enter into the calculation, but it gives an idea of the scale of the problem and its impact. As a rough indication of the future costs that may well be imposed by erosion on agriculture alone, the Food and Agriculture Organization (1984) calculated that, in the absence of soil conservation measures, the period 1975-2000 will witness a decline in agricultural production of about 6 percent in Southwest Asia, 12 percent in Southeast Asia, 25 percent in Africa and Central America, and 78 percent in South America, with an average for the developing world as a whole of almost 19 percent. Of course we must be careful not to overstate the situation. Some observers may ask whether there are any situations at all where a long-term balance persists between depletion and formation of soil. How we have survived several thousand years of widespread agriculture if soil lose through erosion is as serious as suggested above? The main answer is that erosion has become a significant problem only recently because of (1) the present extent and severity of the process, and (b) the lack of new lands for agriculturalists to move into when their erstwhile lands have become over-eroded. In some areas, soil loss simply does not matter: certain the floodplains of Bangladesh are perhaps 50 meters deep in fertile alluvial soil, and are constantly renewed, so a loss of even a centimeter a year is of little consequence. But many if not most areas of the tropics feature shallow infertile soils, so soil erosion can become a critical factor within just a few decades. In short, we face a large problem here. Yet because of the dearth of data, we do not even know how large a problem it really is -- whether in terms of soil volumes being lost on a country-by-country basis, or amounts of soil loss per unit of land area (square kilometer of watershed, hectare of cropland, etc.). Nor do we fully understand the implications for crop productivity under different circumstances, or the rates of reservoir sedimentation, or the sector-by-sector economic costs of erosion (Southgate, 1986). Soil erosion has been a significant phenomenon for decades, yet we have made little effort to document and assess its scope and scale, or its consequences. Indeed our steady indifference towa: soil erosion probably tells us as much about ourselves as about the phenomenon. We simply have not mobilized resource policies, analytic tools, or institutional capacities to help us come to prompt grips with what could rank as one of the greatest problems of our times. Until we can explain this indifference, we may not understand why the farmer likewise ignores this threat to his welfare. Hence we may fail to determine the best ways to intervene in the behavior patterns of the farmer (or the behavior patterns of our institutions). But precisely because the issue has been overlooked for so long, we may hope to recognize some major opportunities to deal with it. In fact, one such opportunity may be emerging already, in the form of documentation with the aid of remote sensing equipment. Two soil parameters -- organic matter and iron oxide -- relate to the reflectance characteristics of eroding soil (Weismiller et al., 1985). As erosion increases in severity, the organic matter content decreases and the iron - 18 - oxide content increases. Through the digital analysis of data from Landsat and SPOT satellites, we can gain some insight into the status of soil resources. Using similar techniques, we can also obtain satellite-imagery data of sediment loads in waterbodies. These inventorying and monitoring techniques will advance our understanding of the Earth's soil cover and what we are doing to it immeasurably. III. EMIRONMENTAL SECTOR WATER Water, whether fresh or salt, is an eminently renewable resource in principle. Its stocks can be recharged either through natural hydrologic cycles or through man-directed means. Yet all too often it is used as a nonrenewable resource (Ambroggi, 1980; Balchin, 1985; Falkenmark, 1984; Lundqvist et al., 1985). Moreover, natural circumstances rarely make water available in the right amount at the right place at the right time. In many parts of the world, precipitation is either excessive or deficient. This immediately complicates our perception of water as a resource. Although the total amount of precipitation per person worldwide now works out to some 16,000 liters per day, it is very unevenly distributed. The Amazon River, for example, discharges about one-sixth of the world's river flows, yet it does so from one of the least populated areas (Amazonia, which is as large as the Sahara, contains as few people as the Sahara). In many other parts of the developing world, human communities experience almost continuous shortages of water. To make matters worse, the demand for water in several parts of the world is increasing several times more rapidly than are population numbers, owing to rising standards of living. By the year 2000, at least one-quarter of all the annually renewing freshwater on the globe will be needed to sustain humankind. The following figures clearly show the imbalance between water needs and supplies in the public health sector alone: the proportion of human communities with ready access to safe drinking water is a mere 4 percent in Ethiopia, 6 percent in Mali, 10 percent in Afghanistan, 11 percent in Nepal, 13 percent in Mozambique, 14 percent in Sierra Leone, 16 percent in Uganda, 18 percent in Paraguay, and 19 percent in Sri Lanka -- to cite but the worst placed countries (UNICEF, 1985). Owing to population growth, 135 million more inhabitants of the developing world are now without access to clean water or adequate sanitation than in 1970. Of course these shortcomings are due as much to the lack of hygienic supplies of water as to outright physical shortages. But in several dry regions -- not only the Middle East but North Africa and the Sahel -- as much as 50 percent (a crucial level) of stable runoff from rivers is already being used, with adverse consequences for public health supplies. (Some details of water shortages, both present and projected, are shown in Figure IV.1.) Water and Disease The significance of these figures is reflected in the incidence of infectious diseases, particularly gastrointestinal diseases in developing nations as opposed to developed ones. In water-short communities, * 19 - household needs are supplied by local streams, lakes, and irrigation channels, many of which are contaminated by pollutants and human excreta. Water-borne pathogens -- which contribute in pe:;icular to typhoid, cholera, amoebic infections, bacillary dysentery, and diarhhea -- account for 80 percent of all disease in developing countries, and for 90 percent of the 14 million child deaths each year. Not only are the consequences of child mortality severe for public health, but they spill over to other sectors, for instance, via their impact on motivation for family planning, and thence onto plans for population growth, with all that implies for development programs overall. Yet the large budget for the United Nations Water and Sanitation Decade (US$300 billion over a period of 10 years) makes no provision for safeguarding water supplies from their principal source, namely forested catchments. Water-related disease is a particularly acute problem in urban communities. By the year 2000 there will be some 60 cities with five million inhabitants or more, and 10 times that number in urban communities. As the hinterlands of Manila and Panama City lose their tree cover and their supplies of domestic water decline in quantity and quality, the urban communities there will face a growing threat of contaminated-water pandemics. Similarly, public health programs in Bangkok, Lagos, Abidjan, and several other conurbations of the humid tropics are being set back because the quantity and quality of their water supplies are declining owing in part (so it is believed; see Section IV) to deforestation in upland catchments. Peninsular Malaysia is a tropical forest territory with some of the highest rainfall on Earth, yet water has been rationed for part of the year in Kuala Lumpur and several other urban areas (Barrow, 1981; Nor and Tang, 1980). Water demand in the Peninsula is expected to double before the end of the century; and if water supplies decline, costs will increase even more rapidly for the Malaysian consumer. Water for Irrigation Water shortages have an even greater impact on agriculture. Irrigated croplands, which now constitute about 18 percent of all arable lands and produce 33 percent of all food, account for 65 percent of the water used worldwide (Food and Agriculture Organization, 1981; Pimentel, 1986; Repetto, 1986b). Since 1950 these lands have inckeased from 94 million to 270 million ha. This expansion has been responsible for about half of the world's increased output of food during the past 35 years. If we are to grow twice as much food during the next two decades, at least half of the increase should come from irrigated croplands. Rather more than half of all irrigated lands are fn developing countries; tropical humid Asia, not counting China, accounts for more than three-fifths of irrigated lands in the developing world. Irrigated agriculture uses seven times more water than the next two water-demanding activities together, namely industry and domestic needs. Yet in several of the better irrigated parts of Asia, particularly certain sectors of the Indian subcontinent -- also Thailand, Indonesia, ard the Philippines -- the Green Revolution has bren losing some of its momentum as farmers can no longer rely on acceptable quantity flows of irrigation water for their multiple crops of bumper harvest rice each year - 20 - (Cruz et al., 1984; Jayal, 1984). In the Ganges River system, dry season water flows have declined by almost one-fifth during the 1970s (Bahugna, 1978). Much of India's success in attaining food sel!-sufficiency has been due to a doubling of the irrigated area since 1960. Yet because of the disruption of major water flows (allegedly through deforestation, see below and next chapter), also because of a surging demand for domestic needs and industry (as well as sheer rise in human numbers), India faces the prospect of an acute water shortage in much of its territory by 1995 (Agarwal and Narain, 1985; Vohra, 1986). In other countries of Southern Asia -- notably Pakistan, Thailand and Vietnam, also the island of Java -- the largest constraint on expanding rice production may prove to be not lack of land, fertilizer, agronomic improvement, and so on, but the lack of irrigation water in appropriate amounts at the right time of year (Colombo et al., 1978; Cool, 1984; Falkenmark, 1984; Gasser, 1981). A similar situation holds for much of Central America (United Nations, 1978). In the last quarter of this century Costa Rica plans to expand its irrigated croplands by 180 percent, El Salvador by 230 percent, Nicaragua by 300 percent, and Panama by 340 percent. Yet in all these countries, deforestation is proceeding apace, and one-third of the region's forest cover has been eliminated in the past three decades. Equally to the point, deforested areas, especially in upland watersheds, have not generaly been put to forms of land use that are sustainably productive, or receive even a modicum of sound management. Rather they are taken over by large- scale ranchers or small-scale cultivators who pursue their agriculture in a manner that is unusually disruptive to dependable year-round water flows from upland cathements. (As emphasized in the first chapter of this paper, it is not deforestation in itself that is antithetical to watershed services; it is the form of land management that occupies deforested lands -- and all too often, the management is notable only by its absence). Nor surprisingly, these countries' water needs are already catching up with stable runoff from watersheds. In several localities the situation is almost as critical as it is in much of the Indian subcontinent. A parallel situation exists in Africa along a broad belt bordering the tropical forests of the equatorial zone and the strip along the Gulf of Guinea (United Nations, 1978). For much of its extent, this forest belt is under pressure from farming populations that progrebsively degrade and destroy the forest cover. Irrigation agriculture is expected to encounter water shortages in several countries before the end of the century. Curiously, and despite the strong emphasis that should be placed on water supplies, the super-scale agricultural programs directed toward expanding irrigation pay little heed to the ultimate source of water supplies, namely forested catchments. As we shall see in Section IV, deforestation in upland areas is becoming a pronounced problem. Watershed Dynamics Water supplies, and hence watershed dynamics, play a central role in several sectors of development. Some 40 percent of the farmers in the developing world live in valleylands and floodplains, and hence are dependent, to some extent at least, on the activities of those 10 percent * 21 - of developing world people who inhabit the upstream catchments of nearby rivers. We still know very little about the extent to which the deforestation of upland catchments disrupts water flows in downstream areas, especially in distant areas (Borthakur, 1978; Easter et al., 1986; Haigh, 1982, which is an annotated bibliography with more than 400 items, and 1984; Hamilton, 1983; Hamilton and King, 1983; High Level Commission on Floods, 1983; Magrath and Grosh, 1985; National Commission on Floods, Government of India, 1980; Rieger, 1976; Singh et al., 1983; Williams and Hamilton, 1982). Forest clearing appears to expand water flows in two ways: runoff increases when forest vegetation is removed; and the deep- rooted trees of the forest are almost invariably replaced by shallower- rooted plants that use less water. This issue is addressed in Section IV, but it is touched on here too, in the phenomenon of flooding. Whether or not deforestation leads to flooding as a result of increased runoff, it surely does so via linkages with erosion. Erosion of topsoil, together with general debris from catchment areas, causes aggradation of stream and river beds, which in turn disrupts water courses and thus increases the chances of flooding. In Honduras, for example, inappropriate upstream land use and deforestation are blamed for damage to agriculture and infrastructure in the Sula River valley amounting to US$33.5 million a year, plus agricultural damage in the Aguan River valley amounting to US$4.5 million (Nations and Leonard, 1986). The worst flooding occurs in the Indian subcontinent, especially in India itself. The floodplains of India's rivers cover 1.25 million km2 or 38 percent of the country, and support more than 300 million people. Flood-prone lands have doubled from 20 million ha in 1970 to 40 million in 1980 (Agarwal and Narain, 1985; Devan and Sharma, 1985; World Bank, 1983). In the Ganges Valley alone, the floodprone lands now total 24 million ha with an average annual flooded area of 8 million ha, including 3.5 million ha of croplands. During the period 1953-1981, the total area affected by flooding has averaged 12.4 million ha per year, with a maximum in 1978 of 19 million ha. The population involved has averaged 27.4 million, with a maximum in 1978 of 70 million. The area of croplands damaged has averaged 3.6 million ha, with a maximum in 1978 of 10 million ha; and the value of crop damage has averaged US$248 million (in real prices), with a maximum in 1978 of just over US$1 billion. The number of cattle lost has averaged 97,400, with a maximum of 618,000 in 1979. Damage to public utilities has averaged US$110 million, with a maximum of US$582 million in 1981, and the total damage to public utilities, housing and other buildings, and crops has averaged US$407 million, with a maximum of US$1.7 billion in 1978. In addition, some 40,000 people drowned during this period. Whereas the average annual value of damage to crops, housing, public utilities, and other property during the period 1974-79 has come to just over US$1 billion (in real terms), the average annual outlay of flood- control works during the same period amounts to less than US$90 million. Control works have been confined mostly to embankments and engineering structures along rivers in the plains, with little in the way of preventive measures undertaken in upper watersheds. No fewer than 18 major catchments in the Ganges basin, covering some 50 million ha, require action of two kinds (World Bank, 1983): downstream flood-control engineering works which * 22 - are needed to protect an estimated 5.5 million ha and will cost US$1.2 billion; and upstream watershed-management measures on 2.8 million ha, which will cost an estimated US$870 million. Both of these proposed outlays are in the form of five-year programs; a 20-year upstream program, tackling problems in some 14 million ha would cost US$3.4 billion. Although the annual rainfall remains substantially the same, the severity of flooding is increasing. The largest amount of damage occurred in the 1978 monsoon season, and there have been subsequent peaks in 1979 and 1981. Damage appears to be caused by the combined effect of excess runoff of rain falling in both the plains and in the Himalayan range, but the damage caused by these two sources of rainwater should be clearly differentiated (World Bank, 1983). In Pakistan, too, principally in the river basin of the Indus, too much water is followed by too little each season (Malik, 1986). The country's populace is concentrated in the central and lower reaches of the river valley, which features the world's largest irrigation system; farming communities attain densities of 2400 or more per km2. In 1976 flooding inundated 80,000 km2 of croplands. Total flooding costs in recent years have amounted to more than US$3 billion. At the same time, there are threats of severe water shortages in the foreseeable future. Already the irrigation network diverts 72 percent of the river's average annual flow, and the present population of 102 million people is expected to reach 170 million by the year 2010. Yet all too little attention is directed toward safeguarding the forested catchments of the Indus in the Himalayas. Similar water regimes (albeit on much smaller scale) are developing in several other great river basins of Southern and Southeast Asia, where the demand for water, especially for irrigation water, is increasing rapidly, as is deforestation of upland catchments (Jayal, 1984). To cite an economist with long-term experience in the region (Cool, 1984), "Water budgets are becoming as critical to the national wellbeing in many Asian nations as the balance of payments and foreign exchange reserves. Great increases in efficiency of water use will become available through (improved) methods of irrigation, through land levelling, and more efficient on-farm controls. But to manage water for maximum productivity will require a discipline which few Asian societies1have demonstrated in recent times. In many areas, water is life. It is also power. As an awareness of its critical importance increases, and as water becomes scarcer and ceases to be a free good, tensions and conflicts may well erupt." Summary and Conclusions Water is extremely important to several development sectors, including agriculture (particularly irrigated agriculture), public health (and thereby family planning and population regulation), and urbanization. It also links up with still other sectors, such as communications (in the form of navigable waterways and port facilities). The problems and opportunities connected with water resources are linked to other sets of problems and opportunities, notably those associated with soil (Section II) and forests (Section IV). - 23 - Although waterflow dynamics in tropical catchments is crucial to our understanding of watershed dynamics, we are still at the stage of trying to ask the right questions. At a time when hundreds of millions of dollars are spent annually on flood-control measures, we need a systematized effort to determine the nature and functions of large-scale watersheds. That we are still far from attaining the understanding we need, is due to our erstwhile approach to water, which tends to be treated as common property and hence is considered nobody's responsibility. The fact that it is a readily renewable resource is taken for granted, with the result that water is wasted on every side *- and in many areas It is now becoming scarce. IV. ENVIRONMENTAL SECTOR: FORESTS Forests of the developing world, which are tropical forests for the most part, are not always thought of as environmental resources. They are typically perceived as standing stocks of wood, whether commercial timber or fuelwood. The well-known report by the Food and Agriculture Organization and the United Nations Environment Programme (1982), Tropical Forest Resources, discusses these forests as sources of material goods, while omitting their environmental services. Yet tropical forests supply myriad services, especially soil protection and the regulation of hydro- systems, the sizable value of which is frequently not appreciated until the forests have been degraded and the services decline. Many of these services are manifested in relation to other resources, notably land and water; and they function in relation to several development sectors in addition to forestry, notably agriculture and energy. As has been pointed out elsewhere (Spears, 1984), "Forestry sub- sector strategy studies rarely take into full account the vital role that forests play in regulating water supplies, protecting soil and water resources, reducing downstream sedimentation, minimizing the risk of flood damage, and ensuring regular supplies of clean and safe drinking water for downstream human and livestock populations. Rarely are the positive effects of forests and trees on agricultural productivity quantified." In this section we examine these key contributions of forests and demonstrate their significance. Of course all this implies that it is important to understand what forests can contribute to the development process. Without this understanding, we shall fail to properly manage and exploit forest ecosystems. With it, we shall be able to expand our benefits from tropical forests. Before we go on to consider forests' environmental services in detail, let us look at a group of material goods that generally receive far less attention than timber or fuelwood, and that suffer much attrition when forests are overutilized for timber or fuelwood (or are cleared to make way for agriculture and settlement), namely, nonwood products. Nonvood Products Nonwood products, sometimes known as minor forest products (although rattan, a leading item, is woody), include all products other - 24 - than timber and fuelwood that can be harvested from tropical forests. They range from damar, sandalwood, resins, kopal, and several essential oils and edible oils, to fruits and nuts, fibers and canes, natural silk, and exudates. Almost all can be harvested with only marginal disturbance of forest ecosystems. Apart from their local use, these products make many contributions to industry. A host of latexes are available (in addition to natural rubber), together with gums, waxes, dyes, and tannins, to name but a few leading categories of industrial materials (Balick, 1979; Myers, 1983a, 1984 and 1988b; Oldfield, 1984; Schultes, 1980; Staritaky, 1980; U.S. National Academy of Sciences, 1975). We are quite likely to use one or another of these ingredients each time we apply a deodorant, an after-shave lotion or a lipstick, each time we use cellophane or dynamite, each time we read a glossy magazine or mail a letter, each time we apply varnish to our furniture or our fingernails, each time we listen to a record or wield a squash racquet, and each time we get out our Gucci accessories or pull on our jogging shoes. These diverse products, also known as minor forest products, turn out to be-not so minor when we consider their variety and value (Lassehuit and van Eerd, 1983). In Peninsular Malaysia, the lowland primary forests contain at least 1,283 non-timber plant species of identified use to humans, constituting roughly one in six of all native species in the Peninsula (Jacobs, 1982). This total does not even take into account many additional food plants that are garnered from the forest in small quantities; nor does it take into account medicines, and it includes only a few items from several other major categories. The true proportion could be as high as one plant species in three. But if we stick with the one-in- six estimate (albeit derived from a meager data base), and extrapolate it to the entire tropical forest biome, we find that a minimum of 15,000 plant species could well be a source of material goods. What are some of the economic values involved? Rattan exports from Indonesia are now worth some US$90 million per year (Cornelius, 1984). In 1981 Indonesia's exports of patchouli oil were earning around US$12 million a year, and exports of related oils brought in more than US$25 million (Tcheknavorian-Asenbauer and Wijesekera, 1982). Indonesia's non- wood products, together with exudates and sundry other products, totalling 80,000 tons in all, brought in foreign exchange totalling around US$200 million in 1982, up from US$28 million in 1973 (Gillis, 1986). Admittedly, the harvest of Indonesia's minor forest products is more widely documented and better quantified in commercial terms than is often the case in other tropical forest countries. Although the data are not characteristic of all tropical forest countries, they reveal the potential that awaits methodical development. Moreover, these figures reflect only a crude estimate of total revenues, which can be compared with export earnings from wood products of US$583 million in 1973 and US$899 million in 1982 (Gillis, 1986). For a more substantive and systematized assessment of minor forest products, consider the case of India (Gupta and Guleria, 1982). In 1977 total net revenues accruing to India's forestry sector, from all sources - 25 - including commercial timber, amounted to US$336 million. Nonwood products accounted for US$134 million, or 40 percent of this total (and their share of forestry exports, 63 percent). Since an estimated three-fifths of all minor forest products are used by local people -- that is, they are consumed on the spot -- they do not enter the cash economy, and hence they are only marginally incorporated into national accounting figures. A realistic figure for the value of nonwood products would therefore be US$200 million, if not more. Among the leading categories in 1977 were medicinals, drugs, and pharmaceuticals, US$38.4 million; lac and lac products, US$19.8 million; gum, resins, and balsalms, US$14.6 million; bamboos, US$6.8 million; and essential oils, US$5.9 million. Equally important, the rate of growth in revenues from India's nonwood products during the period 1970-77 amounted to 15.6 percent per year, which was far ahead of the rate of growth for commercial timber. In addition, nonwood products were generating employment in excess of 70 percent of the 2.3 million man-years in the forestry sector overall. The true figure for employment, including the man-years not counted by official surveys, could have been as high as 4 million. In addition, tropical forests supply "wild meat". In Nigeria, where the moist forest cover has been reduced by at least 90 percent, local people still derive a renewable harvest of almost 100,000 tons a year from grass-cutters (giant rats), small antelopes, and sundry monkeys. On average, this wild meat constitutes one-fifth of all animal protein for people in Nigeria's forest zone. In Zaire it amounts to almost 27 percent, and in Cameroon, the Ivory Coast, and Liberia 70 percent (Ajayi, 1979; Sale, 1983). In Sarawak, local people obtain about 20,000 tons of meat from forest animals each year, which corresponds to an average consumption of about 12 kg per person per year (Caldecott, 1986). To supply the meat from alternative sources would cost at least US$100 million. Heavy commercial logging causes the sustained-yield harvest of this wild meat to fall from about 54 kg per local resident per year to about 2 kg. In Ecuadorian Amazonia, local people have long taken a self- renewing harvest of wildlife products, partly in the form of meat for consumption and partly in the form of skins and hides for export. As much as 85 percent of the animal protein consumed by local people comes from wild animals, notably peccaries, deer, tapirs, pacas, and agoutis, among some 40 species of mamals in all. A sustainable harvest of wild meat can amount to 240 kg per km2, with a market value of about US$1.8 per kg, or almost US$440 overall (Paucar and Gardiner, 1981; see also Vickers, 1984). (Were the harvest to be systematized, and expanded to include birds, turtles, and fish, the minimum potential value could perhaps be increased several times over.) Genetic Resources: Extinction of Secies One almost entirely unexploited category of nonwood products has the potential to become more valuable than all the others put together: namely, the genetic resources of wild species that contribute to medicine, agriculture, and industry. In the medicinal sector, tropical forest plants supply startpoint materials, such as alkaloids and other biochemical - 26 - compounds, for the manufacture of many analgesics, tranquilizers, diuretics, laxatives, and antibiotics. Over the counter sales of these products worldwide amount to some US$20 billion per year (Farnsworth and Soejarto, 1985; Myers, 1983a and 1984; Principe, 1985). The contraceptive pill, for instance, depends principally upon steroidal materials from a wild yam in Mexico and the calabar bean in West Africa. Another notable example is the rosy periwinkle, a native of Madagascar's forests. This plant has supplied two potent drugs for use against Hodgkin's disease, leukemia, and other blood cancers. Commercial sales of the two drugs now top US$160 million worldwide per year, while economic benefits in the United States alone can be estimated at more than US$300 million in a year (Myers, 1987b). Of course the genetic materials contributed by the periwinkle constitute only a small part of the end- product drugs; along the way there have been many other contributions, in the form of field research, laboratory testing and development, marketing and sales, and so on. But the drug company and its team of experts are, in this case, no better than the genetic materials they have to work with; the periwinkle contributes a small but essential component of the ultimate product. According to the National Cancer Institute of the United States, tropical forests may well harbor another one dozen plants with a capacity to generate superstar drugs against cancer (Duke, 1982). Tropical forests also contribute many new foods, plus improved forms of existing foods (Martin and kuberte, 1979; Oldfield, 1981; Samson, 1980; U.S. National Academy of Sciences, 1975 and 1979; Williams et al., 1975). In a montane forest of south-central Mexico, a primitive relative of modern maize was recently discovered in three tiny patches covering a mere 4 ha, apparently its last habitat (Iltis et al., 1979). Not only is the wild species a perennial (unlike all other forms of maize), but it is resistant to at least four of the eight major viruses and mycoplasmas that have hitherto evaded maize breeders (Nault and Findley, 1981); each year these diseases cause at least a 1 percent loss to the world's maize harvest, which is worth more than US$500 million. Even more important, the wild maize, discovered at elevations of 2500 to 3200 meters, is adapted to habitats that are cooler and damper than established maizelands. This attribute may make it possible to expand the cultivation range of maize by as much as one-tenth. All in all, the genetic benefits supplied by this wild plant, which was an endangered species when it was discovered, could translate into several billion dollars per year (Fisher, 1982). Consider, too, the commercial and economic values of medicinal plants, as attested by data from India, the Philippines, and Java (Jacobs, 1982; Perry, 1980). At least 5 percent of all plant species in the tropical forests of these areas are said to serve the cause of medicine, whether through established commercial markets or through localized "green medicine". At least half of these plants occur in lowland primary forests. Such forests are abundant in the Malay Peninsula, Borneo, and New Guinea and it is surmised that each supports some 200 important medicinal species (Jacobs, 1982; Perry, 1980). A plant-derived drug on the world markets can achieve commercial sales ranging from US$30 million to US$250 million a year (Farnsworth and Soejarto, 1985; Myers, 1984). If US$50 million is taken as a safe low average, these three sectors of Southeast Asia could theoretically generate plant-derived medicinals worth US$10 billion per year. - 27 - Of course the plants would contribute only a small part of the overall commercial value, and in some instances drugs would overlap in terms of their therapies. These estimates are therefore highly speculative and probably excessive. Nonetheless, the fact remains that tropical forests offer much scope for sustained exploitation of their medicinal plants (UNESCO, 1981; UNCTAD; 1982). Without doubt, cornucopias of new nonwood products await investigation and development. Recently it was discovered that one of the local rattans in Peruvian Amazonia provide a better-quality fiber than rattans from Asia, and at a fraction of the cost (Gentry, 1984). Needless to say, such new products will become available only if forest species are not eliminated as a consequence of conventional modes of forest exploitation. From this standpoint, we may well argue that tropical forests tend to be overexploited and underutilized (Myers, 1983b). In fact, many wild species of plants and animals are becoming extinct at unprecedentedly high rates. Tropical forests cover only 7 percent of Earth's land surface, but they harbor at least 50 percent, and conceivably 90 percent, of all the Earth's species. Yet, given present and projected rates of forest degradations, we can reasonably estimate that several hundred, and possibly several thousand, species are being eliminated in these forests each year (Ehrlich and Ehrlich, 1981; Myers, 1985; Raven, 1987; Soule, 1986; Wilson and Peter, 1988). In another half century, several million species could disappear. The value of these forests does not end with their nonwood products, and the wild species from which they derive. Equally important are their environmental services. Environmental Services 1. Protection of Soil As already mentioned (Sections II, III), tree cover protects the soil. The multistoried structure of a tropical forest, together with its luxuriant vegetation (often four times as thick as that in a temperate-zone forest), breaks the physical impact of tropical storms. Although tropical forests cover a mere one-sixteenth of Earth's land surface, they receive almost half of all rainfall on land. About one-quarter to one-third of this rainfall reaches the forest floor as drizzle or drip-runoff down tree trunks because it is intercepted by the forest (Jackson, 1971; Jordan and Heuveldop, 1981; Leopoldo et al., 1984; Sellers and Lockwood, 1981). The canopy of an undisturbed dipterocarp forest in Southeast Asia has been found to intercept at least 35 percent of the rainfall there. In contrast, the canopy of a heavily logged forest intercepts less than 20 percent, and that of a tree plantation, such as rubber or oilpalm, only about 12 percent (Ba, 1977; Daniel and Kulasingam, 1974; Low and Goh, 1972). There is also little soil erosion in the forest. But when the forest is replaced with row crops, the rate rises by 100- to 1000-fold (El- Swaify et al., 1982). Data from a number of sites in Malaysia, including upland areas with slopes of various gradients (Burgess, 1973; Douglas, 1978; Leigh, 1973; Leigh and Low; Low and Peh; 1985; Morgan et al., 1982; - 28 - Nik, 1985; Peh, 1981) indicate that erosion on maize croplands and oilpalm plantations can be 11 times higher than in primary forest, 12 times higher on groundnut plantations, 20 times higher on tea plantations, 34 times higher on vegetable croplands, and 45 times higher on bare soil. To be sure, in western Kenya forest cover has been replaced by tea plantations, with little or no erosion once the crop became established. But the positive results can largely be attributed to the highly permeable soils and climatic conditions of the site, so that these findings have limited application to tther areas (Blackie, 1972) Moreover, much of the outcome depends on the types of land management employed. In certain areas, notably with respect to oilpalm and tea plantations, careful management can minimize sedimentation and offer a sound measure of protection for soil and water resources. Trees not only act as a physical barrier to the impact of a tropical downpour and thus stem erosion, but they also help to conserve soil moisture. They need not constitute a regular forest to supply their services. As is well known, mere lines, such as shelterbelts on farmlands, protect the soil against wind erosion. They also protect it against the drying-out effect of hot winds such as those that frequently blow across the Sahel (Anderson, 1986; Bognetteau-Verlinder, 1980; Magrath, 1979; Spears, 1985; Steeds, 1985). They perform this function by helping to reduce evapotranspiration from crops (they further protect crops by reducing the "sand blast" effect of drifting soil particles). Field experience indicates that, on their windward side, shelterbelts are effective across an area five times the height of the trees, and on the leeward side 20 times their height (Anderson, 1986; Magrath, 1979). An investigation in Niger reveals that wind speed at a height of 1 m in the protected area is reduced by 45-80 percent (while the soil-erosion speed of 15-25 km per hour is not reached at all); and soil moisture is markedly increased at a depth of 80 cm (Bognetteau-Verlinder, 1980). Roughly speaking, then, 20 strips of shelterbelt trees, each 5 km long and planted 200 apart, can protect a cropland zone measuring 20 km2 (Anderson, 1986). Such shelterbelts take up only 1-7 percent of the cropland zone but they have a marked effect on crop productivity. Studies have shown a 13-17 percent increase for maize in Egypt, 38 percent increase for wheat in Egypt, up to 49 percent increase for rice in China, 11-59 percent increase for various crops in Niger, and 10-15 percent increase for various crops in India (50-70 percent increase in dry years) (Magrath, 1979; Sartorius and Henle, 1968). A further survey of 22 projects in 6 countries shows increases amounted to 0-10 percent in two cases, 10-20 percent in six, 20-30 percent in seven, and 30-50 percent in three (Anderson, 1986). 2. Hydrological Services On reaching the ground, rainwater penetrates the earth. Field research at Bihar, India, reveals that infiltration under forest cover is 260 mm per hour, under grass only 120 mm, and under crops only 90 mm (Tejwani et al., 1975). Similar rates have been reported from Peninsular Malaysia (Joseph, 1985). The forest soil acts as a kind of sponge, absorbing rainwater before releasing it in regular amounts. A 13-month - 29 - investigation of two small adjacent catchments in Malaysia, one with primary rainforest and the other mainly under rubber and oilpalm plantations, shows that while the peak runoff per unit area of the forested catchments is about half that of the plantations, the low flows are roughly double (Daniel and Kulasingam, 1974). In the Tai forest area of southwestern Ivory Coast, rivers flowing from primary forest release twice as much water halfway through the dry season, and between three and five times as much at the end of the dry season, as do rivers flowing from coffee-plantation zones (Dosso et al., 1981). Other field research reveals that tree cover can reduce water runoff rates as much as 10- or even 25-fold (Charreau, 1972). Thus forested catchment rivers serve to reduce any drought effect for valleyland farmers who depend on uninterrupted supplies of irrigation water for their multiple crops each year. Conversely, reforestation of cleared watersheds can reduce local flooding, as well as sedimentation of downstream water courses, by 90 percent (Pryor, 1982). Of course much depends on the size of the watershed: the larger the area and the more extended the hydrological system, the less marked are the regulatory effects of lpstream forest cover. The effect of deforestation on flooding is also evident in Upper Amazonia (Gentry and Lopez-Parodi, 1980), although deforestation there is not a fraction of that in Southeast Asia and West Africa. Most human settlements in Amazonia are located along rivers, and their lifestyles are closely tied to the annual flood cycles. These farming communities occupy the relatively flood-free sedimentary terraces known as "restingas," which are formed by alluvial deposits bordering the rivers. They pursue their agriculture in critical equilibrium with the floods: crops are often lost when the yearly flood arrives earlier or lasts longer than usual -- as is now apparently the case, in the wake of extensive deforestation in Peruvian Amazonia. Many people downstream from the deforested territories find that the floods are more extensive than they used to be (even though there has been no change in the precipitation amounts), and that they severely damage their crops. Meanwhile, the upland sectors of Peruvian Amazonia, with their generally rich soils, continue to attract millions of subsistence cultivators. The overall situation is much the same in the Amazonian sectors of Colombia and Ecuador. The trees in a forest act like pumps. Tropical trees tend to be deep-rooted plants, and thus tap deeper water stocks than do temperate-zone trees. The treers in a forest in Java draw 2000-2300 mm of water, and those in the Zaire basin draw 1230-1510 mm -- in contrast, those in the savannah draws only 950-1100 am. When a forest is cleared, the water runoff can increase by as much as 150 times. Note, however, that strip mining increases runoff by as much as 1000 times, and urban development some 20,000 to 40,000 times; (see Sundborg, 1983. For further analyses along these lines, see Bonell et al., 1981; Gilmour et al., 1982; Kartawinata et al., 1981; and Lal, 1981.) How far deforestation actually contributes to downstream flooding is debatable. A number of recent studies (Hamilton, 1983; Hamilton and King, 1983; Williams and Hamilton, 1982) suggest that deforestation does - 30 - not so much cause downstream flooding, as intensify it -- because of the reduction in water infiltration, and in the soil's storage capacity, the increase in water runoff, and the like. Thus, although upstream types of land use can be correlated with the frequency of downstream flooding (plus flood damages), the degree of intensification depends on the complex interactions of many factors, notably topography, geology, soil types, vegetation cover, and human activities. The dynamics of watershed systems are such that the localized effects of forest clearing (such as faster runoff of access water, and earlier and higher peak floods) may quickly be overshadowed by other factors (such as the nature and intensity of rainfall, and the size and morphology of river basins). For instance, soil erosion upstream can increase sedimentation in river channels downstream, and thereby indirectly intensify the flooding phenomenon. Moreover, flood damage may increase when the monetary value of crops, structures, and other human property is taken into account, not to mention the larger number of people who usually take up residence in floodplains. 3. Fisheries Connection The silt carried downstream by certain rivers can cause damage to estuarine and offshore fisheries. Significant damage has already been reported in the Philippines, with its 7000 islands and 18,400 km of coastline (which is almost as long as that of the United States). The coral reefs here cover 44,000 km2 and supply about one-tenth of the country's commercial fish catch, and one-tenth of the subsistence catch. However, the reef fisheries are declining in several areas, partly because of sedimentation problems that derive from deforested catchments (Davey and Graham, 1982; Hodgson and Dixon, 1987; Myers, 1988c; White and Wells, 1982). In addition, many mangrove ecosystems appear to be suffocating beneath silt loads that increase year by year. In the late 1970s the country was exporting US$100 million worth of oysters, mussels, clams, and cockles each year, a valuable trade that is now being threatened by the degradation of inland watersheds. In 1980 the contribution of fisheries to GNP (99 percent of water bodies are marine areas) amounted to 4.4 percent (Cleave, 1986). Additional Services 1. Fuelwood-Fertilizer Linkages Since trees in dry zones help to safeguard the soil against dessication, they serve to stave off desertification. Here, as in other parts of the tropics, they also supply a potentially renewable source of fuelwood. When they are not harvested renewably, however, as already mentioned, local people often turn to animal manure and crop residues for fuel, and thereby deplete their source of fertilizer. Well over two billion people, or the majority of developing- country citizens, rely on wood to cook their meals and to heat their houses. The amount of energy consumed for these domestic purposes is - 31 - almost twice that used by agriculture and industry combined. (In most developed countries, in contrast, cooking takes less than 5 percent of all energy used.) Close to 1.5 billion people in developing countries are having difficulty finding fuelvood, or its derivatives such as charcoal. Many are cutting fuelvood faster than it is supplied by regrowth. Worse, at least 112 million people cannot even lay hands on enough fuelvood to meet their minimum daily needs. By the year 2000, the number of people suffering some scarcity of fuelvood may well increase from 1.5 billion to 2.3 billion, and the number facing acute shortages may well grow from 112 million to over 350 million; by the year 2025, these totals may increase to 3.5 billion and 1 billion, respectively. If people burn manure instead, the loss in fertilizer can also create serious problems, depending on the amount of manure applied to each unit of cropland; its contribution as humus; its nitrogen, phosphorus, and potassium content and other major plant nutrients; and above all the yield- response ratios of crops. All of these parameters vary from one eco-zone to another. They also depend on cultural practices, such as the stage of the growing season in which manure is applied to croplands: if it is applied too early, much of the fertilizer effect will be lost through volatalization; if applied too late, it will be lost in storage. Some other factors to consider include the combustion efficiency of manure and its calorific value. Similar considerations apply to crop residues diverted for fuel: the woody stems of chic peas, for example, can provide a useful source of fuel without necessarily detracting from soil fertility. In short, there need not be an inevitable direct tradeoff between the two uses of manure and crop residues. The significance of fertilizer loss in a developing country can be illustrated by figures drawn from Ethiopia. In 1981-82 the country had just under 25 million cattle, and a human population of roughly 32.5 million, which turned out to be far too low an estimate (Newcombe, 1984). Dung production could be reckoned at roughly 23 million tons dry weight (wet weight 152 million tons). In certain parts of the country, 60-90 percent of the dung was used as fuel. If we calculate that just under 8 million tons overall were used as fuel, this translates, with an average grain-response value of US$76 per ton, into agricultural output foregone worth some US$600 million -- or roughly 30 percent of the country's value added in its entire agricultural sector (Pearce, 1986). In Tanzania, the value of fuelwood production nationwide in 1983 imputed from market prices, was TSh 207 million (Peskin, 1984). But had an amount been included to reflect the cost of time spent in seeking fuelwood (remember that long distances may have to be covered because fuelwood stocks are declining), the putative value would have risen to TSh.2746. This figure, plus an estimate for forest depreciation through over- harvesting, would have meant a reduction in Net National Product of 11 percent. Needless to say, these are rough estimates, but "creative calculation" of this kind is urgently needed if we are to assess the linkages between fuelwood shortages and agricultural shortfalls -- and, more generally, the linkages between the three categories of resources discussed in this paper. - 32 - 2. Tree Broweas Livestock Fodder The linkages between forests and agriculture can be further illustrated by the use of tree browse as livestock fodder in Nepal (Chapman and Vithayathil, 1981). In parts of the country, about one-quarter of livestock fodder comes from forests around villages, much of it in the form of leafy branches cut from live trees. Fodder can use up three to five times more forest than is required for fuelwood and timber. In turn, the resulting deforestation means declining fodder stocks and less healthy livestock. Fodder-starved animals pose a serious threat here because Nepal, with the lowest level of tractor use in Asia, depeads on its draught animals for farm work. But the ripple effects that stem from the degradation of one resource such as the tree cover do not stop there. When the output of forest products is reduced by, say, 20 percent, livestock production may drop as much as 7-20 percent; wheat production (because of the lack of animal draught power), 4-10 percent; jute, well over 5 percent; rice, almost 5 percent; and oilseed, 2-4.5 percent. The decline in agricultural output overall can account for 36-42 percent of the decline in a local economy's output. When the calculation includes other activities that are affected by deforestation (such as furniture making), output falls away by another 4-9 percent. Although calculations of this type must obviously be treated w4th caution, they do suggest the extent to which diverse linkages operate between physiobiotic and socioeconomic factors. Summary and Conclusions The-degradation of tropical forests has spillover effects that go far beyond the forests themselves. To deal with this degradation, it is not enough to attend to the traditional concerns of tropical forestry, namely, the trees within the forests. Attention must also be given to the multiple Xinkages with other factors, particularly soil and water. Increasingly, the future of forests will depend on the activities of agriculture, often in areas far removed from the forests: unless agriculture can be made more productive and intensive in existing farmlands, growing throngs of landless peasants will continue to stream into the forests to practice slash-and-burn cultivation. At the same time, the future of agriculture is tied in with the fortunes of forests: we have noted, for instance, the linkages to flood-and-drought regimes. Development efforts should therefore include a systematized consideration of all features of the developing country's landscape -- ecological, economic, and institutional. V. GAPS IN KNOWLEDGE AND RESEARCH PRIORITIES While many policymakers recognize the urgency of introducing environmental considerations into their analyses, they often encounter a marked lack of basic information on certain key factors. Indeed, we are appallingly ignorant about many aspects of the natural-resource base that underpins important sectors of the economics of LDCs. For example, we have only a very rough idea of how much soil is being lost each year, though we - 33 - believe that in many areas it is almost certainly an unacceptably high amount. Equally to the point, we have no more than a preliminary idea of the extent to which soil loss affects crop productivity. Thus policymakers have to work with basic data that are often beset with significant deficiencies of information and understanding. Of course this will always be the case: we shall never have anywhere near-complete information, let alone understanding, as concerns a number of salient issues. The challenge is to supply policymakers with enough information that they can work with an appropriate degree of "precise imprecision." In this concluding chapter, then, we look at some thematic areas that feature significant gaps in our knowledge and that warrant priority research. 1. Assessment of Resource Stocks We are basically ignorant of the rate at which certain critical resource stocks are being depleted. Consider, for example, the case of soil erosion in Nepal; As we have seen in Section III, the estimates of this key factor -- the most widely cited figure is 240 million tons per year -- are often highly speculative. If we could achieve at least a first approximation of the true scale of the problem, we would attain a level of awareness aid understanding that would be far better than our present estimatesathat are not only crude in the extreme but quite likely inaccurate in a manner that could grossly distort public policy. First we need to find out the area of agricultural landailin Nepal -- and how it is expanding; whether human cultivation pressures aie growing, whether mechanized tillage is increasing, and what otherrf subsidiary factors may affect erosion patterns and trends. Next i need to compile sample measurements of soil transport at watershed outlet-points; we could use these to determine some "reasonable" lower and upper.bounds for average annual loses per hectare (insofar as an average figure is a worthwhile working figure). The figures would no doubt be way above 10 tons per ha per year (according to professionals with extensive on-the- ground experience), while surely less than 150 tons (there would soon be little topsoil left at all). We would then incorporate various other considerations, such as synergistic impacts (e.g., from expanding cultivation on ever steeper hillsides) that may be aggravating the situation; plus factors such as land-management and soil-conservation practices (terracing and the like) that may be relieving the situation. Of particular importance would be the amount of natural erosion that is occurring as opposed to man-derived erosion: are the two merely additive or mutually amplificatory? And so on, with basic appraisal of whatever other key factors are relevant and can be factored in without too much effort (insofar as this would be a preliminary and exploratory assessment -- by no means a fine-grain analysis). Having assembled as much data and analysis as possible, we would then conduct the appropriate calculations for an overall assessment. - 34 - Following on these estimates, we need to appraise the related factor of erosion's impact on development sectors such as agriculture and hydropower energy. How much crop productivity is being lost through erosion, and how much can it be compensated for by countermeasures such as increased inputs of fertilizers? (Note that in the case of Bangladesh, for instance, soil erosion, while severe and extensive, is a matter of no real concern, since the country possesses fertile soils that in many areas amount to a many-meters deep cover which will not be anywhere near exhausted through erosion as long as the soil stock is constantly replenished through washoff from upstream catchments.) How far too is erosion causing siltation of irrigation channels? Is it imposing sedimentation of hydropower facilities, also of coastal-fishery ecosystems, to a degree that significantly depletes output? Through these analyses and calculations we would thus come up with a soil-erosion figure that serves an indicative purpose, subject to a host of qualificatory statements that would invariably have to be attached to this final figure. The end product would be a better informed, albeit thoroughly incomplete, estimate -- no more, and no less. True, it would be of little use for exhaustive policymaking, especially at the local level. But it would be much better than the very limited estimates (hardly better than virtually no estimates) that we often have to work with to date. Of course, many agency officials might balk at such an approach, no matter how many qualifications might be attached to the estimates. The approach would ostensibly depart too far from established norms of scientific rigor. But there is also a lack of scientific rigor in declaring that just because we do not have a (normally) sufficient quantity and quality of evidence about some phenomenon, we can/should assert nothing about it. Even in academia, a response along these lines would not always be in strict accord with scientific method in the proper sense of the term. In the development field, it is a still less appropriate response: by default, other people such as peasant communities will employ their own understanding of the situation, derived from their own "data base", and thereby formulate their own findings, draw their own conclusions, make their own decisions, and undertake their own planning in response. So the situation cannot be deferred until an "adequate" body of evidence becomes available. To decide to "wait until ...# is not to wait at all. It is to hand over the decision-making opportunity to other persons and bodies, even though they may be far less competent to grapple with the issues. In other words, to take no action can be to act quite decisively. In this sense, a first estimate of soil loss in Nepal would help us to judge whether there is a "significant" problem of soil erosion in the country. And if we believe there is such a problem -- whether it is marginal, moderate, serious, severe, critical, or catastrophic -- such an estimate might tell us whether it merits no more than further inquiry and research (if that), some prophylactic action, a substantive and urgent program of countermeasures, or immediate and substantive initiatives to both cure the problem and prevent it from recurring. This key question of soil erosion applies to many other countries that feature steeply sloping lands and highly erodible soils. In, for I - 35 - instance, northern India and Pakistan with their Himalayan foothills, the Ethiopian Highlands, the Colombian plateau, much of the Philippines uplands, and montane sectors of Mexico and Central America, we have only rudimentary data on how fast soil cover is being depleted -- and how far this is significant for sectors such as agriculture and hydropower generation, and also for downstream flooding. So Research Priority No. 1 is to engage in such measurements and evaluations as enable us to come-up_-ith substantie_ firstMapproximation estimates of-soilerosion in-unusually erodible__areas: together with an Anraisal of how far it significantly reduces crop productivity. hydropower output. etc. In like manner we need to document and analyze depletion rates for other critical resource stocks such as forests. While we have some fairly good data on the extent and speed of tropical deforestation, they mostly derive from a baseline of the late 1970s -- and we shall not have updated data until FAO releases its new appraisal of Tropical Forest Resources in a few years' time. Meanwhile, it is becoming plain that deforestation rates have recently accelerated in several parts of the biome: in the Philippines (Revilla et al., 1987), Sarawak (Repetto, 1988), Sabah (Phillips, 1987), Thailand (Phantumvanit, 1987), India (Agarwal and Narain, 1985), Madagas,tar (Sussman et al, 1988), Ivory Coast (Gillis, 1988), Colombia (Andrade and Ruis, 1988) and Peru (Dourojeanni, 1988). The most remarkable instance is Brazil, where a recent remote-sensing assessment (Setzer and Pereira, 1988) shows that in 1987 more than 80,0002 kms of Amazonian forest were burned, by contrast with the FAO 1980 estimate of annual deforestation for the entire moist tropics, 76,0002 kas. These findings make it all the more pressing that we get a firmer picture on the ongoing rate of tropical deforestation. Even more to the point, we urgently need informed estimates of likely rates of deforestation into the foreseeable future. This is unduly difficult to determine in light of the many factors in question -- and it is unduly pertinent to establish the situation if policymakers are to make realistic prognosis of the scale and scope of deforestation with its linked impacts on agriculture, water flows and coastal fisheries. A key factor at issue here is the probable rate of growth in the number'of slash-and-burn cultivators, the people who account for the great bulk of deforestation and who appear to be responsible for much of the burning in Brazilian Amazonia. In many, if not most, countries of the biome, e.g., the Philippines (Cruz, 1988), Brazilian Amazonia (Malingreau and Tucker, 1988) and Ivory Coast (Gillis, 1988), the number of slash-and-burn cultivators appears to be increasing at rates between 3.5 and 10 percent per year, i.e., faster and sometimes several times faster than population growth rates at national level. This rapid increase is due to the phenomenon of the "shifted" cultivator, an agent far more destructive of forests than the shifting cultivator of traditional sort. The shifted cultivator is the man who finds himself landless in established farming areas of his country and who senses no survival option but to head for the forests with his machete and matchbox. Yet despite this man's central significance to the future of - 36 - tropical forests, we have only a very preliminary idea of the present numbers overall, let alone the likely numbers in, say, the years 2000 and 2020. Estimates of the present total range from below 200 million to more than 500 million people. There is hardly a greater deficiency in our data base for tropical forests than this one -- yet to our knowledge no agency has a project or even a plan to fill this gap in our knowledge. Hence, Research Priority No. 2: to gain a more immediate and comprehensive assessment of deforestation rates. both present and projected. to&ether with a snntic.apraisal of the main destructive agents at work. the "shifted" cultivators (and an analysis of reasons for the growing numbers -- reasons both proximate and ultimate). Of course tiie considerations set out here apply to other resource stocks besides soil and forests. We also need urgently to gain a more concise and comprehensive assessment of the present status and future outlook of: tropical coral-reefs (there are not even agreed figures on the extent of these reefs, let alone on their species communities and rates of depletion); and wetlands, such as the vast Sudd Swamp in Sudan. These ecological zones are not listed as Research Priorities in order to keep the priority list short. 2. Vegetation/Atmosphere Interactions, It is increasingly apparent that the decline of tropical forests is associated with some change of climatic regimes. Yet we know all too little about the vegetation/atmosphere interactions that lead to shifts in climatic patterns -- except that the interaotions can cause widespread and severe problems for development processes. If, for instance, declining vegetation leads to declining rainfall, the consequences will be felt in agriculture, hydropower energy, and public health, among other development sectors that are highly dependent on water supplies from rainfall. Even a seemingly negligible decline in rainfall, amounting to a few percent per year, can have a major effect on crop yields (Oram, 1986; Parry and Carter, 1985). i A key question, then, is whether human activities can now change climate. Climatplogists are far from being able to supply a definitive answer, but they do know something about the linkages involved. Obviously, not every shower is due to forest influences; otherwise it would never rain in grasslands. Conversely, one of the wettest places in the world, Cherrapuiji, no longer possesses much forest cover. In certain localities, and undeft:certain circumstances, forests do not induce rain, but that does not mean they cannot do so. Nor does evidence of higher rainfall within forest areas mean that forests increase rainfall. Generally speaking, however, it seems that tropical deforestation can cause sizeable changes in local midroclimates, and, if sufficiently extensive, these local changes can add up to region-wide changes in macroclimates. A few observers still assert there is no connection between forest cover and rainfall regimes, nor that there can be any connection. But others (Dickinson, 1982; Hansen and Takahashi, 1984; Hare, 1985; Henderson- Sellers and Gornitz, 1984; Hutchinson and Hicks, 1985; Lockwood, 1980; - 37 - Moore, 1985; Nicholson, 1985; Salati and Vose, 1984; Shukla and Mints, 1982), drawing on more recent data and analyses, believe there is, or could well be, a substantial connection, notably in the humid tropics. The matter is particularly relevant to the environmental dimensions of development, and it reflects some salient gaps in our knowledge. In essence, the forests/climate connection amounts to this. Forests exchange moisture and energy with the atmosphere more intensively than do any other types of land-surface cover. They do so by evapotranspiration, a process that results from the interaction of moisture in the soil; vegetation which transfers moisture from the soil to the atmosphere; and energy, which converts the moisture into water vapor. Most of the energy comes from radiational heating of the surface, and thus depends on surface albedo (or "shininess" of the land surface). In turn, the albedo depends on the vegetation, which in its turn depends on the soil moisture. The average riverwater drainage from all land surfaces each year is only about one-third of the average precipitation, which means that land-surface evapotranspiration must be a significant phenomenon (Shukla and Mintz, 1982). As vegetation is eliminated, there is less possibility for evapotranspiration, and less moisture is dispatched into the atmosphere for recycling as rain. In short, human modification of vegetation cover, if great enough and sufficiently widespread, bears significantly on climate, the precise response depending upon local conditions (Dickinson, 1982; Pinker et al., 1980; Rosenzweig and Dickinson, 1986; Steward, 1984). Amazonia provides a case in point. Water isotopes reveal that between half and four-fifths of the region's moisture remains within the ecosystem. That is, it is constantly transpired by plants into the atmosphere, where it gathers in the form of storm clouds before being precipitated back onto the forest. The forest thus represents a significant source of its own moisture (Lettau et al., 1979; Salati and Vose, 1984; Shuttleworth, 1988). Were Amazonia to be widely deforested, a great deal less moisture would be evapotranspired into the atmosphere -- and such a decrease would surely entrain some marked and irreversible ecological changes in many areas. Even more important, it could trigger a self-reinforcing dessication of the forest cover, whereby declines in moisture stocks would be followed by yet more dessication, and so forth. In the Panama Canal area, steady deforestation since the-start of the century has been accompanied by a steady decline in rainfall (Windsor et. al., 1986). In Guanacaste Province of Costa Rica, associateditrends have persisted for 40 years (Fleming, 1986; Jantzen, 1976). Similar evidence comes from a variety of areas in India (Biswas, 1980; Meher-Homji, 1980; Padmavalli, 1976). In northwestern Peninsular Malaysia, an area widely deforested especially in Penang and Kedah, not only has rainfall declined, but the distribution has become less predictable in both time and space during the past 75 years, particularly since 1960 (Chan, 1986; Goh, 1980). Over the past two decades, some 20,000 ha of paddy ricefields have been abandoned in this "rice bowl" sector of the Peninsula, and another 72,000 ha have registered a marked dropoff in production (the Peninsula's rice output declined by a critical 27.5 percent in 1978). Yet another example of a decline in forest cover being accompanied by a decline in rainfall comes from southern Yunnan, China (Zhang, 1986). - 38 - Of course it is not (yet) possible to postulate a direct causative link between changes in vegetation cover and shifts in rainfall patterns. All manner of other variables are probably at work. Very few problem- specific experiments have been carried out, because of the scale and complexity of the processes in question. Nonetheless, the severe repercussions that may arise from the putative linkages make it imperative to take at least two types of action (Myers, 1988d): 1) mission-oriented research of a scope and scale to match the issue; and (2) a risk-reduction strategy in policymaking that seeks to respond to both the uncertainty inherent in the situation, and to the sizable risk factor (risk is defined here as the adverse outcome of a circumstance, multiplied by the likelihood of the unfortunate outcome actually coming to pass). Even where the prospect of human-caused dislocation of climatic systems is considered slight, the repercussions could prove so disruptive, adverse, and widespread, that such a prospect should be considered worth avoiding unless the costs of doing so prove prohibitively high. In these or any other circumstances in which a salient factor of public policy lies with uncertainty, it would be better to find that we have been approximately right than precisely wrong. Another type of climatic change that is fast becoming far more likely, while remaining ratber more distant in the future than other climatic changes, is planetary warming. For the most part it is not caused by devegetation, although its impact on vegetation will be profound. This planetary warming stems from the "greenhouse effect" of the buildup of carbon dioxide and several other trace gases in the global atmosphere. According to some recent estimates, we can expect to see signs of the warming effect by the end of the century if not before, and we can expect to experience an overall planetary warming of between 2.5 degrees and 5.5 degrees CO by the second quarter of the next century at the latest (Bolin et al., 1986; Hansen et al., 1988; Mintzer, 1987; Schneider, 1987). At the equator, the warming is expected to be only slight, probably less than 1 degree CO, while at the poles it could well be as much as 7 degrees CO or even more. The climatic changes are likely to include some basic shifts in precipitation patterns, so that some areas will probably be wetter than today and other areas drier. The most pervasive general result for vegetation will be a shifting away from the equator of some divisions between major zones of natural vegetation (Emanuel et al., 1985; see also Shugart et al., 1986) -- but only if there remains any natural environments for them to shift into. Most lands suitable for expansion will have been occupied for human purposes, meaning there will be much less possibility for e.g., tropical forests, woodlands and savannahs to expand anywhere. The implications for agriculture, to cite but one development sector, may be far-reaching in a greenhouse-affected world. As the projections rapidly become more probable, steps must be taken while there is still time to anticipate which areas are likely to become less suitable for agriculture within the foreseeable future, which ones more suitable, and which dessicatory trends can be countered through more efficient irrigation systems, drought-resistant seed types, and so on. Yet many long-range policy and planning decisions concerning food production in - 39 - general, and irrigation works, hydropower facilities, and drought problems in particular, are being made today on the assumption that past climatic patterns will extend Into the future. For instance, there are plans to invest US$100 billion during the last 15 years of this century in an extension of current irrigation networks - and the plans are based on the supposition that there will be no large-scale climatic changes in the foreseeable future. Yet mounting evidence, albeit less than conclusive evidence to date, points to possible changes in rainfall regimes in the coming decades. In addition, there is a more localized type of vegetation/ atmosphere interaction that, according to some recent research of very preliminary and exploratory nature (Rodhe and Herrera, 1988), now appears to be significant for a number of countries. It is acid precipitation, already well known in North America and Europe, and now becoming manifest in parts of southern China, with portents of similar problems for forest cover and agricultural crops among other forms of vegetation in sectors of southeast Asia (Java, Sumatra and southern Thailand), southwestern India, parts of West Africa, Central America, northern Venezuela and southern Brazil. Hence Research Priority No. 3: to promote a far greater degree of knowlede and-Understandina of veetation/atmosphere interactions. especially as concerns climate repercussions and pollution imacts within the foreseeable future: and to evaluate the likely conseauences for development sectors such as agriculture. energy and settlements. 3. Ecological Discontinuities We know all too little about the multiple linkages and interactive dynamics that characterize ecosystems -- except we recognize they can be critically important to the status of natural-resource stocks, hence to the contribution of these stocks to development. True, we now accept that ecosystems are more resilient and adaptable than sometimes assumed. We also know that, if they are not subject to unusual external disturbance, they exhibit distinct trends of development, productivity, stabilization and other autogenic phenomena. But when disturbance is of sufficient scope and scale, ecosystems often start to manifest some unanticipated features, which in the jargon of the trade, are referred to as mutliple equilibria, proliferant domains of stability, threshold responses, breakpoints of irreversible injury, and many other discontinuities in their ecological processes (Barrett and Rosenberg, 1981; Brooker, 1987; Ehrlich, 1986; Mooney and Godron, 1983; Odum, 1985). These inflection points, also known as "jump events", occur when conditions suddenly and rapidly become worse. Our growing experience with tropical forests and watershed regimes reveals that certain large-scale ecosystems can absorb stresses over along periods without much outward sign of damage, until they reach a disruption level at which a jump event becomes increasingly likely and ultimately inevitable. (In non-tropical lands we have seen this happen with acid rain and ozone-layer depletion. In both cases, the stresses imposed by these phenomena have built up covertly over extended periods before suddenly revealing themselves in - 40 - critical proportions.) Moreover, when ecological stresses are removed, often at unexpectedly high cost, the system may not return to its former equilibrium state. Instead it may settle into a new equilibrium, one that may be less useful to human purposes and that we may not know how to amend. The point is that ecosystems may work in nonlinear fashion, and across a broad expanse of space and time. The long-term and cumulative consequences of stress will be impaired functioning of the entire ecosystem, and destabilization of its structures. Stress responses will appear in che form of a decline in the "natural" efficiency of resource use (nutrient losses expand as ecosystems become "leaky"), changes in energetics, an increase in parasitism, and a decrease in mutualism and other positive interactions. As human communities and their technologies continue to expand, they will exert increasing pressures on natural resource stocks and ecosystems, and ecological discontinuities will become more common. Hence Research Prioricy No. 4: to identify and define those forms of ecological discontinuities that are most likely to emerge as a result of human activities and that are most capable of imposing significant costs on development rocesses: and then to develop sets of criteria to determine acceptable and non-acceptable levels of human disturbance. plus guidelines to monitor the effects of disturbance -- especially in those ecosystems that are unusuall_ susceptible to human disQrtion. such as tropical forests. coral reefs. wetlands. estuaries. and montane environments. 4. Asymmetry of Evaluation The four items listed above relate essentially to natural-sciences research.rq,Let us now consider, as a final item, a factor that relates to social-sciences research. It can generically be termed "asymmetry of evaluation". When a country is uncertain about the scale of revenues generated by a stock of resources, it tends to play down their importance, even though they may actually be far from insignificant. An institutional bias comes into play, which leads to asymmetric evaluations and hence serves to engender some distorted policy responses. Consider, for example, minor forest products in Indonesia. Despite their sizeable value, these products are considered to be minor resources in both relative and absolute senses. In 1982 they were worth about US$200 million in export earnings (Gillis, 1986). Yet the 1985 Draft Long-Term Forestry Plan mentions them only four times in 88 pages, and then only in passing. This writer has found during the course of four visits to Indonesia (Myers, 1988f) that precisely because there is no concise and comprehensive tabulation of the volume and value of these products (at least as compared with those of timber and products), they are treated with indifference at best, even though a more accurate assessment of the total volume and value of all such products would surely be far larger than is documented. If improvements in the harvesting of these products were to receive the same share of the budget as wood products did in 1982, they would receive about US$75 million per year instead of the present US$4 million per year (Gillis, 1986). Such i the asymmetrical upshot of perceived uncertainty. - 41 - Of course the asymmetry is more pronounced when the uncertainty is real. The trouble is that many natural resources and environmental values are incommensurable, if not intangible entities. They are not priced according to a market because none exists for them. An individual entrepreneur cannot trade in watershed systems, gene pools, or forest services because these are generally considered common property. Consequently, they tend to be misused and overused by individuals who enjoy unrestricted access to them. Since they bear no price tag to indicate society's estimation of their worth, there is no convenient way to determine their value, or whether they have any value at all, or whether their status is stable, declining, increasing, or whatever. Of course this simplifies the situation, but essentially it does not oversimplify it. Instead we tend to emphasize the entities that can be counted, to the detriment of those that also count. The skewedness of reckoning can occasionally become extreme, as in the case of tropical forests. There is a broad institutional apparatus for appraising the commercial value of hardwood timber, as a principal material output of tropical forests: for instance, there are worldwide markets to supply an up-to-date, continuous, quantified indication of what the timber is worth, both to local communities in tropical forest countries and to consumers thousands of kilometers away. But there is next to no way to pick up market signals of what tropical forest ecosystem services are worth while the-forest.remains in being, even though the costs that arise after removing the forest (as a result of soil erosion, the disruption of water flows, etc.) can prove to be high indeed -- often a good deal higher than the revenues gained through harvesting (or rather nonharvesting) of commercial timber from the forest, as illustrated by the flooding problems in valleylands downstream from deforested watersheds in the Indian subcontinent. Because these. environmental costs are not signalled ahead of time, they are completely ignored and few measures are taken to safeguard the forest cover L even though experience shows that the costs of action may well turn out to be only a fraction of the costs of inaction. The asymmetry of evaluation creates even greater distortkons in the policymaking context. In principle, we generally view soil cover as being more valuable than we treat it in practice (this applies to farmers in Indiana and England, as much as to farmers in India and Ethiopia). Yet, because of institutional bias, we act as though we accord an automatically higher value to commensurables (farm produce, even if grown at cost to soil cover) than to incommensurables (soil cover itself). The asymmetry of evaluation comes down heavily in favor of benefits as identified through existing institutional structures, notably the marketplace, but also property rights and legal systems. Were social institutions geared the other way round, the calculus would work out far differently. Suppose, by way of illustration, the institutional framework were to allow "soil supporters" to purchase the goods they seek through competitive markets. Suppose also that, conversely, there was no opportunity to buy goods produced by agriculture. The disbenefits of agriculture would not only be priced in the marketplace, but they would be reckoned as part of the "solid" price-quantity data. They would thereby become part of the parcel of resource costs, insofar as people havs. to be - 42 - paid to be deprived of things they value. On the other hand, the benefits of agriculture would presumably be evaluated in line with present methods of measuring non-market goods. Hence, the problem of uncertainty would apply to agricultural products, with all the difficulties of policy that implies; while an aura of certainty would extend to soil cover and its values, with all the support for policymaking processes that would imply. In other words, prevailing institutional structures often mean that collective goods are more prone to externalities -- whether economic or environmental -- than are private goods. As a result, the most public- spirited people find it difficult to protect "the public's property." Even if "soil supporters" had the wherewithall to counteract the disbenefits of soil-eroding agriculture, the feedback loops of such an arrangement would not necessarily close automatictlly and speedily enough unless the incentives offered were overwhelmingly persuasive. The essential issue here boils down to the question of whether institutions can be made sufficiently &daptive. Hence Research Priority 6: to develop a Dolicvmaking methodology that systematically takes account of the asymmetry of evaluation, This will not be straightforward: it is a Drospect that has long perDlexed environmentalists and natural resource managers. The essential issue is to adapt institutional mechanisms so that they reflect a more realistic and accurate evaluati2n of the *true yorth" of those natural resources that. by virtue of their conon property status. tend to be undervalued axd overoxvloited * 43 - Abernethy, C.L., Ambuklao Reservoir. Phillinines, Hydraultcs Rpoearch Institute, Wallingford, U.K., 1984. Agarval, A. and S. Narain, editors, The State of India's Environment. 1984-85: The Second Citizen's ReRort, Centre for Science and Environment, New Delhi, India, 1985. Aina, P.O., R. Lal and G.S. Taylor, Soil and Crop Management in Relation to Soil Erosion in the Rainforest of Western Nigetia, in J'1U Erosion: Prediction and Control: 75-82, Soil Conservation Society of America, Ankeny, Iowa, 1977. Ajayi, S.S., Utilization of Forest Wildlife in West Africa, Food and Agriculture Organization, Rome, Italy, 1979. Alvarado, L.A., Sedimentation in Madden Reservoir of the Panama Canal Waesbe, Panama Canal Commission, Balboa, Panama, 1985, Allison, F.E., Soil Organic Matter and Its Role in Crop Production, Elsevior, New York, 1973. Ambroggi, W., Water, Scientific American 243: 101-116, 1980. American Society of Agronomy, Soil Erosion and Conservation in the Tropics, ASA S2ecial Publication No. 43, American Society of Agronomy, Madison, Wisconsin, 1982. Anderson, D., Nigeria: Arid-Zone Afforestation and Industrial Plantations Project, The World Bank, Washington, D.C., 1986. Andrade, G.I. and J.P. Ruiz, Amagonia Colombian&. Fescol Publishers, Bogota, Colombia, 1988. Ba, L.K., BLo-Economics of Trees in Native Malayan Forests, Department of Botany, University of Malaya, Kuala Lumpur, Malaysia, 1977. Bahugna, S.L., Himalavan Trauma. Forests. Faults. Floods, Gandhi Peace Foundation, New Delhi, India, 1978. Balchin, W.G.V., Water - A World Problem, International Journal of Environmental Studies 25: 141-148, 1985. Balick, N.J., Amazonian Oil Palms of Promise: A Survey, Economic Botany 33(1), 1: 11-28, 1979. Bernard, G. and L. Christoffersen. Agricultural Residues As Fuel in the ThirdWorld. Earthscan, London, U.K., 1985. Blackie, J.R., Hydrological Effects of a Change in Land Use from Rainforest to Tea Plantations in Kenya, Proceedingsa of WellinWton S=mposium 2: 312-329, 1972. -44 - Blaikie, P.L., Political Economy of Soil Erosion in Developing Countries, Longman, London, U.K., 1985. Bognetteau-Verlinder, Els, Study on Impact of Windbreaks in Majjia Valley, Niger, Department of Sylviculture, CARE/Agricultural University, Wageningen, Holland, 1980. Bolin, B., B.R. Doos, J. Jagel, and R.A. Warrick, editors, The Greenhouse Effect: Climatic Change and Ecosvatems, John Wiley and Sons, New York, 1986. Borthakur, D.N., Problems of Soil Erosion and Floods in the Northeast Regon, Gandhi Peace Foundation, New Delhi, India, 1978. Brown, L.H. and E.C. Wolf, Soil Erosion: Quiet Crisis in the World Economy, Worldwatch Paper 62, Worldwatch Institute, Washington, D.C., 1984. Burgess, P.F., The Impact of Commercial Forestry on Hill Forest of the Malay Peninsula, in E.S. Soepadmo and K.G. Singh, editors, Proceedings of Symposium on Biological Resources and National Development: 131-136, Malayan Nature Society, Kuala Lumpur, Malaysia, 1973. Caldecott, J.0., Hunting and Wildlife Management in Sarawak. Final Report, World Wildlife Fund, Kuala Lumpur, Malaysia, 1986. Carpenter, R.A., ed. Natural Systems for Development: What Planners Need go K . MacMillan Press, New York, 1984. Challeau, P.G.D., River Flow in the Cameroon Highlands, Hydroelectric Technical Memo No. 3, Central Electricity Board, Kuala Lumpur, Malaysia, 1956. Chapman, J. and F. Vithayathil, Some Potential Economic Effects of Forest Product Scarcities in Nepal: An Input-Output Analysis of the Kosi Zone, International Journal of Environmental Studies 18: 23-32, 1981. Charreau, C., Problems poses par 1 Utilisation Agricole des Sols Tropicaux, Agronomie Tropical (France) 27: 905-929, 1972. Cleave, J., Philippines Central Visayas Regional Project, I: The National Fisheries Sector, Working Paper No. 3, The World Bank, Washington, D.C., 1986. Collard, D., D.W. Pearce and D. Ulph, editors, Economics and Sustainable Environments, MacMillan, London, 1987. Colombo, U., D.G. Johnson and T. Shishido, Reducing Malnutrition in Developing Countries: Increasing Rice Production in South and Southeast_Asia, Triangle Paper No. 16, The Trilateral Commission, New York, 1978. - 45 - Conklin, H.C., Ethnoraphic Atlas of Ifugao: A Study of Environment. Culture and Society in Northern Luzon, Yrle University Press, New Haven, Conn., 1980. Constalbe, M., Ethiggian Hiahlands Reclamation Study, Soil Conservation Research Project, United Nations University, Addis Ababa, Ethiopia, and Soil and Water Conservation Department, Ministry of Agriculture, Addis Ababa, Ethiopia, 1985. Cool, J.C., Factors Affecting Pressure on Mountain Resource Systems, Agriculture Development Council, Kathmandu, Nepal, 1984. Cornelius, L., Bamboo and Rattan: Restoring the "Most Neglected" Resources of Southeast Asia, International Development Research Centre, Ottawa, Canada, 1984. Crosson, P., Soil Erosion in Developing Countries: Amounts. Consequences and_Policiel, Resources for the Future, Washington, D.C., 1983. Cruz, .C.J., More-lople Than Trees: The PhiliRpine Case. Manuscript, Institute of Environmental Science and Management, University of the Philippines at Los Banos, Philippines, 1988. Daniel, J.G. and A. Kulasingam, Problems Arising from Large-Scale Forest Clearing for Agricultural Use, Malaysian Forester 37(3): 152-160, 1974. Daryadi, L., Forestry and Agriculture: A Key Connection in Java, Department of Forestry, Jakarta, Indonesia, 1981. Davey, F.B. and M. Graham, Bivalve Culture in Asia and the Pacific, International Development Research Center, Ottawa, Canada, 1982. Davison, G., Personal Communication, Telephone Discussion. Superintendent, Remote-Sensing Division, Royal Aircraft Establishment, Farnborough, U.K., 1986. Dewan, M.L. and S. Sharma, People's Participation as a Key to Hislaan Eco-System Develooment, Center for Policy Research, New Delhi, India, 1985. Dickinson, R.E., Effects of Tropical Deforestation on Climate. Studies in Third World Societies 14: 411-441, 1982. 1 Dixon, J.A., and N.M. Hufschmidt, ede. Economic Valuation Technigues for the Environment: A Case-Study Workbook. John's Hopkins University Press, Baltimore, Maryland, 1986. Dosso, H., J.L. Guillaumt and M. Hadley, The Tai Project: Land Use Problems in a Tropical Forest, Aabio 10: 120-125, 1981. Douglas, I., The Impact of Urbanization on Fluvial Geomorphology in the Humid Tropics, Geo-Sco-Tropica 1: 229-242, 1978. - 46 - Dourojeanni, N.J., SL El Arbol de la Ouina Hablara, Foundacion Peruana para la Conservacion do la Naturaleza, Lima, Peru, 1988. Dudal, R., An Evaluation of Conservation Needs, pp. 3-12 in R.P.C. Morgan, editor, Soil Conservation: Problems and Prosiects, John Wiley and Sons, Chichester, U.K., 1981. Duke, J.A., Contributions of Neotronical Forests to Cancer Research, Economic Botany Laboratory, U.S., Department of Agriculture, Beltsville, Maryland, 1982. Easter, K.W., J.A. Dixon and M.M. Hufschmidt, editors, Watershed Resources Management: An Integrated Framework with Studies from Asia and the Pacific, Westview Press, Boulder, Colorado, 1986. Ehrlich, P.R. and A.H. Ehrlich, Extinction: The Causes and Conseauences of the Disappearance of Sgeies, Random House, New York, 1981. E1-Swaify, S.A., E.W. Dangler and C.L. Armstrong, Soil Erosion by Water in the Tropics, Research Extension Series 024, Coll. Trop. Agric. and Human Resources, University of Hawaii, Honolulu, Hawaii, 1982. El-Swaify, S.A., W.C. Moldenhauer and A. Lo, editors, Soil Erosion and Conservation, Soil Conservation Society of America, Ankeny, Iowa, 1985. Elvell, H.A. An Assessment of Soil Erosion in Zisbabawe. Zimbabwe Science NEM 19(3/4): 27-31, 1985. Emanuel, W.R., H.H. Shugart and M.P. Stevenson, Climatic Change and the Broad-Scale Distribution of Terrestrial Ecosystem Complexes, limatic Change 7: 29-43, 1985. Falkenmark, M., New Ecological Approach to the Water Cycle: Ticket to the Future, bLoh 13: 152-160, 1984. Farnsworth, N.R. and D.D. Soejarto, Potential Consequences of Plant Extinction in the United States on the Current and Future Availability of Prescription Drugs, Econo Botany 39: 231-240, 1985. Fisher, A.C., Economic Analysis and the Extinction of Species, Department of Agriculture and Resource Economics, University of California at Berkeley, California, 1982. Fleming, T.H., Secular Changes in Costa Rican Rainfall: Correlation with Elevation, Journal of Troical Ecology 2 (in press) 1986. Follett, R.F. and B.A. Stewart, editors, Soil Erosion and Crop ProductIvity, Soil Conservation Society of America, Ankeny, Iowa, 1985. - 47 - Food and Agriculture Organization, Agriculture: Toward 2000, Food and Agriculture Organization of the United Nations, Rome, Italy, 1981. Food and Agriculture Organization, Protect and Produce: Soil Conservation for Development, Food and Agriculture Organization of the United Nations, Rome, Italy, 1984. Food and Agriculture Organization and University of East Anglia (Overseas Development Group), Erosion-Induced Loss in Soil Productivity, Food and Agriculture Organization, Rome Italy, and University of East Anglia, Norwich, U.K., 1985. Food and Agriculture Organization and United Nations Environment Programme, Tropical Forest Resources, Food and Agriculture Organization of the United Nations, Rome, Italy, and United Nations Environment Programme, Nairobi, Kenya, 1982. Fournier, F., MaD of Erosion Danger in Africa South of the Sahara: 1xplanatory Note, Commission for 1echnical Cooperation in Africa, European Economic Community, Brussels, Belgium, 1962. Gasser, W.R., Survey of Irrigation in Eight Asian Nations, Foreign Agricultural Economic Report No. 165, Economics and Statistics Service, U.S. Department of Agriculture, Washington, D.C., 1981. Gentry, A.H., A Botanical Inventory of the Eastern Slope of the Andes in Cntral Peru, (mid-project report to Agency for International Development), Missouri Botanical Garden, St. Louis, Missouri, 1984. Gentry, A.H. and J. Lopez-Parodi, Deforestation and Increased Flooding of the Upper Amazon, Sci2e 210, 1354-1356, 1980. Gillis, M., Non-Wood Forest Products in Indonesia, Department of Forestry, University of North Carolina, Chapel Hill, North Carolina, 1986. Gillis, M., Ivory Coast: Public Policies, Resource Management and the Tropical Forests. In R. Repetto and M. Gillis, eds., ulic Policv and the Misuse of Forest Resources, Cambridge University Press, Cambridge, U.K., (in press), 1988. Goh, K.C., Rainfall Trends and Deficiencies in Peninsular Malaysia, Jh Qiental Geographr 24: 25-40, 1980. Green, S. and K. Minkowski, The Lion-Tailed Monkey and Its South Indian Rainforest Habitat, in P. Rainie, editor, Primr Conservation: 299-337, Academic Press, New York, 1977. Greenland, D.J. and R. Lal, editors, Soil Conservation and Management in the Humid Trooics, John Wiley and Sons, New York, 1977. Gupta, T. and A. Guleria, Non-Wood Forest Products in India, IBH Publishing Company, New Delhi, India and Oxford, U.K., 1982. - 48 - Hadley, M# and J.P. Lanley, Tropical Forest Ecosystems: Identifying Differences, Seeing Similarities, Nature and Resources 19: 2-19, 1983. Haigh, M.J., Soil_Krosion and Soil Conservation Research inIndia: An &Mnotated Bibliogranhy, Discussion Papers in Geography, Oxford Rolytechnic, Oxford, U.K., 1982. Haigh, M., Deforestation and Disaster in Northern India, Land Use Policy, July 1984: 187-198, 1984. Hall, G.F., R.B. Daniels and J.F. Foss, Rate of Soil Formation and Renewal Rates in the USA, in Determinants of Soil lass Tolerance, ASA Publication No. 45, American Society of Agronomy, Madison, Wisconsin, 1982. Hamilton, L.S., editor, Forest and Watershed Development and Conservation in Asia and the Pacific, Westview Press, Boulder, Colorado, 1983. Hamilton, L.S. and P.M. King, Tronical Forested Watersheds: Hydrologic and Soils Response to Major Uses or Conversions, Westview Press, Boulder, Colorado, 1983. Hansen, J.E. and T. Takahashi, editors, Climate Processes and Climate Sensitivity, Geoaysical Monograoh No. 29, American Geophysical Union, Washington, D.C., 1984. Hare, F.K., Climate Variations. Drought and Desertification, World Meteorological Organization, Geneva, Switzerland, 1985. Hecht, S.B., Environment, Development and Politics: Capital Accumulation and the Livestock Sector in Eastern Amazonia, World Development 13: 663-684, 1985. Hemming, J., editor, Change in the Amazon Basin: The Frontier After a Decade of Colonization, University of Manchester Press, Manchester, U.K., 1984. Henderson-Sellers, A., The Effects of Land Clearance and Agricultural Practices on Climate, Studies in Third World Societies 14: 443-485, 1981. Henderson-Sellars, A., and V. Gornitz, Possible Climatic Impacts on Land Cover Transformations, with Particular Emphasis on Tropical Deforestation, Climatic Change 6: 231-258, 1984. Higgins, G.M., A.H. Kassam and L. Naiken, Potential Population Suporting Capacities of Lands in the Develooin; World, Food and Agriculture Organization, Rome, Italy, 1982. High Level Commission on Floods, Report on the Emergent Crisis, Ministry of Irrigation, Government of India, New Delhi, India, 1983. - 49 - Hudson, N., Soil Conservation, B.T. Batsford Ltd., London, U.K., 1971. Hudson, N., Soil Conservation (2nd Edition), Cornell University Prds, Ithaca, N.Y., 1981. Hurni, H., Erosion. Productivity. Conservation Systems in Ethiopal Soil Conservation Research Project, University of Berne, Berne, Switzerland. Hurni, and Nuntapong, Agroforestry Improvements for Shifting Cultivation Systems, Mountain Research and Development 3: 338-345, 1983. Hutchinson, B.A. and B.B. Hicks, editors, The Forest-Atmosphere InteractiB, D. Reidel Publishing Co., Boston, Mass., 1985. Iltis, H.H., J.F. Doebley, R.M. Guzman and B. Pazy, Zea diphlonerennis (Gramineae), a New Teosinte from Mexico, Scine 203: 186-188, 1979. Indian Council for Agricultural Research, The Ganges River System: Hydrologic Disruptions. Flooding Threats. and Irrigated - Agriculture, Indian Council for Agricultural Research, New Delhi, India, 1984. Indus Basin Research Assessment Group, Research Issues Affecting Indus Basin Agricultural Develoment PolicX: Report, Ministry of Agriculture, Islamabad, Pakistan, 1978. IUCN, World Conservation Strategy, International Union for Conservation of Nature and Natural Resources, Gland, Switzerland, 1980. Jackson, I.J., Problems of Throughfall and Interception Assessment and a Tropical Forest, Journal of Hydrology 12: 179-189, 1971. Jacobs, M., The Study of Minor Forest Products, Flora Malesiana Bulletin 35: 3678-3782, 1982. Janzen, D.H., The Microclimate Difference Between a Deciduous Forest and Adjacent Riparian Forest in Guanacaste Province, Costa Rica, Brenesia 8: 29-33, 1976. Jayal, N.D., Destruction of Water Resources - The Most Critical Ecological Crisis of East Asia, paper presented at 16th Technical Meeting of IUCN, Madrid, Spain, 5-14, November 1984, Planning Commission of India, New Delhi, India, 1984. Jordan, C.F. and J. Heuveldop, The Water Budget of an Akmazonian Rain Forest, Acta Amazonca 11: 87-92, 1981. Joseph, K.T., Effect of Rural Development on Soils and Hydrology of Tropical Rain Forest Areas, The Environmentalist (Suppl. 10) 5: 36-39, 1985. - 50 - Kirkby, N.J. and R.P.C. Morgan, editors, Soil Erosion, John Wiley and Sons, New York, 1980. Kishk, N.A., Desert Encroachment in Egypt's Nile Valley, in S.A. El-Swaify, W.C. Moldenhauer and A. Lo, Soil Erosion and Conservation, Soil Conservation Society of America, Ankeny, Iowa, 1985. Klemas, V. and M. Hardisky, The Use of Remote Sensing in Global Biosystem Studies, Advanced Soace Research 3: 115-122, 1983. Kovda, V.A., Loss of Productive Land Due to Salinization, Ahbi 12: 91-93, 1983. Lal, R., Soil Erosion on Alfisols in Western Nigeria: I, Effects of Slope, Crop Rotation and Residue Management, Gooderm 16: 363-375, 1976. Lamb, R. and S. Milas, Soil Erosion: Real Cause of the Ethiopian Famine, Environmentl Consideration: 10: 157-159, 1983. Langdale, G.W. and O.D. Schrader, Soil Erosion Effects on Soil Productivity of Cultivate Cropland, in ASA, pp. 41-51, 1982. Larson, W.E., F.J. Pierce and R.N. Dowdy, The Threat of Soil Erosion to Long-Term Crop Production, Science 219: 458-465, 1983. Lasschuit, J.A. and F.A.C.M. van Eerd, Minor Forest Products and Non-Timber Products of the Forest, University of Wageningen, Netherlands, 1983. Lee, R., Forest Microclimatology, Columbia University Press, New York, 1978. Lee, L.K., Land Use and Soil Loss: A 1982 Update, Journal of Soil and Water Conservation 39: 226-228, 1984. Leigh, C.H., Land Development and Soil Erosion in West Malaysia, AUR 5: 213-217, 1973. Leigh, C.H. and K.S. Low, An Appraisal of the Flood Situation in West Malaysia, in E.S. Soepadmo and K.G. Singh, editors, Proceedings of Symoosium on Bilogical Resources and National Develooment: 57-72, Malayan Nature Society, Kuala Lumpur, Malaysia, 1973. Leonard, J., Natural Resources and Economic Development in Central America: A Regional Environmental Profile, TransAction Press, Rutgers, New Jersey, 1986. Leopoldo, P.R., W. Franken, and E. Matsui, Hydrological Aspects of the Tropical Rain Forest in Central Amazon, Interciencia 9: 125-131, 1984. Lettau, H., K. Lettau and L.C.B. Molison, Amazonia's Hydrologic Cycle and the Role of Atmospheric Recycling in Assessing Deforestation Effects, Monthly Weather Review 107(3): 227-238, 1979. - 51 - Lockwood, J.G., Some Problems of Humid Equatorial Climates, Malavaian Journal of Troaical Geography 1: 12-20, 1980. Low, K.S. and K.C. Goh, The Water Balance of Five Catcbments in Selangor, West Malaysia, Journal of Troical Geograh 35: 60-66, 1972. Low, K.S. and C.H. Peh, Conversion of Forest to Tree Crops: General Malavsian Experience, cited in Nik, 1985 (see below). Lundqvist, J., U. Lohm and M. Falkenmark, editors, Strategies for River basin Manaaement, D. Reidel, Dardrecht, Netherlands, and Boston, Mass., 1985. Lyles, L., Possible Effects of Wind Erosion on Soil Productivity, Jgrna of Soil and Water Conservation 30: 279-283, 1975. Magrath, W.B. and M.E. Grosh, A Dynamic Aproach to Watershed Management with Emnhasis on Soil Brosion, Working Paper No. 36, Department of Agricultural Economics, Cornell University, Ithaca, N.Y., 1985. Malik, B.A., Pakistan's Limited Water Resources and growing Demand. International Commission on Irrigation and Drainage, New Delhi, India, 1986. Malingreau, J.P. and C.J. Tucker, Large-Scale Deforestation in the Southern Amazon Basin of Brazil, Mhi 17: 49-55, 1988. Martin, F.W. and R.M. Ruberte, Edible Leaves of the Tropics, Institute of Tropical Agriculture, Mayaguez, Puerto Rico, USA, 1979. McIntyre, J. at a1, Crop-Livestock Interactiors in Africa. The World Bank, Washington, D.C., 1988. McPherson, M.P., Checking Land Destruction: lumediate Gains and Future Benefits, World Resources Institute Journal.1984: 50-57, 1984. Meher-Homji, V.M., The Link Between Rainfall and Forest Clearance: Case Studies from Western Karnataka, Transactions of Institute-of Indian Geographers 2: 59-65, 1980. Melillo, J.M., C.A. Palm, R.A. Houghton, G.M. Woodwell and N. Myers, A Comparison of Recent Estimates of Disturbance in Tropical Forests, Environmental Conservation 12: 37-40, 1985. Milford, R., Personal Communication, (based on 15 years of field experience in agricultural sectors in Mexico), World Bank, Washington, D.C., 1986. Mintzer, I.M., A Matter of Degrees: The Potential for Controllin the Greenhouse Effect. World Resources Institute, Washington, D.C., 1987. - 52 - Mok, S.T., Imoact of Forest Resources Development in the Environment of kWaAA, Forestry Department of Peninsular Malaysia, Kuala Lumpur, Malaysia, 1982. Molofaky, J., C.A.S. Hall and N. Myers, A Coparison of Tropical Forest Su=s, U.S. Department of Energy, Washington, D.C. 1986. Mooney, H.A. and M. Godron, editors, Disturbance and Ecosystems, Springer Verlag, New York, 1983. Moore, P.D., Forests, Man and Water, International Journal of Environmental Studies 25: 159-166, 1985. Morgan, R.P.C., editor, Soil Erosion and Its Control, Van Nostrand Reinhold, New York, 1986. Morgan, R.?.C., T. Hatch and W.S.W. Harun, A Simple Procedure for Assessing Soil Erosion Risk: A Case Study for Malaysia. Neitschrift fur Geomornhologie. Suplement Band 44: 69-88, 1982. Myers, N. CogyMrsion of Tropical Moist Forests (report to the National Academy of Sciences), National Research Council, Washington, D.C., 1980. Myers, N., A Wealth of Wild Species, Westview Press, Boulder, Colorado, 1983a. Hyers, N., Tropical Moist Forests: Over-Exploited and Under-Utilized?, Forest Ecology and Management 6: 59-79, 1983b. Myers, N., The Primary Source, W.W. Norton, New York and London, U.K., 1984. Myers, N., Tropical Deforestation and Species Extinctions: The Latest News, Futures 17: 451-463, 1985. Myers, N., Mass Extinction of Species: A Great Creative Challenge, Albright Lecture in Conservation, University of California at Berkeley, California, 1987. Myers, N., Report to the Government of Indonesia on Sustainable Development of'Certain of Indonesia's Natural Resources, Ministry of Environment and Population, Jakarta, Indonesia. Narayana, D.V.V. and R. Babu, Estimate of Soil Erosion in India, Journal of Irrigation and Drainage Engineering 109: 419-434, 1983. Narayana, D.V.V. and G. Sastry, Soil Conservation in India, in S.A. El- .,Swaify, W.C. Moldenhauer and A. Lo, Soil Erosion and Conservation, Soil Conservation Society of America, Ankeny, Iowa, 1985. NASA, LaRd-Related Global Habitability Science Issues, National Aeronautics ;and Space Administration, Washington, D.C., 1983. - 53 - National Academy of Sciences, 1975, Underexploited Tropical Plants With Promising Economic Value, National Academy of Sciences, Washington, D.C., 1975. National Academy of Sciences, 1979, Tropical Lemmes: Resources for the Future, National Academy of Sciences, Washington, D.C., 1979. National Commission on Floods, Government of India, A Report: Emerent Prohblems With Spgcial Respect to Irrigation, National Commission on Floods, New Delhi, India, 1980. Nations, J. and J. Leonard, Natural Resources in Central America, in A. Maguire and J.W. Brown, editors, Bordering on Trouble: Resources and Politics in Latin America, Adler and Adler, New York, 1986. Nault, L.R. and W.R. Findley, Primitive Relative Offers New Traits for Corn Improvement, Ohio Reort 66(6): 90-92, 1981. Newcombe, K., An Economic Justification for Rural Afforestation: The Case of Ethiopia, Energy Department Paper No. 16, The World Bank, Washington, D.C., 1984. Nicholson, S.E., The Drought Situation in Africa: Report to the U.S. State DeRartment, Department of Meteorology, Florida State University, Tallahassee, Florida, 1985. Nik, A.R., Watershed Management in Malaysia: A Perspective, Wallaceana W. 42: 3-8, 1985. Nor, S.M. and H.T. Tang, Forest Conservation in Malaysia: A Reapnraisal, Forest Research Institute, Kepong, Kuala Lumpur, Malaysia, 1980. Odum, E.P., Trends Expected in Stressed Ecosystems, BioSciene 35: 419-422, 1985. Oldfield, M.L., The Value of the Conservation of Genetic Resources, National Parks Service, Washington, D.C., 1981. Oldfield, M.L., The Value of Conserving Genetic Resources, National Parks Service, Washington, D.C., 1984. Olson, G., Improving Uses of Soils in Latin America, Geoderma 9: 257-267, 1973. Oram, P.A., Sensitivity of Agricultural Production to Climatic Change, Climatic Change 7: 129-152, 1985. Padmavalli, R., Utilization of Major Natural Resources, Proceedingl Symoosium on Resource Development and Planning: 1-19, -Department of Geography, Madras University, Madras, India, 1976. Parrington, J.R., W.H. Zoller and N.K. Aras, Asian Dust: Seasondl Transport to the Hawaiian Islands, Science 220: 195-197, 1983. - 54 - Paucar, A. and A.L. Gardiner, Establishment of a Scientific Research Station in the Yasuni National Park of the ReOublic of Ecuador, National Forestry Program, Ministry of Agriculture, Quito, Ecuador, 1981. Pearce, D.W., The Major Conseauences of Land and Water Mismanagement in Developing Countries, World Bank, Washington, D.C., 1985. Pearce, D.W., The Economics of Natural Resource Management: Issues Paper, The World Bank, Washington, D.C., 1986. Pearce, D.W., The Sustainable Use of Natural Resources in Developing Countries, in R. Kerry-Turner, editor, Environmental Economics Issues: Sustainability. Resource Conservation and Pollution Control (forthcoming) 1987. Peh, C.H., The Suspended and Dissolved Sediment Load of Three Small Forested Drainage Basins in Peninsular Malaysia, Malavsian Forester 44: 438-452, 1981. Pereira, H.C., Land Use and Water Resources in Temperate and Tropical Climates, Cambridge University Press, London, U.K., 1973. Perry, L.M., Medicinal Plants of East and Southeast Asia, MIT Press, Cambridge, Mass., 1980. Peskin, H.M., National Accounts and the Development Process: Illustration withTanzania, Resources for the Future, Washington, D.C., 1984. Peskin, H.M., The Environmental Accounts Framework, Resources for the Future, Washington, D.C., 1987. Peterson, A.E. and J.B. Swan, editors, Universal Soil Loss Equation: Past. Present and Future, Soil Science Society of America, Madison, Wisconsin, 1979. Phantumvanit, D., Thailand Natural Resources Profile, Thailand Development Research Institute, Bangkok, Thailand, 1987. Phillips, C., Preliminary Observations on the Effects of Logging on a Hill Forest in Sabah. In Hadi, Y., K. Awang, N.M. Majid, and S. Mohamed (Eds.), Im2act of Man's Activities on Trooical Upland Forest Ecosystems: 187-215, Faculty of Forestry, Universiti Pertanian Malaysia, Selangor, Malaysia. Pimentel, D. and nine others, Biomass Energy from Crop and Forest Residues, Science 212: 1110-1115, 1981. Pimentel, D., eLal., Land Degradation: Effects on Food and Energy Resources, Science 194: 149-155, 1982. Pimentel, D. and 13 others, World Food Economy and the Soil Erosion Crisis, Environmental Biology Report 86-2, Cornell University Press, Ithaca, N.Y., 1985. - 55 - Pimentel and five others, Deforestation: Interdependency of Fuelvood in Agriculturee Oikas 46: 404-412, 1986. Pinker, R.T., O.B. Thompson and T.F. Eck, The Albedo of a Tropical Evergreen Forest, Quarterl Journalof Royal Meteorological Society 106: 551-558, 1980. Posner, J.L. and M.S. McPherson, Agriculture on the Steep Slopes of Tropical America: The Current Situation and Prospects, World Develoment 10: 341-353, 1982. Principe, P.P., The Value of Biological Diversity Amon Medicinal Plants, Environment Directorate, Organization for Economic Cooperation and Development, Paris, France, 1985. Prospero, J.N., R.A. Glaccun and R.T. Nees, Atmospheric Transport of Soil Dust from Africa to South America, Nature 289: 570-572, 1981. Pryor, L.D., Ecological Hismanagement in Natural Disasters, Man romentalist (Suppl. 2) 2: 1-14, 1982. Quansah, C., The Effect of Soil Type, Slope, Rain Intensity and Their Interactions on Splash Detachments and Transport, Journal of Soil Science 32: 215-224, 1981. Quintiliano, J. and three others, Perdas por Erosso no Estado de Sao Paolo, Bragania 20: 1113-1182, 1961. Rambler, N., editor, Global Biology Research Proaram, National Aeronautics and Space Administration, Washington, D.C., 1983. Randrianarijaona, P., The Erosion of Madagascar, Ambio 12: 308-311, 1983. Raven, P.H., Biological Resources and Global Stability (Speech presented at Presentation of International Prize for Biology, Kyoto, Japan, November 22nd, 1986), Missouri Botanical Garden, St. Louis, Missouri, 1986. Repetto, R., Government Subsidies and Misuse of Forest Resources, World Resources Institute, Washington, D.C., (in press) 1986. h epetto, R., Government Subsidies arA Misuse of Water Resources, World Resources Institute, Washington, D.C., (in press) 1986. Repetto, R., Natural Resource Accounting for Countries with Natural Resource-Based Economies, World Resources Institute, Washington, D.C., 1986. Repetto, R., The Forest for the Trees? Government Policies and the Misuse of Forest Resources. World Resources Institute, Washington, D.C., 1988. 56 - Revilla, A.V., J.A. Canonizado and M.C. Gregorio, Forest Resources Managne RWort. Forest Resources Management Task Force, Resource Policy Group, Manila, Philippines, 1987. Rieger, H.C., Floods and Droughts in the Himalayas and the Ganges Plain, in Mountain Environment and Develooment 13-29, SATA, Berne, Switzerland, 1976. Rijaberman, F.R. and M.G. Wolman, editors, Ouantification of the Effect of Erosion on Soil Productivity in an International Context, Deflt Hydraulics Laboratory, Delft, Netherlands, 1984. Robinson, F.H., A Report on the Panama Ganal Rain Forests, Panama Canal Commission, Balboa, Panama, 1985. Rodhe, R. and 4. Herrera, Acification in Tropical Countries. John Wiley and sons, Chichester, U.K. and New York, 1988. Rogers, P.R., Fresh Water in R. Repetto, editor, The Global Possible: 255-298, Yale University Press, New Haven, Conn., 1985. Salati, E. and P.B. Vose, Amazon Basin: A System in Equilibrium, Science 225: 129-138, 1984. Sale, J.B., The Imnortance and Values of Wild Plants and Animals in Africa, International Union for Conservation of Nature and Natural Resources, Gland, Switzerland, 1983. Samson, J.A., Tropical Fruits, Longman Inc., New York, 1980. Sanchez, P.A., Properties and Management of Soils in the Tropics, John Wiley and Sons, New York, 1976. Sancholuz, L., Land Degradation in Mexican Maize Fields, Doctoral Dissertation, University of British Columbia, Vancouver, Canada, (cited in Pimental atal., 1986), 1984. Sanders, D.W., Relating Potential Productivity to Soil Loss, EA Landand Water Technology Newsletter: 21-25, 1981. Sartorius, P. and H. Henle, Forestry and Economic Development, Praeger Publishers, New York, 1968. Schmidt, B.L., at al., editors, Determinants of Soil Loss Tolerance, Special Publication No. 45, American Society for Agronomy, Madison, Wisconsin, 1982. Schmink, M. and C.H. Wood, editors, Frontier Expansion in Amazonia, University of Florida Press, Gainesville, Florida, 1985. Schumm, S.A. and M.D. Harvey, Natural Erosion in the USA, in Determinants of Soil Loss Tolerance, ASA Publication No. 45, American Society of Agronomy, Madison, Wisconsin, 1982. - 57 - Schultes, R.E., The Amazonia as a Source of New Economic Plants, Economic Botany 33(3): 259-266, 1980. Sellers, P.J., Canopy Reflectance, Photosynthesis and Transportation, International Journal of Remote Sensing 6: 335-1372, 1985. A , Sellers, P.J. and J.G. Lickwood, A Computer timulation of the Effects of Different Crop Types on the Water Balance of Small Catchments Over a Long Time Period, Ouarterly Journal of the Royal Meteorological SocietX 107: 395-414, 1981. Setzer, A.W. and M.C. Pereira, Amazon Biomass Burnings in 1987 and Their Atmospheric Emissions, Science (submitted) 1988. Sfeir-Younis, A., Soil Conservation in Develoing Countries. a Background Report, The World Bank, Washington, D.C., 1986. Shugart, H.H., N.Y., a Antonovsky, P.G. Jarvis and A.P. Sandford, 02, Climatic Change and Forest Ecosystems, in B. Bolin, B.R. Doos, J. Jagel and R.A. Warrick, editors, Greenhouse Effect: Climatic Change and Ecosystems: 475-521, John Wiley and Sons, New York, 1986. Shukla, J. and Y. Mintz, Influence of Land Surface Evapotranspiration on the Earth's Climate, Science 215: 1498-1501, 1982. Shuttleworth, W.J. and 14 others, Observations of Radiation Exchange Above and Below Amazonian Forests, Notes and Correspondence, OaterlX Journal of Royal Meteorological Society, 1984. Shuttleworth, W.J. and four others, Eddy Correlation Measurements of Energy Partition for Amazonian Forests, Quarterly Journal of Royal Meteorological Society 110: 1143-1162, 1984. Sinha, B., Water Resources Management in India: Problems and Perspectives, East-West Center, Honolulu, Hawaii, 1983. Singh, J.S., A.N. Pandey and P.C. Pathak, A Hypothesis to Account for the Major Pathway of Soil Loss from Himalaya, Enironmental ConservAtion 10: 343-345, 1983. Soule, M.E., Conservation Biology: Science of Scarcity and Diversity, Sinauer Associates, Sunderland, Mass, 1986. Southgate, D., Estimating the Downstream Benefits of Soil Conservation in a Hydroelectric Watershed, Department of Agricultural Economics, Ohio State University, Columbus, Ohio, 1986. Southgate, D., F. Hitzehusen and R. MacGregor, Remedying Third World Soil Erosion Problems, American Journal of Agricultural Economics 66: 879-884, 1984. * 58 - Spears, J., Economic Analysis Issues in Bank-Financed Forestry Projects, Agriculture and Rural Development Department, World Bank, Washington, D.C., 1984. Spears, J., Sudan: Forestry Sector Review, World Bank, Washington, D.C., 1985. Staritsky, G., Is There a Future for Medicinal Crops? san 23(2): 8-82, 1980. Steeds, D., Desertification in the Sahelian and Sudanian Zones of West Africa, The World Bank, Washington, D.C., 1985. Steward, J.B., Measurement and Prediction of Evaporation from Forested and Agricultural Catchments, Aricultural Water Management 8: 1-28, 1984. Stocking, M.A., The Measurement, Use and Relevance of Rainfall Energy in Investigations into Erosion, Zeitschrift Geomorphologie N.F. 29: 141-150, 1978. Stocking, M.A., Conservation Strategies for Less Developed Countries, in R.P.C. Morgan, editor, Soil Conservation: Problems and Prospects: 377-384, John Wiley and sons, Chichester, U.K., 1980. Stocking, M.A., Erosion and Soil Productivity: A Review, Food and Agriculture Organization of the United Nations, Rome, Italy and Overseas Development Group, University of East Anglia, Norwich, U.K., 1984. Stocking, M.A., Erosion-Induced Loasin Soil Productivity:. Trends in Research and International Cooperation, Food and Agriculture Organization of the United Nations, Rome, Italy, and Overseas Development Group, University of East Anglia, Norwich, U.K., 1985. Stocking, M.A., ffects of Soil Erosin on Soil Nutrients in Zimbabwe, Food and Agriculture Organization, Rome, Italy, 1986. Sundborg, A., Sedimentation Problems in River Basins, Nature-and Resources (UNESCO) 19: 10-21, 1983. Sundborg, A. and A. Rapp, Erosion and Sedimentation by Water: Problems and Prospects, fthio 15: 215-225, 1986. Sussman, R., 0. Green, and A.H. Centry, A Study of Deforestation in Eastern Madagascar Using Satellite Imagery, Science (in press), 1988. Tcheknavorian-Asenbauer, A. and R.O.B. Wijesekera, Medicinal and Aromatic Plants for Industrial Development, United Nations Industrial Development Organization, Vienna, Austria, 1982. Tejwani, K.G., S.K. Gupta and H.N. Mathur, editors, Soil and Water Conservation Reseac Indian Council of Agricultural Research, New Delhi, India, 1975. * 59 - Townsend, J.R.G. and C.J. Tucker, Continental Land Cover Classification Using Satellite Data, Paper presented at 36th Congress of the International Astronautical Federation, Stockholm, Sweden, October 7-12th, 1985, University of Reading, U.K. and Goddard Space Flight Center, Greenbelt, Maryland, 1985. Troeh, F.R., J.A. 'Hobbs and R.L. Donahue, Soil and Water Consrvation for Productivity and Environmental Protection, Prentice-Hall, Englewood Cliffs, New Jersey, 1980. Tucker, C.J., J.R.G. Townshend and R.L. Donahue, Soil and Water Conservation for Productivity and Environmental Protection, Prentice-Hall, Englewood Cliffs, New Jersey, 1980. Tucker, C.J., J.R.G. Townshend and T.E. Goff, African Land-Cover Classification Using Satellite Data, Science 227: 369-374, 1985. UNCTAD, Trade and Development Report, United Nations Conference on Trade and Development, Geneva, Switzerland, 1982. UNESCO, Proceedings of the Fourth Asian Symosium on Medicinal Plants and Spices, Bangkok, Thailand, September 1980, UNESCO Special Publication, UNESCO, Paris, France, 1981. UNICEF, The State of the World's Children, UNICEF, New York, 1985. United Nations, Water Develooment and Management, Proceedings of the United Nations National Water Conference, 1977, Pergamon Press, Elmsford, New York, 1978. U.S. Department of Energy, Detecting the Climatic Effects of Increasing Carbon Dioxide: Characterization of Information Requirements for Studies of Carbon Dioxide Effects: Projecting the Climatic Effects of Increasing Carbon Dioxide: Direct Effects of Increasing Carbon Dioxide on Vegetation: and Atmospheric Carbon Dioxide and the Global Carbon Cycle, Carbon Dioxide Research Division, U.S. Department of Energy, Washington, D.C., 1985. Vickers, W.T., The Faunal Components of Lowland South American Hunting Kills, Interciencia 9: 366-376, 1984. Vohra, B.B., Land and Water ManaXement Problems in India, Ministry of Home Affairs, New Delhi, India, 1982. Vohra, B.B., Management of Natural Resources: Urgent Need for Fresh Thinkinu, Advisory Board on Energy and Water, Government of India, New Delhi, India, 1986. Wadsworth, F., Death to the Panama Canal, paper presented to International Strategy Conference on Tropical Deforestation, Department of State, Washington, D.C., 1978. - 60 - Warford, J.J., Natural Resource Manaement and Economic Development (Paper presented at IUCN Conference in Ottawa, June 1986), World Bank, Washington, D.C., 1986. Weismiller, R.A., G.E. Van Scoyoc, S.E. Pazar, K. Latz, and M.P. Baumgardner, Use of Soil Spectral Properties for Monitoring Soil Erosion, in S.A. E1-Swaify, W.C. Moldenhauer and A. Lo, Soil Erosion and Conservation, Soil Conservation Society of America, Ankeny, Iowa, 1985. West Pakistan Water and Power Development Authority, Mangla Watershed Managemet Stud. West Pakistan Water and Power Development Authority, Islamabad, Pakistan, 1977. White, A.T., and S.M. Wells, Coral Reefs in the Philippines, QQM 16: 445-451, 1982. Wiggins, S.L., The Economics of Soil Conservation in the Acelhuate River Basin, E1 Salvador, pp. 399-415, in R.P.C. Morgan, editor, oil Conservation: Problems and Prospects, John Wiley and Sons, Chichester, U.K., 1981. Wijewardene, R. and P. Waidyanatha, Conservation Farming: Systems. Techniques and Tools for Small Farmers in the Humid Tropics, Ministry of Agricultural Development and Research, Colombo, Sri Lanka, and the Commonwealth Consultative Group on Agriculture for the Asia-Pacific Region, Commonwealth Secretariat, London, U.K., 1984. Williams, J.T., C.H. Lamoreaux and N. Wulijarni-Soetjipto, editors, Southeast Asian Plant Genetic Resources, BIOTROP, Bogor, Indonesia, 1975. Williams, P.B., Planning Problems in International Water Development, Philip Williams and Associates, San Francisco, Calif., 1983. Williams, J. and L.S. Hamilton, Watershed Forest Influences in the Tropics and SubtroDics: A Selected. Annotated Bibliography, Environment and Policy Institute, The East-West Center, Honolulu, Hawaii, 1982. Wilson, J., Colonization in Rondonia: The Case of Ariquemis, Doctoral Dissertation, University of Florida, Gainesville, Florida, 1985. Wilson, E.0, and F. Peter, editors, Proceedings of National Forum on BioDiversity, National Academy of Sciences and Smithsonian Institute, Washington, D.C., 1987. Windsor, D.M. and A.S. Rand and W.M. Rand, Variation in Rainfall on Barro Colorado Island, in Ecology of a Tropical Forest (Spanish edition), 1986. - 61 - Wiscbmeier, W.H. and D.D. Smith, Predicting Rainfall ErosionLosses - A Guide to Conservation Planning, Agricultural Handbook No- 537, U.s. Department of Agriculture, Washington, D.C., 1978:t' Wolman, M.G., Soil Erosion and Crop Productivity, in R.F. Follettoand B.A. Stewart, editors, Soil Erosion and Crop Productivity, Soil Science Society of America, Madison, Wisconsin, 1985. i3 Woodwell, G.M. editor, The-Role of Terrestrial Vetation in-tba-Qlobal Carbon Cycle: Measurement by Remote Sensing, John Wiley and Sons, New York, 1984. Woodwell, G.M. gt al., Global Deforestation: Contribution to Atmospheric Carbon Dioxide, lgiepec 222: 1081-1086, 1983. Woodwell, G.M. and eight others, Deforestation Measured by Landsat: Steps Toward a Method, Carbon Dioxide Research Division, U.S. Department of Energy, Washington, D.C., 1983. World Bank, Staff Appraisal Renort: India Himalaan Watershed Management Project. Uttar Pradesh, World Bank, Washington, D.C., 1983. World Bank, World Development Report. 1984, The World Bank, Washington, D.C., 1984. World Bank, World Develooment Report 1985, The World Bank, Washington, D.C., 1985. World Bank, Sudan: Forestry Sector Review, World Bank, Washington, D.C., 1985. World Resources Institute and World Bank, Tropical Forests: A Call for Action, World Bank and World Resources Institute, Washington, D.C., 1985. World Resources Institute, World Resources 1987, World Resources Institute, Washington, D.C., 1987.